Perpetrating Genocide: A Criminological Account 9781138648814, 9781315626239

Focusing on the relationship between the micro level of perpetrator motivation and the macro level normative discourse,

693 33 5MB

English Pages [291] Year 2017

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

Perpetrating Genocide: A Criminological Account
 9781138648814, 9781315626239

Table of contents :
Cover
Series Page
Title Page
Copyright Page
Dedication Page
Contents
List of Illustrations
Acknowledgments
Prologue:
Inside Nyamata Church
Visiting perpetrator worlds
Note
Chapter 1: Introduction: An unimaginable and
uncharacteristic act
Imagining the unimaginable
The crime of genocide
The enigma of causation
Methodology and objectives of this book
Theoretical approach
Methodological approach: towards a criminology of genocide
perpetration
Structure of the book
Notes
Works cited
Chapter 2: The emergence of the genocidal
context
Introduction: the genesis of genocide
The prohibition of killing
Catastrophic loss: the context of the genocidal state
Strain and adaptation
Political and economic crises and strain
The drivers of the emergence of the genocidal context
Radicalization
State radicalization: mainstreaming atrocity
Group radicalization: the humanity gap
Individual radicalization: the surrender of the will
Genocide in public and private spaces
Polarization in the economic and political spheres
Violence and social relations
Escalation and restraint
Conclusion: social order and disorder
Works cited
Notes
Chapter 3:
The genocidal context
Introduction: harm-producing relations of power
The genocidal state as a state of moral disengagement
Genocidal contexts compared
Norms facilitating violence
Obedience
Conformity
Urgency
Structures facilitating violence
The legal system
Organizing genocide
Killing in organizations
Violence specialists
Paramilitaries
Prisons and concentration camps
The genocidal state as an organizational culture
State power and the localization of the moral context
Conclusion: structuring choice
Notes
Works cited
Chapter 4: Propaganda:
Communicating the moral context
Introduction
The functions of propaganda
The dehumanization dynamic: scapegoating and normative shifts
Genocidal ideology
Mechanisms of genocidal propaganda
Genocidal discourses
Stigmatizing propaganda
Group-bindingnarratives
Conclusion: state power and propaganda
Notes
Works cited
Chapter 5:
Who kills?
Introduction
Profiling perpetrators
Killing as a necessary evil
The marginal
Joining
The typical perpetrator
Perpetrator coalitions
Professionals and amateurs
Vanguard and reluctant perpetrators
Direct perpetrators and leaders
Pathways to perpetration
Pathological perpetration
Ideological perpetration
Entrepreneurial perpetration
Conformist perpetration
Compromised perpetration
Variable individual responses to genocide
Action roles in genocide
The propensity to rescue: rescuers as outliers
Controls
Conclusion: distance and propensity
Notes
Works cited
Chapter 6:
Deciding to kill
Introduction
Perpetrator opportunity
Victim vulnerability
Decision-making
Social margin of discretion
Role margin of discretion
The immediate margin of discretion
Presence of authority figures
Group dynamics
Distance
Perceived threat
Emotion and intoxication
Risks and incentives
Moral rules
Direct and indirect perpetrators
Norm bandwagons and tipping points
Conclusion: mass violence and individual will
Notes
Works cited
Chapter 7:
Killing
Introduction: from context to killing
Mobilizing perpetrators
The killing inhibition: willing executioners?
Kill, flee, posture, resist?
The boundaries of violence
Forced perpetration
Conclusion
Notes
Works cited
Chapter 8:
Rationalizing killing
Introduction: reconciling with the moral context
The genocidal techniques of neutralization
Neutralization-drift
theory
Appeal to higher loyalties
Denial of the victim
Denial of humanity
Denial of responsibility
Denial of injury
Claim of normality
Claim of inevitability
Claim of relative acceptability
Claim of inner opposition
The denial of autonomy
Assessing moral neutralization
Subcultural narratives in genocide
Conclusion: tools of adaptation
Notes
Works cited
Chapter 9:
Coping with killing
Introduction: adjusting to genocide
The progression of genocide
Escalating violence
The “problem” of the decent Jew next door: overcoming resistance
to killing
Perpetrator R22, Hutu Male, Mbazi Sector, “Ignace”
The end of genocide
Individual recidivism
Effects of killing on perpetrators
Coping mechanisms
The decreasing cost of perpetration
Perpetrator identity
The sadistic shift
Systems of denial
Conclusion: genocidal momentum
Notes
Works cited
Chapter 10: After genocide I:
Memory, trauma, and rehabilitation
Introduction: forgetting genocide
Reforming the criminogenic state
The evolution of perpetrator narratives
Memory
Trauma
Rehabilitation
Conclusion: after genocide
Notes
Works cited
Chapter 11: After genocide II:
Justice
Introduction: transitional justice and the prosecution of
genocide
Prosecuting genocide
Legal ordering
The problem of genocidal intent
The attribution of responsibility
The reproduction of perpetrator narratives
Prosecution as prevention
Reconsidering transitional justice
Conclusion: simple answers for complex questions?
Notes
Works cited
Chapter 12: Conclusion:
Killing without consequence?
Comprehending genocide
The globalization of conscience?
Evil and intention: sleepwalking through the end of
the world
Notes
Works cited
Index

Citation preview

Perpetrating Genocide

Focusing on the relationship between the micro level of perpetrator motivation and the macro level normative discourse, this book offers an in-­depth explanation for the perpetration of genocide. It is the first comparative criminological treatment of genocide drawn from original field research, based substantially on the author’s interviews with perpetrators and victims of genocide and mass atrocities, combined with wide-­ranging secondary and archival sources. Topics covered include: perpetration in organizations, genocidal propaganda, the characteristics of perpetrators, decision-­making in genocide, genocidal mobilization, coping with killing, perpetrator memory and trauma, moral rationalization, and transitional justice. An interdisciplinary and comparative analysis, this book utilizes scientific methods with the objective of gaining some degree of insight into the causes of genocide and genocide perpetration. It is argued that genocide is more than a mere intellectual abstraction – it is a crime with real consequences and real victims. Abstraction and objectivity may be intellectual ideals but they are not ideally humane; genocide is ultimately about the destruction of humanity. Thus, this book avoids presenting an overly abstract image of genocide, but rather grounds its analysis in interviews with victims and perpetrators of genocide in Rwanda, Burundi, Uganda, Bosnia, Cambodia, Bangladesh, and Iraq. This book will be highly useful to students and scholars with an interest in genocide and the causes of mass violence. It will also be of interest to policy-­ makers engaged with the issues of genocide and conflict prevention. Kjell Anderson is an interdisciplinary scholar specializing in the study of mass violence, and is currently an Affiliated Research Fellow of the Centre for International Criminal Justice at the Vrije Universiteit Amsterdam.

Routledge Studies in Genocide and Crimes against Humanity Edited by Adam Jones University of British Columbia in Kelowna, Canada

The Routledge Studies in Genocide and Crimes against Humanity series publishes cutting-­edge research and reflections on these urgently contemporary topics. While focusing on political-­historical approaches to genocide and other mass crimes, the series is open to diverse contributions from the social sciences, humanities, law, and beyond. Proposals for both sole-­authored and edited volumes are welcome. For more information about this series, please visit: www.routledge.com/ Routledge-­Studies-in-­Genocide-and-­Crimes-against-­Humanity/book-­series/RSGCH The Structural Prevention of Mass Atrocities Understanding Risks and Resilience Stephen McLoughlin Constructing Genocide and Mass Violence Society, Crisis, Identity Maureen Hiebert Last Lectures on the Prevention and Intervention of Genocide Edited by Samuel Totten Perpetrating Genocide A Criminological Account Kjell Anderson The United States and Genocide (Re)Defining the Relationship Jeffrey S. Bachman

Perpetrating Genocide A Criminological Account

Kjell Anderson

First published 2018 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 711 Third Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2018 Kjell Anderson The right of Kjell Anderson to be identified as author of this work has been asserted by him in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data Names: Anderson, Kjell Follingstad, 1977– author. Title: Perpetrating genocide : a criminological account / Kjell Anderson. Description: Abingdon, Oxon ; New York, NY : Routledge is an imprint of the Taylor & Francis Group, an Informa Business, [2018] | Series: Routledge studies in genocide and crimes against humanity | Includes bibliographical references and index. Identifiers: LCCN 2017034780| ISBN 9781138648814 (hbk) | ISBN 9781315626239 (ebk) Subjects: LCSH: Genocide–Research. | Genocide–Case studies. | Mass murderers–Research. | Mass murderers–Case studies. Classification: LCC HV6322.7 .A63 2018 | DDC 364.15/1–dc23 LC record available at https://lccn.loc.gov/2017034780 ISBN: 978-1-138-64881-4 (hbk) ISBN: 978-1-31562-623-9 (ebk) Typeset in Times New Roman by Wearset Ltd, Boldon, Tyne and Wear

It has been said that evil is the absence of empathy. If evil is the absence of empathy, then hell is the absence of hope. I dedicate this work to empathy and to hope.

Contents



List of illustrations Acknowledgments Prologue: inside Nyamata Church

  1 Introduction: an unimaginable and uncharacteristic act

x xi xiii 1

Imagining the unimaginable  1 Methodology and objectives of this book  7   2 The emergence of the genocidal context

17

Introduction: the genesis of genocide  17 The prohibition of killing  19 Catastrophic loss: the context of the genocidal state  19 The drivers of the emergence of the genocidal context  23 Radicalization  26 Genocide in public and private spaces  32 Conclusion: social order and disorder  37   3 The genocidal context Introduction: harm-­producing relations of power  43 The genocidal state as a state of moral disengagement  44 Genocidal contexts compared  46 Norms facilitating violence  47 Structures facilitating violence  51 The genocidal state as an organizational culture  60 State power and the localization of the moral context  62 Conclusion: structuring choice  63

43

viii   Contents   4 Propaganda: communicating the moral context

69

Introduction  69 The functions of propaganda  70 The dehumanization dynamic: scapegoating and normative shifts  71 Genocidal ideology  73 Mechanisms of genocidal propaganda  75 Genocidal discourses  78 Conclusion: state power and propaganda  87   5 Who kills? 

93

Introduction  93 Profiling perpetrators  94 Perpetrator coalitions  99 Pathways to perpetration  101 Variable individual responses to genocide  111 Controls  116 Conclusion: distance and propensity  117   6 Deciding to kill

124

Introduction 124 Perpetrator opportunity  125 Decision-­making  128 Social margin of discretion  130 Role margin of discretion  131 The immediate margin of discretion  133 Risks and incentives  138 Conclusion: mass violence and individual will  145   7 Killing Introduction: from context to killing  150 Mobilizing perpetrators  150 The killing inhibition: willing executioners?  154 Kill, flee, posture, resist?  156 The boundaries of violence  158 Forced perpetration  163 Conclusion  165

150

Contents   ix   8 Rationalizing killing

170

Introduction: reconciling with the moral context  170 The genocidal techniques of neutralization  170 Assessing moral neutralization  182 Subcultural narratives in genocide  184 Conclusion: tools of adaptation  185   9 Coping with killing

191

Introduction: adjusting to genocide  191 The progression of genocide  191 Individual recidivism  199 Systems of denial  209 Conclusion: genocidal momentum  211 10 After genocide I: memory, trauma, and rehabilitation

217

Introduction: forgetting genocide  217 Reforming the criminogenic state  219 The evolution of perpetrator narratives  223 Conclusion: after genocide  231 11 After genocide II: justice

236

Introduction: transitional justice and the prosecution of genocide  236 Prosecuting genocide  237 Prosecution as prevention  245 Reconsidering transitional justice  248 Conclusion: simple answers for complex questions?  249 12 Conclusion: killing without consequence?

253

Comprehending genocide  253 The globalization of conscience?  256 Evil and intention: sleepwalking through the end of the world  259

Index

264

Illustrations

Figures 6.1 Perpetration process 8.1 The genocidal techniques of neutralization 9.1 The decreasing cost margin of perpetration

129 173 203

Tables 1.1 1.2 4.1 6.1

Rwanda interview sample – occupation in 1994 Location of Rwandan perpetrator interviews Typology of genocidal propaganda Authority figures at killing sites in Rwanda

12 13 78 133

Boxes 4.1 The moral context of genocidal propaganda 6.1 Assessment of victim vulnerability

78 126

Acknowledgments

This book is a work of both the head and the heart. Such an undertaking would not have been possible without the help of many individuals and institutions. I have benefited greatly from the support of my friends and family, so I must thank them first and foremost. My father, Alan Anderson, was particularly invaluable in closely reading my drafts and providing both substantive and copy­editing suggestions. Ingjerd Veiden Brakstad was a great support throughout, reading my drafts, and undoubtedly making my ideas better. This book began as a thesis, so I must also recognize the guidance and support of my committee, particularly my supervisor, William Schabas, who was instrumental to the success of this project. Thank you also to my internal examiner, Shane Darcy, and my external examiner, Alex Hinton, for their thoughtful input. Frank Chalk and Ray Murphy each read parts of the thesis draft. The Irish Centre for Human Rights provided a supportive and enriching home base for my thesis research, as well as lasting friendships. I was assisted by many other research bodies including: the Montréal Institute for Genocide and Human Rights Studies at Concordia University; the Centre for Human Rights and Citizenship at West Bengal National University of Juridical Sciences; the Centre for Research into Indo-­Bangladesh Relations in Kolkata; the Documentation Centre for Cambodia; the Peace Palace Library; the Bophana Resource Centre; the National Library of India; the Research and Documen­tation Centre (Sarajevo); the Institute for the Research of Crimes Against Humanity and International Law (Sarajevo); the National University of Rwanda (especially Justine Mbabzi and David Mwesigwa); the Commission National pour la Lutte contre le Génocide (CNLG); the Ministry of Internal Security (Rwanda); and the TIG Administration (Travaux d’Intérêt Général, particularly Bizimana Evariste). Thanks also to Bhaskar Tanna, Ruma Pal, and the Ray family for assisting me in India; Monaem Sarker, Hassan Ali, Zead-­Al Malum, and Rafiqul Islam for providing invaluable help during my time in Dhaka; Eric Niragira for your assistance in Burundi; Dennis Martin and Empowering Hands in Uganda; as well as Karlijn Vandervoort and Aonghus Kelly in Cambodia. Iva Vukusic provided some insight into violence in Bosnia, as did Timothy Williams for Cambodia. Erin Jessee has also been a good sounding board and colleague throughout my research.

xii   Acknowledgments I must also acknowledge the work of my interpreters and guides who provided me with invaluable insight, as well as logistical and moral support: in Rwanda – Kassim Mbarushimana, Yassin Nsabimana, Eugène Ndisanze, and Patrick Muhire; in Cambodia Socheata Peou; in Bosnia Mirnes Dervisevic, Sudbin Music (Prijedor) and Berin Milic (Tuzla); in Burundi “Ignace”; and in Uganda, Herimos Odongo. My work was also made possible by the hard work of my research assistants: Koen Kluessien (who conducted archival research on propaganda), Anne Van Nassau (who assisted me with transcription of the Rwandan, Bosnian, and Ugandan interviews), Rick Hoefsloot (who helped to transcribe and analyze the Burundi interviews), and Noyel Ry (who transcribed and translated my Cambodia interviews). I also benefited from the assistance of Marc Sherman, who provided additional bibliographic and index edits, Michael Flynn, who loaned me his computer, and Karen Bähr, who assisted me with historical German census data. Thank you all! Thank you to the NIOD Institute for War, Holocaust, and Genocide Studies for providing a supportive environment for writing my book, especially Nanci Adler, Laura van Kessel, Katrina Cooper, Thijs Bouwknegt, and Uğur Ümit Üngör. I also owe a debt of gratitude to the anonymous reviewers, and the editors at Routledge, particularly Adam Jones for his diligence, patience, and support. Yehonaten Alsheh also provided useful suggestions on the book proposal. Finally, thank you to the many individuals who, by revisiting their difficult memories, ultimately made this book possible.

Prologue Inside Nyamata Church

And when you look long into the abyss, the abyss also looks back at you. (Friedrich Nietzsche, Beyond Good and Evil)

On a hot summer day in 2007 I stood inside Nyamata Church in the rust-­red dust of the countryside near Kigali. Thirteen years earlier, it was the site of a massacre. Thousands of Rwandans from the surrounding commune sought refuge on sanctified ground, but they were to receive no mercy. Instead, they were slaughtered with guns, grenades, and, with the face-­to-face intimacy of clubs and machetes. One woman was raped and then impaled on a wooden stake. Today Nyamata Church is an oasis of avocado trees. At first it is impossible to conceive that it was once a landscape racked with horror. Yet, with guidance, you begin to recognize the signs, oddly pacified through the movement of time. The altar cloth is soaked in dried blood; bullet holes mark nearly every surface; the wall is spattered crimson where infants were smashed against it. I descend beneath the church into Cimmerian underground chambers where ravaged bones are stacked to the roof – skulls bearing witness to human suffering through their cracked shells and jagged voids. These are not mere artefacts, but extinguished human lives. It was entirely beyond my comprehension. What made such extraordinary brutality possible? Where did this malevolent force go? It seems to have disappeared into the ether, yet I know that can’t really be possible. Encountering genocide brought me to the threshold of hell, where I peered inside in some attempt at comprehension without being swallowed myself. Still, there was much I did not understand. Rwandans seemed much the same as people everywhere else. Was this Elysian land of rolling green hills and volcanoes illusory? Could genocide happen in Rwanda again? I want to say no, but I cannot. Rwanda is neither heaven nor hell – it’s just a place in between. And that might be the most frightening thought of all.

Visiting perpetrator worlds Several years ago, I worked with victims of genocide in Rwanda. The sheer scale of the Rwandan genocide, and the involvement of many civilians, led me to the

xiv   Prologue: inside Nyamata Church inescapable conclusion that perpetrators are not extraordinary people. Yet after meeting the victims and hearing their stories of extreme brutalization, genocide also seemed to be an extraordinary act, far removed from our everyday moral beliefs. How are such acts possible? I pursued this question by researching the causes of genocide, interviewing perpetrators and victims of genocide and other atrocities in Rwanda, Burundi, Uganda, Bosnia, Bangladesh, and Cambodia. The results that I obtained from these interviews startled me. The perpetrators gave the impression that although the genocide had happened, it had not happened to them. I had expected that some perpetrators would minimize their guilt, but not that this abrogation of responsibility would be so universal or so total. Apparently, this was a human evil of staggering magnitude without any authors. It was Hannah Arendt’s banality of evil manifested, but it was not limited to bureaucracies of murder such as the Nazi extermination camps. Rather, it was ubiquitous, mundane, and rooted in everyday social tendencies. The apparent normalcy of the perpetrators was grounded not only in their biographies, but also in their psychological imperative to minimize and avoid responsibility for immoral acts. Although politicized hatred was an underlying factor, perpetrators often acted based on other considerations. Not only did they abide by and contribute to the objectification of the victims, they also self-­objectified as nebulous moral actors, responsive only to the expectations of authority and peers. Sitting in an empty fast-­food restaurant in Phnom Penh, I asked a survivor of Tuol Sleng Prison about the public denials of one of the former guards. He sighed deeply, with monsoon lightning flashing outside: “we live in the same country, but in two different worlds.”1 My research sought to explore this perpetrator world, and led me to the conclusion that it was not an exotic place, but an everyday, interior universe. Focusing on the relationship between perpetrator motivation and the normative context, this book offers an explanation for the perpetration of genocide. It is ultimately based on what the perpetrators themselves have told me.

Note 1 Interview C06 (Cambodia), Phnom Penh, November 2008.

1 Introduction An unimaginable and uncharacteristic act

Imagining the unimaginable [Genocide is a crime of] such unprecedented brutality and of such inconceivable savagery that the mind rebels against its own thought image and the imagination staggers in the contemplation of a human degradation beyond the power of language to adequately portray … one cannot grasp the full cumulative terror of murder one million times repeated. (Einsatzgruppen Judgment1) I try not to think about it or I will lose my mind. (Rwandan victim of genocide2) Genocide is an unimaginable act. This is true both on the collective level – where genocide represents the coordinated, intentional extermination of a racial, ethnic, national, or religious group – and on the individual level, where genocide usually entails the assault or murder of other individuals. It is unimaginable for those of us who choose to research it, particularly if we are not witnesses to the crime. We strive to “know” genocide intellectually, but this knowledge will always be inadequate. The judges in the Einsatzgruppen trial attempted to convey some of this horror to us, the bystanders. Yet we remain, thankfully, isolated from the visceral reality of violence as we read these words from the safety of our homes and libraries. Genocide is also unimaginable for many of those who commit it. For the most part, perpetrators never imagined themselves killing. Consequently, it is difficult for them to reconcile what they did with who they are. In effect, they deny or reimagine their crime in order to maintain their sense of themselves as decent people. Genocide, for most perpetrators, is an uncharacteristic act. In approaching the subject of genocide, we are also approaching evil. Evil has a dualistic character: it can refer, as an adjective, to acts that are “profoundly immoral and wicked,” or, as a noun, to “profound immorality and wickedness, especially when regarded as a supernatural force.”3 When studying perpetrators of genocide, we might question whether their acts are evil or their fundamental nature? This book argues that the perpetration of genocide is a product of the

2   Introduction moral context – individual alignment with moral norms as dictated by state power. Evil, then, is transitive rather than eternal; acts of evil are fixed in the moment, even as the consequences of evil may persist. Genocide has become a shorthand sobriquet of human evil, and arguably it possesses a normative weight surpassing any other crime. However, it is this very power of the notion of genocide that confounds our understanding. The discourse is that such evil is not meant to be understood, but only to be feared. We are left with a “rhetoric of awe” and an “unintentional sacralisation.”4 The protagonist in Bernhard Schlink’s The Reader grapples with this dilemma of understanding perpetration: I wanted simultaneously to understand Hannah’s crime and to condemn it. But it was too terrible for that. When I tried to understand it, I had the feeling I was failing to condemn it as it must be condemned. When I condemned it as it must be condemned, there was no room for understanding.5 Religious figures may verbalize genocide using the language of sin or divine malevolence. All this mystification of evil has the effect of making genocide seem an inexplicable and, consequently, inexorable force. It also eternalizes the causes of genocide, rather than locating them in the realm of human social interaction. Genocide is not an act of God, any more than it is a natural disaster or a manifestation of mass hysteria. With the mystification of genocide also comes the mystification of the perpetrators. They are demonized as monsters arising from authentic and metaphysical evil rather than human society. This demonization of the perpetrators contains an aspect of righteous condemnation. Such condemnation is morally justifiable, yet when manifested as this sort of demonization, it is also misleading. Paradoxically, it does not represent a psychological mechanism of responsibility, but rather one of avoidance and denial. “The génocidaires are nothing like you or I; how could they be?” To acknowledge our mutual humanity with the perpetrators is to question our very essence as moral beings. However, a great deal of research reinforces this very point – the perpetrators are, for the most part, like you and me. They were not born malevolent, but were shaped by social forces and circumstances. To seek this understanding is not to frame the actions of the perpetrators in a way that excludes them from the exercise of free will. Although this book will emphasize the influence of the moral context on perpetration, it will not place perpetrators within a framework of inexorable social forces; to do so would be misleading – removing perpetrator agency and individual variance. Not all people, when placed in similar circumstances, will choose to perpetrate. If we are ever to have any hope of ending genocide, we must first understand it. This understanding can then form the basis for preventative actions: actions that represent an affirmation of the worth and dignity of human life. While this book will utilize scientific methods to gain a degree of insight into the causes of genocide, it will humanize genocide, rather than presenting it as an intellectual

Introduction   3 abstraction. To understand genocide, we must open the doors of perception, if only a little. We must imagine the unimaginable. The crime of genocide Genocide is a term that is often used but seldom understood. Lay persons may refer to any great human evil as genocide. This, of course, is a concept so broad as to be completely meaningless, much less a crime which may be prosecuted. However, even among scholars, there are many competing definitions of genocide. Much of the definitional quagmire arises from the diverging purposes of social scientists and lawyers. Genocide, as socially defined, is necessarily broad, attempting to incorporate all aspects of genocide and genocidal behavior for analytical purposes. The legalistic conceptualization of genocide is far more focused, individualized, and restrictive. As a work of criminology, this book applies a social scientific perspective to the study of a legally-­defined crime. The legal definition of genocide is largely encapsulated in Articles 2 and 3 of the 1948 Convention on the Prevention and Punishment of the Crime of Genocide: Article 2 In the present Convention, genocide means any of the following acts committed with intent to destroy, in whole or in part, a national, ethnical, racial or religious group, as such: (a) killing members of the group; (b) causing serious bodily or mental harm to members of the group; (c) deliberately inflicting on the group conditions of life calculated to bring about its physical destruction in whole or in part; (d) imposing measures intended to prevent births within the group; (e) forcibly transferring children of the group to another group. Article 3 The following acts shall be punishable: (a) (b) (c) (d) (e)

genocide; conspiracy to commit genocide; direct and public incitement to commit genocide; attempt to commit genocide; complicity in genocide.

The essential components of genocide are conveyed here, including the objective elements (the criminal acts) and the subjective elements (criminal intent, namely the intent to destroy the group). The Convention definition of genocide has itself

4   Introduction been essentially reproduced in the charters of the International Criminal Court (ICC), International Criminal Tribunal for Rwanda (ICTR), and International Criminal Tribunal for the Former Yugoslavia (ICTY). It has also been replicated in domestic legislation, particularly implementing legislation for the Rome Statute. This domestic legislation may take modified form – for example, the inclusion of political groups in the domestic legislation of several countries including Ethiopia, Bangladesh, and Poland. As this definition shows, genocide is primarily focused on acts intentionally directed at the physical or biological destruction of ethnic, national, racial, or religious groups. Other forms of cultural destruction are not encompassed by this definition, nor is the destruction of other types of groups (such as political, class, or gender groups). The element of intent to destroy a group is what most distinguishes genocide from other crimes, such as crimes against humanity or murder. A regime can commit killing on a massive level, but this is not genocide unless it is directed against one of the protected groups, with the intent to destroy that group in whole or in part. While there are many nuances to this definition, as interpreted by courts, these are the essential elements. Most social scientific understandings of genocide move beyond the definition of genocide codified in the Genocide Convention to incorporate other forms of mass killing and extermination. Some of these definitions emphasize the asymmetrical nature of the violence, while others stress the number of victims. Frank Chalk and Kurt Jonassohn, for instance, define genocide as “a form of one-­sided killing in which a state or other authority intends to destroy a group, as that group and membership in it are defined by the perpetrator.” 6 There is also much debate over which groups (if any) should be a part of the definition. In particular, the exclusion of political and social groups from the legal definition is often criticized by social scientists. While many social scientific definitions may be analytically useful, they lack the precision and rigor to be tenable in criminal law. By necessity, the legal definition of genocide is much more precise, and refers only to certain acts, perpetrated against an ethnic, racial, religious, or national group, with the intent to destroy that group in whole or in part, as such. The liberty and reputation of an individual hangs in the balance in criminal trials. Thus, criminal definitions must be precise in order to be just, and also to meet the requirements of the principle of nulla poena sine lege (no punishment without law). This book will use the legal definition of genocide; however, it will not be bound by this definition, just as critical criminology does not confine itself to legal definitions of criminality.7 Hence perpetrators of (possibly) non-­genocidal mass atrocities are also included here (Uganda and Cambodia). Non-­genocidal episodes of mass killing involve many similar social processes as legally-­defined genocide, particularly at the micro level of individual perpetration. A word must also be said here about the term “perpetrator.” This book uses it in a functional way – to describe individuals who commit acts that directly contribute to genocide. These acts are normally considered to be crimes, but we will also consider perpetrators in cases where there have been no criminal

Introduction   5 convictions. Being a perpetrator is neither a static nor an all-­encompassing category: some perpetrators are also victims. The focus herein is especially on killing, and on rank and file killers, though genocide may take different forms and is often effected by multiple means, including not only killing but sexual violence, the prevention of births, and forced displacement.8 This book seeks to establish a criminological theory of the perpetration of genocide, examining legal phenomena (crimes) through a social lens. The enigma of causation Genocide is a complex social phenomenon; thus, it lacks a direct and definitive cause. Rather, it is necessary to adopt an intuitive approach emphasizing adequate causality – that it was reasonably likely that the outcome would be that one, rather than condition sine qua non (that the act in question could not have happened otherwise). This is the approach taken in this book. It does not argue that the explanation proposed is the only one possible. Rather, it seeks to contribute to the body of theories and empirical data that help us understand the perpetration of genocide. While the analysis in this volume is multilevel, the focus is on the micro level of the individual perpetrator. It will particularly examine the moral universe of perpetrators, and ask what perpetrator narratives tell us about the causes of genocide. Genocide studies is a relatively young field. Genocide has not often been a main subject of inquiry, and is rarely treated extensively in general texts. As Alex Alvarez notes, this reluctance might be due to the perceived complexity of genocide, the perception that it is something that happens in “other” societies beyond those of the researchers, and the difficulties in measuring and testing hypotheses.9 Moreover, criminology and other social sciences, have at times been implicated in genocidal processes; particularly, as a means for the state to label marginalized groups as “criminal.”10 To this list might be added an existential dilemma: that genocide, because of its outrageous and exceptional nature, is particularly difficult for people (including researchers) to comprehend; therefore, there is an aversion to studying it. For instance, Elie Wiesel argues that genocide can only be understood “metaphysically, spiritually, or not at all.”11 As a scholarly work this book already contains the assumption that genocide is a social phenomenon rather than a spiritual or cosmological one. It also accepts that cases of genocide may be compared, as socio-­historical phenomena. There is a paucity of criminological literature on genocide, in spite of the fact that genocide is often viewed as “the crime of crimes.”12 Although there have been studies of mass or collective violence in the United States, mass criminality on the international level has received little attention.13 Furthermore, genocide has been historically neglected by criminologists in their scholarly conferences and publications.14 More generally, criminology has given relatively little attention to state or mass crimes compared with more individualistic offenses. It must be noted, however, that a significant body of

6   Introduction literature in criminology addresses organizational crime. This has some relevance for genocide, but relatively little work has been done specifically on state crimes.15 The application of criminology to genocide studies offers notable possibilities for innovation and insight. Genocide differs in significant ways from “ordinary crimes.” These differences must be borne in mind as we apply criminological theory to the problem of genocide. First, genocide is a political crime, committed pursuant to policy or political objectives. The elimination of the victim group is required for the achievement of these objectives. Second, genocide is a crime of conformity, a state crime committed with the support and acquiescence of the state or state-­ like authorities. Third, genocide is a mass crime with a massive number of perpetrators, victims, and bystanders. This means that genocide occurs, in an institutionalized context, as part of a systematic campaign. It also means that perpetration occurs largely in group settings. Fourth, genocide is an intergenerational crime, where past episodes of violence, and collective memory, inform the perceptions of individual perpetrators. Moreover, the objective of genocide is to preclude the existence of future generations of victims. The attention to genocide within the social sciences and humanities has grown significantly, particularly in the aftermath of contemporary genocides such as those in Bosnia, Rwanda, and Darfur. The advent of international courts with jurisdiction over international crimes has also meant that the legal literature has proliferated. Despite the difficulty of proving causation, there is still a sizable body of literature that seeks to do just that. As befitting the complexity of genocide, this research spans the social sciences and humanities, including sociology, criminology, anthropology, political science, psychology, and history. Nonetheless, most of these books offer syntheses of existing research rather than significant original research. Moreover, although some books do focus on genocidal perpetration at the micro level, such as Ordinary Men by historian Christopher Browning, Killing Neighbors by Lee Ann Fujii, and The Order of Genocide by political scientist Scott Straus, there are few comparative studies on this aspect of genocidal killing.16 A handful of books treat genocide from a criminological perspective. Alex Alvarez, a pioneer in the field, has written several books applying criminological theory to genocide.17 Joachim Savelsberg’s Crime and Human Rights conceptualizes the criminology of atrocity crimes, while Nicole Rafter’s The Criminology of Genocide offer a multi-­level framework for studying mass atrocities from a criminological perspective.18 Darfur and the Crime of Genocide by Hagan, Rymond-­Richmond, and Parker adopts a criminological perspective in examining the Darfur genocide by making use of victimization surveys.19 Beyond these books, there are also several edited volumes featuring chapters on various aspects of the criminology of international crimes, and an increasing volume of journal articles.20 While all of this work undoubtedly contributes to our understanding of genocide, there is a need for empirically-­grounded, comparative studies.

Introduction   7

Methodology and objectives of this book Theoretical approach Most theories of genocide causation focus either on the political context of genocide or on the pathology of individual perpetrators. Few adequately address the link between individual acts and social conditions. Ideology is often presented as being a central factor in the process of transforming “ordinary men” into “willing executioners.” Yet explanations of genocide that focus on top-­down determinants of individual behavior all suffer from a failure to explain the variability of individual behavior. If macro factors are all-­important, then we would expect to see all individuals residing in a genocidal state behave in a similar fashion: as determined and obedient perpetrators. Most theories pay very little attention to the process through which macro factors come to define the behavior of individual perpetrators. Nor do they account for divergence – the numerous individuals who live in the same society but never participate in violence. This book will focus on this “missing link” in genocide scholarship by addressing the interaction of the individual with the context as a cause of individual participation. It will especially focus on the moral universe and perspective of perpetrators of genocide. I argue that most perpetrators, like most people, respond to their surroundings in ways that minimize risk and provide coherence. Some perpetrators are opportunists, using the space that genocide provides to harm others in pursuit of their own interests. For others, perpetration is motivated less by hatred and more by an inability to reckon with the moral rupture which genocide represents. They would rather conform than face ostracism; challenging the existing social order can also undermine self-­identity. Social norms are enforced through social sanctions such as ostracism, but also by direct coercion and other forms of punishment. Norms are created and sustained through compliance with both formal authorities and informal expectations. Genocidal states are criminogenic; the crime of genocide is encouraged or permitted by the state. Therefore, genocide can only be understood within the enabling moral context created by state power. Even perpetrators who may not share the ideological beliefs of the regime may perpetrate genocide when faced with the alternative of resisting state power. Perpetrators of genocide are both criminals and good citizens. Perpetrating Genocide draws on extensive semi-­structured interviews with perpetrators of genocide. These interviews sought to ascertain how perpetrators contextualize and frame their behavior – how they reconcile their transgressive actions with ‘conventional moral rules. The book deploys an original theoretical approach integrating situational action theory with neutralization-­drift theory, as well as other theoretical insights drawn from sociology, psychology, political science, history, economics, and law. Situational action theory posits that crimes occur as a function of propensity (derived from individual moral rules and moral habits, mediated by self-­control) and exposure (the moral context). In normal circumstances, the moral context

8   Introduction prohibits crime, but in genocide the state is itself crime-­producing (criminogenic). Situational action theory normally holds that “individuals who have moral rules and moral habits that correspond to what is stated in the law will tend to have a low crime propensity.”21 However, the opposite may be true in genocide, where legitimate authorities may endorse criminal behavior. The conflict between previously held moral beliefs and the evolved moral context is resolved through perpetrator rationalizations. These techniques of neutralization allow perpetrators to revise and reframe previously held moral rules so that they are consistent with perpetrators’ own actions in the evolved moral context. This moral context is inculcated through state power. In summary, individual action in genocide can only be understood in the context of individual interaction with social expectations. Individuals will adjust their moral beliefs in response to changed conditions, as dictated by coercive and persuasive forms of power. This book argues that the perpetrator’s adjustment to the criminogenic moral context is the central driving force in perpetrating the crime of genocide. Methodological approach: towards a criminology of genocide perpetration In exploring these research questions, a criminology of genocide perpetration is constructed. This means two things: first, the use of criminological theory to examine genocide, and second, the development of an interdisciplinary criminology of genocide incorporating the causes (criminogenic conditions) and control of the individual perpetration of the crime of genocide. Such an enterprise poses several methodological challenges and opportunities. Genocide research is difficult due to the contentious nature of the subject itself: it is often politically sensitive, and this poses risks for the researcher and subjects. Moreover, due to elevated risks, conducting field research during a genocide is essentially impossible. Research must be done after the fact, when less information may be available and memory is time-­altered.22 Interviews illuminate individual perspectives that are often left out of other research methodologies (such as archival research) which tend to privilege accounts from elite actors. Sensitive, micro-­level narrative research, such as that of Lee Ann Fujii, Erin Jessee, and Jennie Burnet in Rwanda, and Timothy Williams and Alexander Hinton in Cambodia, illustrate the complexity of individual participation in ways that macro level conceptual or historical accounts fail to capture. The use of interviews, although a rich source of information and insight, is also methodologically challenging. The subjects may distort facts and omit details.23 Moreover, their retelling of their stories may fulfill some inner psychological need (i.e., for coherence).24 Interviewees may feel pressure to relate what they think the interviewer wants to hear, or to align themselves with broader social expectations. Many distortions are not intentional deceptions; our experience is mediated by our understanding of the world.25 In our search for analytical

Introduction   9 clarity and meaning, we cannot neglect the contradictory and contingent nature of the perpetrator narratives in this book. The interview subjects are viewing past actions partly through the lens of the present. They are also striving for internal coherence and consistency.26 They want to view themselves as good people, as is deemed necessary to maintain healthy selfhood. But they also want me, the interviewer, to recognize their goodness and their humanity. In some sense, the interviewer may act as a proxy for the broader moral community, and the interview subject may minimize her responsibility in order to minimize external moral opprobrium.27 This also often involves the alignment of individual narrative with public discourses. Interviews are a co-­production of the subject and the interviewer. They provide information, but they are also about establishing a relationship between the subject and researcher (as well as the interpreter). The context of the interviews is also important: does the subject feel able to speak freely? Although many interviews were conducted in places of incarceration, they never took place in the direct presence of third parties (other than interpreters). Interpreters can affect the interview results via their proficiency in the language of the interview, cultural familiarity, knowledge of qualitative research, interpersonal skills, and relationship with the researcher. The fact that prisoners are deprived of their liberty as a result of the acts under discussion cannot be discounted as an influence on their narrative.28 The subject may reinterpret or distort her memories. Making sense of inconsistent narratives, in which perpetrators contradict themselves, misremember, and obfuscate, may itself be a hegemonic exercise in creating meaning. Yet this study does not primarily utilize interviews in seeking positivist truth. Rather, interviews are invaluable in illuminating the perspective of the perpetrator. The search for positivist truth can be useful in examining the criminal acts that constitute genocide, but there are no positivist truths to be gleaned about the perpetrator mindset. Philosopher Paul Ricoeur argues that the analysis of narratives may serve several purposes. The first is Objective: the establishment of objective fact. The second is Subjective: the establishment of subjective fact (i.e., facts viewed through a subjective lens). The third is Constitutive: narrative as a way of shaping experience.29 The constitutive view aligns most closely with the research methodology utilized in this book: interviews are seen less as a source of objective or subjective fact, and more as a means for the interview subjects to convey meaning about action or selfhood. My interview analysis incorporates each of these perspectives to seek a critical understanding of both the content and function of narratives, and what these narratives tell us about the perpetration of genocide. As Lee Ann Fujii noted, stories may be “emotionally true” even if they are “factually false,” and this emotional accuracy is more important, in some ways, when assessing perpetrator motivation and rationalization.30 In conducting comparative research, there is a danger of universalizing the particular – ignoring inconsistencies in the search for broader patterns. However, there also many potential advantages to comparative interdisciplinary research, such as developing new theoretical and conceptual models, gathering new comparative and case data, and cross-­fertilizing fields of knowledge.

10   Introduction Rather than writing a history of the genocides in Rwanda, Cambodia, and elsewhere, this book has a thematic focus. It does not employ case studies. Rather, it advances a set of theories intended to explain the social processes involved in genocide perpetration. These theories focus on the micro level; accordingly, they draw from interviews with perpetrators and victims of genocide and crimes against humanity. Taken in isolation, they do not offer a comprehensive theory of genocide. Instead, they offer new explanations for individual perpetration. In comparing the results of the interviews, I found more similarities than differences between the contexts. Social context is crucial in creating an enabling environment for killing. This was true of every example I drew from, although conditions varied between cases, with regard to elements such as coercion, discussed on pp.  130–131, and norms relating to sexual violence, discussed on pp. 162–163. Generally, the perpetrators saw killing as wrong, but their role in killing as one in which they had no other choice. This was true both of perpetrators who had undergone formal judicial processes (as with all my Rwandan interview subjects) and of those who had not (most of my Bosnian and Cambodian subjects). In Rwanda, perpetrator narratives seemed more politicized than in other cases. Perhaps this was a response to the increased pressures placed on Rwandans by the strong post-­genocide narrative in that country. This was especially apparent in the way Rwandan interviewees spoke about justice processes. There is a possibility that the interview subjects in Rwanda developed a kind of master-­narrative through their close interaction in prison, and their participation in indoctrination programs such as Ingando.31 Their reproduction of these narratives may have been a result of internalization or a self-­interested desire to produce socially-­acceptable discourses. Nonetheless, the narratives produced by the interview subjects were in many ways also consistent with narratives existing before and during the genocide. This research employs a methodology that is complex and multi-­level, suiting the complex and multi-­level nature of genocide. The methodological guiding principles are as follows: • • •

Grounded: the research methodology and theory is informed by knowledge gained in situ. Integrative: the research integrates theoretical perspectives from multiple disciplines. Qualitative: the research focuses on the nature of a social phenomenon rather than its extent.

The research included the following components: (1) Semi-­structured narrative interviews with perpetrators of genocide and related atrocities in Rwanda, Burundi, Uganda, Bosnia, and Cambodia. (2) Interviews with victims in these countries, as well as Bangladesh and Iraq.32

Introduction   11 (3) Review of relevant secondary sources and legal documents (treaty and case law). (4) Comparative archival research into genocidal propaganda in Rwanda (1990–1994), Bosnia (1992–1995), and the Netherlands (1934–1945). The grounded principle has shaped the methodology and theoretical approach of this book. For example, initially more emphasis was to be placed on the quantitative analysis of mass numbers of perpetrator interviews (or even perpetrator questionnaires). But after completing several interviews, it became apparent that the production of statistical data from such interviews would be misleading. There is arguably a general tendency for individuals to minimize their responsibility for morally wrongful acts; this tendency may result in skewed data. For example, when asked whether he had killed, one perpetrator initially denied it, then later revised his story. Some subjects were taboo in the interviews – particularly the idea that the perpetrator might have enjoyed killing. Sexual violence was also not spoken about openly, even among perpetrators who admitted to having killed. Among respondents, we see the usual linguistic mechanisms of avoiding agency – the omission of the personal pronoun (“I”) and the attribution of acts to other individuals: “the army used to force people to do labor and then raped the women when they finished” (Burundian soldier).33 Interview subjects also often presented themselves as bystanders rather than perpetrators. This was considered in the interview design by including questions about witnessing as well as participating in violence.34 The interviews adopted a semi-­structured approach. They were conducted with a standard list of questions, but also offered scope to include other questions during the interview. The open nature of many questions gave subjects greater ownership of the interviews, and offered the researcher a greater chance to assess their narrative. This method also incorporated meta-­data such as the disposition of the subjects when answering questions, their falsehoods, evasions, silences, denials, and responses to the interviewer and interpreter.35 The subjects’ selective appropriation – what they left out – may be as remarkable as what they chose to tell. Narratives show how we collect and interpret facts and arrange them cognitively to help us find order and meaning in our reality.36 Almost all interviews were also audio recorded, although interview subjects were also provided with anonymity in the hope of achieving more candid results. The completed interviews were also supplemented by interview data from other sources, such as Michael Mann (the Holocaust and Bosnia), Christopher Browning (the Holocaust), Jean Hatzfeld (Rwanda), Scott Straus (Rwanda), Alex Hinton (Cambodia), and Gitta Sereny (the Holocaust). These sources helped to corroborate, complement, and connect research findings. The most substantial original research for this manuscript consisted of 80 interviews conducted in Rwanda in 2009. However, this study also draws from an additional 60 interviews in Burundi, Uganda, Bosnia, Cambodia, and Bangladesh to provide comparative context. In January 2016, I also travelled to

12   Introduction Iraq, where I interviewed 25 victims from minority communities targeted by the Islamic State. The interviews with perpetrators in Rwanda covered a broad cross-­section of society – see Table 1.1. In order to account for possible gender differences, women were sought out as interview subjects. Of these perpetrator interviewees, 59 were male and nine were female. In each country, I strove to achieve diversity in my sample by randomly selecting interview subjects, as well as by conducting interviews in a variety of locations. Interviews in Rwanda occurred throughout the country in both prisons and prison camps (TIG or Travail d’Intérêt Général) – see Table 1.2. My interviews in Rwanda required the permission of local authorities, whereas in other countries I located interview subjects through local NGOs and investigative work.37 In the prisons and prison camps, I arrived without an appointment, and requested that prison officials allow me to randomly select prisoners convicted of or (in a few cases) charged with genocide. The original research for this book also included an archival study to collect and analyze propaganda from Rwanda, Bosnia, and the Netherlands (particularly propaganda from the Dutch Nazi Party, the NSB, or Nationaal-­Socialistische Beweging in Nederland). This identified similarities and differences in propaganda narratives across several cases of genocide. Propaganda materials were collected from online archives, as well as the archives of the NIOD Institute for War, Holocaust, and Genocide Studies in Amsterdam. Structure of the book This book seeks to examine all facets of individual perpetration of the crime of genocide, beginning with the macro context, before moving to individual variance, perpetration, and individual and collective responses to perpetration. Chapter 2 examines the emergence of the genocidal context as a process of escalating radicalism in response to strain. In Chapter 3, the institutionalization of the moral context is examined in terms of harm-­producing relations of power. These relations are grounded in the moral disengagement of perpetrators (and society) from the victims through the production of social distance, as well as Table 1.1  Rwanda interview sample – occupation in 1994 (n = 68; perpetrators only)38 Occupation

Number and percentage of total

Government official Other professional (i.e., lawyers, doctors) Business Farmer Other manual/low-income labor Unemployed

  9 (13.2)   6 (8.8)   4 (5.8) 41 (60.2)   8 (11.7)   2 (2.9)

Introduction   13 Table 1.2  Location of Rwandan perpetrator interviews Location

Number

Kigali Central Prison Remera Prison Gisenyi Prison Ruhengeri Prison Gitarama Prison Tumba TIG Camp Rugerero TIG Camp Butare (with working prisoners) Rwaza TIG Camp Nyarusenge TIG Camp Mont Kigali TIG Camp Hindiro TIG Camp Nyamata TIG Camp

11 2 2 4 8 4 5 4 5 4 7 6 4

the externalization of responsibility. Chapter 4 analyses the role of discourses in contributing to the moral context of perpetration; it argues that propaganda communicates normativity. Chapter 5 examines the propensity of individual perpetrators, arguing that propensity is a function of moral disengagement and proximity (to perpetrating organizations and individuals), but also that even individuals who are not morally disengaged may still perpetrate to maintain their alignment with the political community. Chapter 6 is focused on the process of perpetration itself – it argues that a perpetrator’s decision to perpetrate is driven by propensity, as well as the perpetrator’s interaction with the moral context. Chapter 7 examines the nature of genocidal killing– the mobilization of perpetrators, the nature of genocidal violence, and the range of options available to perpetrators. It argues that mobilization is not random, but there is an element of randomness – some individuals will find themselves in situations where there are great situational pressures to perpetrate. Chapter 8 examines the discourses used by perpetrators as vocabularies of motive and post-­facto rationalizations, which reduce cognitive dissonance among perpetrators. Chapter 9 considers how genocide is sustained on the individual and collective level. Chapter 10 addresses the period after genocide, especially how individuals and societies come to terms with the perpetration of genocide. Chapter 11 focuses on the prosecution of genocide and contends that the processes that facilitate genocide, especially the criminogenic moral context, must be addressed in transitional justice mechanisms. Chapter 12, the conclusion, explores the implications of this book’s findings in terms of individual moral responsibility and culpability. I argue that genocide is ultimately driven by state power, and by many individuals’ tendency to align themselves with the expectations of their moral community. Reluctant perpetrators confront a system built on domination, in which many good people will behave badly – a system of killing without consequence.

14   Introduction

Notes   1 United States of America v. Ohlendorf et al. “Einsatzgruppen Trial,” Judgment, 3 L.R.T.W.C. 470 (United States Military Tribunal), April 8 1948.   2 Interview R65 (Rwanda), Kigali, October 2009.   3 Google Dictionary, accessed on May 8, 2016.   4 Hoffman, After Such Knowledge, 7.   5 Schlink, The Reader, 157.   6 Chalk and Jonassohn, The History and Sociology of Genocide, 23.   7 Barak, “Crime, Criminology and Human Rights,” 33.   8 Anderson, “Colonialism and Cold Genocide.”   9 Alvarez, Governments, Citizens, and Genocide, 4. 10 Woolford, “Genocide, Affirmative Repair,” 139. 11 Alvarez, Governments, Citizens, and Genocide, 14. 12 Schabas, Genocide in International Law. 13 Yacoubian Jr., “Genocide, Terrorism, and the Conceptualization of Catastrophic Criminology,” 68. 14 Yacoubian Jr., “The Insignificance of Genocidal Behavior,” 10–11. While it remains a relatively small niche, there are now several research networks such as the European Criminology Group on Atrocity Crimes and Transitional Justice within the European Society for Criminology, as well as the Supranational Criminology group, based in the Netherlands. 15 See, for example, the work of Green and Ward in State Crime. 16 See: Browning, Ordinary Men; Fujii, Killing Neighbors; and Straus, The Order of Genocide. 17 See, for example, Alvarez, Genocidal Crimes. 18 Savelsberg, Crime and Human Rights; and Nicole Rafter, The Crime of All Crimes: Towards a Criminology of Genocide (New York: NYU Press, 2016). 19 Hagan, Rymond-­Richmond, and Parker, Darfur and the Crime of Genocide. 20 These edited volumes include: Smeulers and Haveman, Supranational Criminology; Letschert, et al., Victimological Approaches; and Parmentier and Weitekamp (eds.), Crime and Human Rights. 21 Wikström and Svensson, “When does self-­control matter?” 396. Situational Action Theory was first devised by Wikström “Crime as alternative,” 1–37. 22 Wood, “The Ethical Challenge of Field Research.” 23 Monroe, The Hand of Compassion, 283. 24 Monroe, 284. 25 Monroe, 203. 26 Festinger, A Theory of Cognitive Dissonance. 27 In oral history, the construction of such subjectivities is often referred to as “composure,” both in terms of composing a narrative and ensuring the narrative maintains individual composure, that is, “the state of being calm and in control of oneself.” Summerfeld, “Culture and Composure,” 65–93. 28 Fleetwood, “In Search of Respectability,” 52. 29 Presser and Sandberg, “Introduction: What Is the Story?” 4. 30 Fujii, Killing Neighbors, 152. 31 Mironko, “Igitero: means and motive,” 50. 32 Victims herein include both those killed during a genocide, and individual survivors of genocidal violence. 33 Interview Bu3 (Burundi), Bujumbura, September 2009. 34 Erin Jessee identified a similar tendency among her perpetrator interview subjects in Rwanda. See: Jessee, Negotiating Genocide in Rwanda, 154 and 169. 35 Fujii, “Interpreting Truth and Lies,” 153. 36 Monroe, 268.

Introduction   15 37 The Ministry of Internal Security (MINITER), in the case of prisons, and the Executive Secretariat of National Committee of Community Services as an Alternative to Imprisonment, for the TIG prison camps. 38 Twelve interviews with victims in Rwanda were also undertaken, to compare victim and perpetrator perspectives. These interviews focused on the Nyamata region, where several perpetrator interviews were also conducted.

Works cited Alvarez, Alex. Genocidal Crimes. New York: Routledge, 2010. Alvarez, Alex. Governments, Citizens, and Genocide: A Comparative and Interdisciplinary Approach. Bloomington, IN: Indiana University Press, 2001. Anderson, Kjell. “Colonialism and Cold Genocide: The Case of West Papua.” Genocide Studies and Prevention 9, no. 2 (2015). Barak, Gregg “Crime, Criminology and Human Rights: Towards an Understanding of State Criminality.” In State Crime (vol.  I), edited by David O. Friedrichs. Aldershot, UK: Dartmouth Publishing Company, 1998. Browning, Christopher. Ordinary Men: Reserve Battalion 101 and the Final Solution. New York: Harper, 1998. Chalk, Frank and Kurt Jonassohn. The History and Sociology of Genocide. London: Yale University Press, 1990. Festinger, Leon. A Theory of Cognitive Dissonance. Stanford: Stanford University Press, 1957. Fleetwood, Jennifer. “In Search of Respectability: Narrative Practice in a Women’s Prison in Quito, Ecuador.” In Narrative Criminology: Understanding Stories of Crime, edited by Lois Presser and Sveinung Sandberg. New York: New York University Press, 2015. Fujii, Lee-­Ann. “Interpreting Truth and Lies in Stories of Conflict and Violence.” In Surviving Field Research: Working in Violent and Difficult Situations, edited by Chandra Lekha Sriram, John C. King, Julie A. Mertus, Olga Martin-­Ortega, and Johanna Herman. New York: Routledge, 2009. Fujii, Lee-­Ann. Killing Neighbors: Webs of Violence in Rwanda. New York: Cornell, 2009. Green, Penny and Tony Ward. State Crime. London: Pluto Press, 2004. Hagan, John, Wenona Rymond-­Richmond, and Patricia Parker. Darfur and the Crime of Genocide. New York: Cambridge University Press, 2009. Hoffman, Eva. After Such Knowledge: Memory, History, and the Legacy of the Holocaust. New York: PublicAffairs, 2004 Jessee, Erin. Negotiating Genocide in Rwanda: The Politics of History. London: Palgrave, 2017. Letschert, Rianne, Roelof Haveman, Anne-­Marie de Brouwer, and Anthony Pemberton (eds). Victimological Approaches to International Crimes: Africa. Antwerp: Intersentia Press, 2011. Mironko, Charles. “Igitero: means and motive in the Rwandan genocide.” Journal of Genocide Research 6, no. 1 (2004). Monroe, Kristen Renwick. The Hand of Compassion: Portraits of Moral Choice during the Holocaust. Princeton, NJ: Princeton University Press, 2003. Parmentier, Stephan and Elmar Weitekamp (eds.). Crime and Human Rights. Amsterdam: Elsevier, 2007. Presser, Lois and Sveinung Sandberg. “Introduction: What Is the Story?” In Narrative Criminology: Understanding Stories of Crime, edited by Lois Presser and Sveinung Sandberg. New York: New York University Press, 2015.

16   Introduction Savelsberg, Joachim. Crime and Human Rights: Criminology of Genocide and Atrocities London: Sage, 2010. Schabas, William. Genocide in International Law (2nd Edition). Cambridge, UK: Cambridge University Press, 2009. Schlink, Bernard. The Reader. New York: Vintage Books, 1995. Smeulers, Alette and Roelof Haveman. Supranational Criminology: Towards a Criminology of International Crimes. Antwerp: Intersentia, 2008. Straus, Scott. The Order of Genocide: Race, Power, and War in Rwanda. Ithaca, USA: Cornell University Press, 2006. Summerfeld, Penny. “Culture and Composure: Creating Narratives of the Gendered Self in Oral History Interviews.” Cultural and Social History, no. 1 (2004): 65–93. United States of America v. Ohlendorf et al. “Einsatzgruppen Trial,” Judgment, 3 L.R.T.W.C. 470 (United States Military Tribunal), April 8 1948. Wikström, Per-­Olof. “Crime as alternative. Towards a cross-­level situational action theory of crime causation.” In Beyond Empiricism: Institutions and Intentions in the Study of Crime, edited by Joan McCord. New Brunswick, NJ: Transaction, 2004. Wikström, Per-­Olof and Robert Svensson. “When does self-­control matter? The interaction between morality and self-­control in crime causation.” European Journal of Criminology 7, no. 5 (September 2010). Wood, Elizabeth Jean. “The Ethical Challenge of Field Research in Conflict Zones.” Qualitative Sociology, 29, no. 3 (2006). Woolford, Andrew. “Genocide, Affirmative Repair, and the British Columbia Treaty Process.” In Transitional Justice: Global Mechanisms and Local Realities After Genocide and Mass Atrocity, edited by Alexander L. Hinton. Newark, NJ: Rutgers University Press, 2011. Yacoubian, George S. Jr. “Genocide, Terrorism, and the Conceptualization of Catastrophic Criminology.” War Crimes, Genocide, and Crimes against Humanity, vol. 2 (2006). Yacoubian, George S. Jr. “The Insignificance of Genocidal Behavior to the Discipline of Criminology.” Criminal Law and Social Change, no. 34 (2000).

2 The emergence of the genocidal context

Introduction: the genesis of genocide Genocide demands the slaughter of innocents. The state, as a political community, is composed of individuals who must reconcile this slaughter with their own worldviews. Such reconciliation necessitates the transformation of society, institutions, and identity. These transformations ultimately entail the exclusion of categories of persons from the political and moral community. This shift from old norms does not happen overnight. Rather, it occurs in an incremental fashion through both ruptures (moral breakages) and continuities (the perpetuation of conventional norms). Rupture is required to exclude the victim group and legitimize violence against them, while continuity, as embodied in the state, allows individuals within a rapidly changing context to maintain a sense of stability and to contextualize violence as part of the normal functions of the state. Continuity is also present in the persistence of self-­image. In her research on everyday relations between Jews and Non-­Jews in the German-­Dutch border region under Nazi rule, Froukje Damant argues: Normality is a concept with two faces: It is used both in the descriptive as well as in the more evaluative [moral] sense. In a descriptive approach normality is considered as “the familiar and predictable.” This form of normality is based on the presence of patterns and predictability … [moral normality is] the sense of “good and appropriate.” … Moral normality is bound up with an existing moral order, that is, with values of good and evil and of appropriateness.1 In genocide, both forms of normality undergo significant changes. Genocidal states aggressively shift norms around the use of violence, particularly against dehumanized groups. Entire groups are excluded from the community by being labelled deviant or enemy; they are effectively criminalized (subject to the legitimate use of force by state enforcement bodies and their proxies) or categorized as combatants (the legitimate targets of the use of force in the context of an armed conflict). Thus, belonging to such groups becomes an inherently deviant act as the genocidal state wages war against social groups.2 Moral normality is

18   The emergence of the genocidal context fundamentally altered in these circumstances, but descriptive normality, in the sense of what is familiar, also changes to reflect the new morality. The perpetuation of familiar elements (descriptive normality) in the face of radical shifts in moral normality is characterized by Damant as “abnormal normality.”3 The strain that precedes genocide is more than just hardship. It entails a state of existential fear. This fear is derived from uncertainty, as well as the perceived threatening characteristics of the victim group. Strain is externally projected onto out-­groups, resulting in the narrowing of identity, the production of social distance, and moral disengagement. Violence also often escalates in the pre-­ genocide period, as extreme measures become not just palatable but essential. The increasing use of violence represents moments of rupture; Jean Améry used the metaphor of the “the first slap” to describe such an undermining of “faith in humanity.”4 As Cass Sunstein notes: Small shocks to publicly endorsed norms and roles decrease the cost of displaying deviant norms, and rapidly bring about large-­scale changes in publicly displayed judgements and desires. Hence societies experience norm bandwagons and norm cascades. Norm bandwagons occur when the lowered cost of expressing new norms encourages an ever-­increasing number of people to reject popular norms, to a “tipping point” where it is adherence to the old norms that produces social disapproval. Norm cascades occur when societies are presented with rapid shifts towards new norms.5 Genocide entails both norm bandwagons and norm cascades. Perpetrators may perpetrate, despite their personal beliefs, to align themselves with social norms. However, unlike in Cass’s theory, genocidal norm bandwagons are greatly amplified through the exercise of state power. Yet continuities also remain in such periods. In genocide, the fundamental source of continuity is the state. Although radicalized, it still often appears synonymous with the political community, and consequently as a source and controller of legitimate authority (authority which is also accrued through individual deference to state functions). Private relations may also serve as a source of continuity for a time. But these private relationships are also often transformed by public stigmatization and violence. Genocidal states often encroach deeply on the private sphere – a necessary exercise to ensure the breaking of pre-­existing social ties that underpins the total enterprise of genocide. This chapter will conceptualize the processes at work in the emergence of the genocidal context. It begins by analyzing shifting norms around violence, then discussing the state of disequilibrium, or strain, which often precedes genocide. It then conceptualizes genocide as a response to strain through the concentration of power in order to remove obstacles and existential threats. Radicalization occurs through escalating moral breakages and violence. Many individuals seek coherence in such circumstances, adjusting to a new moral normality to maintain coherence and alignment with conventional values.

The emergence of the genocidal context   19

The prohibition of killing Perpetrating genocide contravenes and undermines an array of norms. Most fundamentally, the prohibition on killing is revised to enable the killing of certain categories of people. The prohibition is nearly universal in human societies, with notable exceptions, such as acts of self-­defense and lawful killing in warfare (which is often cast as collective self-­defense). Perhaps, then, this prohibition can be refined to include only the killing of innocents. In Rwanda, for example, customary law dictates kwicha kirazira (“killing is taboo”) and an abominable (ishyano) act. This taboo maintains that “blood spilled in violence or “bad death” can inflict extreme mental or physical illness upon the perpetrator and anyone else who comes in contact with it.”6 For example, one of Erin Jessee’s interview subjects argues that he “will never have peace because he has touched blood.”7 In the Bosnian wars and genocide, the killing of civilians was often militarized in both its organization and its framing. Villages were “cleared” of insurgents in much the same way that Jews were sometimes referred to as “insurgents” during killing operations in the Holocaust. This allowed perpetrators to regard violence as an act of ethno-­religious self-­defense, undertaken in response to the existential threats that other groups posed. Paramilitaries often spoke openly of violence among their peers, but maintained the taboo with their families.8 Acts of extreme violence (such as torture and rape) often took place in private spaces in concentration camps, within hearing of other prisoners but separated from the public. In Cambodia, certain forms of genocidal violence were also rationalized. Once again, this violence often took place out of the public eye (such as in detention facilities, often with music playing to drown out victims’ cries), with the notable exception of punishments for moral offences (like extramarital sex). Those responsible were sometimes publicly executed by commune or district-­ level chhlop (intelligence cadres).9 Cambodia is a predominantly Buddhist country; Buddhist beliefs condemn violence as arising from ignorance and desire.10 These clashed with Khmer Rouge norms of “disproportionate revenge” against class enemies and counter-­revolutionaries, engendering cognitive dissonance.11 Authorizing genocidal killing requires redefining victims as threats and reconstituting murder as self-­defense. This occurs largely through the social production of fear of the other.

Catastrophic loss: the context of the genocidal state Strain and adaptation In the modern era, genocide has generally been facilitated by exceptional social conditions. These conditions entail a state of disequilibrium generated by fundamental socio-­economic and political upheaval and collapse. Under such conditions, individuals may evince a reduced allegiance to conventional moral norms.

20   The emergence of the genocidal context Strain theory (originated by sociologist Robert Merton and further developed by Robert Agnew) argues that society creates aspirations and cultural goals without providing adequate structures or means to realize them.12 This can result in a state of anomie or “normlessness” in which society loses its moral authority. Crime is one adaptation to anomie.13 Agnew argued that the three primary sources of strain are: first, when an individual or collectivity is blocked from achieving a goal; second, when something valued is threatened or lost; and third, when something negative or undesired is imposed.14 The loss of group power or status may produce a sort of collective strain and collective anomie, which may be addressed by extremist ideologies and policies.15 Collective strain may produce moments of “cognitive opening” among individuals, as they become receptive to ideas they may previously not have considered. As Zygmunt Bauman argues, periods of “deep social dislocation” where society seems: so formless – “unfinished,” indefinite and pliable – literally waiting for a vision and a skillful and resourceful designer to give it a form. At no other time does society seem so devoid of forces and tendencies of its own, and hence incapable of resisting the hand of the gardener, and ready to be squeezed into any form he chooses. The combination of malleability and helplessness constitutes an attraction which few self-­confident adventurous visionaries could resist. It also constitutes a situation in which they cannot be resisted.16 These visionaries may be ethnic entrepreneurs who cultivate prejudice to serve their own instrumental ends. In effect, strain is externally projected, through propaganda, onto the victim group. Alongside this scapegoating, there is also an aggrandizement of the in-­group, which builds internal cohesion, but also makes collective and individual identities vulnerable to narcissistic wounds, to perceived attacks and disrespect.17 Ethnic entrepreneurs are also “norm entrepreneurs,” desiring to shift social norms. They attempt this by signaling their desire for social change (for, example through propaganda), seeking allies, and reducing the costs of breaking norms.18 The ideological vanguard of the perpetrators often comes from areas disproportionately affected by the loss of power. For example, in its early days, the Nazi party in Germany was dominated by those from border regions such as East Prussia and Silesia.19 Similarly, two of the top three Khmer Rouge leaders were from border regions that had been occupied by Vietnam and Thailand at various points in Cambodia’s recent history. Many Khmer feared Vietnamese domination, and the Khmer Rouge stoked these fears. The role of borderlanders may be celebrated in propaganda; according to Serbian nationalist politician Biljana Plavsic: The Serbs in Bosnia, and in the border regions in particular, have maintained and sharpened the ability to recognize danger for the nation early on

The emergence of the genocidal context   21 and to develop the mechanisms for its defence. In my family, it was said that the Serbians in Bosnia are much better than those in Serbia … I am a biologist, and therefore I know that these species which live next to others and are therefore constantly threatened by others have the best possibility to adapt.20 Situations of collective strain, when coupled with other factors like extremist ideologies, are ripe for the emergence of genocidal contexts. Indeed, where the out-­group is presented as an obstacle or a threat, this may justify the concentration of power and the use of violence against that group. The imposition of undesirable conditions on the in-­group, for example through a power-­sharing agreement, may also contribute to a perception of existential threat. In-­group cohesion may increase in response to this state of uncertainty.21 Strain theories hypothesize that structural breakdowns lead to an inability of societal institutions to control people’s behavior. Social ecology theories of crime, such as William and Kelling’s “broken window” thesis, argue that as visible disorder increases, so too do fear and isolation.22 This can create a kind of security dilemma on the micro level, in which criminal behavior becomes necessary to combat the criminal behavior of others. Crime in states of social disorganization may be a result of anomie, as well as contributing to the sense of normlessness. In failed states, anomie may contribute to the pervasiveness of ordinary crime and even war crimes (which often arise from indiscipline), but not necessarily crimes against humanity or genocide. These require greater central control, discipline, and systematic design. Genocide is not a situation of normlessness. Rather, the genocidal state represents itself as restoring order in situations of perceived normlessness – it responds to crisis by providing decisive action. Blame for unmet aspirations is deflected onto marginalized groups. In genocide, uncertainty over norms is quickly remedied by the assertion of new norms. Collective crime is one adaptation to collective strain. This crime is political, targeted at the source of strain, namely the victim group. Political and economic crises and strain There are many examples of states that have committed genocide in atmospheres of economic or political decline and strain. The Armenian genocide took place in the aftermath of Ottoman imperial losses in the Balkans and elsewhere. Between 1750 and 1914, the size of the Ottoman Empire was halved. The Holocaust occurred in the aftermath of Germany’s traumatic losses in World War I, and the Rwandan genocide in the context of the decline of the Habyarimana state. Interwar Germany and Austria provide classic examples of extremist political movements arising out of strain. During this period, stagflation racked the German economy – an American dollar was worth 4.28 German Marks in 1914, but 5,349,144 marks in 1923.23 Moreover, although Jews only represented at

22   The emergence of the genocidal context most 3 percent of the German population, they had a disproportionately visible presence in commerce, the arts, and education. In 1930, German Jews owned two-­thirds of German department stores and chain stores, half of all textile firms, and the two largest publishing houses; they also constituted a majority of lawyers in Germany.24 Vienna had a particularly large Jewish presence: 62 percent of lawyers in the city were Jewish, along with 47 percent of physicians, and Jews owned 70 percent of the wholesale and retail businesses in the city.25 Yet prejudice itself cannot arise solely from resentment over a disproportionate share of wealth; without pre-­existing prejudice “Jew” would not be an economically relevant category.26 Similarly, Yugoslavia in the 1990s was on the brink of economic collapse and disintegration, with falling production, huge foreign debt, and inflationary pressures.27 Among Serbian nationalists, demographic shifts in Kosovo became emblematic of the threat posed to the Serbian nation, as Serbs, in the course of hundreds of years, saw their presence in Kosovo reduced to only around 20 percent of the total population.28 In pre-­genocide Rwanda, the invasion by the RPF (Rwanda Patriotic Front) forced the Hutu-­dominated government to share power and pledge that the army would be made 50 percent Tutsi (a great over-­representation). Moreover, Hutu refugees had flowed into Rwanda after the massacres in Burundi in 1972 and 1993, and this greatly increased land pressures. In fact, between 1984 and 1989, the average size of landholdings in Rwanda had shrunk by 12 percent.29 This was only exacerbated by the decrease in commodity prices. Rwanda before the genocide was also enduring severe economic hardships as rapid population growth was coupled with a series of economic shocks between September 1989 and October 1990. Per capita income fell precipitously from US$320 in 1989 to only US$200 by 1993.30 These conditions retarded the upward mobility of young Rwandan men, who struggled to obtain the status of their parents.31 Rwanda was also a very young country, with 57 percent of its population under the age of 20 in 1990.32 Hutu elites, who had benefited from clientelist relations with the Habyarimana regime, saw their power eroding at the hands of the Tutsi-­dominated RPF. The Akazu patronage state further constricted opportunities through its cronyism, whereby people from Habyarimana’s power base of Gisenyi and Ruhengeri communes occupied half of all senior positions in state corporations. Moreover, Gisenyi, Ruhengeri, and Kigali consumed half of the government budget for regional development.33 Even so, state patronage was increasingly limited under the Arusha Accord’s power-­sharing arrangements, while the bureaucracy shrank as a result of structural adjustment.34 There was also an ethnic dimension to the exercise of state power and patronage. For example, in stark contrast to Tutsi domination in the pre-­colonial and colonial periods, there were no Tutsis among the 143 burgomasters (heads of communes).35 Although many of the perpetrators interviewed for this study indicated they were satisfied with their overall living situation in 1994, many also contended that Tutsis had more than their fair share of Rwanda’s wealth.

The emergence of the genocidal context   23 Existential fear is exacerbated by sudden shifts in the socio-­political order. These shifts can occur for myriad reasons, including regime change, war, refugee flows, and economic collapse. The movement of refugees may have profound destabilizing effects as, by their very presence, they disrupt the local ‘ethnic balance of power’ and contribute to the impression that strangers (‘others’) threaten local groups. For example, at the end of World War I, there were around 70,000 recent Jewish immigrants from Poland and Galicia in Berlin alone.36 War, of course, often represents a disruptive force. Everyday life is uprooted, and hard norms soften or disappear. Warfare generates a heightened sense of threat in which draconian measures may be tolerated, and it provides cover from external interference. War can also lead to the militarization of society, providing a compelling rationale for killing (for security or in self-­defense). It normalizes extreme violence, and necessitates the involvement of “violence specialists” such as soldiers, police, and paramilitaries.37 In Rwanda, this militarization of society was embodied in the civil defense program (launched in 1990) with its neighborhood civilian security patrols (Amarondo), and in the distribution of small arms to civil administrators and army reservists. The army was also expanded, and paramilitary groups such as the Interahamwe were formed. After the genocide commenced, many acts of killing were framed through structures and discourses developed before the genocide. Killing Tutsis was presented as rooting out infiltrators, and killing groups were sometimes called Amarondo. In Burundi, the army doubled in size during the civil war (1995–2002), and recruited only Tutsis.38 Warfare can be transmuted into the mass killing of civilians if those civilians are redefined as combatants. Militarization is also a preparatory stage for the mass mobilization of violence by soldiers or civilians. One Rwandan recounts: “Before 1994 I didn’t expect that the Tutsi would be killed, but I saw signs of the killings because they started to give weapons to the people.”39 Militarization expands participation in violence as well as the targets of violence; it represents a shift in the norms and institutions of social control. States of anomie and strain are ripe for genocidal solutions, particularly when they also feature a group perceived to be an existential threat.

The drivers of the emergence of the genocidal context Genocide is not generally an initial political goal. Rather, it is a “last resort” aimed at forcefully consolidating state power in the face of perceived external or internal threats and obstacles. In other words, genocide is more functional than intentional, although some level of planning is always required to produce the genocidal context. The Armenian Genocide, for example, was a systematic policy pursued by the Ottoman Turkish government in response to the perceived threat posed by the Armenian population. It occurred, of course, in the context of an armed conflict, and was driven by fear of Russia (which had invaded the Empire from the east), as well as by a hegemonic Turkish nationalism that was distrustful of ethnic minorities such as the Armenians. It was a draconian response that, in an atmosphere of insecurity and fear, escalated into genocide.40

24   The emergence of the genocidal context When an ethnic group is blocked from achieving a goal (which may be territorial, economic, political, or, more amorphously, a national destiny), this can contribute to a sense of collective strain. Consequently, the removal of the obstacle (in this case the group) may be seen as a solution. The removal is seemingly instrumental, a means to an end. Yet prejudicial ideologies enable the exclusion of the target group from the moral community, and undermine the normal moral rules that would prohibit their destruction. Indigenous groups were often presented as obstacles in colonial genocides, representing a hostile component of the landscape or natural environment. Colonial policies of racial hierarchy also contributed greatly to the hardening and stratification of ethnic identity in many countries, including Rwanda and Burundi. Out-­groups may also be seen as posing an existential threat to the survival of the in-­group. This threat is always driven by stereotypes applied to the victim group, derived from both contemporary events and historical memory. Through the construction of the victim group as an existential threat, each individual member of the victim group becomes tied to an abstraction – the eternal enemy, the barbaric enemy, the inhuman enemy, and so on (see Chapter 4). One female perpetrator in Rwanda recounts: What I thought was … the Tutsis were the enemy of the country of Rwanda. So, for me the Tutsis were not only enemies for Rwanda but also for myself. That is why I was feeling more and more encouraged, so that before the enemy kills you, you better kill them.41 But despite this attempt to tie individuals to negative abstractions, these definitions of the victim may not “stick” until they are reinforced with practical evidence. One interviewee recalls: I used to see teachers separate Tutsis and Hutus, telling Tutsis to stand up … every Tutsi stood up. And since that time, I was seeing differences between Tutsi and Hutu.… But these differences did not mean anything to me – some of them were my friends.… The hatred started when the genocide began.42 Many other interviewees had similar experiences in Rwandan schools: When we were in the class, the teacher taught us how the Tutsi walks, how he looks, even showed a Tutsi to the class. When I was in secondary school … the teacher told us, before we entered the classroom, to separate themselves: the Hutu on this side, the Tutsi on this side and the Twa on this side. Some of my family members were killed and when I was younger … it was in ’73, the teachers used to come in the class and ask the Tutsi to raise up their hands. They chased the Tutsis out.43

The emergence of the genocidal context   25 Threats may also be produced by historical reinterpretation and mythohistory. A history of antagonism and atrocity between groups makes incitement to genocide salient. All history is socially constructed, but some constructions are more exclusionary and closed to alternative interpretations than others. Ethnic entrepreneurs and other leaders provide interpretations of the past that fulfil their contemporary political aspirations. These revisionist interpretations may also be utilized as a propaganda tool to foment hatred and incite violence. In Rwanda, many perpetrators also heard negative stories about Tutsi domination in years past: “The Tutsi used to oppress Hutus, making Hutus work for them for free… They used to beat them, if you disobeyed that would be punished and you would be beaten.”44 Members of victim groups also often draw from historical accounts that produced social distance; a survivor of the Srebrenica massacre in Bosnia recounts, I heard stories from old people, that in the Second World War they came to villages with guns […] so they entered the villages, they take the people to the forest and killed them in the forest. This is what I knew from the stories from older generations.45 This historical discourse produces a kind of continuum of grievance that serves contemporary political purposes. Narratives of violence are often transmitted intergenerationally. Consequently, genocide may itself be a kind of intergenerational crime, fueled by the memories of older generations but committed largely by the young. A Rwandan perpetrator recalls, “When I was a little boy they used to tell me that the Tutsis were bad people who took Hutus as slaves.”46 Such narratives are a primary channel through which historical enmity is transmitted from one generation to the next as the child, ignoring the surface interpretations and rationalizations, hears the note of helpless fury and impotence in the accounts of beloved adults and fantasizes scenarios of revenge against those who have humiliated his family and kin.47 Furthermore, past victimization may contribute to a sense of vulnerability, as well as a perception that the out-­group poses an existential threat. This sense of vulnerability may lead to violent, pre-­emptive “counter-­attacks.”48 The father of General Ratko Mladic (who bears substantial responsibility for the Srebrenica genocide) was murdered by the fascist Ustase during World War II. Yet clearly, a history of violence alone cannot “cause” genocide, as such histories occur throughout the world and underlie many situations which do not descend into genocide. Moreover, although evidence shows that children of offenders (of ordinary crimes) are likelier to themselves be aggressive and to offend, the causes of this apparent intergenerational transmission remain unclear.49 Both the removal of obstacles and the countering of existential threats may also act as a vector for the concentration of authoritarian power. The presence of

26   The emergence of the genocidal context an enemy allows for the suspension of the rule of law and the construction of a security state. When genocide occurs, it is also generally directed by an authoritarian regime. Authoritarian ideologies are often focused on the maintenance of ideological, racial, and even sexual purity. Those who control the levers of power, elites, have a special position in the production of the genocidal context. Low-­level perpetrators feel they lack the power to steer the ship of state – they are more concerned with meeting their needs for sustenance, success, social valuation, and survival. The concentration of power, indeed genocide itself, is legitimized by the presence of enemies, who pose a threat to shared goals and survival.

Radicalization Genocidal radicalization is the normalization of extreme violence against particular groups. Genocidal radicalization entails a process of moral rupture and escalating violence against the out-­group. The key mechanisms underlying this radicalization are the building of internal cohesion and the exclusion of out-­ groups.50 In tandem, they contribute to the erosion of intergroup empathy. Enemies are first delineated, then socially distanced, then destroyed. The state monopolizes the legal codification of deviant behavior – it has the power to distinguish right from wrong, and to define social relationships. Targeted groups are no longer true subjects of the state, with the responsibilities and entitlements of citizens. Rather, they are living in a twilight zone of legality and legitimacy. Interpersonal relationships are also affected by the categorization of deviancy, with individuals who are labelled as deviants defined as hostis humani, enemies of humanity, and excluded from the community. The classification of groups also represents a shrinking of the moral universe, in which entitlements and trust are ethnically determined and extended only to those within the narrowed political community. Every state exercises the right to control its borders, to demarcate the boundaries of the political community, and to define deviance. But in genocidal states, these powers are exercised in extremis; categories of individuals are placed beyond the protection of the law by virtue of their act of being, their personal identity. Concomitantly, the normal complexity and multi-­faceted nature of social relationships is broken down in service of a restricted focus on certain aspects of identity, which become rules of belonging determinant of life or death. The revision of the moral context is ultimately not possible without a survival discourse that presents the victim group as a threat to the ultimate survival of the perpetrator group. This survival discourse is both reality-­based and imagined; neuroses, rooted in prejudice, serve to bind the imagined to the actual, and to depict the victim group as posing a grave danger. The survival discourse drives the breakdown of social ties and the legitimation of violence. How can regular social ties be maintained with a group that represents an existential (and often eternal) threat? In such conditions, regular social interactions are framed as a struggle for survival. One Rwandan perpetrator explains:

The emergence of the genocidal context   27 Yes, of course I was afraid. They [the authorities] used tell me that once they [the RPF] were getting into the country they would kill us – the RPF would kill us. Also, any survivors would be enslaved, and they would bring back the monarchy.51 The revision of language is fundamental to the reframing of the situation, as Chapter 4 demonstrates. The revision of collective moral rules, and the inculcation of a survival discourse, is realized in part through the production of moral rupture through moments of moral shock, such as, Habyarimana’s plane crash on April 6 1994 (attributed by Hutu extremists to the RPF ), or the burning of the Reichstag on February 27 1933 (alleged by the Nazis to be part of a broader communist plot). The burning of the Reichstag formed a pretext for the banning of the Communist party and the restriction of civil liberties, while the assassination of Habyarima led to immediate reprisal attacks against Tutsis and Hutu moderates.52 Such moments challenge and overturn conventional understandings of security within the political community. They convey extreme insecurity and an existential threat, attributed to an out-­group. These ruptures are only revolutionary when societies lack resilience, in contexts of pervasive and cumulative instability. This radicalization of norms occurs discursively, in ideology and propaganda, but the exercise of actual violence also has a communicative function. It makes the dehumanization of the victims and the acceptability of violence seem self-­ evident. The victims’ degraded status is apparent in the state’s use of violence against victims, and/or the failure to punish perpetrators of violence. The witnessing of (repeated) acts of public marginalization and violence against defined groups communicates normative messages about the need to exclude those groups from the political community. In many cases, authoritarian regimes also demand the increased participation of civilians in state activities, including the use of violence. State radicalization: mainstreaming atrocity Genocidal contexts are characterized by polarization and extremism in all facets of economic, political, and social life. Genocidal radicalization entails the shifting of norms from the conventional, wherein state violence is restricted to the maintenance of law and order, to new conditions in which violence against civilians is deemed acceptable or even imperative. Such a process entails a kind of “criminalization” of entire social groups; belonging to such groups becomes a deviant act in and of itself. Criminalization entails the definition of the group as deviant, its exclusion from full citizenship rights, and the expanding lawful (and unlawful) use of force against the group. Although many Rwandan Tutsis were citizens, their legitimacy was constantly challenged in political discourse. This criminalization is usually broadly communicated, even while acts of violence may still be restricted to violence specialists.

28   The emergence of the genocidal context Radical ideological discourses authorizing or even requiring violence often precede actual violence, but also occur alongside it, functioning as an organizing principle and mode of moral neutralization. The survival discourse is central to genocidal ideology; the destruction of the out-­group is depicted as a means of “self-­defense.” Even while radicalization is often driven by top-­down normative expectations, produced through discourses and state-­sanctioned acts of violence, it may also be facilitated by bottom-­up processes, particularly excessive violence at the local level. The existence of a perceived existential threat drives the reconfiguration of bargaining power among political factions. In Rwanda, for example, there was a radical turn with the advent of multiparty politics after the RPF intervention in 1990 – a state of intensifying polarization and discourse both among Hutu factions and between Hutus and Tutsis. Extremist groups present themselves as a stabilizing force, providing certainty in a period of chaos. They may also gain legitimacy, as the Nazis did, through the direct exercise of power. Their shift from violent opposition to government also entails the institutionalization of violence. This transformation was evident in the dissolution of the Sturmabteilung (SA) in Nazi Germany, and the shift to the more professional Schutzstaffel (SS) to implement state terror. The achievement of political power does not mean that the Nazis, Khmer Rouge, or Hutu Power factions in Rwanda lost their radical ideological orientation, but rather that this radicalism became institutionalized – a highly criminogenic situation for the crime of genocide. The state itself becomes radical – directed partly (but not exclusively) at the destruction of specific populations. Group radicalization: the humanity gap In genocide, our normally complex individual identities are simplified. There is a kind of narrowing of awareness so that: (a) one layer of identity (e.g., “ethnicity”) and one group affiliation within this layer (e.g., “Hutu”) eclipses all others, and (b) other groups within the same identity “layer” are alienated. This aggrandizement of the in-­group and devaluation of the out-­group creates a “humanity gap” – a social distance between perpetrator and victim. Genocide becomes possible in this space. Dehumanization can also be an act of self-­affirmation: the perpetrators’ selves expand into the space created by the negation of the victims.53 Moreover, a “hardening of identity” is evident. The layers of identification most relevant to genocide (race, ethnicity, religion, nationality) are already less fluid, and more politicized, than many other social categories. However, during genocide, one relevant source of identity becomes determinant of life and death. As one perpetrator recounts, “After the death of Habyarimana … it wasn’t the game of politics but the game of race.”54 Another remembers that “before the genocide, we saw the Tutsi as friends, without any problem. But then [when] the genocide came, they taught us that the Tutsi are not friends. They taught us that the Tutsi are the cockroaches.”55 Another argues: “It was a good relationship

The emergence of the genocidal context   29 without any problem before the genocide. But the war came and separated us … it was something that came suddenly.”56 Similarly, in Bosnia ethnic antagonism was not pervasive before the war, but prejudices emerged amidst the conflict.57 Violence itself contributes to the production of social distance. In Rwanda, the situation was transformed in the lead up to the genocide, to the point that “Each group supported their race.”58 During the civil war, a perpetrator remembers, his Tutsi friends “started to change, started to cut off our relationship. From this fact, I began to believe that maybe they were [RPF] accomplices.”59 Similarly, “before the RPF attacked, we [Hutus and Tutsis] were the same.”60 Another respondent remembers, “in 1994 things changed and I started to think that Tutsis were different from Hutus.”61 Suspicions intensified greatly during the civil war, with widespread rumors of Tutsi-­owned bars poisoning their Hutu customers and vice-­versa. This led to bars effectively being segregated by ethnicity.62 In Burundi a similar calculus prevailed. According to a former (Tutsi) CNDD-­FDD fighter: “before the war, I lived in a Hutu-­dominated community. At home, we were not allowed to speak of Hutu/Tutsi differences. It was only during the crisis that I started to see those differences.”63 Violence is an act of negation that requires a justification. It redefines relationships – for some neighbors it may no longer matter that they cultivate the same fields, attend the same church, and studied in the same secondary school. By this point, these sources of identity have been subsumed by the extremism and the survival discourse that characterizes genocide. Many perpetrators are personally acquainted with their victims. One survivor of the 1971 mass atrocities in Bangladesh (from the Jallad Khana killing site) recalled that he asked the people attacking him, his neighbors, “Why are you beating me?” They responded: “because you are Bengali. We don’t know you anymore.”64 Similarly, a Rwandan perpetrator remembers “in truth, it came to me only afterward: I had taken the life of a neighbor … at the fatal instant, I did not see in him what he had been before.”65 This “interpersonal amnesia” is part of wider social processes that narrow awareness. These are likely to have begun long before genocide occurs, but during genocide the narrowing of awareness becomes more extreme, and the humanity gap widens. Humans belong to groups because membership in such groups tells us who we are and what we should do. Every society has a concept of “us and them,” and group memberships are also exercises in boundary-­making. Yet “us” and “them” are not value judgments until they have been married to moral concepts such as “good” and “bad.” Xenophobia, the fear of outsiders, is common, and seems to be grounded in deeper processes of evolutionary psychology; it eases the stigmatization of out-­groups. Leaders cultivate the humanity gap through direct appeals and institutional structures of enforcement and socialization, such as the police and the education system. The narrowing of awareness is facilitated by threat and uncertainty (strain), and by the presence of an already devalued, hence easily scapegoated, victim group. Groups are especially vulnerable to scapegoating and negative social valuation when there are already strong ethnic dimensions to economic,

30   The emergence of the genocidal context social, and political life. Intense competition and polarization between groups seems to increase the saliency of group membership. Group interaction and competition can be a social amplifier: the pre-­existing beliefs of individuals are augmented as members interact without the moderating effect of outside influences.66 Thus, over time beliefs may become more extreme. The genocidal state is a closed system in which extremist views are not moderated by alternative viewpoints (the presence of such alternative views greatly undermines obedience).67 In the case of genocide, outbidding may also occur, in which leaders with ethnic constituencies adopt ever more strident and exclusionary positions to prove their loyalty and gain the support of their group. The viewpoints and emotions of other group members may also grow more closely aligned through the processes of conformity and group-­think. In this way, the opinions of group members become more extreme and closely aligned. This group radicalization effect also underlines the narrowing of awareness. Overgeneralization in situations of danger is adaptive, as it can help to identify threats.68 In other words, the narrowing of awareness provides clarity by simplifying complexity.69 Such “primal thinking” also includes “personalization” – assuming that all stimuli are directed at you; “selective abstraction” – a focus on small details, stripped of their context, that might indicate a threat; and “dichotomous judgment” – the division of others into opposed categories (e.g., friends and enemies).70 The reductionist, binary thought patterns that characterize genocide facilitate killing by greatly reducing the capacity for empathy. The narrowing of awareness is also accompanied by a process of “hostile framing.” Hostile framing theory, developed by Aaron Beck, posits that certain negative characteristics come to be associated with out-­groups through stigmatization and stereotyping. These characteristics are actually cognitions such as “evil,” “bad,” “suspicious,” “dirty,” or “untrustworthy.” Through hostile framing, all the acts of the victim group are interpreted in light of these essential negative characteristics. Indeed, even non-­action may be seen as evidence of plotting and conspiracy-­making. In genocidal contexts, the state becomes the embodiment of the group, whether the group is defined by ethnicity, religion, class, or political affiliation. Moreover, the state, as the embodiment of the people, also becomes a subject not only of obedience, but of veneration.71 In Rwanda, the hardliners gained the upper hand because they controlled state and parastatal structures.72 In December 1991, a military commission identified the country’s “principal enemy” as “Tutsis inside or outside the country.”73 Rwandan elites were feeling increasingly besieged – in the 1993 Arusha Accords, the MRND (Mouvement républicain national pour la démocratie et le development – the longtime governing party) received only a third of the posts in the transitional government. Thus, in Rwanda we see a process of ideological radicalization, as the saliency of ethnicity increased in the context of a heightened “fear of the other” following the RPF invasion; this reduced the space for moderates, ultimately resulting in the eradication of alternative viewpoints. However, the recourse to genocide was not automatically produced by

The emergence of the genocidal context   31 ideological radicalization. Rather, extremists fully seized control of the state through a series of political struggles in a rapidly evolving context after the crash of Habyarimana’s plane.74 These processes of narrowing awareness and hostile framing place the victims outside of the moral universe, with its systems of moral reciprocity and obligations rooted in notions of shared humanity. Individuals are depersonalized and consigned to a negative category. This state of moral disengagement is institutionalized in genocidal states. The ethnic group is reconstituted as a “norm community,” centered on an “ethnic norm” – a set of rules for members of the ethnic group to follow, especially in their behavior towards ethnic rivals.75 Those who violate the ethnic norm may be ostracized or cast out of the group. The ethnic norm is not “natural,” but constituted through political processes of group radicalization. The existence of ethnic norms, communicated through propaganda, is essential for mobilizing genocide. Individual radicalization: the surrender of the will As radical groups become mainstream, they are also able to recruit individuals from the mainstream. This process shifts the balance of risks to the point where non-­membership in the radical group becomes riskier than membership. The state profoundly influences individual identity and action. This does not happen uniformly. Rather, it occurs through a process of individual interpretation and negotiation. The state draws its power from its dominion over individuals; therefore, the concentration of power in state hands also requires the apparent sublimation of individual will (though not individual self-­interest). Under the narrowed awareness of genocide, personal identity becomes particularly determinant of social relations. A genocidal context is one in which individual autonomy is continuously and severely eroded, even while, paradoxically, it presents opportunities for self-­aggrandizement. Genocide also displays simultaneous processes of devaluing the selfhood of victims and aggrandizing that of perpetrators. In other words, the perpetrator’s self grows into the space of the victim, and the group self into that of the perpetrator. The victims’ personal selves are destroyed, while the perpetrators’ become linked to an “exaggerated group self ” through collective action in pursuit of ideological goals. Thus, genocide represents an attack on selfhood involving extreme deindividuation and dehumanization. The individual victim is targeted for destruction based on their collective identity. One might ask why individuals acquiesce to the genocidal state. To enforce an ethnic norm, coercion is required, yet focusing on this coercive element misses the subtler, seductive incentives that individuals gain by aligning themselves with the community. The fear of isolation is certainly a powerful motivator, as isolation often places individuals in a situation of fundamental vulnerability. Belonging to the group is an act of survival, but it also provides a more metaphysical comfort – the sense that one is not alone. This effect may be even stronger in close-­knit groups such as military units. Such groups have sharp

32   The emergence of the genocidal context demarcations of insiders and outsiders and strong norms of group loyalty. These groups also often reward those who best demonstrate loyalty and consonance with group norms. Psychologist Erich Fromm has argued that the certainty of authoritarian ideology satisfies our need to “escape from freedom.”76 The problem with freedom is that it offers not only the potential for individual realization, but also the possibility of individual failure and anomie. The individual habit of obedience to the state, as a moral norm, greatly facilitates the surrender of individual will. The state provides certainty, even if individual doubts persist. Through surrender of individual will to the community, individuals may also achieve transcendence. Our lives are underscored by a fear of mortality (the sense that our selfhood is tied completely to our mortal bodies), and groups provide the promise that our selves will live on beyond the death of our bodies. Groups institutionalize transcendence through the remembrance of martyrs and a sense of higher purpose. Perpetrators do not lose all sense of themselves, but strive to maintain their alignment with the group. Violence changes risk and incentive structures for individual action, and politicizes private relationships. Self-­identity is often transformed to coincide with the revised moral context. This adaptation helps to prevent the development of self-­condemning self-­identity, a psychologically damaging state. The alignment of individual identity and individual ethics with the group increases propensity, by drawing individuals closer to the moral logic of the genocidal state.

Genocide in public and private spaces Through escalating acts of persecution and escalating civilian involvement, violence increasingly moves from the public sphere to the private. In the periods before genocide, the hostile framing of the victim group is actualized through institutionalized discrimination in the public sphere. This occurs because the hegemonic power of the state determines rules of belonging and constructs a system of legal rights and entitlements. Judicial and law-­enforcement institutions play a particularly strong role in dictating the nature of economic, political, and social relations and entitlements. Where violence occurs publicly against members of the victim group, it may be especially powerful. Yet information about privately occurring violence (for example, torture in detention facilities, disappearances) often enters the public sphere, and serves to reinforce messages about the acceptability of violence against the stigmatized victim group, as well as fear of the coercive power of the state. The narrowing of awareness and hostile framing of members of the victim group contribute to the breakdown of social relations in the private sphere. Where the victim group is perceived as immoral, individuals will avoid inter-­ group interaction.77 Yet relations in the public sphere are also transformed, becoming increasingly polarized and extreme. Victim groups are removed from positions of power and influence, contributing to their increasing marginalization

The emergence of the genocidal context   33 and to the power differential between perpetrators and victims. Competition between in-­group elites may be a primary source of polarization once out-­groups are driven out of the political sphere. Elites may also act out of a feared loss of privilege. A former sector counsellor in Rwanda argues: “I was protecting my income – I didn’t realize I was doing something wrong.”78 The fact of polarization further contributes to the perception of grave crisis, fundamental uncertainty, and existential threat. This perception drives people towards the group and allows for the concentration of authoritarian power in the hands of the state’s leadership. 79 Polarization in the economic and political spheres Polarization, discrimination, and conflict in the public sphere are both indicative of a transformation in social norms and a source of this transformation. Individual anxieties about a loss of economic power are existential in a very real sense; without money, there is no security. Economic activity may be ethnicized; indeed, it may be a key location of inter-­group interaction and conflict. Forms of economic engagement where there are no benefits to cooperation, can lead to solutions involving the physical elimination of competition and obstacles.80 The victim group is seen as an obstacle to collective self-­realization, but also as an existential threat to perpetrators’ survival. This is particularly true of land disputes, since land serves as an anchor for physical security as well as a potent symbol of entitlements. Competition over scarce resources may become militarized and racialized, and economic opponents may be dehumanized.81 Moreover, increasing government control over the private sector in authoritarian states may lead to a situation in which the production of wealth depends on political access, as in both Burundi and Rwanda. Systems of patronage increase the intensity of competition, and when they are linked to ethnic identities, provide incentives for the elimination of rival ethnic groups. Political contexts of existential fear contribute to the concentration of power, intensification of political competition, and escalation of state violence against the enemy group. In the atmosphere of intense polarization immediately before the genocide in Rwanda, “you could hate somebody for belonging to a rival party even if you followed the same policy … somebody with no party had a problem.”82 Experimental evidence further demonstrates that in-­group identification increases in the presence of existential fear.83 Identity exclusivism and politicization may be an initial response to perceived threats, but when these threats are deemed a menace to the survival of one’s community, then more violent responses become necessary. In this fearful environment, false accusations proliferate. Serbian Orthodox Bishop Filaret appeared on Serbian television holding a child’s skull in his hands and telling viewers the girl had been beheaded with an axe in front of her mother (a false story).84 Such atrocities demand a proportional response; the commission of atrocities as acts of vengeance, deterrence, and self-­defense. Slobodan Milosevic repeatedly raised the specter of an impending genocide against the Serbs,

34   The emergence of the genocidal context echoing historical atrocities against the group. In situations where the group perceives its options as limited the balance of power may be shifted towards those proposing radical solutions.85 Thus, genocidal politics are defined by polarization and tension, as well as the projection and attribution of atrocities onto the victim group. In pre-­genocidal Rwanda (1990–1994), the Tutsi and RPF were often accused of atrocious acts, and politics became increasingly polarized (and dangerous) as extremists entered mainstream politics, producing an escalating denunciation of treacherous elements. The presence of an enemy within requires a closing of ranks in order to identify and expunge the danger.86 Genocidal politics are often highly nationalistic; rigid rules of belonging are developed as ethnic nationalist movements seek to geographically constitute ethnic identity. Genocide and ethnic cleansing may sometimes occur as an extreme response when competing ethnic groups have plausible and legitimate claims to the same territory.87 The social order asserts the right to purge itself of all alien influences.88 In this sense, genocide may be a “fundamental mechanism for the unification (ethnicization) of the national state.”89 Nationalism effectively binds the individual to the group. Nonetheless, ethnic cleansing and genocide are often a “last resort,” an outcome of internal political competition.90 Extreme violence is normalized in the genocidal context, as an extension of the state’s role of providing security. Violence and social relations The state monopolizes legitimate violence, and can use this violence in certain prescribed ways, in defense of collective interests.91 In genocidal societies, however, the boundaries of violence shift and unlimited violence against entire social groups becomes permissible. This is often presented as an act of self-­ defense. In such societies, people grow accustomed to the use of violence. Underlying genocide is a utilitarian assumption that the state has the right to decide which (human) sacrifices are necessary to attain its goals.92 The only question is how many lives the state can take, not whether it has the right to take lives.93 However, perpetrators can argue that such deaths do not represent a loss of “life,” as targeted individuals are essentially dehumanized. Extreme violence in genocidal regimes is always viewed as instrumental and essential. Common Khmer Rouge slogans exhorted, “when pulling out weeds, remove them roots and all” and “no gain in keeping, no loss in weeding out.”94 This utilitarian view extends to the treatment of mortal remains. The Nazis plundered corpses to remove gold teeth from the deceased, while the Khmer Rouge placed corpses at the bottom of fruit trees and dumped executed bodies in rice paddies for use as “fertilizer.”95 Furthermore, participation in and exposure to violence may desensitize individuals and ease the commission of further violent acts. The widespread use of violence does more than habituate individuals; it also terrorizes potential opponents of genocide and reinforces the impulse to obey authority. The “brutalization thesis” posits that the more people

The emergence of the genocidal context   35 are exposed to “legitimate” violence, the more they desire to commit violence themselves.96 Yet even with the state’s near-­monopoly on legitimate violence, inflicting violence on a massive scale requires an ideological justification. The use of violence in the periods preceding genocide may also act as a kind of experiment, testing limits and further marginalizing victims. Non-­state actors often commit the first acts of public violence. The Rwandan genocide, Holocaust, and Armenian genocide were all preceded by massacres and other acts of political violence. One Tutsi survivor recalls that the Interahamwe and soldiers used to grab and kidnap the young boys of this area and took them to be killed. For the parents, it was confusing – because they did not know where the kids went. Sometimes boys who survived decided to leave, to escape, and to join the RPF.97 Violence often occurs incrementally in the period preceding genocide, through processes of cumulative radicalization and escalation. With the exercise of authoritarian power, private space comes increasingly to resemble public space, and the moral rules governing the treatment of the victim group shifts. Yet civil society institutions, such as churches, may still serve as guardians of the conventional moral order. Such institutions may respond to genocide in several ways: they may cooperate with, and be coopted by, the genocidal state; they may resist its genocidal designs; or they may simply ignore violations of the “old” moral rules. In the case of Rwanda, the Catholic church hierarchy was often passive during the genocide, but some individual priests did protect victims. (Seven of nine bishops were Hutu, but 65 percent of the clergy were Tutsi.)98 Hitler’s personal secretary, Karl Brandt, recalls the dictator saying that if war were to break out, the final solution should be implemented, as the resistance from churches would be diminished in wartime.99 In fact, as Helen Fein’s research shows, churches could play a crucial role in slowing the implementation of Nazi policies in areas where Germany did not exercise direct control.100 The witnessing of public acts of violence, moments of rupture, poses a dilemma for individuals, forcing them to become perpetrators or bystanders. A Rwandan remembers: “the group of men came to my home at night, with knives. Because I had a Tutsi wife I decided to go with them. They took someone and killed him. I watched.”101 Continuity, the legitimacy of the state, must be reconciled with these ruptures. The primary means by which individuals do this is “just-­world thinking” – the assumption that those who are punished are guilty.102 In an experiment, Lerner and Simmons demonstrated the just-­world hypothesis when observers reacted to a woman receiving electric shocks, not with sympathy, but with the conviction that she must have done something wrong.103 Therefore, victims must have done something to deserve their suffering. Challenging this assumption means fundamentally challenging the fundamental legitimacy of the state, which, after all, functions as a kind of social glue, holding people together and protecting shared interests and individual entitlements. Challenging state legitimacy often carries high risks.

36   The emergence of the genocidal context During the establishment of this new moral context, perpetrators may perceive no distinct break with the previous social order, as the process unfolds incrementally until the new morality is normalized. On the micro level, breaks with the previous social order constitute individual moral “tipping points.” Through escalating commitments, individuals may find themselves behaving in unforeseen ways. The tension between new norms and old produces individual acquiescence as a means of moral resolution and coherence. Public violence acts as a propaganda of the deed, communicating norms about victims. Shortly after Habyarimana’s assassination Tutsi elites were dragged down the streets of a Rwandan town and killed with machetes and nail-­studded clubs.104 Such violence was terrifying for Tutsis and communicated a strong message; according to one of my interview subjects, a primary school teacher: “to not stop people from killing showed it was okay.”105 When individuals are attacked by state agents, it is tempting to assume that they must have behaved wrongly. Victims are therefore obviously outside the moral community. They are deviants. Yet the message disseminated by state persecution and violence extends beyond just-­world thinking. Propaganda emphasizes that victims have done something wrong as members of criminalized social groups. Their marginalization precedes the genocide. One Tutsi survivor of the Rwandan genocide recalls that “they used to throw stones on our roof, and at school [Hutu] children used to call us ‘cockroaches’ and ‘snakes’.”106 In Cambodia, however, the social groups targeted by mass killing did not exist prior to the rise of the Khmer Rouge. Groups such as so-­called “new people” were invented from scratch. As Tajfel demonstrated, the process of categorization itself seems to lead to bias.107 Yet this invention of social groups also attests to the protean nature of these groups – “new people” became a meaningless category after the Khmer Rouge fell in 1979. Escalation and restraint In conceptualizing the emergence of the genocidal context, we must account not only for processes of escalation, but also for processes of restraint. Like individual perpetration, the genocidal context is not inevitable, but rather a product of decisions taken by individuals in response to political and economic circumstances and interests. For example, Leo Kuper has argued that genocide did not occur in South Africa, despite racist ideology, because state elites understood that the economy depended on the cheap labor of black South Africans.108 Periods of escalation may be followed by periods of de-­escalation, and seemingly inevitable genocides may never occur. The failure to stop genocide may not be attributable to the inevitable momentum of genocidal processes, but rather to the weaknesses of inhibitory factors and institutions. These inhibitory factors include: limits on the power to effectuate genocide; countervailing views and sources of power; and intervention by third parties.109 The period before genocide often includes both phases of escalation and phases of retreat.110 We must consider, for example, the role of genocidal

The emergence of the genocidal context   37 massacres – episodes of mass killing that are highly localized and restrained. For example, the Hamidian and Cilician Massacres preceding the Armenian genocide did not result in a total campaign of extermination against the Armenians of Ottoman Turkey. Similarly, Rwanda experienced a series of massacres in the period of 1990–1994 (after the RPF invasion). While the Cilician massacres shattered the sense of security of many Armenians – they became more aware of their position of vulnerability within the Ottoman Empire – they were followed by a period of retreat.111 The massacres were more repressive than genocidal, and broader violence was also constrained by external factors. State elites will not commit genocide unless it is within their power, furthers their interests, and aligns with the “founding narrative” of the state.112 Limitations may arise from other internal factors (such as conflicts, alternative sources of power, and the state’s inability to exercise a monopoly on the use of force). Equally, there may be external limitations (controls) on the exercise of state power. This may have prevented the massacres in Rwanda in 1959 or in the Ottoman Empire in 1904–1905 from developing into broader killing campaigns.113 Killing is facilitated when perpetrators believe their actions will not be stopped or punished by any external force. When there is no reasonable prospect of internal or external control, genocide can be committed with impunity. External assistance to the victims may be unlikely simply because of a lack of political will. A retired Bosnian Army commander recalls that the UN forces in Bosnia had “very shiny uniforms, but they were just chicken … mostly they sold cigarettes.”114 Bosnian Serb military commander General Ratko Mladic used the unlikelihood of external intervention to terrorize the Muslims of Zepa, saying: “nothing, nobody, neither Allah, nor the United Nations can help you out … I am now your God.”115 The presence of affine populations in neighboring states may prove a disincentive to genocide by decreasing the vulnerability of the intended victims. Alternatively, if a state’s genocidal intentions are externally supported (as those of the Bosnian Serbs were by Serbia proper), the costs of genocide decrease.

Conclusion: social order and disorder The emergence of the genocidal context occurs as a contingent response to perceived strain. Victims are seen as an existential threat to the perpetrator group, and this legitimizes the concentration of power, the removal of the victims from the moral community, and the escalation of violence. The pre-­genocidal context is a state of disequilibrium requiring a corrective: genocide. Genocide is a process, rather than a set of static preconditions that produce particular behaviors. The disruption of old norms does not occur overnight, but incrementally. Nonetheless, genocide, unlike war, represents a significant rupture of the old social order as members of the victim group cease to be regarded as fellow citizens.116 Individuals in a genocidal context seek coherence, an explanation for the crisis the state faces. They may find this coherence partly in continuity and the preservation of the state. It may be easier for individuals to accept

38   The emergence of the genocidal context explanations that sustain their alignment with the in-­group and the state. Without such continuity and legitimacy, the state will struggle to command individual obedience. This continuity exists in most cases of genocide. In Rwanda, elements of the interim government seized control of state institutions within days after Habyarimana’s death.117 One could argue that the collapse of the Habyarimana state demonstrated the weakness of the Rwandan state. Yet many of the elites who controlled the country after April 7 1994 held powerful state positions before the genocide. The Khmer Rouge regime failed in this respect – Cambodia was a state without continuity. The declaration of Year Zero, the emptying out of cities, and the creation of new social categories marked a profound rupture. The state broke down rather than evolving. Social rupture and the restructuring of group relations facilitate the mobilization of violence by strengthening in-­group solidarity and stigmatizing the out-­group. The emergence of the genocidal context makes possible the discursive and structural foundations of individual perpetration. Individual perpetrators are often acting to advance their own interests, within an extraordinary setting in which they perceive their options as limited.

Notes    1    2    3    4    5    6    7    8    9   10   11   12   13   14   15   16   17   18   19   20   21   22   23   24   25   26

Damant, “Living in an Abnormal Normality,” 35. Shaw, What is Genocide? Damant, 35. Amery, At the Mind’s Limits. Sunstein, “Social Norms and Social Rules,” 9. Jessee, Negotiating Genocide, 169. Jessee, 169. Personal communication with Iva Vukusic, April 2017. Personal communication with Timothy Williams, April 2017. Hinton, Why Did They Kill?, 60. Hinton, 74 and 94. Strain theory has some commonalities with economic deprivation theory in criminology, which claims that poverty is criminogenic. On the latter, see, for example: Pratt and Cullen, “Assessing Macro Level Predictors,” 393. Wallace, “Anomie Theory,” 21–22. Agnew, Pressured into Crime. This collective strain is similar to Runciman’s concept of fraternal deprivation. See: Jacoby, Understanding Conflict and Violence, 107. Bauman, Modernity and the Holocaust, 14. Hinton, 244. Sunstein, 23. Midlarsky, The Killing Trap, 54. Semelin, Purify and Destroy, 35. Abrams and Hogg, “Collective Identity,” 446. Lanier and Henry, Essential Criminology, 222. Midlarsky, 138. Midlarsky, 139. Midlarsky, 139. Goldhagen, Hitler’s Willing Executioners, 41.

The emergence of the genocidal context   39 Pavkovic, The Fragmentation of Yugoslavia, 77. Pavkovic, 49. Midlarsky, 164. Des Forges, Leave None to Tell the Story, 159. African Rights, Rwanda: Death, Despair and Defiance, 20. Semelin, Purify and Destroy, 15. African Rights, 22. African Rights, 18–19. African Rights, 22. Midlarsky, 51. Straus, The Order of Genocide, 7. Daley, Gender and Genocide in Burundi, 95. Interview R29 (Rwanda), Rugerero TIG Camp, August 2009. Mann, The Dark Side of Democracy, 148. Interview R73 (Rwanda), Kigali Central Prison, October 2009. Interview R78 (Rwanda), Nyamata TIG Camp, October 2009. Interviews R28 (Rwanda), Rugerero TIG Camp, August 2009; R38, Rwaza TIG Camp, August 2009; and R64, Hindiro TIG Camp, October 2009.   44 Interview R25 (Rwanda), Tumba TIG Camp, August 2009.   45 Interview B06 (Bosnia), Srebrenica, April 2009.   46 Interview R29 (Rwanda), Rugerero TIG Camp, August 2009.   47 Kakar, The Colours of Violence, 39.   48 Staub, The Psychology of Good and Evil, 298.   49 Conger, et al. “Angry and aggressive behavior across three generations, 143–160.   50 Anderson, “Mainstreaming Atrocity.”   51 Interview R42 (Rwanda), Nyarasenge TIG Camp, August 2009.   52 Chalk and Jonassohn, The History and Sociology of Genocide, 340.   53 Hinton, 249.   54 Interview R10 (Rwanda), Gitarama Prison, July 2009.   55 Interview R37 (Rwanda), Rwaza TIG Camp, August 2009.   56 Interview R69 (Rwanda), Hindiro TIG Camp, October 2009.   57 Bergholz, “Sudden Nationhood,” 701.   58 Interview R30 (Rwanda), Butare, August 2009.   59 Interview R40 (Rwanda), Nyasenge TIG Camp, August 2009.   60 Interview R60 (Rwanda), Mont Kigali TIG Camp, October 2009.   61 Interview R58 (Rwanda), Rugerero TIG Camp, September 2009.   62 Jessee, 164.   63 Interview Bu5 (Burundi), Bujumbura, September 2009.   64 Interview Ba1 (Bangladesh), Jallad Khana, Dhaka, September 2008.   65 Hatzfeld, Machete Season, 47.   66 Waller, Becoming Evil, 35.   67 For example, in one variation of the Milgram experiment two experimenters disagreed in front of the subject and the subject showed substantially reduced compliance to the demands of the authority figures.   68 Beck, Prisoners of Hate, 72.   69 Hinton, 243.   70 Beck, 72.   71 Brannigan, Beyond the Banality of Evil, 83.   72 Straus, The Order of Genocide, 89.   73 Straus, The Order of Genocide, 25.   74 Guichaoua, From War to Genocide, 192–214.   75 See Sunstein, 14; and Bhavnani, “Ethnic Norms and Interethnic Violence,” 652.   76 Fromm, Escape from Freedom.   77 Brambilla, et al., 811.

  27   28   29   30   31   32   33   34   35   36   37   38   39   40   41   42   43

40   The emergence of the genocidal context   78 Interview R75, (Rwanda), Kigali Central Prison, October 2009.   79 For the relationship between uncertainty and group entitativity see: Castano, Leidner and Slawuta, 2.   80 Daley, 9.   81 Daley, 9.   82 Interview R16 (Rwanda), Gitarama Prison, July 2009.   83 Castano, et al., “I Belong, Therefore, I Exist,” 135.   84 Cekic, The Aggression against the Republic of Bosnia and Herzegovina, 250.   85 Horowitz, The Deadly Ethnic Riot, 75.   86 Semelin, Purify and Destroy, 31.   87 Mann, 6.   88 Horowitz, Taking Lives, 343.   89 Horowitz, Taking Lives, 354.   90 Semelin, Purify and Destroy, 31.   91 Weber, Politics as a Vocation.   92 Horowitz, Taking Lives, 365.   93 Horowitz, Taking Lives, 365.   94 Locard, Pol Pot’s Little Red Book, 77 and 206.   95 Locard, 204.   96 Lanier and Henry, 87.   97 Interview R50 (Rwanda), Nyamata, September 2009.   98 Semelin, Purify and Destroy, 88.   99 Weale, Army of Evil, 176. 100 Fein, Accounting for Genocide, 67. 101 Interview R10 (Rwanda), Gitarama Prison, July 2009. 102 Lerner and Simmons, “Observer’s reaction to the ‘innocent victim’,” 203–210. 103 Smith and Mackie, Social Psychology, 159. 104 Jessee, 101. 105 Interview R21 (Rwanda), Ruhengeri Prison, August 2009. 106 Interview R48 (Rwanda), Nyamata, September 2009. 107 Tajfel, “Experiments in intergroup discrimination,” 96–102. See also: Waller, 241. 108 Kuper, Genocide: Its Political Use in the Twentieth Century, 208. 109 Mayersen, On the Path to Genocide, 206–215. 110 Mayersen, 216–226. 111 Mayersen, 67. 112 Straus, “Retreating from the Brink.” 113 Mayersen, 208–213. 114 Interview B08 (Bosnia), Sarajevo, April 2009. 115 Semelin, “Analysis of a Mass Crime,” 366. 116 Semelin, Purify and Destroy, 98. 117 Guichaoua, 224–225.

Works cited Abrams, Dominic and Michael A. Hogg. “Collective Identity: Group Membership and Self-­Conception.” In Blackwell Handbook of Social Psychology: Group Processes, edited by Michael A. Hogg and R. Scott Tindale. Malden, USA: Blackwell Publishers, 2001. African Rights. Rwanda: Death, Despair and Defiance. London: African Rights, 1995. Agnew, Robert. Pressured into Crime: An Overview of General Strain Theory. Los Angeles: Roxbury Publishing, 2006. Amery, Jean. At the Mind’s Limits: Contemplation by a Survivor on Auschwitz and its Realities Bloomington: Indiana University Press, 1980.

The emergence of the genocidal context   41 Anderson, Kjell. “Mainstreaming Atrocity: Radicalization in Genocide and Terrorism.” In Rethinking Security in the Twenty-­First Century, edited by Dan Jacob. New York: Palgrave Macmillan, 2016. Bauman, Zygmunt. Modernity and the Holocaust. Ithaca, New York: Cornell University Press, 2000. Beck, Aaron T. Prisoners of Hate: The Cognitive Basis of Anger, Hostility, and Violence. New York: Harper, 2000. Bergholz, Max. “Sudden Nationhood: The Microdynamics of Intercommunal Relations in Bosnia-­Herzegovina after World War II.” American Historical Review 118, no. 3 (2013). Bhavnani, Ravi. “Ethnic Norms and Interethnic Violence: Accounting for Mass Participation in the Rwandan Genocide.” Journal of Peace Research 43, no. 6 (2006). Brambilla, Marco, Simon Sacchi, Stafano Pagliaro, and Naomi Ellemers. “Morality and intergroup relations: Threats to safety and group image predict the desire to interact with out-­group and in-­group members.” Journal of Experimental Social Psychology, 49, no. 5 (2013). Brannigan, Augustine. Beyond the Banality of Evil. London: Oxford, 2013. Castano, Emanuele, Bernhard Leidner, and Patrycja Slawuta. “Social identification processes, group dynamics and the behavior of combatants.” International Review of the Red Cross 90, no. 870 (2008). Castano, Emanuele, Vincent Yzerbyt, Maria-­Paolo Paladino, and Simona Sacchi. “I Belong, Therefore, I Exist: In-­group Identification, In-­group Entitativity, and In-­group Bias.” Psychology and Personality Bulletin 28, no. 2 (2002). Cekic, Smail. The Aggression against the Republic of Bosnia and Herzegovina. Sarajevo: Institute for the Research of Crimes against Humanity and International Law, 2005. Chalk, Frank and Kurt Jonassohn. The History and Sociology of Genocide. London: Yale University Press, 1990. Conger, R. D., T. Neppl, K. J. Kim, and L. Scaramella. “Angry and aggressive behavior across three generations: A prospective, longitudinal study of parents and children.” Journal of Abnormal Child Psychology 31, no. 2 (2003): 143–160. Daley, Patricia O. Gender and Genocide in Burundi: The Search for Spaces of Peace in the Great Lakes Region. Oxford: James Curry, 2008. Damant, Froukje. “Living in an Abnormal Normality: The Everyday Relations of Jews and Non-­Jews in the German–Dutch Border-­region, 1933–1938. In The Holocaust and European Societies: Social Processes and Social Dynamics, edited by Frank Bajohr and Andrea Low. London: Palgrave Macmillan, 2016. Des Forges, Allison. Leave None to Tell the Story: Genocide in Rwanda. New York: Human Rights Watch, 1999. Fein, Helen. Accounting for Genocide: National Responses and Jewish Victimisation during The Holocaust. New York: The Free Press, 1979. Fromm, Erich. Escape from Freedom. New York: Holt Paperbacks, 1994. Goldhagen, Daniel. Hitler’s Willing Executioners. New York: First Vintage Books, 2001. Guichaoua, Andre. From War to Genocide: Criminal Politics in Rwanda, 1990–1994. London: University of Wisconsin Press, 2015. Hatzfeld, Jean. Machete Season. New York: Picador Books, 2005. Hinton, Alexander. Why Did They Kill? Cambodia in the Shadow of Genocide. London: University of California Press, 2005. Horowitz, Donald. The Deadly Ethnic Riot. New Delhi: Oxford University Press, 2002. Horowitz, Irving Louis. Taking Lives: Genocide and State Power. New Brunswick, USA: Transaction Publishers, 2002.

42   The emergence of the genocidal context Jacoby, Tim. Understanding Conflict and Violence. New York: Routledge, 2008. Jessee, Erin. Negotiating Genocide in Rwanda: The Politics of History. London, Palgrave Macmillan, 2017. Kakar, Sudhir. The Colours of Violence. Chicago: University of Chicago Press, 1996. Kuper, Leo. Genocide: Its Political Use in the Twentieth Century. Middlesex, UK: Penguin Books, 1981. Lanier, Mark and Stuart Henry. Essential Criminology. Boulder, USA: Westview Press, 2004. Lerner, M. J. and C. H. Simmons, “Observer’s reaction to the ‘innocent victim’: Compassion or rejection?” Journal of Personality and Social Psychology 4, no. 2 (1966): 203–210. Locard, Henri. Pol Pot’s Little Red Book: The Sayings of Angkar. Chiang Mai: Silkworm Books, 2004. Mann, Michael. The Dark Side of Democracy. New York: Cambridge Books, 2005. Mayersen, Deborah. On the Path to Genocide: Armenia and Rwanda Reexamined. New York: Berghahn Books, 2014. Midlarsky, Manus. The Killing Trap: Genocide in the Twentieth Century. Cambridge, UK: Cambridge University Press, 2005. Pavkovic, Aleksandr. The Fragmentation of Yugoslavia: Nationalism and War in the Balkans. New York: St. Martin’s Press, 2000. Pratt, Travis C. and Francis T. Cullen. “Assessing Macro Level Predictors and Theories of Crime: A Meta-­Analysis.” Crime and Justice 32 (2005). Semelin, Jacques. “Analysis of a Mass Crime: Ethnic Cleansing in Yugoslavia, 1991–1999.” In The Spectre of Genocide: Mass Murder in Historical Perspective, edited by Robert Gellately and Ben Kiernan. New York: Cambridge University Press, 2003. Semelin, Jacques. Purify and Destroy: The Political Uses of Massacre and Genocide. New York: Columbia University Press, 2007. Shaw, Martin. What is Genocide? Cambridge: Polity Press, 2007. Smith, Elliot R. and Diane M. Mackie. Social Psychology. New York: Psychology Press, 2007. Staub, Ervin. The Psychology of Good and Evil: Why Children, Adults, and Groups Help and Harm Others. New York: Cambridge University Press, 2000. Straus, Scott. “Retreating from the Brink: Theorizing the Dynamics of Mass Violence and Restraint.” Perspectives on Politics 10, no. 2 (2012). Straus, Scott. The Order of Genocide: Race, Power, and War in Rwanda. Ithaca, USA: Cornell University Press, 2006. Sunstein, Cass R. “Social Norms and Social Rules” John M. Olin Law & Economics Working Paper 36. Chicago: University of Chicago, 1995. Tajfel, Henri “Experiments in intergroup discrimination.” Scientific American, 223, no. 5 (1970). Wallace, Richard A. “Anomie Theory.” In Encyclopedia of Criminology and Deviant Behavior, edited by Clifton D. Bryant. Philadelphia: Brunner-­Routledge, 2001. Waller, James. Becoming Evil. New York: Oxford University Press, 2002. Weale, Adrian. Army of Evil: A History of the SS. New York: New American Library, 2012. Weber, Max. Politics as a Vocation. Philadelphia: Fortress Press, 1965.

3 The genocidal context

Introduction: harm-­producing relations of power The genocidal state is a crime-­producing system. Within this criminogenic state, the crime of genocide is made possible by systematized social distance, extreme violence, and the externalization of moral responsibility. The perpetrator and the victim belong to different moral universes. This divergence, the humanity gap, is institutionalized in state structures that morally disengage perpetrators from victims. The state produces genocide, through discursive and organizational structures that marginalize the victims and place them outside the moral community. Once genocidal destruction begins, these harm-­producing relations of power are manifested in extremis. Despite the power of the state in producing a moral context and institutionalizing norms, it is crucial to remember that individual perpetrators do not uncritically receive norms from above. Rather, the mobilization of genocide, and the production of genocidal ideology, are often interactive processes in which low-­ level actors play a role in the implementation and direction of policy. The dualistic nature of the genocidal state (as a force for rupture, but also a source of continuity) is crucial to the adaptation of perpetrators. By virtue of its monopoly on the use of force and its ability to levy tax and extract resources, the state has unrivalled power and legitimacy. Moreover, the state is predicated on a presumed legitimacy; we accept its existence as the natural order of things. Those who do reject or resist the state often find themselves defined as deviants, subject to the coercive exercise of state power. The international community is still largely a community of states rather than citizens. This powerful assumption of legitimacy greatly magnifies the power of the state and the impetus for individuals to align themselves with its norms and demands. As one perpetrator in Rwanda put it: “Yes, you obey your leaders. Otherwise you must have your own country, if you don’t follow your leaders.”1 Genocide is about the institutionalization of public and private harm against groups defined as deviant. This represents a shift in the moral context, which is also reflected in moral habits or practices. As Wikström and Treiber write: “An actor will habitually choose to harm others if he/she sees doing so as the only viable alternative for action.”2 Thus the production of harm fundamentally

44   The genocidal context structures perpetrator choices through the institutionalization of revised moral norms. Beyond the revision of moral norms, the state also externalizes responsibility through the institutionalization of violence. To exercise these functions and powers, the state enacts and enforces laws, while also creating organizations to implement policies. These serve as points of intersection between the state and its citizenry. In a sense, then, as harm is institutionalized, there is also a shift in values to reflect the new reality, while still not discarding conventional values altogether. To oppose the state could mean abandoning many of the social goods that the state is said to protect. Constitutive criminology defines crime as “the harm resulting from humans investing energy in harm-­producing relations of power” and suggests that crimes “are no less than people being disrespected.”3 Genocide is arguably the ultimate harm-­producing power relation, in which the victims suffer “social death” and are subject to destruction, while the perpetrators enjoy impunity.4 Rather than crime being a complete break from conventional values, perpetrators may perpetrate genocide partly out of deference and fidelity to conventional institutions. Thus, even perpetrators with a relatively low propensity to engage in criminal behavior may perpetrate genocide in response to the altered moral context. For many perpetrators, moments of perpetration may be a choice between a definite wrong (contravening the state) and a possible harm (killing persons whose death is demanded by trusted institutions and peers). Genocide is rooted in these harm-­producing relations of power, established both institutionally and discursively. The institutionalization of harm is enabled by the creation of an extreme power differential between perpetrators and victims, as well as a state of moral disengagement (in which killing can occur without consequence). On the micro level, perpetration is made possible by the distancing and disengagement of perpetrator from victim, as well as the individual’s desire to seek coherence and conformity. Genocide requires the application of extreme asymmetrical force in service of an ideological goal (the complete destruction of the dehumanized other). As a mass crime genocide involves a high level of coordination and systematic design. Therefore, genocide is not possible without institutional relations of harm and an enabling moral context. This chapter considers the institutionalization of harm in the genocidal state, in terms of norms and institutions of social control. It also examines the role of specific institutions in producing criminal behavior. The institutionalization of norms in the genocidal state alters the perpetrator’s perceived range of options, encouraging individual perpetration of the crime of genocide.

The genocidal state as a state of moral disengagement The state creates a power differential between victims and perpetrators through the institutionalization of harm. This power differential also involves moral distancing through the aggrandizement of the perpetrator and the devaluation of the victims. Through these processes, the social distance between the perpetrator and

The genocidal context   45 victim increases, opening a space for harm to the victim group. Victims must be vulnerable and excluded from the moral community. Moral distancing requires the moral disengagement of the perpetrators. As Albert Bandura demonstrates, acts of harm are facilitated greatly by processes in which the perpetrator does not feel morally responsible towards his/her victims.5 Moral disengagement is the absence of empathy; empathy is negatively related to both aggression and anti-­social behavior more generally.6 The genocidal context is also a system of moral disengagement, in which the perpetration of harms against certain social groups is accepted or even expected. Of course, moral disengagement occurs in many other contexts, from petty crime to crimes against humanity. What sets genocidal moral disengagement apart is, first, that the violence is endorsed by the state; and second, that the violence is total – directed at the destruction of ideologically-­stigmatized social groups. The state’s endorsement of acts of extreme harm, which it would normally prohibit and counter, presents a uniquely dangerous situation. Through totalizing ideology and policy, individual members of the victim group are denied escape. Through discourses of justification and structures of power, perpetrators are given an opportunity to kill without consequence, or as Gunther Anders wrote about the Holocaust, “the chance for unpunished inhumanity.”7 Only victims who have been devalued by the genocidal state may be killed without the fear of violating conventional norms. In Soldaten, a landmark study of German “ordinary” perpetrators, Neitzel and Welzer observe: It was the coupling of murder and morality – the realization that unpleasant acts were necessary and the will to carry out these acts in defiance of feelings of human sympathy – that allowed the perpetrators to see themselves as “respectable” people, as people whose hearts, in Hoss’s mind, “were not bad.”8 The maintenance of a positive self-­image is essential for mass perpetration, even among perpetrators who may not share the goals of the regime or the desire to kill. The genocidal state opens up an opportunity space for acts of grave inhumanity.9 This opportunity space is rooted in harm-­producing relations of power. Power relations are produced and reinforced discursively through propaganda and other narratives that devalue and stigmatize the victim group. These power relations are then institutionalized in the genocidal state. The institutionalization of power relations includes the development of coercive institutions, increasing levels of violence against the victim group, and the marginalization of victims. Moral identity is not clearly threatened in genocidal states, because the exercise of state power (and the prerogative to obey it) is itself seen as moral (aligned with conventional values). Indeed, fidelity to the state may seem a reasonable solution to the problem of strain – authoritarian states often make this argument explicitly. Conventional values, in the form of adherence to state power, may be seen as comforting, even while alignment with state power increasingly involves

46   The genocidal context violence; this violence, it must be remembered, is against a threatening “other.” Through its treatment of these groups, often involving acts of public persecution or violence, the state makes clear that these groups are to be excluded. Moments of moral rupture may conflict with conventional values, creating confusion in some citizens. However, even as the state produces confusion and uncertainty, it also provides an explanation for these acts through the re-­ categorization and re-­characterization of victim groups. It may be easier to assume that the use of violence against stigmatized groups is justified than to discard conventional social structures. Similarly, the state may make abundantly clear, through its focus on the identification of traitors to the community, that abandoning the state means placing oneself outside the community and exposing oneself to ostracism or worse.

Genocidal contexts compared Genocide usually occurs in situations of strain; extreme violence is called for to address this disequilibrium. Consequently, violence escalates in tandem with deteriorating social relations. However, even with these common conditions, the form that genocide takes varies. In particular, one can distinguish between specialized and participatory genocides. In the former, individual participation in killing is often relatively limited; those who participate are mainly specialists (bureaucrats and violence professionals); and violence is largely hidden from public view. This does not mean the public lacks awareness of the persecution. Rather, the killing generally occurs out of the public eye. In specialized genocide, norms authorizing killing are selectively disseminated, although the victim group is still publicly stigmatized as an existential threat to the survival of the in-­group. The Holocaust, the genocide of the Herero, and the Bosnian genocide largely fit this model (despite variations in the degree of civilian involvement). In the Nazi case, this occurred, for instance through extermination campaigns directed against Jews in eastern Europe. In participatory genocides, on the other hand, mass participation in killing is encouraged, many civilians are involved in killing, and killing occurs publicly. Accordingly, norms authorizing genocidal killing are widely disseminated to the public. Rwanda in 1994, Burundi in 1988, and Indonesia in 1965 largely fit this model, although all three cases also employed violence specialists. Moreover, some cases of genocide do not neatly fit into either model (e.g., the Armenian Genocide and the Cambodian Genocide). Despite variations, these cases illustrate broad patterns of participation. Estimates place the number of Germans and Austrians who were “direct agents” (killers) in the Nazi Genocide at around 203,500 (200,000 men and at least 3,500 women); this includes those involved in shootings, ghetto liquidations, and concentration camp gassings.10 These perpetrators were drawn from a total population (as of the last census before the war, in May 1939) of 57,004,268 adults.11 Therefore, approximately 0.36 percent of the adult population in the Greater German Reich participated directly in killing, or

The genocidal context   47 approximately 0.73 percent of adult German and Austrian males. In contrast, Scott Straus calculates between 175,000 and 210,000 perpetrators, amounting to 7–8 percent of the adult Hutu male population.12 Thus, the participation rate pf adult males in the Rwandan Genocide was approximately ten times that of the Holocaust. These are crude estimates, and the Rwandan figures problematically do not include women. Yet they still point to a significant difference in the model of killing utilized. This specialized/participatory dichotomy has important implications for gauging the size of the normative community. In both Germany and Rwanda, hate propaganda was prevalent. However, in Rwanda, the norm of killing Tutsis was communicated much more broadly. The norm and demand to kill is conveyed through direct incitement (orders and exhortations), propaganda, and practices attesting to the presence of the norms. Organizational and normative structures limit divergence from genocidal norms. In practice, it is not easy to separate these categories, as government organizations are imbued with normative authority. In specialized mass killing, the perceived range of individual options (the role margin of discretion: see p. 131) assumes greater importance, as participation in killing is limited largely to specialists. Conversely, in more participatory genocides, the social margin of discretion (perceived society-­wide options) is more evident, as demands for killing extend across the ethnic group. In both cases, stigmatization through propaganda facilitates the general acquiescence to violence and coercion towards the victim group.

Norms facilitating violence Obedience The state’s structuring of individual choice in genocide is greatly facilitated by norms of social control. These exist in numerous contexts, yet genocide, with its centralization of power in response to a perceived threat, heightens their influence. Norms must also be accompanied by sanctions that regulate social relations by punishing disobedience. Over time, norms become moral habits, internalized through repetitive action. Obedience is one such moral habit. For example, Kurt Mobius, a guard in Chelmo camp (in Nazi Germany) writes: Haupsturmfuhrer Lange said to us that the orders to exterminate the Jews had been issued by Himmler and Hitler. We had been drilled in such a way that we viewed all orders issued by the head of state as lawful and correct. We police went by the phrase, “whatever serves the state is right, whatever harms the state is wrong.” I would also like to say that it never even entered my head that these orders could be wrong.13 The state of moral disengagement that characterizes genocide is also a state of habitual moral disengagement produced by norms of obedience.

48   The genocidal context Obedience to authority is a fairly universal human tendency and one that is ingrained in us from a very early age. As children, we must obey our parents. There are two aspects to such early demands for obedience: the need to obey the specific rule invoked, and the general principle that rules are to be obeyed.14 Schools are systems of conformity where those who obey rules are rewarded and those who disobey are punished. We spend most of our first 20 years as a “subordinate element in an authority system.”15 As Stanley Milgram notes, job promotions have the effect of simultaneously rewarding obedience and further entrenching the hierarchical system.16 Obedience is often implicit, and authority is communicated through symbols of authority such as flags and uniforms. Obedience to authority was most famously demonstrated in the experiments of Stanley Milgram, in which research subjects displayed an apparent willingness to inflict physical harm on others at the behest of an authority figure.17 In the experiments, obedience substantially decreased when the authority figure left the room (it also increased as the victim was placed further away from the teacher – a confirmation of the importance of distance in the denial of responsibility).18 One would expect that the effects of obedience would be even stronger in genocide, given the presence of state power. Moreover, in genocide, normativity is often strongly communicated through propaganda and bolstered by positive reinforcement and conformity. The Milgram experiments also demonstrate, however, that compliance may be limited in the absence of institutional power or urgency. Disobedience may be punished through the direct exercise of coercive power (i.e., violence at the hands of state agents, including the deprivation of liberty), but many perpetrators will also habitually demonstrate obedience to maintain their ties to the group and its conventional values. Ultimately, it may be easier to neutralize previously held moral norms than to isolate oneself from the group. For many perpetrators, obedience may not be grounded in the fear of coercive power, but rather in other forms of pressure and influence arising from norms of masculinity and the duty to comrades. An Einsatzgruppe perpetrator argued: Himmler issued an order stating that any man who no longer felt able to take the psychological stresses should report to his superior officer.… In my view this whole order was an evil trick; I do not think I would be wrong to say it bordered on the malicious – for after all, which officer or SS man would have shown himself up in such a way? Any officer who had declared that he was too weak to do such things would have been considered unfit to be an officer.19 Obedience may also take place in an indirect way, in which perpetrators do not directly obey authority, but perform the actions that they think authority figures desire. Perpetrators may act in the spirit of orders rather than in strict accordance with them. This is what Holocaust historian Ian Kershaw dubbed “working towards the Fuhrer.”20 For example, a former LRA soldier testified in the Ongwen Trial at the International Criminal Court: “You know, when you’re

The genocidal context   49 assigned to a particular job there are times when you may go over and above what you’re supposed to do as an individual, so you may do what you haven’t been authorized to do….”21 We must consider, as we will below, that cultures of obedience can be created; the state functions as an organizational culture in which norms governing the treatment of others are established, even in the absence of direct commands. Obedience to authority results in a recalibration of responsibility, in which the emphasis is placed on the superior’s right to command, rather than the actor’s consent.22 The (low-­level) perpetrator transforms himself from a subject into an object of the will of others. The nature of organizations may reinforce this tendency, as every actor believes he or she is acting under external influence. This leads to a responsibility vacuum in which moral authority is incapacitated without being openly challenged.23 As will be argued in Chapter 5, close social bonding with the state increases compliance with the state’s genocidal demands, which may include perpetration, or acquiescence (remaining a bystander). Conformity Conformity, the convergence of individuals with a social norm, strongly affects individual behavior.24 Conformity is rooted in our need for connectedness and the construction of a shared social reality, expressed in ideology and moral norms. The group exerts a pervasive influence on individuals, providing cues for the moral context. Humans, as social animals, react to the behavior of others, and often mirror it to gain social approval. People seek to belong to groups in order to know who they are and what they should do. As one Rwandan genocide perpetrator put it, “we liked being in our gang.”25 To sustain the benefits of group membership (i.e., feelings of strength, anonymity, and transcendence), individuals must retain their connection to the group. Thus, individuals will go to great lengths to protect the group from physical and rhetorical attacks. Cognitive psychologist Aaron Beck labels such behavior “groupism.”26 This collective counterpart of egoism leads to events being interpreted in terms of the group’s interests and beliefs. It also subordinates personal interests to the interests of the group, and opposes the interests of the out-­group. The same psychological mechanism that produces joy at personal triumphs also operates when the group triumphs.27 Notions of duty to comrades may also motivate perpetrators to participate in genocide. The need for acceptance from one’s in-­group can be a powerful motivator. Psychologist Bruce Hood writes: Excluded individuals engage in behaviors that increase their likelihood of being reconciled back into the group. We are more likely to mimic, comply with requests, obey orders, and cooperate with others who don’t deserve it. We become obsequious to the extent that we will agree with others who are obviously in the wrong.28

50   The genocidal context Experiments show that people will intentionally give the wrong answer to a question to more closely align themselves with the expressed preferences of other group members.29 In genocide, conflicts between group norms and individual norms may be reconciled through moral neutralization – discourses stipulating that the victims are not innocent and that the same moral rules do not apply to them. In such contexts, it may ultimately be easier for perpetrators to neutralize previously held moral norms than to isolate oneself from the group. For most perpetrators, then, a desire to conform lies at the core of the decision to perpetrate. Urgency A sense of urgency greatly facilitates social control, as well as intensifying demands for individual alignment with collective goals. As noted in the previous chapter, genocide is normally accompanied by a survival discourse that presents the state and community as facing an immediate threat from internal and external enemies. This state of fear demands the centralization of power and the setting aside of normal rules; it is, to draw from Giorgio Agamben, a “state of exception.”30 Such states of exception often coincide with the declaration of states of emergency or martial law, enabling the state to use force in previously unlawful ways to counteract a grave danger to the polity. Emergency powers often involve setting aside human rights, a possibility that is foreseen, to some extent, in human rights conventions, with their list of derogable rights (such as freedom of speech). However, the state derogates even non-­derogable rights in genocide, partly by defining civilians as enemy combatants, or criminals. For example, the Enabling Act of March 23 1933 gave the German Nazi government the power to enact legislation for four years, even if it diverged from the constitution. A few months later, Hitler declared that “the party has now become the state,” and in December a law was passed declaring the unity of the Nazi Party and German state. In 1937, the Nazi minister of Justice Franz Gurtner argued that: National Socialism substitutes for the conception of formal wrong the idea of factual wrong: it considers every attack against the welfare of the people’s community, every violation of the requirements of the life of the nation as wrong.… The judge interprets the existing regulations not literally but according to their spirit and basic thoughts.31 Genocidal contexts are always states of exception – the exercise of state power is far more expansive than in normal situations. The victim group cannot be targeted without first defining it as a threat to the political and social community. The state of exception demands collective endeavor; individuals who do not align themselves with state power can thus also be seen as enemies of the political community.

The genocidal context   51 To concentrate power in its hands, the state often represses or coopts civil society and media institutions. This concentration of power may not be directed at the realization of genocide, but rather constitutes an aspect of broader authoritarian governance. Where all power flows through the state, there is often a lack of alternative norms and power bases to challenge the norms inculcated by the state, even when they include the persecution or murder of groups. This was true in Nazi Germany, in Cambodia under the Khmer Rouge, and even, arguably, within the state of polarized multi-­party democracy in Rwanda during the early 1990s.

Structures facilitating violence Genocide is a complicated endeavor involving the coordination of numerous individual acts of perpetration. Even cases such as Rwanda, which unfolded rapidly, were implemented largely through existing state structures. The legal system In making laws, states also set out norms of behavior; in genocide, this pertains especially to the treatment of certain categories of persons. Law, in this context, is less about shared goods, and more about demarcations of difference. Those who are placed outside the political community receive little or no protection from the state. Indeed, in stripping legal entitlements from persons belonging to marginalized groups, the state simultaneously communicates norms of valuation and weakens the security of the group. The group is placed in a situation of vulnerability where it occupies geographical space through the forbearance of the political community to which it no longer belongs. The legal marginalization of groups often includes denationalization – the restriction of citizenship rules to exclude the victim group.32 This is an explicit statement that the group no longer belongs to the community and is no longer extended state protections. When citizenship is withdrawn, or withheld, the group may be excluded from many other legal entitlements – including employment, land purchase, and the right to continue to reside within the state. This is the case today in Burma, where the Muslim Rohingya minority are not recognized as full citizens, but labelled as “Bengalis” (indicating they are from Bangladesh). This is the state exercising its power, through the law, to legitimize certain groups while delegitimizing others. The law may also directly exclude groups from participation in political life (through barring them from public office, for example), as well as restrict their economic opportunities (by prohibiting land ownership, exacting special taxes, preferential employment quotas, and so on). Land-­ownership restrictions are especially damaging, as with land comes not only the ability to make a living, but also a normative statement of belonging to a particular geographical space. In extending the dissolution of social ties between groups to the private sphere, genocidal states often restrict fraternization between the victim group

52   The genocidal context and the wider population (e.g., through anti-­racial miscegenation laws, such as the Nuremberg laws in Nazi Germany). This has the effect of further isolating the victim group and contributing to the narrowing of social relationships. Population control measures may extend beyond anti-­miscegenation laws to include other measures to prevent births among the targeted group. This may include the forcible transfer of children, for example through slavery or residential schools, as well as laws preventing births through forcible sterilization or abortion. Such laws may be directed at the destruction of the victim group. They ultimately arise from demographic neuroses – if the victim group is able to outnumber “us,” they may seize control of the state. Beyond changing the law to reinforce the persecution of the victim group, the non-­application of laws may lead to what Becker calls the “conventionalization of crime.”33 Through non-­enforcement, certain criminal acts cease to be deviant, at least when directed against designated out-­groups. This effect may be particularly strong when the state itself is committing acts of violence. Through such mechanisms, the law provides an enabling framework for harms. The criminalization of social groups occurs both through the law and through the de facto removal of legal protections from members of the victim group. Beyond its normative functions, law may also be utilized to provide cover for existing acts.34 Legal scholar Jennifer Balint argues that the law is never directly used to authorize murder, but rather contributes to the legitimacy of genocidal practices.35 The decision to commit genocide is “generally not public, nor officially legal.”36 This does not mean, of course, that the destruction of the victim group is never in the public sphere, rather that it often occurs euphemistically. There is often a “dual state” in which traditional judicial bodies continue to function, providing the pretense of “normative consistency, legal hierarchy, and internal coherence,” alongside the “prerogative state,” which acts outside the law.37 In Rwanda, the dual state was embodied by the interim government which tried and failed to control extremist factions, thus creating an environment where “anything that was not explicitly forbidden and sanctioned was permissible.”38 The continuation of the judicial system, alongside other bureaucratic structures, facilitates the moral ruptures of genocide by providing a sense of moral continuity and the legitimacy of state power. Organizing genocide Genocidal campaigns require the coordination of large numbers of perpetrators, and such coordination involves hierarchical organizations (as “machines for controlling human behavior”) and the division of tasks.39 Many of these organizational structures may seem most directly relevant to state actors, such as army officers, police, politicians, and bureaucrats. However, the existence of bureaucratic and coercive institutions has a considerable effect on civilian perpetrators and bystanders, who may not undergo extensive training but still feel that their options are constrained by the existence of a powerful, threatening, and far-­reaching state.

The genocidal context   53 Genocidal states may institutionalize moral norms in several ways. These include state propaganda, which serves the primary purpose of communicating moral norms and expectations; “the propaganda of the deed,” in which the actions (or non-­actions) of state actors convey a normative message to the audience; and finally, the regulation of social relations and group entitlements. Hierarchy Sociologist Zygmunt Bauman argued that within bureaucracies there is a “delegitimation of all but inner-­organizational rules as the source and guarantee of propriety, and thus denial of the authority of private conscience, become the highest moral virtue.”40 A bureaucratic abstraction and distancing from outside moral norms ensues, wherein the moral norms of the state are institutionally and publicly relevant whereas private belief systems are not. Bauman described such structures of authority as “system power”: System power involves authorization or institutionalized permission to behave in prescribed ways or to forbid and punish actions that are contrary to them. It provides the “higher authority” that gives validation to playing new roles, following new rules, and taking actions that would ordinarily be constrained by pre-­existing laws, norms, morals, and ethics.41 In general, cognitive dissonance, the discomfort that arises when actions do not accord with previously held beliefs, is mitigated when responsibility for actions is attributed to someone else.42 This principle was employed in extremis in Argentina where superior officers in the junta actually signed release forms for each kidnapping absolving the direct perpetrators of responsibility.43 Similarly, in Nazi Germany a message from Hitler to the troops on the Eastern Front promised “no German participating in action against bands or their associates is to be held responsible for acts of violence either from a disciplinary or judicial view.”44 Within the Khmer Rouge, the externalization of responsibility through hierarchy was extreme – with only those ranked above you considered to be part of the regime (Angkar). Research shows that dissonance effects are strongest when individual actions have serious consequences and/or people feel personally responsible for their actions.45 Genocide, like warfare, is a form of organized violence, and for such violence to be controlled, it must be directed by a hierarchical structure. This hierarchy need not be formalized, but even if it is only de facto, perpetrators must understand their place and responsibilities in it. Hierarchies stratify power, with victim groups placed on the bottom, and thus vulnerable to power exercised by their superordinates. Hierarchical organization removes moral agency from perpetrators, as responsibility is externalized and subject to the requirement to obey those above.

54   The genocidal context Compartmentalization Compartmentalization is another structural factor that facilitates moral disengagement. It involves a functional division of labor in which perpetrators each perform a specialized task contributing to the overall genocidal campaign. These ‘fragments of perpetration’ allow the perpetrator to feel that their own contribution is insignificant, and this, in turn, reduces feelings of cognitive dissonance and moral culpability. A Rwandan perpetrator, who was part of a group that killed three people, recounts: “I did not feel pity because the killers had guns or machetes and they came and took the victims, so that the victims could be killed. I had nothing to do except to watch how they were killed.”46 States, like large corporations, are “structurally and procedurally complex,” with limited scope for personal responsibility.47 John Lachs described the “mediation of action,” in which other actors are interposed between the individual and their action, acting as mediators.48 This increases the distance between principals and the consequences of their acts. As Raul Hilberg wrote of the Nazi bureaucrats, “they could destroy a whole people by sitting at their desk.”49 Bureaucrats are technicians, executing hierarchically-­defined tasks within their range of competence. Albert Speer once said that he “exploited the phenomenon of the technician’s often blind devotion to his task … without any scruples.”50 Ironically, when Speer was asked about the use of slave labor by the Nazi regime, he responded: “I don’t know anything about it and it was not a concern of mine.” Asked whose concern it was, he replied: “It would have been the concern of various agencies.”51 Similarly, Nuon Chea (“Brother Number Two” in the Khmer Rouge) maintained: If they ask me in court who killed the people, I will say I was in charge of the legislative body and education, so the killing was the problem of government administration, which was the responsibility of Pol Pot and Son Sen.52 Bureaucracies that produce harm on such a massive scale often allow perpetrators to maintain partial ignorance of atrocities. Killing in organizations In the genocidal state, as in all states, the state exercises a monopoly over the use of force. The use of sanctioned violence in states in normally confined to state agents such as the army and police. Participation in genocide may take the form of: professional participation (those experienced in violence, acting within organizations dedicated to violence), or amateur (civilian) participation (often within a context of widespread participation in killing). Even in genocides that make extensive use of specialists, mass support is often cultivated. Most genocides contain a mix of both types of perpetrators: violence specialists may take a leading role, but mass participation is still useful. For example, in the Holocaust, killing occurred at the hands of violence specialists, such as Einsatzgruppen and

The genocidal context   55 SS formations (in mass executions and gassings). But many amateur perpetrators also participated in killing actions such as the mass executions on the Eastern Front.53 In Rwanda, the army and extremist paramilitaries played a central role, but the scale of the killing could not have been attained without the mass participation of many amateurs. In genocide, the state’s monopoly of violence often breaks down, so that quasi-­state actors (e.g., paramilitaries) and civilians are extensively involved in violence. This serves several purposes, including force multiplication, the diffusion of culpability, and deniability. The military and other state agents may actually force civilians to perpetrate; according to one Rwandan perpetrator: “They sent someone from the military who was in charge of influencing the others to participate.… We were afraid. If we did not participate we could be hunted.”54 Violence specialists Army and police Repressive control in authoritarian states requires a large police and security apparatus to maintain compulsory obedience.55 Genocide is often led by state agents within these institutions. Such individuals already have expertise and socialization in the use of violence. Professional violence organizations, state-­ sanctioned organizations that use violence to maintain order and security, often involve rigid hierarchy and extensive training and ideological indoctrination. They also often make use of ritualized initiation ceremonies and oaths, distinctive clothing (uniforms), the memorialization of the dead as martyrs, and an organizational culture stressing the primacy of the group. 56 These special characteristics, along with state-­granted legitimacy, increase behavioral conformity by reducing goal variance. For example, a convicted perpetrator in the former Yugoslavia notes: I was a member of Yugoslav People’s Army. It was a legal institution in ex-­ Yugoslavia.… And I was a part of the state system. I was forced by that system to go there to defend the state attributes that system imposed to me itself. […] I never doubted in the legality of that act.57 Militaries control and limit violence in same ways that societies do: through both formal and informal controls: these can include, for example, military law, formalized hierarchies, the norm of obedience, the awarding of medals, and norms of honor.58 A Yugoslav army memo noted that paramilitaries undermined “military morale,” because their “primary motive was not fighting against the enemy but robbery of private property and inhuman treatment of Croatian civilians.”59 Violence must only be used in prescribed ways against designated targets. The perpetration of genocide does not require great technical skill. This is demonstrated by the large number of amateur perpetrators in cases like Rwanda. Nonetheless, many of those perpetrators who kill the most are soldiers or

56   The genocidal context paramilitaries who have undergone extensive training. Such training aims to inculcate obedience to authority and a conditioning to kill. According to one Rwandan perpetrator: “We were taught how to use the guns, how to react during the training so that we could be able to use those weapons. The soldiers were obliged to teach us how to use a gun, how to shoot.”60 Training may involve the brutalization of the killers themselves through techniques such as humiliation, harsh discipline, sleep deprivation, and physical exhaustion. At the same time, perpetrators are made aware of their important role in history – a role that can only be actualized through brutality toward the dehumanized other. This brutality becomes the main mechanism for the reinforcement or restoration of self-­worth. Socialization to killing may involve the gradual exposure to violence and the development of conditioned responses. For example, the Interahamwe trained on human-­shaped dummies, while the training of Serb soldiers took an even more direct form: some recruits were taken to farms to wrestle pigs to the ground, pull back their heads, and slit their throats.61 This mode of killing – cutting the throat – was frequently used by Serb forces in the Bosnian war. Military training seeks to overcome resistance to such acts, and includes not just Pavlovian and operant psychological conditioning, but also the erosion of moral identity through depersonalization and isolation from the outside world.62 Soldiers are effective perpetrators not only because of their training, but because of their norms of obedience and loyalty to comrades. Research conducted among American soldiers and Marines in Iraq shows that group cohesiveness increases in combat; moreover, 41 percent of soldiers and 44 percent of Marines support torture to save the life of a soldier or marine.63 Marginalization may have lethal effects on the battlefield. The police are not generally as disciplined as the military, and not as focused on killing. But they still are still state agents in an organizational culture, with a norm of hierarchical obedience. They are generally the institution with primary responsibility for the enforcement of laws: state agents who are authorized to use force against the citizenry. As such, the police are nearly always involved in the commission of genocide. This involvement can take various forms, such as the enforcement of laws producing harm, the mobilization of civilians, and the round-­up and killing of victims.64 The police, as symbols of state authority, also communicate important messages about the acceptability of violence, or inaction in the face of violence. They are, of course, also armed and often well-­trained in the use of violence. The police and military alike are often involved in the training and supply of civilian perpetrators. Paramilitaries Regimes’ use of non-­state or quasi-­state actors (such as paramilitary groups) often indicates a willingness to engage in illegitimate forms of violence, such as genocidal killing. Many paramilitary groups, such as the Sans Échec in Burundi, behave much like gangs, complete with heavy juvenile participation. Such

The genocidal context   57 groups are often composed of marginalized individuals who are given a chance, through participation in the group, to increase their status. In some cases, militias, like gangs, exist in a power vacuum, and individuals “will gravitate to joining the only real social organization present in their social environment.”65 Young males are overrepresented in such groups, as they are in violent crime in general.66 Paramilitaries may advance their own self-­interests while simultaneously contributing to state political goals.67 They may also allow their members to access an alternate social system where they have greater power and agency. For the genocidal regime, paramilitaries act as force multipliers: they can be quickly mobilized (due to their limited training) and free up regular forces for conventional battle.68 Criminological research on gangs shows that individuals with few opportunities are the most likely to join gangs as a means of gaining opportunities and security. As noted, many militia groups that participate in atrocities operate much like gangs, with initiation rituals and the promise of increased status and personal enrichment. The genocidal state may also provide opportunities for marginalized individuals through the de facto recognition of paramilitary groups, or through the formal state apparatus itself. Violent paramilitary groups, such as the White Eagles (Bosnia) or Interahamwe (Rwanda), may be legitimized. The state may also “reward” members of such groups by allowing them to commit self-­enriching criminal acts (e.g., pillage, plunder, and rape). Violence organizations, as well as other less formalized groups of perpetrators, may become “transgressive communities.” These are groups which, through violence, have become alienated from conventional norms. Acts of transgression may create a sense of unity, with members of the group forming interrelationships and interdependencies of validation. Fear and shame increase the cohesion of the group. Furthermore, individuals may feel pressured to use violence to maintain their status within the group. 69 Indeed, the performance of violence is fundamental to group membership and identity. Acts of initiation and coerced perpetration may forcefully implicate people in violence and integrate them in the transgressive community. This also features in other forms of mass atrocity; for example, one interviewee, who was a former member of the Lord’s Resistance Army (LRA), recounted that upon being abducted he was forced to kill one of his friends with an axe. The body was then strung up in a tree and drained of blood, which the conscripts were forced to drink.70 Such acts serve to alienate perpetrators from society at large, and from their prior moral beliefs. Societal rejection may increase commitment to deviant behavior.71 Moreover, the perpetration of transgressive communities may reduce the need for other actors to participate in genocide. Research into criminal gangs is informative for understanding the dynamics of paramilitary groups. There are three major criminological perspectives on the relationship between gangs and individual acts of violence. The first perspective, selection, argues that people who are already violent join gangs; the second, facilitative, posits that socialization in gangs makes ordinary individuals violent; the third, enhancement, argues that individuals who already have some

58   The genocidal context propensity for crime become more violent once they are in gangs.72 Transgressive communities in genocide most commonly reflect the third tendency. Individuals who are already somewhat marginal choose to join these groups. Yet we cannot discount the possibility that even individuals without a high propensity for violence may find themselves in situations where they are members of a criminal organization that was not initially criminal. Once within the gang, violence is often escalated through exposure to it, as well as through group cohesiveness, interaction with “highly delinquent peers,” and organization.73 Gangs, like paramilitary groups, may become the primary group, more important even than family in terms of individuals’ self-­identity and support.74 They may also offer an opportunity for individuals to “express masculinity,” particularly when men feel disempowered or emasculated.75 Group approval becomes the supreme factor, and violence an expression of loyalty and commitment. Where genocide is limited to specialists, extreme normative change authorizing killing need also only be focused on these narrower norm communities. Although fewer personnel are involved, their level of socialization to killing (and their obedience) is generally deeper than with civilians. Violence amateurs: civilians We have already highlighted the role of state agents in genocide, but genocide may also involve civilian participation. We will highlight two key processes for mass participation in genocide: escalating commitments and forward panic.76 In a pre-­genocidal social formation, a general simplification of social life is often evident. Individual action is increasingly constrained by the state, enemies are identified, and communal efforts are demanded to counteract an urgent and growing threat. Effectively, the state imposes itself increasingly upon the private sphere to ensure obedience; this alignment between public and private reduces dissonance and divergence between state policy and popular opinion. Authoritarian governments often seek to strengthen the intensity of individual identification and interaction with the state and its institutions. Before participation in violence takes place, prospective perpetrators (and bystanders) may have participated in many non-­violent acts at the regime’s behest. This participation need not be fully voluntary, but broadly speaking, it centers on an adherence to rules. Social psychologists have found that, through the “foot in the door” phenomenon, compliance will increase greatly if significant requests are preceded by relatively insignificant ones. For this strategy to be effective, the initial request must be meaningful and require effort; compliance with it must be seen as purely voluntary.77 In the case of genocide, self-­perception inferences are crucial in aligning perpetrators with the moral context; the perpetrators must adjust their worldview, through their acts, to regard their actions as just and meaningful. Even small acts (such as a perfunctory “Heil Hitler”) can increase identification with the regime.78 Civilians are often cumulatively exposed to acts of public violence against the victim group.

The genocidal context   59 Sociologist Randall Collins argues that violence may also be produced through a process he calls “forward panic.” This characterizes situations in which a period of tension is followed by a sudden overturning of power relations, usually reflecting the vulnerability of an opponent. Collins cites the examples of police beatings following car chases, or the massacres of civilians and hors de combat soldiers that sometimes follow the taking of a military objective. The mechanism of forward panic may help to explain the occurrence of spontaneous, mob violence in genocide, as occurred, at least to some extent, in the Rwandan genocide and the Gujarat pogrom (2002). In genocide, such tension is socially produced, particularly when it comes to locating the source in the victim group. The onset of genocide may represent a moment in which perceived power imbalances are suddenly overturned. Mass participation in violence often occurs where periods of prolonged political tension and insecurity (strain) are followed by a fundamental shift in the power relationship between victims and perpetrators. In these circumstances, the victims are rendered vulnerable – they are disempowered and may be targeted with impunity. The mobilization of mob violence may also represent a shift for participants, from tension to release and even ecstasy, as violence is interpreted as a settling of scores. The conditions that enable forward panic are still, of course, state-­created. Forward panic may also be partly institutionalized (e.g., in the so-­called “riot systems” of South Asia) by assuring would-­be perpetrators of impunity for their actions. Jews in Poland often knew about impending pogroms – Jewish communal authorities had for centuries designated special funds for the appeasement of local authorities (who would then head off the pogrom).79 Although forward panic may offer a partial explanation for some episodes of mob violence in genocide, it still does not explain specific patterns of mobilization. Prisons and concentration camps The concentration of victims is often an early stage on the path to genocide. It delegitimizes victims’ claims to territory; moreover, it renders them more vulnerable by placing them in areas which are often marginal. It also contributes to victim’s social segregation and marginalization. For example, the daughter of the Nazi governor of the Warthegau (Poland) described the Łódź ghetto to her fiancé: It’s really fantastic. A whole city district totally sealed off by a barbed-­wire fence […]. You mostly just see riff-­raff loafing about. On their clothes, they have a yellow Star of David both behind and in front.… You know one really can’t have any sympathy for these people. I think that they feel very differently from us and therefore don’t feel this humiliation and everything.80 Concentrating victims also makes them easier to kill – rather than hunting diffuse groups of victims, the state confines them to a manageable area. Detention facilities may be used as a tool of repression, to hold victims before they are

60   The genocidal context killed, or to kill them through means such as executions, poisonous gas, torture, exposure, or starvation. In concentration camps, the omnipotence of the perpetrator over the life of the victim is made manifest.81 The victims are transformed into objects of revulsion as they are stripped of culture and reduced to a position of “bare life.” Richard Glazar, a survivor of the Treblinka Nazi concentration camp, writes: Obviously no ordinary standards of emotion or behavior can apply; because all of existence, for us [the prisoners] especially, and up to appoint, at least by reflection, also for them [the guards], was reduced to a primeval level: life and death … one must not forget their incredible power, their autonomy within their narrow, and yet, as far as we were concerned, unlimited field; but also the isolation created by their unique situation by what they – and hardly anyone else even within the German Nazi Community – had in common.82 The perpetration of genocide may be seen as a self-­affirming manifestation of the omnipotence of the perpetrator and the powerlessness of the victim. Nowhere is this dichotomy more apparent than in the highly-­structured environment of concentration camps and institutionalized torture systems. A former Tuol Sleng prisoner remembers: The guards beat the prisoners like they were not human.… If I have to compare between human and animal, we were treated worse than their pets. They could look straight at us; we didn’t dare to look back at them – we couldn’t even look at them! There are no better words to describe the conditions we encountered.83 Within camps, victims are entirely deprived of their privacy. Privacy and its absence also signal personhood and power – the prisoners will undress in front of the guards but the converse is never true. Within the “camp world,” perpetrators are gods, exercising limitless power over their dehumanized victims. Camps are often isolated from wider society and maintain their own strictly-­ regimented order. Concentration camps and certain prisons may function as institutionalized perpetration systems – institutions whose primary function is perpetration.

The genocidal state as an organizational culture Genocidal states seek to produce certain types of behavior that would conventionally be defined as criminal. Such states tend to be authoritarian, militarized societies in which great pressures are brought to bear on individuals to obey and conform. The criminogenic state also uses coercion and a narrowed discourse to limit the perceived options available to potential perpetrators; they must believe that doing the bidding of the regime is their best or even only option.

The genocidal context   61 Many perpetrators of genocide are members of organizations. Organizational crimes are the outcome of organizational goals and constraints. Ideology provides organizational goals to genocidal organizations and constrains individual action to prescribed organizational norms. These organizational norms accrete into organizational culture, which forms the context for the behavior of all members of the organization. Other sources of organizational culture include routines (the everyday practices of the organization), rituals (training programs and ceremonies), stories (the narratives and myths which prevail within the organization), symbols, control systems (systems of promotion and punishment), and power structures (those who are the most powerful/most senior are also the most closely associated with desired values).84 We know that genocide has the effect of objectifying the victim, but this is also usually accompanied by a self-­objectification on the part of perpetrators. This self-­objectification renders us mere objects of the machine and frees us from moral conflicts. Thus, self-­objectification involves both the denial of feeling and the denial of agency. Self-­objectification can be accomplished through complete immersion in organizational culture. Certain organizations function as criminogenic systems: that is, certain features of their internal structures play a role in generating criminal activity within the system. The genocidal state is one such criminogenic organization. Needleman and Needleman distinguish between “crime coercive” and “crime facilitative” systems.85 The genocidal state and genocidal organizations are both crime coercive systems, because crime flows directly from organizational goals, rather than being incidental to them. Organizational culture can structure the decision-­making of individuals, by redefining moral rules and causing them to shift from individualism to “groupism.” The moral context also acts as an organizing force and principle for carrying out violence: it communicates norms and decentralizes power to the local level. The inculcation of values in genocidal states also means that the desires of the state may be communicated indirectly to individual perpetrators, even in the absence of direct commands. This aspect of genocidal mobilization is often overlooked in criminal trials, which seek out formalized (or at least de facto) chains of command. Rather, the situation is often more akin to “working towards the Führer” – the inculcation of organizational goals at the lower level. This means that individual perpetrators (especially mid-­level ones) are often placed in a position where they must act, within their range of effective control, to advance broader organizational goals. For example, the Nazi German military doctrine of Auftragstaktik (mission command) gave broad directives while delegating authority to the local level.86 There is always such a process of local interpretation and improvisation in response to national objectives. This improvisation may be focused on realizing local or individual goals, but it may equally promote alignment with central power and public opinion.

62   The genocidal context

State power and the localization of the moral context Genocide is a dynamic and interactive process produced through central coordination, local improvisation, and contingent responses to shifting circumstances. Many scholars argue that genocide is incremental and preceded by preparatory stages, including hate speech, the establishment of institutions of discrimination, and the concentration of victims.87 But these linear models fail to capture the complexity and fluidity of genocide. To enact genocide, the state must be able to mobilize perpetrators on the local level. It does so partly by communicating moral norms and inculcating a climate of existential fear. But the state must also be able to impact upon the local level. Thus, states that do not possess such a capability, i.e., failed states, cannot enact genocide. There are several factors that shape the ability of the state to implement genocide at the local level. These include the cohesion of national power, the local saliency of appeals to violence, and power relations at the local level (local rivalries and power structures). The state must possess effective control over its territory. States that do not possess effective control lack the ability to implement genocide. McDoom demonstrates, for the Rwandan genocide, a correlation between areas controlled by the ruling elite and the onset of genocide.88 The organizing power of genocidal discourses may be reduced when the state is strongly opposed by other groups claiming legitimacy. However, Brehm’s findings on Rwanda indicate that areas broadly under the control of the MRND, but in which power was contested by the RPF or other political opponents, may have experienced higher levels of violence.89 Perhaps, then, violence in such cases functions as a means of asserting control in the face of external or internal threats. As for the saliency of genocidal appeals at the local level, this varies depending on context. Where there is a history of violence between the audience and the target group, this increases saliency.90 Similarly, local conflicts contribute to the notion that the victim group poses an existential threat. There may be an alliance between local and central actors, with the former providing personnel for the central conflict, while the latter give local actors the necessary leverage over their local rivals.91 Consider again the example of Rwanda. Throughout the pre-­colonial era, the country was essentially a feudalistic society, with strongly entrenched social strata and norms of obedience. This continued in certain respects during the post-­colonial era. Scott Straus rightly notes that the Rwandan state has an “unusual depth and resonance at the local level.”92 After the RPF invasion in 1990, a new level of administration was set up, the nyumbakumi (Swahili for “ten houses”), further contributing to a situation in which, as Augustine Brannigan writes, there was a “direct line of political control from the president’s office right down to the individual hamlets.”93 Moreover, the mobilization of the population in Rwanda was easier because of the tradition of corvée or forced or communal labor (ubureetwa and umuganda, respectively). One perpetrator remembers, “we used to be in a group and those groups used to work in shifts.”94 It is striking that the killing euphemism

The genocidal context   63 used most often by génocidaires was “work” (urakora). The word Interahamwe itself originally referred to umuganda work parties. Cambodia was also a feudalistic society where the king was even said to have the power to control rainfall.95 Patronage relationships were also very important in Cambodia as patrons provided security and benefits to their clients, in return for their unquestioning loyalty.96 Acceptance of the given social order was considered as the best path to a satisfying existence.97 The Khmer Rouge exercised pervasive control over its citizens. One might further argue that the extreme obedience demanded by the Khmer Rouge was relatively easy to command in Cambodia – a society that was already extremely hierarchical, with separate pronouns and vocabulary for different ages and classes, as well as extensive patron-­client relationships.98 There are also many rituals in Khmer culture associated with deference to rank, including a system of bows and non-­ verbal greetings, not sitting higher than your “superiors,” avoiding their eyes, and not walking in front of them.99 While the Khmer Rouge supposedly sought to eliminate this stratification, it introduced fresh forms of hierarchy as the so-­called “new people” were degraded and dehumanized.100 The authoritarian political system was also one of ritual obedience. For example, Khmer Rouge slogans insisted: “The Angkar orders, execute!” and “He who protests is an enemy; he who opposes is a corpse!”101 The coupling of strong norms of obedience (structured through the genocidal state and communicated through propaganda) with extraordinary state reach and an absence of independent civil society institutions to mitigate state power, greatly facilitates the mobilization of genocide. Brannigan calls this state of nearly absolute control “administrative closure.”102 This chapter has focused particularly on the role of the state in fomenting genocide. But can genocide also be citizen-­led and state-­supported? Local leaders can certainly enact policies that surpass the intentions of the central regime. Such radical local policies may demonstrate that violence on a larger scale is possible. This may have been the case in Nazi Germany, where The Reichsgau Wartheland implemented the mass killing of Jews in response to both local conditions and perceived regime objectives. Yet even where genocide is influenced by such bottom-­up processes, the state still provides an opportunity space and moral context facilitating a recourse to extreme “solutions” at the local level, particularly by devaluing and disempowering the victim group.

Conclusion: structuring choice Unlike murder, genocide is endorsed by mainstream authorities in the societies in which it occurs. Genocide, then, is not internally deviant (deviant in the society in which it occurs), but only externally deviant (deviant within the global community). The state itself acts as a criminogenic system: its internal characteristics generate criminal activity through harm-­producing power relations. Yet genocidal contexts are also often characterized by a tension between social control and social disorganization. The state emerges as a response to

64   The genocidal context strain, and attempts to govern amid extreme social disruption and, usually, warfare. The criminogenic context may also differ depending on individuals’ positions in society, as I will explain in Chapter 6. A tension exists between old norms and those that emerge in the genocidal context. Individuals, including perpetrators, must adapt to the emergent context and reconcile it with their own moral beliefs, experiences, and interests. The state contributes to the framing of available options by externalizing responsibility, devaluing victims, and desensitizing perpetrators through the institutionalization of new moral habits. By these means, the individual consequences of perpetration are significantly reduced.

Notes    1    2    3    4    5    6    7    8    9   10

Interview R31 (Rwanda), Butare, August 2009. Wikström and Treiber, “Violence as Situational Action,” 88. Milovanovic and Henry, “Constitutive Criminology,” 272. Card, “Genocide and Social Death,” 63–79. Bandura, “Moral Disengagement in the Perpetration of Inhumanities,” 193–209. Miller, and Eisenberg, “The relation of empathy.” Neitzel and Welzer, Soldaten, 147. Neitzel and Welzer, 147. See: Brantingham and Brantingham, Environmental Criminology. Lower, Hitler’s Furies, 142. These figures do not include non-­German/Austrian perpetrators.   11 Census of Germany, May 17 1939. This figure includes the Saar, Austria, Sudetenland, and Memelland. It is also important to note that I calculated this figure on the basis of a definition of the age of majority as being >18 years old rather than the definition of >21, which was used in Germany at the time. The total population was 79,375,281.   12 Straus, “How many perpetrators were there in the Rwandan genocide?,” 94. Straus’ estimate is of “active males” 18–54 years old and does not include women.   13 Klee, Dressen and Riess, The Good Old Days.   14 Milgram, Obedience to Authority, 136.   15 Milgram, 137.   16 Milgram, 138.   17 See: Waller, Becoming Evil, 104; and Browning, Ordinary Men, 172.   18 Obedience increased to nearly total if the subject performed a subsidiary task and did not directly administer the shocks (a confirmation of the power of compartmentalization). Moreover, when people were not under direct surveillance “cheating” occurred and people administered lower shocks (avoidance behavior in a situation of information asymmetry).   19 Klee, Dressen, and Riess, 82.   20 Ian Kershaw, “ ‘Working Towards the Führer’.”   21 Prosecutor v. Dominic Ongwen, 21.   22 Bauman, Modernity and the Holocaust, 162.   23 Bauman, 164.   24 This was demonstrated in the famous Asch experiment (1951) where tested individuals knowingly gave a wrong answer to conform to the responses of their peers.   25 Hatzfeld, Machete Season, 12.   26 Beck, Prisoners of Hate, 144.   27 Beck, 146.

The genocidal context   65   28 Hood, The Self Illusion, 190.   29 See, for example: Asch, “Opinions and Social Pressure.”   30 Agamben, State of Exception. Agamben also argues that the state of exception may in some ways be the normal condition in states and that the supremacy of the law may itself be a fiction (87).   31 Balint, Genocide, State Crime, and the Law, 42.   32 Arendt, The Origins of Totalitarianism.   33 Carson, “The Conventionalization of Early Factory Crime,” 39.   34 Chen and Yeh, “The Construction of Morals,” 84–105.   35 Balint, 34.   36 Balint, 35.   37 Balint, 43.   38 Guichaoua, From War to Genocide, 225.   39 Alvarez, Governments, Citizens, and Genocide, 106 (Footnote 110).   40 Bauman, 22.   41 Bauman, 226.   42 Smith and Mackie, Social Psychology, 278.   43 Staub, The Roots of Evil, 84.   44 Waller, 188.   45 Aronson, The Social Animal, 212.   46 Interview R35 (Rwanda), Rwaza TIG Camp, August 2009.   47 Alvarez, Governments, Citizens, and Genocide, 104.   48 Bauman, 24.   49 Bauman, 24.   50 Lifton, The Nazi Doctors, 494.   51 Overy, Interrogations, 475.   52 Chon and Thet, Behind the Killing Fields, 162.   53 The use of local auxiliaries such as Lithuanians, Latvians, and Ukrainians was ostensibly done in part to spare the Einsatzgruppen personnel from the psycho­logical burden of killing.   54 Interview R61 (Rwanda), Mont Kigali TIG Camp, October 2009.   55 Horowitz, Taking Lives, 342.   56 Alvarez, Governments, Citizens, and Genocide, 94.   57 Rauschenbach, Staerklé and Scalia, “Accused for Involvement in Collective Violence,” 228.   58 Rhodes, Masters of Death, 26–27.   59 Mueller, “The Banality of ‘Ethnic War’,” 49.   60 Interview R55 (Rwanda), Rugerero TIG Camp, September 2009.   61 Alvarez, Governments, Citizens, and Genocide, 96.   62 Glover, Humanity: A Moral History of the 20th Century, 51. See also Grossman, On Killing, 252–255.   63 Castano, Leidner, and Slawuta, “Social identification process,” 268.   64 In the Holocaust reserve police battalions were dispatched from Germany to the eastern front to directly participate in killing.   65 Rothe and Mullins, “Genocide, War Crimes and Crimes against Humanity in Central Africa,” 151.   66 Alvarez, Genocidal Crimes, 91.   67 Rothe and Mullins 152.   68 Alvarez, Genocidal Crimes, 90.   69 Beck, 272.   70 Interview U10 (Uganda), Gulu, July 2009.   71 Kaplan, “Deviant Identity and Self-­Attitudes,” 113.   72 Stretesky and Pogrebin, “Gang-­Related Gun Violence,” 86–87.   73 Bouchard and Spindler, “Groups, gangs, and delinquency,” 922.

66   The genocidal context   74   75   76   77   78   79   80   81   82   83   84   85   86   87   88   89   90   91   92   93   94   95   96   97   98   99 100 101 102

Stretesky and Pogrebin, 88. Oliver, The Violent World of Black Men. Waller, 205–207. Smith and Mackie, 274–275. Staub, The Roots of Evil, 17. Gross, “Opportunistic Killings,” 19. Lower, 85. The Zimbardo Prison Experiment demonstrated the important role that rigidly controlled institutions such as camps may have in structuring the relationship between perpetrators and victims. Sereny, Into that Darkness, 178. Interview C06 (Cambodia), Phnom Penh, November 2008. Pettinger, Mastering Organizational Behavior, 189. Needleman and Needleman, “Organizational Crime,” 518. Weale, 3. Stanton, “The Ten Stages of Genocide.” See also Chalk and Jonassohn, The History and Sociology of Genocide, 261; and Hatzfeld, 66. McDoom, “Predicting violence within genocide,” 41. Brehm, “Subnational Determinants,” 23. McDoom, 41. Kalyvas, The Logic of Violence in Civil War, 383. Straus, The Order of Genocide, 8. Brannigan, Beyond the Banality of Evil, 94–95. Interview R02 (Rwanda), Kigali Central Prison, July 2009. Staub, 198. Hinton, Why Did They Kill?, 114. Bit, The Warrior Heritage, 64. Hinton, 184; see also Bit, 22. Hinton, 188. The “new people” were generally those from urban areas. Locard, Pol Pot’s Little Red Book, 122 and 204. Brannigan, 95.

Works cited Agamben, Giorgio. State of Exception. Chicago: University of Chicago Press, 2005. Alvarez, Alex. Genocidal Crimes. New York: Routledge, 2010. Alvarez, Alex. Governments, Citizens, and Genocide: A Comparative and Interdisciplinary Approach. Bloomington, USA: Indiana University Press, 2001. Arendt, Hannah. The Origins of Totalitarianism. New York: Schocken Books, 1951. Aronson, Elliot. The Social Animal (8th edition). New York: Worth Publishers, 1999. Asch, Solomon E. “Opinions and Social Pressure.” Scientific American 193, no. 5 (1955). Balint, Jennifer. Genocide, State Crime, and the Law. New York: Routledge, 2012. Bandura, Albert. “Moral Disengagement in the Perpetration of Inhumanities.” Personality and Social Psychology Review 3, no. 3 (1999). Bauman, Zygmunt. Modernity and the Holocaust. Ithaca, New York: Cornell University Press, 2000. Beck, Aaron T. Prisoners of Hate: The Cognitive Basis of Anger, Hostility, and Violence. New York: Harper, 2000. Bit, Seanglim. The Warrior Heritage: A Psychological Perspective on Cambodian Trauma. El Cerrito, USA: Seanglim Bit, 1991.

The genocidal context   67 Bouchard, Martin and Andrea Spindler. “Groups, gangs, and delinquency: Does organization matter?” Journal of Criminal Justice, 38, no. 5 (2010). Brannigan, Augustine. Beyond the Banality of Evil. London: Oxford, 2013. Brantingham, Paul and Patricia Brantingham. Environmental Criminology. Prospect Heights, Illinois: Waveland Press, 1991. Brehm, Hollie Nyseth. “Subnational Determinants of Genocidal Killing in Rwanda,” Criminology 55, no. 1 (February 2017). Browning, Christopher. Ordinary Men: Reserve Police Battalion 101 and the Final Solution. New York: Harper, 1998. Card, Claudia. “Genocide and Social Death.” Hypatia 18, no. 1 (Winter, 2003). Carson, W.G. “The Conventionalization of Early Factory Crime.” International Journal of the Sociology of Law, 7, no. 1 (1979). Castano, Emanuele, Bernhard Leidner, and Patrycja Slawuta. “Social identification processes, group dynamics and the behavior of combatants.” International Review of the Red Cross 90, no. 870 (2008). Census of Germany, May 17 1939. Chalk, Frank and Kurt Jonassohn. The History and Sociology of Genocide. London: Yale University Press, 1990. Chen, Donald L. and Susan Yeh. “The Construction of Morals.” Journal of Economic Behavior and Organization, vol. 104 (August 2014). Chon, Gina and Sambath Thet. Behind the Killing Fields: A Khmer Rouge Leader and One of His Victims. Philadelphia: University of Pennsylvania Press, 2010. Glover, Jonathan. Humanity: A Moral History of the 20th Century. New Haven: Yale, 2001. Gross, Jan T. “Opportunistic Killings and Plunder of Jews by their Neighbors – a Norm or an Exception in German-­occupied Europe?” Amsterdam: NIOD Institute for War, Holocaust and Genocide Studies, 2014. Grossman, Dave. On Killing. New York: Back Bay Books, 1996. Guichaoua, Andre. From War to Genocide: Criminal Politics in Rwanda, 1990–1994. London: University of Wisconsin Press, 2015. Hatzfeld, Jean. Machete Season. New York: Picador Books, 2005. Hinton, Alexander. Why Did They Kill? Cambodia in the Shadow of Genocide. London: University of California Press, 2005. Hood, Bruce. The Self Illusion: How the Social Brain Creates Identity. Toronto: HarperCollins, 2013. Horowitz, Irving Louis. Taking Lives: Genocide and State Power. New Brunswick, USA: Transaction Publishers, 2002. Kalyvas, Stathis. The Logic of Violence in Civil War. New York: Cambridge, 2006. Kaplan, Howard K. “Deviant Identity and Self-­Attitudes.” In Encyclopedia of Criminology and Deviant Behavior, edited by Clifton D. Bryant. Philadelphia: Brunner-­ Routledge, 2001. Kershaw, Ian. “ ‘Working Towards the Führer’ Reflections on the Nature of the Hitler Dictatorship.” In The Third Reich, edited by Christian Leitz. London: Blackwell, 1999. Klee, Ernst, Wili Dressen, and Volker Riess (eds.). The Good Old Days: The Holocaust as Seen by its Perpetrators and Bystanders. Old Saybrook: Konecky and Konecky, 1991. Lifton, Robert Jay. The Nazi Doctors. New York: Basic Books, 2000. Locard, Henri. Pol Pot’s Little Red Book: The Sayings of Angkar. Chiang Mai: Silkworm Books, 2004.

68   The genocidal context Lower, Wendy. Hitler’s Furies: German Women in the Nazi Killing Fields. London: Vintage, 2014. McDoom, Omar Shahabudin. “Predicting violence within genocide: A model of elite competition and ethnic segregation from Rwanda.” Political Geography, no. 42 (2014). Milgram, Stanley. Obedience to Authority: An Experimental View. New York: Harper Perennial Classics, 2004. Miller, P.A. and N. Eisenberg. “The relation of empathy to aggressive and externalizing/ antisocial behavior.” Psychological Bulletin, 103, no. 3 (1988). Milovanovic, Dragan and Stuart Henry. “Constitutive Criminology: Origins, Core Concepts, and Evaluation.” Social Justice 27, no. 2 (Summer 2000). Mueller, John. “The Banality of ‘Ethnic War’.” International Security 25, no. 1 (2000). Needleman, Martin L. and Carolyn Needleman. “Organizational Crime: Two Models of Criminogenesis” The Sociological Quarterly 20, no. 4 (Autumn 1979). Neitzel, Sonke and Harald Welzer. Soldaten. Frankfurt: Verlag Fischer, 2011. Oliver, William. The Violent World of Black Men. New York: Lexington Books, 1994. Overy, Richard. Interrogations: The Nazi Elite in Allied Hands, 1945. New York: Penguin Books, 2001. Pettinger, Richard. Mastering Organizational Behavior. New York: Palgrave, 2000. Rauschenbach, Mina Christian Staerklé, and Damien Scalia. “Accused for Involvement in Collective Violence: The Discursive Reconstruction of Agency and Identity by Perpetrators of International Crimes.” Political Psychology 37, no. 2 (2016). Prosecutor v. Dominic Ongwen, Document ICC-­02/04-01/15-T-­48-ENG (trial transcript), March 7 2017. Rhodes, Richard. Masters of Death: The SS-­Einsatzgruppen and the Invention of the Holocaust. New York: Alfred A. Knopf, 2002. Rothe Dawn L. and Christopher W. Mullins. “Genocide, War Crimes and Crimes against Humanity in Central Africa: A Criminological Exploration.” In Supranational Criminology: Towards a Criminology of International Crimes, edited by Alette Smeulers and Roelof Haveman. Antwerp: Intersentia, 2008. Sereny, Gitta. Into that Darkness. London: Vintage Books, 1983. Smith, Elliot R. and Diane M. Mackie. Social Psychology. New York: Psychology Press, 2007. Stanton, Gregory H. “The Ten Stages of Genocide.” www.genocidewatch.org/genocide/ tenstagesofgenocide.html. Accessed May 22 2016. Staub, Ervin. The Roots of Evil. Cambridge: Cambridge University Press, 1989. Straus, Scott. “How many perpetrators were there in the Rwandan genocide? An estimate.” Journal of Genocide Research 6, no. 1 (March 2004). Straus, Scott. The Order of Genocide: Race, Power, and War in Rwanda. Ithaca, USA: Cornell University Press, 2006. Stretesky Paul B. and Mark R. Pogrebin. “Gang-­Related Gun Violence: Socialization, Identity, and the Self.” Journal of Contemporary Ethnography 36, no. 1 (2007). Waller, James. Becoming Evil. New York: Oxford University Press, 2002. Weale, Adrian. Army of Evil: A History of the SS. New York: NAL Caliber, 2012. Wikström, Per-­Olof H. and Kyle H. Treiber. “Violence as Situational Action.” International Journal of Conflict and Violence 3, no. 1 (2009).

4 Propaganda Communicating the moral context

Introduction Those who can make you believe absurdities can make you commit atrocities. (Voltaire) The state speaks to its citizens through propaganda. This propaganda may, in some cases, have the power to transform individual consciousness and shape attitudes, but its main function is to communicate norms (acceptable opinions and actions). In genocidal contexts, the means of mass communication are often entirely controlled by the state. Consequently, propaganda, even when it comes from ostensibly private sources, represents the opinion of the state. Jean Stoetzel argues that individuals may have two separate opinions on the same subject: a private opinion only shared with people close to them, and a public opinion exercised in the public realm and shaped by public discourse.1 People may act on this public opinion even if it does not conform to their private opinion; over time, public actions may lead to the transformation of private beliefs, as individuals draw self-­perception inferences and rationalize their acts. Eventually, public opinions may be transformed into private opinions. Arthur Greiser, Nazi Gauleiter of Wartheland in occupied Poland, amply demonstrated this public opinion/private opinion dichotomy when he argued before a Polish court that his “official soul” had carried out his crimes, but his “private soul” had always been opposed.2 Through the mechanism of conformity, individuals often seek to align themselves with public opinion as expressed through propaganda. The success of propaganda is not reliant on the transformation of (private) opinion. Rather, state power ensures compliance with the new moral context (public opinion), despite diverging private opinions. This chapter addresses the role of propaganda in genocide, drawing from archival research into propaganda in the Netherlands under Nazi occupation, Rwanda during the civil war and genocide periods, and Serbia during the Balkans wars of the 1990s. It analyzes narratives found in genocidal propaganda and considers the importance of propaganda in the perpetration of genocide. It focuses particularly on the reframing of moral rules to exclude, demonize, and

70   Propaganda eventually attack the targeted group. It argues that propaganda communicates the moral context – providing guidance to individual perpetrators of violence.

The functions of propaganda Hate propaganda and incitement to violence are outlawed in many domestic and international laws. These regulations are based on the premise that such propaganda is a threat to the rights of others. Even where propaganda does not directly persuade, it provides justifications and a normative framework for acts of perpetration. Moreover, pervasive narratives “othering” the victim group, framing them as “the enemy,” may have a subconscious effect, even on those whose do not identify with prejudicial beliefs. Therefore, individuals in such societies are arguably more receptive to appeals based on hate than individuals in other societies, due to their continuous exposure to such views from legitimate authorities (and their lack of exposure to alternative perspectives). Hate propaganda communicates new moral norms. These norms are especially likely to contribute to violence when propaganda is coupled with direct appeals from authority figures, as well as institutional settings structuring individual decisions. Genocidal propaganda does more than just preach to the converted. It also provides inspiration, authorization, and normative expectation to the reluctant. For example, one perpetrator in Burundi recalls: “I heard FRODEBU [party] songs that said we are going to make a new Burundi after the death of Ndadaye. A Burundi without Tutsi.”3 Senders of genocidal propaganda strive to reverse or reduce the moral costs of killing, subsequently providing perpetrators with vocabularies of motive and absolution. Most theories of propaganda utilize a sender-­receiver model to analyze the methods and structures of propaganda.4 This model is based on the notion that propaganda involves a person or group (sender) who intentionally seeks to persuade an individual or group (receiver) with a particular message. However, it should be noted that not all attempts to persuade may be considered propaganda. Many definitions of propaganda also stipulate that the sender must be a government, but this excludes the importance of non-­governmental agents or proxies. It would be more accurate to state that propaganda is sent on behalf of a political organization or cause; thus, the ultimate sender is a political institution, organization, or cause.5 Moreover, individuals and organizations may produce propaganda that they believe aligns with state objectives. In other words, normative frames act as an organizing principle, not only for violence, but also for the production and dissemination of propaganda. The sender tends to be a formal group with internal organization, whereas receivers do not generally possess any real level of organization. Where the sender lacks legitimacy, propaganda may emphasize persuasion and assertions of legitimacy alongside normative communication (this was the case with Nazi propaganda in the Netherlands).6 The members of the receiver group are potential supporters of the sender. They are not linked formally to each other, but at the same time, they do share socially-­significant characteristics such as ethnicity,

Propaganda   71 race, sex, or age. Group cohesion thus needs to be strengthened, alongside other propaganda objectives. The reframing of moral rules in propaganda occurs through the production and reinforcement of myths, as well as through direct incitement to violence. Myth-­reinforcing propaganda creates the conditions for propaganda to be plausible and comprehensible. Incitement, on the other hand, functions more as a call to action: “the RTLM told me to kill the Tutsi,” notes one Rwandan perpetrator.8 Dehumanizing propaganda serves to increase the saliency of the authority figures’ messages by reshaping perpetrator perceptions of the victim and communicating moral norms. Propaganda constitutes a kind of ecosystem: it creates normative justifications for violence – vocabularies of motive – while the use of violence also creates a need for propaganda as post-­facto exoneration and coherence. The individual needs to frame her actions in such a way that they remain consistent with her notions of moral selfhood. Thus, propaganda preserves the perpetrator’s self-­ identity by externalizing responsibility and providing justification for acts of violence. It defines victims as perpetrators or combatants, legitimizing the use of lethal force against them in alleged self-­defense. Genocidal propaganda, like all propaganda, socializes people incrementally. 7

The dehumanization dynamic: scapegoating and normative shifts Genocide involves shifts in public policy, but also cognitive and cultural shifts. As argued in Chapter 2, the prelude to genocide is also a period of cognitive constriction in which a narrowing of awareness occurs alongside dichotomous processes of aggrandizement and devaluation. Such processes derive from state policy, but they are also facilitated by propaganda that seeks to produce social distance and shift norms. A genocidal campaign to kill every individual in a group is not possible without some form of ideological justification. This does not mean that every perpetrator is highly prejudiced. Yet when perpetrators receive relentless messages from trusted sources dehumanizing the victim group, it is likely that this will alter or erode their previously held views. Dehumanization is generally preceded by some form of scapegoating or moral blaming. Yet scapegoating is not enough to generate genocidal solutions. Rather, the group must be dehumanized and excluded from the moral community. Scapegoating is accomplished in four stages: first, an urgent social problem is identified; second, an identity group is negatively stereotyped; third, the social problem is attributed to the stereotyped group; and fourth, action against the stigmatized group is presented as being imperative. The first stage in scapegoating occurs when an urgent social problem is identified – such as the loss of resources necessary for survival, or a plot to destroy the in-­group. This social problem must pose an existential threat to group survival. This problem also provides a sort of “cognitive opening” in which circumstances drive individuals to consider new explanations and actions.

72   Propaganda Second, the group is labelled as “bad,” and stereotyped accordingly. Such stereotyping draws heavily on xenophobia (i.e., the mistrust of perceived strangers) and on the perceived position of the targeted group vis-­à-vis the person doing the stereotyping. Eric Hoffer writes: “the ideal devil is a foreigner. To qualify as a devil, a domestic enemy must be given foreign ancestry.”9 Hitler called the Jews foreign; Hutu nationalists portrayed Tutsis as foreign; Turkish nationalists emphasized the foreignness of persecuted minorities; Bosniaks (Bosnian Muslims) were often described by Serb nationalists as “Turks”; and during the French Revolution, aristocrats were portrayed as Germans. Stereotyping is a means for its authors to exercise power over the stereotyped. Stereotyping is closely related to what psychologists call the fundamental attribution error – we tend to attribute our own actions to circumstances, but essentialize others’ actions as derived from inborn characteristics. Stereotyping can shape our threat perception (in snap judgments based on superficial processing). For example, experiments occurring after the police shooting of black Haitian Amadou Diallo in New York indicated that, when viewing the same image, people thought that blacks were holding guns and whites were holding tools.10 Stereotypes are often persistent; people will ignore or downplay the importance of information that challenges their stereotype. They may also actively elicit information that confirms their stereotypes.11 Third, the out-­group is depicted as the source of threat to the in-­group. At this stage, the out-­group also becomes abstracted. It is no longer a collection of individuals, but rather an entity (object) that poses a threat. This negative abstraction is then applied to every member of the victim group, the “metaphysical other.” Finally, the out-­group is denigrated and dehumanized as action is taken to address the threat. In this final stage, action will likely be counselled against the out-­group – in the most extreme case, genocide. These stages of scapegoating are not necessarily sequential – they may occur simultaneously. Scapegoating is psychologically useful because it offers solutions to perceived problems.12 Moreover, anger is far more likely to result when we think that a person or group is at fault, rather than structural conditions.13 Dehumanization fundamentally seeks to separate individuals from their inborn markers of humanity, thus driving a wedge into social relationships that transcend group membership. Language, as an ontological system, is crucial in this process. One way that dehumanization occurs is through technicalization – a discursive strategy of power involving the use of technical language and euphemisms to describe the treatment of the victims. Such technicalization facilitates the perpetration of genocide by rendering genocide purely functional, insulating perpetrators from the moral implications of their actions. Another way to denigrate the victim group as a lesser form of life is by associating them with social unpalatability. For example, victims may be linked to disgust, death, dirtiness, perversion, or degeneracy. Psychologist Jamie Arndt conducted an experiment in which one group of test subjects simply read an essay about foreigners, while

Propaganda   73 the other group read the same essay after the word “death” was subliminally flashed on a screen in front of them. After this minimal symbolic association, the second group was much more hostile to foreigners.14 Ethnic groups are themselves cognitions (mental images), and by associating two cognitions – for example, “Tutsi” and “greedy” – behavior towards individual members of the stigmatized group may be modified. Memories of past behavior and beliefs also contribute to hostile framing, leading us to interpret all events in light of immutable, negative characteristics.15 If we believe the other group has knowingly violated a rule, this may also generate anger and other negative emotions.16 In general, there are two forms of dehumanization: first, objectification (the rendering of individuals as passive objects without human characteristics or merit), and second, animalization (comparing individuals to “lesser,” animal forms of life). In animalization, individuals may also be abused in the same manner as animals – as was the case with the “Gypsy hunts” from the sixteenth to the nineteenth century in Europe. The Khmer Rouge also made the “new people” perform tasks normally reserved for animals, such as pulling a plough. The use of euphemisms is crucial in dehumanization. It is well-­known that Tutsis were called inyenzi (cockroaches) during the Rwandan genocide, but prusak (“Prussian”) is a synonym for “cockroach” in Polish, just as Russe (“Russian”) is the German equivalent.17 Markers of difference may even be inscribed on the body, as was the case with prisoner tattoos in Auschwitz and Tuol Sleng prison in Cambodia.18 Dehumanization may also be imposed through customs and clothing (e.g., prison uniforms). Dehumanization is often a self-­fulfilling prophesy, as perpetrators consign the victims to situations in which they will manifest the desired characteristics – for example, filthy and emaciated concentration camp inmates. Marginal groups and classes often perform roles in society related to death or other unclean tasks.19 Thus, victims conform to, and thereby validate, perpetrators’ image of them.20 With action comes the transformation of hatred from a cognitive construct to a physical reality. Action also creates the need for justification: the perpetrator must come to believe that the victim deserved mistreatment. The dehumanization of victims transforms them into valueless objects that may then be “acted upon” by perpetrators.

Genocidal ideology Genocidal ideology infuses the perpetration of genocide with meaning. Ideology can be defined broadly as any coherent set of socio-­political ideas accepted by individuals or peoples. Genocidal ideology is related to other nationalist and racist ideologies, yet it also has certain distinct features, such as survival discourse – the notion that the victim poses a mortal threat. For example, around half of perpetrator respondents in Rwanda reported being afraid of the Tutsi-­ dominated rebel group, the RPF. Interviewees recalled:

74   Propaganda The MRND [governing party] was telling people to open their eyes and do something because the Tutsis are planning. People used to say the RPF would vivisect us. They taught us that if the RPF took power they would destroy the country. We thought if the RPF took power they would kill all Hutus.21 Other common features of genocidal ideology are claims of in-­group purity and out-­group contamination, grievances over contemporary and historical wrongs, and direct incitements to genocidal killing. Purity, in genocidal ideology, entails eliminating negative elements and promoting positive elements. Genocidal incitement is often euphemistic – the elimination of the enemy group is present in coded language. Ideology makes such incitement comprehensible and persuasive. By linking genocidal propaganda with existing cultural knowledge, perpetrators are provided with vocabularies of motive.22 By contrast, the type of hate propaganda seen in violent conflicts generally does not include propaganda directly calling for the elimination of the enemy group. Ideologies directed at the total transformation of society are often extremely utilitarian. Human lives are inconsequential in the quest for higher purpose, and thus the application of violence may be boundless. Khmer Rouge ideology was high-­modernist: it focused on perpetual progress, mastery of nature, and human emancipation through science.23 As one survivor of Democratic Kampuchea recounts,  we were told repeatedly that in order to save the country, it was essential to destroy all the contaminated parts … It was essential to cut deep, even to destroy a few good people rather than chance one “diseased” person escaping eradication.24 The ideology of the Khmer Rouge also emphasized anger against the so-­called oppressor classes. Genocidal ideology and propaganda serve a need. They provide individuals with a sense of identity, inspiration (including hope), certainty, and rationalizations for morally repugnant acts.25 Ideology conveys the perpetrators’ place in history; it ennobles them, and allows the mortal self to associate with the eternal. Through identification of the enemy, the solution to pressing problems becomes apparent. Furthermore, ideology can provide a powerful justification for atrocity, as individuals come to feel they are acting in conformity with a collective belief.26 Those with a greater stake in the system (mid- and high-­level perpetrators) may have a greater need for ideology to lend coherence to their actions. Rauschenbach, Staerklé, and Scalia’s research on high-­level Rwandan and Yugoslav perpetrators found that they were often preoccupied with their place in history.27 Ideology also has imperialistic tendencies: true believers in an ideology

Propaganda   75 will try to implement, proselytize, and impose it. Perhaps the most important function of ideology is to frame individual action alternatives. The urgency of the threat demands individual participation, while ideology simultaneously justifies and exonerates individual acts of perpetration. 28

Mechanisms of genocidal propaganda Propaganda works. Advertising demonstrates that human attitudes are subject to external influence. Genocide would not be possible without propaganda as a tool of mass mobilization, influence, and communication. People seek out propaganda because, like ideology, it reinforces their views and provides coherence, and the promise of privileged knowledge. For example, the prologue of the infamous anti-­Semitic propaganda document, The Protocols of the Elders of Zion, declares that “the protocols provide the key to a host of disturbing and seemingly insoluble riddles.”29 For receivers, propaganda is useful – it identifies threats, provides guidelines for actions, validates prejudice, filters information, and justifies actions already taken. Propaganda also fulfills psychological needs for mastery, self-­esteem, and belonging to a social group. 30 The development of mass media has greatly increased propaganda’s reach and potency. It allows individuals to receive the same message simultaneously, while still feeling autonomous, as in a darkened cinema.31 It is important that propaganda make each person feel personally addressed and valued, while still focusing on their connection to the broader group.32 In order to be effective, propaganda also must be total. It must use all forms of media (with their differing effects), and foreclose discursive space for opposing views. It must be continuous and immersive, so as to suppress personal reflection and the penetration of alternative ideas. The use of slogans is also essential to distill genocidal ideology to an easily comprehensible form. The categorical nature of such slogans excludes alternatives, nuance, and doubt.33 Propaganda often uses words connected to emotions and virtues.34 Yet propaganda not only appeals to the emotion of the receiver group; it also makes good use of reason. Contrary to popular belief, the conveyed message is not necessarily based on lies.35 In the end, the sender seeks at least an appearance of credibility; outright lies make this much harder. Another frequent technique is to accuse intended victims of committing atrocities – thereby legitimizing atrocity against this group and framing violence as responsive. For example, at one public meeting in Gikongoro during the Rwandan genocide, a local leader erroneously claimed that a Tutsi man had cut off the fingers of a young Hutu girl, and that the RPF had killed his own wife and children.36 Similarly, a former LRA fighter recalls: The commanders … sat every evening and told us, that [we are fighting] because we want to overthrow the government … second, Ugandan soldiers are sleeping with their mothers, abusing their children, and causing a lot of atrocities. That is the reason why we must fight tirelessly.37

76   Propaganda Moreover, much hate propaganda contains a “backhanded compliment”; a supposedly positive out-­group trait is depicted as negative. For example, in Rwanda and Burundi, Tutsi women were often portrayed as being very beautiful – but this beauty was a trap designed to ensnare Hutu men, who would then be seduced and betrayed. There is no great distance between the compliment of cleverness and the slander of deviousness. Genocidal propaganda provides perpetrators with external rationalizations for their actions. These eventually become internal rationalizations as they are embedded in perpetrators’ subconscious. The crucial moment comes when external ideas become internal attitudes. One Rwandan convicted of genocide related: “they were using songs and propaganda to drive hate into our heads – to make the Hutus believe that the Tutsis were just there to take their things.… We did not think it was hate – just reality.”38 Radio in Rwanda served a further purpose, according to a former RLTM deejay: “there was a system during the genocide whereby we used to communicate … directing where Tutsi were hiding. So, we were supposed to say [this] on the radio, so that Interahamwe would be aware of these places.”39 Propaganda is not always a top-­down phenomenon. It may also be disseminated horizontally, as group members indoctrinate each other. However, when similar propaganda is found in multiple locations, it may indicate that a coordinated campaign is underway. Genocidal messages are interpreted based on the perpetrators’ own motivations and experience, as well as the local cultural context.40 Propaganda disseminates an ideology to render certain acts acceptable (for example, to neutralize resistance to killing), but the social context also enables (regulates) the penetration of propaganda.41 Beyond individual dissemination via mass media (for example, a person reading a newspaper, listening to the radio, or receiving a text message), propaganda may also be received in a group setting, such as a political rally. When propaganda is received collectively, conformity effects are likely much stronger, as individuals imbibe cues from their reference group (respected authorities as well as peers). Propaganda may also be disseminated indirectly, through rumors, as individuals discuss the messages they have received, and listeners disseminate this information to others, often in slightly altered form. In Rwanda, information and rumors were often disseminated through face-­to-face conversations in local bars.42 Radio played an especially important role in the dissemination of genocidal propaganda in Rwanda. Only 29 percent of Rwandans had radios, but many Rwandans gathered in groups to listen to their friends’ sets. Among my 80 respondents in Rwanda, about two-­thirds listened to RTLM (Radio-­Télévision Libre des Milles Collines, a radio station promoting hatred against Tutsis), while a third read Kangura (a periodical containing hate speech). A former Interahamwe argues: “All Hutus listen to RTLM; it was their own radio.” Another perpetrator argued that RTLM was “unavoidable.”43 RTLM was not just a purveyor of hate; many were also attracted to the station “for the music.”44 Radio is distinct from other media, in that individuals often listen to it in group settings. Perhaps this social dimension reinforces the propaganda effect. It is interesting

Propaganda   77 to note that 38 percent of interviewees listened to radio in a group setting, one perpetrator even reporting that he listened to the RTLM at MRND meetings.45 In some areas prior to the genocide, radios were also distributed freely by local authorities.46 RTLM was a private station, but it had numerous government ties, and also often used the frequency of the government station, Radio Rwanda.47 Print media may have also played an important role in conditioning Rwandans for the genocide. Sixty-­six percent of Rwandans were literate in 1994, and newspapers were widely distributed, even in rural areas.48 Effective propaganda must spur people to action (or deliberate inaction). Thus, the presence of a group or institution to organize action increases propaganda’s efficacy. This is one reason why externally generated propaganda (originating outside the group) is of limited effectiveness; it is propaganda without structures of implementation. (It is also less trusted than internal propaganda.)49 The more vibrant the group (with feelings of belonging, common objectives, and structure), and the more individuals are integrated with the group, the more effective propaganda will be.50 A document used by the Rwandan government during the genocide, called “Note Relative à la Propagande d’Expansion et de Recrutement,” argued that propaganda would be more effective if events were created to lend credence to it.51 In the absence of institutional structures of state power, propaganda is of limited effectiveness.52 Consequently, it is easier to establish a correlation between genocide and genocidal propaganda than the unique causal power of propaganda. Acting on propaganda makes it immediately clear that the moral context created by propaganda is more than rhetorical. It is a framework for action – a normative guide for citizens. For the state to utilize propaganda to communicate and shift norms, it must first seize control of propaganda instruments. By doing so, the government ensures that its narrative is the only one heard, or at least the dominant one. In December 1990, Radio Rwanda was transformed when its director, Christopher Mfizi, was replaced by an extremist intellectual, Ferdinand Nahimana. In March 1992, Radio Rwanda broadcast a fraudulent list (supposedly obtained from RPF/ Tutsi rebels) of prominent Rwandans ‘marked for death.’ Other extremist media outlets, including RTLM and Kangura magazine, were also closely tied to the MRND government. The Bosnian case evinced a similar seizure of the means of propaganda. In spring 1992, Yugoslavian President Slobodan Milosevic dispatched paramilitary troops and technicians to seize a dozen television transmitters in predominantly Serbian areas of northern and eastern Bosnia. As a result, more than half of Bosnians began to receive TV broadcasts controlled by Belgrade rather than Sarajevo.53 Croatian broadcasts also turned increasingly nationalistic during this period, with appeals for Croatians to take up arms before the Serbs launched their ‘campaign of extermination.’54 Croatian Television also stirred up fears about a (Bosniak) Muslim state in which Croatian Catholics would be subjugated. The media content on both Serbian and Croatian sides was dehumanizing propaganda. Serb media described Croats as “vampirical Ustase,” while Croat media depicted Serbs as “beasts in human form” and “bearded animals on two legs.”55

78   Propaganda

Genocidal discourses Certain patterns in genocidal discourse became apparent during the research undertaken for this book.56 Propaganda narratives can generally be divided into two overarching categories: group-­binding propaganda (directed at the in-­group and designed to build support for collective action); and stigmatizing propaganda (directed at creating social distance from the out-­group). In practice, both categories contribute to social distance (the “humanity gap”), through aggrandizement on one hand, and devaluation on the other. These categories can, in turn, be broken down into several sub-­narratives – see Table 4.1. This typology draws partly from the “Story-­Based Theory of Hate” developed by American psychologists Sternberg and Sternberg, although they do not account for group-­binding propaganda. Hate propaganda, they argue, tends to disseminate certain “stories,” such as portrayals of the target as an impure other, enemy of god, stranger, controller, faceless foe, immoral, bringer of death, greedy enemy, barbarian, criminal, torturer, murderer, seducer/rapist, animal/pest, powermonger, subtle infiltrator, destroyer of destiny, and deceiver.57 These stories tend to have a beginning, an end, and a plot that generates hatred. Hate propaganda, then, is not static, but parallels events on the ground. Stigmatizing and group-­binding propaganda produces an ethnic norm governing the treatment of the out-­group by the in-­group. Stigmatizing propaganda Let us begin our analysis with stigmatizing propaganda narratives. The primary purpose of such propaganda is to define the enemy. Where the out-­group is associated with death, threats to safety, or threats to individual’s national or gender identity support for discrimination, aggression, and revenge increases.58 The definition of the enemy is founded on several questions – see Box 4.1: Box 4.1 The moral context of genocidal propaganda • • • •

Who is our enemy? Why are they our enemy? What must we do to them? Why must we do this?

Table 4.1  Typology of genocidal propaganda Stigmatizing propaganda

Group-binding propaganda

1  The Eternal Enemy 2  The Hidden Enemy 3  The Inhuman Enemy 4  The Foreign Enemy 5  The Barbaric Enemy

1  The Omnipotent Leader 2  The United Group 3  The Glorious Mission

Propaganda   79 In answering these questions, a devaluing moral context is produced that stipulates the identity of victims, ideological justifications for violence, the collective intention of perpetrators, motive, and even the means to implement genocide. The Eternal Enemy narrative essentializes the targeted group by presenting it as an implacable and timeless enemy. Group members are not to be trusted because they are an eternal threat, and/or because they have performed acts of atrocity in the past. Consequently, it recasts violence against the group as pre-­ emptive (or reactive) self-­defense, and legitimizes revenge as restorative justice for past atrocities. The following example from the NSB (Dutch Nazi Party) invokes the crucifixion of Jesus Christ to characterize Jews as Christ killers, while simultaneously presenting fascism as a bulwark against Jewish evil:  For 2,000 years you people yelled “crucify him” and it became your ruination, now you call to the successors of the crucified: “crucify them! and this too shall be your ruination! Jews of the Netherlands, our word shall not be in vain.… Fascism is resolute!”59 Similarly, in an article published in Politika on June 29 1989, Slobodan Milosevic invoked historical continuity to argue that Serbs were in a battle (also indicating grave danger): The Kosovo myth unified the entire Serbian people scattered throughout Yugoslavia.… Today, six hundred years later, we find ourselves at battle and facing battle once again. This is not a battle with weapons although weapons may yet be used.60 Bosnian Serb journalist Vlado Slijepcevic maintains this historicist view of the conflict, writing that the “Krajina liberators are breaking the strongest Ustase and jihad fanatics’ citadel in occupied Jajce.…”61 The use of the term Ustase invokes memory of World War  II, when Croatian fascist Ustase paramilitaries targeted Serbs Historization was also prevalent in Rwanda. Both the recent past (the civil war) and the more distant past (the pre-­colonial periods and post-­independence massacres) were invoked to demonstrate the essential evil of Tutsis. A Kangura article from 1990 reminded readers of the threat facing Hutus at the hands of Tutsis: “After the infamous massacres of Hutus in 1965, 1972–1973 and 1988 by minority Tutsi regimes in Burundi, the Hutu people are today, more than ever before, threatened with extermination.”62 Another Kangura article, two years later, presented the same picture of ongoing Tutsi oppression of Hutus: The situation, however, will not improve because Hutus will still be oppressed and will always be killed. Meanwhile, the Tutsis will act in the same manner: Will they actually accept being led by people who remember the graves of their parents, grandparents, brothers and sisters and who are still being killed simply because they were born Hutu!63

80   Propaganda An RTLM broadcast a few months before the genocide warns listeners that Tutsi wickedness “is not something recent, it is their usual behavior”: In Gatengerane sector, in early 1991, they killed citizens using knives. In Cyumba commune, in Gishambashayo in Rubaya sector, the Inkotanyi killed people, saying that they refused to show them where the Rwandan Army was. After having killed them with knives, they locked them in houses. The population thought that those people had fled, but after some days, they started to smell a bad smell around the house of so-­and-so and then, they discovered that those people had died in their houses. So, the wickedness that they showed in Ruhengeri is not something recent, it is their usual behavior; guerrilla fighters always show that kind of viciousness. So, what happened in Ruhengeri is not a new thing. In any case, such wickedness must be condemned, whatever you are fighting for.64 A former RTLM deejay discusses the strategies the station used to incite genocide: One mission of the RTLM was to focus on that issue, the 1959 revolution. Telling people, you need to keep this, please don’t mess up. Because they want to come back here and kill us and restore the monarchy. Don’t allow them to be back here; and here within the country are their friends and their relatives. So, these ones also, they are helping them. And some of their children have already joined the RPF.65 A sector counsellor similarly notes: “Talking about the RPF ’s desire to restore the monarchy was a political strategy.”66 Such discourses present a picture of violence and conflict set apart from time – it is perennial, because of the essentially evil nature of the targeted group. The Hidden Enemy narrative presents the target group as scheming, plotting secretly against the interests of the in-­group. This can include conspiracy theories and infiltration discourses. The indeterminacy of the out-­group’s evil is precisely what makes it ubiquitous and powerful. The hidden enemy discourse may also contribute to a pervasive paranoia, as in Cambodia, where, according to one perpetrator, “the enemy was invisible.”67 The narrative is especially prominent in attempts to mobilize mass killing (especially through participation). It blurs the line between civilians and armed combatants – in effect all civilians are armed combatants. Rwanda, like many other genocides, evinced an obsession with rooting out internal enemies; the term Ibyitso (accomplices) was often employed for propaganda purposes. The idea of infiltration may also violate notions of the fair fight – producing an enemy without honor. The following example from the Netherlands presents World War II as a battle against “Jewish World Powers”: “The hour of the last decision for Europe has come. The Führer has called his army and the armies of Finland and Romania to battle against the Jewish world powers in the East, against the USSR.”68 The

Propaganda   81 Jewish world-­conspiracy narrative was pervasive in Nazi propaganda, in the Netherlands and elsewhere: “The Englishmen-­Bolshevik dance to the tune of the Jewish clique.”69 Conspiracy narratives also often invoke the demographic threat posed by minority groups, and depict it as a form of infiltration: The Germans and Hungarians have personally experienced what it means when the Jews are in power. Then one can also understand that many Dutchmen, who feel for their country and people and think about the future, see with great fear the increase of Central European Jews in our country.70 Infiltration and conspiracy are said to be the modus operandi of Jews: “They [the Jews] do not like to work in the full light, but rather stay in the background.”71 In Bosnia, there was similar fear-­mongering about secret Muslim plans: … there are a secret network and secret message channels within the SDA [a Bosniak party] and the Muslim Bosniak Organization, and there have been various kinds of speculation about the 15,000 armed Muslim troops, who have … army equipment that has never been returned.72 Hutu extremists in Rwanda often raised the specter of a “Tutsi-­Hima conspiracy” to control the African Great Lakes region. The Tutsi threat was also presented through infiltration discourses. Perhaps this is not surprising, considering that the main power bases of opposition groups were among Tutsi exiles in neighboring countries, as well as abroad. The term Inkotanyi, slang for RPF fighters, was used to describe all Tutsis in Rwanda. The implication was that all Tutsis were involved in violence, that all Tutsis were threatening and legitimate targets of violence: … the Cockroaches Inkotanyi are trying to infiltrate themselves and they’re lucky because they return to where they have come from. If we happen to know where they have infiltrated in the daytime, we will chase them with tough means.73 Another perpetrator reasons: “we killed the Tutsis not because they were Tutsis … they said they were allies of the RPF.”74 In the following example, dated a month prior to the genocide, the Tutsis were presented as weevils (a pest) as well as infiltrators: At RTLM, we have decided to remain vigilant. I urge you, people of Biryogo, who are listening to us, to remain vigilant. Be advised that a weevil has crept into your midst. Be advised that you have been infiltrated, that you must be extra vigilant in order to defend and protect yourself.75 This Tutsi infiltration supposedly extended beyond military operations, infecting all aspects of life:

82   Propaganda Without being aware, the enemy is scoring points. He is very patient and subtle, moreover. The enemy is gaining ground and is continuously infiltrating our ranks, in all areas of national life. Be it with scholarships to study abroad, be it with success at each level, one same category of students allegedly better or more intelligent than the other, simply because they belong to a particular ethnic group. They might win the battle in the institutions of higher learning. And if this tide lulls us to deeper sleep, chances are that the children of the majority will suffer for who knows how long? This is a real crisis! We have reached a deadlock! We must realize that we are in hot soup!76 Tutsi infiltration was thus presented as a grave and urgent danger. Khmer Rouge propaganda similarly invoked a “concealed enemy boring from within.”77 The Inhuman Enemy narrative dehumanizes the target group, thereby placing them outside of the community of moral obligation. The enemy is presented as animalistic or otherwise non-­human. By means of such narratives, the sender is also implicitly arguing that the enemy is implacable and incomprehensible. In Dutch Nazi (NSB) propaganda, Jews were depicted as leeches: Moreover, we know them as the demolishers of our people’s community, as the puppets of the Jewish capitalism, as the double-­faced friends of Churchill and late Chamberlain, as the leaches [uitzuigers] of our “Volk’s” life, the people who undermine our power of resistance, the thieves of our national greatness.78 The enemy might also be seen as toxic, possessing inherently dangerous characteristics; their very presence poses the risk of contamination:79 What is race? Different “Volk” or people and races. Race is life attitude. […] we do not want a “mass” in which the individuals do not understand each other due to inherited differences. We do not want a “mass” that, through different natures, will bring our culture, our entire people to ruination.80 There are many similar examples of dehumanizing narratives from Rwanda: “So, Inkotanyi, instead of being exterminated, running all over this town, eating dirt, eating flour, eating unripe coffee, accept the negotiations because your time is over.”81 Here Tutsis are presented as dangerous animals: “They must pay back everything they spoiled together with [the] RPF. They acted like wild animals and you all know such animals are killers.”82 The essential nature of the Tutsis as cockroaches means that the ‘infestation’ must be exterminated: If you are a cockroach you must be killed, you cannot change anything, if you are Inkotanyi you cannot change anything. No one can say that he has captured a cockroach and the latter gave him money, as a price for his life, this cannot be accepted. If someone has a false identity card, if he is

Propaganda   83 Inkotanyi, a known accomplice of R.P.F. don’t accept anything in exchange, he must be killed.83 In such cases, propaganda was used to direct violence. Tutsis were also presented as essentially evil: “True, the Inkotanyi are synonymous with demons, and they have no desire to put an end to the massacre of their fellow countrymen who ardently support peace and noble justice.”84 The Foreign Enemy narrative also creates social distance by emphasizing the foreignness of the enemy group. This functions as a kind of distancing, while simultaneously removing the group from the political community, and delegitimizing their political and property claims. If they are unwanted interlopers, they have no right to benefits from the political community, or to geographic space: The Jews have been roaming Europe for centuries but never seemed able to truly merge with the people of countries they lived in. They remained strangers in the house of their host. This is because the Jew knows no national interests, only international [ones]. This results in great dangers for every people and every state.85 A perpetrator in Rwanda recalls: I heard it so many times … that the Tutsis were from Ethiopia … during the genocide they killed them [Tutsis] and put the dead bodies in the river so that they go back to Ethiopia quickly, by the shortcut.86 Before the genocide, a much subtler propaganda discourse, evident in President Habyrimana’s development speeches, implied that only peasants (Hutus) were real Rwandans.87 A similar agrarian romanticism was present in both Khmer Rouge and Nazi ideology. Barbaric Enemy narratives produce an image of the enemy as being barbaric, immoral, and committing grave atrocities. This propaganda often functions as a sort of psychological projection – what Alison Des Forges called “mirror politics.” You accuse the enemy of acts that you yourselves are committing. The Khmer Rouge, for example, accused the Yuon (Vietnamese) of committing “genocide against our Kampuchean race.”88 The moral implications of barbarism are once again a social distancing of the enemy, but also a quid pro quo authorization of reprisal attacks. Out-­group immorality is key mechanism in generating perceived threat.89 Such narratives therefore transform all perpetrator group violence into self-­defense or justice-­seeking. These narratives were prevalent in the Balkan wars, as in this Serb media report from 1993: There are the eyewitness accounts of Serb women who have passed through Croat camps. They were provided using ordinary words; mankind has not devised words that are horrible enough for what they say. Many of them are pregnant, some of them are infected with AIDS, several of them have

84   Propaganda committed suicide, some are under psychiatric treatment, and the majority do not want to talk, cannot remember. After all, all of these living tragedies, written out by hand, animals stamped with the [Croatian] checkerboard flag – they are all the same, and they are all old. “What else can we do to you to make you understand that we are not your brothers?” one of the right-­hand men wondered back in the last war. However, the story has been perfected: this time they are raped, beaten, killed … and accused of raping, beating and killing.90 The Croatians in this excerpt are presented as rapists, acting in the name of “the checkerboard [Croatian] flag.” Barbarism narratives often involve rumors of extreme cruelty: “In my hand I am holding gold teeth. I was told that Croat criminals are pulling them out of Serb civilians.”91 Such rumor-­based narratives also often involve barbaric acts against women and children: “It seems that Muslim extremists invented the most horrific crime on the planet. Last night they fed Serb children to the lion at the Sarajevo Zoo.”92 According to another broadcast on Radio Television of Serbia: Do you have an example that they [Croats] killed someone, cut his throat, or similar examples of such crimes? Interviewee: I left earlier, I didn’t see such things, but I heard from others that there was torturing. Reporter: Like what? Interviewee: Well, slaughtering, they were cutting off fingers, pulling fingernails off children … we have found children in pots ready to be baked. We discovered beheaded soldiers … Reporter: They have no mercy for anyone, do they?93 Reporter:

Merciless enemies lose any claim to merciful treatment. Thus, we see again the reversal of victimization. There are similar examples from Rwanda: The Inkotanyi use several tactics. As soon as they see that they have lost the armed war, they will use their sisters’ bodies as well as that of their wives and mothers. This has already happened [;] it is appalling! If a young man fornicates with his sister, he is capable of doing so with anyone.94 These acts of incest ascribe a profound immorality to out-­groups. Once again, the enemy is depicted as merciless: “When the Inyenzi catch a cellule [local government] committee member, they skin him alive.”95 And again: He told me that when Inyenzi killed Kayuku they disemboweled him and put stones in his belly. That was almost the same thing for Bucyana whose neck was cut off. And someone else came and crushed his body with a car.96 Through such projection, fear of the enemy is greatly increased. The survival discourse comes to the fore – that the very existence of the in-­group is threatened

Propaganda   85 not only by the actions of the out-­group, but by its very nature. The acts of the perpetrators are once again recast as self-­defense: You see … we set up these security units, because as we mounted patrols, those predators were … circumventing the roadblocks and attacking people’s houses, for instance, your house. So, we put a stop to all that.97 This radio message was broadcast during the Rwandan genocide – the mass killing of Tutsis – yet one gains the impression that Hutu extremists are merely acting in self-­defense in response to Tutsi violence. Group-­binding narratives Group-­binding narratives are focused not on the external group per se, but rather on the positive characteristics and mission of the in-­group. Alongside ideology, they facilitate the inculcation of an ethnic norm and reduce free-­riding and goal variance in genocidal violence.98 The Omnipotent Leader creates a figure of the mythical leader, thereby externalizing responsibility, while also bestowing the leader’s perceived positive qualities upon all group members. For example, while administering the Hitler Oath in 1934 Rudolph Hess implored: “Through your oath you bind yourselves to a man who — that is our faith — was sent to us by higher powers. Do not seek Adolf Hitler with your mind. You will find him through the strength of your hearts!” The United Group narrative stresses the unity of the group against external enemies, while simultaneously exhorting all group members to participate in the genocidal campaign. Such unification narratives are often coercive – group members who do not participate are ostracized, or even deemed traitors to be cast out or killed. For example: The Serbs and Montenegrins have not remained alone these past few days. Each home in Serbia is by their side. And each home in Serbia is ready to leave tomorrow for Kosovo […] because they [i.e., the “Albanian chauvinists”] know and must know also in the future that they will not nor will never occupy a single inch of Serbian territory.99 This passage, written by Slobodan Milosevic, presents Serbs as a united people fighting in defense of their homeland. The more negative aspects of unification narratives involve drawing group boundaries and identifying traitors. The Ten Commandments of the Bahutu in Rwanda are a powerful example of such a narrative:   (1) Every Hutu male should know that Tutsi women, wherever they may be, are working in the pay of their Tutsi ethnic group. Consequently, shall be deemed a traitor: •

Any Hutu male who marries a Tutsi woman;

86   Propaganda • •

Any Hutu male who keeps a Tutsi concubine; Any Hutu male who makes a Tutsi woman his secretary or protégée.

  (2) Every Hutu male must know that our Hutu daughters are more dignified and conscientious in their role of woman, wife or mother. Are they not pretty, good secretaries and more honest!   (3) Hutu women, be vigilant and bring your husbands, brothers and sons back to their senses.   (4) Every Hutu male must know that all Tutsi are dishonest in their business dealings. They are only seeking their ethnic supremacy. “Time will tell.” Shall be considered a traitor, any Hutu male: • • • •

who enters into a business partnership with Tutsis; who invests his money or State money in a Tutsi company; who lends to, or borrows from, a Tutsi; who grants business favors to Tutsis (granting of important licenses, bank loans, building plots, public tenders …) is a traitor.

  (5) Strategic positions in the political, administrative, economic, military and security domain should, to a large extent, be entrusted to Hutus.   (6) In the education sector (pupils, students, teachers) must be in the majority Hutu.   (7) The Rwandan Armed Forces should be exclusively Hutu. That is the lesson we learned from the October 1990 war. No soldier must marry a Tutsi woman.   (8) Hutus must cease having pity for the Tutsi.   (9) The Hutu male, wherever he may be, must be united, in solidarity and be concerned about the fate of their Hutu brothers; • • •

the Hutu at home and abroad must constantly seek friends and allies for the Hutu Cause, beginning with our Bantu brothers; they must constantly counteract Tutsi propaganda; the Hutu must be firm and vigilant towards their common Tutsi enemy.

(10) The 1959 social revolution, the 1961 referendum and the Hutu ideology must be taught to Hutus at all levels. Every Hutu must propagate the present ideology widely. Any Hutu who persecutes his Hutu brother for having read, disseminated and taught this ideology shall be deemed a traitor.100 Such a narrative marginalized Tutsis completely. Moreover, the Hutu Ten Commandments also reinforced gender norms, such as the importance of procreative, nurturing, and service-­oriented roles for women, and militarized ones for men. (This and similar ideological documents may have been more directly addressed to urban elites. One perpetrator notes: “I did not know about the Ten Commandments because I was in a village, but I did hear them in prison,” while another argues “that was something for the cities, not the countryside.”101)

Propaganda   87 Finally, narratives emphasizing The Glorious Mission speak to the nobility of the perpetrator group’s endeavor. They present the nation as glorious, possibly blessed by God, and all the group members as imbued with this same patina of nobility: Ennobling propaganda was used in the Balkans: “what we were discussing here can no longer be called politics, it is a question of our fatherland.”102 In the midst of the Rwandan civil war in 1991, Kangura exhorted: We are now in a critical phase. I am saying this because the war that we thought was almost over has taken a turn for the worse. You are all aware of how the Head of State has endeavored to restore peace in Rwanda, but in vain.103 Such narratives depict the enemy as implacable, pervasive, unjust, and ruthless, and the in-­group as noble and united, fighting against steep odds but assured of victory. The seed of moral neutralization is contained in these narratives, but they also contribute to the reframing of the moral context. The enemy’s barbarism, brutality, and evil mean that normal rules do not apply. “We must destroy them, or they will destroy us,” the narratives imply. Even those perpetrators who are not seduced by such arguments imbibe the sender’s message that the destruction of these groups is desirable and expected by the state.

Conclusion: state power and propaganda Propaganda plays a crucial role in the perpetration of genocide; it defines the enemy, absolves perpetrators, and incites violence. Propaganda is an exercise in decision-­priming: it does not dictate perpetrator action, but it helps to shape perpetrators’ perspectives in ways that narrows action alternatives, encouraging the commission of violence. The perpetrator begins to understand that violence against the enemy group is acceptable, or even expected. Propaganda communicates the moral context. It makes clear who the enemy is, why they are the enemy, and what must be done about them. Propaganda builds social distance and the power differential between victims and perpetrators by alienating, marginalizing, and dehumanizing victims while simultaneously aggrandizing the perpetrator group. In a sense, then, propaganda may have the effect of directing violence, even if that violence is produced by state power. Propaganda, we must remember, is also authored by state power. As a social crime, genocide requires a shared language of malevolence, and propaganda provides such a language. It is propaganda that incarnates resentment in a human form, and gives purpose to enmity.

Notes    1    2    3    4

Ellul, Propaganda: The Formation, 280. Arendt, Eichmann in Jerusalem, 127. Interview Bu8 (Burundi), Bujumbura, August 2009. See, for example, the definitions utilized in Jowett and O’Donnell, Propaganda & Persuasion; Speier, “The Rise of Public Opinion”; Lippmann, The Phantom Public; Herman and Chomsky, Manufacturing Consent.

88   Propaganda    5    6    7    8    9   10   11   12   13   14   15   16   17   18   19   20   21

  22   23   24   25   26   27   28   29   30   31   32   33   34   35   36   37   38   39   40   41   42   43   44   45   46   47   48   49   50   51

Ross, “Understanding Propaganda,” 19. The NSB (National Socialist Movement) was essentially the Dutch Nazi party. Ross, 19. Interview R26 (Rwanda), Gisenyi Prison, August 2009. Hoffer, The True Believer, 93. Smith and Mackie, Social Psychology, 178. Smith and Mackie, 182. Staub, The Roots of Evil, 17. Beck, Prisoners of Hate, 31. Berreby, Us and Them, 232. Beck, 94. Beck, 93. Waller, Becoming Evil, 247. Hinton, Why Did They Kill?, 224. Berreby, 233. Goldhagen, Hitler’s Willing Executioners, 172. Interview R01 (Rwanda), Kigali Central Prison, July 2009; R04 (Rwanda), Kigali Central Prison, July 2009; Interview R21 (Rwanda), Ruhengeri Prison, August 2009; Interview R27 (Rwanda), Gisenyi Prison, August 2009; and Interview R29 (Rwanda), Rugerero TIG Camp, August 2009. Hinton, 30. Hinton, 48. Hinton, 155. Staub, The Psychology of Good and Evil, 299. Ellul, 200. Mina Rauschenbach, Christian Staerklé, and Damien Scalia, ‘Accused for Involvement in Collective Violence: The Discursive Reconstruction of Agency and Identity by Perpetrators of International Crimes,” Political Psychology 37, no. 2 (2016). Ellul, 197. Cohn, Warrant for Genocide, 89. Ellul, 148–149. Ellul, 8. Ellul, 6. Locard, Pol Pot’s Little Red Book, 2. Sternberg and Sternberg, The Nature of Hate, 132. Ross, 20. Omaar, “The Leadership of Rwandan Armed Groups,” 253. Interview U10 (Uganda), Gulu, July 2009. Interview R1 (Rwanda), Kigali Central Prison, July 2009. Interview R73, Kigali Central Prison, October 2009. Hinton, 295. Ellul, 63. Jessee, Negotiating Genocide in Rwanda, 164. Interview R75 (Rwanda), Kigali Central Prison, October 2009; and Interview R15 (Rwanda), Gitarama Prison, July 2009. Interview R09 (Rwanda), Kigali Central Prison, July 2009. Interview R55 (Rwanda), Rugerero TIG Camp, September 2009. Des Forges, Leave None to Tell the Story, 67. Des Forges, 69. Des Forges, 67. Ellul, 21. Ellul, 50. Des Forges, 66.

Propaganda   89   52 Straus argues that in Rwanda propaganda had little effect on direct mobilization, but that it did catalyze hard-­liners, reinforce messages, and frame public choice. See Straus, “What Is the Relationship,” 609–637.   53 Price, “Bosnia,” 5.   54 Price, 5.   55 Glover, Humanity: A Moral History, 51. See also Dave Grossman, On Killing (New York: Back Bay Books, 1996), 130.   56 The material for this research was garnered from several sources: documents, posters, and other images from the NIOD archives; RTLM transcripts and Kangura issues from RwandaFile (Jake Freyer, editor) and Aegis Trust’s Rwandan Genocide Archive; and speech transcripts and media documents from the Foreign Broadcast Information Service. All Dutch language NSB propaganda was translated by Koen Klussien and is listed according to the dossier number in the NIOD archives.   57 Sternberg and Sternberg, 84.   58 Littman and Levy Paluck, “The Cycle of violence,” 90.   59 “Wat praat gij van anti-­semitisme?,” NIOD archive, dossier 123–591 (1935–1941).   60 Milosevic quoted in De la Brosse, “Political Propaganda,” 13.   61 Thompson, Forging War, 53. This excerpt was written by Bosnian Serb journalist Vlado Slijepcevic on Dnevnik 2, October 29 1992.   62 “He Who Lives by the Sword Will Die by the Sword,” Kangura no. 3 (July 1990), RwandaFile, accessed March 13 2015, http://rwandafile.com/Kangura/k03a.html.   63 B. Ndekezi, “Power Based on the Multiparty System Has Never Been in the Interests of the Tutsi,” Kangura 30 (January 1992), RwandaFile, accessed February 17 2015, http://rwandafile.com/Kangura/k30v.html.   64 L. Mpambara, (trans.), “RTLM 0039,” RwandaFile (November 24 1993), accessed 15 February 2015, http://rwandafile.com/rtlm/pdf/rtlm0039.pdf.   65 Interview R73 (Rwanda), Kigali Central Prison, October 2009.   66 Interview R74 (Rwanda), Kigali Central Prison, October 2009.   67 Interview C05 (Cambodia), countryside near Phnom Penh, November 2008.   68 “Het uur der eindbeslissing voor Europa is gekomen,” NIOD 249–056 A1.   69 NIOD 249–056 A1.   70 J. Boelen, “Jodenvervolging en Christenvervolging,” NIOD 123–591 (Amsterdam 1938).   71 “De positie van de Joden in Nederland,” NIOD 123–591 (1935–1941). Article on the idea that there have hardly been any Jewish politicians, followed by percentages of Jews in high ranks/well-­paid jobs.   72 M. Marjanovic, “The Muslim Boycott of Military Service,” FBIS (January 29 1993) in Vojska – military paper.   73 Nsengiyumva, F. (trans.), “RTLM 0011,” RwandaFile (May 28 1994), accessed March 13 2015, http://rwandafile.com/rtlm/pdf/rtlm0011.pdf.   74 Interview R02 (Rwanda), Kigali Central Prison, July 2009.   75 “RTLM 0340 – extract,” RwandaFile (March 14 1994), accessed February 15 2015, http://rwandafile.com/rtlm/pdf/rtlm0340x.pdf.   76 “Editorial: Rwanda – Why Should the Solidarity of the Majority Cause Sleepless Nights?” Kangura no.  35 (May 1992), RwandaFile, accessed February 23 2015, http://rwandafile.com/Kangura/k35b.html.   77 The Revolutionary Flag, “Pay Attention to Sweeping Out,” 4.   78 “De reactie als verdrukte onschuld,” NIOD 249–056 A1.   79 See Neilsen, “Toxification,” 83–95.   80 “Ras of ‘chaos’: de grondslagen van de scheppingsorde. Ons ‘bloed’ tegenover het joodsche bolsjewisme,” NIOD 249–0564 A1.   81 “RTLM 0065,” RwandaFile (April 14 1994), accessed February 14 2015, http:// rwandafile.com/rtlm/pdf/rtlm0065.pdf.

90   Propaganda   82 “RTLM 0115,” RwandaFile (April 22 1994), accessed February 15 2015, http:// rwandafile.com/rtlm/pdf/rtlm0115.pdf.   83 F. Nsengiyumva, (trans.), “RTLM 0011” RwandaFile (28 May 1994), accessed February 15 2015, http://rwandafile.com/rtlm/pdf/rtlm0011.pdf.   84 R. Zayana, “The Inkotanyi’s New Strategy of Widespread Destabilization,” Kangura no. 8 (January 1991), RwandaFile, accessed February 15 2015, http://rwandafile.com/ Kangura/k08a.html.   85 C. Thoen, “Jodenrede – radio,” NIOD 123–591 (1942).   86 Interview R2 (Rwanda), Kigali Central Prison, July 2009.   87 Verwimp, “Development ideology.”   88 The Revolutionary Flag, “The National Duties of All of Us,” 1   89 Brambilla, et al., “Morality and intergroup relations,” 820.   90 O. Bogavac, “In the name of a thousand years of culture: rape, beat, kill! Suffering of the underage aggressors,” FBIS, (February 27 – March 12 1993) in Duga.   91 RTS correspondent during Vukovar battle: Anes, “Balkans: Media and War Crimes.”   92 RTS broadcast by Rada Djokic in 1992: Anes, unpaginated.   93 RTS broadcast: Anes, unpaginated.   94 “Editorial: Hutus Should Help Kangura Defend the Hutus; I will Pay the Price if Necessary, but I Will Warn the Hutus,” Kangura no. 19 (July 1991), RwandaFile, accessed February 15 2015, http://rwandafile.com/Kangura/k19a.html.   95 M. V. Uwamahoro, (trans.), “RTLM 0142,” RwandaFile (November 29–30 1993), accessed February 19 2015, http://rwandafile.com/rtlm/pdf/rtlm0142.pdf.   96 C. Zikuliza, (trans.), “RTLM 0192,” RwandaFile (3 April 1994), accessed 20 February 2015, http://rwandafile.com/rtlm/pdf/rtlm0192.pdf.   97 M. V. Uwamahoro, (trans.), “RTLM 0009,” RwandaFile (May 17–18 1994), accessed February 15 2015, http://rwandafile.com/rtlm/pdf/rtlm0009.pdf.   98 Franciso Gutiérrez Sanín and Elizabeth Jean Wood, “Ideology in civil war: Instrumental adoption and beyond,” Journal of Peace Research 51, no. 2 (2014).   99 De la Brosse, 13. Slobodan Milosevic in Politika, February 6 1990. 100 Hassan Ngeze, “Appeal to the Bahutu Conscience,” Kangura no. 6 (December 1990), RwandaFile, accessed May 30 2016, http://rwandafile.com/Kangura/k06a.html. 101 Interview R12 (Rwanda), Gitarama Prison, July 2009; and Interview R53 (Rwanda), Mont Kigali TIG Camp, September 2009. 102 This was a speech Slobodan Milosevic made to the executive of the League of Communists of Serbia on returning from Kosovo in 1987. See Thompson, 53. 103 “Editorial: Hutus Should Help Kangura Defend the Hutus; I will Pay the Price if Necessary, but I Will Warn the Hutus,” Kangura no. 19 (July 1991), RwandaFile, accessed March 25 2015, http://rwandafile.com/Kangura/k19a.html.

Works cited Anes, Alic. “Balkans: Media and War Crimes. The International Relations and Security Network (2009). Accessed June 19 2017, www.css.ethz.ch/en/services/digital-­library/ articles/article.html/102376/pdf. Arendt, Hannah. Eichmann in Jerusalem: A Report on the Banality of Evil. New York: Penguin Books, 2006. Beck, Aaron T. Prisoners of Hate: The Cognitive Basis of Anger, Hostility, and Violence. New York: Harper, 2000. Berreby, David. Us and Them: Understanding Your Tribal Mind. New York: Little Brown & Co, 2005. Brambilla, Marco, Simon Sacchi, Stafano Pagliaro, and Naomi Ellemers. “Morality and intergroup relations: Threats to safety and group image predict the desire to interact

Propaganda   91 with out-­group and in-­group members.” Journal of Experimental Social Psychology, 49, no. 5 (2013). Cohn, Norman. Warrant for Genocide. London: Serif, 2005. De la Brosse, Renaud. “Political Propaganda and the Plan to Create a ‘State for all Serbs:’ Consequences of Using the Media for Ultra-­Nationalist Ends.” Report compiled for the International Criminal Tribunal for the Former Yugoslavia, 2003. Available at: www. stadensomoffrades.se/de_la_brosse_political_propaganda_pt1.pdf. Des Forges, Allison. Leave None to Tell the Story: Genocide in Rwanda. New York: Human Rights Watch, 1999. Ellul, Jacques. Propaganda: The Formation of Men’s Attitudes. New York: Vintage Books, 1973. Glover, Jonathan. Humanity: A Moral History of the 20th Century. New Haven: Yale, 2001. Goldhagen, Daniel. Hitler’s Willing Executioners. New York: First Vintage Books, 2001. Herman, Edward and Noam Chomsky. Manufacturing Consent: The Political Economy of the Mass Media. New York: Pantheon, 1988. Hinton, Alexander. Why Did They Kill? Cambodia in the Shadow of Genocide. London: University of California Press, 2005. Hoffer, Eric. The True Believer: Thoughts on the Nature of Mass Movements. New York: Harper Perennial Modern Classics, 2002. Jessee, Erin. Negotiating Genocide in Rwanda: The Politics of History. London, Palgrave Macmillan, 2017. Jowett, Garth S. and Victoria O’Donnell. Propaganda & Persuasion. Los Angeles: Sage Publications, 2015. Lippmann, Walter. The Phantom Public. New York: Macmillan, 1927. Littman, Rebecca and Elizabeth Levy Paluck. “The Cycle of Violence: Understanding Individual Participation in Collective Violence.” Advances in Political Psychology 36, suppl. 1 (2015). Locard, Henri. Pol Pot’s Little Red Book: The Sayings of Angkar. Chiang Mai: Silkworm Books, 2004. Neilsen, Rhiannon S. “ ‘Toxification’ as a more precise early warning sign for genocide than dehumanization? An emerging research agenda,” Genocide Studies and Prevention 9, no. 1 (2015) 83–95. Omaar, Rakiya. “The Leadership of Rwandan Armed Groups Abroad with a Focus on the FDLR and RUD/URANA.” Kigali, Rwanda: Rwanda Demobilization and Reintegration Commission, 2008. Price, Munroe E. “Bosnia,” in Munroe E. Price (ed.), Restructuring the Media in PostConflict Societies: Four Perspectives,” Cardozo Online Journal of Conflict Resolution 2, no. 1 (February 2001). Ross, Sheryl Tuttle. “Understanding Propaganda: The Epistemic Merit Model and its Application to Art.” Journal of Aesthetic Education, 36, no. 1 (2002). Sanín, Franciso Gutiérrez and Elizabeth Jean Wood. “Ideology in civil war: Instrumental adoption and beyond.” Journal of Peace Research 51, no. 2 (2014). Smith Elliot R. and Diane M. Mackie, Social Psychology. New York: Psychology Press, 2007. Speier, Hans. “The Rise of Public Opinion.” In Propaganda, edited by Robert Jackall, New York: NYU Press, 1994. Staub, Ervin. The Roots of Evil. Cambridge: Cambridge University Press, 1989. Staub, Ervin. The Psychology of Good and Evil: Why Children, Adults, and Groups Help and Harm Others. New York: Cambridge University Press, 2002.

92   Propaganda Sternberg, Robert J. and Karin Sternberg. The Nature of Hate. New York: Cambridge University Press, 2008. Straus, Scott. “What Is the Relationship between Hate Radio and Violence? Rethinking Rwanda’s ‘Radio Machete’,” Politics & Society 35, no. 4 (2007). The Revolutionary Flag, July 7 1978. Thompson, Mark. Forging War: The Media in Serbia, Croatia and Bosnia-­Hercegovina. London: Article 19, 1994. Verwimp, Philip. “Development ideology, the peasantry and genocide: Rwanda represented in Habyarimana’s speeches.” Journal of Genocide Research 2, no. 3 (2000). Waller, James. Becoming Evil. New York: Oxford University Press, 2002.

5 Who kills?

Introduction Genocide involves mass killing, yet not every individual kills. Why? I have focused thus far on the powerful role played by the state in creating, communicating, and institutionalizing the moral context of genocide. We must not discount, however, the variation that exists in individual responses to the moral context. How do we account for these differences? In this chapter, we will focus on propensity – individual characteristics that may influence the likelihood of perpetration. We will also consider individual variance, particularly in terms of alternatives to perpetration. Situational action theory argues that crime is a function of exposure (the moral context) on one hand, and propensity on the other. Propensity, according to situational action theory, is primarily a matter of individual orientation towards moral rules and habits. Wikström and Svensson argue: Moral rules are rules about what is the right or wrong thing to do in a particular circumstance, whereas ‘moral habits’ refers to automated responses to familiar circumstances based on a moral habituation to act in a particular way as a reaction to the particular circumstance.1 In the case of genocide, both moral rules and moral habits are heavily shaped by the discourses and institutions of state power. Moral rules are often communicated through both propaganda and the actions of state agents, while moral habits are inculcated through repeated obedience to state demands. For violence specialists, moral habits may also be developed through specific indoctrination processes. However, not all individuals interpret and enact state demands in the same manner. Individuals are not robots to be “programmed” by the state, nor are they vessels to be filled with state messages. Rather, all individuals have a degree of agency. Consequently, they act as entrepreneurs in interpreting and implementing the demands of the state, and responding to the genocidal context. There are some individuals, of course, who resist the state entirely, choosing instead to rescue others or resist authority. Wikström and Svensson further elaborate:

94   Who kills? an individual’s propensity to commit a particular type of crime depends on his or her tendency to see that particular act of crime as an action alternative. Individuals are likely to show variation in their propensity to see particular types of crimes as action alternatives depending on their preferences.…2 In other words, individual perceptions and preferences differ in crime, as they do in everything else. Many individuals may desire wealth but few will steal to obtain it. In Chapter 6 we will further explore the decision to commit genocide; here we are primarily concerned with how individual characteristics, positionality, and perception shape self-­control and choice. In the case of “normal crimes,” such as theft or murder, self-­control brings individuals into alignment with the demand of the state to abstain from criminal behavior. In genocide, however, self-­control, in the form of abstention from violence, may bring individuals into direct conflict with the state. Genocide, in a sense, represents continuity with conventional authority, even while participation in genocide may involve a break from other pre-­existing moral norms. This break is accomplished in part through the incremental escalation of violence and moral disengagement from victims. Consequently, when applying situational action theory to genocide, we must consider that in genocide the moral context (exposure) is far more powerful than with ordinary crimes. Individuals who resist criminogenic contexts are defying the state itself. Therefore, propensity in genocide is a function of individuals’ interactions with the moral context, particularly their proximity to the regime and other perpetrators, and their level of moral disengagement. Those with higher propensity are more likely to become involved in perpetration.

Profiling perpetrators Killing as a necessary evil What kind of person commits genocide? Answering this question is far from simple. In considering whether perpetrators are extraordinary, we must also consider the ‘normalness’ of perpetration. In almost all cases, it is a relatively small, but significant, subset of the “perpetrator group” that is directly involved in the perpetration of genocide. Among perpetrators, there is an even smaller group that leads violence, vanguard perpetrators, as I will explain below. In Rwanda, for example, Straus estimates that 75 percent of the killing was done by only 10 percent of the génocidaires.3 However, if we venture beyond this group of immediate perpetrators, we find many other individuals who are involved in acts of persecution against the targeted group, acts that facilitate the broader genocidal process. Moreover, there are individuals who never actively perpetrate, who are aware of the atrocities occurring in their midst but fail to take any measures to stop them. In major German cities around 30 percent of people knew about the Holocaust and a further 10 percent had “suspicions.”4 Such passive

Who kills?   95 bystanders are not actively involved in perpetration, but their inaction facilitates perpetration; indeed, it may even communicate the normative message that perpetration is not a deviant act, but rather something “normal.” The institutionalization of perpetration, moral disengagement, and moral neutralization contributes greatly to the conventionalization of perpetration, even while direct individual perpetration (killing) may still be seen as an unpleasant or unsavory task. Participation in genocide, then, may be at once unusual and conventional. The act of perpetration may remain unusual, even while it is tolerated and justified by the state in the revised moral context. An inherent tension is present in the idea of genocidal killing. Genocidal ideology’s focus on purity entails getting rid of undesired elements while also maintaining and facilitating the positive characteristics of the group. Participation in violence is often seen as a form of debasement, or, at best, recognized as a difficult but necessary task. Violence, then, is not inherently ennobling, though self-­ sacrifice is. Violence in many genocidal movements and ideologies is not the focus, but rather a means to an end; killing is a utilitarian “necessary evil,” rarely foregrounded in ideology or identity. The marginal Violence may sometimes be outsourced to paramilitaries or criminals; such groups remain on the margins of the state. For armies, the use of paramilitaries circumvents norms of military honor and conduct. The phenomenon of forced perpetration also shows that violence often remains an unattractive proposition, even in genocidal contexts. Forced perpetration spreads culpability for violence, as well as involving others in impure, dirty work. Paramilitarism also allows good citizens to keep the blood literally and figuratively off their hands. Perpetrators may be drawn voluntarily to their roles by self-­selection (e.g., sadists seek out roles that give them an opportunity to commit sadistic acts). Even in specialized killing organizations, such as the SS, many individuals were transferred to the service for mundane reasons, such as their unfitness for active combat because of injury.5 Moreover, sometimes people with a record of violence (such as convicted criminals) are recruited by perpetrating forces because of their habituation to the use of violence. In Bosnia, thousands of prison inmates were released after accepting offers of reduced sentences and war booty in exchange for joining paramilitary groups.6 There was a looting hierarchy among these groups, with the largest and most powerful groups, such as Arkan’s Tigers and the White Eagles, gaining preferential access to the most valuable goods, and local militias being left with stripping the wires out of walls and dismantling doorframes.7 Bosnian Serb paramilitary groups were said to be 80 percent composed of criminals, while significant underworld figures were also involved in Croat and Muslim paramilitary groups.8 This gave marginal individuals the chance to become heroes. Indeed, experts have contended that most atrocities in Bosnia were committed by men with criminal records.9

96   Who kills? In the Armenian genocide, some criminals were also pardoned so that they could take part in the violence.10 Similarly, in Darfur some Janjawiid were “Arab” criminals who had been released from prison; their participation in the militia gave them the opportunity to commit “legitimate” acts of violence.11 Convicted poachers were also recruited to an SS unit as a means of rehabilitation.12 On the other hand, more violent individuals (such as known psychopaths) were excluded from killing duties in the SS, largely for ideological reasons: it was feared that they would commit “excessive” acts of violence, such as torture, unbefitting the SS. Prunier argues that in Rwanda, the paramilitaries “had tended to recruit mostly among the poor. As soon as they went into action, they drew around them a cloud of even poorer people, a lumpenproletariat of street boys, rag-­pickers, car-­washers and homeless unemployed.”13 Yet most perpetrators in genocide do not have a marginal background or criminal record. Only a small number of Holocaust perpetrators had prewar criminal records, except for Kapos (prisoner foremen).14 None of the perpetrators interviewed for this study had criminal records prior to the genocide. Joining The reasons for joining organizations involved in genocide are many, including personal enrichment, excitement, power, status, and even a lack of options. This lack of choice may be circumstantial (such as a job transfer to a killing unit), but there might also be a security dilemma present. In situations of pervasive violence, whether gang violence or armed conflict, the ubiquity of membership in armed groups means that joining an armed group is often perceived as the only way to obtain security. An ex-­soldier in Burundi explains: “there were two parties in the conflict, the government and those who were against it. Everyone had to join one group.”15 A rebel on the other side of the violence concurs: “Joining the CNDD was a refuge for me, because I was afraid of getting killed by the army. I had to protect myself.”16 Another rebel notes, “I joined because they were massacring Hutus. I felt obliged to join.”17 A soldier (and former Sans Échec paramilitary) argues, “I joined the army because I saw that our people were being killed. I wanted to protect Tutsis.”18 Of course, some soldiers and perpetrators are forcibly conscripted, but polarization also contributes to the impetus to join armed groups. Social disorganization may also motivate individuals who perceive a lack of opportunity and upward mobility, particularly unemployed young men. As one Burundian soldier notes, “it was work. I had to support my family.”19 The situation of these individuals is made all the worse by patriarchal norms of providing for others. Young men may also be overrepresented among genocide perpetrators, and more prone to aggression generally (for physiological or sociological reasons). Perpetrators may also see joining a genocidal group as a way to exercise mastery and increase their security in situations of uncertainty and strain. Recruitment for paramilitary organizations may exploit existing networks (as happened with the recruitment of members of the Red Star Football Fan Club

Who kills?   97 for Arkan’s Tigers, or the links between government-­financed football fan clubs and the Interahamwe in the Rwandan genocide). Proximity to other perpetrators appears to be a strongly criminogenic factor. McDoom notes that in the Rwandan genocide, “participants often resided in neighborhoods with concentrations of other participants”; having a family member who is a perpetrator is even more criminogenic, perhaps the ultimate form of proximity.20 The typical perpetrator Logically, the greater the degree of mass participation in genocide, the more one would expect perpetrators to be broadly representative of their societies. Genocide may be implemented by relatively small cadres of hard-­core believers. It may equally be undertaken by ordinary people who are mobilized by various means, including coercion, conformity, ideology, and personal interest. In cases of limited mass participation, the moral context may also necessarily be limited to the open endorsement of violence within the group. In Rwanda, there were around 175,000–210,000 direct perpetrators.21 These perpetrators did not differ significantly from non-­participants in marital status, fertility rates, occupation education, religion or their social relationships with Tutsis.22 Many of those interviewed for this book reported having positive relations with Tutsis before the genocide (mixed Hutu-­Tutsi families were also common). Although genocide seems to be disproportionately committed by men,23 both children and women also participate in genocide. Many Khmer Rouge perpetrators were children; this reflected the Khmer Rouge’s rejection of the concept of childhood, and the relative ideological malleability of children, who are not yet fully cognitively developed.24 The LRA in Uganda made extensive use of forcibly conscripted children. One female ex-­LRA (herself a former child abductee) argued that children are easily corrupted: “it was very easy to deal with the kids. Because it is very easy to change a kid’s mind.” Another goes so far as to say that “kids are the worst people on earth.”25 In 1976, the staff of the Khmer Rouge S-­21 prison consisted of only 25 personnel over the age of 25, 108 between 18 and 22 years of age, and 20 under the age of 18.26 In Rwanda, thousands of children were charged with genocide, including a seven-­year-old.27 One Rwandan interviewee recalls children assisting with the killing: “I had no power to stop them at that time. Those children did not kill him, they just cried out and they screamed until the mature people came and killed the victim.”28 In the Holocaust, one German female camp guard was only 15 years old, although the average age was 26.29 There is much criminological evidence that young men are overrepresented among perpetrators of crime.30 Age curves peak particularly early for violent crimes. Yet in genocide, the prevalence of the young may not be universally true. Browning cites an average age of 39 for Reserve Police Battalion 101 (a unit involved in mass executions on the Eastern Front). Mann arrived at an average age of 32–41 among SS perpetrators, while Straus’s interviewees in Rwanda had an

98   Who kills? average age of 30–39.31 In war crimes trials at the Court of Bosnia and Herzegovina, the average age was similarly, 30.32 Brehm, et al.’s research further shows that the median age of those tried for genocide in Rwanda was 34. The median age for my Rwanda perpetrator interviewees was 33.8. They account for this older age range by arguing, as I do, that genocide differs from ordinary crimes, as it is a crime committed in conformity with the state and conventional values. Genocide, in other words may be “framed through a lens of duty and honor” rather than being a deviant act.33 Nonetheless, the ages of the perpetrators I interviewed in some other cases tended to be younger – Burundi, Uganda, and Cambodia all involved many child perpetrators; the median age of my Uganda interviewees was 12.2 (at the time they “joined” the LRA), and one of my Cambodia interview subjects was around 15 at the time he first became a perpetrator. Women are also always involved in genocide, in a range of roles. They are more likely to play supporting roles, such as acting as informants or providing moral and material support, but they also sometimes kill.34 The divergence of men and women in killing roles is likely grounded in patriarchal norms. Brehm, et al.’s research on the Rwandan genocide demonstrates that women were underrepresented in genocide, relative to ordinary homicide, and that in the genocide, women were more likely to be involved in incitement and property crime rather than direct killing.35 In 1998, there were 1,200 women imprisoned in Rwanda for involvement in the genocide.36 Verwimp estimates (based on his sample) that less than 10 percent of adult perpetrators in Rwanda were women.37 The relative paucity of female perpetrators was also true of Khmer Rouge Cambodia. In the notorious Security Prison 21 (S-­21) there was only one female interrogator.38 In The Holocaust, there were around 3,500 female concentration camp guards in the Nazi system, as well as 13 million female Nazi Party members.39 When you consider the overall adult female population in Germany in the last census before the war, as well as the statistics on male participation I gave on pp. 46–47, this means that German and Austrian men were approximately 60 times more likely to participate in genocide than German and Austrian women.40 These gender discrepancies in perpetration may not be explained through in-­born characteristics; meta-­analyses show that gender differences in the propensity for cruelty are “much smaller than commonly believed.”41 Moreover, recent research on sexual violence in the wars in the eastern regions of the Democratic Republic of the Congo indicate that women were perpetrators of sexual violence in 41 percent of cases with female victims and 10 percent of cases with male victims.42 Women involved in such extreme violence are often granted a degree of agency, but only as part of a movement directed against the “other”; they are not free to challenge patriarchal norms. Erna Petri, a German woman who killed several Jewish children, argued: I lived among men who were in the SS and carried out shootings of Jewish persons. I seldom came into contact with other women, so that in the course of this time I become more hardened, desensitized … I wanted to show them that I, as a woman, could conduct myself like a man.

Who kills?   99 While Petri was likely immersed in an environment of expressive violent masculinity, her statement here downplays the agency and initiative she showed in murdering several children, while completely alone, not allowing herself “to be swayed” by the pleas of her child victims.43 Once convicted, women are sometimes seen as doubly deviant, because they had transgressed their “natural” role as gentle mothers and caregivers; however, there may also be a reluctance to convict them for the same reasons.44

Perpetrator coalitions Rather than the mass mobilization of ordinary men, a more accurate picture of genocidal mobilization stresses the formation of genocidal coalitions comprising individuals and groups who participate for disparate reasons yet still facilitate the state-­driven genocidal campaign.45 Such a view of genocidal mobilization and perpetration also recognizes the multiplicity of individual motives and forms of violence that we see during genocide. The use of diverse means to enact genocide is the norm rather than the exception. For example, in the Rwandan genocide, killing was undertaken by a mix of ordinary civilians mobilized to participate, state agents (such as police and army), and paramilitaries (Interahamwe, Impuzamugambi). The army and Interahamwe often coordinated their violence, as did paramilitaries and the Yugoslav Army (JNA) in Bosnia.46 Paramilitaries may be used to attack soft targets, to commit atrocities, to mobilize civilians, and to mop up after military operations. Similarly, the Holocaust was implemented by highly-­indoctrinated SS camp guards, but also by Einsatzgruppen perpetrators of mass shootings, as well as civilians acting as bureaucrats, or as direct killers in the pogroms in the East. As Langerbein notes, the majority of the Einsatzgruppen were not “ordinary men,” but rather committed Nazis with longstanding ties to the regime. This does not mean, however, that they weren’t psychologically ordinary.47 One would expect the characteristics and propensity of perpetrators to vary significantly between violence professionals and civilians. Even among professional perpetrators who are socialized to kill, significant diversity is evident. One war correspondent who witnessed the execution of Jews in Latvia during the Holocaust wrote: As a journalist, what interests me are the people who had to carry out such an action … I saw SD personnel weeping because they could not cope mentally with what was going on. Then again, I encountered others who kept a score-­sheet of how many people they had sent to their death … who today can determine which ones were those who wept as they carried out their duties and which ones kept a score-­sheet?48 To profile perpetrators, then, we must consider the diversity of roles and of individual motive. As argued throughout this work, perpetrator motivation is multivariate. Both high-­level perpetrators and low-­level perpetrators may be partially

100   Who kills? motivated by status anxiety, particularly in situations of insecurity and change. The strong role of the state in structuring perceived individual action alternatives also contributes to a situation in which many individuals participate without strong intentionality. The profiling of “typical” individual perpetrators may seem to be an impossible task; rather, broadly-­speaking, we can categorize perpetrators according to several dichotomous traits. Professionals and amateurs First, there is the distinction between professional and amateur killers, which we have already highlighted. Professional perpetrators (specialists) often belong to organizations such as the army, police force, and paramilitaries. It is important to note, however, that sometimes these organizations, such as the Interahamwe, do not initially focus on killing. Rather, they gradually escalate acts of violence while binding members to the group and norms of obedience. Vanguard and reluctant perpetrators Second, one can distinguish between vanguard and reluctant perpetrators. Eager perpetrators, or vanguard perpetrators as I term them, often play a disproportionate role in killing, by initiating killing, repeating genocidal acts, and intimidating others to kill. Such perpetrators may be particularly useful to the state. A Rwandan perpetrator observes that “People who were enthusiastic killers were embraced by the authorities.”49 Reluctance may partly depend on internal moral controls, but can also be grounded in moral inhibition or visceral disgust. Yet even visceral disgust may be transformed into self-­pity through discourses emphasizing the burden of violence or the “necessary evil.” Violence professionals are less likely to be reluctant, as are those with conditions that predispose them to moral disengagement, such as psychological disorders. In fact, the primary role of the Interahamwe in the Rwandan Genocide may have been to coerce reluctant civilians into perpetrating genocide. Direct perpetrators and leaders Finally, we can distinguish between direct perpetrators and leaders. Direct perpetrators must overcome confrontation-­tension and fear, as well as visceral disgust, to kill. By contrast, leaders face none of these challenges but can view killing abstractly, rather than as an actual act. Leaders also exercise power, which gives them a level of agency that direct perpetrators may lack. This agency, found more broadly among elites including many mid-­level perpetrators, also requires greater ideological justification, while lower-­level perpetrators may participate simply because they feel they have no alternative. Lower-­level perpetrators may also participate because propaganda has made them authentically fear the enemy group. Leaders can be roughly divided into entrepreneurs (leaders using identity for instrumental reasons) and ideologues (ideological fundamentalists). For

Who kills?   101 entrepreneurs, violence is a “means of political control … a policy alternative to accommodation, one that is adjusted to the level of threat presented by the opposition.”50 For ideologues, on the other hand, human rights abuses are not a means to “achieve or protect power, but a policy consequence of [the leader’s] … core values.”51 For example, shortly before his arrest, Nuon Chea argued: But I am not a killer. Only those who wanted to destroy us and make problems for the Khmer Rouge were smashed. Those bad cadre and traitors killed our people and we killed only those who killed our people. Why should we keep them? If we did not smash the enemy there would be no Cambodia today.52 Genocidal leadership often combines these two, non-­mutually exclusive, elements: the decision to commit genocide may be a “rational” response to a perceived threat, yet also be grounded in prejudicial beliefs. Leaders act as “norm entrepreneurs” in shifting beliefs and expectations regarding the use of violence. New leaders often also emerge in genocide as individuals demonstrate a special capability. Local leaders are important to the implementation of genocide, often determining individual membership in the group. Leaders are also often present at killing sites, conveying the normativity of genocidal killing even more clearly.

Pathways to perpetration The intent to destroy the group requires a level of moral disengagement. The state actively encourages this by instituting harm-­producing relations of power, as well as disseminating discourses that emphasize the existential threat posed by the victim group. The genocidal state facilitates acts of aggression and harm, although even these acts are often constrained by certain norms, namely sanctioned forms of violence. State power opens an opportunity space for such morally-­disengaged acts of aggression. Moreover, the state seeks to increase moral disengagement by creating institutions that structure and externalize choice. The entire genocidal state may promote moral disengagement by rendering harmful acts imperative. The genocidal state often recruits violence specialists from among individuals who are already prone to moral disengagement. These individuals may, in some cases, have a psychological abnormality. They may also have experienced violence, whether as a victim or a perpetrator. Those who have been victimized may also be motivated to seek vengeance for their prior victimization. Those who have already engaged in violence may be habituated to its use. The state also produces social states of moral disengagement through its normative messages and institutional structures. Moral habits are formed partly through violence, as identity evolves in tandem with participation in violence. In this section, we will consider propensity as a product of moral disengagement and proximity, arguing that moral disengagement contributes to reduced

102   Who kills? self-­control as criminogenic, state-­derived, moral rules are internalized by perpetrators. However, not all perpetrators are morally disengaged. Some will perpetrate simply as a means of maintaining ties to the political community. Proximity, understood as the social and spatial distance of the individual to perpetrators and victims before and during the genocide, also affects propensity. This proximity involves degrees of intimacy with individuals and organizations involved in genocide, as well as a lack of intimacy with the designated victims of genocide. Thus, proximity also has interactive effects with moral disengagement. Rather than categorizing individuals in terms of persistent characteristics, it is more useful to think of pathways to perpetration: (1) (2) (3) (4) (5)

Pathological Ideological Entrepreneurial Conformist Compromised

These are motives for individual involvement in violence, ordered in terms of moral disengagement (with the most morally disengaged pathway being pathological). These individual pathways are grounded partly in propensity, but individual pathways also change according to circumstances. Through the act of killing, even those following conformist and compromised pathways may become increasingly morally disengaged. Drawing from Smeulers, the pathway to perpetration may also relate to individual orientation towards criminality and deviancy prior to genocide. Those already involved in crime are more likely to be motivated by pathologies; law-­abiding citizens, are more likely to perpetrate for conventional or ideological reasons (or as coerced, compromised perpetrators); finally, borderline individuals, those who are deviant, socially-­alienated, and maladapted, but not criminal, may perpetrate for entrepreneurial or ideological reasons.53 Consider first, pathological acts of perpetration that are grounded in the enjoyment of violence itself. These may be committed by individuals predisposed to such acts, yet even perpetrators who are not initially pathological may commit pathological acts of perpetration. Genocide represents an opportunity space for the acting out of violence. Second, perpetration may also be motivated by ideology. The killing of individual victims is seen as entirely derived from their essential negative essence, as stipulated in ideological discourses. Thus, ideologically motivated perpetrators are also willing to harm victims through their dedication to furthering ideological objectives. Third, some acts of perpetration are primarily entrepreneurial, driven by a desire for self-­enrichment. Where these entrepreneurial acts of perpetration extend beyond looting to killing victims for their goods, then they are rooted also in the perpetrator’s moral disengagement from the victim.

Who kills?   103 Fourth, some acts of perpetration are motivated by conformity: alignment with state power and perceived mainstream norms. Conformist and compromised pathways to perpetration may not be grounded in much moral disengagement at all, but rather in proximity to the state and other perpetrators. Finally, perpetration may be driven by victimization. Criminologists Alette Smeulers and Barbora Hola call these perpetrators “compromised” – they are forced to commit acts of perpetration because of their vulnerable position (e.g., child soldiers, members of the victimized group).54 In a sense, these five pathways are also ranked according to perpetrators’ projected levels of cognitive dissonance. It is expected that pathological perpetration involves little value conflict, as pre-­existing values already justified the use of violence. Ideological perpetration, by contrast, may involve some conflict between ideological goals and empathy towards victims. Conformist perpetration may incur the greatest degree of moral conflict as these acts draw individuals away from “universal values,” even when they are based on fulfillment of other moral obligations, namely loyalty and deference to the political community. Many génocidaires may have the intention to kill or victimize the person standing in front of them, while not necessarily possessing an intent to destroy the group. This is because of the diversity of individual motives that generate intent. Most génocidaires are also not truly “reactive” offenders, where individuals offend because they fail to control emotional responses, but rather are “instrumental” offenders, committing violence in premeditated and systematic ways. However, propaganda may work to reframe instrumental violence as reactive. Let us consider these categories more closely, keeping in mind that they are not mutually exclusive; individuals may occupy multiple categories, whether simultaneously or over time. Pathological perpetration Psychological factors Some acts of perpetration are rooted in the enjoyment of violence itself. Such acts may often be committed by “high moral disengagers,” as Albert Bandura argues, who are “more readily angered and behave more injuriously than those who apply moral self-­sanctions for detrimental conduct.”55 These individuals are also less likely to feel guilty after engaging in behavior that harms others.56 For some individual killers, violence may actually be an end, even if the group’s ideology itself is more focused on genocide as a means to achieve worthy goals. Not all evil is banal. Some perpetrators choose to kill because of a profound hatred of the victim group or a desire to inflict pain on other human beings. One Rwandan perpetrator confessed that “I supported the Hutu power movement because it called for the killing of Tutsis.”57 Philip Zimbardo found in his Stanford Prison Experiment that about one third of the “prison guards” (perpetrators, in a sense) were “cruel and tough,” less than 20 percent were “good,” and the

104   Who kills? remainder (around 50 percent) were rule-­oriented.58 This may roughly coincide with the breakdown for genocide perpetrators (vanguard perpetrators are a minority as are rescuers, while the majority of individuals are bystanders or reluctant perpetrators). Browder also concludes that “reluctant executioners” form a majority among perpetrators.59 Psychologist Stephen Baum estimates that 15–20 percent of individuals are perpetrators, 60–70 percent are bystanders, and 15–20 percent are rescuers and calls this proportion the “Bell Curve of Hate.”60 Baum also argues that those with low social development (maturation) are more likely to be perpetrators. Moral disengagement may relate, in some cases, to mental illness or abnormal socialization. However, contrary to popular perception, most génocidaires do not suffer from psychological abnormalities, although sadism and psychopathy may be disproportionately represented among vanguard perpetrators.61 There may be some perpetrators who are morally disengaged due to psychological or physiological conditions which contribute to reduced empathy or impulse control. For example, individuals with personality disorders such as psychopathy, borderline, narcissism, and schizophrenia may exhibit atypical responses to the suffering of others, as may individuals with conduct disorder, schizophrenia, bipolar disorder, and dissociative disorder.62 There are also studies that link aggressive behavior to psychological characteristics such as narcissism and low self-­esteem.63 These conditions are too rare to explain the mass perpetration of genocide. Nonetheless, pathologies may contribute to perpetration by certain individuals, who may be generally morally disengaged even before the onset of genocidal conditions. Sadists need to experience power over others (sometimes through cruelty). To sadists, living beings become objects of control. The tendency for over-­control is fundamental to all harm-­producing relations of power and all perpetration in genocide; even non-­sadistic individuals may come to behave sadistically (see the Sadistic Shift on pp. 208–209). One apparently sadistic “injector” who worked as an assistant to a doctor in a concentration camp embodied this urge when he stated: I had an interesting dream – a dream that after my death I would live in a special place where I would rule, where I would be sent thousands of people to kill personally with phenol injections. It was a wonderful dream.64 Sadists require a response from their victims, or they will escalate violence – they greatly fear a loss of control. Bullies display a similar relationship with their victims. A 2008 study at the University of Chicago indicated that youths who are bullies experience a positive physiological response when they watch others suffer.65 Josef Stalin was a classic sadist; he enjoyed exercising arbitrary and total control over others, and ensured that they always knew he was in control.66 Many individuals possess some sadistic tendencies, but these are usually balanced by other traits.67 Psychopaths are people who are self-­obsessed, disconnected from others, lack self-­reflection, and, in some cases, are asocial, aggressive, and highly impulsive.68

Who kills?   105 General studies of crime show that violent recidivism is much more common among psychopaths than among other individuals.69 Psychopathy is intimately linked to moral disengagement as “psychopathic individuals present a breakdown in moral socialization and impairment in the affective system.”70 Psychopaths lack empathy, and have difficulty forming lasting bonds with others. For example, a Bosnian perpetrator recounted strangling several prisoners of war because it was “too much of a hassle” to hand prisoners over to the Red Cross.71 Most genocide perpetrators are not psychopaths, but psychopaths might be overrepresented among perpetrators (there is a paucity of research in this area). Kren and Rappoport estimate that only 10 percent of SS men working in concentration camps could be diagnosed as psychologically abnormal.72 The fact that few genocide perpetrators have severe psychological problems is evinced by the normal lives that many of them lead after perpetration.73 A possibility exists that the conditions present during genocide may increase the prevalence of psychopathy and sadism.74 In his study of former SS personnel, psychologist John Steiner argued that some people are “sleepers” with certain dispositions and characteristics that lie dormant until activated by circumstances.75 Thus, the sadistic dispositions of sadists may be latent until they are provided with the opportunity to act sadistically. American serial killer Arthur Shawcross had the opportunity during the Vietnam War to torture, kill, and mutilate; he later said: “I was never happier.”76 In all forms of mass violence, anti-­social elements emerge to engage in opportunistic violence such as killing, rape, theft, and kidnapping. Beyond psychological disorders, we must also consider the influence of broader processes of socialization. Perpetrators arrive at the moment of perpetration with all the “baggage” of their experience and socialization. Such factors may impact upon the perpetrator’s interpretation of the situation and her perceived range of action alternatives. Psychologists argue that there is a relationship between an individual’s upbringing and her propensity for aggression. Abnormal socialization and development processes (usually originating in early childhood) can cause personality disorders and dispositions thereto (i.e., latent personality disorders).77 Punishment (especially corporal punishment), parental rejection, hostility, and violence in the home all increase the likelihood of aggression. Moreover, studies show that physical punishment and the withdrawal of affection decrease children’s concern for the welfare of others.78 Slobodan Milosevic had a history of depression in his family, and a severely disrupted childhood – both his parents committed suicide. A high percentage of adult criminals were abused as children.79 Some psychologists have argued that Hitler’s supposed abuse as a child accounted for his highly neurotic behavior.80 In some circumstances, over-­ permissiveness (an absence of consequences for bad behavior) may also be associated with aggression as it constitutes a sort of positive reinforcement.81 Attachment theorists, such as John Bowlby, argue that children separated from the affection of their mothers in their early years have difficulty bonding with others and typically lack empathy.82 Social learning theories of criminology

106   Who kills? assert that we seek out role models and model our behavior on them; this latent behavior may then be triggered by events.83 Edwin Sutherland’s social ecology theory of criminology posits that, while social disorganization is criminogenic, crime is nevertheless learned from role models.84 Notorious Bosnian Serb militia leader Zeljko Ranatovic (Arkan) fled an abusive and alcoholic father at the age of nine, entering a life of juvenile delinquency and crime. In 1973, he was arrested for assault, but family connections led to the charges being dropped.85 Shortly afterwards, he was recruited into the secret police and began a new career harassing and killing Yugoslav political exiles throughout Europe. He also committed bank robberies during this period and was convicted of armed robbery in Belgium, The Netherlands, and Germany. He was on trial in Sweden, but escaped when members of his gang drew their weapons in court.86 In the late 1980s, he became head of the fan club for the football team Red Star Belgrade (Arkan’s pastry shop was opposite Red Star Stadium), and this group of football hooligans eventually formed the basis of the paramilitary group Arkan’s Tigers. Alvarez argues that it is not surprising that that Arkan’s Tigers were formed from a football fan club, as there is a connection between the phenomenon of sports hooliganism and such militia; both types of group are linked to identity, territory, and masculinity.87 Such transgressive communities may also generate increasingly radical behavior to demonstrate loyalty and toughness. People with depression and low self-­esteem may be attracted to violence as a means of relieving boredom.88 However, the usual psychological and psychoanalytical typologies of offenders (i.e., those with a weak ego or superego, neurotics, and so on) are of limited relevance here, given the extent of participation in genocide and the political nature of the crime. Theodore Adorno posited that there is an “authoritarian personality” – individuals who had a strict upbringing with dominant and disciplinarian fathers – that makes some people, and some countries, more susceptible to fascist ideologies.89 Evidence from interviews with SS members imprisoned for genocide and related crimes showed that many had unsatisfactory family relations with authoritarian fathers who practiced corporal punishment.90 Moreover, SS members tended to have a strong authority orientation and a preference for hierarchical systems.91 Similarly, Ervin Staub argues that many youth in post-­World War I Germany grew up in deprivation and instability, with absentee fathers, and that the Nazis fulfilled a need for an authoritarian father-­presence.92 Some have even argued that Adolf Hitler was himself the proxy authoritarian father, and that this paternal infallibility was formalized in the Führer principle.93 Jonathon Glover and Robert Jay Lifton both subscribe to this view, contending that “many Nazis had an authoritarian upbringing” and that the concept of “proper, controlled, relatively impersonal behavior … infuses German culture and character.”94 However, the theory of authoritarian personality has been roundly criticized as overly simplistic and deterministic.95 As Haritos-­Fatouros noted of torturers, most perpetrators “do not have to have a certain kind of personality, only exposure to certain kinds of psychological, social, and political conditions.”96

Who kills?   107 Genocide involves perpetrators from varied family backgrounds and social strata. Thus, the most important socialization factor for genocide may not be parenting at all, but the experience of perpetrators with violence and prejudice. According to Cornish and Clarke’s Reasoning Criminal Model, criminogenic backgrounds include: direct and vicarious experience of crime; self-­perception; conscience and moral attitudes; and foresight (planning).97 Negative experiences at the hands of the enemy group may lead the perpetrator to infer that all members of that group possess negative characteristics; previously victimized groups are overrepresented among genocide perpetrators.98 In Burundi, for example, seven out of 11 interviewees indicated that they had family killed by members of the opposite group (Hutus killed by Tutsis or Tutsis killed by Hutus) in prior episodes of violence.99 Michael Mann argues that violence is often perpetrated by men who have been socialized during an earlier phase of violence. For example, the early Nazi cadres were dominated by World War I veterans who formed paramilitary organizations such as the Freikorps and Wehruerband.100 Almost 30 percent of high-­ ranking Waffen SS officers had Freikorps experience, while many SA were former Wehreurband.101 The SA, however, was dominated by unemployed young men. Moreover, many radicals hailed from the so-­called lost territories in the east. A study of over 600 Nazi biographies showed that first-­generation Nazis came largely from anti-­Semitic and authoritarian backgrounds (focused on the restoration of German honor after World War I); second-­generation Nazis were more ideologically-­oriented and less focused on the wartime defeat; and a subset of second-­generation Nazis, the “marcher-­fighters” (those prone to street violence such as SA), often came from fatherless and/or abusive homes.102 Some génocidaires may also have learned impunity through the commission of violent acts that went unpunished.103 Biophysical factors We must also consider what role, if any, biophysical factors – such as genetics and medical conditions – may play in perpetration. As with psychological factors, the effects of biophysical factors are not relevant for the majority of perpetrators, who are perpetrating a political crime within a certain social context. Moreover, in criminology in general, the search for biological explanations for crime has proven largely fruitless. Certain individuals may have a greater genetic predisposition to aggression, but this is in no way determinative.104 Genes do not really code for behavior; the precise outcome of development cannot be predicted from starting conditions alone.105 The presence of brain injuries or abnormalities may, however, contribute to aggressive behavior.106 Cerebospinal disorders and certain viruses (e.g., the Borna virus) have also been linked to neuro-­behavioral and emotional problems, while other theories link aggression to factors such as an extra Y chromosome, defective genes, biochemical imbalances, low intelligence, brain chemical disorders, perinatal trauma, and low levels of the neurotransmitter serotonin.107

108   Who kills? Research in the field of perpetrator psychology and physiology is limited. But from broader research on aggression, we can posit that acts of violence are easier to commit for some individuals than for others. These perpetrators may already be morally disengaged from the victims of their acts. Such perpetrators may be particularly suited for egregious acts of cruelty. Moreover, it is difficult to isolate purely physiological sources of violence. Often there are interactive effects between environmental and biological explanations (for example, individuals may be socially isolated because of mental or physical defects).108 Genocide, as a mass crime, requires more than just pathological perpetration. Ideological perpetration Moral disengagement and perpetration may also be produced through adherence to ideologies that encourage hatred or fear towards the victims. As with domestic hate crimes, bias against the victim group is present in most perpetrators (about half of Rwandan perpetrators interviewed for this study said they supported the Hutu Power movement). However, it may not often be the primary motivating factor for violence.109 For some perpetrators, ideology may only function as a post-­facto justification. As noted previously, many genocidal leaders are ideological, while lower-­ level perpetrators may participate for more mundane reasons. The leaders of the Khmer Rouge, for example, were mostly well-­educated and ideologically motivated. However, even low-­level perpetrators operate within normative frameworks that are generated partly by ideological discourses and state policies. Ideology is essential to genocide. It ties individuals to negative cognitions, and it seeks to transform killing into an abstract act – a condition that is, in and of itself, conducive to moral disengagement. Both ideological and conformist acts of perpetration often require a push from state authorities, in the form of incitement, inducement, or coercion. Entrepreneurial perpetration Perpetration may also be driven by economic incentives. Around a third of Rwandan perpetrator interviewees admitted to having stolen from the victims.110 This theft may have been opportunistic, rather than a primary motive, yet it was a primary motive for some: I did not even want to kill all those people. I only wanted to take things, and I took a mattress because it was close to my house … I was afraid that guys would follow me asking why I was not killing others.111 However, the threat posed by the RPF to local elites may have motivated some to participate; one of the first policies implemented by the interim government was to pay the wages of the heads of cellules (who had not been paid for several months).112 Moreover, among more morally disengaged perpetrators, theft may sometimes motivate killing.

Who kills?   109 In the Rwandan genocide, material motives were arguably insufficient to prompt most perpetrators to begin killing. But for many perpetrators, they certainly determined the choice of target. One Rwandan interviewee recounted: “everybody rich was being attacked.”113 These same perpetrators may not have chosen to kill purely for material enrichment, but they likely did decide to take advantage of the opportunity space of genocide to enrich themselves. Consider the example of rioters who loot shops. There are those individuals willing to break a store window to obtain goods. Others would lack the courage to transgress norms by actively initiating a theft, but would still reach through a broken window to grab an iPhone. One perpetrator in Rwanda argues: “I took things because other people were taking things.”114 He adds: “it was said that wealthy people were RPF accomplices.” Another respondent stated, The wealthy people and the intellectuals were the first to be attacked in my community … they used to tell me that the wealthy people were the ones who supported the RPF. And the intellectuals would take power if the RPF came.115 Wealthy people were attacked first in many communities “because we thought if we attacked the poor, we could not get anything from the poor. But if we attacked rich people, we could get something from them.”116 These narratives are corroborated by Verwimp’s research, which shows that “members of households who owned large land properties” were more likely to be killed.117 Armies also often loot, generally without killing those who own the possessions. A soldier in the Burundi army remembers: “often we attacked because we needed to survive, we attacked farms to steal the cows, to steal the food. But we didn’t kill the guards of the farm if they didn’t resist.”118 The targets of violence remain marginalized victims. According to one perpetrator in Burundi, “the Sans Échec killed, they stole many things, they also killed often just for robbing the dead. But it was always the Hutu [who were victims].”119 Conformist perpetration Many acts of perpetration, both amateur and professional, are rooted in ties to conventional values. Perpetrators may not be fully morally disengaged, but participate in genocide regardless to align themselves with the social context. They may set aside their empathy for the victims out of obedience to state authority and loyalty towards the in-­group. Such acts of perpetration, within their enabling moral context, are prosocial according to the inner rationality and normativity of the perpetrator group. This is because genocide is a crime driven by the state or state-­like structures. Normally crime is a break from the social order, but this is not true of genocide. Travis Hirschi’s Social Bond theory argues that criminal behavior is less likely when strong social bonds exist. This is an overriding assumption of many criminological theories. Social bonds consist of attachment (caring about others);

110   Who kills? commitment (commitment to conventional behavior); involvement (time and energy devoted to participation in conventional activities); and belief in the moral validity of norms.120 In genocidal societies, however, social bonding may, in some cases, be criminogenic. Those most closely bound to the mainstream may be the most likely to commit genocide (as a crime of obedience). In genocide, attachment (caring about others) is transformed, through a narrowing of awareness, into caring about one’s own group (groupism). The greater one’s identification with the in-­group, the more likely it is that one will perpetrate to further the goals and bolster the security of the in-­group. Non-­participation may be seen as an act of disloyalty. Commitment and involvement in genocide is still commitment towards conventional behavior, but killing itself has become partly conventional. Commitment and involvement, then, are reflected in support for, participation in, or acquiescence to genocide. Involvement is also deepened by learning through doing (socialization to violence through exposure to violence). Finally, belief represents devotion to a new set of norms established by genocidal propaganda. Perpetrators in genocide are often not typical criminals. They may not be marginal members of society or individuals involved in anti-­social activities or deviant subcultures. Education, for example, provides no immunization to evil: many genocide perpetrators are well-­educated. One study of the SS found that two-­thirds held university degrees, and a third had doctorates (many in law).121 Moreover, only 16 percent had suffered the loss of a parent before the age of 19, while only 24 percent had their employment disrupted (by unemployment, business failure, or embezzlement charges).122 Many génocidaires are the “absolute antithesis of ‘the criminal’ as conventionally conceived.”123 They may be intelligent, focused on productive activities, and closely bonded with the mainstream. The proximity of perpetrators to other perpetrators, but also to the state, may be a factor in increasing propensity. Many perpetrators strive to maintain their social bonds to the political community before, during, and after genocide. The moral disengagement involved in conventional acts of perpetration may be limited compared with pathological or ideological perpetration. These individuals may be “going through the motions,” rather than fully committing themselves to acts which harm others, or focused on technical tasks. Accordingly, one would expect that conformist perpetration entails the highest levels of cognitive dissonance, as well as higher levels of “posturing” and other behavior that stops short of full participation in genocidal killing. Compromised perpetration Compromised perpetration differs from other categories in that it is rooted in victimization, rather than moral disengagement or fidelity to the existing social order. Compromised perpetrators are also victims. Perhaps they perpetrate due to coercion, or are members of the victim group who collaborate with perpetrators, or are victims who hide “in plain sight” by feigning membership in the perpetrator group. One of my interview subjects, for example, was a Tutsi who joined

Who kills?   111 the Interahamwe at the behest of his friend, an Interahamwe leader, to masquerade as Hutu, thereby preventing his own victimization.

Variable individual responses to genocide Action roles in genocide Beyond perpetration, individuals may assume many different roles in a genocidal context. Genocide is a time of great confusion for many. Old moral norms have been discarded, revised, or set aside in deference to new norms. This normative shift is accompanied by the social exclusion of specific groups, and the escalation of violence. Yet even as the state, in response to strain, seeks to provide certainty to its citizens many of those citizens may struggle to adapt their behavior to the new norms. Their interpretations of the situation may involve weighing competing obligations. The failure of some to adapt may be rooted in the persistence of old, universal norms of reciprocity that preclude harming others. Adjusting to the genocidal state is a process of interpretation, adaptation, and innovation. Individuals retain some scope for action, even in the face of the demands of the genocidal state. This range of options – the action alternatives – differs from person to person and society to society due to propensity, as well as positionality. Other than joining or otherwise endorsing the genocidal state, individuals may also choose to remain on the sidelines, or to actively resist violence. However, the social space available to diverge from community norms and demands shrinks in genocide. Pressures for conformity and obedience are great, and internal cohesion is constructed partly by rooting out in-­group members who are deemed disloyal, treacherous, or imposters (concealed members of the victim group). Perpetration in genocide is just one possibility among a range of roles available to individuals. This range is a matter of personal perception, even as it is linked to the broader context. Thus, it is not fixed, but changes in response to internal beliefs and external relationships. Once individuals become perpetrators, further acts of perpetration may carry a reduced moral cost, as well as spurring changes in self-­perception and self-­identity. To abandon a perpetrator role may be to acknowledge that the perpetration was morally wrong or unnecessary. Perpetration does not exist in a vacuum. Rather it is, in part, a response to acts and omissions by other actors. Experimental evidence also shows that the presence of observers who apparently approve of aggression may cause subjects to behave more aggressively.124 Morality is arguably always situational. Although we may adhere closely to deontological notions of right and wrong, we also treat others according to our relationship with them, as well as their position in the broader community. Analyses of roles in genocide similarly emphasize the relationships between the parties, as well as the broader moral context. Genocidal roles are fluid, responding to both internal and external influences and controls.

112   Who kills? Individuals may shift between roles based on opportunity (reacting to a situation), doubt (feeling remorse or uncertainty about perpetration), moral bandwagoning (where individuals change their own behavior to respond to shifting norms in society), or coercion. Given the nature of genocide as a state crime, perpetrators are often presented with opportunities for increased status and material goods. Opportunity may also, for example, motivate perpetrators to rescue others in exchange for payment. Doubt may lead individuals to cease perpetrating and become bystanders or rescuers. If perpetration involves negative feelings, negative results, or other forms of punishment or ostracism, this may contribute to perpetrator doubt. The shift between roles in genocide may also contain an element of randomness – individuals may be moved from a non-­perpetrating role to a perpetrating role through a job transfer, for example, or simply by being in the wrong place at the wrong time. This randomness, or “genocidal serendipity,” constitutes a recognition that, although the act of perpetration is intentional, the entry into the moment of perpetration may not be foreseen or deliberate. Rather than moving “cleanly” from one role to another, individuals may accumulate roles. By assuming a new role (e.g., perpetrators who begin rescuing), individuals may substantially alter their identity, security, and criminal culpability (rescuing could be considered a mitigating factor in sentencing, for example). It may also be quite difficult for individuals who shift roles to shift back. Becoming a perpetrator often involves transgressing fundamental moral and legal norms. One cannot simply undo acts of perpetration through other acts. Roles, can therefore accrete to individuals during and after genocide. Such role shifting engenders friction, in the form of additional risks and changed reward/ incentive structures, logistical challenges (changing one’s job or location, gathering provisions to give to victims, etc.,), moral tension (through becoming a perpetrator, for example), and, for new perpetrators, conformity costs (the difference between individual and group preferences). Most typologies of roles are overly static and deterministic, based on characteristics rather than actions. It is not entirely unusual for individuals to change roles; however, it may be that the repetitive commission of certain acts (e.g., killing) involves a hardening of roles. The act of perpetration often reframes the situation, as perpetrators rationalize and neutralize their actions to avoid cognitive dissonance. This contributes to a shift in positionality, even as perpetrators avoid self-­identifying as such. Individual identity in genocide is an organizing principle for violence. One becomes a victim simply through one’s membership in an identity group. Furthermore, perpetration is institutionalized, and these institutional structures create the power context that enables acts of victimization and perpetration, as well as recidivism. If an individual holds a position in an organization carrying out killing, then there are strong structures present that encourage recidivism. This situation differs, of course, for ordinary civilian perpetrators, though their action alternatives may also be constrained by institutional power.

Who kills?   113 Individuals may respond to the demands of authority in several different ways, including acquiescence (conformity with demands that the individual does not support), endorsement (conformity with demands that the individual supports), internalization (perceiving a demand as internally generated rather than externally imposed), and resistance (refusing a demand or avoiding its implementation). Norms, structures of control, and individual propensity act as intervening variables in the consideration of preferences. Acquiescence, endorsement, and internalization all lead to an individual becoming a perpetrator or a passive bystander. Acquiescence to the general context is at the heart of bystander behavior, while endorsement and internalization both entail some involvement in perpetration (even if it is merely a cognitive alignment with collective objectives). Perpetration itself can take different forms. There are direct perpetrators who commit acts of violence; accomplices, who aid direct perpetrators (e.g., “I never killed anybody … when the group wants to kill the Tutsi we went together”); encouragers, who incite others (e.g., “promoters told us where the cockroaches were and we had to go find them”); informers, who provide information on the location of victims; and leaders, who order and organize perpetrators.125 These roles appear broadly to follow gender lines. Many female perpetrators, for example, act as informers or accomplices rather than as direct perpetrators. The people who most eagerly support genocide are often those who feel more acutely threatened by the victim group, are closely bonded to the mainstream (state), or are already habituated to the use of violence.126 In the proximal context, perpetrators are those who kill or assault, while in the distant context perpetration involves ordering and organizing violence. Those not present at the scene of perpetration may also be beneficiaries of violence. They may expropriate property, or occupy spaces vacated by victims, such as job openings, political positions, and houses. By definition, bystanders ignore victims or fail to assist them. Depending on the level of mass participation, and perhaps their ability to effectuate genocide, bystanders may be seen by the regime as either irrelevant or a threat. One could argue that bystander behavior facilitates perpetrators’ actions: to borrow Ian Kershaw’s famous formulation, “The road to Auschwitz was built by hate, but paved with indifference.”127 Bystanders may also be pressured to become involved, or be victimized by perpetrators who view their inaction as a form of opposition. They may refrain from acting out of an absence of empathy, or because they fear the consequences of action. They may also drift into other roles, such as those of perpetrators or rescuers, during the course of genocide. Many individuals retreat into the private sphere during the rise of the genocidal state and the genocide itself. This privatization allows individuals to avoid involvement and moral conflicts. In her research, Damant recounts how some people in Westmünsterland began to avoid social contact with their Jewish neighbors during the Nazi period – not out of a desire to marginalize those individuals, but out of feelings of discomfort or moral conflict. Social relations often become a source of great stress during these periods, as individuals struggle to reconcile competing

114   Who kills? cognitions of appropriate behavior towards the victim group, particularly with regard to people who are personally known to the victims. Retreatism may also be a way of maintaining “normalcy” and ties to conventional values. There is a certain comfort in being willfully blind to oppression. But not all bystander passivity involves retreatism or privatization. Rather, many individuals seek to carry on their lives as usual. Political activism is not the normal state of affairs for most people. One might question whether the extreme context of genocide requires positionality – is it possible to remain neutral in any moral or actual sense? The range of action alternatives doubtlessly narrows during genocide, as the genocidal state requires greater involvement, and violence becomes more pervasive. The act of violence itself may not only require implementation, but also mass culpability, if it is to be rationalized. The propensity to rescue: rescuers as outliers Much of this work has been concerned with individual perpetrators of genocide. Yet to fully understand perpetration, it is informative to consider the other side of the coin. Why do some individuals resist genocide while others do not? Are there unique characteristics or factors that impel individuals to opt out of genocide, and instead rescue victims or challenge the legitimacy of the genocidal state? Or is the default reaction of individuals passivity and non-­intervention? Altruistic behavior in the genocidal culture seems to be exceptional. In fact, it may be considered anti-­social in the context of mass, sanctioned perpetration: “sometimes if they found you hiding a Tutsi in your house they obliged you to kill the Tutsi. And that is how I got involved in the violence.”128 Thus, such behavior cannot be driven by rational choice (as killing often is the rational choice); rather, it is more a function of self-­identity.129 A moderately positive self-­worth is associated with empathy, whereas low self-­esteem and extreme egotism are not.130 Rescuers tend to have a more expansive moral community – those to whom they feel morally obligated.131 Our ability to live up to our ideals reflects how deeply these ideals are integrated with our sense of self.132 Those with a “pro-­social” value orientation (a positive view of humanity, concern about others’ welfare, and a sense of personal responsibility) are more likely to help others.133 The most extreme contrast with these pro-­social traits can be found in sociopaths, who may display essentially the opposite characteristics (a negative view of humanity, a lack of empathy, and an absence of personal accountability). During the narrowing of awareness that characterizes genocide, potential rescuers, like perpetrators, respond differentially to the plight of victims, depending on the saliency of certain cues, such as the strength of the connection between rescuer and victim. Moreover, the well-­known experiments of Latané and Darley found that the more bystanders are present, the more responsibility is diffused, and the less likely that any one will intervene.134 The presence of multiple victims may not increase (and may even decrease) the likelihood of intervention.135 Bystanders are more likely to help if it is deemed socially appropriate to

Who kills?   115 do so. Conversely, bystanders will also be less likely to intervene if others seem unconcerned, if values conflict (for example, obedience to authority and the obligation to help those in need), or if the costs of intervention are high. Conflicting values are almost always present in genocide, as are high intervention costs, so the prospect of intervention is low. Not helping others may also bear certain external costs (social sanctions) or internal costs (guilt or shame). In the case of genocide, external costs are not a factor in perpetration (as genocide is normative for its society), while internal costs may be minimized by neutralization and structural factors. Who, then, acts as a rescuer? We can hypothesize that the propensity to rescue is inversely related to the propensity to perpetrate. In other words, rescuers are those who are more morally engaged with victims. This moral engagement may arise from removal from structures of conformity and cohesion, from personal ties to specific victims, or from moral identity. A Rwandan respondent notes, “I saved my brother’s wife because she was a member of the family.”137 Those removed from moral disengagement, i.e., structures of conformity and cohesion, may be disproportionately the marginal, the rebellious, or those who oppose the state or conventional values. We can also hypothesize that rescuers are those with strong competing norms that facilitate divergence from genocidal demands. Perhaps rescuer behavior is only possible among individuals whose internal controls are strong enough to withstand external influence. When such individuals participate in killing, they likely experience strong cognitive dissonance, despite moral neutralization. The altruistic do not allow group identity to overwhelm their personal identity despite the narrowing of awareness. Helping others may reinforce our self-­image and help to resolve moral inconsistencies. Perhaps this is also why some perpetrators choose to act as rescuers alongside their acts of murder. One Rwandan perpetrator even argued that he used his position as an Interahamwe to rescue, “there are so many people I helped, because after being a killer, I gained their [the Interahamwe’s] trust.”138 The line between rescuing and perpetration illustrates that not all rescuing is altruistic, at least not in the sense that it reflects a universal ethic. Rescuing is often dictated by personal ties, and perpetration may even be partly driven by the desire to protect specific individuals. For example, one Rwandan perpetrator interviewed for this study killed others in order that his colleagues would not suspect him of hiding his Tutsi girlfriend: 136

One day I did not go to Amarondo, and they came to my house to charge me a fine, and I paid a fine of 500. She was there inside my house – they did not think that the girl was inside my house. From that time, she obliged me to join others so that they could not know that she was hiding in my house.139 He may have perpetrated many times in order to rescue a single individual. In Rwanda, there were also acts of “rescue” in which the rescued individual was subsequently sexually assaulted.140

116   Who kills? It also seems, based on my interviews and other sources, that sometimes individuals will perpetrate in groups but rescue when alone. This may occur because sole perpetrators are removed from the social pressures present in the group. Rescue, then, can also be an act of opportunism. Acts of rescue may function as a kind of deviance within the genocidal state, an exception to “normal” values. A Bosnian Serb general (Bosnian Army) interviewed for this book argued, “To be good, to do something good in bad times, it takes more. It takes courage.”141

Controls Situational action theory posits that internal and external controls affect how a person chooses among action alternatives, when at least one alternative involves breaching a rule of conduct. However, controls are only relevant in cases where there is a “breach of moral rule guidance”; in other words, where there is a conflict between moral rules.142 The theory distinguishes between self-­control (internal controls) and deterrence (external controls). We will consider here the role of these forms of control in genocide, particularly considering the mechanisms of moral disengagement discussed earlier. In genocide, external controls in the form of deterrence are of limited relevance, as the state itself seeks to produce crime. External controls are not present in the genocidal state; there is no authority to interdict and punish crimes, and to protect victims. External control, in such circumstances, can only be provided by authorities other than the state – namely international or supranational bodies with the capacity to intervene. But, as we note in Chapter 11 deterrence is also predicated on the certainty of swift punishment – a condition that does not exist with regard to genocide. The “Global Community” is at best a weak system of social control. With genocide, as with ordinary crimes, self-­control involves adhering to prior moral rules, and abstaining and desisting from crime. However, self-­control in genocide and adherence to moral values may actually involve breaching the law, or, at least, contradicting legitimate authorities. Self-­control depends on individual propensity, but is also affected by environmental stresses, including peer pressure and intoxication. The ability to exercise self-­control is also conditional on moral beliefs and moral identity. However, moral rules in genocide are dictated by state power. In fact, exposure (the moral context) is much more powerful than in normal crimes. This is because the state has resources of power and legitimacy far exceeding those of individuals or deviant subcultures. Moreover, in the case of habitual action (moral habits), individuals may perceive only one causally effective alternative. Consequently, controls have limited relevance because there is no moral rule guidance. One would expect, therefore, that individuals who are less heavily socialized (perhaps especially conformist perpetrators and individuals with strong moral identities) may exhibit greater self-­control and a lower propensity to violence. A lack of self-­control has also been linked experimentally to moral disengagement.143 Self-­control, then, is eroded by moral disengagement, which is in turn socially produced by the state. Non-­conforming behavior is risky when evil is institutionalized and normalized.

Who kills?   117

Conclusion: distance and propensity Individual participation in genocide results from the complex interaction of contextual factors, experiential factors, and individual personality. Despite the intense situational factors present, the decision to participate in genocide remains a decision. This is demonstrated well by the different response of individuals in similar social circumstances. This chapter has argued that propensity is a function of the individual’s interaction with the context, particularly moral disengagement and proximity – that is, the perpetrator’s closeness to victims and to other perpetrators. Proximity to the state increases compliance, while moral distance from the victims, the humanity gap, reduces moral conflict. Relationships are of crucial importance to genocidal propensity. While much attention has been paid to perpetrators’ moral disengagement from victims, many may also perpetrate merely to maintain their relationships with peers and the wider community. Those perpetrators who have higher propensity have lower tipping points for participating in genocide. In many cases, they may even kill thoughtlessly, as the path of least resistance. 

Notes    1    2    3    4    5    6    7    8    9   10   11   12   13   14   15   16   17   18   19   20   21   22   23   24   25   26   27

Wikström and Svensson, “When does self-­control matter?,” 396. Wikström and Svensson, 396–397. Scott Straus, “How many perpetrators were there,” 95. Johnson and Reuband, What we knew, 369. Goldhagen, Hitler’s Willing Executioners, 337. John Mueller, “The Banality of ‘Ethnic War’,” International Security 25, no.  1 (2000), 49. Andreas, “The Clandestine Political Economy,” 35. Mueller, “The Banality of ‘Ethnic War’,” 50–51. Mueller, 52. Mann, The Dark Side of Democracy, 183. Flint and de Waal, Darfur: A Short History, 103–104. Weale, Army of Evil, 274. Prunier, The Rwanda Crisis, 231–232. Mann, 223. Interview Bu1 (Burundi), Bujumbura, September 2009. Interview Bu2 (Burundi), Bujumbura, September 2009. Interview Bu6 (Burundi), Bujumbura, September 2009. Interview Bu10 (Burundi), Bujumbura, September 2009. Interview Bu3 (Burundi), Bujumbura, September 2009. McDoom, “Who killed in Rwanda’s genocide?,” 461–462. Straus, 85–98. Williams, “The Complexity of Evil,” 74. Williams is citing Fujii (2008) and McDoom (2009). The proportion of males in Gacaca Court trials in Rwanda, for example, was 94.5 percent men for killing, and 89.2 percent for looting. Meng-­Try and Sim, Victims and Perpetrators?, 12. Interviews U06 and U07 (Uganda), Gulu, July 2009. Chandler, Voices from S-­21, 33. Jamieson, “Genocide and the Social Production of Immorality,” 144.

118   Who kills?   28 Rwanda Interview R69, Hindiro TIG Camp, October 2009.   29 She was posted at the Gross-­Rosen camp on the German-­Polish border. See: Lower, Hitler’s Furies, 15.   30 See, for example, Ulmer and Steffensmeier, “The age and crime relationship.”   31 Brehm, “Subnational Determinants,” 10.   32 Brehm, 29.   33 Brehm, 27.   34 Smeulers, “Female perpetrators”; see also: Hogg, “Women’s participation.”   35 Brehm, Uggen and Gasanabo, “Age, Gender, and the Crime of Crimes,” 28.   36 Mann, 463.   37 Nicole Rafter, The Crime of All Crimes: Towards a Criminology of Genocide (New York University Press, 2016), 162.   38 Rafter, 170.   39 Lower, 142.   40 As noted in Chapter 3, these estimates are crude and limited by a lack of comprehensive data. The estimate of the number of female participants is based only on camp guards, when considered out of a total adult female population of 29,621,141 at the time. My sources are Lower, 142; and the Census of Germany, May 17 1939, respectively.    41 Bandura, “A Murky Portrait of Human Cruelty,” 225.   42 Johnson, et al., “Association of Sexual Violence,” 553.   43 Lower, 133 and 155.   44 Jessee, “Rwandan Women No More,” 73; Hogg, 81–83.   45 Gerlach similarly wrote of “coalitions for violence.” See: Gerlach, “Extremely Violent Societies,” 458–467.   46 See, for example, Edina Bećirević’s account of the attack on Zvornik in Genocide on the Drina River, 91.   47 Langerbein, Hitler’s Death Squads, 182.   48 Klee, Dressen and Riess, 129.   49 Interview R39 (Rwanda), Nyarusenge TIG Camp, August 2009.   50 Mitchell, Agents of Atrocity, 32.   51 Mitchell, 41.   52 Chon and Thet, Behind the Killing Fields, 161–162.   53 Smeulers, “Perpetrators of International Crimes,” 242.   54 Smeulers and Hola, “Criminology Discovers International Criminal Law,” 10.   55 Bandura, et al., “Mechanisms of Moral Disengagement,” 371.   56 Bandura, et al., 371.   57 Rwandan Perpetrator Interview R67, farmer, Hindiro TIG Camp, October 25, 2009.   58 Browning, Ordinary Men, 171.   59 Browder, “Perpetrator Character and Motivation,” 495.   60 Baum, The Psychology of Genocide, 105.   61 See, for example, Kren and Rappoport, The Holocaust and the Crisis of Human Behavior.   62 See, for example: Cox, et al., “The balance between feeling and knowing; and Decety, et al., “Atypical empathic responses.”   63 Smith and Mackie, Social Psychology, 480.   64 Lifton, The Nazi Doctors, 264.   65 University of Chicago (press release), “Bullies may Enjoy Seeing Others in Pain,” November 7, 2008.   66 Fromm, The Anatomy of Human Destructiveness, 321.   67 Fromm, 324.   68 American Psychiatric Association, Diagnostic and statistical manual of mental disorders.   69 Andrews and Bonta, The Psychology of Criminal Conduct, 219.

Who kills?   119   70   71   72   73   74   75   76   77   78   79   80

De Aguirre, “Neurobiological bases of aggression,” 228. Interview B05 (Bosnia), Tuzla, April 2009. Alvarez, “Adjusting to Genocide,” 156. Alvarez, Governments, Citizens, and Genocide, 22. Weber, Amunts and Schneider, “Structural Brain Abnormalities,” 25. Staub, The Roots of Evil, 133. Baumeister, Evil: Inside Human Violence, 224–225. Lanier and Henry, Essential Criminology, 129–130. Staub, The Roots of Evil, 72. Staub, The Psychology of Good and Evil, 200. Kressel, Mass Hate, 110–111. Other studies of Hitler such as Alice Miller’s examination in For Your Own Good argue that Hitler came from an abusive background and that he was a sadist with anal-­retentive impulses.   81 Staub, The Psychology of Good and Evil, 201.   82 Lanier and Henry, 134.   83 Lanier and Henry, 144.   84 See, for example: Sutherland, Principles of Criminology.   85 Mann, 404.   86 Alvarez, Genocidal Crimes, 80.   87 Alvarez, Genocidal Crimes, 107.   88 Fromm, 278.   89 Glover, Humanity, 330–331. His discussion is based on Adorno, et al., The Authoritarian Personality.   90 Staub, The Roots of Evil, 111.   91 Staub, The Psychology of Good and Evil, 301.   92 Staub, The Roots of Evil, 114   93 Miller, For Your Own Good.   94 Glover, 331; Lifton, 392.   95 See, for example: Martin, “The Authoritarian Personality,” 1–26.   96 Haritos-­Fatouros, “Cruelty: A Dispositional or Situational Behavior,” 230.   97 Lanier and Henry, 91.   98 See Mann; see also Midlarsky, The Killing Trap.   99 An additional four interviewees responded “don’t know” to the question. 100 Mann, 195. 101 Mann, 195. 102 Kressel, 128. 103 Midlarsky, 43. 104 Staub, The Roots of Evil, 35. 105 Bateson, “Is Aggression Instinctive?” in Groebel and Hinde, Aggression and War, 44. 106 Herbert, “The Physiology of Aggression” in Groebel and Hinde, 62–69. See also: Eslinger, “Neurological and Neuropsychological Bases of Empathy,” 193–199; and Stein, “The Neurobiology of Evil,” 306. See also: Kluger, “What Makes Us Moral,” 36. 107 Ellis, “Biological Factors,” 27. See also: Lanier and Henry, 102. 108 Brookman, Understanding Homicide, 67–68. 109 McDevit, “Hate Crime Offenders,” 307. 110 Interview R55 (Rwanda), Rugerero TIG Camp, September 2009. 111 Interview R31 (Rwanda), Butare, August 2009. 112 Verwimp, “An economic profile of peasant perpetrators,” 321. 113 Interview R8 (Rwanda), Kigali Central Prison, July 2009. 114 Interview R34 (Rwanda), Rwaza TIG Camp, August 2009. 115 Interview R37 (Rwanda), Rwaza TIG Camp, August 2009. 116 Interview R55 (Rwanda), Rugerero TIG Camp, September 2009. 117 Verwimp, “Testing the Double-­Genocide,” 438. 118 Interview Bu9 (Burundi), Bujumbura, September 2009.

120   Who kills? 119 120 121 122 123 124 125

Interview Bu10 (Burundi), Bujumbura, September 2009. Hirschi, Causes of Delinquency. Herbert, National Socialist Extermination Policies. Mann, 222–223. Friedrichs, “The Crime of the Century?,” 29. Horowitz, The Deadly Ethnic Riot, 352. Interview R34 (Rwanda), Rwaza TIG Camp, August 2009; Interview R3 (Rwanda), Nyarusenge TIG Camp, August 2009. 126 Mann, 8–9. See also: Hoffer, The True Believer, xiii and 47–59. 127 Kershaw, Popular Opinion and Political Dissent, xxi. 128 Interview R67 (Rwanda), Hindiro TIG Camp, October 2009. 129 Monroe, The Hand of Compassion, xi. 130 Staub, The Psychology of Good and Evil, 358. 131 Monroe, 229. 132 Monroe, 223. 133 Staub, The Psychology of Good and Evil, 16. 134 Staub, The Psychology of Good and Evil, 74. 135 Miller, Cushman and Hannikainen, “Bad Actions or Bad Outcomes?,” 585. 136 Staub, The Psychology of Good and Evil, 125–130. 137 Interview R29 (Rwanda), Rugerero TIG Camp, August 2009. 138 Interview R02 (Rwanda), Kigali Central Prison, July 2009. 139 Interview R62 (Rwanda), Mont Kigali TIG Camp, October 2009. 140 Jessee, Negotiating Genocide, 173; Mueller, 56. 141 Interview B10 (Bosnia), Sarajevo, April 2009. 142 Wikström and Svensson, 397. 143 Gabbiadini, et al., “Interactive Effect of Moral Disengagement,” 456.

Works cited Adorno, T. W., Else Frenkel-­Brunswik, Daniel J. Levinson and R. Nevitt Sanford. The Authoritarian Personality. Norton: New York, 1950. Alvarez, Alexander. “Adjusting to Genocide: The Techniques of Neutralization and the Holocaust.” Social Science History 21, no. 2 (Summer 1997): 139–178. Alvarez, Alexander. Genocidal Crimes. New York: Routledge, 2010. Alvarez, Alexander. Governments, Citizens, and Genocide: A Comparative and Interdisciplinary Approach. Bloomington, USA: Indiana University Press, 2001. American Psychiatric Association. Diagnostic and statistical manual of mental disorders: DSM IV-­TR. Washington, DC: American Psychiatric Association, 2000. Andreas, Peter. “The Clandestine Political Economy of War and Peace in Bosnia.” International Studies Quarterly, 48, no. 1 (2004). Andrews, D.A. and James Bonta. The Psychology of Criminal Conduct. Cincinnati: Anderson Publishing, 1994. Bandura, Albert, Claudio Barbaranelli, Gian Vittorio Caprara, and Concetta Pastorelli. “Mechanisms of Moral Disengagement in the Exercise of Moral Agency.” Journal of Personality and Social Psychology 71, no. 2, (1996). Bandura, Albert. “A Murky Portrait of Human Cruelty.” Behavioral and Brain Sciences 29, no. 3 (2006). Baum, Stephen K. The Psychology of Genocide: Perpetrators, Bystanders, and Rescuers. London: Cambridge University Press, 2008. Baumeister, Roy F. Evil: Inside Human Violence and Cruelty. New York: W.  H. Freeman, 1997.

Who kills?   121 Bećirević, Edina. Genocide on the Drina River. New Haven, CT: Yale University Press, 2014. Brehm, Hollie Nyseth. “Subnational Determinants of Genocidal Killing in Rwanda.” Criminology 55, no. 1 (February 2017). Brehm, Hollie Nyseth, Christopher Uggen, and Jean-­Damascene Gasanabo. “Age, Gender, and the Crime of Crimes: Toward a Life-­Course Theory of Genocide Participation.” Criminology 54, no. 4 (2016). Brookman, Fiona. Understanding Homicide. London: Sage, 2005. Browder, C. G. “Perpetrator Character and Motivation: An Emerging Consensus?” Holocaust and Genocide Studies, 17, no. 3 (2003). Browning, Christopher. Ordinary Men: Reserve Battalion 101 and the Final Solution. New York: Harper, 1998. Chandler, David. Voices from S-­21: Terror and History in Pol Pot’s Secret Prison. Chiang Mai, Thailand: Silkworm Books, 2000. Chon, Gina and Sambath Thet. Behind the Killing Fields: A Khmer Rouge Leader and One of His Victims. Philadelphia: University of Pennsylvania Press, 2010. Cox, C. L., L. Q. Uddin, A. Di Martino, F. X. Castellanos, M. P. Milham, and C. Kelly. “The balance between feeling and knowing: affective and cognitive empathy are reflected in the brain’s intrinsic functional dynamics.” Social Cognitive and Affective Neuroscience 7, no. 6 (August 2012). De Aguirre, María Inés. “Neurobiological bases of aggression, violence, and cruelty.” Behavioral and Brain Sciences 29, no. 3 (2006). Decety, J., K. J. Michalska, Y. Akitsuki, and B. Lahey. “Atypical empathic responses in adolescents with aggressive conduct disorder: a functional MRI investigation.” Biological Psychology 80, no. 2 (2008). Ellis, Lee. “Biological Factors Associated with Criminal/Antisocial Behavior.” In Encyclopedia of Criminology and Deviant Behavior, edited by Clifton D. Bryant. Philadelphia: Brunner-­Routledge, 2001. Eslinger, Paul. “Neurological and Neuropsychological Bases of Empathy.” European Neurology 39, no. 4 (1998). Flint, Julie and Alex de Waal. Darfur: A Short History of a Long War. New York: Zed Books, 2005. Friedrichs, David O. “The Crime of the Century? The Case for the Holocaust.” Crime, Law and Social Change, 34, no. 1 (2001). Fromm, Erich. The Anatomy of Human Destructiveness. Greenwich, USA: Fawcett, 1979. Fujii, Lee Ann. “The Power of Local Ties: Popular Participation in the Rwandan Genocide.” Security Studies 17, no. 3 (2008). Gabbiadini, Alessandro, Paolo Riva, Luca Andrighetto, Chiara Volpato, and Brad J. Bushman. “Interactive Effect of Moral Disengagement and Violent Video Games on Self-­Control, Cheating, and Aggression.” Social Psychological and Personality Science 5, no. 4 (2014). Gerlach, Christian. “Extremely Violent Societies: An Alternative to the Concept of Genocide, Journal of Genocide Research 8, no. 4 (2010). Glover, Jonathan. Humanity: A Moral History of the 20th Century. New Haven: Yale, 2001. Goldhagen, Daniel. Hitler’s Willing Executioners. New York: First Vintage Books, 2001. Groebel, Jo and Robert A. Hinde (eds.). Aggression and War: Their Biological and Social Bases. New York: Cambridge University Press, 1993. Haritos-­Fatouros, Mika. “Cruelty: A Dispositional or Situational Behavior in Man?” Behavioral and Brain Sciences 29, no. 3 (2006). Herbert, Ulrich. National Socialist Extermination Policies. New York: Berghahn Books, 2000.

122   Who kills? Hirschi, Travis. Causes of Delinquency. Berkeley: University of California, 1969. Hoffer, Eric. The True Believer: Thoughts on the Nature of Mass Movements. New York: Harper Perennial Modern Classics, 2002. Hogg, Nicole. “Women’s participation in the Rwandan genocide: mothers or monsters?” International Review of the Red Cross 92, no. 877 (March 2010).  Horowitz, Donald. The Deadly Ethnic Riot. New Delhi: Oxford University Press, 2002. Jamieson, Ruth. “Genocide and the Social Production of Immorality.” Theoretical Criminology, 3, no. 2 (1999). Jessee, Erin. “Rwandan Women No More: Female Genocidaires in the Aftermath of the 1994 Rwandan Genocide.” Conflict and Society: Advances in Research, 1, no. 1 (2015). Jessee, Erin. Negotiating Genocide in Rwanda: The Politics of History. London, Palgrave Macmillan, 2017. Johnson, Eric and Karl-­Heinz Reuband. What we knew: Terror, mass murder, and everyday life in Nazi Germany. London: John Murray, 2005. Johnson, Kristen, Jennifer Scott, Bigy Ruhita, Michael Kisielewski, Jane Asher, Ricardo Ong, and Lynn Lawry. “Association of Sexual Violence and Human Rights Violations with Physical and Mental Health in the Eastern Democratic Republic of Congo.” Journal of the American Medical Association 304, no. 5 (August 2010). Kershaw, Ian. Popular Opinion and Political Dissent in the Third Reich: Bavaria 1933–1945. London: Oxford University Press, 1983. Klee, Ernst, Willi Dressen, and Volker Riess (eds.). The Good Old Days: The Holocaust as Seen by its Perpetrators and Bystanders. Old Saybrook, USA: Konecky and Konecky, 1996. Kluger, Jeffrey. “What Makes Us Moral,” Time, December 3, 2007. Kren, George M., and Leon Rappoport. The Holocaust and the Crisis of Human Behavior. New York: Holmes & Meier, 1994. Kressel, Neil J. Mass Hate: The Global Rise of Genocide and Terror. New York: Westview Press, 2002. Langerbein, Helmut. Hitler’s Death Squads: The Logic of Mass Murder. College Station: Texas A & M University Press, 2004. Lanier, Mark and Stuart Henry. Essential Criminology. Boulder, USA: Westview Press, 2004. Lifton, Robert Jay. The Nazi Doctors. New York: Basic Books, 2000. Lower, Wendy. Hitler’s Furies: German Women in the Nazi Killing Fields. London: Vintage, 2014. Mann, Michael. The Dark Side of Democracy: Explaining Ethnic Cleansing. New York: Cambridge Books, 2005. Martin, John Levi. “The Authoritarian Personality, 50 Years Later: What Lessons Are There for Political Psychology?” Political Psychology 22, no. 1 (March 2001). McDevit, Jack. “Hate Crime Offenders: An Expanded Typology.” Journal of Social Issues 58, no. 2 (2002). McDoom, Omar Shahabudin. “Who killed in Rwanda’s genocide? Micro-­space, social influence, and individual participation in intergroup violence.” Journal of Peace Research 50, no. 4 (2013). McDoom, Omar Shahabudin. The Micro-­politics of Mass Violence: Authority, Opportunity, and Security in Rwanda’s Genocide. London: London School of Economics, 2009. Meng-­Try, Ea and Sorya Sim. Victims and Perpetrators? Testimony of Young Khmer Rouge Comrades. Phnom Penh, Cambodia: Documentation Center of Cambodia, 2001. Midlarsky, Manus. The Killing Trap: Genocide in the Twentieth Century. Cambridge, UK: Cambridge University Press, 2005.

Who kills?   123 Miller, Alice. For Your Own Good: Hidden Cruelty in Child Rearing and the Roots of Violence. New York: Farrar, Straus, Giroux, 1983. Miller, Ryan, Fiery A. Cushman, and Ivar A. Hannikainen. “Bad Actions or Bad Outcomes? Differentiating Affective Contributions to the Moral Condemnation of Harm.” Emotion 14, no. 3 (2014). Mitchell, Neil J. Agents of Atrocity: Leaders, Followers, and the Violation of Human Rights in Civil War. Houndsmills, UK: Palgrave Macmillan, 2004. Monroe, Kristen Renwick. The Hand of Compassion: Portraits of Moral Choice during the Holocaust. Princeton, USA: Princeton University Press, 2003. Mueller, John. “The Banality of ‘Ethnic War’.” International Security 25, no. 1 (2000). Prunier, Gérard. The Rwanda Crisis: History of a Genocide. New York: Columbia University Press, 1995. Smeulers, Alette, and Barbora Hola. “Criminology Discovers International Criminal Law.” Unpublished paper, February 15, 2009. Smeulers, Alette. “Female perpetrators – ordinary and extra-­ordinary women.” International Criminal Law Review 15, no. 2 (2015). Smeulers, Alette. “Perpetrators of International Crimes: Towards a Typology.” In Supranational Criminology: towards a criminology of international crimes, edited by A. Smeulers and R. Haveman. Antwerp: Intersentia, 2008. Smith, Elliot R., and Diane M. Mackie. Social Psychology (3rd edition). New York: Psychology Press, 2007. Staub, Ervin. The Psychology of Good and Evil: Why Children, Adults, and Groups Help and Harm Others. New York: Cambridge University Press, 2002. Staub, Ervin. The Roots of Evil. Cambridge: Cambridge University Press, 1989. Stein, Dan J. “The Neurobiology of Evil: Psychiatric Perspectives on Perpetrators.” Ethnicity and Health 5, no. 3 (2000). Straus, Scott. “How many perpetrators were there in the Rwandan genocide? An estimate.” Journal of Genocide Research 6, no. 1 (March 2004).  Sutherland, Edward. Principles of Criminology. Philadelphia: J. B. Lippincott, 1947. Ulmer, Jeffrey T., and Darrell Steffensmeier. “The age and crime relationship: Social variation, social explanations.” In The Nurture Versus Biosocial Debate in Criminology: On the Origins of Criminal Behavior and Criminality, edited by K. Beaver, B. Boutwell, and J. C. Barnes. Newbury Park: Sage, 2014. University of Chicago (press release). “Bullies may Enjoy Seeing Others in Pain: Brain Scans Show Disruption in Natural Empathetic Response.” November 7, 2008. Available at: https://news.uchicago.edu/article/2008/11/07/bullies-may-enjoy-seeing-otherspain-brain-scans-show-disruption-natural-empathet. Verwimp, Philip. “An economic profile of peasant perpetrators of genocide, Micro-­level evidence from Rwanda.” Journal of Development Economics, 77, no. 2 (2005). Verwimp, Philip. “Testing the Double-­Genocide Thesis for Central and Southern Rwanda.” Journal of Conflict Resolution, 47, no. 3 (2003). Weale, Adrian. Army of Evil: A History of the SS. New York: NAL Caliber, 2012. Weber, Sabrina, Katrin Amunts, and Frank Schneider. “Structural Brain Abnormalities in Psychopaths – A Review.” Behavioral Sciences and the Law, 26, no. 1 (2008). Wikström, Per-­Olof, and Robert Svensson. “When does self-­control matter? The interaction between morality and self-­control in crime causation.” European Journal of Criminology 7, no. 5 (September 2010). Williams, Timothy. “The Complexity of Evil: A Multi-­Faceted Approach to Genocide Perpetration.” Zeitschrifft für Friedens- und Konfliktforschung, 3, no. 1 (2014).

6 Deciding to kill

Introduction Genocide is the sum of innumerable acts of individual criminality and cruelty. While most of the perpetrators interviewed for this book claim that they did decide to participate, they argue that they did so in a situation of a fundamental lack of choice. For example, one Rwandan perpetrator explained that “when ordinary people receive orders from the authorities, they have no choice but to follow them.”1 Another perpetrator, this time in Burundi, took passivity (self-­objectification) to another level by simply stating “one is obliged to accept these things.”2 The moral context, the social conditions that underlie genocide, acts as a form of decision-­priming for perpetrators. Yet, the variability of individual responses demonstrates that perpetration is not fully determined by the context. Rather, perpetration is a product of the relationship between the individual and the context. To kill, perpetrators must first be presented with an opportunity space in which killing becomes a viable option. This opportunity space is provided by the state through its discursive and institutional structures. Moreover, the enactment of state policies marginalizing the victims communicates the devalued status of victims to the perpetrators, while also generating the power differential necessary for genocidal killing. Within this opportunity space, perpetrators are presented with a range of alternatives; these alternatives are themselves bounded by the context, as well as by the propensity of perpetrators. Furthermore, perpetration also requires the physical means to perpetrate and an immediate context conducive to perpetration – for example, an empowered perpetrator, within a supportive crowd, armed with a deadly weapon, facing a disempowered and symbolically appropriate victim. In ordinary crimes victim vulnerability may exist in highly specific contexts, whereas in genocide vulnerability is institutionalized and perpetration is systematized. This chapter will build on the discussion of exposure (the moral context) and propensity to focus on perpetrator decision-­making. It draws from situational action theory to disaggregate the context (exposure) and determine how exposure, and other situational factors, affect the perpetrator’s perceived range of options. Subsequently, it considers the ranking of action alternatives in relation to perpetrator values and interests, as well as the mobilization of killing. Finally, options in the moment of perpetration itself will be examined.

Deciding to kill   125 The perpetration of genocide occurs in the shadow of state power. Individuals who are not members of the victim group are often confronted with demands to participate in genocidal killing, or at least to ignore such killing. The decision to perpetrate in this context is often rendered as a kind of non-­decision. Yet those who become perpetrators are often maximizing their own benefits in the context of genocide. To some individuals perpetration is an opportunity, while to others it is a catastrophe. Crime is a product of propensity and exposure. Propensity is determined largely by the individual’s moral rules and moral habits, as well as her ability to exercise self-­control; yet we must also address the role of factors beyond moral rules in the commission of crime. In particular, we must consider the role of individual agency in perpetration. This chapter argues that perpetrator decision-­ making is a product of the individual’s perceived options, as filtered through their understanding of their environment.

Perpetrator opportunity Genocide involves setting aside conventional moral norms prohibiting killing. Thus, it requires an opportunity space to produce acts of harm against the victims. This opportunity space is provided largely through the harm-­producing relations of power created by the genocidal state. Perpetrator opportunity is a multifaceted phenomenon comprising factors such as impunity, capability, context, and the vulnerability of the victims. First, perpetrator opportunity is facilitated by apparent impunity. This impunity is derived largely from state messages, as well as the lack of negative sanctions for genocidal acts. In genocide, the rule of law may still be operational, but only in service of a restructured moral order. Second, the perpetrators must have the capability to commit genocide. This capability is rooted in the power differential between perpetrators and victims. The perpetration of genocide also requires a certain degree of efficiency; perpetrators must perpetrate effectively and repeatedly. Third, the opportunity to perpetrate is provided by a context of violence. This lends legitimation to perpetrator acts; even if perpetrators are motivated by self-­enrichment, their acts will still be part of a politicized campaign of violence. Individual motives can differ from collective ones, but still substantially contribute to the realization of collective goals. The prevalence of violence also provides perpetrators with a degree of anonymity, as many others are also killing or present in the same group of killers as the perpetrator. A context of dehumanization greatly facilitates genocide, designating victims as legitimate and even essential targets of violence. At some point, killing may become conventional, easing moral conflicts and facilitating bystander passivity. The conventionalization of killing is facilitated by normative justifications for killing (i.e., moral neutralization), but also by the involvement of non-­marginal individuals, such as state agents and “ordinary” civilians.

126   Deciding to kill Finally, opportunity arises from the vulnerability of victims. Routine activity theory posits that the presence of motivated offenders and suitable targets greatly increases the likelihood of crime.3 For our purposes, we can distinguish between general suitability and specific suitability. In genocide, all individual members of a marginalized group targeted for extermination are generally suitable as targets. One of the key functions of propaganda is to render victims symbolically attractive targets for killing. When we consider specific suitability, the suitability of the individual target at the moment of the crime, it is logical that more vulnerable victims are, the more likely they are to be deemed suitable targets for killing. The most common risk factors identified for victimization are exposure (physical visibility and accessibility of the victim), guardianship (“the effectiveness of those persons or objects who prevent violations from occurring”), and attractiveness (“the material or symbolic desirability of persons or targets to potential offenders”).4 The state motivates perpetrators, defines victims as suitable, and ensures that guardians do not interfere with perpetration. In genocide, all members of the targeted group have a high degree of symbolic desirability (attractiveness). They are viewed as dangerous vermin to be exterminated for the health of the body politic. Most victims of genocide are also exposed, as the power of the state is far-­reaching. This exposure, however, varies according to local conditions. In Rwanda with its population density and mass participation the exposure of victims was great. Victim vulnerability The vulnerability of the victims reduces the cost of perpetration. In his work on ethnic riots, Donald L. Horowitz argues that perpetrators are risk-­averse, only attacking after calculating target resistance and police response.5 This is equally true for genocide: if victims are strong and prepared to resist, this makes perpetration much more difficult, and even dangerous. Perhaps more importantly, if the victims were protected by authorities, then violence would be unlikely. A victim I interviewed in Rwanda recalled that her house was, at first, protected by a Rwandan soldier who was a close family friend. The Interahamwe avoided the house until the soldier left, at which point she was raped and her children were murdered. Individual victims must be vulnerable, along with the victim group as a whole; individual vulnerability is closely connected to the vulnerability of the group, especially during the narrowing of awareness that characterizes genocide. Victim vulnerability may be assessed on the basis of several factors. See Box 6.1: Box 6.1 Assessment of victim vulnerability (a) Will the victim resist? (b) Does the victim have the means to resist? (c) What is the prospect of impunity?

Deciding to kill   127 Evaluating whether the victim will resist requires an assessment of demeanor. Seemingly passive victims represent attractive targets to génocidaires, just as they do to bullies. Aggression in animals generally only occurs if there is an attractive target to attack (i.e., a subordinate target).6 This vulnerability is rooted in institutional structures. In Einsatzgruppen operations in Eastern Europe during the Holocaust, victims often outnumbered the perpetrators by as much as 50 to 1.7 In the Soviet Union in 1941, a force of 135,000 guards controlled a gulag of approximately 1.5 million prisoners, while in Cambodia 70,000 Khmer Rouge soldiers and 1,000 political cadres dominated eight million people.8 The Bosnian town of Višegrad was “substantially controlled for years” by Milan Lukić and 15 men; through atrocious violence they forced 14,500 Muslims to leave the town.9 Such a disparity is common in genocide. It reflects the vulnerability of victims, and their passivity (the difficulty of resisting “legitimate” and powerful authorities). The means of resistance may be assessed based on the size of the victim group (normally they constitute a minority), the weaponry they possess, and their prospects of outside support (often the only guardianship available to victims of genocide). Groups with affine populations in neighboring states are less vulnerable than isolated groups. Controls from within the state are lacking because of the criminogenic nature of the genocidal state. Victimization is not possible without victim vulnerability. Often victims of genocide seem quite passive, sometimes even literally digging their own graves. One Rwandan perpetrator recalled: “some of them [the victims] knew they would be killed and they just sat down and waited.”10 Victims may be passive out of fear or resignation. Genocide, by its very nature, is asymmetrical – the force arrayed against victims is overwhelming. Moreover, fear may have a paralyzing effect on victims –what psychologists call the fear paralysis reflex. Perpetrators often cultivate this fear. A Srebrenica survivor recounts: we were here … in the woods, with the Serb [paramilitaries] up on the other side. And the Serbs would call us by names and tell us “let me come over to you … I will scalp your head … we will rape all of your women.”11 Victims may also be passive because of their isolation; with no-­one to respond to their plight, they have no reason to hope. This can operate as a cycle: the lack of guardianship and external controls contributes to passivity, consequently feeding the vulnerability of victims. As one member of an Einsatzkommando related: “Many [victims] … came without resistance to the execution area. It is not as though they had any alternative.…”12 Victim passivity is fundamentally a product of the relationship of victims to other actors. Perpetrators instill fear and resignation in victims, while the failure of bystanders to intervene compounds these factors and underlies victim isolation and vulnerability. Like perpetration, victimization is a process of habituation. The power differential between perpetrator and victim sometimes allows acts of perpetration, such as rape, to occur without physically overpowering the victim.

128   Deciding to kill Even as perpetrators produce the power differential that contributes to victim passivity, they also interpret this passivity as indicative of victim inferiority. According to Franz Stangl, former commandant of the Sobibór and Treblinka death camps, his treatment of the victims: has nothing to do with hate. They were so weak; they allowed everything to happen – to be done to them. They were people with whom there was no common ground, no possibility of communication – that is how contempt is born. I could never understand how they could just give in as they did. Quite recently I read a book about lemmings, who every five or six years just wander into the sea and die; that made me think of Treblinka.13 Stangl thus generated a self-­fulfilling prophecy by defining the position of victims and drawing personally convenient inferences from it. The victims of genocide are profoundly alone, “socially dead,” and set apart from the rest of humanity. They are among the least socially powerful actors.14 Generally, victims of crimes and human rights violations must rely on the state for redress, but the nature of the genocidal state dictates that, without regime change, victims of genocide must seek redress from their personal and institutional victimizers.15 The perpetrator’s act of killing may also function as an act of definition; the very fact of the victim’s death is “proof ” of their degraded value, according to perpetrator logic. Finally, impunity makes victims more vulnerable, as do circumstances in which victimization is easily concealed. As one Rwandan perpetrator recounted, “I knew we would not be prosecuted, because we had the power.”16 Concealment, of course, is often embedded in and facilitated by the regime. The commission of other, non-­genocidal acts of violence (e.g., war) makes the concealment of genocidal violence far easier.

Decision-­making In a sense, every decision we make is the product of how we, as distinct individuals, respond to our environment. If decisions were purely objective, made solely based on the surrounding context irrespective of subjective, individual perceptions, then every individual would react in predictable ways to the same stimulus. Needless to say, this is not the case. Any decision-­making model for genocide perpetration must address this complex interaction between personal and environmental factors. The rationality of decision-­making is socially embedded and individually variable. Even individuals with identical social roles (e.g., police officers) will behave differently due to differences in individual propensity, as well as their immediate influences. Conscious decision-­making may be bypassed entirely by spontaneous actions and moral habits (including operant conditioning). Moreover, individual decision-­making is shaped by the information available to individuals in the moment of perpetration, which is, in turn, shaped by power.17 Propaganda also influences perpetrator decision-­making by clarifying the moral context. It is

Deciding to kill   129 impossible in practice to separate “rationality” from social norms. Rather, “individual rationality is a function of social norms.”18 Situational action theory, which conceptualizes crime as an interaction between propensity and exposure, focuses on the nexus between individual and environment. See Figure 6.1. This is in keeping with the perpetration model of this book. Propensity and exposure function as decision-­priming, preparing perpetrators to make genocidal decisions. The perpetrator then assesses the situation through various lenses such as his or her immediate situation and role in society, as well as the costs and benefits of participation and non-­participation. For example, years of dehumanizing propaganda may change how a perpetrator defines a situation. When an event such as political collapse occurs, it is interpreted in light of the propaganda embedded in the subconscious of the perpetrator. The presence of triggers, such as incitement to violence, orders from authority figures, and peer involvement may place the perpetrator under additional pressure to participate in genocide. In situations where value costs are minimal, i.e., moral conflict is neutralized, perpetrators may also consider other types of costs and benefits. In the final stage, the perpetrator considers his options, which compose the “perpetration matrix.” The consideration of options generally occurs subconsciously and rapidly; decision-­priming ensures that perpetrators have often basically chosen their action before being presented with a decision. Kelman and Hamilton rightly assert that obedience to authority sometimes means that “once a demand is seen as legitimate, the person acts as if he were in a non-­choice situation.”19 In many cases, the perpetrator is able to kill, while avoiding the decision to kill; in effect, the decision that perpetrators take is to conform, or to place themselves in the killing situation. Perpetrators have several options when placed in such a situation: they may kill, leave, resist participation in killing, or posture (feign participation).

Decision priming

Lenses

Options

• Exposure (context, propaganda) • Propensity

• Social, role, and immediate margin of discretion • Risks and incentives

• Kill • Flee • Posture • Resist

Perpetration matrix

Figure 6.1  Perpetration process.

130   Deciding to kill In a sense, genocide features a connivance between individual will and the collective will of the group, as expressed in institutional structures and ideologies. This does not imply that perpetrators are automatons. Rather, they retain power to act in conformity with or contrary to social influences.

Social margin of discretion The moral norm authorizing killing, is not universally and uniformly experienced in the genocidal context. We must disaggregate the moral context and recognize how it is bounded by positionality. The range of action alternatives available to perpetrators is determined by several factors, ranging from the macro to the micro: the social margin of discretion, the role margin of discretion, the immediate margin of discretion, and the propensity of individual perpetrators. Perpetrators assess and rank their action alternatives according to the overall social context and their role within it.20 Ordinary perpetrators do not passively receive and act on orders or norms received from above, they act as independent agents in interpreting options, filtered by situational and personal factors. The social margin of discretion is the perceived range of choices available to the perpetrator based on the society in which they live, while the role margin of discretion is the perceived range of choices available to the perpetrator based on their roles. The moral rules and expectations applied to individuals are informed by their positionality. In genocidal societies, the victim group is extensively dehumanized, and strong vertical and horizontal pressures to conform are imposed on all individuals within the group (this is especially true in participatory genocides). It can be very difficult to resist such pressures. The perpetrator may believe that he or she has no choice other than to perpetrate. As one perpetrator in Rwanda recounts, “I felt like I was supposed to kill.”21 Once the genocidal state has emerged, and the moral rules prohibiting violence against the victimized group are revised, the perpetrator may perceive his or her choices to be limited. In effect, the perpetrator has a limited social margin of discretion. This does not mean that the perpetrator lacks the ability to choose whether to perpetrate, but rather that the perceived range of choices is narrower than it would be for an individual in a different society. To act against one’s in-­group and legitimate authorities is a deeply asocial, self-­marginalizing, and potentially dangerous decision. The danger posed by disobedience differs from society to society, and even among genocides. In Rwanda and Cambodia, the killing of disobedient individuals was relatively common, whereas in Nazi Germany such acts were rare or nonexistent. A former Khmer Rouge group leader argues that “everyone was afraid,” while, according to a former Tuol Sleng guard, “We were told that we have the Angkar’s rules and if we disobey then we would be accused as an enemy against Angkar.”22 When asked who the enemies of the Khmer Rouge were, the group leader responded: “anyone who committed wrongdoings … you

Deciding to kill   131 did not even know what your mistake was.” This “coercive environment” was cited by the Khmer Rouge tribunal as a mitigating factor in the criminal responsibility of Kang Kek Iew (aka Duch, who directed Tuol Sleng Prison).24 Even where disobedience is not punishable by death, disobedient individuals may still face negative consequences; former Einsatzgruppen member Albert Hartl argues that “as a general rule such people could not expect to be promoted in the foreseeable future.”25 But postwar German prosecutors searching Nazi records found no case “in which refusal to carry out an order would have entailed an objective danger for the life and limb of the recipient of that order.”26 For example, when Erst-­Boje Ehlers of the Einsatzgruppen refused an order, he was transferred to an office job.27 Conflicts in moral rule guidance between emergent genocidal rules and the old rules may be effectively neutralized (see Chapter 8). Such adaptation may prove easier for many perpetrators than the alternative of challenging the state and their peers, and possibly experiencing shame for their own wrongful acts. The social margin of discretion also relates to proximity – in some contexts the state has greater reach and more direct impact on the private sphere. Moreover, some cases of mass killing involve less civilian participation. In these instances, the state’s campaign to destroy victims may be limited to certain groups of professional killers. Yet even with limited mass participation, the marginalization of and violence toward the victim group is usually widely understood.28 Perhaps, then, it is more useful to speak of moral contexts of genocide, rather than a single moral context. The general moral context is distinguished from that of the transgressive community; each context carries its own norms and expectations of individual behavior. In any case, the moral context of individuals varies in relation to their proximity to other perpetrators. It may also be localized – with normative expectations and moral habits depending on conditions in the local community. These moral contexts function as norm communities, and those who do not belong may be sanctioned or remove themselves voluntarily (for example through acts of resistance).29 In specialized genocides, where participation is more limited, there is less need for enforcement (coercion). The functional division of labor creates a greater social margin of discretion because killing is mostly limited to specialized units. 23

Role margin of discretion The effects of proximity on perpetration is represented in the role margin of discretion. For example, if perpetrators are violence specialists, such as a police officers, they may have internalized the norm of obedience, and believe it is their duty to carry out the orders of superordinates, even if these extend to harming others. Moreover, repeated obedience to orders, or repeated acts of violence against the victim group, may contribute to the formation of moral habits. This extreme commitment and identification with duty is also a technique of neutralization (the denial of autonomy).30 Perpetrators may also be heavily influenced by the organizational culture that characterizes their institution.

132   Deciding to kill Perpetrators may be situated within subcultures of violence. These transgressive communities have embraced the new norms of violence against the victim group. Perpetrators who perform roles in such organizations are under great social pressure to perform extreme and demonstrative acts of violence. Indeed, perpetrators who self-­select by joining such organizations have invested their identity and prestige in performative violence, toughness, and perhaps the transgression of old moral norms. The perpetrator’s role and social position may also increase perpetrators’ tendency to externalize their own agency. Cultural criminology argues that subcultures (i.e., transgressive communities) create their own reality. They draw from and challenge the wider culture, and provide a reference group for criminals. Subcultures such as the Interahamwe and SS may represent more extreme values than prevail in the mainstream, but these will still fall within the social margin of discretion that authorizes extreme violence against devalued groups. In a sense, each organization has its own moral context. Even within organizations, expectations may differ depending on role. Cooks, for example, may not be expected to take part in executions. It is important to remember, however, that the role margin of discretion may change as the individual’s role changes in an organization, or in the broader society. Beyond occupational and subcultural contexts, other social roles may also impact the perceived range of options available to perpetrators. For example, youth and the elderly may not be expected to participate directly in killing. Moreover, gender roles may also shape individual options. In Rwanda, for example there is a “dialectic of male/female distinctions” in which women are deemed to be reserved, submissive, and maternal, and men are dominant, strong, and logical.31 Moreover, in many cultures men are still held primarily responsible for the protection of the family and group. Thus, reflecting patriarchal norms, boys and men are often expected to participate in violence, while women may be excluded from killing or political decision-­making. Roles also influence the individual’s relationship with other actors – members of victim groups may, for instance, seek to increase their social distance from police and other potential perpetrators. Moreover, race, ethnicity, religion, nationality, and political orientation are all social identities that powerfully impact the perceived range of options available to individuals. The social and role margins of discretion are closely linked with individual alignment to social groups. This alignment occurs through conformity, but duty also plays an important role; individuals are expected to demonstrate loyalty to the group by conforming and fulfilling collective objectives. They have a duty to society (the social margin of discretion) as well as a duty towards their peers (the role margin of discretion). Notions of role and duty are both enforced by others’ expectations. The duty to peers may be an especially strong motivation in highly cohesive social formations, such as military units. In many cases, the duty to peers can suffice to motivate atrocious acts.

Deciding to kill   133

The immediate margin of discretion We can deepen our understanding of perpetrator positionality by moving beyond the society and group to consider the immediate context of perpetration. In a sense, it is propensity and the social and role margin that brings individuals to the moment of perpetration. The situational influences in the immediate context are the most directly determinative of killing, while also reflecting prior choices. Immediacy, in this sense, is both temporal (of the moment) and spatial (in that place). Immediate influences include the presence of authority figures, group dynamics, distance, threat perception, emotion and intoxication. Many of these factors also influence individuals’ capacity to exercise self-­control and adhere to their moral beliefs. We will examine each of these elements in turn, beginning with the presence of authority figures. Presence of authority figures Genocide, as a state crime, is permeated by authority. It is common for authority figures to be present in the environment of perpetration. Even when authority figures are not present, the legitimation of violence may be communicated through propaganda, as well as by state symbols (e.g., the provision of uniforms to non-­state actors, the display of the national flag). When paramilitaries or civilians participate, they are often led by or connected with authorities. In Rwanda, authority figures such as military, police, state officials, or church leaders were usually present at killing sites. See Table 6.1. Perpetration that is sanctioned by legitimate authority (in furtherance of elucidated state goals and policies) differs substantially from perpetration that is contrary to state objectives. Disobedience to the demands of the state and political community also means non-­conformity with peers. Isolated individuals are often marginalized by the community. Yet even in the absence of negative social sanctions, non-­conformity can be frightening for many people. Knowledge of and compliance with “situational etiquette” (the action deemed appropriate in a particular situation) greatly eases social relations. Breaking ranks with comrades is an asocial act.32 Moreover, the refusal of an unpleasant communal obligation is both a form of duty-­ shirking, leaving others to do the “dirty work,” and an expression of moral Table 6.1  Authority figures at killing sites in Rwanda Soldiers/Police Politicians/Government officials Church leaders None of the above

  4 (8%) 20 (40%)   6 (12%) 15 (30%)

Notes “Where you saw killings, were any of the following types of people present?” (Fifty perpetrator and victim respondents answered the question; multiple responses were possible.)

134   Deciding to kill reproach. For example, in the Józefów massacre during the Holocaust, non-­ shooters tended to attribute their behavior to being “too weak” rather than “too good.”33 Shame is a strong mechanism of social control: people seek to avoid negative social evaluation by conforming with peers. Bonds of camaraderie may also motivate perpetrators to take up their “share” of unpleasant tasks such as killing. Genocide perpetrators may drift into perpetration because they fear being socially ostracized. This is reflected in the claim of normality technique of neutralization (see p. 178). In some cases, individuals may be directly pressured to kill, with the implication that non-­compliance will have serious negative consequences. Yet this seems relatively rare. Rather, most obedience is habitual, cultural, or self-­ interested. In a secretly-­taped conversation at an interrogation center, a former Auschwitz guard, Eugen Horak, recounted to another soldier: When the sergeant major came along for volunteers for a firing squad, the majority of us did not wish to go. Then he gave us orders, “You, and you,” and the men had no alternative. It’s quite right too that they had no chance of examining the moral implications of the order. Orders must be carried out.… Those people lose all feeling.34 Thus, obedience in many cases is habitual, completely bypassing normal rational-­choice processes. Obedience can produce a kind of “subjective necessity.”35 In specialized killing organizations such as the SS, obedience is habitual. But the norm of obedience involves more than conformity and social pressures. Obedience is often backed and legitimized by the power of the law itself. One Rwanda perpetrator stated: “I killed them because it was policy … it was like a rule – I had no choice,” while another said, “we thought at the time we were doing something wrong, but the authorities were ordering us to do it.”36 In fact, 60 percent of the Rwandan perpetrators interviewed claimed to have been forced to perpetrate. Individuals may face a choice between committing a definite wrong (by violating the rule of law) and a possible wrong (by killing a person condemned by legitimate authorities).37 In such cases, killing becomes the preferred alternative. Authorities may directly communicate orders to kill, but in other cases, such demands are implicit in the moral context and organizational culture. Group dynamics Most genocidal killing is done collectively. This is unsurprising when we consider that killing is perpetrated in the name of the group. Most perpetrators are situated in a supportive group, usually composed of other individuals who kill, sometimes alongside an audience. This audience may actively encourage killing, but even when it does not, its presence may be seen as a tacit endorsement. Consider, for example, the pogroms against Jews by Polish peasants during World War II: Jan Gross describes “an aggressive, criminal crowd” accompanying the

Deciding to kill   135 killers. As Lee Ann Fujii notes, witnessing killing is not necessarily an act of abstention, rather the watchers are part of the performance of genocide.39 Groups also provide cues about appropriate levels of violence. The difficulty of resisting a new (genocidal) moral context is magnified by the presence of a group of conformist peers. Individuals making decisions often also rely on a “reference group” of trusted sources, but such information grows increasingly inaccurate.40 Outrageous rumors often form the basis of group action. Indeed, the out-­group’s transgression of profound moral norms often demands retaliatory action that in other contexts might be illegitimate or immoral. Killing may become a means to demonstrate loyalty to the group, particularly in transgressive communities. One genocide perpetrator states: “the jeering of colleagues is awful to overcome.… When the killings begin, you find it easier to ply the machete than to be stabbed by ridicule and contempt.”41 Similarly, Alfred Krumbach, a first lieutenant in the Einsatzgruppen who had fainted twice at the sight of bodies in police training, feared fainting again in front of his colleagues.42 The desire to conform greatly magnifies the power of groups to influence individual behavior. Social psychologists point to a coaction effect –people work faster when they see others work – as well as an audience effect (also known as evaluation apprehension); they work faster when observed by others.43 It is reasonable to believe that such effects are also present in cases of genocidal killing that occur in group settings. Indeed, several interview subjects explicitly stated that their participation in genocide was influenced by the genocidal behavior of their peers. As one respondent explained, “if you saw your friend killing people, you also thought you should get involved.”44 In “performance” cases, as when a crowd gathers to watch two men who appear about to fight, individuals may be considered responsible for violence by their mere presence.45 However, to be held criminally responsible, people in such settings must be aware of the effect of their presence. Both experiments and anecdotal evidence also show that the spell of individual conformity can easily be broken when the individual being pressured receives support from any of his or her peers.46 The group also diffuses responsibility and lends perpetrators a measure of anonymity. For example, the affidavit by Otto Ohlendorf (head of Einsatzgruppe D) states “In Einsatzgruppe D I never sanctioned shootings by individuals. I always gave orders for several people to shoot simultaneously, in order to avoid any individual having to take personal responsibility.”47 Such an anonymity effect allows people to behave in ways that they would not normally if they were faced with the prospect of being singled out and ostracized. Effectively, they can adhere more easily to a revised moral context when they are accompanied by numerous co-­perpetrators. One of my interview subjects in Rwanda used a bow and arrow to shoot in the leg a Tutsi man fleeing a mob of Hutus. This hobbled the man; the crowd overtook him, and he was hacked to death. The interviewee was adamant that he had not killed the victim, despite his obvious role. The group setting and compartmentalization of tasks provided him with deniability. 38

136   Deciding to kill The more closely a perpetrator is bonded with the group, the lower the costs of perpetration, and the less likely it is that individuals will feel responsible. When it lacks autonomy, perpetrator decision-­making is divorced from the consequences. Group cohesion tends to increase in times of tension and conflict, such as genocide. Moreover, in strong groups, all individuals must participate or appear to participate. An individual who has experienced a strong degree of de-­ individuation, such as someone in a professional violence organization, may experience a reduction of moral costs as individual identity is eroded. The withdrawal of autonomy and embedding of individuals in group settings also renders self-­control, i.e., non-­conformity, more difficult. Distance The greater the physical distance between perpetrator and victim, the less perpetrators experience the consequences of their acts, and the less moral conflict is engendered. In general, killing is much more difficult at bayonet/machete range than it is at mechanical sight (gun scope) range, and killing at gun scope range is more difficult than killing from an airplane.48 Modern technology has resulted in a poor correlation between psychological guilt and degrees of moral responsibility.49 Moreover, the costs of killing decrease as less effort is required from the perpetrator – it is easier to kill someone by clicking a mouse than to overpower them and force a knife into their body. As one Rwandan perpetrator argues, “killing with a gun is a game compared to the machete; it’s not so close up.”50 Perceived threat When perpetrators perceive victims as threats to their personal safety, whether in the short or long term, this may greatly facilitate the decision to perpetrate. Moreover, the killing act, when directed against individuals deemed a “threat” to the perpetrator or social group, is transformed into an act of self-­defense. This neutralizes dissonance over conflicting sets of moral rules. It also redefines the situation so that perpetration is recast as an act of self-­defense, thus it can be committed without challenging the values of the perpetrator. Propaganda discourses are notably important here. According to psychologists, threats and frustration are the two main sources of aggression.51 Victims may be deemed to pose a threat to perpetrators either on a personal level (this person is a threat to me) or on a generalized level (this person is from a group which threatens my group). Ideology connects these two notions of threat, linking the “personal victim” with the “metaphysical victim.”52 A reversal of victimization occurs in which victims are perceived as perpetrators. The external projection of strain, “victim-­ blaming,” also contributes to the threat motif.

Deciding to kill   137 Emotion and intoxication Everyone was afraid, even Interahamwe. There were the people who really enjoyed killing, like the marijuana users or the boozers.53 (Rwandan perpetrators) The act of perpetration is generally not banal for the perpetrators. Rather, most act in a state of heightened emotional arousal. When perpetrators feel anger towards their victims, certain biophysical responses are triggered that ease perpetrators’ tasks, and help them bypass ordinary decision-­making entirely. Some studies indicate that aggression, the intent to harm others, is linked to anger and other negative emotions (e.g., frustration, fear, pain), although anger is often simply expressive.54 Aggression may also be an effective means to suppress anxiety.55 The frustration-­aggression hypothesis posits that aggression may be caused by non-­attainment of an expected goal (strain). An individual provoked to anger often views the actions of their target as only the most recent provocation in a long series.56 Thus, the anger of génocidaires may be born from propaganda dehumanizing and framing the victim group as hostile. Omar McDoom found that a significant minority of his Rwandan perpetrator respondents claimed they killed to avenge the assassination of Habyarimana.57 Some individuals are also chronically aroused – which often leads them to interpret situations as threatening.58 Anger may significantly cloud the rationality of perpetrator decision-­making.59 Emotional arousal tends to limit people’s potential to carefully process information. Such emotional states reduce self-­control, rendering moral disengagement and aggression more likely.60 Furthermore, groups are arguably less capable of anger-­suppression than individuals.61 For example, psychologists point to the phenomenon of “emotional contagion,” in which we mimic the emotions of others.62 When people project hate against the dehumanized other (or take pleasure in tormenting them), these emotions may exert a powerful effect on others. Some criminologists also posit that some perpetrators of violent crime enter a state of “red-­out,” in which they are enraged and even disassociated from their normal self.63 Intoxication may have the effect of lowering perpetrator inhibitions; this can influence the perpetrators’ assessment of moral costs, as well as reducing self-­ control. Studies show that alcohol makes people more aggressive.64 A survivor of Omarska concentration camp in Bosnia recounts that “the beating was constant; they were always completely drunk.”65 In Rwanda, the military sometimes distributed drugs and alcohol to ordinary perpetrators so that “they started enjoying killing.”66 Alcohol was also commonly used before, during, and after killings in the Holocaust. A Jewish survivor of Treblinka recalls seeing “SS-­men who held a pistol or truncheon in one hand, [and a] whiskey bottle in the other.”67 Moreover, generous alcohol rations were distributed to SS forces in the East; this was likely intended to promote social bonding, but was also used during

138   Deciding to kill executions, with Einsatzgruppen shooters in Kamenets-­Podolsk taking breaks from shooting children to drink schnapps before returning to the firing line.68 The effect of intoxicants is recognized in many juridical systems as an excuse; however, the intoxication defense is generally unavailable when the individual foresees the disinhibiting effects of intoxication and voluntarily consumes intoxicants (perhaps even as an intentional strategy to gain “Dutch courage” for committing a crime). Both emotional arousal and intoxication may facilitate moral disengagement.

Risks and incentives Almost all perpetrators are rational. This does not mean that their decisions are correct, or even beneficial to them. Individuals assess their options based on flawed information, flawed interpretations, and limited intellectual power. However, from the perspective of perpetrators acting in the moment, they are participating in genocide because it advances their interests, or because they feel they have no other (or better) option. Perpetrator rationality is embedded in power structures – non-­participation may involve challenging the state. Recent iterations of rational-­choice theories of crime acknowledge that criminal behavior is socially situated, with decisions made in a context of limited knowledge.69 Even in an enabling organizational culture, genocide is facilitated by moral neutralization that distances perpetrators from universal moral values, above all the prohibition against killing. Perpetrators may not deliberate at length, but neither do they act as martyrs. They are still satisficing agents, selecting the best available option. Genocide often occurs in a fluid context, requiring rapid decision-­making. Thus, decisions tend to be intrinsic and intuitive – following the path of least resistance. This helps to explain why many non-­ideological perpetrators kill. They do so because, in the genocidal context, obeying legitimate authority is the easiest option. It is often said that studies of genocide must explain why people kill. We might better ask why, in environments that strongly encourage killing, some choose not to kill. Hannah Arendt wrote, “under the conditions of the Third Reich only ‘exceptions’ could be expected to react normally.”70 Perhaps adapting to changing social circumstances may better represent normalcy than self-­sacrificing opposition to state policies? Individuals in a genocidal context will assess their alternatives based on risks (potential losses incurred through participation or non-­participation) and incentives (potential benefits incurred through participation or non-­participation). These risks and incentives are interpreted through subjective lenses, including individual propensity and exposure. Risks for perpetrating genocide are lowered substantially by the social conditions present in the genocidal context. Propaganda communicates norms, urges the participation of relevant in-­groups (whether they are violence specialists or civilians), and sets out the risks involved in individual variance and resistance (non-­participation). Risks are understood differentially, based on an individual’s propensity and position (role and immediate margin) within society.

Deciding to kill   139 The prospect of impunity, both through domestic endorsement and external apathy, makes perpetration much less costly. The risks involved in participation are physical (in overcoming the resistance of victims and other intervening parties), emotional (in that harming others may be upsetting and stressful), and moral (the dissonance engendered by contravening conventional norms and personal beliefs). The risks involved in non-­participation are centered on state authority (being punished by the state), physical violence (being killed or abused by authorities), and ostracism (being isolated from the social group). If you are Paul, a 20-year-­old Rwandan male in 1994, you have to weigh your fear of the Interahamwe who come to your door against your moral and visceral disgust at the idea of killing. A collective action problem is present among reluctant perpetrators and bystanders in genocide: they all privately disagree with genocidal killing, yet they feel that they alone would not have the power to overcome state-­sanctioned mass violence. This could be overcome if these individuals were able to coordinate their resistance, but such coordination would involve substantial risks – especially for the first movers – with no guarantee of success. From the perspective of the perpetrating regime, a collective-­action problem may also arise in mobilizing free-­riders – individuals from the ethnic group who should be involved in violence, but avoid full participation. Genocide also involves a moral hazard, as the physical, emotional, and moral risks are shifted from high-­level authorities to low-­level perpetrators. These high-­level authorities often do not face the consequences of perpetration directly. Risks are higher for low-­level perpetrators where the social margin of discretion is restricted (i.e., participatory genocides) or where they are in positions with a restricted role or immediate margin (e.g., they are in an army unit tasked with participating in killings, and in a killing environment with authority figures who demand compliance). Moral risks are, of course, relative to perpetrators’ own perspectives – their moral rules and degree of self-­control. Perpetrators with high propensity will experience lower value costs for killing, particularly in an enabling environment. The weight given to moral risks varies from individual to individual; those who assign strong weight to such costs may resist perpetration. The presence of incentives may mitigate or override risks for some individuals. Incentives for participation are always related to individual needs or wants, such as for money, sex, status, excitement, and approval. The “destructive needs satisfaction” of genocide involves the perpetrator satisfying his needs by stealing selfhood, as well as physical and material well-­being, from the victims. Financial incentives may also play a role in the ranking of action alternatives. These financial incentives may include immediate benefits, such as goods looted from the victims or payments received for their act of perpetration; or more long­term benefits, such as possible promotion within bureaucratic structures. Sex may also motivate some perpetrators, in much the same way that it motivates “ordinary” rapists: they receive sexual gratification as well as a rush from

140   Deciding to kill exercising dominance over the victims. Of course, this dominance need not take a sexual form; sadists may also “need” to kill as a way of gratifying their need to control other people. Involvement in genocide may also be exciting, as individuals earn the approval of their peers and enhance their social status. By contrast, in the genocidal context, non-­participation may lead to social ostracism or even physical jeopardy. Non-­participants may be branded as deviants, transgressing the emergent moral order. Consider the following account by Alison Des Forges of the risks and incentives present in the Rwandan Genocide: They (Burgomasters) directed or permitted communal police, militia, or simply other citizens to burn down houses and to threaten the lives of those who refused to join in the violence. They also offered powerful incentives to draw the hesitant into killing. They, or others solicited by them, provided cash payments, food, drinks and, in some cases, marijuana to assailants. They encouraged the looting of Tutsi property, even to the point of having the pillage supervised by the communal police.… In several places police reprimanded those people who wanted only to pillage and not to kill.… One of the most important resources for the burgomaster in enlisting participants was his authority to control the distribution of land, a much desired and scarce source of wealth for the largely agricultural population. Hutu who had attacked Tutsi in the 1960s had acquired the fields of the victims. A generation later, people again hoped to get more land by killing or driving Tutsi away … another participant commented that people were cultivating lands taken from victims “to reward themselves for the work they had done.” As usual, “work” meant “killings.”71 In genocide, unfortunately, the incentives for non-­participation are scant, usually centering on moral identity. Some individuals receive benefits from flouting norms (through prohibited violence or altruism), as their self-­image is built on rebellion or moralism.72 Incentives for participation are more important for some individuals than for others. Pathologically-­driven perpetrators will likely become involved in killing irrespective of incentives. Incentives may, however, reinforce the decision-­making of ideologues, entrepreneurs, and conformists. They may consider, for example, the degree of effort required, the amount and immediacy of the reward, and the likelihood and severity of punishment.73 Incentives provide tangible benefits for involvement, while also demonstrating that participation in killing is correct and that participants will receive rewards from legitimate authorities. How does all this determine who participates in genocide? High-­propensity individuals (those with low self-­control, high moral disengagement, and high proximity) may consider incentives for participation, while interpreting the risks of participation as minimal. Low-­propensity individuals (those with high self-­ control, low moral disengagement, and low proximity) may not highly value

Deciding to kill   141 incentives for participation, but they will participate anyway, out of fear of the risks of not participating. The result is that some low-­propensity individuals participate, and most high-­propensity individuals do also. The rates of participation increase among highly-­structured roles like soldiers and concentration camp guards. Individuals with lesser self-­control, those who are morally disengaged from the victims, and those more driven to maintain their ties to the political community, will rank alternatives in a manner that favors perpetration. Where moral rules conflict, creating uncertainty about the “correct” action, then other factors may enter into play, such as risks and incentives that tip individuals toward perpetration. Perpetrators who are already morally disengaged from the victims may also seek such incentives, but they will not be primary motives for perpetration. Moral rules Moral rules or “normative attitudes can work as a moral filter and delete normatively proscribed actions from the set of alternatives under consideration.”74 In other words, material benefits are only relevant to those who consider such incentives important. Moreover, normativity is also socially situated in community values and community participation in violence. In some cases of moral conflict, though, material benefits outweigh moral costs. One Rwandan perpetrator remembers: On the way home from my field in July of 1994, I met people distributing beef, but they would not give me any. So, the next day I tried again, and they would still not give me any. So, then I joined.… Nobody was pressuring us. We pressured ourselves because we wanted property.75 Active perpetrators are less likely to prioritize moral considerations over other benefits. In effect, they have already transgressed a moral norm, and must rationalize their transgression to maintain their self-­image. Once a preference is decided, there is a strong incentive to justify it.76 Perpetrators of genocide, whether active or passive, must maintain some sense of moral consistency. They, like other individuals, strive to avoid cognitive dissonance. Such dissonance occurs when an individual holds two incompatible cognitions (ideas). It may also occur, for example, when an individual’s actions do not match their beliefs (cognition), producing discomfort and making it much more difficult to act in situations of profound inner moral resistance. Individuals may work to avoid cognitive dissonance by various means, including the exclusion or reimagining of disconfirming information, changing cognitions so they are more compatible with existing ones, or creating new cognitions that help to bridge the gap.77 The techniques of neutralization can both restructure existing cognitions and introduce new cognitions. Before people undertake an action, they may also assess whether the intended act accords with their beliefs. We may think here of moral risks: actions that exemplify values are

142   Deciding to kill a sort of deposit in the positive self-­image “bank,” while actions that clash with values represent a deduction or even an attack on one’s self-­image. The exception to this principle is perpetrators within transgressive subcultures who may revel in a kind of deviant, “bad” self-­image. Killing contravenes several universal moral norms, and therefore genocide may engender cognitive dissonance. Nonetheless, situational factors increase moral disengagement and externalize responsibility, as well as changing the perpetrator’s view of the victim. Perpetrators also self-­objectify, removing their own agency and consequently neutralizing moral costs. If the victim is no longer seen as innocent, or if involvement is seen as involuntary, then dissonance becomes less likely. Thus, moral neutralization lowers the value costs of engaging in violent acts. Ideology may also alter values analysis as perpetrators come to restructure their own values, and thus their assessment of the situation. The coercive pressures of the genocidal state distort normal mechanisms of self-control. Behavioral economists Douglas Nelson and Eugene Silberberg argue that where the price to individuals of being able to express their own values and interests is low; they [individual values and interests] will loom large in the choices made; but where the price one pays for expressing one’s own ideology, or norms, or preferences extremely high, they will account much less for human behavior.78 In genocide the price for expressing divergent norms is often death. In times of uncertainty, costs and benefits may not be clear, especially given the limited information available to individuals. Perpetrator self-­control in genocide is a deviant act, whereas in normal crimes self-­control generally involves acquiescence to mainstream moral and legal norms. Perhaps, most crucially, the conflicted perpetrator, who is not content to act merely out of passive conformity, must also see the individual victim as dangerous – linked to the metaphysical victim. The decision to perpetrate, like all other decisions, is made not in terms of absolute values, but rather according to the ranking of values. Participation in genocide may come to carry fewer perceived costs than defying the state and betraying one’s own group.79 Direct and indirect perpetrators Direct perpetrators (those present at the killing site) can be distinguished from indirect perpetrators (those not present). These twin categories of perpetrator roles present differing options. Both types of perpetrators exist within the same overall moral context, yet the decision to commit genocide, for indirect perpetrators such as leaders, is a matter of policy-­making rather than a visceral act. Direct perpetration, then, is a product of an individual’s role margin, which could be altered by serendipity (for example, the individual who stumbles across a group of killers with their victims in a participatory genocide).

Deciding to kill   143 Direct perpetrators must confront victims; they must wield the weapon. It should not be surprising, then, that indirect perpetrators seem to be more ideological, while direct perpetrators are more driven by situational factors. For example, in Rwanda very few direct perpetrators were even familiar with ideological treatises such as the “Ten Commandments of the Hutu.”80 Indirect perpetrators rarely see the consequences of their perpetration. Thus, they can maintain the purity of their ideology without being concerned about the implications of their beliefs. It may be easier for indirect perpetrators, such as “desk murderers,” to maintain a state of moral disengagement from the victim group. They are not faced with the consequences of perpetration. They also likely suffer fewer moral conflicts and less trauma. Himmler, upon witnessing the execution of 200 Jews in Minsk, was on the verge of physical collapse.81 Of course, indirect perpetrators are also not located at the scene of perpetration. Thus, they are not subject to immediate contextual factors, such as intoxication. Nonetheless, other factors are relevant to the decision-­ making of indirect perpetrators. Consider the following case: I was head in the sector, so when the genocide started … those people who were on top of me said that you should make some roadblocks.… So, I put up some roadblocks. They were inspected by people who were trained and who had guns.… When the president died, that is when they started the violence, along with people from Habyarimana’s region. And they were calling this revenge: that is when they started killing the Tutsis. I wanted to stop the violence, but I found out that it was something beyond me.… I was a leader, so my position did not allow me to be on the roadblock … so with my own hands I did not kill anybody … I never witnessed somebody killing another person. But as someone who was head in the sector, I used to get reports all the time from the cells … telling me that it is all so bad. By the time I realized that the guns I had distributed were being misused, I found myself in genocide. I have played a role in this genocide. They pressured me to kill Tutsi. When I said “no” they even tried to shoot me. I was in my office. They sent Interahamwe to kill me … and then when I was not killed, they said “you should not come back to the office, you should remain at home.” So, when I remained at home, that is when they started killing, during my absence.82 This perpetrator was a mid-­level leader whose overriding concern appears to have been maintaining his position. He claims to have opposed the genocide, but did not take significant steps to stop it. Indirect perpetrators may participate for purely instrumental reasons (the desire to gain or maintain power or material benefits) or for ideological reasons (the desire to implement a program of political/social change). This distinction between entrepreneurial and ideological leaders roughly corresponds to Max Weber’s typology of instrumental rationality versus value rationality.83 One may then place genocidal leaders on a continuum of instrumentality, with some leaders (Pol Pot, for example)

144   Deciding to kill exhibiting less of it, but being more ideologically-­driven. Yet all leaders of genocide display both instrumental and ideological motives; even instrumental acts are facilitated by underlying exclusionary ideologies. Norm bandwagons and tipping points One of the main mechanisms promoting individual participation in genocide is norm bandwagoning, particularly in more participatory cases. Individuals acquiescence to genocidal processes once it becomes clear those processes represent the new mainstream. They align themselves with whatever norm they believe is going to be predominant.84 In the case of genocide, state power transforms emergent norms into dominant ones. Philip Verwimp draws from Timur Kuran to provide a convincing model of how preference falsification, “the practice of misrepresenting one’s genuine wants under perceived social pressures,” leads to mass participation: “a genocidal bandwagon is set in motion because the publicly declared preference of each individual is enough to draw the next individual on board.”85 In other words, individuals will participate in genocide if they believe it conforms with dominant norms, and this perception is based, in part, on the declared perceptions of others (public opinion). Propaganda then plays a crucial point in communicating the dominant norm and public opinion. This, in turn, lowers individual tipping points for participation in genocide. We must also recall that genocidal propaganda stokes fears of the other, presenting an image of an implacable and irrational enemy who could strike at any moment. The strain preceding genocide lends further credibility to this argument. We can hypothesize that individual tipping points vary, depending on propensity and margins of discretion. In other words, individuals are more likely to “tip” if they occupy certain roles (i.e., belong to certain norm communities), and are in close proximity to other perpetrators (who may act as a reference group or enforce norms coercively). We can thus conclude that those closest to the regime and to other perpetrators have lower thresholds. They may perceive propaganda as more credible, and may have already adopted the ideology of the regime by means of self-­selection (through pathological or ideological pathways). Tipping-­ point mechanisms are most relevant for conformist pathways to perpetration – these perpetrators would likely not join genocide without the presence of a perceived norm bandwagon. Some individuals, however, will never “tip,” as it poses too great a moral cost. It is also interesting to consider more broadly the relationship of individual decisions to the group enterprise of genocide – that is, of individual motives to collective ones. The atrocities that occur during genocide differ from those in war, because in genocide discipline and group solidarity actually enable rather than restrain abuses against civilians, while in war discipline usually reduces illegitimate violence.86 Therefore, in both genocide and other forms of mass violence, the individual’s decision to commit violence is profoundly shaped by the group context, but one that actually encourages the commission of atrocities. One Rwandan perpetrator remembers: “First there was an order, and if I did not

Deciding to kill   145 join the group they would have come to take my property.”87 Once again, non-­ genocidal alternatives become costlier, and genocide becomes less costly (at least in the short term).

Conclusion: mass violence and individual will Individual perpetration results from a complex interaction between contextual factors and individual propensity. Despite the intense situational factors present in genocidal contexts, participation in genocide still requires a decision on the part of the perpetrator. Individuals choose to perpetrate genocide, but this choice is not made in isolation. Rather, it is heavily influenced by the social environment. Perpetrators voluntarily become objects of external forces or, in the case of leaders, themselves generate those forces. Perpetrators’ options are shaped by their society, their roles within that society, and the immediate context of genocidal acts. Moreover, the costs of genocide are shifted in such a way that perpetration becomes the least costly option. This is especially relevant for conflicted perpetrators. When they feel unsure of the correct course of action, they may privilege their own self-­interest over empathy for the victims. Perpetration ultimately functions as a system, governed by normative and organizational structures, and consisting of the perpetrators, their accomplices, and willfully blind bystanders.

Notes   1 Interview R61 (Rwanda), Mont Kigali TIG Camp, October 2009.   2 Interview Bu6 (Burundi), Bujumbura, September 2009.   3 Cohen and Felson, “Social Change and Crime Rate Trends,” 588–608.   4 Cohen, Kluegel, and Land, “Social Inequality and Predatory Criminal Victimization,” 505–524. See also: Birbeck and LaFree, “The Situational Analysis of Crime and Deviance,” 126.   5 Horowitz, The Deadly Ethnic Riot, 386–387.   6 Groebel and Hinde, Aggression and War, 13.   7 Valentino, Final Solutions, 36.   8 Valentino, 36.   9 Mueller, “The Banality of ‘Ethnic War’,” 54. 10 Interview R1 (Rwanda), Kigali Central Prison, July 2009. 11 Interview B06 (Bosnia), Srebrenica, April 2009. 12 Klee, Dressen, and Riess, The Good Old Days, 78. 13 Sereny, Into that Darkness, 232–233. 14 Kauzlarich, “Victimisation and Supranational Criminology,” 451. 15 Kauzlarich, 451. 16 Interview R28 (Rwanda), Rugerero TIG Camp, August 2009. 17 Beyond rational choice models, where the individual picks the best choice from a menu, “bounded rationality” models assume that the information available to the individual is limited. In her work on cults Janja Lalich argues for a model of “bounded choice” where the information available to individuals is also limited by power. See: Lalich, Bounded Choice, 257–258. 18 Sunstein, “Social Norms and Social Rules,” 6. 19 Kelman and Hamilton, Crimes of Obedience.

146   Deciding to kill 20 See also: Steiner, “The role margin as the site for moral and social intelligence,” 61–75. 21 Interview R26 (Rwanda), Gisenyi Prison, August 2009. 22 Interview C04 (Cambodia), Kampong Chhnang, October 2008; Interview C05 (Cambodia), countryside near Phnom Penh, November 2008. 23 Interview C04 (Cambodia), Kampong Chhnang Province, October 2008. 24 Hinton, Man or Monster?, 239. 25 Klee, Dressen, and Riess, 85. 26 Mann, The Dark Side of Democracy, 268. 27 Mann, 268. 28 For example, in major German cities around 30 percent of people knew about the Holocaust and a further 10 percent suspected that mass killing was occurring. See: Johnson and Reuband, What we knew, 369. 29 Sunstein, 14. 30 Hinton, Why Did They Kill?, 268. 31 Burnet, Genocide lives in Us, 44–45. 32 Browning, Ordinary Men, 185. 33 Browning, 185. 34 Overy, Interrogations, 199. 35 Mann, 269. 36 Interview R30 (Rwanda), Butare (with prison work group), August 2009; and R34, TIG Camp (near Musanze), August 2009. 37 Baumeister, Evil: Inside Human Violence, 267. 38 Gross, “Opportunistic Killings and Plunder of Jews by their Neighbors,” 8. 39 Fujii, Killing Neighbors, 175. 40 Staub, The Psychology of Good and Evil, 389. 41 Hatzfeld, Machete Season, 226. 42 Langerbein, Hitler’s Death Squads, 178. 43 Butler and McManus, Psychology, 112. 44 Interview R22 (Rwanda), Tumba TIG Camp, August 2009. 45 May, Genocide: A Normative Account, 161. 46 Smith and Mackie, 317. 47 Klee, Dressen, and Riess, 60. 48 Grossman, On Killing, 99–141. 49 Glover, Humanity, 100. 50 Hatzfeld, 22. 51 Staub, The Roots of Evil, 35. 52 Bauman, Modernity and the Holocaust, 189. 53 Interview R8 (Rwanda), Kigali Central Prison, July 2009; Interview R61 (Rwanda), Mont Kigali TIG Camp, October 2009. 54 Anger may be manifested in physical sequelae such as muscle tensing. Seymour Feshbach, “The Bases and Development of Individual Aggression,” in Groebel and Hinde, 80. 55 Fromm, Anatomy of Human Destructiveness, 224. 56 Horowitz, The Deadly Ethnic Riot, 538. 57 McDoom, The Micro-­Politics of Mass Violence, 285. 58 Maier-­Katkin, Mears, and Bernard, “Towards a Criminology of Crimes Against Humanity,” 241. 59 The same argument has been made with respect to “ordinary” hate crimes. See: Piquero and Tibbets, Rational Choice, 267. 60 Smith and Mackie, 486. 61 Jacob M. Rabbie, “Group Processes as stimulants of Aggression,” in Groebel and Hinde, 150. 62 For example, when a baby cries, others will also start to cry; when a person yawns, other people will feel the urge to yawn. See: Beck, Prisoners of Hate, 146.

Deciding to kill   147 63 Herve, Cooper, and Yuille, “Memory Formation in Offenders,” 65. 64 It is not entirely clear whether alcohol facilitates aggression directly by paralyzing inhibitors in the brain, indirectly through the cognitive changes which accompany alcohol consumption, or whether alcohol is linked to aggression through predispositions. See Bushman and Cooper, “Effects of Alcohol on Human Aggression,” 348. 65 Alvarez, Governments, Citizens, and Genocide, 24. 66 Interview R1 (Rwanda), Kigali Central Prison, July 2009. 67 Westermann, “Stone Cold Killers,” 9. 68 Westermann, 4 and 7. 69 Piquero and Tibbet, 265. 70 Arendt, Eichmann in Jerusalem, 27. 71 Des Forges, Leave None to Tell the Story, 236–237. 72 Sunstein, 13. 73 Cornish and Clarke, The reasoning criminal. 74 Kronenberg, Heintze, and Mehlkop, “The Interplay of Moral Norms and Instrumental Incentives,” 283. 75 Interview R40 (Rwanda), Nyarusenge TIG Camp, August 2009. 76 Hood, The Self Illusion, 157. 77 Aronson, The Social Animal, 183. 78 Nelson and Silberberg, “Ideology and Legislator Shirking.” 79 Hood, 157. 80 Only around a third of my Rwandan interview respondents were familiar with the “10 Commandments of the Hutu” and most of these were relatively well-­educated (i.e., government officials and other professionals). 81 Midlarsky, The Killing Trap, 187. 82 Interview R03 (Rwanda), Kigali Central Prison, July 2009. 83 See: Max Weber, The Theory of Social and Economic Organization. New York: Oxford University Press, 1947. 84 Verwimp, Peasants in Power, 183. 85 Verwimp, Peasants in Power, 190. 86 Humphreys and Weinstein, “Handling and Manhandling Civilians,” 433. 87 Interview R35 (Rwanda), Rwaza TIG Camp, August 2009.

Works cited Alvarez, Alex. Governments, Citizens, and Genocide: A Comparative and Interdisci­ plinary Approach. Bloomington, USA: Indiana University Press, 2001. Arendt, Hannah. Eichmann in Jerusalem: A Report on the Banality of Evil. New York: Penguin Books, 2006 (originally published 1963). Aronson, Elliot. The Social Animal (8th edition). New York: Worth Publishers, 1999. Bauman, Zymunt. Modernity and the Holocaust. Ithaca, New York: Cornell University Press, 2000. Baumeister, Roy F. Evil: Inside Human Violence and Cruelty. New York: W.  H. Freeman, 1997. Beck, Aaron T. Prisoners of Hate: The Cognitive Basis of Anger, Hostility, and Violence. New York: Harper, 2000. Birbeck, Christopher and Gary LaFree. “The Situational Analysis of Crime and Deviance.” Annual Review of Sociology 19 (1993). Browning, Christopher. Ordinary Men: Reserve Battalion 101 and the Final Solution. New York: Harper, 1998. Burnet, Jennie. Genocide lives in Us: Women, Memory, and Silence in Rwanda. Madison: Wisconsin University Press, 2012.

148   Deciding to kill Bushman, Brad J. and Harris M. Cooper. “Effects of Alcohol on Human Aggression: An Integrative Research Review.” Psychological Bulletin 107, no. 3 (1990). Butler, Gillian and Freda McManus. Psychology: A Very Short Introduction. New York: Oxford University Press, 1998. Cohen, Lawrence and Marcus Felson. “Social Change and Crime Rate Trends: A Routine Activity Approach.” American Sociological Review 44, no. 4 (1979). Cohen, Lawrence E., James R. Kluegel, and Kenneth C. Land. “Social Inequality and Predatory Criminal Victimization: An Exposition and Test of a Formal Theory.” American Sociological Review 46, no. 5 (October 1981). Cornish, Derek Blaikie and R. V. G. Clarke. The reasoning criminal: rational choice perspectives on offending. Berlin: Springer-­Verlag, 1986. Des Forges, Allison. Leave None to Tell the Story: Genocide in Rwanda. New York: Human Rights Watch, 1999. Fromm, Erich. The Anatomy of Human Destructiveness. Greenwich, USA: Fawcett, 1979. Fujii, Lee-­Ann. Killing Neighbors: Webs of Violence in Rwanda. New York: Cornell, 2009. Glover, Jonathan. Humanity: A Moral History of the Twentieth Century. London: Yale University Press, 2001. Groebel, Jo and Robert A. Hinde (eds.). Aggression and War: Their Biological and Social Bases. New York: Cambridge University Press, 1993. Gross, Jan T. “Opportunistic Killings and Plunder of Jews by their Neighbors – a Norm or an Exception in German-­occupied Europe?” Amsterdam: NIOD Institute for War, Holocaust and Genocide Studies, 2014. Grossman, Dave. On Killing. New York: Back Bay Books, 1996. Hatzfeld, Jean. Machete Season. New York: Picador Books, 2005. Herve, Hugues, Barry S. Cooper, and John C. Yuille. “Memory Formation in Offenders: Perspectives from a Biopsychosocial Model of Eyewitness Memory.” In Narrative Criminology: Understanding Stories of Crime, edited by Lois Presser and Sveinung Sandberg. New York: New York University Press, 2015. Hinton, Alexander Laban. Man or Monster? The Trial of a Khmer Rouge Torturer. London: Duke University Press, 2016. Hinton, Alexander Laban. Why Did They Kill? Cambodia in the Shadow of Genocide. London: University of California Press, 2005. Hood, Bruce. The Self Illusion: How the Social Brain Creates Identity. Toronto: HarperCollins, 2013. Horowitz, Donald. The Deadly Ethnic Riot. New Delhi: Oxford University Press, 2002. Humphreys, Macartan and Jeremy M. Weinstein. “Handling and Manhandling Civilians in Civil War.” American Political Science Review 100, no. 3 (August 2006). Johnson, Eric and Karl-­Heinz Reuband. What we knew: Terror, mass murder, and everyday life in Nazi Germany. London: John Murray, 2005. Kauzlarich, David. “Victimisation and Supranational Criminology.” In Supranational Criminology: Towards a Criminology of International Crimes, edited by Alette Smeulers and Roelof Haveman. Antwerp: Intersentia, 2008. Kelman, Herbert C. and Lee Hamilton. Crimes of Obedience. New Haven, USA: Yale University Press, 1990. Klee, Ernst, Willi Dressen, and Volker Riess (eds.). The Good Old Days: The Holocaust as Seen by its Perpetrators and Bystanders. Old Saybrook: Konecky and Konecky, 1991. Kronenberg, Clemens, Isolde Heintze, and Guido Mehlkop. “The Interplay of Moral Norms and Instrumental Incentives in Crime Causation.” Criminology 48, no. 1 (2010).

Deciding to kill   149 Lalich, Janja. Bounded Choice: True Believers and Charismatic Cults. London: University of California Press, 2004. Langerbein, Helmut. Hitler’s Death Squads: The Logic of Mass Murder. College Station: Texas A & M University Press, 2004. Maier-­Katkin, Daniel, Daniel Mears, and Thomas J. Bernard. “Towards a Criminology of Crimes against Humanity.” Theoretical Criminology 13, no. 2 (2009). Mann, Michael. The Dark Side of Democracy: Explaining Ethnic Cleansing. New York: Cambridge Books, 2005. May, Larry. Genocide: A Normative Account. New York: Cambridge University Press, 2010. McDoom, Omar. The Micro-­Politics of Mass Violence: Authority, Security, and Opportunity in Rwanda’s Genocide. (PhD thesis). London School of Economics and Political Science, 2008. Midlarsky, Manus. The Killing Trap: Genocide in the Twentieth Century. Cambridge, UK: Cambridge University Press, 2005. Mueller, John. “The Banality of ‘Ethnic War’.” International Security 25, no. 1 (2000). Nelson, Douglas and Eugene Silberberg. “Ideology and Legislator Shirking.” Economic Inquiry 25, no. 1 (1987). Overy, Richard. Interrogations: The Nazi Elite in Allied Hands, 1945. New York: Penguin Books, 2001. Piquero, Alex and Stephen G. Tibbets (eds.). Rational Choice and Criminal Behavior. New York: Routledge, 2002. Sereny, Gitta. Into that Darkness: An Examination of Conscience. London: Vintage Books, 1983. Smith, Elliot R. and Diane M. Mackie. Social Psychology. New York: Psychology Press, 2007. Staub, Ervin. The Psychology of Good and Evil: Why Children, Adults, and Groups Help and Harm Others. New York: Cambridge University Press, 2002. Staub, Ervin. The Roots of Evil. Cambridge: Cambridge University Press, 1989. Steiner, John M. “The role margin as the site for moral and social intelligence: The case of Germany and National Socialism,” Crime, Law and Social Change, 34, no. 1. (2000). Sunstein, Cass R. “Social Norms and Social Rules.” John M. Olin Law & Economics Working Paper 36. Chicago: University of Chicago, 1995. Valentino, Benjamin. Final Solutions: Mass Killing and Genocide in the 20th Century. Ithaca, USA: Cornell University Press, 2004). Verwimp, Philip. Peasants in Power: The Political Economy of Development and Genocide in Rwanda. New York: Springer, 2013. Westermann, Edward B. “Stone Cold Killers or Drunk with Murder? Alcohol and Atrocity during the Holocaust.” Holocaust and Genocide Studies 30, no. 1 (Spring 2016).

7 Killing

Introduction: from context to killing Killing requires more than the decision to participate. To enact genocide, individual decisions must be married to institutional and organizational structures that enable killing. It is not enough to create a normative context, or to tip individuals towards killing. How is the critical mass to kill mobilized? In this chapter, we will investigate these questions in terms of structures and processes of mobilization. We will also explore the forms that violence takes. Once individuals take the decision to kill, how is this decision implemented in the killing environment? We will posit a distinction between abstract killing (the idea of killing) and actual killing. This distinction between abstraction and reality leads to challenges that perpetrators must overcome in the killing environment. In the abstract, a perpetrator may want to kill all Armenians, but in actuality they may find it difficult to shoot the child in front of them. This reluctance may stem from a sense of disgust, which may be morally grounded or merely visceral. In a sense, all genocidal killing must marry abstraction to actuality. Overcoming disgust, whether visceral or moral, may also lead to a hardening of perpetrator identity. Ultimately, the final decision to kill is always made on the spot. Many perpetrators find it difficult to kill, and may seek other options. Once killing begins, the forms it takes become a fundamental aspect of the enactment of genocide. Killing involves extreme violence to advance certain ideological goals. Individual perpetrators may, however, choose a range of killing methods that extend well beyond the narrow requirements of ideological objectives. The coercion of perpetrators is also a key mechanism to ensure the mobilization of genocidal killing.

Mobilizing perpetrators Most individuals will not act to perpetrate genocide without structures and agents of mobilization. These recruit individuals, bring them to the location of killing, and demand the perpetration of violence. Genocide is not a spontaneous eruption of hate – it would not occur without elites deciding to organize perpetration.

Killing   151 In her brilliant book Killing Neighbors, Lee Ann Fujii extensively researches the mobilization of genocide at the local level in Rwanda. She argues, as I do here, that genocide is a matter of central coordination and local improvisation. Local elites and “ordinary” perpetrators alike act as entrepreneurs, interpreting and responding to the context; in Rwanda “national-­level power holders created a script and gave their backing to local-­level power holders who directed the script in their communities.”1 Centralized power, then, is an essential element for genocide – creating the moral context through its normative messages, as well as its institutionalization of moral rules and moral habits. Ambiguity around ethnic identity in Rwanda increased the power of local elites, who were able to interpret “Tutsiness” in ways that facilitated directing killings. This led to a situation in which many Hutus who “looked Tutsi” were killed, as well as Hutus seen as traitors to the Hutu cause. As one interviewee argued, “no Hutu can physically know he killed a Tutsi; no Tutsi can physically know he killed a Hutu.”2 Another perpetrator recalls this power of local elites to determine ethnic identity: Hutus in the north believe that they are more Hutu than those in the rest of the country … at the start of the genocide they were only killing Tutsis, then they began to kill Hutus who looked kind of Tutsi.3 This also created space for a number of Tutsis to take up the genocidal killing of Tutsis as a way of proving their Hutu bona fides. One of my Tutsi interview subjects was “recruited” into the Interahamwe by his friend, who was the head of the Interahamwe in his area, in order to protect him: “It was not something I planned for, it came abruptly. I was on the way home and I ran into my friend. What happened, happened. I found myself on the roadblock with a gun and I was drunk.”4 He killed four Tutsi victims during the genocide. For individuals with suspect loyalty, killing becomes an act of self-­definition. A Hutu I interviewed in Burundi similarly cloaked his identity, joining the Tutsi militia Guardiens de la Paix: “even my colleagues still think that I am Tutsi.”5 The tension between organization and randomness (genocidal serendipity) in genocidal participation is not unusual in cases where perpetrators are mobilized en masse (such as in Rwanda or in the pogroms in the East during World War II). For example, one Rwandan interviewee recalls perpetrators coming from different places to meet and then set out on killing missions, but goes on to state that they would gather others they met along the way.6 Another remembers, I was at home and then they came to take me so that I could attend a meeting – I went as a person who is going to attend a meeting. Then we did not attend a meeting but we killed people with the group.7 A degree of certainty over the target of genocidal killing is necessary for genocide. As one Rwandan perpetrator remembered, “the MRND told me about the difference between Hutu and Tutsi so we could distinguish them at

152   Killing roadblocks.”8 This delineation renders the individual members of the group subject to death. One survivor of Srebrenica argues, “if he [the perpetrator] sees a cow in the field, he would kill even the cow, because it was a Muslim cow.”9 Ethnic groups must first be delineated before they can be destroyed. Target-­ selection may therefore be influenced in part by markers of difference, such as clothing, language, accent, or physical characteristics.10 However, the treatment of each individual member of the group need not be identical to align with the goals of the genocidal regime. There must be space for autonomy at the local level, to afford a degree of empowerment to perpetrators. The diversity of means and policies at the local level also means that genocide often exhibits many inconsistencies in the treatment of individuals. These inconsistencies occur despite the power of the moral context. We must remember that individuals do not uncritically absorb the moral context. Rather, they act as interpreters and entrepreneurs. For genocidal regimes, it is essential to keep these inconsistencies to a minimum, but divergent behavior by some individuals may still be balanced by the perpetration of others. Strong social bonding and loyalty to the state may increase the possibility that an individual will perpetrate genocide. In his detailed comparative research on the onset of genocide throughout Rwanda, Scott Straus found that the most significant correlating variable for the onset of genocide in Rwanda was support for the MRND party.11 Straus concluded that, despite the dissemination of hate speech throughout Rwanda, variables such as ethnic hatred and deprivation were insignificant for the onset of genocide. This would seem to indicate that genocide is truly a “crime of obedience” driven by social bonding. Yet the choice of Tutsis as the target still indicates that prior dehumanization of Tutsis likely played a significant role. The actual mobilization of killing in Rwanda was “socially embedded in a set of dense ties.”12 In other words, civilian participation in perpetrator coalitions is often dependent on pre-­existing social networks. Moreover, the recruitment of individuals for the coalition flows partly from the propensity of individual perpetrators, as well as their position relative to others involved in killing. Members of professional or semi-­professional (military or paramilitary) groups of killers have developed moral habits and conditioning that facilitate perpetration, but individuals who are closely bonded to the regime or to other killers (through membership in organizations or interpersonal ties) may also be mobilized. Thus, mobilization may occur through state agents, party and movement members, individual self-­selection, and other local networks. In communities with substantial mass participation by “ordinary people,” there will be greater pressure to join, and a correspondingly greater likelihood of involvement in genocide. Often violence in Rwanda was organized by bringing people together in public meetings: The chief of the Interahamwe used to organize the meeting. During the meeting, they obliged the people to go to the roadblock so that no enemy could pass at the roadblocks … it was mandatory to go to the roadblock. If a

Killing   153 person refused to go to the roadblock he could be considered an accomplice or enemy … that is why I had to go.13 In many cases recruitment occurred through direct face-­to-face contact or even intimidation: “One day they [local political leaders] came to me and told me, ‘today you have to kill all the Tutsis in their houses.’ ” Similarly, “the mayor came to the area to take the people who were supposed to kill others and persuaded them. That is how I got involved in the violence.”14 Areas with higher population density likely involve denser social networks, bringing state and co-­ perpetrators into closer proximity.15 Once individuals are mobilized, there must be some level of centralized organization and control for genocide to occur. Leaders are dependent on followers to implement their decisions. For leaders (principals), there are three central problems in their relationship with their followers (agents): 1) goal variance, in which agents have different goals than the principal; 2) information asymmetry, in which agents have an information advantage over the principal, and 3) ambiguity, in which agents do not correctly understand the intentions of the principal.16 Where leaders lack adequate data (information asymmetry), agents may act on their divergent goals, and some individuals may evade full participation in killing. Similarly, if agents do not understand orders correctly, they may act in contravention of them. State propaganda is crucial in communicating appropriate behavior to agents/followers. Thus, control for genocidal states is ultimately a matter of ensuring that local elites “buy in” to the genocidal endeavor. These elites must understand the expectations of the state. Moreover, they must have the power and resources to mobilize individuals on the local level. From the perspective of leaders, a free-­rider problem may arise when group members refuse to participate in genocide. This may be reduced by stressing the primacy of ethnic identity (the narrowing of awareness), providing ideological justifications, and applying enforcement mechanisms. The presence of an ethnic norm may ease mobilization, but it is important to remember that ethnic groups are neither natural nor cohesive, and that ethnicity is more of a political and social process than a static category. Bhavnani writes by specifically delineating which actions are “right” in periods of heightened group conflict (and, by implication, which actions are “wrong” when groups cooperate), ethnic norms can effectively reduce these [enforcement and monitoring] costs and generate collective compliance.17 There appears to be a correlation between ethnicity and the escalation of violence in conflicts. This can be explained by a reduction in free-­riding (through the easier identification of in-­group members and free-­riders, as well as the group’s ability to extend credible promises to in-­group members).18 Even those who do not strongly identify with the in-­group may still join to protect their family, and avoid the risks of non-­participation.19 Ideology may also facilitate mobilization

154   Killing by clarifying incentives for involvement and risks for non-­involvement, socializing individuals to group norms (obedience, conformity, and urgency), and increasing individual commitment.20 Ideology may also increase group capacity for violence by boosting morale and identifying enemies. Enforcement is key to mobilizing violence in genocide – it ensures that even reluctant perpetrators will view the risks of non-­participation or rescue as excessive. In Rwanda for example, Des Forges notes that Hutus who protected Tutsis were “raped, beaten or killed”; these cases “were widely known in local communities and often led other Hutu to refuse or end assistance to Tutsi.”21 Beyond coercion, in-­group policing often takes the form of social ostracism. In-­group deviants “may attract particularly negative reactions,” leading some in-­group to try to re-­establish their membership credentials by denigrating the out-­group.22 The presence of an ethnic norm reduces the costs of in-­group policing. Challenges of mass mobilization may be resolved using paramilitary groups (as occurred to some degree among Serb forces in the Bosnian civil war).23 Moreover, with increasing specialization (and decreased participation) comes a lesser need for enforcement. The functional division of labor ensures a greater social margin of discretion and reduced role margin, as killing is mostly limited to specialized units.

The killing inhibition: willing executioners? Now that we have analyzed the decision-­making of perpetrators, we must also consider the act of killing itself. One of the central debates in perpetrator studies is how difficult perpetrators find it to kill. Some theorists, like Daniel Goldhagen, argue that génocidaires are “willing executioners” who are driven by hate to kill their victims. In contrast, other scholars, like Christopher Browning, argue that many perpetrators are reluctant. The evidence gleaned from my research appears to support Browning’s view. What, then, is the nature of this inhibition? In answering this question, the work of military psychologist Dave Grossman is informative. Grossman concludes that killing is quite difficult, and that many individuals will intentionally miss their targets, but that almost any person can learn to kill if conditioned to do so.24 This is corroborated by experimental evidence demonstrating that individuals are averse to harming others. This is an action aversion (i.e., distress experienced through the act itself ), rather than an outcome aversion (distress experienced through the harmful outcome).25 Violating this harm aversion may result in physiological reactions associated with negative stress.26 Randall Collins similarly finds that the dominant emotions present in violent confrontations are tension and fear; this makes the “performance of violent acts inaccurate and incompetent.”27 He also argues, with Grossman, that pathways exist to overcoming this confrontation tension. He identifies five mechanisms by which confrontation tension is reduced: (1) attacking the weak; (2) audience-­oriented, staged, and controlled fair fights (e.g., a boxing match); (3) confrontation-­avoiding remote violence (such as a drone attack); (4) confrontation-­avoiding by deception

Killing   155 (such as killing someone from behind); and (5) confrontation-­avoiding by absorption in technique (a focus on technical performance rather than the implications of the act). Many of these techniques are embedded in genocidal institutions and processes, particularly moral disengagement and the externalization of responsibility. Genocidal killing is focused on the weak – it is only possible because of the enormous power differential between perpetrators and victims. Deception is also used in killing processes in which perpetrators do not view their victims, or deceive them. An example is the pretense that the Nazi gas chambers were only showers. Absorption in technique is also present in specialized killing processes through compartmentalization, the functional division of labor, and technicalization. Finally, remote violence is also utilized in genocidal killing, by means of any process that increases the physical distance between victims and perpetrators. Killing structures serve to isolate perpetrators from their victims through social and physical distancing. SS execution squads had little or no contact with victims beforehand, to minimize feelings of guilt.28 Why is killing difficult? Perhaps the conundrum arises partly from the moral prohibition on killing that is ingrained from a very young age. But perhaps it also derives from human intelligence itself: we recognize something of ourselves in our victims, and it is this capacity for empathy that must be overcome (through dehumanization, for instance). A Japanese soldier in World War II remembers: My opponent was bound to a tree, just staring and staring at my face. I aimed at that place, and then I went and I stabbed in the left part of the chest, but after hitting, my hand slipped! I did it like this but since it was so terrifying [my hand] slipped.29 The perpetrator’s hand slipping was more than just an absence of technique; it reflected moral habits and conflicts. After witnessing the mass killing of 100 Jews in 1941, Einsatzgruppen General Erich von dem Bach-­Zelewski protested: “Look at the eyes of the men in this Kommando, how deeply shaken they are! These men are finished for the rest of their lives. What kind of followers are we training here? Either neurotics or savages!”30 We must also consider the possibility that the source of the killing inhibition in some individuals is not moral so much as it is disgust with the visceral nature of killing. Another SS officer went so far as to argue that “the people really worthy of pity were we, the liquidators, because our men were in worse nervous condition than those who had to be shot.”31 This is not empirically accurate, but one can cite Roy Baumeister’s magnitude gap – the fact that most perpetrators (of all crimes) place much less emphasis on their crimes than do victims.32 Perpetrators also tend to speak of their crimes in minimalistic, detached language.33 Yet one reason the Nazis shifted from direct killing, via executions and gas vans, to industrialized killing was the psychological toll that repeated killings took on the perpetrators. A psychiatrist with the Einsatzgruppen estimated that 20 percent of troops suffered psychological “decomposition” (half of these from

156   Killing “moral issues” and half from the “unpleasantness” of the task).34 The use of alcohol and other intoxicants is also indicative of the difficulty some perpetrators experience while killing. One Nazi doctor who worked in a concentration camp recalls: “One got drunk every time … a certain number of bottles were provided for each selection.”35 Yet the killing inhibition is not equally present for all perpetrators; those who are more morally disengaged are less likely to be constrained. A witness in the Eichmann trial stated that “the variety of their [SS  men] reactions to killing was quite extensive … some get almost hysterical,  some close to a nervous breakdown and some just go on shooting and killing.”36 In searching for overarching explanations for perpetration, we must not minimize the diversity of perpetrators’ propensities, motivations, and responses to killing.

Kill, flee, posture, resist? Individuals in the killing environment are presented with a range of options, including to kill, to flee the scene (avoid killing), to posture (simulate killing), or to resist (refuse to kill entirely). The option that a perpetrator chooses depends on his or her perceived action alternatives. Avoidance behavior (posturing or flight) is often tolerated among groups of killers. One perpetrator in Rwanda explains, “if someone presented a little excuse, we would offer to take on his part of the job that one time.”37 Another perpetrator recounts: “those who wanted to chat, chatted, those who wanted to dawdle, dawdled – if they could avoid being noticed.”38 Avoiding killing in Rwanda might mean paying a fine.39 In Burundi, a perpetrator remembers: “Sometimes during combat I hid so as not to have to take part. So, I never killed someone, but I saw it happen.”40 According to Otto Horn (an SS guard in Treblinka): I shirked as much as possible – I always volunteered for night-­duty so as to avoid the other things … I used to go and just sit behind one of the barracks and snooze. I didn’t want to see anything. Yes, I think several people felt like I did. But that was the most positive thing one could do – you know, play possum.…41 This sort of avoidance behavior seems quite common among potential perpetrators. In his history of the Jozefow massacre (in Poland during World War II), Christopher Browning reports that about a dozen men (out of 500) accepted commanding officer Major Trapp’s offer to not take part in a mass execution. Moreover, “almost tacitly everyone refrained from shooting infants or small children,” and Some policemen who did not request to be released from the firing squads sought other ways to evade. Non-­commissioned officers armed with submachine guns had to be assigned to give so-­called mercy shots “because both from excitement as well as intentionally” individual policemen “shot

Killing   157 past” their victims … others … hid in the Catholic priest’s garden until they grew afraid that their absence would be noticed … others hung around the marketplace because they did not want to round up Jews during the search. Still others spent as much time as possible searching the houses so as not to be present at the marketplace, where they feared being assigned to a firing squad. A driver assigned to take Jews to the forest only made one trip before he asked to be relieved. “Presumably his nerves were not strong enough” [commented the man who took over his driving duties].42 One participant in the massacre recalls: I myself took part in some ten shootings, in which I had to shoot men and women. I simply could not shoot at people anymore … at the end I repeatedly shot past … other comrades were also relieved sooner or later because they simply could no longer continue.43 One might attribute such avoidance behavior to the fact that many of the participants in the massacre were not violence professionals. Yet, as Dave Grossman argues, there is a long history of professional soldiers intentionally “shooting past” their would-­be victims (this was confirmed in the firing studies of SLA Marshall on the lethality of American soldiers in the Pacific theatre during World War  II).44 Moreover, there is significant evidence that destructive aggression occurs in conjunction with momentary or chronic emotional withdrawal (e.g., killing occurs when inhibition is lowered).45 However, improved training techniques (socialization and conditioning) have greatly reduced posturing and flight among soldiers in combat.46 Accordingly, one would expect that posturing and flight would be greatest in genocide among amateur, rather than well-­trained and experienced perpetrators. The presence of avoidance behavior indicates that moral costs and moral identity fluctuate over time for individual perpetrators. They retain awareness of old moral values even as these have been superseded by new, institutionalized norms. Perpetrators with low value costs (high propensity) are more inclined to kill, while other perpetrators may kill, but may also fall elsewhere on the spectrum of action alternatives – supporting, posturing, resisting, or fleeing. The range of perceived options available to each perpetrator will vary from moment to moment in accordance with long-­term factors (propensity and exposure), adaptation, and immediate contextual factors. Moreover, even in the immediate environment of killing, perpetrators maintain some scope for tactical agency. To kill, perpetrators must overcome their confrontation tension and become habituated to the act. This habituation occurs partly through participation in violence, but the devaluation and disempowerment of victims also lowers risk and creates conditions conducive to killing. Moreover, euphemisms for killing are often employed, such as “the final solution” in the case of the Holocaust, and “communal labor” or “work” (umuganda or gurakora) in Rwanda. Moreover, in Rwanda, killing was sometimes referred to as “guhiga” or “hunting,” and

158   Killing hunting terms such as gushorero (to herd) and kuvumbara (to flush out of hiding) were common in perpetrator narratives of killing.47 For some perpetrators, killing is a de-­realized, dreamlike, morally hollow act. One perpetrator in Rwanda told me that killing “was like a game. It was like pushing a chair. It was so simple,” while another said, “there was no problem killing them – they were enemies.”48 When the meaning of killing is sublimated, there is no need to overcome tension or feelings of empathy for the victims. One is just a vessel, carrying out actions devoid of meaning. This is not dissimilar to Milgram’s agentic state, but it is not a fixed condition; rather, it is one that perpetrators “drift” into. Moral doubts remain for most perpetrators, but the “dreamtime” of genocide is a state of moral disengagement. This moral disengagement is more sustained for some perpetrators than for others. The act of killing also evinces an internal rationality. Killing makes sense within the values of “the group.” According to one former LRA member: “right now you consider killing as a crime, because you are not there … when you are in the bush, there is no religion, so you consider killing as not a crime.”49 Yet the justification of killing does not make it easy for most perpetrators. Rather, it eases the shame felt after the fact. One Rwandan perpetrator recalls, with agonizing clarity, the first killing: Still I do remember the first person who looked at me at the moment of the deadly blow. Now that was something. The eyes of someone you kill are immortal, if they face you at the final instant. They have a terrible black color. They shake you more than the streams of blood and the death rattles, even in the great turmoil of dying. The eyes of the killed, for the killers, are his calamity if he looks into them. They are the blame of the person he kills.50 Violence, however, becomes easier for most perpetrators.

The boundaries of violence The state creates the demand and opportunity space for violence, but this does not explain the specific forms that violence takes. The act of killing, as noted, is an act of self-­definition. It is applied to the victims against their will, and demonstrates their utter devaluation. Violence may contribute to the dehumanization of the victims, reinforcing the social distance, the humanity gap, between them and perpetrators. The infliction of violence also serves to produce ethnic difference and social difference. In Rwanda, markers of ethnic difference were themselves targets of violence: the legs or noses of Tutsi were cut to symbolize that difference.51 One of my Rwandan victim interviewees said that her husband, who was killed in 1990 after being accused of being an RPF collaborator, was first made to eat grass (like a Tutsi cow) before being tortured to death. Killing represents an act of closure that delineates insiders and outsiders, regarding both the out-­ group and traitors within the in-­group. As one Rwandan perpetrator argues, “we killed them because they were suspected, as accomplices.”52

Killing   159 Violence is also an act of self-­definition for perpetrators presented with the opportunity to exercise control over others. A sadistic shift is often evident in genocide, in which even perpetrators who cannot be clinically classified as sadistic begin to behave sadistically. Violence shapes individual identity through mechanisms of post-­facto rationalization, as perpetrators try to make sense of the violence they have wreaked. Moreover, genocide also grants perpetrators a scope for creativity. This creativity is driven by a form of extreme narcissism, in which personal aggrandizement and “achievement” take precedence over victims’ suffering. As victims respond to their suffering this may also provide positive feedback for certain perpetrators. As Alex Hinton argues, perpetrators “remain active subjects who construct meaning and assert their self-­identity through their violent practices.”53 Perpetrators can also employ a cruel humor to signify their own moral disengagement and their mastery over victims. One perpetrator in Burundi, a former CNDD-­FDD fighter who served as a prison guard, recalls: The prisoners were asked if they wanted something warm or cold to eat for dinner. Naturally they said they wanted something warm. The guards then brought them a plate with their feces. If they didn’t eat it, they would be killed.54 Mass killing may also be encouraged by creating a “festival of cruelty” in the killing environment, one that celebrates “gratuitous and inventive cruelties.”55 This may be accompanied by a mood of elation and creativity.56 For example, one witness from the notorious Tuol Sleng prison in Cambodia recalls young guards kicking the heads of dead prisoners’ corpses, “frivolously for fun.”57 Killing may be accompanied by humor or even glee. Peter Maas reported in Bosnia “an odd enthusiasm on the part of the torturers, who laughed, sang, and got drunk while inflicting their crimes. They weren’t just doing a job, they were doing something they enjoyed …”58 Satisfactions available to perpetrators in such an environment include an enhanced reputation and standing among peers, alleviation of boredom, and a sensation of mastery. “Kills” may also represent trophies for perpetrators, whether by giving them a competitive advantage vis-­à-vis their colleagues, through the severing of victims’ body parts, or through theft of victims’ possessions. Genocide becomes a competition, and success in that competition a source of pride. Some perpetrators enjoy killing, particularly those with high levels of moral disengagement. Such enjoyment appears to be unusual and it also violates a powerful taboo; many interview subjects seemed taken aback or offended when I asked them if they ever enjoyed killing. There were, however, a handful of perpetrators who admitted to enjoying killing. For example, a perpetrator in Rwanda told me that “the reason why I loved killing was because everywhere in the country, they were killing.”59 A striking difference in perspective between victims and perpetrators is evident here: survivors often think that all perpetrators enjoy killing (“of course they enjoyed it … they enjoyed it because they wanted to take over the properties

160   Killing of Tutsis”), whereas perpetrators are either reluctant to admit to such enjoyment or truly do not feel it.60 Genocidal violence is theatrical. As Michael Ignatieff argues, genocide contains the promise of a world without enemies; the extreme and symbolic violence of genocide creates such a world in microcosm for its audience.61 It is a theatrical repudiation of the worth of the victim group, and a message to the future. Genocidal violence may include such highly symbolic and ritualistic acts such as the disembowelment of pregnant women (the destruction of the “procreative capability”), assaults on the head (the “decapitation of the intellect”), the mutilation of breasts (severing the link between mother and child and degrading the honor of the group), and forced incest (a reversal of the laws of nature and an attack on the reproductive function of the group).62 Such acts make the devaluation of the victim group and the omnipotence of the perpetrator manifest. Violence may also take the form of blasphemous acts – perhaps following existing cultural knowledge, or contributing to the general environment of normlessness, of extreme anomie. One victim recounts perpetrators “crucifying” a victim: “They used to say that Tutsi looked like Jesus and Jesus was killed on the cross, so it was like they should be killed like Jesus.”63 Acts of humiliation also contribute to the incremental habituation of perpetrators to harming victims. Franz Stangl, former commandant of Sobibór and Treblinka camps, argues that cruelty served “to condition those who had to carry out the policies. To make it possible for them to do what they did.”64 Once the victims are transformed from living beings into corpses, they are further distanced from humanity in general and perpetrators in particular. Stangl recalls: I remember [Christian] Wirth standing there, next to the pits full of blue-­ black corpses. It had nothing to do with humanity – it couldn’t have; it was a mass – a mass of rotting flesh. Wirth said “What shall we do with this garbage?” I think unconsciously that started me thinking of them as cargo.65 Jacques Sémelin argues that this psychological distancing can be accomplished either by “extreme bodily violence” or by mechanized killing, such as the gas chambers in the Holocaust.66 Individual victims who attempt to re-­humanize themselves by reaching out to the perpetrators, begging for their lives in some cases, suffer further distancing violence. In his interrogation, Ohlendorf noted: Such persons who had to carry out these shootings lost any feeling and respect for human life, and when these people who had to be shot were emotionally affected or excited, then beatings would result as a matter of course.67 Green and Ward argue that these extreme acts of violence may also demonstrate perpetrators’ commitment to their ideology and their peers.68 Within the group, the infliction of violence may be competitive and even festive, as a celebration of boundless (im)morality. In Uganda one interview subject recalled an incident of cannibalism: “We came upon a witch doctor, who was a

Killing   161 Karamojong. We roasted him, slaughtered him, and he was cooked and given to others, to eat. To our fellow soldiers.”69 A similar case of cannibalism was documented in the Ongwen trial at the ICC where witness p-­330, a former LRA fighter, remembered: “Then they [the LRA soldiers] started cutting civilians, cutting them into pieces and the body pieces were cooked in a pot and people ate the meat.”70 Though such practices seem rare in genocide (Hinton documents a case of cannibalism in Cambodia), they might occur as part of the transgressive community’s divorce from traditional norms.71 The determination of individual targets is partly a function of victim vulnerability and central coordination. But perpetrators also have considerable scope to choose individual victims (assuming they belong to the targeted group). Those targeted may be rivals or superiors. Genocide, then, becomes an arena for the expression and “resolution” of conflicts. It is an opportunity to reverse the dominant-­subordinate dynamic and inflict violent retribution for narcissistic wounds. A victim in Rwanda remembers that wealthy people were especially cruelly treated. They were “tortured, seriously, and some of them were killed like how Jesus was killed – crucified.”72 Genocide thus takes the form of an act of vengeance and “rebalancing.” Consider, for example, the following report from a Sergeant-­Major in the Wehrmacht about a mass killing in Lithuania: In the square I saw civilians, some in shirtsleeves, others wearing other types of clothing, beating other civilians to death with iron bars. I was not able to tell whether the victims were Jews. Someone, however, remarked at the time that these were the Jews who had swindled the Lithuanians before the Germans had arrived. I heard from some soldiers standing nearby, who I questioned, that the victims were being beaten to satisfy a personal desire for vengeance.73 Similarly, a perpetrator in Uganda said (when asked why people killed): “A lot of [it was for] revenge, because some of them, they give you somebody that killed your sister, that killed your family members. So, you get that courage. To have that revenge at least, and pay back.”74 In Bosnia, a camp guard reportedly mocked a prisoner who, as a judge, had issued him a traffic ticket 20 years before.75 The act of denunciation may be enough to have a neighbor removed – a tempting way to permanently resolve a dispute.76 In Rwanda, some killings targeted individuals seen as “trouble-­makers” or envied for their success.77 The resolution of interpersonal conflicts during mass violence is an under-­researched subject, although it is commonly represented in micro level studies. The fact that victims could sometimes “buy” the safety of their loved ones, at least temporarily, seems a common practice in genocide. For example, in Rwanda, “Someone who is Hutu, who has Tutsi wife, he must come to sacrifice one cow, just to pay for the death.”78 Another perpetrator in Rwanda recounts that if victims could not pay, they would take a door or a (corrugated tin) roof as payment; if victims lacked these valuable possessions, they were killed. “We only killed one person because the others gave us money.”79 Such payments

162   Killing attest to opportunities for escape, but they also demonstrate that the complete powerlessness of victims, and the total power of perpetrators. Yet genocide is not always a morality-­free zone. Rather, ideological movements may attempt to impose boundaries on unsanctioned violence in accordance with their internal ideological logic. In a secret recording between two prisoners at an interrogation center, a former Auschwitz guard notes:  “Extermination is one thing but there is no need to torture your victims beforehand.”80 Moreover, one perpetrator in Rwanda argued that, in his area, “we knew that we were supposed to destroy and burn the houses of Tutsi, not kill them.”81 Genocide does not occur in a state of anarchy, but in a situational context in which moral rules have been revised to exclude certain categories of persons. Even in such cases, certain forms of violence may be proscribed (killing innocent members of your in-­group, for example), or defined as unsavory (excessive acts of violence such as torture or rape). This is not true in all instances, of course; but varies from case to case. The group’s internal ethic delineates acceptable forms of violence, as well as imposing control over “acceptable” violence, and hence the boundaries of moral disengagement and exclusion. For example, from the perspective of the genocidal state, both profiteers and rescuers may act out of incorrect motives, whether by perpetrating for incorrect reasons, or resisting perpetration. In the Third Reich, the personal theft of Jewish property by government employees was occasionally punished by execution (although it was considered the theft of state property rather than individual property).82 Yet even incorrect acts of perpetration (such as those for material or sexual motives) may still be accepted in furtherance of group objectives. Personal enrichment may also function as a reward for perpetration in certain cases. Fascinatingly, Himmler wrote in a memorandum in May 1940 that deportation was the best method to resolve the “Jewish question,” “if one rejects the Bolshevik method of physical extermination of people out of the inner conviction as un-­German and impossible.” Hitler responded by writing “very correct” in the margin.83 Similarly, a few months later, Heydrich argued that “biological extermination … is undignified for the German people as a civilized nation.”84 The same argument (the civilized nature of the people precluding genocide) is also used to deny genocide. The Khmer Rouge and Islamic State both officially disapprove of sexual violence. However, in both the Islamic State and the Khmer Rouge, not all forms of sexual violence are prohibited; indeed, forced marriage is presented as a fulfilment of religious/ideological ideals. An Êzidî woman said of being sold in a slave market: “it was like a fashion show. The men tapped their foot when they wanted to buy that one.”85 In the Khmer Rouge, Code 6, the rule against immoral offences, prohibited all sex outside of marriage (including rape and adultery) but endorsed forced marriage.86 Sexual violence is, of course, also common in genocide, to undermine the survival and dignity of the group and individuals targeted. This sexual violence may, in some cases, also be a form of opportunistic crime, rather than strategic in nature. A female former CNDD-­informant argues that

Killing   163 without making a distinction between Hutu and Tutsi, the soldiers abducted young women to give as a present to their superiors … but there were others who raped because of hate. They first raped the women and afterwards killed them. These were two different things. It is the difference between sexual desire and rape.87 Both forms of violence here constitute rape, but the second is more explicitly focused on the immediate, rather than eventual, destruction of the group.

Forced perpetration When you are captured, they give you somebody to kill. They give you somebody to kill whether he or she had done something wrong or not. You are supposed to kill. If you refuse to kill, they said they would kill you. (Former LRA fighter88) During genocide, perpetrators may seek to force or goad non-­participating individuals into killing. Acts of forced perpetration serve the purpose of ensuring that everyone has blood on their hands. The greater the degree of mass participation, the more the killing is conventionalized, and the less perpetrators are seen as contaminated or otherwise deviant. Moreover, mass perpetration is a form of burden sharing: everyone needs to do their share of the dirty work. This diffuses shame and guilt. Forced perpetration was common practice among LRA soldiers in Uganda who, after all, were mostly abducted children rather than recruits. Three former abductees recall: I was not a real soldier but I was also given a gun. So, I was also trained using the gun, how to use it. When the fighting broke out I also fought.… When I was abducted, at first, I was kept as a prisoner and mistreated. And if you lasted for long without escaping, there were saying: “You are now becoming one of the members.” … They see that you have no interest in going back home. You are enjoying what you are doing, so they start training you.… (U01) When you escape as a prisoner and you are recaptured, you will be killed in front of the public. So that others should take that one as an example that they should not escape.89 (U03) He [Joseph Kony] was so much interested in using children … because children can easily do what you want, they are easy listeners. They can listen very well … they can be corrupted. One thing that he would do to make them so calm, whoever tries to escape, that child will not be killed. Of

164   Killing course, they know where the abducted children are from. So, they bring that child, up to near where he or she was abducted.… Then they would bring all the family members of the child and kill them together as all the children watch … and tell them that whoever tries to escape like he or she will be killed like their families.90 (U06) Most of these people fighting for the LRA, they are always returned … [when they find escapees] they kill that person, cut that person to pieces.… You have visions in you, you have bad feelings in you.… You end up fighting, you say I will fight until my last day also comes.91 Perpetrators in such groups are bound together by the act of transgressing norms – an act that often evokes a mix of shame and empowerment. Forced perpetration may also serve as a test of individual loyalty. One of my Rwandan Hutu interviewees was forced to kill because he had a Tutsi wife: In the military camp, where I was, they brought a person for me to kill, to test me. I had never killed anyone and they knew it, so they wanted me to kill in order to prove I was with them. The person I killed was not a Tutsi, but he was an Interahamwe who had stolen things … I was in the MRND and had a Tutsi wife and among the Ten Hutu Commandments you are not supposed to marry a Tutsi – so I was paralyzed.92 In this excerpt, we see forced perpetration used to generate loyalty, but also the limits on criminality during genocide (the individual executed was a perpetrator who was deemed to have broken rules by stealing from others). One villager remarked of soldiers abetting violence: “since they had guns they could have used them. Instead they led the villagers, who were then told to do the killing.”93 As already noted, the element of coercion differs from case to case. Episodes of mass violence with substantial civilian participation often contain strongly coercive demands. Within the LRA most members greatly feared the consequences of disobedience. One former LRA fighter recalls the consequences of escape: One of the captives tried to escape, and he was brought back […] and there were other who had tried to escape before, and those ones were killed. And those killing were watched, so that they can see and point to them, that whoever tries to escape will be killed as those ones.94 Another former fighter notes: Those who did not want to kill, they would be singled out. Put them in the open and order others to kill them. So that they demonstrate and become a lesson for others. So, the next you do not refuse.95

Killing   165 Another one of my interview subjects reported receiving 100 lashes for an escape and then being knifed in the throat: “I was at the point of death.”96 Once these abductees were integrated in the LRA, they felt a greater capacity to interpret and question their orders: “I had a lot of fear in defying the orders from above. But it was also possible that I would defy some orders, now that I was being incorporated in the system.”97 Fear was instilled through brutal disciplinary practices, as well as the cultivation of an aura of divine power centered on charismatic leader Joseph Kony. Even individuals who did not like Kony often feared his supernatural powers: We believe that Joseph Kony had holy spirits. Why? […] Joseph Kony would tell us, that in one-­week time UPDF [Uganda People’s Defence Forces] will attack our camp. And sometimes we would be reminded of certain thing that would happen in the next hour and then it happens.98 Even individuals who are coerced into joining violent groups may come to increasingly identify with the group. In fact, a study by Aronson and Mills concludes that participants who passed through severe initiations before joining a group had stronger ties to the group than those who had mild initiations or none at all.99 Moreover, groups who recruit by coercion are more likely to engage in sexual violence than other groups.100

Conclusion As this chapter has demonstrated, genocidal mobilization and killing are complex processes involving many possible outcomes. While genocide is certainly planned, this is not true of all individual acts of killing or participation in killing. A tension also exists between the boundlessness of killing (as a form of moral freedom) and the boundedness of the ideological goals and vision of the perpetrator group. The same normative context that authorizes genocidal killing also dictates the forms that killing should take. Yet individual perpetrators often engage in excessive acts of violence in pursuit of their own pleasure. Potential perpetrators have several options in the killing environment: not only killing and refusal to kill, but also posturing and flight. These options are constrained by the immediate margin of discretion. Where a smaller margin exists (e.g., where authority figures and other in-­group members are present as enforcement mechanisms), then individual variance is reduced. Considering the act of killing raises the subject of rationalization – what happens afterwards? How do individual perpetrators, especially reluctant perpetrators make sense of what they have done? How is repeated killing facilitated? These questions will be considered in the next two chapters.

166   Killing

Notes    1    2    3    4    5    6    7    8    9   10   11   12   13   14   15   16   17   18   19   20   21   22   23   24   25   26   27   28   29   30   31   32   33   34   35   36   37   38   39   40   41   42   43   44   45   46   47   48   49

Fujii, Killing Neighbors, 124. Interview R11 (Rwanda), Gitarama Prison, July 2009. Interview R1 (Rwanda), Kigali Central Prison, July 2009 Interview R2 (Rwanda), Kigali Central Prison, July 2009. Interview Bu4 (Burundi), Bujumbura, September 2009. Interview R14 (Rwanda), Gitarama Prison, July 2009. Interview R61 (Rwanda), Mont Kigali TIG Camp, October 2009. Interview R55 (Rwanda), Rugerero TIG Camp, September 2009. Interview B06 (Bosnia), Srebrenica, April 2009. Komar, “Variables Influencing Victim Selection in Genocide.” Straus, The Order of Genocide, 61. Fujii, 129. Interview R62 (Rwanda), Mont Kigali TIG Camp, October 2009. Interview R30 (Rwanda), Butare, August 2009; Interview R37 (Rwanda), Rwaza TIG Camp, August 2009. McDoom, “Antisocial Capital,” 890. Mitchell, Agents of Atrocity, 45. I have added the principle of “ambiguity” to Mitchell’s original typology. Bhavnani, “Ethnic Norms and Interethnic Violence,” 658. Eck, “From Armed Conflict to War,” 373. Eck, 374. Sanín and Wood, “Ideology in Civil War.” Bhavnani, 666. Hogg, “Social Categorization, Depersonalization, and Group Behavior,” 67. Mueller, “The Banality of ‘Ethnic War’,” 48. See: Grossman, On Killing. Miller, Cushman, and Hannikainen, “Bad Actions or Bad Outcomes?,” 586. Cushman, et al., “Simulating murder,” 2–72. Collins, “Micro and Macro Causes of Violence,” 11. Weale, Army of Evil, 319. Dawes, Evil Men, 66. Lifton, The Nazi Doctors, 159. Mann, The Dark Side, 247. Baumeister, Evil: Inside Human Violence, 19. Baumeister, 42–43. Mann, 271. Lifton, 193. Sivan, 20:56. Hatzfeld, Machete Season, 12. Hatzfeld, 13. Hatzfeld, 15. Interview Bu4 (Burundi), Bujumbura, September 2009. Sereny, Into that Darkness, 167. Browning, Ordinary Men, 62–63 (emphasis in original). Browning, 65. Grossman, xiii. Fromm, The Anatomy of Human Destructiveness, 147. See Grossman. Mironko, “Igitero: means and motive,” 52. Interview R2 (Rwanda), Kigali Central Prison, July 2009; Interview R29 (Rwanda), Rugerero TIG Camp, August 2009. Interview U01 (Uganda), Gulu, Uganda, July 2009.

Killing   167   50   51   52   53   54   55   56   57   58   59   60   61   62   63   64   65   66   67   68   69   70

Hatzfeld, 22. See, for example: Malkki, Purity and Exile. Interview R57 (Rwanda), Rugerero TIG Camp, September 2009. Hinton, Why Did They Kill?, 31. Interview Bu9 (Burundi), Bujumbura, September 2009. Fred E. Katz cited in Browning, 208. Katz in Browning, 208. Ysa, “Prak Khan and Interrogation at S-­21,” 14. Mueller, 55. Interview R28 (Rwanda), Rugerero TIG Camp, August 2009. Interview R66 (Rwanda), Kigali, October 2009. Ignatieff, “The Danger of a World without Enemies.” Malkki, 92–93. Interview R51 (Rwanda), Nyamata, September 2009. Sereny, 101. Sereny, 201. Semelin, Purify and Destroy, 297. Overy, Interrogations, 191. Green and Ward, State Crime, 181–182. Interview U04 (Uganda), Onyama Returnee Camp, July 2009. Prosecutor v. Dominic Ongwen, Document ICC-­02/04-01/15-T-­53-Red-­ENG (trial transcript), March 14 2017, 13.   71 Hinton, 294.   72 Interview R48 (Rwanda), Nyamata, September 2009.   73 Klee, Dressen, and Riess, The Good Old Days, 34.   74 Interview U09 (Uganda), Gulu, July 2009.   75 Pervanic, The Killing Days, 156–7. Cited in Kalyvas, The Logic of Violence, 373.   76 See: Kalyvas, 330–364.   77 André and Platteau, “Land relations under unbearable stress,” 40.   78 Interview R40 (Rwanda), Nyasenge TIG Camp, August 2009.   79 Interview R55 (Rwanda), Rugerero TIG Camp, September 2009.   80 Overy, 199.   81 Fujii, 162.   82 Lower, Hitler’s Furies, 101.   83 Valentino, Final Solutions, 175.   84 Rees, Auschwitz, 45.   85 Interview Ir7 (Iraq), Refugee Camp near Dohuk, January 2015.   86 De Langis, “This is now the most important trial.”   87 Interview Bu5 (Burundi), Bujumbura, September 2009.   88 Interview U9 (Uganda), returnee camp near Gulu, July 2009.   89 Interview U01 (Uganda), Gulu, July 2009.   90 Interview U03 (Uganda), Onyama Returnee Camp, July 2009.   91 Interview U06 (Uganda), Gulu, July 2009.   92 Interview R01 (Rwanda), Kigali Central Prison, July 2009.   93 African Rights, Rwanda: Death, Despair and Defiance, 998.   94 Interview U02 (Uganda), Onyama Returnee Camp, July 2009.   95 Interview U03 (Uganda), Onyama Returnee Camp, July 2009.   96 Interview U10 (Uganda), Gulu, Uganda, July 2009.   97 Interview U02 (Uganda), Onyama Returnee Camp, July 2009.   98 Interview U03 (Uganda), Onyama Returnee Camp, July 2009.   99 Aronson and Mills, “The Effect of Severity.” 100 Cohen, “Explaining Rape During Civil War.”

168   Killing

Works cited African Rights. Rwanda: Death, Despair and Defiance. London: African Rights, 1995. André, Catherine and Jean-­Philippe Platteau. “Land relations under unbearable stress: Rwanda caught in the Malthusian trap.” Journal of Economic Behavior and Organization 34, no. 1 (1998). Aronson, Elliot and Judson Mills. “The Effect of Severity of Initiation on Liking for a Group.” Journal of Abnormal and Social Psychology, 59, no. 2 (1959). Baumeister, Roy F. Evil: Inside Human Violence and Cruelty. New York: W.  H. Freeman, 1997. Bhavnani, Ravi. “Ethnic Norms and Interethnic Violence: Accounting for Mass Participation in the Rwandan Genocide.” Journal of Peace Research 43, no. 6 (2006). Browning, Christopher. Ordinary Men: Reserve Battalion 101 and the Final Solution. New York: Harper, 1998. Cohen, Dara Kay. “Explaining Rape during Civil War: Cross-­National Evidence (1980–2009).” American Political Science Review, 107, no. 3 (2013). Collins, Randall. “Micro and Macro Causes of Violence.” International Journal of Conflict and Violence 3, no. 1 (2009). Cushman, Fiery, Kurt Gray, Allison Gaffey, and W. B. Mendes. “Simulating murder: The aversion to harmful action.” Emotion 12, no. 1 (2012). Dawes, James. Evil Men. Cambridge: Harvard University Press, 2014. De Langis, Theresa. “ ‘This is now the most important trial in the world’: A New Reading of Code #6, The Rule Against Immoral Offences Under the Khmer Rouge Regime.” Cambodian Law and Policy Journal 3 (December 2014). Eck, Kristine. “From Armed Conflict to War: Ethnic Mobilization and Conflict Intensification.” International Studies Quarterly, 53, no. 2 (2009). Fromm, Erich. The Anatomy of Human Destructiveness. Greenwich, USA: Fawcett, 1979. Fujii, Lee-­Ann. Killing Neighbors: Webs of Violence in Rwanda. New York: Cornell, 2009. Green, Penny and Tony Ward. State Crime. London: Pluto Press, 2004. Grossman, Dave. On Killing. New York: Back Bay Books, 1996. Hatzfeld, Jean. Machete Season. New York: Picador Books, 2005. Hinton, Alexander. Why Did They Kill? Cambodia in the Shadow of Genocide. London: University of California Press, 2005. Hogg, Michael A. “Social Categorization, Depersonalization, and Group Behavior.” In Blackwell Handbook of Social Psychology: Group Processes, edited by Michael A. Hogg and R. Scott Tindale. Malden, USA: Blackwell Publishers, 2001. Ignatieff, Michael. “The Danger of a World without Enemies.” The New Republic, February 21 2009. Kalyvas, Stathis. The Logic of Violence in Civil War. New York: Cambridge University Press, 2006. Klee, Ernst, Willi Dressen, Volker Riess (eds.). The Good Old Days: The Holocaust as Seen by its Perpetrators and Bystanders. Old Saybrook, USA: Konecky and Konecky, 1996. Komar, Debra A. “Variables Influencing Victim Selection in Genocide.” Journal of Forensic Science 53, no. 1 (2008). Lifton, Robert Jay. The Nazi Doctors. New York: Basic Books, 2000. Lower, Wendy. Hitler’s Furies: German Women in the Nazi Killing Fields. London: Vintage, 2014. Malkki, Liisa H. Purity and Exile: Violence, Memory, and National Cosmology Among Hutu Refugees in Tanzania. London: University of Chicago Press, 1995.

Killing   169 Mann, Michael. The Dark Side of Democracy: Explaining Ethnic Cleansing. New York: Cambridge Books, 2005. McDoom, Omar Shahabudin. “Antisocial Capital: A Profile of Rwandan Perpetrators’ Social Networks.” Journal of Conflict Resolution 58, no. 5 (2014). Miller, Ryan, Fiery A. Cushman, and Ivar A. Hannikainen. “Bad Actions or Bad Outcomes? Differentiating Affective Contributions to the Moral Condemnation of Harm.” Emotion 14, no. 3 (2014). Mironko, Charles. “Igitero: means and motive in the Rwandan genocide.” Journal of Genocide Research 6, no. 1 (2004). Mitchell, Neil J. Agents of Atrocity: Leaders, Followers, and the Violation of Human Rights in Civil War. Houndsmills, UK: Palgrave Macmillan, 2004. Mueller, John. “The Banality of ‘Ethnic War’.” International Security 25, no. 1 (2000). Overy, Richard. Interrogations: The Nazi Elite in Allied Hands, 1945. New York: Penguin Books, 2001. Pervanic, Kemal. The Killing Days. London: Blake, 1999. Rees, Laurence. Auschwitz: A New History. London: BBC Books, 2005. Sanín, Francisco Gutiérrez and Elisabeth Jean Wood. “Ideology in Civil War: Instrumental Adoption and Beyond.” Journal of Peace Research 51, no. 2 (2014). Semelin, Jacques. Purify and Destroy: The Political Uses of Massacre and Genocide. New York: Columbia University Press, 2007. Sereny, Gitta. Into That Darkness. New York: Vintage Books, 1983. Sivan, Eyal (director). “The Specialist: Portrait of a Modern Criminal.” Documentary film. Kino Films. DVD release date: October 22 2002. 128 minutes. Straus, Scott. The Order of Genocide: Race, Power, and War in Rwanda. Ithaca, USA: Cornell University Press, 2006. Valentino, Benjamin. Final Solutions: Mass Killing and Genocide in the 20th Century. Ithaca, USA: Cornell University Press, 2004. Weale, Adrian. Army of Evil: A History of the SS. New York: NAL Caliber, 2012. Ysa, Osman. “Prak Khan and Interrogation at S-­21,” Documentation Centre for Cambodia, Searching for the Truth, no. 10 (October 2000).

8 Rationalizing killing

Introduction: reconciling with the moral context Genocide, the extermination of human beings en masse according to imputed group identity, requires the revision of moral rules. Victims must be excluded from the moral community, and perpetrator acts must be reframed as acceptable or even desirable. As previously argued, propaganda, ideology, and state structures are essential for the revision of the moral context. Moreover, through words and deeds, the genocidal state communicates ideal behavior to citizens, who may then become perpetrators. Perpetrators need ideology and propaganda to provide them with comprehension of the current situation, a framework for action, and justification for their acts. Yet there is still a need for individual perpetrators to reconcile their acts with the previous moral framework. Genocide does not require that all perpetrators are true believers; acquiescence and rationalization of wrongful acts are enough. Propaganda greatly eases perpetration by providing external justification for actions. Although internal justification is more likely to result in changed attitudes/ behavior, external justification allows perpetrators to operate without fundamentally threatening their self-­image. That is, perpetrators may believe that their actions were not the result of negative personal characteristics, but rather because of deception or compulsion by external forces.1 Thus, external justification, through propaganda, minimizes cognitive dissonance. Propaganda also communicates acceptable and expected actions. This process is facilitated by the individual’s need to frame his action in such a way that it remains consistent with his notions of moral selfhood. Distress may also be caused by the friction between conventional and emergent norms – a state that requires adaptation to avoid a psychological breakdown. Through techniques of neutralization, perpetrators can reconcile their acts in an altered moral context with past notions of right and wrong. Thus, the techniques of neutralization are more than just post-­facto rationalizations; they are criminogenic vocabularies of motive.

The genocidal techniques of neutralization Evil in the Third Reich had lost the quality by which most people recognize it – the quality of temptation.2 (Hannah Arendt)

Rationalizing killing   171 Neutralization-­drift theory Societies impose a general prohibition on private killing, with extraordinary exceptions such as self-­defense. Therefore, as a general principle, genocide is deviant, a malum in se (“bad in itself  ”) act, yet is positively normative in the societies in which it occurs. This is because genocide is a crime driven by the state or state-­like structures. The social order has been redefined through state power to render genocide an imperative. The positive normative nature of genocide for particular social communities is demonstrated by the fact that even after the end of World War II, many helpers of Jews in Poland were reluctant to reveal themselves, because they had “broke[n] the socially approved norm” of anti-­Semitism.3 Neutralization provides justifications for individuals to objectify themselves as servants of the group, institution, or state. Yet if genocide is normative within its society, it remains universally (externally) deviant. For example, an individual on trial for killing mentally ill persons in Frankfurt during World War II claimed that “this directive (the “Hitler Directive”) had partially suspended the general prohibition on killing.”4 But how exactly is the prohibition on killing neutralized, and how is morality reversed in cases of genocide and crimes against humanity? The criminological theories of Gresham Sykes and David Matza provide a solid basis for our analysis. Sykes and Matza’s neutralization-­drift theory posits that certain rationalizations operate both as post-­facto justifications for deviant conduct and as vocabularies of motive (pre-­perpetration authorization). The techniques of neutralization may provide “a culturally feasible path of action,” while primary motivation has other sources – factors such as conformity and anticipated rewards, for example. In fact, neutralization may occur even before the perpetrator has contemplated action. Sykes and Matza argue that techniques of neutralization create a state of drift in which people can move easily between delinquency and the mainstream. Deviancy occurs when the bond between action and legal norms is neutralized.5 However, genocidal behavior might not be deviant within the genocidal state. Consequently, neutralization entails a break from previously held (universal) moral values – the state of deviancy may encompass the entire genocidal state. The techniques can only be effective if they invoke and amplify pre-­existing beliefs among the population. These techniques of neutralization (such as the denial of responsibility and claim of relative acceptability) are a means for perpetrators to maintain consonance and coherence with the society in which they live. This is true both during genocide (when the techniques may constitute a vocabulary of motive) and after genocide (when the techniques may rationalize acts that are now deemed deviant within the altered moral context). In mass crimes, the motives of individual perpetrators are culturally situated and impossible to separate from social structures. Thus, rationalizations allow perpetrators to do what they have already chosen to do. If perpetrators commit deviant acts without neutralization, then they may develop a persistent “deviant identity”

172   Rationalizing killing with concomitant feelings of self-­rejection; therefore, neutralization is essential.6 Both state and horizontal propaganda drive neutralization, but it may also occur in a very subtle fashion, when authority figures condone violence with statements such as “it was understandable.”7 Moral neutralization theory is closely related to several other psychological and criminological theories. Neutralization drift theory first developed because of dissatisfaction with subcultural theory, which posited that all criminals rejected conventional values when joining criminal groups. This is one of the attractive elements of applying moral neutralization theory to genocide – the idea that all génocidaires are deviants simply does not make sense. Rather, an ongoing tension between pre-­genocidal values and the genocidal state characterizes many perpetrators. Albert Bandura argues that the selective activation and disengagement of internal controls permits different types of conduct based on the same moral standards. He writes: “self-­sanctions can be disengaged by reconstituting the conduct, obscuring personal causal agency, misrepresenting or disregarding the injurious consequences of one’s actions, and vilifying the recipients of maltreatment by blaming and devaluing them.”8 Leon Festinger’s theory of cognitive dissonance, first published the same year as moral neutralization theory, similarly deals with reconciling competing cognitions. Festinger calls the incompatibility of these cognitions a state of “cognitive dissonance” and argues that “one way to reduce dissonance is to change one’s opinions and evaluations to bring them closer in line with one’s actual behavior.”9 Increasingly, dissonance theory focuses on ego defense rather than solely cognitive consistency.10 Liau, Barriga, and Gibbs focus on individuals’ “self-­serving distortions.” Primary distortions are self-­centered attitudes, thoughts, and beliefs, while secondary distortions serve the primary distortions: they are “pre or post transgression rationalizations that serve to ‘neutralize’ conscience or guilt.”11 One of these rationalizations – assuming the worst – bears a close resemblance to Beck’s hostile framing theory. Assuming the worst is defined as “gratuitously attributing hostile intentions to others, considering a worst-­case scenario for a social situation as if it were inevitable and assuming that improvement is impossible in one’s own or others’ behavior.”12 Moral neutralization theory has been adapted to address many different offenses, including property crimes and white-­collar crimes. I have developed a modified and expanded version of Sykes and Matza’s techniques of neutralization that is particularly relevant to the crime of genocide – see Figure 8.1. The first three techniques represent the reversal of morality, while the latter seven are primarily concerned with the reduction of moral cost. In other words, the former authorize and inspire action, while the latter simply reinforce belief in the contingent moral viability of the committed act. The techniques reducing moral cost are more likely to be used as post-­facto mechanisms, although they also reflect the institutional frameworks of perpetration. Ideology and notions of duty can contribute to the reversal of morality, as can the projection of negative

Rationalizing killing   173 11 12 13 14 15 16 17 18 19 10

Appeal to higher loyalties Reversal of morality Denial of the victim (justifications) Denial of humanity Denial of responsibility Denial of injury Claim of normality Reduction of moral cost Claim of inevitability (excuses) Claim of relative acceptability Claim of inner opposition Denial of autonomy

Figure 8.1  The genocidal techniques of neutralization.

characteristics onto the victims. The fundamental importance of the techniques of neutralization is that they help to reconcile self-­concept with action. Perpetrators can maintain their view of themselves as good people if their actions are considered morally correct and justified. The techniques of neutralization may also allow the separation and coexistence (without dissonance) of the normal self and the perpetrator self. In other words, the same person commits the acts of both the normal self and the perpetrator self, but the perpetrator self does not reflect the true (normal) self, because the perpetrator self is only acting for the reasons supplied by the techniques of neutralization: “I am not doing these bad things out of choice – this is not who I am.” This prevents the development of a deviant self-­image. The distinction between perpetrator self and normal self is not sealed or cast in stone, rather it wavers and fluctuates. Even as they kill most perpetrators maintain some consciousness of wrong-­doing. Perpetrators are often aware of the negative implications of their behavior.13 In my findings, this sense of moral transgression seems to have oscillated among many perpetrators, one of whom reported that “one day we felt we were doing something wrong and the next day we didn’t.”14 However, the separation of selves witnessed in interviews may be less related to separate selves, and more to the desire to present an appropriate public self.15 The fulfillment of social norms applies, after all, to both the process of perpetration and the interview setting. However, during genocide, many perpetrators do lose sight of the wrongfulness of their behavior. For example, three Rwandan perpetrators argue: “For me, my acts were not wrong at that time;” “I thought that what I was doing was good and that it would make me famous;” and: “If I’d known I was doing something wrong I wouldn’t have gotten involved.”16 These statements, of course, may also be reflective of post-­facto regret. Yet, in the presence of the universal prohibition on killing, such sentiments are only possible with the restructuring of morality that neutralization encourages. About half of Rwandan perpetrator respondents reported that they felt at that time that their actions were wrong, with one recounting: “One day I told my friend I was doing something terrible. I felt like it was the end of the world!” and another saying, “some believed they were committing a sin.”17 These results

174   Rationalizing killing represent well the complexity of perpetrator motivation and the drift from universal moral values that neutralization facilitates – even when “drifting” perpetrators do not usually completely sever their ties to pre-­existing value systems. Neutralization originates from three sources: the state (propaganda messages, actions as models to emulate, and the externalization of responsibility), peers (social pressures and emulation), and perpetrators themselves (internal justifications to fulfil the tendency to minimize guilt for wrongful acts). Propaganda may directly provide perpetrators with messages of inspiration and authorization, yet perpetrators may also arrive at some of these techniques independently, to reduce their sense of guilt and maintain a positive self-­image. For example, perpetrators who feel shame after participating in killing may revise their own history, and come to believe that they had no choice other than to participate. Neutralization may also prevent these shameful feelings from arising in the first place. Moreover, social structures such as hierarchical organizations and group settings may ease processes of neutralization. For instance, perpetrators acting in a group may feel as though they are less responsible than if they were acting alone. Let us examine each of these neutralization techniques in turn. Appeal to higher loyalties We did bad things for a good cause.18 I was protecting my country.19

(Perpetrators in Burundi)

Perpetrators may frame their participation in atrocities as a form of altruism, believing that their deeds were done in service of others, such as their family, their ethnic group, their country, or their ideology. The appeal to higher loyalties is often a utilitarian argument – placing one moral principle or virtue over others (that is, group survival and solidarity over the sanctity of human life). When coupled with extreme dehumanization, perpetrators may come to believe that their intended victims are not fully human. The appeal to higher loyalties also involves replacing one set of moral values with another. Genocide requires that many individuals act in coordination to accomplish an overarching goal, so it is unsurprising that many perpetrators feel their violence was committed for a higher purpose. In Rwanda, about half the perpetrators interviewed felt their actions were beneficial to their ethnic group, while more than a third felt they were protecting their country. Propaganda often appeals to such sentiments, sometimes using guilt and threats to isolate and intimidate non-­participants. The state’s mobilization of genocide may itself indicate to perpetrators that their participation serves a higher purpose. Denial of the victim Because of our understanding, we thought that they were bad people, they were dangerous.20

Rationalizing killing   175 Denial of the victim justifies perpetration by depicting the victim as not a victim at all. Often this is closely related to a survival discourse – the victim is seen as actually a perpetrator and a threat to the survival of the perpetrator group. One perpetrator in Rwanda described how Hassan Ngeze’s propaganda in Kangura demanded that “we should hate the Tutsis and kill them before they killed us,” while an RTLM on-­air personality recounted her own foreboding rhetoric that “there are Tutsis outside who want to come back to kill us. There are Tutsis inside helping them. They have already sent their sons.”21 Perpetrators see themselves as victimized. In this context, all perpetrator actions are interpreted as self-­defense, hence morally legitimate. In the words of one Rwandan perpetrator: “I felt they were enemies of Rwanda and myself,” thus, according to another, “there was no problem” killing them.22 This reversal of victimization is frequently invoked in genocidal propaganda. During the Bosnian war, Serb media framed the conflict as “a preventive war against a genocidal [Bosniak] force.”23 Denial of the victim may also take the form of alleging that the victim had the power to escape his or her suffering but chose not to do so, because of a stubborn character, greed, or passivity. A final variation of the denial of victims is the “just world hypothesis”: because we live in a just world, the victims must have done something to deserve their suffering.24 “We asked ourselves what must they have done to deserve to be killed?”25 The just world hypothesis is based on both belief in the goodness of human nature and moral myopia – an inability to consider the suffering of others. There are strong overtones of obedience to authority as well. Genocide usually involves the participation of legitimate authority figures, and this strengthens just-­world assumptions about the victim. Even victims may not challenge the legitimacy of the regime’s actions; studies of concentration camp inmates in the Holocaust show that some prisoners did not challenge the justness of the system, but only argued that they had been mistakenly incarcerated.26 Just-­world thinking effectively reverses the burden of proof. Those accused by the state are guilty unless proven innocent – but demonstrating innocence is hardly possible. Institutionalized torture systems, such as at Tuol Sleng Prison, embody just-­ world thinking. The mere fact of arrest implies presumed guilt, and it is the role of the torturer to extract a confession, even if the prisoner did not actually commit a crime. Denial of humanity Denial of humanity justifies violence on the grounds of victims’ allegedly reduced or absent humanity. Once the moral context has changed, victims are excluded from the moral community and not eligible for the norm of reciprocity. For example, one perpetrator in Rwanda argued: “I heard [on the radio] that the Tutsis were killers and also that they had long tails and long ears.”27 Other génocidaires in Rwanda recounted: “they told us they [the Tutsis] were animals” and “[propaganda said] cockroaches have come. Wake up so you can fight these people … even if you go to your plantation to dig, go with a gun … kill Tutsis wherever you

176   Rationalizing killing find them.”28 Victims also recall being called “snakes” and “cockroaches” before the genocide.29 The denial of humanity (dehumanization) allows perpetrators to deviate from norms without directly assaulting the norms themselves – they are depicted as simply inapplicable in the “emergency” circumstances. Sykes and Matza also argue that “the potential for victimization would seem to be a function of the social distance between the juvenile delinquents [perpetrators] and others.”30 By denying humanity, perpetrators seek to create this social distance (the humanity gap), not only through propaganda but through discriminatory legal regimes, as around land ownership laws and citizenship requirements. Denial of responsibility We had to join the others in the group so that we would not be killed ourselves. We were supported by the government so I didn’t think about the victims. We didn’t feel pity because we were told to kill them. For those of us who can’t read or write we support the leaders. If I had tried to stop the killings I would have been killed too. The person who did something really wrong was the person who told us to kill the Tutsi. I never went because I wanted to – I was forced.31 The denial of responsibility is a denial of individual volition that underlies many other techniques of neutralization. This includes circumstances in which the perpetrator acts within a system of control or absent personal volition. It includes justifications such as intoxication (I was too drunk to consider the morality of my actions), superior edicts (I was only following orders), and coercion/duress (I was forced to do what I did). Studies show that alcohol consumption may also lead individuals to anticipate fewer consequences for bad behavior.32 Most moral and legal responsibility involves intentionality, and to deny responsibility is to deny such criminal intent. Rather than thinking about “the horrible things I am doing (or have done)” the perpetrator thinks about “the horrible things I had to do.”33 Perpetrators of genocide also frequently invoke duress and obedience to authority. My study found that 61.3 percent of Rwandan perpetrators reported that the genocide was caused by obedience to authority, while 71.9 percent of perpetrators believed that they had no choice but to kill. One Rwandan perpetrator argues: “the authorities obliged us to kill the Tutsi or else we would be killed.”34 Intoxication and duress are viable criminal defenses in certain circumstances, but superior orders do not apply in cases of genocide or crimes

Rationalizing killing   177 against humanity, which are legally considered to be “manifestly unlawful” (i.e., the perpetrator knows that they are wrongful, and the harm caused by perpetration is graver than the harm avoided). Intoxication may also be used as an excuse, removing individual agency. The effects of obviating direct responsibility are clear. In Rwanda one perpetrator, who used a bulldozer to crush as many as 2,000 people taking shelter in a church, reasoned: “I did not leave my house voluntarily to kill people. I was taken there and told to do it. So, I don’t feel so bad about it.”35 In some organizations (and states), responsibility is “free-­floating,” with every member of the organization convinced that they are acting at another’s behest.36 Free-­floating responsibility means that moral authority is incapacitated without being openly challenged or denied.37 Denial of injury I killed without knowing who was being killed … I just shot into the forest.… When you’re in combat you shoot just like everyone else … I never killed a civilian, but I am not sure. We were told to shoot everything that moved, so maybe I killed without knowing. There were cases where civilians were killed, but it was always an accident. It was involuntary.38 The denial-­of-injury technique of neutralization revolves around claims that the supposed victim was not actually hurt, that it was not the perpetrator’s intention to hurt the victim, or that the perpetrator did not cause the victim’s injury or death. Denial of injury may also assert that violence was a rightful (even lawful) form of retaliation or punishment.39 In deploying such denial techniques, the perpetrator may use euphemisms (e.g., “work” rather than “killing” in Rwanda). Historian Raul Hilberg examined thousands of Nazi documents and determined that the only time the word “kill” was used was in an order concerning dogs.40 This illustrates well the ubiquitous use of euphemistic language in genocide. Laboratory studies, meanwhile, have demonstrated the disinhibitory power of euphemistic language.41 Compartmentalization, the functional division of labor, also supports the denial of injury. It is quite possible to fulfill one’s task within an organization, such as passing on messages, and not feel responsible for the entirety of the atrocity. Even Franz Stangl (the camp commandant of Treblinka and Sobibór) was somehow able to claim: “I never intentionally hurt anyone myself.”42 Similarly, when Otto Moll, a leader of the guards in Auschwitz was asked “how many people do you estimate went through the operation which you are responsible for – how many victims?” he responded: When you use the words – “you were responsible” – I want to emphasize again that I do not wish to have that word applied in any way to the actual

178   Rationalizing killing killing of people, as I was not responsible for the actual physical ending of their lives and I will not admit that as it is not the fact.43 Here the perpetrator, Moll, leans on the functional division of labor to disclaim responsibility. Legally, individuals may still be held responsible in such circumstances under mechanisms such as the joint criminal enterprise theory of liability, complicity, and co-­perpetration. Distancing – that is, increasing the physical or emotional distance between perpetrator and victim– also eases denial of injury. One interviewee in Rwanda reported shooting someone fleeing from a mob in the leg with a bow and arrow, causing the injured person to be killed by the mob; yet he remained adamant that he was not responsible for the victim’s death.44 The effects of distance hold for all types of killing, not just genocide. Studies have shown a correlation between altitude and attitude among air crews during the Vietnam War. Those in high-­altitude bombers (such as B-­52s) spoke almost exclusively in terms of professional performance; fighter-­bomber pilots witnessed the consequences of their attacks, and had an inclination to explain and rationalize their actions; the personnel of helicopter gunships saw everything and experienced the full set of psychological consequences felt by ground personnel.45 Killers also frequently express a sense of derealization, in which they feel they are playing a game, acting in a movie, or living in a dream. One Rwandan génocidaire recounts that killing “was like watching a movie – I didn’t have much in my head,” while another claims that “It was like a game and nobody thought about the consequences.”46 Derealization may also be accomplished through alcohol or “doubling” – the creation of a somewhat separate perpetrator self that insulates the perpetrator from the reality and consequences of perpetration. Many nurses and doctors involved in the euthanasia programs in Nazi Germany claimed they were not fully aware of the consequences of their actions, that the doses of medication they were administering would kill their patients.47 Denial of injury is often facilitated by entire systems of denial, comprised of institutionalized structures, practices, and cultures that provide perpetrators with the tools for denial. These may be direct justifications, or, more indirectly, a cognitive space that allows them to deny participation. Claim of normality Killing Tutsis was not only normal, it was fashionable.48 By claiming normality, perpetrators redefine their actions as no longer transgressive, because they have become “normalized” – many other perpetrators are performing the same actions.49 Moreover, they are doing so without negative sanctions, and often receive positive reinforcement. There is a strong tendency for individuals to conform, and this certainly plays a role in perpetrator justifications. The “co-­action” effect may also operate where people work faster when they see others also working.50 During genocide, killing may very well be “normal” behavior while

Rationalizing killing   179 avoiding killing may be “shirking duty.” Killing, in fact, is regularized and institutionalized – governed by explicit or implicit norms. Another version of the claim of normality is to assert that the existence of universal, ubiquitous evil renders individual responsibility a negligible factor. This may take the form of one of two arguments: first, a metaphysical argument (the world is full of evil, so what I’m doing is not that bad or is even normal), and second, condemnation of the condemners: “you (my potential accuser) have done worse things yourself than what I am now doing.”51 When evil is no longer extraordinary, it loses its deviant quality. Claim of inevitability Who was I to stop the killing? The time for Tutsis was over. The people were supposed to die. I heard about people killing on the radio. One person could not stop this. My participation didn’t mean much – those people would be killed if I did nothing. I had no power to prevent the genocide.52 This common technique of neutralization is grounded in perceived impotence: “the crime was going to occur whether or not I participated, so my contribution to the criminal endeavor made no difference whatsoever.” The perpetrator also believes that the consequences of their decision to participate are reduced because they lacked power to stop the genocide from occurring. In his detailed recounting of the Józefów massacre, Christopher Browning quotes a policeman who argued that “without me the Jews were not going to escape their fate anyway.”53 In Rwanda, 88.7 percent of the perpetrators interviewed for this study reported that they believed, at the time, that the victims would be killed irrespective of their participation. Such beliefs further erode perpetrator agency and reduce the perceived magnitude of harmful acts (even though genocide is ultimately the product of many individual acts). As a state crime, driven by state power, genocide ensures that the alienation of individuals from a sense of agency is both profound and ubiquitous. The claim of inevitability is driven by asymmetries of power, particularly that of the state and the resulting power differential between perpetrators and victims. Claim of relative acceptability There were soldiers who behaved badly and killed people.54

180   Rationalizing killing The claim of relative acceptability reframes the perpetrator’s actions as being less deviant because they are less harmful than the alternatives. The claim includes two different sub-­techniques: the lesser harm – my violence was less extreme than others, so I am less morally culpable and less “evil”; and killing as compassion (mercy killing) – I am killing the victims to save them from future suffering. As an example of the lesser harm, Willi Mentz, an executioner at Treblinka concentration camp, argued that he was a compassionate man who never tormented his victims and complied with their wishes by leaving the decision of “who would be next” entirely to them.55 Another perpetrator, this time in Rwanda, argued that his killing was more compassionate than that of his colleagues, because he used a gun rather than the hammers and knives other perpetrators were using.56 Among Nazi perpetrators of the Holocaust, SS officer Heinrich Hesse justified a killing by stating: One of the Jewish people killed by me was a Jewish woman aged between 20 and 30, I cannot remember exactly. She was a beautiful woman. I was glad to be able to shoot her so she did not fall into the hands of the Untersturmführer. But please don’t take that to mean that I enjoyed it.… My only thought was that if I had to do something I should cause the person as little pain as possible.57 By means of this exonerating comparison (to use Bandura’s term), killing is presented as an act of mercy, albeit through an additional gendered lens. In contrast, killing as compassion transforms the acts of the perpetrator from cruel to kind. For instance, at the Józefów massacre, one man who shot only children argued: “I reasoned with myself that after all without its mother the child could not live any longer. It was … soothing to my conscience to release children unable to live without their mothers.”58 Similarly, in Rwanda, when mother of six children Juliana Mukankwaya bludgeoned other children to death “she was doing the children a favor, because they were orphans who faced a hard life.”59 Numerous examples of “mercy killing” are also found in the testimonies of Nazi doctors who participated in the Holocaust. It seems the claim of relative acceptability is almost universal. Fully 96.6 percent of the perpetrators interviewed in Rwanda maintained that other people committed worse crimes than they had. This finding is also consistent with my comparative interviews outside Rwanda. Mercy killing occurs regularly but more rarely. In Rwanda, around a quarter of my respondents reported that it was better for them to kill victims quickly to prevent them from being tortured by someone else, with one perpetrator reporting “we thought that if we didn’t kill these people they would be more cruelly killed by the Interahamwe.”60 One can also agree with the genocidal objective while disagreeing with the means employed. For example, Ervin Josting, the commander of Mainz-­Finthen Military Airfield in Nazi Germany, privately disclosed to a comrade: “I quite agree the Jews had to be turned out, that was obvious, but the manner in which it was done was absolutely wrong, and the present hatred is the result.”61

Rationalizing killing   181 Claim of inner opposition The claim of inner opposition acknowledges that an individual participated in violence, but did so in a state of inner opposition. While one still bears responsibility, this is reduced, along with its moral cost, because one did not participate eagerly. In other words, one may perform a criminal act with a criminal mind, but supposedly without having a criminal heart. In fact, such criminal conduct, some perpetrators assert, may be merely a ruse intended to dupe the “real criminals. For example, Dr. Otto Bradfisch, a former Einsatzgruppen official who had presided over the deaths of at least 15,000 people, claimed that he had been “inwardly opposed” to his actions.62 Another perpetrator in Rwanda argued that “I had to show them I was together with them but I didn’t necessarily support them.”63 The claim of inner opposition is also especially prevalent as a rationalization after the fact. The denial of autonomy I did not kill anybody.… The group in which I was in killed those six victims.64 The final technique of neutralization is perhaps the most extreme. In denying autonomy, the perpetrator completely denies his or her role in the violence, and instead claims: “I did not kill – the group killed.” In Cambodia, when prisoners asked their captors why they were being detained and tortured, they were told that Angkar (the state) was responsible and that they should “ask Angkar.”65 Similarly, Hans Frank argued in 1935: “I have no conscience; Adolph Hitler is my conscience.”66 The individual perpetrator self-­objectifies to remove their agency and moral culpability. A Rwandan perpetrator denied having killed, but when pressed, responded: “they were killed by us. It’s not the same.”67 Rwandan perpetrators often used the term “igitero” (“a group of people assembled to launch an attack”) to describe their collective attacks, with violence attributed to the group rather than the individual: “That igitero cut his wife on different parts of her body, and she died after three days.”68 Group attacks in Rwanda also reduced the risk of spiritual contamination posed by contact with victims’ blood.69 The denial of autonomy is consistent with Freud, Le Bon, and Zimbardo’s views on crowd psychology, wherein the individual ceases to exist (deindividuates) and cedes all rationality to the crowd.70 This perspective is probably overly simplistic – individuals may join a crowd to express their identity, and they often still feel autonomous and rational in crowds – but the larger the crowd, the easier it is for an individual to disclaim agency.71 Much of the violence in the Rwandan genocide was perpetrated by crowds. Based on information gathered during my interviews, it would seem that, generally speaking, the spatial position of perpetrators in the crowd correlates with their level of motivation. The most extreme and eager perpetrators (vanguard perpetrators, as discussed in Chapter 5) were at the forefront and bore the brunt of moral responsibility. Conformists, on the

182   Rationalizing killing other hand, who lingered around the fringes of crowds, felt less morally responsible. According to a Rwandan perpetrator “when I was in the group the last person did not use to kill the victims, but the first people, the ones who were in the front, were the ones who killed the victims.”72 Groups diffuse responsibility and contribute to de-­individuation, the reduction of individuality in favor of the group. The wearing of uniforms, masks, or hoods also greatly increases the de-­ individuation effect Another facet of the denial of autonomy is role conformity, whereby the perpetrator argues that they are not responsible (or acting autonomously) because they were simply performing a role. This argument is especially powerful when the role is sanctioned by legitimate authorities, as is usually the case in genocide. When asked why he had acted against his innermost convictions, Ludwig Sprauer, a doctor in Nazi Germany responsible for the euthanasia deaths of at least 3,000 of his patients, answered: “My God, I was a civil servant!”73

Assessing moral neutralization It is difficult to determine whether the techniques of neutralization are merely post-­facto rationalizations. When conducting interviews after the fact, it can be difficult to determine whether perpetrator neutralization derives from alignment with contemporary social expectations, general tendencies to minimize moral guilt, or processes originating during the genocide itself. Excuse-­making may also allow perpetrators to indicate to their audience that they remain aligned with the universal social order, even though they may have violated it.74 Indeed, the narratives of perpetrators about their past transgressions are shaped by the society that they presently inhabit. Therefore, such narratives must be examined critically, and, where possible, corroborated. Some theorists also posit that rationalizations indicate that perpetrators feel their victims are owed an explanation.75 This need would not exist without universal prohibitions on killing. Yet such theorizing minimizes the possible pre-­perpetration effects of neutralization. There is considerable empirical support for the effects of neutralization prior to perpetration including longitudinal studies.76 Based on the evidence at hand, it seems that techniques of neutralization can contribute to the production of social distance, and to the perpetrator’s quest for consistency and coherence during periods of moral rupture. Perhaps neutralization reflects a tendency to minimize moral guilt, and to avoid guilt-­inducing actions. But the specific forms such neutralization takes are often dictated by ideology and social structures.77 Moral neutralization discourses may also contribute to the mentality of perpetration as a “socially-­generated psychological space for action.”78 A particularly strong correlation exists between the techniques of neutralization and propaganda narratives. The inhuman enemy narrative meshes with the denial of humanity, while narratives of the hidden enemy, historical enemy, foreign enemy, and barbaric enemy facilitate denial of the victim. Group-­binding propaganda narratives communicate the claim of higher loyalties, while also impressing upon perpetrators that violence is normal, inevitable, and committed

Rationalizing killing   183 by the group rather than individuals. The reduction of moral costs also flows partly from propaganda narratives, but is additionally located in institutional structures that facilitate moral disengagement and an externalization of personal responsibility. These structures also limit the perceived range of available alternatives. However, we can also hypothesize that direct invocation of ideological justifications (“I killed the Jews because they were evil”) would be the narratives most intimately linked to the moral context at the time of perpetration, therefore the likeliest to disappear after perpetration. The question is whether transgression would occur without neutralization. Although no linear causality can be shown, the findings indicate that transgression is eased or even erased by the enabling moral environment provided by state ideology (messages of inspiration and authorization). This ideology of motive, when combined with a genocidal infrastructure, is difficult to resist. It is also important to note that neutralization is, to some degree, normal. As Maruna and Copes note, “taking full responsibility for every personal failing does not make a person ‘normal,’ it makes them extraordinary (and possibly at risk of depression).”79 In our schema, this statement may be truer for techniques that reduce moral costs than for those that reverse morality – the complete denial of the victim, or of their humanity, is not “normal.” A feedback loop is evident in techniques of neutralization. The criminal act is legitimized by propaganda (vocabularies of motive), but perpetration also creates the need for propaganda (rationalization). Eric Hoffer correctly notes that “we cannot pity those we have wronged, nor can we be indifferent towards them. We must hate and persecute them or else leave the door open to self-­contempt.”80 This tendency is a function of our egoistic perspective of the world. Such self-­ justification may contribute to a restructuring of morality. The more a goal becomes part of one’s self-­definition, the more it is transformed into a psychological need.81 For a technique of neutralization to be effective, it must have plausibility, and perpetrators must have a strong will to believe.82 Thus, neutralization “enables crime but it does not require it” – the will of the perpetrator remains important.83 Neutralization serves an important function in allowing perpetrators to remain “committed to the dominant normative system” while so qualifying “imperatives that violations are ‘acceptable’ if not right.”84 This maintenance of ties to the dominant normative system is a crucial factor in the perpetration of genocide. Perpetrators who followed pathological or ideological pathways to perpetration are more likely to use techniques that reverse morality, while conformist perpetrators may reduce moral costs, retaining some ambivalence about the moral rightness of genocide. For compromised perpetrators, narratives naturally focus entirely on their lack of volition, by denying responsibility. The techniques emphasized vary case by case, in relation to public discourses. The techniques not only resolve moral conflicts, but also conventionalize killing and render it comprehensible. Post-­facto systems of denial flow directly from deniability during perpetration. These systems of denial are characterized by deceptive planning and implementation, including the use of euphemisms and coded commands, the

184   Rationalizing killing concealment of human remains, and the destruction of incriminating orders.85 The facts are not completely ignored, but their meaning is altered. These systems of denial function best when we are not even aware of them, such as in cases where the truth is simply withheld. In Rwanda, there is a norm among some Hutus to “be silent” (Ceceka) when called to testify at Gacaca. Such denials draw on “shared cultural vocabularies” – collusion between people (often within organizations) to back up each other’s denials.86 Perpetrators effectively hide from the implications of their actions: present and future, personal and societal. The law “contains the seeds of its own neutralization” – in the form of justifications and excuses.87 Even if perpetrators are not aware of justifications as legal defenses, these forms of negation of the offence may be rooted in customary beliefs.88 The law simply reflects and codifies these beliefs. Dr. Valentin Falthauser, a psychiatrist who, as director of the Kaufbeuren Mental Institution, was responsible for thousands of euthanasia killings in Nazi Germany, was sentenced by a post-­war German court to only three years’ imprisonment for manslaughter. The court argued that Falthauser had committed his crimes out of “compassion” for his patients, which was “one of the noblest motives of human conduct.”89 In effect, the court endorsed his claim of killing as compassion. Certain techniques of neutralization do conform roughly to criminal defense strategies, such as the claim of universal evil with tu quoque, denial of responsibility with duress or intoxication, and denial of the victim with subjective self-­ defense. Through subjective self-­defense the perpetrator makes a legal argument (as justification) that they committed their crime out of the mistaken belief that they were acting in self-­defense. This argument was raised by a defense counsel in the Einsatzgruppen case, Dr. Rudolf Aschenauer, who said that the defendants had committed acts “in presumed defense on behalf of a third party” (the Reich) and “under conditions of presumed necessity to act for the rescue of the third party from immediate, otherwise unavoidable danger.”90 The court rejected his arguments. This is not surprising, as self-­defense generally requires an objective element, as well as demonstrated necessity and proportionality. The plausibility of such victim fear-­mongering often lies not in actual power relations, but in discourses of subterfuge and hidden conspiracies. In the case of genocidal societies, one can also say that, in some cases, acts are legally but not psychologically criminal, because of the factor of collective approval and encouragement.91

Subcultural narratives in genocide It is important to note that the discourses used by perpetrators to describe their acts are highly dependent on the context in which they are speaking. They relate partly to the overall post-­genocide context – who is held responsible for the acts of atrocity? What is the collective memory narrative? We will discuss these issues further in Chapter 10, but it is important here to consider the role of subcultures in moral neutralization. In violent subcultures, particularly those I have labelled “transgressive communities,” perpetrators may neutralize being good

Rationalizing killing   185 rather than being bad. These perpetrators have already rejected conventional values (at least to some degree); thus, the incentives are greater to celebrate behavior that is conventionally viewed as wrong (like killing) and to neutralize conventionally good behavior (such as showing empathy towards victims). Discourses that may be subcultural at one point (such as those espoused by the Nazi Party in Germany before they assumed power) may eventually become conventional, as they are mainstreamed and normalized through state power and propaganda. We cannot assume that all perpetrators will feel the pull of conventional values. In other words, not all perpetrators will experience a state of cognitive dissonance requiring neutralization. Some perpetrators reject conventional values by killing, and indeed some have rejected conventional values even before they kill. This may be especially true for killers in paramilitary organizations, who are often recruited from criminal ranks. The moral context of genocidal states is not always completely coherent, much less universally understood or adhered to. Even as the state attempts to reduce individual variance through consistent messaging and institutional structures, moral norms unique to specific (subcultural) groups still obtain. The norms might differ more for transgressive communities than for civilian perpetrators. This could reflect standards of professionalism, but such organizations could conversely use violence in ways proscribed for civilians. Moreover, individuals in contexts where violence is mainstreamed may well not see themselves as active perpetrators of violence, while individuals in deviant subcultures may celebrate their acts of violence.92 The layering of violence in society – from transgressive communities, to soldiers, to civilian perpetrators, to supportive bystanders – mirrors the layering of moral norms. Despite overarching principles (e.g., those communicated through state propaganda), norms are not understood uniformly across society. Processes of individual translation are present, and groups often establish their own norms and interpretations of overarching norms.

Conclusion: tools of adaptation The techniques of neutralization are a crucial link between ideology and action. They are tools of adaptation that allow perpetrators to adjust to an evolved moral context, thereby maintaining their adherence to state authority and ties to the group. Paradoxically, perpetration involves contravening fundamental values (such as the prohibition of killing) which contribute to self-­image. Yet this occurs, in part, for the purpose of maintaining ties to the group, which is another fundamental facet of self-­image (the mirror self ). Moreover, the groups involved in genocide (racial, ethnic, national, and religious) tend to be ascribed at birth and relatively fixed. Isolation from one’s own group, particularly in the highly-­polarized context of genocide, is not only tremendously stressful but potentially lethal.93 Consider also that, by means of propaganda and techniques of neutralization, the in-­group seeks to persuade its members that previously held moral rules are no longer applicable. Ultimately, it becomes easier to neutralize or ignore old moral norms than to isolate oneself from the group.

186   Rationalizing killing The techniques of neutralization also broaden the range of perceived options available to perpetrators. As such, they serve a “decision-­priming” function. Perpetrators act within structures of legitimate authority against an already dehumanized victim group. When confronted by overwhelming social pressures, and state power, most perpetrators will seek rationalizations to ensure they maintain a consistent sense of self in the face of profound upheaval.

Notes   1 Staub, The Roots of Evil, 209–210.   2 Arendt, Eichmann in Jerusalem, 150.   3 Gross, “Opportunistic Killings and Plunder,” 190.   4 De Mildt, In the Name of the People, 122.   5 Matza, Delinquency and Drift, 101.   6 Kaplan, “Deviant Identity and Self Attitudes,” 111.   7 This was the statement made by Gujarat Chief Minister Narendra Modi in response to the 2002 pogrom against Muslims in Gujarat.   8 Bandura, et al., “Mechanisms of Moral Disengagement,” 364.   9 Festinger, A Theory of Cognitive Dissonance, 98. 10 Greenwald and Ronis, “Twenty years of cognitive dissonance.” 11 Liau, Barriga, and Gibbs, “Relations between self-­serving cognitive distortions,” 301. 12 Liau, Barriga, and Gibbs, 302. 13 Scully and Marolla, “Convicted Rapists’ Vocabularies of Motive,” 530. 14 Interview R37 (Rwanda), Rwaza TIG Camp, August 2009. 15 See Goffman, “On Face-­work.” 16 Interview R73 (Rwanda), Kigali Central Prison, October 2009; Interview R55 (Rwanda), Rugerero TIG Camp, September 2009; and Interview R58 (Rwanda), Rugerero TIG Camp, September 2009. 17 Interview R78 (Rwanda), Nyamata TIG Camp, October 2009; Interview R29 (Rwanda), Rugerero TIG Camp, August 2009. 18 Interview Bu2 (Burundi), Bujumbura, September 2009. 19 Interview Bu3 (Burundi), Bujumbura, September 2009. 20 Interview R62 (Rwanda), Mont Kigali TIG Camp, October 2009. 21 Interview R39 (Rwanda), Nyarusenge TIG Camp, August 2009; Interview R73 (Rwanda), Kigali Central Prison, October 2009. 22 Interview R73 (Rwanda), Kigali Central Prison, October 2009; and Interview R29 (Rwanda), Rugerero TIG Camp, August 2009. 23 Irwin and Becirivic, “Visegrad in Denial,” 6. 24 Lerner and Simmons, “Observer’s Reaction to the ‘Innocent Victim’.” 25 Interview R35 (Rwanda), Rwaza TIG Camp, August 2009. 26 Fromm, The Anatomy of Human Destructiveness, 85. 27 Interview R4 (Rwanda), Kigali Central Prison, July 2009. 28 Interview R36 (Rwanda), Rwaza TIG Camp, August 2009; Interview R15 (Rwanda), Gitarama Prison, July 2009. 29 Interview R48 (Rwanda), Nyamata, September 2009. 30 Sykes and Matza, “Techniques of Neutralisation,” 665. 31 Interview R60 (Rwanda), Mont Kigali TIG Camp, October 2009; Interview R55 (Rwanda), Rugerero TIG Camp, September 2009; Interview R67 (Rwanda), Hindiro TIG Camp, October 2009; Interview R11 (Rwanda), Gitarama Prison, July 2009; Interview R17 (Rwanda), Gitarama Prison, July 2009; Interview R26 (Rwanda), Gisenyi Prison, August 2009; and Interview R79 (Rwanda), Nyamata TIG Camp, October 2009. 32 Littman and Levy Paluck, “The Cycle of violence,” 92.

Rationalizing killing   187 33 Semelin, Purify and Destroy, 255. 34 Interview R67 (Rwanda), Hindiro TIG Camp, October 2009. 35 African Rights, Rwanda: Death, Despair and Defiance, 405. 36 Zygmunt Bauman, Modernity and the Holocaust (Ithaca, New York: Cornell University Press, 2000), 163. 37 Interestingly, in the post-­war German Euthanasia Case the court reasoned that the individual defendants suffered from “inertia of the will” for conforming with the goals of the Nazi state, whereas in the most famous Nazi propaganda film, “The Triumph of the Will,” the ideal of will was embodied in a collective endeavor of individuals acting in service of the organic state. See: Riefenstahl. 38 Interview Bu9 (Burundi), Bujumbura, September 2009; and Interview Bu6 (Burundi), Bujumbura, September 2009. 39 Sykes and Matza, 668. 40 Waller, Becoming Evil, 189. 41 Bandura, et al., 365. 42 De Mildt, 300. 43 Overy, Interrogations: The Nazi Elite in Allied Hands, 393. 44 Interview R79 (Rwanda), Nyamata TIG Camp, October 2009. 45 Lifton, The Nazi Doctors, 495. 46 Interview R78 (Rwanda), Nyamata TIG Camp, October 2009; and Interview R22 (Rwanda), Tumba TIG Camp, August 2009. 47 See De Mildt, 181; and Lifton. 48 Interview R4 (Rwanda), Kigali Central Prison, July 2009. 49 For more on the Claim of Normality see: Coleman, The Criminal Elite. 50 Gillian Butler and Freda McManus, Psychology: A Very Short Introduction (New York: Oxford University Press, 1998), 112. 51 Sykes and Matza, 668. 52 Interview R01 (Rwanda), Kigali Central Prison, July 2009; Interview R02 (Rwanda), Kigali Central Prison, July 2009; Interview R18 (Rwanda), Ruhengeri Prison, August 2009; Interview R39 (Rwanda), Nyarusenge TIG Camp; Interview R62 (Rwanda), Mont Kigali TIG Camp, October 2009; and Interview R76 (Rwanda), Kigali Central Prison, October 2009. 53 Browning, Ordinary Men, 73. 54 Interview Bu3 (Burundi), Bujumbura, September 2009. 55 De Mildt, 263. 56 Rwandan Interview R4 (perpetrator), unemployed/Interahamwe, Kigali Central Prison, July 7 2009. 57 Klee, Dressen, and Riess, The Good Old Days, 201. 58 Browning, 73. 59 Baumeister, Evil: Inside Human Violence, 308. 60 Interview R37 (Rwanda), Rwaza TIG Camp, August 2009. 61 Neitzel and Welzer, Soldaten, 125. 62 Arendt, 127. 63 Interview R17 (Rwanda), Gitarama Prison, July 2009. 64 Interview R70 (Rwanda), Hindiro TIG Camp, October 2009. 65 Hinton, Why Did They Kill?, 238. 66 Overy, 500. 67 Interview R30 (Rwanda), Butare, August 2009. 68 Mironko, “Igitero: means and motive in the Rwandan genocide,” 51 and 54. 69 Jessee, Negotiating Genocide, 169. 70 See, for example, Zimbardo, “The Human choice.” 71 Latane and Darley, for example, conclude that individuals are much less likely to feel responsible as independent agents when in a group. See: Latane and Darley, The Unresponsive Bystander. See also: Reicher, “The Psychology of Crowd Dynamics,” 203.

188   Rationalizing killing 72 Interview R60 (Rwanda), Mont Kigali TIG Camp, October 2009. 73 The court released him after only a short term because he didn’t have a “criminal personality” and had acted out of a “misunderstood conception of duty.” See De Mildt, 106. 74 Maruna and Copes, “What Have We Learned,” 252. 75 Maruna and Copes, 253. 76 Alvarez, “Adjusting to Genocide,” 139–178. 77 This tendency to minimize moral (and legal) guilt has been shown in numerous studies including one that demonstrated, through anonymous personal histories, that child molesters grossly underreported their crimes even after conviction. See: Goldhagen, Worse than War, 171. 78 Gudehus, “Social Theoretical Considerations.” 79 Maruna and Copes, 227. 80 Hoffer, The True Believer, 95–96. 81 Staub, 15. 82 Baumeister, 307. 83 Minor, “Techniques of Neutralisation,” 295–318. 84 Sykes and Matza, 667. 85 Cohen, States of Denial, 14. 86 Cohen, 64. 87 Matza, 61. 88 Matza, 74. 89 De Mildt, 99. 90 De Mildt, 10. 91 Le Bon, The Crowd, 105. 92 Katherine Blee’s work on the Ku Klux Klan is informative here. See: Blee, “Questions of Methods and Ethics,” 6. 93 In some cases, this isolation may arise from circumstances beyond personal choice – for example, individuals who may be “stranded” between groups due to intermarriage.

Works cited African Rights. Rwanda: Death, Despair and Defiance. London: African Rights, 1995. Alvarez, Alexander. “Adjusting to Genocide: The Techniques of Neutralisation and the Holocaust.” Social Science History 21, no. 2 (Summer 1997). Arendt, Hannah. Eichmann in Jerusalem: A Report on the Banality of Evil. New York: Penguin Books, 2006. Aronson, Elliot. The Social Animal (8th edition). New York: Worth Publishers, 1999. Bandura, Albert, Claudio Barbaranelli, Gian Vittorio Caprara, and Concetta Pastorelli. “Mechanisms of Moral Disengagement in the Exercise of Moral Agency.” Journal of Personality and Social Psychology 71, no. 2, (1996). Baumeister, Roy F. Evil: Inside Human Violence and Cruelty. New York: W.  H. Freeman, 1997. Blee, Kathleen M. “Questions of Methods and Ethics in the Study of U.S. White Supremacists.” Paper presented at the “Approaching Perpetrators Workshop, May 14–16, University of British Columbia, 2014. Browning, Christopher. Ordinary Men: Reserve Battalion 101 and the Final Solution. New York: Harper, 1998. Coleman, James W. The Criminal Elite: The Sociology of White-­Collar Crime (5th edition). New York: St. Martin’s, 2002.

Rationalizing killing   189 Cohen, Stanley. States of Denial: Knowing about Atrocities and Suffering. Cambridge: Polity Press, 2008. De Mildt, Dick. In the Name of the People: Perpetrators of Genocide in the Reflection of their Post-­War Prosecution in West Germany. The Hague: Martinus Nijhof, 1996. Festinger, Leon. A Theory of Cognitive Dissonance. Stanford: Stanford University Press, 1957. Fromm, Erich. The Anatomy of Human Destructiveness. Greenwich, USA: Fawcett, 1979. Goffman, Ervin. “On Face-­work.” Psychiatry: Interpersonal and Biological Processes 18, no. 3 (1955). Goldhagen, Daniel. Worse than War. New York: PublicAffairs, 2009. Greenwald, A. G. and D. L. Ronis. “Twenty years of cognitive dissonance: Case study of the evolution of a theory.” Psychological Review, 85, no. 1 (1978). Gross, Jan T. “Opportunistic Killings and Plunder of Jews by their Neighbors – a Norm or an Exception in German-­occupied Europe?” Amsterdam: NIOD Institute for War, Holocaust and Genocide Studies, 2014. Gudehus, Christian. “Social Theoretical Considerations towards an Understanding of War Crimes.” In War Crimes Trials and Investigations: A Multidisciplinary Introduction, edited by Jonathan Waterlow and Jacques Schuhmacher. London: Palgrave Macmillan, 2017. Hinton, Alexander. Why Did They Kill? Cambodia in the Shadow of Genocide. London: University of California Press, 2005. Hoffer, Eric. The True Believer: Thoughts on the Nature of Mass Movements. New York: Harper Perennial Modern Classics, 2002. Irwin, Rachel and Edina Becirivic. “Visegrad in Denial.” London: Institute for War and Peace Reporting, December 1998. Jessee, Erin. Negotiating Genocide in Rwanda: The Politics of History. London: Palgrave Macmillan, 2017. Kaplan, Howard B. “Deviant Identity and Self Attitudes.” In Encyclopedia of Criminology and Deviant Behavior, edited by Clifton D. Bryant. Philadelphia: Brunner-­ Routledge, 2001. Klee, Ernst, Wili Dressen, and Volker Riess (eds.). The Good Old Days: The Holocaust as Seen by its Perpetrators and Bystanders. Old Saybrook: Konecky and Konecky, 1991. Latane, Bibb and John Darley. The Unresponsive Bystander: Why Doesn’t He Help? New York: Appleton, 1970. Le Bon, Gustav. The Crowd. London: Dover Publications, 2002. Lerner, Melvin and Carolyn Simmons. “Observer’s Reaction to the ‘Innocent Victim’: Compassion or Rejection?” Journal of Personality and Social Psychology 4, no.  2 (1966). Liau, Albert K., Alvaro Q. Barriga, and John C. Gibbs. “Relations between self-­serving cognitive distortions and overt vs. covert antisocial behavior in adolescents.” Aggressive Behavior 24, no. 5 (1998). Lifton, Robert Jay. The Nazi Doctors. New York: Basic Books, 2000. Littman, Rebecca and Elizabeth Levy Paluck. “The Cycle of Violence: Understanding Individual Participation in Collective Violence.” Advances in Political Psychology 36, suppl. 1 (2015). Maruna, Shadd and Heith Copes. “What Have We Learned from Five Decades of Neutralization Research?” Crime and Justice 32 (2005). Matza, David. Delinquency and Drift. New Brunswick, USA: Transaction, 2009.

190   Rationalizing killing Minor, William. “Techniques of Neutralisation: A Reconceptualization and Empirical Examination.” Journal of Research in Crime and Delinquency, 18, no. 2 (1981). Mironko, Charles. “Igitero: means and motive in the Rwandan genocide.” Journal of Genocide Research 6, no. 1 (2004). Neitzel, Sonke and Harald Welzer. Soldaten. Frankfurt: Verlag Fischer, 2011. Overy, Richard. Interrogations: The Nazi Elite in Allied Hands, 1945. New York: Penguin Books, 2001. Reicher, Stephen. “The Psychology of Crowd Dynamics.” In Blackwell Handbook of Social Psychology: Group Processes, edited by Michael A. Hogg and R. Scott Tindale. Malden, USA: Blackwell Publishers, 2001. Riefenstahl, Leni (director). Triumph of the Will. 1935, re-­released on DVD by Synapse Films, 2006. Scully, Diana and Joseph Marolla. “Convicted Rapists’ Vocabularies of Motive: Excuses and Justifications.” Social Problems 31, no. 5 (June 1984). Semelin, Jacques. Purify and Destroy: The Political Uses of Massacre and Genocide. New York: Columbia University Press, 2007. Staub, Ervin. The Roots of Evil. Cambridge: Cambridge University Press, 1989. Sykes, Gresham M. and David Matza. “Techniques of Neutralisation: A Theory of Delinquency.” American Sociological Review 22, no. 6 (December 1957). Waller, James. Becoming Evil. New York: Oxford University Press, 2002. Zimbardo, Philip. “The Human choice: Individuation, reason, and order versus deindividuation, impulse, and chaos.” In Nebraska Symposium on Motivation 1969, edited by W. J. Arnold and D. Levine. Lincoln: University of Nebraska Press, 1970.

9 Coping with killing

Introduction: adjusting to genocide Genocide is a totalizing process. Given its aims to destroy the victim group, genocide requires the removal of obstacles to perpetration, as well as the repetition of perpetration. In this chapter, we will examine the progression of genocide. How does genocidal killing evolve over time on the individual and the collective level? Participation in violence may itself be transformative for individual and group identities. Individual perpetrators must adjust to killing. This adjustment is a matter of habituation, but it may also involve reframing the understanding of perpetration to maintain individual identity. This conventionalization of killing also leads to a further radicalization of the private and public spheres. For many perpetrators, killing becomes easier over time as they habituate to violence, are rewarded for it, and adjust to having transgressed moral norms. The resistance of reluctant perpetrators is overcome through intimidation, rationalization, and coercion. But not all perpetrators must overcome significant moral conflicts. Genocide may be empowering and enjoyable for some perpetrators – an exciting state of normlessness and freedom. This chapter will analyze systemic and individual mechanisms that sustain genocide and facilitate recidivism.

The progression of genocide Escalating violence Genocide often unfolds in an incremental manner: from persecution, to isolated acts of violence, to mass violence. These stages may also represent progressive habituation to moral disengagement and escalating violence. As noted elsewhere, genocidal ideology is normally not focused on mass killing as an end in itself. Rather, violence is seen a transitory stage to achieve “positive” objectives. The process of genocide is neither linear nor inevitable. Instead, genocidal violence is often preceded by both periods of escalation and de-­escalation. As we have established in Chapter 2, pre-­genocidal contexts are usually characterized by significant strain. In a sense, then, genocide and crimes against

192   Coping with killing humanity are at the most extreme end of the concentration and exercise of coercive power in response to threat. If prejudice is actualized as violence, it may occur first through proxies (non-­government forces). By using such proxies, deniability is maintained. The use of violence in this early moment may act as a kind of experiment – gauging public support and opposition to violence against the targeted group. If violence is to become more systematic, it is usually elites who are targeted first. The targeting of these individuals renders victims more vulnerable, while simultaneously communicating that average members of the victim group may protect themselves from exposure to violence through deference to the regime. As violence escalates and expands into new categories of victims and perpetrators, there is a concomitant escalation in propaganda discourses. Victim culpability and threat are brought into focus, and eventually the existential threat posed by the victim group necessitates the endorsement of their destruction. The beginnings of mass violence generally involve violence specialists; these same individuals may subsequently contribute to mass participation through inciting and intimidating others to participate. In some cases, forced perpetration is used as a means of expanding the circle of perpetrators. Such an expansion of perpetration is necessary for genocide on a logistical level. However, it also contributes to shifting the normative (moral) context towards the conventionalizing of killing (thereby reducing the risks of participation). As argued elsewhere in this book, genocide is both a moment of rupture (the breaking of conventional norms) and a form of continuity (manifested in fidelity to conventional institutions). Genocide may be relatively participatory (as in Rwanda, where much of the killing was locally directed), or it may be more specialized (as in Nazi Germany, where there were at least 10,000 concentration camps, including ghettos, with hundreds of thousands of personnel working at killing institutions).1 However, even the Holocaust involved localized violence; it was a “mosaic composed of discrete episodes, improvised by local decision-­makers.”2 This is always true of genocide: local perpetrators, with myriad motives, each commit discrete acts (such as killing or rape). The accumulation of such acts, with some central coordination and messaging from the state, constitutes genocide. Mere mob violence is not enough to effect total annihilation. The sustained, organized nature of genocide is one of the characteristics that distinguishes it from other forms of mass violence, such as pogroms. Sabini and Silver argue: “thorough, comprehensive exhaustive murder required the replacement of the mob with a bureaucracy, the replacement of shared rage with obedience to authority.”3 However, we must be careful not to over-­universalize the Holocaust. The case of Rwanda demonstrates that bureaucratic control can exist comfortably alongside mob violence, and rage alongside obedience to authority. These elements are not mutually exclusive. In Rwanda, the killings originated with a central kernel of extremists who exercised a degree of control and influence over their local counterparts. This degree of control is reflected in the consistency of genocidal propaganda, and the involvement of state agents (such as soldiers and administrators). The death of the

Coping with killing   193 president presented an opportunity space that was then exploited by extremists. Moderates were marginalized or killed as extremists from other areas of the country moved into areas where killing was resisted.4 The killings were at first limited to Kigali and “a few communes in Gisenyi (the president’s commune), Gikongoro, and Kibungo”; killing did not spread to other areas until after visits by the interim authorities.5 For example, genocide never occurred in Giti Commune, because the RPF entered the area before the hardliners gained the upper hand.6 Consider, for example, the following representative accounts of Rwandan political leaders detailing the process of genocide in Rwanda. An MRND official recalls: “It all started in October 1993 with a meeting in Nyamirambo Stadium. After this the intensive training of Interahamwe in camps began … in December 1993 lists were prepared of people who had to die.”7 The Interahamwe were indoctrinated at mass rallies and attended three-­week long training camps.8 Similarly, a sector counsellor was at lots of political meetings where they were talking about genocide. It was my duty to call all local leaders at the cell level and order them to kill. A list had been prepared in 1991 – the people on the list were mostly elites, wealthy people, and Hutu opposition.9 The crash of Habyarimana’s plane was the catalyst to enact these extremist plans. Another sector counsellor remembers: “I heard before the genocide [in meetings] that after Habyarimana was killed there would be no Tutsis left in Rwanda … When I heard that Habyarima’s plane had crashed I thought ‘now it’s time’.”10 Perhaps the full potential of genocidal destruction was only realized by the extremists once violence was already underway. A “northern political-­ military faction” progressively seized control of the state, possibly without “any central decision-­making” involved.11 The Rwandan government also used previously existing infrastructure, such as security committees, bureaucracy, and the media. This included posting public notices in traditional sites for such communications. As explained in Chapter 3, there was also a cultivated equivalency between genocidal killing and umuganda (the norm of regular communal labor).12 Public notices included work details (with schedules and names of required participants), just as with umuganda. The use of umuganda-­like structures also demonstrated continuity between the radical state committing genocide and the Rwandan state. Burgomasters also dispatched subordinates for house-­to-house searches. As noted previously, authority figures of some kind were often present at killings. Anthropologist Charles Mironko also argues that killing in Rwanda drew from pre-­existing social practices centered on mobilizing civic defenses in cases of attack (guterwa): When a person is attacked, s/he shouts for help and those who live on the same hill or hamlet are socially and morally obligated to come to that person’s aid. But during the genocide, this practice was used not to help the victims, but rather to assemble the attackers.13

194   Coping with killing Thus, perpetration in Rwanda drew from pre-­existing social practices for civic mobilization. Bosnia also exhibited an effort to align local actors with the state: central authorities (such as the state security apparatus) sent officials to each municipality looking for trusted allies, who were then told that the Muslims needed to be cleared out for security reasons.14 This trusted person could be the mayor, the local police chief, or a hospital administrator. There often seems to have been little in the way of a formal chain of command in evidence. In both Rwanda and Bosnia, genocide involved a mix of civilians and violence specialists, including paramilitaries. When violence begins, it may be limited to marginal members of the out-­ group (such as thieves), or “treacherous elements” such as opposition leaders and intellectuals. A Bosniak in Višegrad remembers: “In the beginning when they entered Višegrad they were asking and looking for intellectuals and they were their first targets … and leaders of course.”15 Eventually authorities may argue that the entire “tree” must be removed root and branch to prevent the spread of disease. In fact, the criminalized groups may cease being “them” and become “it” at some point, as they are dehumanized and objectified. The first movers in genocidal mass violence are often those with high propensity, including those tied closely to the state – such as formal agents and ideologues (including party members). These individuals may be most readily mobilized due to their socialization, institutionalization, and moral disengagement. To this list, we might add the “unlucky” – individuals who are in the wrong place at the wrong time, and thus vulnerable to the exercise of state authority. As violence spreads from paramilitaries and non-­state actors to the state and civilians, killing becomes increasingly conventionalized. Society and the state itself also become radicalized. Radical groups, such as paramilitary forces or extremist political parties, may themselves come to constitute the state. These non-­state actors may become de facto state agents in the context of genocide. For example, in Rwanda, the Interahamwe and the RTLM, both private actors, came increasingly to be associated with the will, authority, and legitimacy of the state. This shift in the normative and legal framing of violence may also contribute to shifts on the individual level, whereby ordinary people come to behave sadistically when placed in situations of omnipotent power over their devalued victims. Legal theories of genocide are often built on fictions – that genocide consists of a clear plan prepared years in advance by authorities, who then command their subordinates to participate. Such models of genocide are oddly ahistorical, resembling early intentionalist perspectives on the Holocaust. The fictional nature of these genocides is borne out by the difficulty courts often have in finding evidence to substantiate these kinds of theories. The failure to establish wide-­ranging conspiracies often leads to a resort to remote liability theories, such as the “type 3” joint criminal enterprise. The problem with judicial theories of responsibility is inherent in the (often fruitless) search for evidence of clear genocidal commands and intent. This quest assumes that violence is a highly-­structured process, rather than one rooted in normative expectations and interpretations.

Coping with killing   195 The “problem” of the decent Jew next door: overcoming resistance to killing Even in criminogenic states where the state has authorized killing and created structures of moral disengagement, there remains a possibility of rekindling former collegiality, of remembering old loyalties. For those who care about humanity, this is a hopeful sign. Yet for the perpetrating state, it represents goal variance: a problem to be overcome. Consider, for example, the following 1938 Report from Saxony: When the people read of the measures taken against the Jews in the big cities, then they approve of them. But when a Jew of their circle of close acquaintances is affected, then the very same people moan about the terror of the regime. Then compassion stirs in them again.16 In his Posen speech in October of 1943 Himmler also derided this tendency, saying: “The Jewish people is to be exterminated,” says every party member. “That’s clear, it’s part of our program, elimination of the Jews, extermination, right we’ll do it!” And then they all come along, the 80 million good Germans, and each one has his decent Jew. “Of course, the others are swine, but this one is a first-­class Jew’.”17 In Rwanda, Alison Des Forges observed that: “some [reluctant Hutus] might massacre strangers in churches or at barriers, knowing only that they were Tutsi, and refuse to attack neighbors, knowing that they were Tutsi but knowing also that they were not enemies.”18 One Rwandan génocidaire argued: “We saved the lives of some people because you can’t kill someone you love.”19 Yet this is not entirely true, as the general prevalence of intimate violence shows. How can people simultaneously act as murderers and rescuers? Individuals are complex: they can simultaneously demonize and persecute a group while exempting individual members of the group. Bauman argued that the source of “the problem of the decent Jew” lies in separating the metaphysical Jew (the collective cognition) from the personal Jew (the individual known to the perpetrator).20 One could use the same terms in other genocides: the metaphysical Tutsi (insidious underminer of the Hutu race, plotting to dominate the African Great Lakes), or the metaphysical Armenian (a greedy, treacherous threat to the Turkish people). To make the metaphysical other stick to the personal other, a complete social death of the victim group must be engineered. Hans Mommsen writes: Only after cumulative discrimination legislation had pressed Germany’s Jews into the role of social pariahs, completely deprived of any regular social communication with the majority population, could deportation and extermination be put into effect without shaking the social structure of the regime.21

196   Coping with killing The persistent personalization of the victims, however, allows some perpetrators to participate in genocide as both perpetrator and rescuer (killing strangers while rescuing family members, for example). Moreover, perpetrator resistance to killing personalized victims can also be overcome by tasking perpetrators with killings outside their home region (as was common in Cambodia and Rwanda). Even though genocide is a sustained campaign, individual intent may fluctuate during a genocide. Consider the case of Goran Jelisić, a Bosnian Serb who held a position at the Luka camp during the Bosnian war. Jelisić allowed a Muslim to escape after playing Russian roulette with him, in spite of expressing the desire to “exterminate” all Bosnian Muslims and calling himself the “Serbian Adolf.”22 Moreover, Clément Kayishema (the prefet of Kibuye) incited genocide, yet rescued 72 Tutsi children after a massacre.23 George Ruggiu also incited genocide, but rescued Tutsis, as (purportedly) did many of the perpetrators interviewed for this study (almost all Rwandan perpetrators interviewed claimed to have also rescued people).24 Such behavior is possible because individuals who are known personally may be excluded from the ranks of the dehumanized Other. To prepare for actual killing, many perpetrators are socialized through propaganda or training. Consider, for example, the following case study: Perpetrator R22, Hutu Male, Mbazi Sector, “Ignace” 25 The perpetrator, whom we will call “Ignace,” was a 34-year-­old father of four at the time of the genocide. He was not politically active before the genocide – he supported Habyarimana, but, he says, only because it was popular. I never heard about this [Hutu power] before … but since 1990 I started to hear that the Tutsis wanted to come back … I never had a fear of the Tutsis but since 1990 … there were rumors from the government members … they used to say that RPF have returned and are killing people, are hurting people. After that I became afraid … I heard that people were killed in Burundi and that others were dropped in the rivers. They used to say “bad things” about the Tutsi on the radio, for example that the Tutsi were “the enemy,” however I did not believe this. We never expected that something would happen, but there were signs. Sometimes a neighbor would be taken to prison for being a Tutsi, others were killed … I never expected it would actually happen. I got involved in the genocide when the mayor from my district called upon us and told us to prevent thieves from escaping. So, they used to call Tutsis thieves. After, in the night, a soldier came and asked us “what are you doing here?” Then we said “We are just watching these refugees. The mayor told us to watch over them.” Then the soldier said: “are you crazy? You don’t have to watch them – Tutsis are your enemy. You kill them, not just watch them.” This soldier started to teach us how Tutsis were the enemies. Then he said “you are a soldier now. I’m going to kill them. I’m going to shoot them.

Coping with killing   197 Then you will watch to see if anyone escapes or survives. Then you will kill them.” The military used to say “if you don’t kill these guys, their relatives are outside the country and hoping to get in. So, if you don’t kill them they will kill you.” The mayor also began to incite violence saying, “these guys, these Tutsis, in 1959 stole all your cows and ran away. Now is the time for revenge. It’s your time, steal their cows and you make sure that after stealing his cow you kill him … if you want to kill a snake in a tree chop down the tree” [Kinyarwanda proverb]. My first one – I can feel it – but it was like when you see your friends playing football you’re always attracted to the game. […] The better way was to kill them quickly … because we wanted to steal their property … If you saw your friend killing people, you also felt you had to get in front … sort of like a game. Nobody would recognize what happened until after, the consequences. As time went on killing became easier: It was like killing fish. We had no idea about what was happening or what was going to happen.… Of course, I was afraid because if I wouldn’t obey I would have been beaten.… We knew the Tutsis were enemies. Then we were supposed to kill the ones who were inside the country, so that those on the outside would stop. It was confusing. Only authorities knew everything and they lied to us. […] My father told me I was doing something wrong; even my wife got mad at me. In this excerpt, we see several features of participation in genocide – ideological justifications for killing, propaganda on the expected treatment of the victim group, escalating commitments, and incremental habituation of the perpetrators. We also see the use of violence specialists (the army in this case) and legitimate authorities (the mayor) to overcome resistance to killing. A state of confusion is evident, caused by the uneasy coexistence between old norms and new; the norms against killing persevere in the private sphere, as the perpetrator’s father disapproves of his participation in killing. We have a similar example from another respondent who recalls: “My father used to advise us [his children] if one of us was to try to kill someone or being involved in someone being killed – you will never be my child.”26 Another perpetrator reports that “the Interahamwe were sent to sensitize us” and, eventually, “people asked me why I had stopped killing.”27 Yet for other perpetrators the public norm of involvement in killing overrode moral conflicts and countervailing information coming from the private sphere: In the killings, we were supported by the government … that is why people did not think many things. We did not think about the victims … I did not believe I was doing something wrong. I thought that what I was doing was good and that it was something that would make me famous … I used to drink while at the roadblocks.28

198   Coping with killing Another perpetrator recalls “before the RPF invaded, our neighbors attacked Tutsis and we defended them. But after the RPF invaded, intellectuals told us why we should hate them … everything changed after the plane crash of Habyarimana.”29 Coercive force and persuasion are often used to overcome resistance: Where I was living during the genocide, nobody killed the victims there. All victims fled, they went to the church. After hiding in the church nobody attacked them […]. After arriving in the church the people who were living in the neighboring sectors came with clubs and machetes. They found us there in the area. The victims who were in the church, they had the policeman who were in charge of their security. After arriving at the church those people who were from the neighboring area they just came to tell us to help them, because they wanted us to participate and support them to fight against the victims who were hiding themselves in the church.30 Other perpetrators note “where ordinary people were not killing the military was there,” and “the neighboring district came to attack the Tutsi in our area and we tried to defend those Tutsi … but later the intellectuals in our area influenced us so that we could kill the Tutsi.”31 Perpetration becomes a tautological act of self-­definition. The victims are killed because the state says they are bad, and because the victims have been killed they must have been bad. The abstractions applied to the group are also applied, by implication, to every individual victim. Moreover, where perpetrators are reluctant, more motivated perpetrators may be brought in. As resistance to killing is overcome, even reluctant perpetrators come to believe that killing is conventional. The end of genocide If genocide is a process unfolding progressively over time, then we might ask how does genocide end? There are two central possibilities: “genocide interrupted” and “genocide completed.” Genocide is often interrupted through external intervention. In contrast, in cases of completed genocides, genocide is integrated into memory as a passing phase, a miscalculation, or something that simply never occurred. As regimes move towards the completion of genocide, a certain momentum may build. The totality of genocidal ideology and policy certainly contains an element of momentum, as the destruction of the group is not accomplished until the group is completely destroyed. This act of destruction is also an act of creation; mass death acts as a sort of tabula rasa for the imposition of a new vision. This tabula rasa element of genocide was central to Nazi policy, in which Lebensraum meant the removal or killing of tens of millions of “Slavs” and Jews, to resettle ethnic Germans. Moreover, in colonial genocides the land was seen as “Terra Nullius” (empty land) to be settled and civilized. The indigenous

Coping with killing   199 inhabitants were landscape – a hostile element to be tamed. Similarly, the Islamic State, through its declaration of a caliphate, is seeking to create a land belonging only to their iteration of Islam; an act of creation that necessitates the destruction of all heretical elements such as the Êzidî, Shi’a, and Christians.32 The momentum of genocide is contained within this vision of creation through destruction. Genocide is presented not only as a necessary evil, but also as a divine burden. Thus, there is an impetus to complete this task, as there may be for any task already undertaken. The “goal gradient” theory developed by psychologist Kurt Lewin contends that the closer one is to a goal, the greater the incentive for continuation, and the greater the disincentive to abandon the goal.33 The interruption of goal-­directed behavior is a source of tension, and the closer an individual is to attaining their goal, the greater the tension.34 The goal gradient is reinforced by bureaucratic goal displacement: the tendency of bureaucracies to undertake ever-­expanding projects to justify their existence.35 In other words, genocide may be perpetuated by the desire of perpetrators to complete the mission they have embarked on. The interruption of the mission may also call into question the necessity of the act in the first place – a challenge to revised perpetrator self-­concepts. The interruption of genocide also presents the possibility for revenge, particularly in situations of seemingly retaliatory and cyclical genocide (such as in the Former Yugoslavia, Rwanda or Burundi): in one perpetrator’s words, “I thought that if the Tutsi would not be killed, the Tutsis would kill us the next time.”36 When comparing contemporary genocides, it seems that once genocide is embarked upon it generally does not stop of its own accord. The Rwandan genocide, Bangladesh genocide (1971), Holocaust, and Cambodian genocide, were all terminated by foreign military intervention. Moreover, pre-­genocidal pogroms, such as in Rwanda from 1990–1994, are usually only contained through government intervention. It is difficult for victim groups to end genocide, because they normally confront a tremendous power differential.

Individual recidivism Effects of killing on perpetrators The perpetration of genocide on the individual level is also an evolving process. Of course, this recidivism, like the initial act of perpetration, is situated in structures of power (or “harm-­producing relations of power”). Yet we must go deeper to understand the specific mechanisms that impact the recidivism of individual perpetrators, beyond moral neutralization. Any such discussion must begin with an acknowledgement that many perpetrators find their role difficult.37 There is a fundamental difference between abstract and actual killing. Genocide is a deeply disturbing act for some perpetrators, even though perpetration itself may aim to maintain a coherent worldview. One Rwandan perpetrator recounts: “At that time I would sit at night and

200   Coping with killing think: my God, I killed those people.… What am I doing?”38 Perpetrator trauma and memory will be analyzed in detail in Chapter 10. The psychological stresses involved in killing are reduced through the diffusion of responsibility within the group (both the immediate group and the criminogenic state more broadly). The act of killing is also situated in a political context that externalizes and neutralizes individual culpability and shame. For example, in the Nazi publication Tätigkeits-und Lagebericht, one finds the following account of an Einsatzgruppe killing squad: All the men coped with the tough physical stress well. No less considerable were the extreme psychological demands made on them by the large number of liquidations. The morale and self-­possession of the men was kept up by personally reminding them constantly of the political necessity [of what they were doing].39 Even with such ideological indoctrination, perpetration is challenging for many; SS Obersturmbannführer Ernst Ehlers remembers “I was frantic with worry as to how I could be released from having to take part in this Einsatz.”40 Similarly, Albert Hartl notes that some SS men assigned to firing squads “were overcome by uncontrollable fits of crying and suffered health breakdowns.”41 One would expect that perpetrator trauma would vary depending on perpetrators’ level of moral disengagement, as well as their habituation to killing. Those less morally disengaged, such as conformist perpetrators, might feel greater stress in the act of killing. However, these perpetrators, as well as compromised perpetrators, may also avoid stress through recontexualizing their killing as involuntary. We can hypothesize that the more highly motivated (morally disengaged) perpetrators are, the less dissonance they experience. But less motivated perpetrators may more easily externalize responsibility. The features of genocidal states, such as the centralization of authority, the ideological climate of necessity, and mass participation all have the effect of creating institutional contexts that minimize individual culpability and guilt. Moreover, recidivism is also facilitated by institutionalized perpetration systems, such as concentration camps, that are purpose-­built for perpetration, and in which perpetrators are placed close to potential victims. Coping mechanisms Rationalization and ideological self-­justification The neutralization of the killing inhibition is not a complete neutralization. Most perpetrators still experience psychological consequences in the short term and the long term. The prohibition/inhibition on killing is not completely removed. It seems logical to posit that the psychological consequences for direct perpetrators are greater than for indirect perpetrators (even though indirect perpetrators, such as leaders, may be responsible for many more deaths than direct perpetrators).

Coping with killing   201 There are various ways that the psychological and moral impact of killing may be reduced, including through rationalization, psychological defense mechanisms, and substance abuse. Post-­facto justifications for killing or recidivism are also a means to maintain coherence. In a sense, the techniques of neutralization may be especially useful as explanations of persistence or desistance in crime; the perpetrators commit their acts in a setting of intense social pressures and then, with more time for reflection, must rationalize their acts in order to re-­perpetrate.42 Agnew concludes that “longitudinal data suggests that neutralization may be a relatively important cause of subsequent violence.”43 Moreover, Hirschi concludes that neutralization may be part of the “hardening process,” i.e., the conventionalization of killing.44 Yet, as argued previously, moral neutralization underscore the definition of the victims and the situation, and this facilitates violence in the killing moment. We have already extensively discussed perpetrator rationalizations in Chapter 8, but we can add two techniques of neutralization here that are especially important post-­facto: the ledger metaphor, and the denial of genocide.45 In the ledger metaphor, the perpetrator reduces the magnitude of their transgression by arguing that they have done more good than harm in their life. For example, one perpetrator in Rwanda said of his (Tutsi) wife: “She knows that I am a good man.” Genocide denial may also be a form of neutralization, as perpetrators reduce the magnitude of the violence to reduce their own perceived moral responsibility. Violence is moved from a genocide frame (characterized by the massive and asymmetrical application of force against vulnerable victims) to a conflict frame (characterized by equivalency between two belligerent parties). This shift is often eased by the use of combat-­related terms. For example, in Rwanda genocidal violence was regularly described as igitero, derived from the infinitive gutera (“to wage a war”). Thus, violence committed by the igitero was recast as combat rather than genocide.46 Research shows that viewing aggressive acts will decrease aggression if “the aggression seems excessive or gratuitous,” but viewing violence perceived as justified will actually increase aggression.47 In general, neutralization and ideological justifications serve to restructure morality to render past transgressions justifiable. The letters home of SS Obersturmführer Karl Ketsch illustrate the role of ideological justifications in easing the killing inhibition (and helping perpetrators cope with killing): The sight of the dead (including women and children) is not very cheering. But we are fighting this war for the survival or non-­survival of our people.… Here in Russia, wherever the German soldier is, no Jew remains. You can imagine that at first I needed some time to get to grips with this.48 In this excerpt, we see two key components of adjustment to perpetration: the survival discourse as an ideological justification and the incremental habituation to violence – “I needed some time.” Goebbels himself wrote in his diary (on March 27 1942):

202   Coping with killing The procedure is a pretty barbaric one and not to be described here more definitely. Not much will remain of the Jews.… The prophecy which the Führer made about them for having brought on a new world war is beginning to come true in a most terrible way. One must not be sentimental in these matters. If we did not fight the Jews, they would destroy us. It’s a life-­ and-death struggle between the Aryan race and the Jewish bacillus.49 Here the existential threat of the Jews is made manifest while also blaming the victims for their role in bringing violence on themselves. Killing is also presented as a necessary evil, rather than something to be spoken of openly. The self-­delusions of perpetrators are more than just duplicitous – they are a means of self-­preservation. Psychological dissociation and substance abuse In addition to justifications, perpetrators may cope with the impact of killing through other mechanisms. Psychological defense mechanisms (such as doubling and denial) may insulate perpetrators from the consequences of their acts. This ability of some people to psychologically compartmentalize is well illustrated by the diary of Dr. Johann Paul Kremer, a doctor in Auschwitz: September 4, 1942 … present at special action [selection] in the women’s camp … the most horrible of horrors … September 6 … today, Sunday, an excellent dinner: tomato soup, half a chicken with potatoes and red cabbage […] October 10 … I took and preserved … material from quite fresh corpses, namely the liver, spleen, and pancreas … October 11 … today, Sunday, we got for dinner quite a big piece of roast hare with dumplings and red cabbage.50 However, the formation of multiple self-­identities is also indicative of mental disorder.51 It seems that intoxication may play a role in reducing inhibitions and facilitating violence.52 Substance abuse may also enable perpetrators to alleviate their difficult emotional state. In a sense, alcohol and drugs produce a state of derealization, in which violence seems less consequential. This intoxication may take place before the crime is committed or afterwards, as a coping mechanism. For example, one participant in a Nazi massacre of Jews in Poland says that after they returned to their barracks, “generous quantities of alcohol were provided, and many of the policemen got quite drunk.” Another policeman says, “most of the other comrades drank so much solely because of the many shootings of Jews, for such a life was quite intolerable sober.”53 Moreover, 88.2 percent of Rwandan perpetrators interviewed for this study reported that they, or other perpetrators they witnessed, were drinking or using drugs.

Coping with killing   203 The decreasing cost of perpetration For many individuals, the psychological cost of participation in killing seems to decrease over time. This is because killing generally becomes easier as perpetrators become habituated to it. Although the factors causing individual perpetration are generally still present, the sense of risk diminishes. Moreover, even killing that is initially motivated by obedience eases subsequent voluntary killing.54 It is increasingly framed as consonant with conventional moral values; moreover, perpetrators likely become more efficient as they incorporate best practices into the killing process. The marginal cost of killing may also decrease: that is, the additional value (moral) costs may decline per person killed. Additionally, inefficiency costs (decreased productivity in killing due to a lack of technical proficiency) are reduced as perpetrators practice killing. Legally speaking, perpetrators do, of course, often receive longer sentences based on the number of their victims. Yet even this marginal cost decreases after a point. The process can be represented graphically – see Figure 9.1. The decreasing cost margin of perpetration results from several processes, including: (1) killers’ euphoria, (2) learning by doing, (3) positive feedback, and (4) secondary transgression. However, the decreasing cost margin may not endure in the long term.55 Some individuals may suffer exhaustion or other forms of psychological breakdown. A former Sans Échec paramilitary from Burundi explains: “I saw the horrors and it changed my mind.”56 Later in the interview, he noted: “after the war some Sans Échec went to prison, [while] others went mad because of what they had done.” Perhaps the cost margin of perpetration may also be dependent on initial levels of propensity and moral disengagement. (i)  Learning by doing/habituation It is a weakness not to be able to stand the sight of dead people; the best way of overcoming it is to do it more often. Then it becomes a habit.…57 (SS Obersturmfuherer Karl Kretschmer of Sonderkommando 4a – in a letter home)

Cost of perpetration

Incidence of perpetration

Figure 9.1  The decreasing cost margin of perpetration.

204   Coping with killing For many perpetrators, killing becomes easier as they grow accustomed to it – it becomes a moral habit. One LRA perpetrator in Uganda recounts: You are taken as a new recruit. To make you be confident enough, you are made to kill your brother … you are brainwashed, you know now that even when I kill there is nothing wrong with it. So, then killings become part of your life.58 A Rwandan perpetrator recounts that killing “was just like a game, for me. It was something which was done each and every day. It was simple … you had no fear that you would be arrested.”59 Even so, perpetrators feel the need to justify their actions and these justifications will lead to an increasing radicalization and isolation from regular norms. A perpetrator on the Eastern Front writes: It’s nice that Herr Kern is going to France. I think he would have been too weak for the East, though people do change here. People soon get used to the sight of blood, but blutwurst [blood sausage] is not very popular around here.60 Pressure exists within the group to demonstrate “toughness” as a form of loyalty, but also evident, at times, is an understanding of the demands of being involved in killing. This may provide space for perpetrators to escape, posture, or otherwise avoid direct involvement in killing. Another perpetrator from the Holocaust, Walter Mattner, a Viennese police clerk in the Einsatzgruppen, wrote to his wife: When the first truckload arrived, my hand was slightly trembling when shooting, but one gets used to this. When the tenth load arrived, I was already aiming more calmly and shot securely at the many women, children, and infants. Considering that I too have two infants at home, with whom these hordes would do the same, if not ten times worse, the death we gave to them was a nice, quick death compared to the hellish torture of thousands upon thousands in the dungeons of the GPU [the Soviet secret police].61 In this document, we see both the claim of higher purpose and the claim of relative acceptability. Thus, the very rationalizations that justify perpetration will also lower the costs of recidivism. Prior experience with killing, even of animals, may also make killing easier for some perpetrators. Rwandans were already accustomed to using machetes on the farm to slaughter animals, and according to one perpetrator: “in the first few days someone who had already slaughtered chickens – and especially goats – had an advantage, understandably.”62 Humans have a tremendous capacity to adapt to their circumstances. At its best, this adaptation is manifested in the will of concentration camp inmates to struggle for survival in the face of horrific circumstances. At its worst, the

Coping with killing   205 initially reticent concentration camp guard becomes numb to killing. This habituation evolves over time. For instance, in order to acclimatize men to killing, the Einsatzgruppen massacres in Eastern Europe normally commenced with the killing of men, before progressing to women and children. The killing of men may be easier than the killing of women and children, as men are traditionally considered to be combatants or potential combatants, and thus may be seen to pose a more manifest threat. The opponent process theory of Richard Solomon helps to explain habituation. Solomon argues that psychological processes are homeostatic: the initial response to killing (the alpha process, visceral aversion) is followed by a beta process (a restorative process including feelings such as euphoria or pleasure). Initially the beta process is weak, but over time the process is reversed – aversion weakens and pleasure strengthens.63 There is a dulling of empathy. Guilt acts as a mechanism of social control by moderating our actions; yet, in cases of genocide this sense of external guilt is removed. The desensitization of the perpetrator runs parallel to the dehumanization of the victim. Moreover, through positive reinforcement, perpetrators may develop a conditioned response, such that perpetration occurs as a habitual reaction. Some perpetrators may also become numb to the effects of killing, going through the motions: “I was blank during the genocide,” as one Rwandan perpetrator reports.64 Habituation, however, does not mean perpetrators become completely comfortable with perpetration. A perpetrator in Burundi notes: At first, I was very uncomfortable to be in the armed group, but step by step I got used to it. But I never got used to killing people. I never felt at ease to having to commit such barbaric acts.65 (ii)  Killer’s euphoria and empowerment Some perpetrators achieve a neurophysiological “high” from killing.66 This is like the euphoria felt by some soldiers in war, and some hunters. It is produced by an increase in levels of endogenous opiates.67 Such a high “rewards” and positively reinforces violent behavior. Killing can also provide perpetrators with a sense of empowerment, as killing is the ultimate expression of the omnipotence of the killer and the powerlessness of the victim. Control by dominants over subordinates may also have a physical pay-­off through endocrinal and metabolic mechanisms.68 Individuals may actually enjoy killing. Some evolutionary psychologists argue that there exists a “pain-­blood-death complex” – animals (including humans) may receive certain positive reinforcements from the actual process of predation/killing. The victims’ physical sequalae – attempts at escape, screams, death throes, blood – may trigger euphoric feelings in some perpetrators.69 As gruesome as this is, it may assist our understanding of cruelty in genocide.

206   Coping with killing (iii)  Positive reinforcement Killer’s euphoria is an internal reward for killing, but rewards often come from external sources (positive reinforcement). This reinforcement may take the form of peer (or authority) approval or rewards. These rewards may be direct, such as the distribution of money, property, or alcohol, or indirect, such as future opportunities for career promotions. For example, one Rwandan perpetrator recounts that “people who were enthusiastic killers were embraced by the authorities,” while another claims: “when you killed someone you were rewarded. If you were a civilian, you would instantly become an army officer with rank the next day.”70 Similarly, a former Tuol Sleng prisoner remembers: “There was a sort of competition [amongst the guards]. If you are absolute with the revolution you’ll earn higher ranks.”71 One commanding officer in Rwanda paid perpetrators 50,000 Rwandan francs for “a job well-­done.”72 Reinforcement has the effect of lessening the impact of the killing act, as well as promoting re-­perpetration. It may also contribute to the development of a conditioned response to killing. When actions are confirmed through positive reinforcement, social conditioning is greatly magnified.73 Moreover, experimental data shows that “when participants perceived their violent behavior to be normative, they experienced less self-­reported distress after killing than when they perceived their actions to be socially invalidated.”74 Finally, the absence of serious consequences for perpetrators may also provide validation for their wrongful behavior. Violence may be inflicted as way of testing limits, determining which acts will be permitted. (iv)  Secondary transgression hypothesis Finally, the marginal cost of perpetration is lowered through what can be called the secondary transgression hypothesis. This posits that once a rule has been transgressed and the perpetrator has crossed the boundary into being a rule violator, there is little incentive to cease violations of the rule and return to “ordinary” values. In other words, with the first act of perpetration (particularly with very serious crimes such as genocidal killing and rape), an individual has been labelled as a “perpetrator” (or worse a “murderer”), and there is no going back. Therefore, the first act of perpetration is the costliest, and frequent perpetration involving many victims transforms perpetration from a personal act to an abstraction. A perpetrator who has already killed ten people has little to lose by killing an eleventh. Many perpetrators do re-­perpetrate. One Cambodian perpetrator interviewed for this study is alleged to have killed as many as 2,000 people by his own hands (executions with a blunt instrument), while there is evidence that Josef Klehr, an SS sergeant, killed 475 people personally.75 Even if there was a moral inhibition against killing present in these perpetrators initially, the performance of repeated acts of perpetration would be impossible without its sublimation or elimination.

Coping with killing   207 Perpetrator identity I admit and recognize my obedience at that time, my victims, my fault, but I fail to recognize the wickedness of the one who raced through the marshes on my legs, carrying my machete. That wickedness seems to belong to another self with a heavy heart.76 (Rwandan perpetrator) Perpetrators may redefine their perceptions to achieve consistency with their actions. This redefinition of identity may occur both in cases of forced transgression of norms and in situations of voluntary participation in violence. The new self may be monstrous by ordinary standards: one German soldier involved in the racial war being waged on the Eastern Front wrote a plaintive letter to his wife, saying “one has to be ruthless and unmerciful. Don’t you have the impression that it’s not me but a different person who is speaking to you?”77 This redefinition of the self often involves sublimation of the individual into the new group self. The loyalty to the group and to comrades generates a sense of duty, and provides an absolution for individual wrongdoing. Moreover, “violent behavior increases group identification” and “group members who perpetrate high levels of violence likely stay longer in their violent group.”78 Individuals’ self-­image is reflected in their engagement in a given activity. If individuals find their actions to be personally enriching, they will restructure their perceptions to interpret the activity as more worthwhile, and the benefits they derive as more deserving. Concomitantly, those harmed by these actions (the victims) will be devalued.79 In other words, individuals participating in genocide will modify their views to conform with their actions, to maintain their psychological integrity. The persistence of the self also overrides new, and contradictory, information. Killing may also gain a certain momentum as individuals who have already killed demand conformity from their peers. Conformity in killing demonstrates social equality. Furthermore, considering that genocide is normative and encouraged by the state, individuals may also try to outdo each other in realizing the collective normative goal of exterminating the enemy. Accomplishing a socially­endorsed task (such as “doing your job”) can bring emotional and material rewards, while the desire to do your job well may become an end in itself. The sequential nature of perpetrator actions also means that the cessation of killing may imply recognition of the wrongful nature of acts committed to that point.80 Thus, in a sense, it is psychologically easier for perpetrators to continue killing and to maintain their positive self-­image (within a redefined moral landscape) than to challenge their beliefs and confront cognitive dissonance. Robert Jay Lifton posits that some perpetrators create a separate perpetrator self, and transfer conscience and volition to that self as a means of maintaining moral identity.81 If this is true, one could argue that moral neutralization contributes to the distancing of the perpetrator self from the normal self. The perpetrator self grows to be invested in the revised moral context,

208   Coping with killing even while the normal self remains connected to conventional morality. Over time, the perpetrator self may become the ordinary self, particularly for those (pathological) individuals who are already disposed towards violence. The reestablishment of the self, after the moral restructuring that enables acts of atrocity, can be traumatic. Perpetrator selves have eclipsed normal selves, producing a lesser conflict of values and a reduced need for self-­justification. Although the concept of the formation of separate selves is well-­established in philosophy (in the work of Daniel Goleman and Richard Lazarus, for example), and in psychology in terms of dissociative disorders, clinical psychologists such as David Spiegel note that “switching behavior from one alternate identity to another is relatively uncommon.”82 The sadistic shift Violence shifts moral values. Some perpetrators, even those with low propensity, may enjoy the freedom from social constraints granted in genocide: “If you have shown your interest, that you are willing. You won, you can train, you become their soldier. You have freedom. You cannot be suppressed all the time.”83 Similarly, a perpetrator in Rwanda said of the Hutu power ideology: I supported it and liked it … I liked it because it gave us freedom.… Freedom in the sense of walking freely, nobody could ask you to show him your ID, you could walk free, you could speak free.84 Killing becomes an exercise of freedom – both a response to freedom from conventions and a way to free oneself from convention. Emile Durkheim argues that external social control is necessary because we lack internal controls; thus, the breakdown of external controls leads to insatiable desires. Several perpetrators in Rwanda spoke of this state of moral freedom (and seeming normlessness) as one in which they had an “animal heart”: “I still had that animal heart” and “The perpetrators believed God did not exist.… At that time people had ‘animal hearts’.”85 This wild state of being was frightening for some perpetrators: “Because of what’s been done, these guys were looking like they were crazy. They looked as though they could kill each other.”86 Another remembers: I thought that it was the end of the earth; it was the Day of Judgment. Because the people killed others without pity, they just killed them because of who they were. For me it was like the end of the earth.87 This setting aside of norms may, however, be exciting for some perpetrators. Killing may become an addiction. While participating in extreme violence, some perpetrators undergo a sadistic shift, in which they start to exhibit sadistic behavior even if they cannot be psychologically diagnosed as sadists. It is difficult to diagnose individuals as

Coping with killing   209 sadists (exhibiting Sadistic Personality Disorder), where their sadistic acts only occur within a unique institutional context.88 However, many perpetrators go beyond a mere duty to kill, and begin to murder with cruel inventiveness. A perpetrator in Burundi, who himself tortured prisoners, argues “these were men who had lost all emotions, who had lost all humanity.”89 In the women’s concentration camp at Ravensbrück, one survivor recounted how inmates would make a game of guessing how long it would be until “humane” guards would begin to act sadistically.90 It seems that some people behave in outrageous ways when given the opportunity to do so. One such opportunity is presented by perpetrator omnipotence in the context of genocidal killing. Mass participation in violence also reconfigures moral frameworks – as individuals adjust their values to account for the conventionalization of violence.

Systems of denial Many of the structures used in the perpetration of genocide also function as systems of denial. These exist both within the state and within the broader cultural framework. Genocidal systems of denial may include elements such as euphemistic language, coded commands, the destruction of documents, the removal of physical structures, and the concealment of human remains. For example, 80.6 percent of the Rwandan perpetrators interviewed for this study made some effort to conceal the bodies of those they had killed. Ideology may also play an important role in denial by privileging abstract ideological concepts over actual human suffering. These systems of denial operate through shared cultural understandings, and many endure long after genocide has occurred. The denial of the crimes against the victims is also a denial of the valuation of the victims. As objects, they are to be acted upon – not considered human victims of crimes. In Nazi Germany, these systems of denial were vigorously enforced by the state. For example, a statement by the Commander in Chief of the German 6th Army Generalfeldmarschall von Reichenau admonishes: The conclusion of the report in question contains the following sentence, “In the case in question, measures against women and children were undertaken which in no way differ from atrocities carried out by the enemy about which the troops were continuously being informed.” I have to describe this assessment as incorrect, inappropriate, and impertinent in the extreme. Moreover, this comment was written in open communication which passes through many hands. It would have been far better if the report had not been written at all.91 Similarly, an SS Untersturmführer, Max Taubner, was convicted on May 24 1943 for taking photographs of the murder of Jews: The accused should not be punished because of the actions against the Jews as such. The Jews have to be exterminated and none of the Jews that were killed is any great loss.… Real hatred of the Jews was the driving motivation

210   Coping with killing for the accused. In the process, he let himself be drawn into committing cruel actions in Alexandriya which are unworthy of a German man and an SS officer.… By taking photographs of the incidents or having photographs taken, by having these developed in photographic shops and showing them to his wife and friends, the accused is guilty of disobedience. Such pictures could pose the gravest threat to the security of the Reich if they fell into the wrong hands.92 In this excerpt, we see several interesting aspects of the genocidal process: the concealment of crimes, but also the normative limits that continue to exist during genocide. A tension is often evident in genocide, and in repressive states more broadly, between the strategic revealing and concealment of acts of violence. As already argued, violence in the public sphere can send an important normative message about the status of victims. However, genocidal regimes may also have to conceal certain forms of violence to minimize the risks that the exposure of such crimes might pose, either internationally or by opposition groups domestically. In some cases, the concealment of crimes may require the commission of further crimes: a perpetrator in Burundi remembers “if I saw somebody who knew me when I was doing an attack, they had to be killed.”93 Concealment is much easier in situations in which the group is isolated (i.e., without external observers), or where confusion predominates, such is in war, where it can be difficult to distinguish atrocities from legal and “legitimate” forms of killing such as targeting enemy combatants. The following order issued to a reserve police unit participating in genocidal killing in Eastern Europe during the Holocaust, is illustrative: Confidential! (1) By order of the higher SS and Police Leader … all male Jews between the ages of 17 and 45 convicted as plunderers are to be shot according to martial law. The shootings are to take place away from cities, villages, and thoroughfares. The graves are to be levelled in such a way that no pilgrimage site can arise. I forbid photographing and the permitting of spectators at the executions. Executions and grave sites are not to be made known. (2) The battalion and company commanders are especially to provide for the spiritual care of the men who participate in this action. The impressions of the day are to be blotted out through the holding of social events in the evenings. Furthermore, the men are to be instructed continuously about the political necessity of the measures.94 In this excerpt, we see the importance of deniability, but also the provision of an ideological rationalization for killing, the use of coping mechanisms, and positive reinforcement for the perpetrators. The approach of the Ottoman Turkish government to the genocide of the Armenians is instructive, in that concealment was embedded in the command structure. The “dual-­track” approach involved, on the one hand, deportation orders sent through official channels from the Interior Ministry to provincial

Coping with killing   211 governors, and, on the other hand, coded commands for killing issued orally by the party’s responsible secretaries (who travelled to the provinces to convey the regime’s demands).95 Those who refused to participate were often dismissed.96 Beyond this structure of concealment, the bodies of the victims were often disposed of. A coded telegram sent on July 21 1915 to several provincial and district governors instructed: “the dead who remain on the roads are to be removed and their corpses are to be thrown into the valleys, lakes, or rivers.”97 The treatment of bodies may also signal the denial of social honor. In Rwanda, victims were sometimes buried, but at other times were left in the open. One political leader convicted of genocide in Rwanda reports: “People who tried to bury bodies would be killed. Their intention was to show they [the victims] weren’t like human beings.”98 Another interviewee recalls that bodies were thrown in holes that were normally used for the corpses of dead animals.99 Similarly, in the Lord’s Resistance Army, the corpses of enemies were left to rot while the corpses of LRA members were treated with respect.100 Bodies may also be intentionally degraded: one survivor of the massacre at Karama Church in Rwanda recalls seeing male and female Tutsi bodies stripped and symbolically positioned as though they were having sex: according to the perpetrators, “these Tutsis need to have sex one last time.”101 Systems of denial (at the time of perpetration) are often stronger in specialized genocides, as the regime seeks to limit awareness of killing to the relevant organizations.

Conclusion: genocidal momentum This chapter has demonstrated that the process of violence evolves over time, on both individual and collective levels. Genocide itself is a constitutive process – as violence unfolds, it also creates meaning. The identities of perpetrators adjust to internalize or contextualize violence. Violence is a process of habituation and routinization, but also of logistical problem-­solving. Perpetrators seek to enact violence more effectively and to overcome the resistance of individuals opposed to it. The existence of opposition is itself challenging to the self-­concept of perpetrators, who need to believe that they are engaged in a productive and morally valid act. Therefore, the moral and material context of genocide is partly created through individual participation in violence. Propensity also likely increases via participation in violence, as individuals rationalize their roles, become habituated to them, and function as de facto state agents. Individual acts of murder create the momentum of genocide but the pressure on individuals to perpetrate also increases through the collective actualization of genocidal killing.

Notes    1    2    3    4

Goldhagen, Hitler’s Willing Executioners, 167–168. Gross, Neighbors: The Destruction of the Jewish Community, 81. Quoted in Bauman, Modernity and the Holocaust, 90. Straus, The Order of Genocide, 90–91.

212   Coping with killing    5    6    7    8    9   10   11   12   13   14   15   16   17   18   19   20   21   22   23   24   25   26   27   28   29   30   31   32   33   34   35

  36   37   38   39   40   41   42   43   44   45   46   47

Guichaoua, From War to Genocide, 167. Straus, 242. Interview R1 (Rwanda), Kigali Central Prison, July 2009. Alvarez, Genocidal Crimes, 90. Interview R75 (Rwanda), sector counsellor, Kigali Central Prison, October 27 2009. Interview R74 (Rwanda), sector counsellor, Kigali Central Prison, October 27 2009. Guichaoua, 235. 87.6 percent of Rwandan perpetrators interviewed for this study participated in Umuganda prior to the genocide. Mironko, “Igitero: means and motive in the Rwandan genocide,” 53. Goldhagen, Worse than War, 127. Interview B09 (Bosnia), Višegrad, April 2009. Goldhagen, Hitler’s Willing Executioners, 122. It is also interesting to remember that there was opposition in Germany to the murder of some Jews (German Jews and war heroes for instance) because they were “good Jews.” Bauman, 187. Des Forges, Leave None to Tell the Story, 262. Interview R67 (Rwanda), Hindiro TIG Camp, October 2009. Bauman, 189. Quoted in Bauman, 190. Prosecutor v. Goran Jelisic, para. 102. See also: Behrens, “A moment of kindness?,” 128. Prosecutor v. Clement Kayishema, and Obed Ruzindana. Prosecutor v. Georges Ruggiu. This conversation has been slightly restructured, for the sake of clarity. Interview R22 (Rwanda), Tumba TIG Camp, August 2009. Interview R42 (Rwanda), Nyarusenge TIG Camp, August 2009. Interview R57 (Rwanda), Rugerero TIG Camp, September 2009. Interview R55 (Rwanda), Rugerero TIG Camp, September 2009. Interview R60 (Rwanda), Mont Kigali TIG Camp, October 2009. Interview R72 (Rwanda), Hindiro TIG Camp, October 2009. Interview R25 (Rwanda), Tumba TIG Camp, August 2009; and Interview R60 (Rwanda), Mont Kigali TIG Camp, October 2009. Moradi and Anderson, “The Islamic State’s Êzîdî genocide in Iraq.” Staub, The Psychology of Good and Evil, 330. Staub, 330. Bauman, 17. An example of this phenomenon may be the expanding vision of the Nazi regime in its Jewish policy – moving from the creation of Jewish protectorate in Poland, to a protectorate in Madagascar, then a protectorate in the USSR (beyond the Archangel-­Astrakhan line), and finally extermination. Interview R26 (Rwanda), Gisenyi Prison, August 2009. MacNair, Perpetrator-­Induced Traumatic Stress; and Nordstrom, “Deadly myths of aggression,” 151; see also: Grossman, On Killing. Interview R02 (Rwanda), Kigali Central Prison, July 2009. Klee, Dressen, and Riess, The Good Old Days, 61. Klee, Dressen, and Riess, 82. Klee, Dressen, and Riess, 84. Maruna, and Copes, “What Have We Learned,” 71. Agnew, “The Techniques of Neutralisation and Violence,” 572. Hirschi, Causes of Delinquency, 208. For more information on the ledger metaphor see: Klockars, The Professional Fence. Mironko, 51. Kosloff, Greenberg, and Solomon, “Considering the roles of affect,” 232.

Coping with killing   213   48   49   50   51   52   53   54

Klee, Dressen, and Riess, 163. Semelin, Purify and Destroy, 235. Lifton, The Nazi Doctors, 292. Presser and Sandberg, “Introduction: What Is the Story?,” 8. See: Bushman and Cooper, “Effects of Alcohol on Human Aggression.” Browning, Ordinary Men, 69 and 82. Martens, Kosloff, and Eckstein Jackson, “Evidence That Initial Obedient Killing Fuels.”   55 Differences in susceptibility to conditions such as Post-­Traumatic Stress Disorder (PTSD) also appear to be more of a product of individual differences in adaptability as opposed to the magnitude of the act of perpetration.   56 Interview Bu10 (Burundi), Bujumbura, September 2009.   57 Rhodes, Masters of Death, 220.   58 Interview U04 (Uganda), Onyama Returnee Camp, July 2009.   59 Interview R02 (Rwanda), Kigali Central Prison, July 2009.   60 Klee, Dressen, and Riess, 167.   61 Sémelin, 250.   62 Hatzfeld, Machete Season, 37.   63 Dutton, The Psychology of Genocide, 116.   64 Interview R73 (Rwanda), Kigali Central Prison, October 2009.   65 Interview Bu09 (Burundi), Bujumbura, September 2009.   66 See, for example, Nell, “Cruelty’s Rewards.”   67 Gove and Wilmoth, “The Neurophysiologic High,” 238.   68 Herbert, “The Physiology of Aggression,” 63.   69 Dutton, 143.   70 Interview R39 (Rwanda), Nyarusenge TIG Camp, August 2009; and Interview R75 (Rwanda), Kigali Central Prison, October 2009.   71 Interview C06 (Cambodia), Phnom Penh, November 2008.   72 Omaar, “The Leadership of Rwandan Armed Groups,” 234.   73 Sternberg and Sternberg, The Nature of Hate, 38.   74 Webber, et al., “Using a Bug-­Killing Paradigm.”   75 Mann, The Dark Side of Democracy, 253.   76 Hatzfeld, 48.   77 Heer and Nauman, War of Extermination, 332.   78 Littman and Levy Paluck, “The Cycle of Violence,” 95.   79 Staub, 326.   80 Milgram, Obedience to Authority, 149.   81 Lifton, 30.   82 See Spiegel and Cardena, “Disintegrated experience,” 832. See also Goleman, Vital Lies, Simple Truths; and Lazarus, Fifty Years of the Research and Theory.   83 Interview U01 (Uganda), Gulu, July 2009.   84 Interview R62 (Rwanda), Mont Kigali TIG Camp, October 2009.   85 Interview R02 (Rwanda), Kigali Central Prison, July 2009; and R29 (Rwanda), Rugerero TIG Camp, August 2009.   86 Interview R30 (Rwanda), Butare, August 2009.   87 Interview R60 (Rwanda), Mont Kigali TIG Camp, October 2009.   88 Kaminer and Stein, “Sadistic personality disorder,” 482.   89 Interview Bu9 (Burundi), Bujumbura, September 2009.   90 Alvarez, 107.   91 Klee, Dressen, and Riess, 153.   92 Klee, Dressen, and Reiss, 202.   93 Interview Bu9 (Burundi), Bujumbura, September 2009.   94 Browning, 14.   95 Akcam, The Young Turk’s Crimes against Humanity, 194.

214   Coping with killing   96   97   98   99 100 101

Akcam, 195. Akcam, 200. Interview R17 (Rwanda), Gitarama Prison, July 2009. Interview R30 (Rwanda), Butare, August 2009. Interview U01 (Uganda), Gulu, July 2009. Interview R48 (Rwanda), Nyamata, September 2009.

Works cited Agnew, Robert. “The Techniques of Neutralisation and Violence.” Criminology, 32, no. 4 1994. Akcam, Taner. The Young Turk’s Crimes against Humanity. Princeton: Princeton University Press, 2013. Alvarez, Alex Genocidal Crimes. New York: Routledge, 2010. Bauman, Zygmunt. Modernity and the Holocaust. Ithaca, New York: Cornell University Press, 2000. Behrens, Paul. “A moment of kindness? Consistency and genocidal intent.” In The Criminal Law of Genocide: International, Comparative and Contextual Aspects, edited by Ralph J. Henham and Paul Behrens. Aldershot, UK: Ashgate, 2007. Browning, Christopher. Ordinary Men: Reserve Battalion 101 and the Final Solution. New York: Harper, 1998. Bushman, Brad J. and Harris M. Cooper. “Effects of Alcohol on Human Aggression: An Integrative Research Review.” Psychological Bulletin 107, no. 3 (1990). Des Forges, Allison. Leave None to Tell the Story: Genocide in Rwanda. New York: Human Rights Watch, 1999. Dutton, Donald G. The Psychology of Genocide, Massacres, and Extreme Violence: Why Normal People Come to Commit Atrocities. Westport: Praeger Security International, 2007. Goldhagen, Daniel. Hitler’s Willing Executioners. New York: First Vintage Books, 2001. Goldhagen, Daniel. Worse than War: Genocide, Eliminationism and the Ongoing Assault on Humanity. New York: PublicAffairs, 2009. Goleman, Daniel. Vital Lies, Simple Truths: The psychology of self-­deception. London: Simon and Schuster, 1996. Gove, Walter R. and Charles Wilmoth. “The Neurophysiologic High That Would Appear to Positively Reinforce Criminal Behavior.” In Encyclopedia of Criminology and Deviant Behavior, edited by Clifton D. Bryant. Philadelphia: Brunner-­Routledge, 2001. Gross, Jan T. Neighbors: The Destruction of the Jewish Community in Jedwabne, Poland. London: Penguin Books, 2002. Grossman, Dave. On Killing. New York: Back Bay Books, 1996. Guichaoua, Andre. From War to Genocide: Criminal Politics in Rwanda, 1990–1994. London: University of Wisconsin Press, 2015. Hatzfeld, Jean. Machete Season: The Killers in Rwanda Speak. A Report. New York: Picador Books, 2005. Heer, Hannes and Klaus Nauman. War of Extermination: The German Military in World War II 1941–1944. New York: Berghan Books, 2000. Herbert, J. “The Physiology of Aggression.” In Aggression and War: Their Biological and Social Bases, edited by Jo Groebel and Robert A. Hinde. New York: Cambridge University Press, 1993.

Coping with killing   215 Hirschi, Travis. Causes of Delinquency. Berkeley: University of California, 1969. Klee, Ernst, Willi Dressen, and Volker Riess (eds.). The Good Old Days: The Holocaust as Seen by its Perpetrators and Bystanders. Old Saybrook: Konecky and Konecky, 1991. Kaminer, Debra and Dan J. Stein. “Sadistic personality disorder in perpetrators of human rights abuses: A south African case study.” Journal of Personality Disorders 15, no. 6 (2001). Klockars, Carl B. The Professional Fence. New York: Free Press, 1974. Kosloff, Spee, Jeff Greenberg, and Sheldon Solomon. “Considering the roles of affect and culture in the enactment and enjoyment of cruelty.” Behavioral and Brain Sciences 29, no. 3 (2006). Lazarus, Richard. Fifty Years of the Research and Theory of RS Lazarus: An Analysis of Historical and Perennial Issues. New York: Psychology Press, 2013. Lifton, Robert Jay. The Nazi Doctors: Medical Killing and the Psychology of Genocide. New York: Basic Books, 2000. Littman, Rebecca and Elizabeth Levy Paluck. “The Cycle of Violence: Understanding Individual Participation in Collective Violence.” Advances in Political Psychology 36, suppl. 1 (2015). MacNair, Rachel. Perpetrator-­Induced Traumatic Stress: The Psychological Consequences of Killing. Westport: Praeger, 2002. Mann, Michael. The Dark Side of Democracy: Explaining Ethnic Cleansing. New York: Cambridge Books, 2005. Martens, Andy, Spee Kosloff, and Lydia Eckstein Jackson. “Evidence That Initial Obedient Killing Fuels Subsequent Volitional Killing Beyond Effects of Practice.” Social Psychological and Personality Science 1, no. 3 (2010). Maruna, Shadd and Heith Copes. “What Have We Learned from Five Decades of Neutralization Research?” Crime and Justice 32 (2005). Milgram, Stanley. Obedience to Authority: An Experimental View. New York: Harper Perennial Classics, 2004. Mironko, Charles. “Igitero: means and motive in the Rwandan genocide.” Journal of Genocide Research 6, no. 1 (2004). Moradi, Fazil and Kjell Anderson. “The Islamic State’s Êzîdî Genocide in Iraq: The ‘Sinjâr Operations’.” Genocide Studies International 10, no. 2 (March 2017). Nell, Victor. “Cruelty’s Rewards: The Gratifications of Perpetrators and Spectators.” Behavioral and Brain Sciences 29, no. 3 (2006). Nordstrom, Carolyn. “Deadly myths of aggression.” Aggressive Behavior 24, no.  2 (1998). Omaar, Rakiya. “The Leadership of Rwandan Armed Groups Abroad with a Focus on the FDLR and RUD/URANA.” Kigali, Rwanda: Rwanda Demobilization and Reintegration Commission, 2008. Presser, Lois and Sveinung Sandberg. “Introduction: What Is the Story?” In Narrative Criminology: Understanding Stories of Crime, edited by Lois Presser and Sveinung Sandberg. New York: New York University Press, 2015. Prosecutor v. Clement Kayishema, and Obed Ruzindana. (Case No. ICTR-­95–1-T). Trial Chamber Judgement. May 21 1999. Prosecutor v. Goran Jelisic. (Case No. IT-­95–10). Trial Chamber Judgement. December 14 1999. Prosecutor v. Georges Ruggiu. (Case No. ICTR-­97–31-I). Trial Chamber Judgement. June 1 2000.

216   Coping with killing Rhodes, Richard. Masters of Death: The SS-­Einsatzgruppen and the Invention of The Holocaust. New York: Alfred A. Knopf, 2002. Semelin, Jacques. Purify and Destroy: The Political Uses of Massacre and Genocide. New York: Columbia University Press, 2007. Spiegel, David and Etzel Cardena. “Disintegrated experience: the dissociative disorders Revisited.” Journal of Abnormal Psychology 100, no. 3 (1991). Staub, Ervin. The Psychology of Good and Evil: Why Children, Adults, and Groups Help and Harm Others.. New York: Cambridge University Press, 2002. Sternberg, Robert J. and Karin Sternberg. The Nature of Hate. New York: Cambridge University Press, 2008. Straus, Scott. The Order of Genocide: Race, Power, and War in Rwanda. Ithaca, USA: Cornell University Press, 2006. Webber, David, Jeff Schimel, Andy Martens, Joseph Hayes, and Erik H. Faucher. “Using a Bug-­Killing Paradigm to Understand How Social Validation and Invalidation Affect the Distress of Killing.” Personality and Social Psychology Bulletin 39, no. 4 (2013).

10 After genocide I Memory, trauma, and rehabilitation

Introduction: forgetting genocide Genocide is often seen as an uncharacteristic act, from the perpetrator’s perspective, and as a defining moment, from society’s perspective. Where societies do not address genocide, a social void remains – a moment of mass murder which cannot be easily explained. The quest for coherence contributes to how genocide is viewed after the fact – how it can be excluded from self-­identity, or rationalized in a way that makes it comprehensible and “uncharacteristic.” Killing, in some respects, may not be a very durable norm: once the urgency and survival discourse driving genocide is gone, the norm authorizing killing also disappears (although it may be replaced by norms rationalizing past killings). However, the social ruptures produced by genocide do not end with genocide itself. Rather, they endure in the years and even decades to follow. As noted in Chapter 9, genocide often ends with external intervention. Yet in those cases where genocide is not interrupted, the regime and society must either enforce collective amnesia or rationalize mass killing as a legitimate use of force. Both exigencies are continuations of the genocidal process. The first possibility, amnesia, arises from the concealment and euphemistic framing of crimes during perpetration. The completion of genocide facilitates amnesia, as it allows perpetrators (including the state) to transcend violence and return to conventional values and practices. For example, the genocides against indigenous peoples in the Americas and elsewhere are forgotten or downplayed by the institutional and individual descendants of the perpetrators, and the memory-­holders of violence, the victims, remain marginalized – a position realized through the exercise of violence in the first place. Moreover, the descendants are beneficiaries – wealth has been built on not only the act of perpetration but the act of forgetting. The post-­facto rationalization of mass killing is in some way a continuation of the rationalizations that feature during genocide. The victims are culpable for the treatment meted out to them: they are not victims at all, their destruction was necessary, or it was an excessive act committed by a few misguided individuals unrepresentative of the political community. Such collective amnesia and rationalization is only possible in contexts in which the perpetrator group is victorious – in which they retain control over the instruments of power and have confronted

218   After genocide I no real moment of reckoning. Sadly, history is replete with examples of such scenarios; accountability for genocide may be more the exception than the rule, although there is an increasing recognition that genocide is a deviant act within the international system and that states themselves may be deviant actors through committing genocide.1 The forgetting or rationalization of genocide on an institutional level informs individual responses to genocide. The state continues to define normativity and the rules of identifying with, and belonging to, the community. When perpetration is forgotten, justified, or even glorified, individual perpetrators will remember their crimes differently than when there is regime change and societal transformation after genocide. For example, the memory of the Indonesian mass killings of communists in 1965 remains largely suppressed or celebrated (as depicted brilliantly by Joshua Oppenheimer in his films “The Act of Killing” and “The Look of Silence”). This absence of community shaming means that some perpetrators continue to feel pride and even boast about their acts of killing. This has profound effects on victims, who remain situated within the social context which produced their victimization. Moreover, the perpetrator group may police those who challenge narratives of past violence. This contributes to a profound sense of victim vulnerability. It also leaves victims outside the moral community; thus, the moral ruptures represented by the act of perpetration are sustained. For victims, the undermining of conventional values during genocide may also destroy the thin veneer of faith in the goodness of others. One Rwandan victim argues: “Hutu used to pretend to be good people, and I would have allowed my daughter to marry them. But now I cannot allow that, because I recognize how bad they are,” while another says, “because of what happened to me I now never feel safe.”2 In cases where the regime committing genocide is deposed, there may be a process of transitional justice.3 We use the term here in a broad sense to encompass a set of practices and policies that aim to transform a society after mass atrocities. Transitional justice mechanisms have proliferated in recent years. In the past quarter-­century, truth commissions have been formed in 40 states; there are international tribunals with jurisdiction over 125 states, and domestic transitional justice proceedings in scores of others. Alternatively, a negotiated settlement may lead to an amnesty being granted to some or all perpetrators. Negotiated settlements, such as occurred in South Africa, often result in limited accountability supplemented by restitution to victims. There was also a kind of negotiated transition in Bosnia, but the presence of external powers demanding justice incentivized the prosecution of atrocity crimes. Regime change and transitional justice also result in a significant period of disruption as the state attempts, once again, to shift moral frameworks. This may entail institutional and legal reforms, accountability procedures, lustration (removing perpetrators and those involved with the ancien régime from positions of power), and vetting (preventing those implicated in violence from occupying state positions). These three scenarios – regime displacement, regime victory, and negotiated settlement – also have implications for how individual

After genocide I   219 perpetrators remember and frame their crimes. The image of the perpetrator in society is closely tied to transitional justice strategies. In post-­Nazi Germany, perpetrators were depicted as exceptional – a view that accorded with the very limited pursuit of criminal accountability. In contrast, in Rwanda, perpetrators have sometimes been presented as representative of all Hutus – a perspective that flows from the government’s maximal prosecution strategy, as well as its need to repudiate old ethnic power structures, and buttress new ones. In Germany, the “major offenders” were those who committed crimes “out of political motives” and those who “participated in killings, tortures, or other cruelties.”4 Moreover, “activist” offenders and “lesser offenders” were judged and offered limited punishment “according to their character” and to their ability to “fulfil their duties as a citizen.”5 Consequently, many alleged perpetrators were exonerated or had their proceedings abandoned. By 1949, in the American Zone, the proportion was 35.2 percent, yet by 1950 in the British Zone, 87.8 percent of cases had been abandoned or the subjects designated as “non-­incriminated persons.”6 The French Zone was lower, at 52.7 percent, and for Federal Territory proceedings it was 67.7 percent.7 Perhaps this should not be surprising when we consider that a survey of Germans conducted in the American zone of occupation in 1945 found that 20 percent professed to “fully supporting” Hitler’s Jewish policy and another 19 percent were “generally in favor but felt that he had gone too far.”8 In this chapter, we will discuss the aftermath of genocide – including how the past can be dealt with to rebuild society and prevent future crimes. The discussion begins with a consideration of how to dismantle the criminogenic state. It then analyzes perpetrator narratives, both individual and collective, as well as perpetrator trauma and rehabilitation. The period after genocide is a challenging one where many of the processes that underscore genocide may continue, if they remain unchecked. The perpetration of genocidal violence has significant consequences for both individual and society.

Reforming the criminogenic state Our foregoing analysis of genocide concluded that the genocidal state is criminogenic, and that this plays a key role in generating and facilitating genocide perpetration. Any consideration of genocide prevention must proceed from this understanding; consequently, the criminogenic state must be reformed. In the period following genocide, sources of both continuity and rupture are evident. Considering that the moral context itself has turned criminogenic, a return to a conventional moral context necessarily entails further moral revision, a further moment of rupture. This rupture is not as significant as the initial transformation facilitating genocide, for it entails a return to conventional values rather than a challenge to those values. As a Rwandan perpetrator attests: “my view of the Tutsis changed during the genocide; afterwards it returned to normal.”9 Nonetheless, the state must be transformed, former elites must be removed and possibly punished, and institutions must be dismantled. Beyond public institutions, the “deep state” may also be a focus of transformational efforts,

220   After genocide I undermining the power base of the genocidal elites. Social and economic inequalities may also need to be addressed to fulfil the requirements of social justice and reduce the vulnerability of the victim group. Such transformations will always involve a modicum of social disruption, and may have detrimental effects for individuals and groups who benefited from the previous system. One Rwanda perpetrator remarks: “I was afraid at the end of the genocide when the new government was announced.”10 In transforming the state after genocide, we must also consider which sources of continuity should be maintained, and which displaced. This consideration may involve weighing short-­term stability against the long-­term benefits of accountability. The polity, as constituted by an extremist ideology (the genocidal state), is a vector for the marginalization of targeted groups, and the mobilization of violence against these groups. Demagogues use politics to reshape the moral context – to redefine moral rules and the scope of moral protections. Ethics, rather than being universal, are bounded by community. In the criminogenic state, the commission of grave crimes becomes positively normative or acceptable, as moral rules are revised to facilitate mass atrocities. Genocide can only be understood within the enabling moral context created by state power. Individuals adjust their moral rules in response to changed conditions, as dictated by coercive and persuasive forms of power. If mass atrocities, such as genocide, are driven by criminogenic states exercising their normative and coercive power over individuals, then how can these enabling structures be dismantled? The power structures enabling genocide extend beyond institutions, and permeate society itself – manifested in harm-­producing relations of power. How can politics be reconfigured after mass atrocities in ways that reduce the likelihood of future violence? And what role can transitional justice mechanisms play in this reconfiguration? To assess the impact of courts and tribunals on the reconstitution of the polity after mass atrocities, we can hypothesize that potential impacts may include the de-­legitimization of extremist political factions and leaders, and a consequent reconfiguration of electoral politics towards moderate factions. It may also feature the removal of extremist leaders (and perpetrators) through trials and lustration processes. Other indirect effects may include normative demonstration and deterrence, the restoration of respect and rights to the victimized community, and the revision of destructive narratives (hate propaganda and mythohistory). Each of these areas is subject to claims by proponents of transitional justice, but are these claims valid? Transitional justice does not always delegitimize leaders. Rather, some may use transitional justice (especially where it is externally imposed) to bolster their legitimacy, presenting themselves as defiant against the forces of imperialism and outside interference. This has been the case, to some extent, with Kenyan president Uhuru Kenyatta, who was elected in Kenya while under investigation by the ICC, or Serbian nationalist politician Vojislav Šešelj, who frequently presents himself as a sort of Serbian martyr to international persecution (although he was acquitted by the ICTY).

After genocide I   221 The ICC indicted several Kenyans in relation to the 2007–2008 election violence. One would expect that such charges would isolate those indicted, seriously damaging their political reputations; yet the 2013 Kenyan presidential election was won by the indicted Uhuru Kenyatta; alongside his running mate, William Ruto, who was also indicted. During their campaign, Kenyatta and Ruto successfully leveraged the political impact of the ICC charges by crafting a narrative of Kenyan persecution at the hands of the great powers. This narrative evoked Kenya’s history of brutal colonial rule, and the Kenyan government successfully mobilized political support among many other African states, as well as the African Union. Transitional justice efforts are, almost by definition, highly selective. Often those who are the most responsible are not tried, or are tried and acquitted due to evidentiary shortcomings. The removal of extremists and those involved in violence through lustration processes may also be highly problematic. In Germany’s post-­World War II de-­Nazification campaigns, individuals responsible for crimes sometimes effectively bought their exoneration (by purchasing denazification certificates on the black market). In contrast, de-­Ba’athification in Iraq was so wide-­ranging and indiscriminate that many former Ba’ath party officials, who may or may not have been responsible for crimes, were marginalized and subsequently recruited by anti-­government forces, such as the Islamic State. Nonetheless, removing extremists from positions of power may help to shift discourses within the state and the wider society. Transitional justice may also produce indirect effects on the criminogenic state and genocidal context. The dismantling of the criminogenic state is ultimately a matter of undoing harm-­producing relations of power and the criminogenic moral context. Ideally, these efforts will lead to a reconfiguration of institutionalized power structures and relationships. Consequently, the range of action alternatives for individuals in post-­genocidal societies should disincentivize participation in genocide or other harmful acts. A primary goal in undoing harm-­producing relations of power must be the reintegration of victims into the moral community. This demarginalization demands that victims and victim groups must be recognized by the state as possessing the full range of rights and entitlements that citizenship entails. It also means, of course, that harmful practices against the victim group must end immediately – including all acts of marginalization and persecution. Legislation must be enacted and enforced to prohibit hate speech and incitement, in accordance with international law. Racist discourses contribute to the process of marginalization, as well as the mobilization of killing itself. Thus, actively enforcing the prohibition on hate speech and incitement to genocide is essential to the prevention of genocide. Finally, the crimes of the past must be recognized, and those responsible must be punished. The juridical processing of harm recognizes the intrinsic dignity and humanity of the victims, as does the reestablishment of a rule of law not bounded by ethnicity. Courts and collective-­memory practices re-­narrate victims’ trauma in a manner appropriate for public consumption and suitable for political

222   After genocide I and policy objectives.11 Yet for some victims, “collective memory of the general can feel like private forgetting of the individual,” especially where individual memories diverge from collective narratives.12 The recognition of crimes also occurs through other mechanisms, such as truth commissions, official apology, memorialization, and restitution. The post-­genocide state must build a more equitable and democratic future. Economic and political inequalities must be ameliorated. The instruments of state power must also reflect the overall composition of the population, with special consideration to the representation of underrepresented groups. Harm-­ producing relations of power are created in part through the absence of alternative discourses and sources of power. Reforming institutions after genocide involves both removing leaders involved in atrocities, as well as the reform of core institutions. Measures must also be taken to address organizational cultures that encourage the production of harm. Violence must be limited by legal frameworks: it cannot be directed at entire populations, and must always be subject to legislative and judicial review. More generally, the narrowing of awareness in genocide must be reversed through more inclusive, civic forms of nationalism, as well as the democratization of power and discourse. Power must neither be ethnicized nor monopolized. Genocide is an expression of absolute intolerance: perpetrators hold the very existence of the targeted group to be intolerable. Thus, tolerance is the antithesis of genocide. This tolerance must be ingrained in the “national ideology” of the state and in prevailing concepts of citizenship. Nation-­states, founded on the notion that the nation is synonymous with the state, are premised on an intolerance of alternative identities. In contrast, states rooted in civic nationalism emphasize common citizenship as at the core of state identity, and designate all bearers of citizenship as full bearers of rights. Efforts must be made to engender cooperation and interdependence in states that exhibit a high degree of social segregation (such as ethnic polarization). Studies show that even extended intergroup contact (with friends of friends who are in the out-­group) provides important information on how to behave towards members of out-­groups.13 The challenges in rebuilding social relations after genocide are enormous. A Srebrenica survivor argues: If I had the opportunity now to go to kill ten Serbs, and after I have killed ten Serbs, they put me in jail for 30 years, I would accept it. Because in this moment, the Serbs which have done crimes, they are coming to jail, and they do a short time there, 10 to 20 years for [killing] 10,000 people. It would be acceptable to me to kill ten Serbs and to go to jail for 30 to 40 years because of that. And even then, it would not bring the dead back.14 We cannot demand that victims forgive. Thus, holding perpetrators individually accountable for genocide is essential to rebuilding trust in the system.

After genocide I   223

The evolution of perpetrator narratives Narratives constitute stories about what happened during genocide and why. In the post-­genocide period, perpetrator narratives evolve in response to the broader social context. These narratives also reflect perpetrators’ self-­image, their audience (whom they are addressing), and instrumentality (the purpose of the narrative, aimed at maximizing personal standing). Narratives that emphasize the rightness of genocide (pathological or subcultural narratives) will persist when genocide remains justified by the state and society. This section considers how perpetrators remember past crimes, individually and collectively. It also examines the revision of narratives, and asks whether such revision assists in rehabilitating perpetrators and pursuing broader goals of transitional justice and peacebuilding. Memory Collective memory Collective crimes engender collective memory, and memory itself constitutes a sort of moral framework, a way of viewing past harms. Collective memories also function as a way of constituting or reconstituting moral fields: acts of memory may seek to celebrate, justify, condemn, or forget harm done by and to collectivities. Thus, where collective memory is contested, morality is also contested – the existence of multiple contested memories means that moral frameworks are also multiple and contested. The contestation of moral frameworks may have fundamental implications for the ways in which power is used in the present, particularly regarding the continuing exclusion and persecution of victim groups. Yet contestation also reflects a critical space in society to discuss, and question, the past. The configurations of collective memory shift and evolve over time to meet the needs of the political present. Repudiating genocide is in the interest of regimes that represent a reconfiguration of elites and a break from the past. Alternatively, where the genocidal regime successfully completes genocide, collective memory will be manifested in discourses of denial or reversed morality. Through its transitional justice policies, “the state involved is trying not only to illuminate the past, but also to justify the present.”15 Genocide may also represent a kind of tabula rasa in cases of regime change – the incoming regime is presented with space in which to build a new future and new identities on the ruins of the past. Through transitional justice the perpetrators are identified and condemned in an official record-­making process. Trials, despite their drawbacks, present the possibility of reintegrating victims and perpetrators into the same moral framework (the law). This will not occur where justice is perceived as unfair; rather, in such circumstances, trials may contribute to the marginalization of new groups and further conflict. Notions of collective blame alienate those within the “perpetrator group” who might be inclined to contribute to the peacebuilding project. They may also

224   After genocide I contribute to collective shaming, without reintegration – a situation which alienates individuals and groups rather than rebuilding the polity.16 What happens in the courtroom will also influence narratives outside it. Perpetrators orient themselves by these narratives. Where perpetrators have successfully completed genocide, and remain in power, perpetrator neutralizations will persist, emphasizing the reversal of morality and not just the reduction of moral costs. In other words, they will continue to justify genocide. This is, to some degree, the way that the crimes of 1971 are framed in Pakistan, the 1965 mass killings of communists in Indonesia, and the Guatemalan genocide of 1981–83 by much of the political establishment in Guatemala. Collective memory narratives denying genocide often claim equivalency: that all sides committed atrocities, and thus the moral culpability is evenly distributed. This equivalency erodes efforts to come to the terms with the past, as it denies the collective nature of victimization in genocide. Moreover, it is often claimed that the past is ambiguous – that it is unclear what happened. Such collective denial narratives are present in almost all cases of genocide denial, including the Turkish denial of the Armenian genocide, which has become formalized in state structures. Descendants of perpetrator groups often seek to “exonerate their ancestors’ misdeeds.”17 Moreover, studies show that when it is the in-­group that commits atrocities, “individuals dehumanize the victims more and show a lesser tendency … to punish the (in-­group) perpetrators.”18 Memory also contextualizes and neutralizes individual acts of violence. Where the collective context excuses individual perpetration, individual perpetrators will persist in viewing their acts as rightful. This places the victims and members of the victim group in a situation of vulnerability. The individuals who victimized them remain free, communicating implicitly the rightfulness of the acts of the perpetrators and the continuing exclusion of victims from the moral community. Moreover, even uninvolved members of the “perpetrator group” (the collectivity identified with the perpetrators) are often faced with collective responsibility for the crimes committed. In societies with contested and competing memories of mass violence, these memories are often instrumentalized, meeting the needs of groups for reframing their behavior and experiences in ways that contribute to a positive self-­image. Therefore, these competing memories also function as competing moral contexts, within which the individual is often presented as a heroic victim, a perpetrator with no choice, or an unaware bystander. Individual memory Individual memory of genocide also flows from this collective context. Unlike most other crimes, genocide is collective, meaning that memory is also politicized and collectivized. Individual perpetrators incorporate external narratives (produced and disseminated collectively) into internal narratives of the past (memory). This incorporation of narratives parallels the localization of violence in genocide – collective narratives provide scripts for individuals to interpret in articulating their own memories of violence.19 Where the genocidal regime is

After genocide I   225 deposed from power, perpetrators account for their own crimes, as well as those of their peers, and must overcome previously internalized narratives, and systems of denial. Deception, accordingly, is also produced collectively, for reasons that are instrumental to the collectivity. The collective nature of perpetration may also facilitate perpetrator neutralization – as perpetrators can reframe their crimes as relatively insignificant, and themselves as lacking agency, in the collective context of mass perpetration. Individual memory, therefore, is also subject to political instrumentality through formal narrative control – that is, attempts by authorities to shape and control narratives.20 Where perpetrators have been displaced from power, formal narrative control repudiates genocide, which has the effect of strengthening techniques of neutralization that emphasize the reduction (rather than the reversal) of morality. In other words, the overall conduct is seen to be wrongful, so individuals emphasize their lack of agency. A shift is evident over time from the reversal of morality to the reduction of moral costs. This shift also represents the reconstitution of moral frameworks. However, in many transitional justice processes (including most trials and truth commissions), perpetrators are given a reduced sentence or even granted amnesty in return for confessions. Such procedures may encourage people to come forward to provide information about past violations, yet they also encourage false confessions. This is not always accepted by victims: I am not satisfied with what the government did, because they took the Interahamwe to the prison. Then after they released them because of confession. But it was not really a confession from the heart, it was pretending […] then they come out and they kill the witnesses. The government has not taken the problem seriously.21 The realignment of perpetrator narratives may of course be instrumental, but even in such cases, professing responsibility may eventually lead to a transformation of perpetrator identities. In the early days of Rwanda’s transitional justice process, those who confessed were often ostracized by other prisoners, though they received benefits within the prison system.22 But bandwagon effects are also present with regard to perpetrator confessions. “Confessional chains” may occur where the confession of one perpetrator encourages others to come forward.23 Confessions may produce conflicts in society, but ultimately, they may also provoke critical reflection on the past.24 Even when they do occur, confessions are generally full of contradictions, as with my interview subjects; perpetrators argue that their conscience demanded they come forward, and then deny having killed.25 On a more physiological level, perpetrator memories naturally vary. Research on perpetrators of “ordinary” violent crimes suggests that it is much more common for perpetrators to experience difficulty in remembering, compared with victims and bystanders.26 For some perpetrators a discrepancy between the internal emotional state at encoding and normal state may account for memory loss.27 In other words, perpetrators often encode (record) memories of

226   After genocide I perpetration during moments of heightened emotional arousal, and this may contribute to memory loss. Some perpetrators of violence may experience so-­called “red-­outs” – states of heightened emotional arousal that contribute to derealization or dissociation during perpetration. For individuals who experience trauma-­ related memory loss, memories are usually partial but not temporally-­defined. In other words, perpetrators who profess to have lost all memory of the moment (or day) of perpetration are generally malingering. Such malingering is very common in perpetrator narratives, as is the reframing of violence more generally through moral neutralization. Shame also appears to have a negative effect on working (short-­term) memory functions, however one would not expect this to impact memories of events that happened months or years ago.28 Perpetrator denials of past wrongful behavior are ubiquitous, and unsurprising if one considers the need to maintain a sense of self in relation to prevailing moral norms. Self-­deception is common not only among perpetrators, but among humans more generally.29 Pauline Nyiramasuhuko, for example, was convicted of genocide in Rwanda, but argued that “it was the Tutsi who massacred the Hutus.” One of my interview subjects denied being aware that genocide was even occurring, even though he lived in Rwanda at the time, while a Cambodian respondent claimed he was only “transporting prisoners to Choeung Ek [the killing fields],” when in fact he was also leading the executions.30 Denial may also be linked to the role margin of discretion and the externalization of responsibility. Low-­level perpetrators may argue that they were unaware of broader patterns of violence, while high-­level perpetrators contend that they could not be aware of all the actions performed by their subordinates. On the collective level, the state often mitigates its responsibility by seeking to decontextualize individual crimes – the perpetrators acted autonomously rather than as part of institutions – while individual mitigation often emphasizes context, that is, the existence of a much larger perpetration system. Trauma Collective trauma There is little research or academic literature on genocide perpetrator trauma. The act of perpetration is itself morally repugnant, and thus it seems dissonant to consider the trauma suffered by some perpetrators, as being worthy of concern. However, perpetrator trauma and rehabilitation must be considered in any transitional justice framework, as part of rebuilding societies after genocide and preventing future violence. Much has been written about the notion of cultural trauma among victim groups, but is it possible that members of the group identified with perpetration may also experience a form of cultural trauma? How might this shape collective and individual identity? The state acts on behalf of the broader political and cultural community. Yet it does so without their full consent. Can the crimes of the state then be imputed to individual members of the political community? This is

After genocide I   227 a difficult question to answer. Undoubtedly, mass acquiescence facilitates genocide. The bystander paradox remains: we cannot necessarily expect individuals to array themselves against state power, particularly where that state power has already demonstrated a willingness to kill entire populations and categories of persons. Yet resistance, when manifested by many individuals, would likely be powerful enough to stop mass atrocities. The mere existence of the question of collective responsibility may say something about the possibility of cultural trauma among perpetrators. Cultural trauma among victims posits that individuals may experience trauma through their membership in a collectivity targeted for genocidal destruction.31 Perhaps, then, the imputation of collective responsibility for a horrendous crime can also manifest itself as trauma among members of the perpetrator group. Although it took decades, Germany’s attempt to confront the Nazi past may also have muted German nationalism and collective identity. It is difficult to generate a collective sense of pride when the collectivity has engaged in moral outrages – crimes against humanity, to use the legal parlance. The group has become identified with evil, and identification with the group thus produces shame. This trauma may be manifested on the family level, as individuals come to realize that a parental figure perpetrated or was complicit in genocide. This may produce cognitive dissonance, as individuals seek to reconcile positive views of their group and family with acts of tremendous evil. Individual trauma Individual perpetrator trauma is also a taboo subject. But it seems clear that many perpetrators do experience such trauma. This is evident in perpetrators’ attempts to rationalize their actions, along with other post facto coping mechanisms to ease recidivism and mitigate psychological injuries. According to a former LRA “bush wife”: “[participating in violence] affects a lot of people; [some] come back completely traumatized because they were forced to kill innocent people. And now those people are being affected by the innocent blood that they shed in the bush.”32 Little research has been conducted on post-­traumatic stress among perpetrators of genocide and other mass atrocities. An Einsatzgruppen psychiatrist estimated PTSD (Post Traumatic Stress Disorder) prevalence at 20 percent within his unit, while a study of Rwandan genocide perpetrators concludes that only 5 percent displayed PTSD symptoms.33 This can be compared to a study of victims of the Cambodian genocide that showed PTSD rates of 62 percent and depression rates of 51 percent.34 Turning to research on ordinary violent crimes, we see that it is not unusual for offenders to experience PTSD.35 For instance, one study reported that almost half of perpetrators of violent crimes reported “intrusive memories” of their violent offences, with intrusion being more likely for those with lower levels of anti-­social beliefs.36 Perhaps international crimes are less conducive to post-­ traumatic stress, given the institutionalized externalization of responsibility, as well as the politicized nature of the crimes themselves. Individuals are less likely to feel distress. Yet killing remains traumatic for many perpetrators irrespective

228   After genocide I of these factors. Moreover, veterans of other forms of political violence, such as armed conflict, often experience PTSD, as well as heightened rates of alcohol abuse, anxiety disorder, and suicide. This psychological trauma may even serve as a “push factor” for committing crimes.37 PTSD is more likely among soldiers who kill in combat than among those who do not.38 Perhaps most concerning, many perpetrators of genocide may experience heightened levels of appetitive aggression: they feel a “need” to commit aggressive acts.39 This is positively correlated with gender (men exhibit a higher propensity) and with the number of offences committed.40 It also corresponds with higher ranks in the armed group and repeated reenlistment.41 In other words, for perpetrators habituated to killing, it may be difficult to leave the experience behind. Even if genocide carries political justifications, killing remains, in some respects, a personal act. Perpetrators may avoid trauma through moral neutralization to avoid developing a deviant self-­identity. Such neutralization “can mitigate both social disapproval and a personal sense of failure.”42 However, for perpetrators who already regard themselves as violent, it will be much easier to integrate the “perpetrator self ” into their overall self-­concept. Thus, we can hypothesize that in such cases, there is a reduced need for neutralization and a lesser risk of dissonance.43 Perpetrators who feel shame (inward-­looking negative feelings about their acts) are more likely to suffer from depression than those who feel guilt (victim-­oriented feelings).44 For many perpetrators of violence, the most traumatic memories relate to the moment when the meaning of the events changes: for example, the point at which a victim stops moving and breathing.45 In this moment, the perpetrator has fully transitioned to becoming a murderer. For some, this profound moment of moral transgression might be eased by reference to the moral context in genocide. Yet the reality of ending the life of another human being remains. In the short term, perpetrator resilience to trauma is highly problematic, as it may facilitate repeated killing. In the long-­term, however, resilience may have social benefits, when perpetrators return to their communities after incarceration. Rehabilitation Collective rehabilitation As a collective crime, genocide requires collective rehabilitation. This may be accomplished through efforts to come to terms with the past, such as the prosecution of perpetrators, official apologies, and reparations for victims. Collective rehabilitation, of course, also implies that responsibility is shared. As the embodiment of the political community that committed atrocities in the first place, collective rehabilitation is a state-­directed project. In a sense, then, collective rehabilitation is an inward-­facing project in which the perpetrator (the state) rehabilitates itself for the crimes committed. This rehabilitation may also be partly externally imposed, as in Germany after World War II, where the Allied powers themselves tried German leaders for crimes of state.

After genocide I   229 Rehabilitation cannot require restitution, as true restitution for genocide is impossible by definition – the victims are dead. According to a survivor of the Rwandan genocide: I wonder how people who used to share drinks … how could people do something like kill each other? Considering what they did, nothing can be repaired … because of that there is nothing to do but pardon them. Nothing can replace what we lost.46 Of course, individual and collective reparations can be paid to families of victims, or to the community; but such compensation may be rejected. When asked about the causes of genocide, Rwandan perpetrators and victims both cited bad leadership but perpetrators also emphasized material motives, while victims rejected this reasoning entirely and identified hate, evil, and bad morals as being key factors. For victims then, the perpetration of genocide is seen as more of a characteristic and less of a transient act. Individual rehabilitation Rehabilitating individual perpetrators is a daunting prospect. Rehabilitation is grounded in what Francis Allen calls “the rehabilitative ideal” – “the idea that criminal behavior has certain identifiable causes and that by treating offenders and addressing these causes (through interventions), the risk of becoming a re-­ offender may be reduced.”47 Reducing the risk of recidivism is challenging, considering the role of state power in perpetration. However, as this book argues, many perpetrators would be unlikely to perpetrate without the presence of a facilitative (and rare) context. High-­level perpetrators are also responsive to the historical context, yet they differ in their additional power to generate the context. So how might perpetrators be rehabilitated? The approach taken by international criminal tribunals to the rehabilitation of perpetrators is somewhat fragmentary. In determining whether perpetrators are rehabilitated (and eligible for early release), the ICTY and ICTR draw from a variety of loosely-­conceptualized indicators: conduct in prison, future prospects, reflection on crimes (including remorse), and personal characteristics such as age, prior conduct, and mental health.48 Conduct in prison was cited in 87.3 percent of decisions, while personality traits were cited in only 12.7 percent of cases.49 While some consistency is evident in the factors considered important by the tribunals, there is little clarity as to how these factors can be measured, or whether they are even important to the success of perpetrator reintegration after release. Perhaps, as Shadd Maruna argues, to desist from crime ex-­offenders must create coherent prosocial identities in story form: what Maruna calls “redemption scripts.”50 The creation of this prosocial identity then requires the revision of moral narratives justifying perpetration. The narratives of ex-­convicts who desist from further crime have three important distinguishing features: they establish (or re-­establish) core beliefs and a sense of the “true self ”; they maintain a positive

230   After genocide I view of free will and the ability to control one’s destiny; and they express a desire to contribute to society.51 Many of the perpetrators interviewed for this book do articulate a sense that perpetration is at odds with their beliefs, as well as a desire to contribute to society. However, the sense of free will is still frequently absent. Rather, they evince a mood of fatalism, arguing that they had no choice but to perpetrate, and indeed may still lack power over their lives. To be sure, this fatalistic mood is rooted in the need to disclaim responsibility, but the influence of criminogenic state power also makes perpetrator fatalism somewhat understandable, as does the limited social power of some perpetrators. Paradoxically, perpetrators in positions of authority may have a broader scope for denial, as they often never witness the murderous consequences of their actions and instructions. Many perpetrators in Rwanda told me that they felt they were better persons than in 1994: “I am different, because at the time of the genocide I was taught to kill, to commit the crime. But now I have been taught to be a good person, to have reconciliation with my country.”52 Another claims, “I am [a] different person now because I know God … I’ve changed completely.”53 Even for those who consider themselves victims of circumstances, such redemption stories may be powerful. Fully 95 percent of the perpetrators I interviewed in Rwanda argued that their personality had changed since the genocide. Yet narratives of denial may also be psychologically functional, allowing individuals to maintain positive self-­identities. There is evidence that excuses (including moral neutralization strategies) do work; they protect the self from threats, increase self-­esteem, decrease depression, improve physical health, and ease social relationships.54 Prosocial identities do not necessarily involve perpetrators admitting that they are bad people. Moreover, there is countervailing evidence that accepting responsibility and/or expressing remorse reduce recidivism.55 We must also consider the differing forms of moral neutralization. Those based on the reversal of morality (such as “they are evil”) are likely more associated with “persistent criminality,” while seeking to reduce moral costs may serve to distance the individual from prior crimes (the perpetrator self ).56 Liau et al., argue that self-­serving distortions “protect the self from blame or negative self-­ concept,” while self-­debasing distortions are those associated with a negative self­concept.57 Certain types of neutralization may, in some ways, be a positive sign. It indicates that perpetrators are moving away from subcultural narratives of violence and seeking to align themselves with broader societal ones.58 However, as we know, unthinking conformity with societal norms may also be dangerous. If Lifton is correct in his “doubling” hypothesis, rehabilitation may also involve eliminating the perpetrator self. Perhaps this process is facilitated by the full confession of wrongful acts, reconciling the perpetrator self with the normal self. This elimination of perpetrator selves is not an explicit goal of accountability processes such as trials, yet legal procedures may reshape self-­identity to some degree. The reconstitution of individual identity is a goal of some rehabilitation programs after conviction. For example, in Rwanda, perpetrators are required to undergo extensive “re-­education” through the ingando program, where they are presented with a historical and civic narrative of Rwandan history and identity.

After genocide I   231 The avoidance of moral and legal accountability is a central theme of this book, and more broadly reflects the process of genocidal perpetration. This tendency to avoid the moral dissonance engendered through killing is common to both soldiers and perpetrators of atrocities. The military often employs euphemisms like “collateral damage” and “tonnage” (in the case of ships destroyed) to describe killing.59 Soldiers telling stories of atrocities often include a negative character, by contrast with whom their own behavior appears more honorable.60 This tendency has also been identified among criminals. Bandura labelled it “exonerating comparison,” while Sykes and Matza referred to “the claim of relative acceptability.” A Rwandan perpetrator spoke at length of those who committed worse atrocities than he, concluding: “I killed one person, and I am proud to say that.”61 Another prisoner who appeared concerned about being evaluated negatively admitted to having killed “at least six” individuals, but refused to give me an exact number of those he had killed.62 Prisoners often construct moral hierarchies, perhaps to diminish the moral toxicity of their acts of transgression and establish some claim to moral superiority.63 In their book Soldaten, based on the secretly-­recorded conversations of German prisoners of war in Britain, Neitzel and Welzer noted that “by differentiating himself from others, the typical storyteller was able to find a space within a larger criminal endeavor in which he could not be accused of behaving immorally.”64 Moral hierarchies also seem to exist among perpetrators of genocide; this may explain the reluctance of most perpetrators to admit to acts of sexual violence (even when they confess to killing), as well as the apparent taboo against admitting that one took pleasure in killing. In Rwanda, many génocidaires also blamed the most extreme acts of violence (“slow deaths”) on socially-­ostracized groups such as the homeless.65 We can hypothesize here that the prevention of genocide requires the (re)establishment of a moral community valuing victims and where offenders recognize the harmful effects of their act.66 The rehabilitation of individual perpetrators of genocide may be contingent on constructing new narratives of involvement in violence that claim actual culpability for atrocities.

Conclusion: after genocide This much is clear: genocide does not end with killing. For victims and perpetrators alike, its ramifications may be felt for a lifetime. For both sets of actors, memory and trauma are situated in broader processes of accounting for the past. Genocidal trauma lives in a dreamtime, where the past has not really passed, where suffering cannot ever be truly undone. Processes of genocidal distancing and victimization may continue long after genocide has abated. This is particularly true when the perpetrators have “won” or are not held responsible. In this scenario, victims may remain outside of the moral community, and memory itself becomes an affront to the sensibilities of the perpetrators, who would usually rather forget. “That was the story of the past,” retorted a former Khmer Rouge prison photographer to my repeated questions.67

232   After genocide I

Notes   1 Anderson, “State Deviancy and Genocide.”   2 Interviews R49 and R44 (Rwanda), Nyamata, September 2009.   3 Transitional justice is defined by the International Center for Transitional Justice as: “the set of judicial and non-­judicial measures that have been implemented by different countries in order to redress the legacies of massive human rights abuses.” These measures may include prosecutions and non-­judicial measures such as truth commissions, reparations for victims, and institutional reforms. See, for example: Thoms, Paris, and Ron, The Effects of Transitional Justice Mechanisms; and Van der Merwe, Baxter, and Chapman, Assessing the Impact of Transitional Justice.   4 Control Council Directive No. 38 (October 12, 1946).   5 Davies, Rationalising the Irrational, 15.   6 The London Agreement, August 8 1945, in Ruhm von Oppen, Documents on Germany under Occupation, 50.   7 Fürstenau, Entnazifizierung, 227.   8 Midlarsky, The Killing Trap, 185.   9 Interview R58 (Rwanda), Rugerero TIG Camp, September 2009. 10 Interview R40 (Rwanda), Nyasenge TIG Camp, August 2009. 11 Dawes, Evil Men, 31. 12 Dawes, 32. 13 Turner, et al., “A Test of the Extended Intergroup Contact Hypothesis,” 856. 14 Interview B06 (Bosnia), Srebrenica, April 2009. 15 Simon, Law, War & Crimes, 91. 16 Braithwaite and Makkai, “Reintegrative Shaming and Compliance,” 361–385. 17 Vollhardt and Bilewicz, “After the Genocide,” 4. 18 Castano, Leidner, and Slawuta, “Social identification processes,” 262. 19 For more on the connection between collective and individual memories see: Bartlett, Remembering. 20 Herve, Cooper, and Yuille, “Memory Formation in Offenders,” 52. 21 Interview R66 (Rwanda), Kigali, October 2009. 22 Tertsakian, Le Chateau, 397 and 404. 23 Fuoss, “Lynching Performances, Theatres of Violence.” 24 Payne, Unsettling Accounts. 25 Tertsakian, 412. 26 Christianson, “Searching for Offenders’ Memories,” 15. 27 Christianson, 15. 28 Cavalera and Pepe, “Social emotions and cognition,” 462. 29 Sahdra and Thagard, “Self-­Deception and Emotional Coherence,” 213. 30 Prosecutor v. Pauline Nyiramasuhuko et al., (Judgement and Sentence); Interview C05 (Cambodia), countryside near Phnom Penh, November 2008. 31 Meierhenrich, “The Trauma of Genocide.” 32 Interview U09 (Uganda), Gulu, July 2009. 33 MacNair, Perpetrator-­Induced Traumatic Stress, 48; and Schaal, Heim, and Elbert, “Posttraumatic Stress Disorder,” 930. Verduin, et al., “Intimate Partner Violence,” 1847, also conclude that perpetrators report more mental health problems in Rwanda than victims. 34 Marshall, et al., “Mental Health of Cambodian Refugees,” 576. 35 Christianson, 5. 36 Evans, et al., “Intrusive Memories in Perpetrators of Violent Crime,” 138. 37 Walklate and McGarry, “Competing for the ‘trace’,” 187 38 Maguen, et al., “The Impact of Killing in War on Mental Health.” 39 Schaal, Heim, and Elbert, 939–941. 40 Schaal, Heim, and Elbert.

After genocide I   233 41 Hermenau, et al., “Growing up in armed groups,” 7. 42 Maruna and Copes, “What Have We Learned,” 232. 43 Christianson, 10. 44 Tangney, Wagner, and Gramzow, “Proneness to Shame, Proneness to Guilt,” 476. 45 Christianson, 106. 46 Interview R47 (Rwanda), Nyamata, September 2009. 47 Kelder, Hola, and van Wijk, “Rehabilitation and Early Release,” 1184. 48 Kelder, Hola, and van Wijk, 1186. 49 Kelder, Hola, and van Wijk, 1186. 50 Maruna, Making Good. 51 Maruna, 88. 52 Interview R36 (Rwanda), Rwaza TIG Camp, August 2009. 53 Interview R04 (Rwanda), Kigali Central Prison, July 2009. 54 Schlenker, Pontari, and Christopher, “Excuses and Character,” 16. 55 See Bagaric and Amarasekara, “Feeling sorry? — Tell Someone Who Cares: The Irrelevance of Remorse in Sentencing,” 364–376; Ware and Mann, “How Should ‘Acceptance of Responsibility’ Be Addressed in Sexual Offending Treatment Programs?,” 279–288. 56 Maruna and Copes, 227. 57 Liau, Barriga, and Gibbs, “Relations between self-­serving cognitive distortions,” 18. 58 Maruna, 144. 59 Neitzel and Welzer, Soldaten, 53. 60 Neitzel and Welzer, 82. 61 Interview R32 (Rwanda), Butare, August 2009. 62 Interview R24 (Rwanda), Tumba TIG Camp, August 2009. 63 Presser and Sandberg, “Introduction: What Is the Story?,” 26. 64 Neitzel and Welzer, 82. 65 Jessee, 173–174. 66 Hola, et al., “Does Remorse Count?” 67 Interview C01 (Cambodia), Siem Reap, October 2008.

Works cited Anderson, Kjell. “State Deviancy and Genocide.” In Genocide: New Perspectives on its Causes, Courses, and Consequences, edited by Ugur Umit Ungor. Amsterdam: Amsterdam University Press, 2016. Bagaric, Mirko and Kuma Amarasekara. “Feeling Sorry? — Tell Someone Who Cares: The Irrelevance of Remorse in Sentencing.” The Howard Journal of Crime and Justice 40, no. 4 (2001). Bartlett, Frederic C. Remembering. A study in experimental and social psychology. Cambridge: Cambridge University Press, 1932. Braithwaite, John and Toni Makkai. “Reintegrative Shaming and Compliance with Regulatory Standards.” Criminology 32, no. 3 (1994), 361–385. Castano, Emanuele, Bernhard Leidner, and Patrycja Slawuta. “Social identification processes, group dynamics and the behavior of combatants.” International Review of the Red Cross 90, no. 870 (2008). Cavalera, Cesare and Alessandro Pepe. “Social emotions and cognition: shame, guilt and working memory.” Procedia: Social and Behavioral Sciences, no. 112 (2014). Christianson, Sven A. “Searching for Offenders’ Memories of Violent Crimes.” In Narrative Criminology: Understanding Stories of Crime, edited by Lois Presser and Sveinung Sandberg. New York: New York University Press, 2015.

234   After genocide I Control Council Directive No. 38 (October 12, 1946), in Official Gazette of the Control Council for Germany, No.  11, October 31 1946, 184; available at German History in Documents and Images, http://germanhistorydocs.ghi-­dc.org/pdf/eng/Denazification% 202%20ENG.pdf. Accessed May 18 2016. Davies, Jessica. Rationalising the Irrational: How Perpetrators Comprehend the Phenomenon of Genocide. (MA thesis). Amsterdam: University of Amsterdam, 2015. Dawes, James. Evil Men. Cambridge: Harvard University Press, 2014. Evans, Ceri, Anke Ehlers, Gillian Mezey, and David M. Clark. “Intrusive Memories in Perpetrators of Violent Crime: Emotions and Cognitions.” Journal of Consulting and Clinical Psychology 75, no. 1 (2007). Fuoss, Kirk. “Lynching Performances, Theatres of Violence.” Text and Performance Quarterly 19, no. 1 (January 1999). Fürstenau, Justus. Entnazifizierung. Ein Kapitel deutscher Nachkriegspolitik. Luchterhand: Neuwied and Berlin, 1969. Hermenau, Katharin, Tobias Hecker, Anna Maedl, Maggie Schauer, and Thomas Elbert. “Growing up in armed groups: Trauma and aggression among child soldiers in the DR Congo.” European Journal of Pyschotraumatology, no. 4 (2013). Herve, Hugues, Barry S. Cooper, and John C. Yuille. “Memory Formation in Offenders: Perspectives from a Biopsychosocial Model of Eyewitness Memory.” In Narrative Criminology: Understanding Stories of Crime, edited by Lois Presser and Sveinung Sandberg. New York: New York University Press, 2015. Hola, Barbora, Joris van Wijk, Francesca Constantini, and Armi Korhonnen. “Does Remorse Count? ICTY Convicts’ Reflections on Their Crimes in Early Release Decisions,” International Criminal Justice Review (2018). Jessee, Erin. Negotiating Genocide in Rwanda: The Politics of History. London: Palgrave, 2017. Kelder, Jessica, Barbora Hola, and Joris van Wijk. “Rehabilitation and Early Release of Perpetrators of International Crimes: A Case Study of the ICTY and ICTR.” International Criminal Law Review, 14, no. 6 (2014). Liau, Albert K., Alvaro Q. Barriga, and John C. Gibbs. “Relations between self-­serving cognitive distortions and overt vs. covert antisocial behavior in adolescents.” Aggressive Behavior 24, no. 5 (1998). MacNair, Rachel. Perpetrator-­Induced Traumatic Stress: The Psychological Consequences of Killing. Westport, CT: Praeger, 2002. Maguen, Shira Thomas J. Metzler, Brett T. Litz, Karen H. Seal, Sara J. Knight and Charles R. Marmar. “The Impact of Killing in War on Mental Health Symptoms and Related Functioning.” Journal of Traumatic Stress 22, no. 5 (2009). Marshall, Grant, Terry Schell, Marc Elliot, S. Megan Berthold, and Chi-­Ah Chun. “Mental Health of Cambodian Refugees 2 Decades after Resettlement in the United States.” Journal of the American Medical Association 294, no. 5 (2005). Maruna, Shadd. Making Good: How Ex-­Convicts Reform and Rebuild Their Lives.” Washington, D.C.: American Psychological Association, 2001. Maruna, Shadd and Heith Copes. “What Have We Learned from Five Decades of Neutralisation Research?” Crime and Justice 32 (2005). Meierhenrich, Jens. “The Trauma of Genocide.” Journal of Genocide Research 9, no. 4 (2007). Midlarsky, Manus. The Killing Trap: Genocide in the Twentieth Century. Cambridge, UK: Cambridge University Press, 2005.

After genocide I   235 Neitzel, Sonke and Harald Welzer. Soldaten: On Fighting, Killing, and Dying. The Secret WWII Transcripts of German POWS. Frankfurt: Verlag Fischer, 2011. Payne, Leigh A. Unsettling Accounts: Neither Truth nor Reconciliation in Confessions of State Violence. London: Duke University Press, 2008. Presser, Lois and Sveinung Sandberg. “Introduction: What Is the Story?” In Narrative Criminology: Understanding Stories of Crime, edited by Lois Presser and Sveinung Sandberg. New York: New York University Press, 2015. Prosecutor v. Pauline Nyiramasuhuko et al. (Judgement and Sentence). International Criminal Tribunal for Rwanda, June 24 2011, ICTR-­98–42-T. Ruhm von Oppen, Beata (ed.), “London Agreement, August 8, 1945.” In Documents on Germany under Occupation, 1945–1954. Oxford: Oxford University Press. Sahdra, Baljinder and Paul Thagard. “Self-­Deception and Emotional Coherence.” Minds and Machines Journal for Artificial Intelligence, no. 13 (2003). Schaal, Susanne, Lale Heim, and Thomas Elbert. “Posttraumatic Stress Disorder and Appetitive Aggression in Rwandan Genocide Perpetrators.” Journal of Aggression, Maltreatment, and Trauma, 23, no. 9 (2014). Schlenker, Barry R., Beth A. Pontari, and Andrew N. Christopher. “Excuses and Character: Personal and Social Implications of Excuses.” Personality and Social Psychology Review 5, no. 1 (2001). Simon, Gerry. Law, War & Crimes, War crimes Trials and the Reinvention of International Law Cambridge: Polity Press, 2007. Tangney, June Price, Patricia Wagner, and Richard Gramzow. “Proneness to Shame, Proneness to Guilt, and Psychopathology.” Journal of Abnormal Psychology 101, no. 3 (1992). Tertsakian, Carina. Le Chateau: The Lives of Prisoners in Rwanda. London: Arves Books, 2008. Thoms, Oskar N. T., Roland Paris, and James Ron. The Effects of Transitional Justice Mechanisms: A Summary of Empirical Research Findings and Implications for Analysts and Practitioners. Ottawa: University of Ottawa Centre for International Policy Studies, 2008. Turner, Rhiannon, Miles Hewstone, Albert Voci, and Christina Vonofakou. “A Test of the Extended Intergroup Contact Hypothesis: The Mediating Role of Intergroup Anxiety, Perceived In-­group and Out-­group Norms, and Inclusion of the Out-­group in the Self.” Journal of Personality and Social Psychology 95, no. 4 (2008). Van der Merwe, Hugo, Victoria Baxter, and Audrey R. Chapman. Assessing the Impact of Transitional Justice: Challenges for Empirical Research. Washington, DC: United States Institute of Peace, 2008. Verduin, Femke, Esther A. N. Engelhard, Theoneste Rutayisire, Karien Stronks, and Willem F. Scholte. “Intimate Partner Violence in Rwanda: The Mental Health of Victims and Perpetrators.” Journal of Interpersonal Violence 28, no. 9 (2013). Vollhardt, Johanna Ray and Michal Bilewicz. “After the Genocide: Psychological Perspectives on Victim, Bystander, and Perpetrator Groups.” Journal of Social Issues 69, no. 1 (2013). Walklate, Sandra, and Ross McGarry. “Competing for the ‘trace’ ” The Legacies of war’s violence(s).” In Criminology and War: Transgressing the Borders, edited by Sandra Walkate and Ross McGarry. New York, Routledge, 2015. Ware, Jayson and Ruth E. Mann. “How Should ‘Acceptance of Responsibility’ Be Addressed in Sexual Offending Treatment Programs?” Aggression and Violent Behavior, 17, no. 4 (2012).

11 After genocide II Justice

Introduction: transitional justice and the prosecution of genocide How can criminal justice respond to the complex challenges posed by perpetration? Genocide is arguably the most difficult crime to prosecute. The challenges flow from the nature of genocide, as a crime of conformity, the complexity of genocidal processes, and the shortcomings of the law. Legal processes do not align well with the social reality of genocide. This misalignment means that individuals responsible for genocide are sometimes not successfully prosecuted, also that those prosecuted are often subject to theories of criminal responsibility that are divorced from genocidal conditions. Where the law is vindictive rather than nuanced, it may create new harm-­producing relations of power. In this chapter, we will consider several of the challenges in prosecuting genocide, including the attribution of genocidal intent and individual criminal responsibility, and the effects of legalized narratives. We must also consider the purpose of prosecution. Is it merely punitive, or can it also contribute to preventing genocide through deterrence? Transitional justice is often seen as a way of generating dialogue about the crimes of the past. Yet transitional justice is a process of moral closure – it is not the opening of a conversation, rather its conclusion. We know that courts make judgments, but so do truth commissions and commissions of inquiry in their final reports and recommendations. This process of moral closure also contributes to establishing and consolidating collective memory. Transitional justice, it must be noted, is also part of the exercise of power. Even where transitional mechanisms encourage grassroots participation, they are still generally established by the state to serve political ends, alongside other goals. The establishment or non-­establishment of transitional justice mechanisms corresponds closely to broader societal considerations as to the valuation of the victim group(s). This chapter will consider the prosecution of genocide in terms of several critical issues, including: the problem of genocidal intent, the attribution of responsibility, and the deterrence of future crimes. It will argue that effective transitional justice must be responsive to the specific process by which the crime of genocide was perpetrated.

After genocide II   237

Prosecuting genocide Genocide, though a grievous crime, is rarely prosecuted. The paucity of prosecutions is due to the relative rarity of the crime, the difficulty in proving genocidal intent, and the political impediments to prosecution, namely states’ reluctance to prosecute themselves or submit to international legal processes. Several challenges to prosecuting genocide merit further exploration. Legal ordering The law as an epistemological system has its own logic, and its own way of processing facts (legal ordering). Positivist truths are sought, yet these truths are constituted through legal epistemologies, particularly the labelling of crimes. The nature of legal processes means that certain types of results are produced in service of narrow goals – to establish the culpability of individuals for specific acts, within a given time period and territorial space. I will examine several of the characteristics of legal ordering here to illustrate juridical notions of perpetration. First, the law has an internal rationality. It is a vocational world and culture with norms of fairness, professionalism, and efficiency. This means that trials must be conducted before a neutral and objective arbiter, according to a set of previously known rules (the legality principle), in order to produce a result. Trials are often adversarial, pitting a prosecution against a defense. One could also argue that there is a performative and dramaturgical aspect of trials. They are directed at an audience, and in the case of genocide, they also seek to produce a sense of moral outrage, personified in the figure of the perpetrator. However, trial participants are expected to control their emotions in the name of judicial decorum. Physical movement is highly regulated in the courtroom, and violations of court etiquette are met with verbal reprimands, formal sanctions (finding contempt of court or turning off the microphone), and nonverbal actions (ignoring an unruly lawyer).1 The trial seeks to create an orderly space to adjudicate “wild behavior.”2 This internal rationality of trials also establishes boundaries of evidentiary inclusion and exclusion. Any information not produced according to the internal standards of the court, and not relevant to the specific task of individualizing responsibility, will be excluded. Second, courts are always bounded by their jurisdiction, whether temporal, geographical, or subject-­matter (crime). In domestic courts, the geographical space is clear (the state or one of its sub-­units), but international courts are only able to investigate crimes and attribute criminal responsibility to territories and individuals over which they have jurisdiction. Courts also exercise jurisdiction only over certain well-­defined criminal acts. The different forms of jurisdiction function to classify, delineate, and exclude forms of violence. Genocide will only be investigated when it is within a court’s jurisdiction, and violent acts that fall outside this jurisdiction will be excluded from the judicial process. The law is embedded in the political practice through which its norms were created — in the case of genocide, by international treaty negotiations. Thus, the exclusion of

238   After genocide II other groups, such as political groups, from the Genocide Convention was a political decision, which came to be applied in the supposedly depoliticized sphere of the criminal trial. Third, courts strive to bring coherence and clarity. This is particularly difficult in the case of international crimes which often involve complex, interactive, dynamic, and even international processes. The search for coherence necessarily produces simplified narratives and interpretations of genocide that reduce uncertainty. Of course, there are elements of ambiguity in almost every judgment issued by a court – such as dissenting opinions and inconsistencies in legal reasoning – but this ambiguity is not held to be ideal. Transitional justice processes avoid complex truths, giving perpetrators an opportunity to deny their crimes.3 Perpetrators may think that the judicial perpetrator does not resemble them at all, and in a sense, they are correct: the one-­dimensionality of perpetrator representations distorts reality. Fourth, courts individualize responsibility, partially removing perpetration from its collective context. Genocide is necessarily a collective crime, facilitated by the social pressures generated by groups. Nonetheless, criminal courts are not truly designed to assess collective responsibility. They deal with individuals as discrete entities. The attribution of individual responsibility for collective acts is often achieved through modes of criminal liability that tie individuals to collective acts, such as joint criminal enterprise, co-­perpetration, command responsibility, and conspiracy. Courts also often look to structures of control, such as hierarchies and chains of command, to impute responsibility to individuals. The attribution of individual criminal responsibility means that courts construct binary and mutually-­exclusive categories (those who are culpable as perpetrators, and those who are not). Such classifications reduce the complexity and fluidity of individual motivation. The simultaneity principle dictates that the perpetrator need only have possessed intent at the moment of perpetration; yet in practice it is difficult for courts to establish intention without patterned behavior. Another assumption is at play here as well: that of planning and traceable intent and decisions. Every act is connected to a broader plan, rather than serendipitous or spontaneous. Taner Akcam writes: There is also an implicit teleology [in determinations of genocidal intent] … which takes the last point of destruction (Auschwitz, Der Zor) as a starting point and looks backward to violent expressions in the perpetrators’ early speeches and writings, treating them as a “serious declaration of intent.” 4 Every act of violence is seen as intentional, connected, and coherent. Thus, through judicial ordering, we are presented with an image of genocide as centrally planned and implemented by deliberate and determined perpetrators. The problem of genocidal intent Judicial ordering also shapes how the law addresses perpetrator intent in genocide. Unlike most international crimes, excepting the crimes against humanity of

After genocide II   239 persecution and apartheid, genocide has a two-­fold intent requirement. First, the perpetrator must possess a general intent to commit one of the underlying acts of genocide (e.g., killing, the prevention of births, the forcible transfer of children). Second, the perpetrator must evince a special intent to destroy an ethnic, racial, national, or religious group. This special intent element has traditionally been interpreted as requiring that perpetrators of genocide purposefully intend (and attempt) to destroy an ethnic, racial, national, or religious group in whole or in part. The rationale behind this purpose-­based interpretive lens is clear: the desire of the perpetrator to destroy the group is what distinguishes genocide from other international crimes, which may include the same underlying acts, but with a different criminal purpose. This purpose is reflected in United Nations General Assembly Resolution 96 (I), “The Crime of Genocide,” which argues that “Genocide is a denial of the right of existence of entire human groups, as homicide is the denial of the right to live of individual human beings.”5 It is this collective-­homicide aspect of genocide that best defines it. The central difficulty with genocidal intent has always centered on proof. Without a perpetrator’s confession of guilt, courts are reliant upon deriving inferences from the context of perpetration. For example, in the ICTR’s Akayesu case, the Chamber argued: intent is a mental factor which is difficult, even impossible, to determine. This is the reason why, in the absence of a confession from the Accused, his intent can be inferred from a certain number of presumptions of fact. The Chamber considers that it is possible to deduce the genocidal intent inherent in a particular act charged from the general context of the perpetration of other culpable acts systematically directed against that same group, whether these acts were committed by the same offender or by others.6 Courts have developed analytical frameworks to determine genocidal intent. The most elaborate was the one used in the Milorad Trbić case at the War Crimes Chamber of the Court of Bosnia. Trbić was a reserve captain in the Army of Republika Srpska (VRS or Bosnian Serb Army) who was involved in the Srebrenica massacre. In its judgment, the Trial Chamber upheld the view that genocidal intent may be inferred from the totality of circumstances, including the ideology that gave rise to genocidal acts.7 The Chamber set out an elaborate test for determinations of genocidal intent, expanding on the reasoning used in Stupar, et al., “Kravica” (Court of Bosnia and Herzegovina).8 The contextual test proposed by the Chamber was also derived from the findings in Akayesu (ICTR), Kayishema and Ruzindana (ICTR), Musema (ICTR), and Jelisić (ICTY).9 The test includes the following factors: (1) The general context of events in which the perpetrator acted including any plan to commit the crime; (2) The perpetrator’s knowledge of that plan; and

240   After genocide II (3) The specific nature of the perpetrator’s acts, including the following: (a) (b) (c) (d) (e) (f ) (g) (h) 4

no acts to the contrary for genocidal intent; single mindedness of purpose; efforts to overcome resistance of victims; efforts to overcome the resistance of other perpetrators; efforts to bar escape of victims; persecutory cruelty to victims; ongoing participation within the act itself; repetition of destructive acts, i.e., more than one act or site;

The acts themselves (The “Kravica Test”): (a) (b) (c) (d) (e) (f  ) (g) (h)

the number of victims; the use of derogatory language toward members of the targeted group; the systematic and methodical manner of killing; the weapons employed and the extent of bodily injury; the methodical way of planning; the targeting of victims regardless of age; the targeting of survivors; and the manner and character of the perpetrator’s participation.10

The elements in this test are cumulative, rather than requisite; thus, the more elements are present, the greater the evidence of genocidal intent. The Chamber found that Trbić was “calm,” “remorseless,” and “compliant” with regard to his role in the Srebrenica genocide.11 Moreover, he actively and continuously participated in the genocidal campaign in Srebrenica, even exercising initiative to ensure its success.12 Trbić killed several prisoners himself in each detention location, to overcome the resistance of remaining victims as well as that of fellow perpetrators (by providing a demonstrative example).13 He also barred victims’ escape by blindfolding them and binding their hands.14 Trbić demonstrated persecutory cruelty towards the victims by detaining them in deplorable conditions prior to their executions and “leaving them to contemplate their fate.”15 He also used derogatory language demonstrating his “indifference to the prisoners.”16 The killing was well-­planned and methodical, with detention places set out and burial equipment prepared in advance.17 In addition, the survivors of the initial killings were targeted.18 The Trbić test involves several normative and empirical assumptions about the nature of genocidal perpetrators: they are determined, deliberate, unwavering, and motivated by hatred towards the victim group. For example, the test cites important indicators such as “single mindedness of purpose,” “systematic and methodical manner of killing,” “ongoing participation,” “no contrary acts,” and the “use of derogatory language.” These factors seem to be deployed uncritically, with little empirical evidence or reference to the substantial social-­scientific, historical, and psychological literature on perpetrators. In short, it is a juridical construction of intent that fails to account fully for the nature of individual genocidal intent, embedded as it is in social and institutional structures.

After genocide II   241 Genocide is not possible without coordination. Its perpetration has an overarching aim or “collective intent,” and an overarching justification or “collective motive.” This collective motive is represented ideologically, though it is not shared by all perpetrators. We can distinguish between merely aggregated individual intent, in which perpetrators act in parallel (coincidental action), and collective intent, in which perpetrators knowingly work towards the same goal as part of an organized campaign (concerted action). Legal philosopher Larry May argues that collective intent may be shown “if there has been communication or interaction among these individuals, and the communication or interaction has influenced these individuals to come to hold similar intentions.”19 In genocide, communication and interaction are facilitated by institutional structures and ideologies. In a sense, individual action is structured through state power; indeed, genocide itself is institutionalized in the moral framework of criminogenic, genocidal states. In such a context, as we have seen, committing genocide becomes a normative act. Where it is committed by authority figures, or permitted by authorities, it is evident that such acts are accepted or even expected. Several other important aspects are relevant for a criminological understanding of genocidal intent. Both motive and intent are variable and fluctuate over time. In a given instance, perpetrators may desire to inflict genocide, while in the next they may feel remorse for their wrongful act. Even with the restructuring of morality, perpetrators may experience doubt and behave inconsistently. Thus, it is rare for perpetrators to exhibit “single mindedness of purpose.” Perhaps there are perpetrators who exhibit such singlemindedness, but they are likely the exception rather than the rule. The Trbić Test might consider perpetrator acts of rescue as “acts contrary” to genocidal intent. However, it is not at all contradictory that genocidal intent may exist alongside the intention to rescue victims, in circumstances where the abstraction of victims fall away and they become personalized once more.20 It is also not unusual for perpetrators to behave inconsistently: Erna Petri, mentioned earlier in this book, fed several starving Jewish children before murdering them. Similarly, expressing remorse in no way demonstrates that a perpetrator did not intend to commit genocide. Rather, such remorse may be a means for perpetrators to reconcile their acts with a restored moral context, in which genocidal violence is once again considered to be deviant. Remorse is also often considered a mitigating factor in sentencing for genocide and other international crimes. However, there is little real consideration in international criminal trials of the veracity and sincerity of remorse, or even what it consists of. In cases such as Stevan Todorović, a former Bosnian police chief, the guilty plea statement explicitly indicates remorse, even while Todorović emphasizes his lack of agency, externalizing responsibility and using the passive voice: “I became aware that Croats and Muslims had suffered a great deal.”21 Remorse may also be a manifestation of conflicted intent. Yet even the acts of conflicted perpetrators are essential to the actualization of genocidal destruction. One perpetrator remembers:

242   After genocide II All the time I knew that what I was doing, the group I was moving with, was totally wrong. But I did it to serve my family.… Instead of me losing a brother or losing some family members, let me go and save my family … every evening I could come back, and sit and see that I am moving in a wrong group, what we are doing is wrong, that I never thought on the same line as the people I was moving with.22 The key understanding, then, is that individual action may not be persistent, even while the overall campaign is sustained. Even killing for highly personalized reasons – such as self-­enrichment, a “personal desire for vengeance,” or sadism23 – may further the cause of genocide. It is also questionable whether genocide requires discriminatory elements such as “persecutory cruelty” or the “use of derogatory language towards members of the targeted group,” as set out in the Trbić Test. Indeed, conformist perpetrators may knowingly act to advance a genocidal campaign, while maintaining neutral views towards the victim group. In such cases, they may be motivated more by conformity or personal enrichment. The use of racial epithets per se is also not indicative of genocidal intent.24 Such epithets and dehumanizing language are commonly brought to bear in non-­genocidal conflicts. The elements of the Trbić Test citing “efforts to overcome the resistance of victims,” “efforts to overcome the resistance of other perpetrators,” and “efforts to bar the escape of victims” are certainly characteristic of vanguard perpetrators. However, such individuals are not representative of all perpetrators of genocide. While one may develop a gradient of moral responsibility based on the willingness of individual perpetrators, even relatively unenthusiastic perpetrators do kill, and, the murders they commit contribute significantly to the genocidal enterprise. The Trbić test demonstrates the need to move away from purpose-­based approaches to genocidal intent towards knowledge-­based approaches that account for typical patterns of perpetrator participation. It is unrealistic to expect that all (or even most) perpetrators of genocide have the intent to destroy, given the complexity, variability, and multiplicity of motive and intent. Rather, responsibility for genocide should encompass those who knowingly act in furtherance of genocidal campaigns, irrespective of their personal motives or purpose. To elaborate further, knowledge of the genocidal plan/genocidal context is dependent on the nature of the genocide itself. Pertinent questions include: does the genocide involve mass participation? And is an underlying genocidal ideology reflected in public discourse? One must also consider the role of the perpetrator: Is the perpetrator aware of the commission of genocidal acts, such as a police officer taking part in deportations, or a train driver transporting prisoners or supplies to concentration camps? Knowledge must be contextual and prima facie. Legal interpretations of genocidal intent often depict perpetrators as extraordinary actors, whose hateful motivations lie, at least partly, outside of ordinary human socialization and motivation. For example, the Jelisić Appeals Chamber found that it is “the borderline unbalanced personality who is more likely to be

After genocide II   243 drawn to extreme racial and ethnical hatred than the more balanced modulated individual without personality defects.”25 The purpose-­based approach to genocidal intent, and the image of perpetrators constructed by courts, flows directly from legal ordering. In seeking individual criminal responsibility and coherent narratives, courts greatly simplify the complexity and contingency of individual participation in genocide. This sim­plified view of perpetrator motivation is produced to meet the needs of legal ordering, which essentially eliminates the variability of individual intent and ascribes collective intent to each individual perpetrator. The attribution of responsibility In prosecuting genocide, we locate the responsibility for a collective crime in the individual perpetrator. One might ask whether it makes sense that a onetime criminogenic state should prosecute crimes that it previously encouraged. Does this shift the blame from the state to individuals, and away from collective complicity? In a sense, it does. Widespread responsibility may be obscured by the prosecution of a few token perpetrators who become symbols of all wrongdoing, thus exculpating others who have committed similar harms. Perhaps, then, beyond individuals, it is also important to “prosecute” the state (under the law of state responsibility). Genocide, after all, represents a policy and pattern that can only be realized through state power. Individual crimes would not exist without state criminality. While we cannot exonerate individuals, mechanisms must be found to ensure that the state and collective dimensions of perpetration are taken into account. Individuals may, of course, be prosecuted as members of criminal organizations. The importance of groups in the commission of genocide is evident in the collective aspects of the prosecution of international crimes (e.g., the use of the joint criminal enterprise mode of liability and co-­perpetration), as well as the joinder of similar cases into organizational trials.26 For example, the International Criminal Tribunal for Rwanda (ICTR) has held two “military trials,” a “media trial,” and an “MRND trial,” while the post-­Nuremberg Control Council Number Ten trials there were Krupp, I. G. Farben, High Command, Einsatzgruppen, and RuSHA (Rasse- und Siedlungshauptamt- Schutzstaffel) cases. At the “Khmer Rouge Tribunal” in Cambodia there is a “government” case. The Nuremberg judgment went so far as to declare entire organizations as “criminal,” such as the Nazi party leadership, Gestapo, and SS. The mass character of crimes against humanity and genocide has also informed the Elements of Crimes of the ICC. This document declares that genocide must occur within a “pattern of similar attacks,” and that crimes against humanity must be pursuant to a “state or organizational policy.” Organizations can also be held responsible for harm through the mechanism of civic liability. Jennifer Balint argues that the attribution of such liability may occur in three stages: first, the designation of key members of institutions as individually criminally liable; second, the transformation of institutions by

244   After genocide II assigning responsibility for institutional actions; and, finally, the monitoring of these processes by internal and external bodies.27 Any preventive strategy for genocide must address the authority-­driven nature of the crime. Authorities normally responsible for the prevention and punishment of crime themselves perpetrate it during genocide. A distinction is often made between aiding and abetting (secondary responsibility for those present at the scene of the crime) and procuring and counselling (tertiary responsibility for those not present), yet both roles are essential for the commission of the mass crime of genocide.28 Drumbl argues that we can distinguish between “conflict entrepreneurs” (overall leaders of violence), “leaders” (who give orders, but are themselves subject to orders), and “killers” (who carry out orders).29 Overall leaders are self-­motivating, and also motivate others; thus they must be held most responsible for genocide. In response to the question, “Who is responsible for the genocide?” a former Khmer Rouge group leader argued: “Those upper leaders. They pressured people along the line, and if you disobeyed you would disappear at night.”30 The law must somehow address these complex moral contexts. Does it make sense to speak of individualizing guilt at all when genocide is obviously a collective endeavor? International criminal law presumes that individuals inherently recognize orders to commit genocide and crimes against humanity as manifestly unlawful. Yet not all individuals recognize the wrongfulness of their acts, particularly when acts of violence against certain groups are incrementally conventionalized and rationalized by legitimate authorities. Perhaps the key still lies in individual propensity and variability: we know that individuals respond to the moral context in a differential way. Not all individuals, when placed in similar conditions, will choose to perpetrate genocide. Yet, the “reasonable person” test still challenges our assumptions about criminal responsibility for genocide, particularly regarding individuals who are conformist perpetrators, rather than ideological or pathological ones. It would probably overstate the case to say that in genocide, the reasonable person might perpetrate rather than rescue. The idea of an overarching environment of coercion is too imprecise for criminal law to bear. Although it is common in genocide, coercion does not exist for all individuals in all circumstances. Of course, direct coercion (e.g., being forced to perpetrate at gunpoint) is still a legitimate defense in trials for genocide, even if superior orders are not. “I would be killed if I did not participate because it was the order of the government at the time,” argued one Rwandan perpetrator.31 We have posited that individual roles are variable and malleable. How, then, do these role shifts affect culpability? If you conspire to commit genocide, then voluntarily abandon your criminal purpose, this may be utilized as a defense. However, once you have perpetrated a single act of murder in a genocidal context, you have crossed a legal threshold and become a perpetrator of genocide. Acts of rescue after this initial act of perpetration may still be considered as a mitigating factor in sentencing.

After genocide II   245 The reproduction of perpetrator narratives A final challenge in prosecuting genocide must be raised here. The judicial process, with its adversarial exchange, may reproduce perpetrator narratives. Nobody enters the courtroom as a perpetrator, and those accused of genocide deserve a vigorous defense, to ensure both individual and systemic fairness. Ideally, alleged perpetrators are free to advance their own position, within the confines of judicial procedure, and this narrative will be repudiated by the court if it is shown to be false. But this ideal assumes that courts will always decide correctly, and that their judgments will outweigh other narratives advanced in the trial process. One need not look far to find examples of courts endorsing false perpetrator narratives. In the recent Šešelj acquittal at the ICTY, the ethnic cleansing of Bosniaks was found to be a strategy to protect civilians in a war zone.32 Courts may also provide a political platform for the accused to question the legitimacy of the justice system, as well as to advance a broader, denialist narrative. Consider, for example, the self-­representations in court of Slobodan Milošević, Radovan Karadžić, Ratko Mladić, and Charles Taylor. The techniques of neutralization are themselves sometimes reproduced by the defense during trials. This may be traumatizing to victims, who may feel that perpetrator accounts are being given too much voice. When Cambodian perpetrator Duch read a poem during his trial at the ECCC, the victims’ representative responded: “We are not here, sir, for cultural edification! […] I am talking about twelve thousand dead at S-­21!”33 Victims, for their part, may have very limited scope in criminal proceedings to advance their own perspective on the violence that they suffered.

Prosecution as prevention The promise of deterrence theory is that criminal sanctions do more than just punish serious deviant acts. They also dissuade potential future offenders by increasing the risk and decreasing the rewards of criminal behavior. Classical criminologist Césare Beccaria argued that for deterrence to be effective, swift and appropriate punishment must be certain.34 Severe punishment will not offer a strong deterrent effect if the criminal believes that impunity is a significant possibility. Unfortunately, in the international criminal-­law context, impunity is still pervasive, especially given the absence of a true supranational authority in the international system (possibly apart from the UN Security Council). Furthermore, as state crimes, perpetrators of genocide and crimes against humanity act at the behest of the state, and thus are protected by state power. Rwandan perpetrators recall: I did not expect to be in prison. I was arrested because the regime lost power. If the old regime had stayed in power my actions would not have been wrong.

246   After genocide II We did not feel guilty for the victims because we were told to kill the victims. The Burgomaster and the counsellors used to tell us, if we did not kill the Tutsi they would kill us. So, we had to do so. Yes, I remember some of them were my neighbors.35 Thus, the state-­directed nature of genocide may preclude any notion of deterrence. General deterrence (to deter all potential criminals in society) can be contrasted with specific deterrence, aimed at individual criminals. A central element of specific deterrence is incapacitation – making it impossible for criminals to commit crimes by imprisoning them. It is held that isolating certain individuals makes crime less likely, and society safer.36 In the case of the ICC, the mere issue of arrest warrants may have had a deterrent effect by restricting the movement of leaders likely to commit international crimes, although the Bashir case calls this into question. Specific deterrence may also include such measures as restrictions and conditions on convicted perpetrators after their release. General deterrence is predicated on three notions: that perpetrators are rational actors; that the rewards available to them can be measured, so that punishments can be devised to counteract this gain; and finally, that perpetrators understand and are aware of potential risks. Criminogenic societies are often closed systems that rationalize criminal acts that would be considered irrational in an “ordinary” context. Perpetrators of genocide are also likely less entrepreneurial than perpetrators of ordinary crimes. Many are serving a role within a larger, legitimized (state-­directed) joint criminal enterprise. Deterrence in the context of genocide is also problematic, because the perpetrators are usually sheltered by a state that itself is committing genocide. Furthermore, general deterrence theory does not account for individual differences in personality and motivation. Criminological research shows that sanctions do have the power to deter criminal conduct, but only under certain circumstances. They are effective where individuals are closely bonded to the broader social community, where they perceive the system as fair and legitimate, and where they feel shame at having committed a wrongful act.37 Thus, the effectiveness of deterrence varies depending on the individual and his or her relationship to the broader society. Unfortunately, many of these criteria are not met for perpetrators in genocidal contexts perhaps indicating that courts might have a weak deterrent effect. Many perpetrators interviewed for this study claim to have been unfairly convicted. This may be true in some cases, particularly when one considers the political instrumentality of many transitional justice processes, as well as the inconsistencies and incapacity of many judicial systems. In Rwanda, many of these claims center on interpersonal disputes or the perceived desire of Tutsis (or others) to “steal” land or other property from convicted Hutus. One convicted perpetrator alleges: “some people who did kill, they got jealous and they said, ‘why should we be in prison alone?’ ”38 Another recounts “in most cases those who were put to prison, it was because of the belongings.”39 In the case of another perpetrator he attributes his conviction to an interpersonal dispute:

After genocide II   247 I have never killed.… But in 1992 I had a fight with this guy from where I came from. So, they have been having grudges from 1992. So, when he saw me out of prison he said: “you have to go back in prison.” I went.…40 This sense of unfairness can be manifested as a backlash against the system. One Rwandan perpetrator argues: “Before the genocide I thought that Tutsis and Hutus were brothers but now, because of what happened to me, I think they’re totally different.”41 Sherman argues that if systems are deemed to be unfair on the collective level, such as where systemic discrimination or oppression exists, then defiance can also occur collectively. In other words, individuals may come to see a system as unworthy of their compliance, because they belong to an identifiable group confronting discrimination.42 The other elements that can generate defiance – weak bonding and an absence of shame – are also inherent in genocide. As previously argued, genocide is a crime of conformity committed pursuant to the political goals of criminogenic states. If perpetrators are not clearly aware that their behavior is wrongful, how can they be deterred? Perceptual deterrence theories tell us that the effectiveness of sanctions depends on their certainty and severity.43 There are, unfortunately, many examples of impunity for genocide. Of the Rwandan perpetrators interviewed, 80 percent did not fear being prosecuted at the time of killing, with one responding: “No, because killings had already taken place since the RPF invasion, and those who were responsible were put in jail and then released by the MRND.”44 This illustrates one of the central problems with deterrence – the certainty of many genocidal elites that their hold on power is secure, and thus they will never have to face international or domestic justice. In November 1945, Nuremberg Prosecutor Robert Jackson sagely observed that Wars are started only on the theory and in the confidence that they can be won. Personal punishment, to be suffered only in the event that the war is lost, will probably not be a sufficient deterrent to prevent a war where the war-­makers feel the chances of defeat are negligible.45 As one Rwandan génocidaire argued: “we didn’t think we would be prosecuted because we thought we would win the war.”46 In the context of genocide, the opportunity space for perpetration comes from the environment of impunity created by the state. For instance, a Rwandan perpetrator told me: “we killed him [the victim], because we thought that there wouldn’t be any consequences. The leaders were doing the same thing.… If one person who killed was prosecuted everyone else would have stopped killing.”47 With the collective action problems involved in resisting genocide, deterrence would have to occur in a widespread and simultaneous manner (collectively) to be effective. Moreover, in transgressive communities, with their strong internal norms and isolation from society, deterrence may be even more challenging.

248   After genocide II Beyond deterrence, the prosecution of genocide may have other preventive effects. These include: restoring the rule of law, incapacitating perpetrators, interdicting inchoate forms of criminality (preventing further crimes), and isolating offenders. Despite the challenges involved in prevention, most perpetrators are risk-­averse, and with increased prosecutions, the risk of punishment would increase, with a corresponding greater reluctance to offend.

Reconsidering transitional justice When we examine perpetration in this more rounded manner, several implications arise for transitional justice mechanisms. First, legal processes must take account of perpetration in its diverse forms. Perpetrators must be disaggregated according to the causes and contexts of their perpetration: compromised perpetrators differ significantly from ideologues in their motivation, and, possibly, their continuing fidelity to criminal behavior. Second, legal processes are not adequate to undertake the transformation of society that must occur after genocide. It is accepted wisdom that the prosecution of international crimes contributes to the prevention of such crimes. Few scholars question this supposition, and even fewer consider how this preventive effect might operate. Yet, any serious consideration of this assumption swiftly leads us to the limited power of the law as an instrument to resolve social problems. Martin Luther King, Jr. recognized these limitations, but argued: “it may be true that morality cannot be legislated, but behavior can be regulated. The law may not change the heart, but it can restrain the heartless.”48 This power to restrain the heartless, through deterrence, is hotly debated. Even if prosecution does not deter the commission of crimes, perhaps it has other effects. As the Polish delegate to the International Covenant on Civil and Political Rights (ICCPR) drafting conference said, the law also comprises “many aspects of an educative character.”49 Perhaps the law can, gradually, help to create new social norms. Most institutions of justice depersonalize the relationship between victim and perpetrator. The state “reconstitutes the suffering of the victim as its own through forensically reworking the suffering as a ritualized truth contest.”50 This legal recognition of victims’ suffering may also play an essential role in peacebuilding. The prosecution of genocide is just one element of a broader transitional justice system. Given their role in perpetration, it makes sense to prosecute high-­ level perpetrators (leaders), as well as leading low-­level perpetrators (vanguard perpetrators), to the full extent of the law. These perpetrators both kill and lead other perpetrators to kill, through a combination of orders, normative expectations, and the demonstrative effect of violence itself. For specific deterrence to prevail, these individuals must be removed from society. In contrast, low-­level perpetrators who do not take a lead in killing may be prosecuted, but with the aim of eventually rehabilitating and reintegrating them in society. Merely serving a prison sentence does not suffice for rehabilitation. Instead, efforts must be made to provide these individuals with opportunities to

After genocide II   249 change their self-­identity and view of past violence. This may also involve restorative justice measures and reparations for victims. Beyond the prosecution of individuals, the criminogenic state must also be dismantled and brought to account for its collective responsibility. This may occur via assistance and reparations to victims, official apologies, revision of official histories (particularly in the education system), the amelioration of harm-­producing relations of power, and the establishment of bodies to ascertain the truth about past violence, especially the fate of individual victims and their remains. Individual and collective reparations may also be attractive transitional justice strategies, considering that they require perpetrators (and responsible institutions) to compensate their victims. Unfortunately, this is not always a workable option. In Rwanda, the courts have awarded millions of dollars in compensation to victims, but they are seldom able to collect, as many perpetrators are indigent. The Rwandan government has wisely also “sought to strike a balance between the survivor’s need for restitution and the perpetrator’s ability to pay” by capping and excluding certain aspects of compensation (such as the house inhabited by the perpetrator and her family).51 Finally, more attention must be devoted to the rehabilitation and reintegration of perpetrators, most of whom will eventually return to their home communities. Such individuals face a host of difficult adjustments, including overcoming shame and fear, and dealing with economic challenges.52 According to Miroslav Kvocka, a former guard and ICTY convictee, “you have a reputation as a war criminal, the age [old age], and you do not have some extra occupation. My occupation is police officer and where can I seek for a job?”53 Feelings of shame and guilt are both also intimately connected to the transformed social context that perpetrators encounter.

Conclusion: simple answers for complex questions? For the courts, genocide remains a conundrum, individualizing guilt for collective acts, attributing motive and intent to confused and inconsistent perpetrators. In seeking clarity, judicial processes reduce the complexity of perpetration, and instead create a perpetrator fiction to suit judicial requirements. As Mégret notes, “tribunals are the tools of society’s condemnation.”54 This condemnation and stigmatization is effectively the purpose of international tribunals – only by condemning perpetrators can tribunals fulfill their purpose. Mégret (drawing from Durkheim) argues that “stigma is both an expression of the inherent moral blameworthiness assigned to the perpetrators of certain heinous acts and a way of constituting the society that assigns this blameworthiness.”55 The international community, then, is also a moral community. Those who violate the moral order are symbols of the worthiness and resilience of that order. Those with the “intent to destroy” are surely among the most insidious criminals: The authors of a genocide would be pursued not only because they have killed hundreds of thousands of people, but, even more so, because they have violated the moral order of the human race by attempting to destroy one of its member parts.56

250   After genocide II Genocide is a crime that cannot be undone. If the victim group is substantially destroyed, it may not be revived. The victims themselves are gone, and the survivors and perpetrators will never be the same. All things considered, there are no one-­size-fits-­all solutions to prosecution. Rather, authorities should consider the pattern of perpetration and the needs of society (including victims) in designing an appropriate transitional justice strategy. It may be the case that the cruelty of the act merits punishment, even where the banality of the perpetrator seems to indicate otherwise.

Notes   1 Hinton, Man or Monster?, 230 and 234.   2 Hinton, 234.   3 Davies, Rationalising the Irrational, 14.   4 Akcam, The Young Turk’s Crimes, 137.   5 United Nations General Assembly Resolution 96 (I), “The Crime of Genocide.”   6 Prosecutor v. Akayesu (TC), para. 523.   7 Prosecutor v. Milorad Trbic, paras 197–199.   8 Prosecutor v. Milos Stupar et al.   9 Prosecutor v. Akayesu, para. 523; Prosecutor v. Kayishema and Ruzind, para. 527; See also: Prosecutor v. Musema, para. 166; ProsecutorJelisic, para. 47. 10 Prosecutor v. Trbic, para. 202. 11 Prosecutor v. Trbic, paras 807, 825. 12 Prosecutor v. Trbic, para. 810. 13 Prosecutor v. Trbic, paras 812–813. 14 Prosecutor v. Trbic, para. 814. 15 Prosecutor v. Trbic, para. 816. 16 Prosecutor v. Trbic, para. 819. 17 Prosecutor v. Trbic, para. 822. 18 Prosecutor v. Trbic, para. 824. 19 May, Genocide: A Normative Account, 119. 20 Glover, Humanity: A Moral History, 37–38. 21 Prosecutor v. Todorović, paras 89–92. 22 Interview R04 (Rwanda), Kigali Central Prison, July 2009. 23 Klee, Dressen, and Riess, The Good Old Days, 60. 24 Research done by criminologists John Hagan, Wenona Rymond-­Richmond and Patricia Parker relied partly on the use of racial epithets as evidence of genocidal intent. See: Hagan, Rymond-­Richmond, and Parker, Darfur and the Crime of Genocide. 25 Judgment, Jelisic (IT-­95–10-A), Appeals Chamber, July 5 2001, x 70. 26 Balint, “Dealing with International Crimes,” 318. 27 Balint, 169. 28 May, 161. 29 Drumbl, Atrocity, Punishment, and International Law, 25. 30 Interview C04 (Cambodia), Kampong Chhnang Province, October 2008. 31 Interview R34 (Rwanda), Rwaza TIG Camp, August 2009. 32 Trial Judgment Summary for Vojislav Šešelj, 5–6. 33 Cruvellier, The Master of Confessions, 314. 34 Lanier and Henry, Essential Criminology, 76. 35 Interview R09 (Rwanda), Kigali Central Prison, July 2009; Interview R58 (Rwanda), Rugerero, September 2009; Interview R67 (Rwanda), Hindiro TIG Camp, October 2009. 36 Lanier and Henry, 86. 37 Sherman, “Defiance, Deterrence, and Irrelevance,” 448.

After genocide II   251 38 Interview R07 (Rwanda), Remera Prison, July 2009. 39 Interview R05 (Rwanda), Kigali Central Prison, July 2009. 40 Interview R06 (Rwanda), Remera Prison, July 2009. 41 Interview R20 (Rwanda), Ruhengeri Prison, August 2009. 42 Sherman, 467. 43 Nagin, “Criminal Deterrence Research,” 7. 44 Interview R01 (Rwanda), Kigali Central Prison, July 2009. 45 Mennecke, “Punishing Génocidaires,” 325. 46 Interview R62 (Rwanda), Mont Kigali TIG Camp, October 2009. 47 Interview R37 (Rwanda), Rwaza TIG Camp, August 2009. 48 Farrior, “Molding the Matrix,” 56. 49 Farrior, 57. 50 Downes, et al., Crime, Social Control and Human Rights, 425. 51 Waldorf, “Goats & Graves,” 532. 52 Tertsakian, Le Chateau, 455. 53 TV Justice Magazine, “Episode 26: life of the Convicted War Criminal.” 54 Mégret, “Practices of Stigmatization,” 289. 55 Mégret, 290. 56 Destexhe, Rwanda and Genocide, 64.

Works cited Akcam, Taner. The Young Turks’ Crimes against Humanity. Princeton: Princeton University Press, 2013. Balint, Jennifer. “Dealing with International Crimes: Towards a Conceptual Mode of Accountability and Justice.” In Supranational Criminology: Towards a Criminology of International Crimes, edited by Alette Smeulers and Roelof Haveman. Antwerp: Intersentia, 2008. Cruvellier, Thierry. The Master of Confessions: The Making of Khmer Rouge Torturer. New York: HarperCollins, 2011. Davies, Jessica. Rationalising the Irrational: How Perpetrators Comprehend the Phenomenon of Genocide. (MA thesis). Amsterdam: University of Amsterdam, 2015. Destexhe, Alain. Rwanda and Genocide in the Twentieth Century. New York: New York University Press, 1995. Downes, David, Paul Rock, Christine Chinkin, and Conor Gearty (eds.). Crime, Social Control and Human Rights From moral panics to states of denial. Essays in honour of Stanley Cohen. Portland, USA: Willan Publishing, 2007. Drumbl, Mark. Atrocity, Punishment, and International Law. Cambridge: Cambridge University Press, 2007. Farrior, Stephanie. “Molding the Matrix: The Historical and Theoretical Foundations of International Law Concerning Hate Speech.” Berkeley Journal of International Law 14, no. 1 (1996). Glover, Jonathan. Humanity: A Moral History of the 20th Century. New Haven: Yale, 2001. Hagan, John, Wenona Rymond-­Richmond, and Patricia Parker. Darfur and the Crime of Genocide New York: Cambridge University Press, 2009. Hinton, Alexander Laban. Man or Monster? The Trial of a Khmer Rouge Torturer. London: Duke University Press, 2016. Klee, Ernst, Willi Dressen, and Volker Riess (eds.), The Good Old Days: The Holocaust as Seen By its Perpetrators and Bystanders. Old Saybrook: Konecky and Konecky, 1991.

252   After genocide II Lanier, Mark and Stuart Henry. Essential Criminology. Boulder, USA: Westview Press, 2004. May, Larry. Genocide: A Normative Account. New York: Cambridge University Press, 2010. Mégret, Frédéric. “Practices of Stigmatization.” Law and Contemporary Problems 76, nos. 3–4 (2013). Mennecke, Martin. “Punishing Génocidaires: A Deterrent Effect or Not?” Human Rights Review 8, no. 4 (July 2007). Nagin, Daniel S. “Criminal Deterrence Research at the Outset of the Twentieth Century.” Crime and Justice 23 (1998). Prosecutor v. Jean-­Paul Akayesu, Judgment, Case No. ICTR-­96–4-T, September 2 1998 (“Akayesu Judgment (TC)”). Prosecutor v. Goran Jelisic (Case No. IT-­95–10), Trial Chamber Judgment, December 14 1999. Prosecutor v. Clement Kayishema, and Obed Ruzindana (Case No. ICTR-­95–1-T), Trial Chamber Judgment, May 21 1999. Prosecutor v. Alfred Musema (Case No. ICTR-­96–13-T). Trial Chamber Judgement. January 27 2000. Prosecutor v. Milos Stupar et al. (Case no. X-­KR-O5/24). First Instance Verdict. July 29 2008. Prosecutor v. Stevan Todorović (Case No. IT-­95–9/1-S), Sentencing Judgement, July 31 2001. Prosecutor v. Milorad Trbic (Case no. X-­KR-07/386). First Instance Verdict. October 16 2009. Sherman, Lawrence W. “Defiance, Deterrence, and Irrelevance: A Theory of the Criminal Sanction.” Journal of Research in Crime and Delinquency, 30, no. 4 (1993). Tertsakian, Carina. Le Chateau: The Lives of Prisoners in Rwanda. London: Arves Books, 2008. Trial Judgement Summary for Prosecutor v. Vojislav Šešelj. International Criminal Tribunal for the Former Yugoslavia. March 31 2016. TV Justice Magazine. “Episode 26: Life of the Convicted War Criminals.” March 16 2012. United Nations General Assembly Resolution 96 (I). “The Crime of Genocide.” December 11 1946. Waldorf, Lars. “Goats and Graves: Reparations in Rwanda’s Community Courts.” In Reparations for Victims of Genocide, War Crimes and Crimes against Humanity, edited by Carla Ferstman, Mariana Goetz, Alan Stephens. Leiden: Brill, 2009.

12 Conclusion Killing without consequence?

Comprehending genocide Genocide can be understood. If we are to conceive of genocide as evil, and there are surely strong reasons for doing so, then we must consider it a human evil, rather than grounding our discussion in the language of mysticism. Genocide is largely wrought by ordinary men and women. This is apparent to most people who have researched genocide, yet still this proposition remains deeply troubling. If we condemn the act of genocide, and condemn the perpetrators, are we also condemning ourselves? What, if anything, separates us from them? Hannah Arendt famously spoke of the banality of evil; perhaps it is more useful to speak of the humanity of evil. Evil is not banal, it is extraordinarily cruel for the victims, and difficult for many perpetrators, but it is also unmistakably human. If the prosecutor in the Eichmann trial was correct to argue that Eichmann “was born human but he lived like a beast in the jungle,” we can ask: are such perpetrators beasts in the shape of men or men in the shape of beasts? In other words, is evil located in their selves, or in their actions? The evidence leads us toward a single conclusion, awesome in its terrible possibility: the potential for killing resides within each of us. The key to this potential rests in the social processes that shape perpetration. In studying these processes through perpetrators’ framing of their moral universe, we come closer to understanding genocide. Genocide is at once universal (grounded in human potentialities) and contingent on context. However, these contexts are never fully determinative. Nor do they exist independently of human volition, which simultaneously produces and responds to its environment. Evil is something you wouldn’t imagine doing; and yet evil is done.1 We all make concessions to power. This book has argued that perpetration is a differential response to the moral context. The advantage of this approach is that it accounts for acts of perpetration beyond those directly ordered by the state or other authority figures. Perpetrators interpret the context in a way that reflects their own interests. Foremost among these is aligning their behavior with their environment, particularly the expectations of authority and their peers. Non-­conforming individuals risk ostracism and expulsion from the community. Alienation from the community places

254   Conclusion: killing without consequence? individuals in a situation of profound vulnerability; the identification and obliteration of traitors occurs alongside the destruction of the victim group. Propensity, then, is a product of moral disengagement and proximity to other perpetrators. The moral context is formed through a combination of harm-­producing discourses and moral habits – the actual practices of state power. These moral rules, moral habits, and normative expectations are produced gradually over time. The reconstitution of the state in a manner that produces social distance and harm to the victim group usually reflects internal political processes in response to perceived strain. The emergence of the genocidal context, and of genocide itself, occurs through moral ruptures and continuities. Ruptures are crucial in producing the disequilibrium which enables the revision of moral rules and the concentration of power, as well as communicating information about the relationship between the political/moral community and the victim group. However, continuity is equally crucial to producing genocide; it is continuity, in the form of links to the past moral and institutional order, which enables the community to accept the incremental revision of moral rules as a means of countering existential threat. In both the pre-­genocide and genocide periods, a conventionalization of killing takes place. Killing becomes an extension of the normal exercise of state power; the state monopoly of violence is directed against specific groups, allegedly to maintain order and defend the community against internal and external enemies. The state generates moral disengagement by institutionalizing harm and discourses produce social distance. Through the conventionalization of killing, norms shift and individuals and institutions adjust to the new moral order. Yet this “new normal” is not universally accepted or understood. For some perpetrators, genocide represents a time of moral conflict, while for others it may present an opportunity. The inherent disruption creates an opportunity for the exercise of violence in service of personal ends. These personal ends may include the pleasure some individuals take in inflicting violence and exercising power over other human beings, as well as the fulfillment of career ambitions, the overturning of local elites, and the settling of personal scores. For those perpetrators who experience moral conflict, mitigation can be found in moral neutralization. The techniques of neutralization, as disseminated by propaganda, cultural beliefs, and regulated norms of behavior (including the tendency to minimize our own moral culpability), allow individuals to conform without threatening their cognitive integrity. Cognitive dissonance results from a clash between actions and beliefs, but the techniques of neutralization allow individuals to restructure their beliefs and neutralize their dissonance. The broad comparative scope of this book illustrates the universal presence of these justifications and their importance in structuring perpetrator action. Perpetrators disclaim responsibility by citing the ideological and normative context of perpetration. Furthermore, perpetrators mitigate the moral (and legal) magnitude of their crimes by externalizing their agency. They seek to maintain their self-­ image by depicting their actions as justifiable, unavoidable, and nobler than those of other individuals placed in the same situation.

Conclusion: killing without consequence?   255 Power structures greatly facilitate this process, which seeks to stave off both external and internal judgment. The state demands obedience, and most individuals will comply, whatever the moral cost. State structures also place individuals in a situation of omnipotence over their disempowered victims. In such circumstances, a sadistic shift may occur in individuals who have not previously exhibited such behavior. This sadism is often forward-­looking, exceeding the immediate demands of genocidal leaders and organizations. Yet the state itself functions as a criminogenic system: genocide becomes normative, and the state seeks to produce the crime of genocide. This criminogenic state inculcates a sort of genocidal organizational culture personified in specialized killing organizations. The nature of genocide as a collective crime, often perpetrated in the presence of supportive (or passive) peers, also eases the killing process. Not every individual responds in the same way to a given context. The range of action alternatives varies depending on individual’s social conditions and role, their personal orientation towards genocidal violence, and their proximity to other perpetrators (including state agents). At the micro level of analysis, the similarities between cases of genocide may be the strongest, even while macro level differences exist in the unfolding of genocide. In criminogenic contexts, individual decision-­making becomes distorted. The genocidal state substantially reduces the costs of participation in genocide, and increases the costs of non-­participation. Moreover, individual participation in killing is rewarded, and once the taboo has been broken, the marginal costs of recidivism decrease substantially for many perpetrators. In other words, killing becomes ever easier. Normative shifts influence both the perceived range of options and their potential consequences. Individuals become perpetrators when they reach their tipping point – as defined by their propensity, margins of discretion, and interpretation of the context, as well as the expressed preferences of others. The actual mobilization and coordination of violence occurs through an interactive and dynamic process. Local and individual interests must, to some degree, align with central power and objectives. Therefore, the moral context arises from an interaction between the state and other actors; yet genocide can only occur in the shadow of state power. Many low-­level perpetrators seem to be motivated more by alignment with group norms than by profound hatred of the other. Structures of obedience (such as the modern state itself ) shape individual behavior. In many cases, obedience and conformity may suffice to ensure compliance with killing demands. However, the denigration of the victim is also important in transforming ordinary people into killers. This is where the contextual nature of genocide is crucial. Genocidal messages will not be equally persuasive and comprehensible in all contexts. Rather, they must draw upon existing cultural understandings and experiences. The act of violence has individual and collective effects. In both cases, perpetrator identity will shift to interpret violence as justified and inevitable. These shifts in individual and collective identity must also be addressed in the post-­ genocide period, together with the broader processes that produced genocide.

256   Conclusion: killing without consequence? Genocide, then, can be best understood as a product of context and individual propensity, mediated by moral rules, particularly those that include groups in the moral community or exclude them from it. The rest of this chapter will draw from the findings of this book to consider broader questions: how do normal moral frameworks differ from perpetrators’, and what can our encounters with perpetrators can tell us about the human experience and the prevention of genocide?

The globalization of conscience? There is no love between men.

(Rwandan génocidaire2)

In this book, we have identified moral disengagement, the shrinking of the moral community of perpetrators, as central to genocide. But how expansive is our own moral community? Identity is fundamentally a social phenomenon – a matter of human imagination. This sense of identity is both egoistic and relational. It is egoistic in that it seeks to locate itself in the world, and relative to others who hold the same identity and different ones. Identity, then, is also complex and fluid. Yet there can be something exclusionary about asserting an identity. Doing so can introduce notions of moral stratification and relative worth. Humankind divides itself into pseudo species (groups), each of which holds itself to be the center of its own universe of conscience and valuation. All other groups are not only outsiders but inhuman.3 This triumph of socially-­constructed speciation over genetic speciation is uniquely human, and exceptionally destructive. We can conceive of identity as primarily vested in one of three “places”: the individual, the group, and the universe.4 Similarly our ethics can be individual, group-­centered, or universal. In genocide, both identity and ethics become group-­centered. Individuals transform themselves from individuals into components of a group, and extend their ethical obligations only to other assumed group members. As Michael Ignatieff has said, “ethics follows ethnicity.”5 The term “conscience” has its etymological roots in the Latin con (with) and scientia (knowledge). In other words, conscience represents a sort of self-­ awareness or consciousness. With the emergence of modern English and German in the sixteenth century, “conscience” ceased to be used interchangeably with “consciousness,” and came to mean a sort of inner virtue that could and should guide our actions.6 This idea of conscience is represented in the preamble of the UN Charter, which states that all human beings “are endowed with reason and conscience, and should towards one another in a spirit of brotherhood.” There are two aspects to this notion of conscience: the “inner voice” that tells us what we shall or shall not do, and to whom we shall and shall not do it.7 In this sense, our conscience is, like our identity, both egoistic and relational. Our perceptions, as well as our actions, shape our moral identity – our image of the person we want to be.8 The character of Glaucon in Plato’s Republic

Conclusion: killing without consequence?   257 argues that people’s self-­interest is best served by seeming moral; their actual morality is irrelevant.9 Can this theory also be extended to bystander states and individual perpetrators? Do they benefit from seeming moral (moral posturing)? This might offer a partial explanation for both the empty promises of the international community in cases like Srebrenica, and the rationalizations of perpetrators. Even universal moral principles such as “thou shalt not kill” are rendered fluid by the human imagination, subject to reinterpretation and dilution. Propaganda drives this process, as does our tendency to reshape our world in deference to the maintenance of our positive self-­image. Laws can also have a normative power in sanctioning actions that might normally violate an individual’s moral identity. However, not all moral principles are equally fluid. A distinction may be drawn between acts that are mala in se (bad in and of themselves) and mala prohibita (bad because they are prohibited). Journalist Jeffrey Kluger uses the example of a schoolchild who knows not to eat in class because the teacher has said it is wrong (mala prohibita). In contrast, that same child will likely hesitate when the teacher tells them to push another child off a chair (mala in se).10 This aversion to harming other people may be innate. But even if morality is consistent, moral behavior is not. Socialization is needed to shape our decisions. Genocidal killing is, of course, a mala in se deviant act and a violation of universal moral norms. Genocide and other mass hate crimes are important because they attack the very humanity of people and peoples. This attack on humanity requires a human response, a response of humanity as a collective. In spite of this compelling logic, our moral outrage when faced with such atrocities is often muted or, at least, not transformed into meaningful attempts to enact solutions. Why is this so? We all belong to moral communities. We define some people as being “us,” while the remainder of humankind are labelled as “them” or “others.” In constructing this dichotomy, those deemed alien are disassociated from ourselves. This disassociation also contains aspects of devaluation. How can we reconcile the worth of “others” with our egoistic view of the world? Roméo Dallaire (the commander of the UN mission in Rwanda during the genocide) queried: “Are all humans human? Or are some more human than others?”11 Carl Schmitt, a political theorist favored by the Nazis, might have responded to Dallaire (as he did to others) that “not every being with a human face is human.”12 Our moral community encompasses those who are close to us and those within our sphere of concern. This moral community entails reciprocity and mutual accountability.13 “The universe of moral obligation, far from being universal, is bounded by community.”14 History is replete with atrocities committed against those groups that have been placed outside of the moral community and devalued. Greek moral philosophy excluded barbarians and slaves; colonial ideologies excluded peoples deemed to be “primitive” and less civilized. Nazism extended reciprocity only to those within the same imagined racial community (so-­called Aryans). Relativism, sometimes employed in the service of tolerance, can also be used to dictate that moral laws are relative, that moral obligations are determined by community membership. This may be equally true for perpetrators

258   Conclusion: killing without consequence? and for the rest of us, although for perpetrators, “others” are subject not only to indifference but to death. In Civilization and its Discontents, Sigmund Freud wrote: If I love someone he must deserve it in some way.… He deserves it if he is so like me in important ways that I can love myself in him; and he deserves it if he is so much more perfect than myself that I can love my idea of my own self in him. Again, I have to love him if he is my friend’s son, since the pain my friend would feel if any harm came to him would be my pain too – I should have to share it. But if he is a stranger to me and if he cannot attract me by any worth of his own or any significance that he may have acquired for my emotional life, it will be hard for me to love him. Indeed, I should be wrong to do so, for my love is valued by all my own people as a sign of my preferring them, and it is an injustice to them if I put a stranger on par with them. But if I am to love him (with this universal love) merely because he, too, is an inhabitant of this earth, like an insect, an earth-­worm, or a grass snake, then I fear that only a small modicum of my love will fall to his share.…15 There is truth to what Freud is arguing here, that the basis of compassion lies in human relationships. Can strangers be loved? Can they belong to our moral community? By even posing this question, we fundamentally challenge liberal notions of universal ethics. The notion of “others” itself hardens the boundaries of the moral community. Distant bystanders’ perceptions of suffering, and thus of appropriate responses, are dictated by available information and the way it is presented. As Michael Ignatieff observes, “the media mediates between the suffering of strangers and the conscience of those in zones of safety.”16 Interestingly, the magnitude of atrocities, and the appropriate level of empathy, are dictated less by the number of people victimized and more by the nature of the victimization (e.g., if it is especially gruesome).17 Moreover, the greater the perceived magnitude of atrocities, the more robust the solutions favored.18 Images of suffering demand action. Otherwise, as Susan Sontag observed, we are merely acting as voyeurs.19 Unlike other senses, sight can be turned off (we have lids on our eyes, but not on our ears).20 We can change the channel. Within their immediate moral community, most people feel a common empathy and sense of moral obligation. The further we venture away from this community, the less this obligation is felt. There is a certain unreality to atrocity occurring on the other side of the world. Its victims are remote, and they are left diminished.21 We can be bystanders in the most distant sense, evading the empathetic distress and sense of obligation that might accompany evil occurring in our midst. In the modern age, our awareness of atrocity is greater than ever before. Yet humankind has perpetuated its emotional and moral distance. We cannot see beyond our own tribal boundaries. The human species is still relatively young;

Conclusion: killing without consequence?   259 perhaps our orientation towards such thinking is an evolutionary anachronism in the global age. Although there has been a kind of globalization of awareness, this has not been matched by a globalization of the moral community. Our moral identity, our moral community remains highly localized and internalized.22 A citizen of Sarajevo recalls: In October 1991, I was here in my nice apartment in peaceful Sarajevo when the Serbs invaded Croatia, and I remember when the evening news showed footage of the destruction of Vukovar just a couple of hundred miles away, and I thought to myself, “oh how horrible,” and switched the channel. So how can I be indignant if someone in France or Italy or Germany sees the killing taking place here day after day on their evening news and says, “Oh, how horrible,” and looks for another program? It’s normal, it’s human.23 To end genocide, a globalization of the human conscience may be required, an expansion of our moral horizons. Individualistic identity must be married to universalistic ethics. The suffering of others must be personalized to fully realize our human rights. Our circle of obligation, our moral community, must encompass all of humanity. Only then might all people be judged as equally human, and only then might genocide end.

Evil and intention: sleepwalking through the end of the world At the time, we didn’t understand the value of human life, and that’s why ordinary people became involved in violence.24 Genocide is an extreme, far-­reaching project; evil on a grand scale. Yet great evil is not always accompanied by great intentions.25 Genocide is not accidental, yet it is often a matter of moral tipping points that are crossed unthinkingly, or for reasons incommensurate with the breathtaking consequences of genocidal destruction. Innumerable individual acts produce genocide. For victims, genocide frequently represents, in a very real sense, the end of their world. Murder may appear a great moral threshold, even more so repeated murder. Yet we find many perpetrators seemingly sleepwalk through the end of their victims’ worlds. Paradoxically, this unconscious evil rebels against thousands of years of accumulated moral norms. The state arguably exists to protect its citizens, and yet also functions as an instrument of human destruction. The genocidal state of exception proves the rule – the state sets aside norms; individuals adjust; the state’s power is always prerogative. Can the state ever behave altruistically? States, like corporations, are centered on the advancement of internal interests. The complacency of perpetrators and bystanders also must not pass unchallenged. Even the surrender of free will is a willful act. A victim of the Khmer Rouge argues that perpetrator defenses of “just following orders”:

260   Conclusion: killing without consequence? would just be their excuses in order to reduce their weight of guilt. I honestly don’t want to hear such words. Although I agree that there were those who just obeyed orders to do such things, the question is: have you voluntarily offered yourself to be ordered?26 Thus, free will is essential for the protection of human life from the tyranny of the genocidal state. “The person who did something really wrong was the person who told us to kill the Tutsi,” rationalized one Rwandan génocidaire; yet it can be exceedingly difficult to locate this person. One may be left with the absurd impression of a perpetratorless genocide.27 Killing is greatly facilitated when it is divorced from consequence, and when perpetrators, as Sykes and Matza write, succumb to a “mood of fatalism.”28 The law, with its strict taxonomies, seeks to reduce complex social phenomena to binary judgments. One is either guilty or not guilty; one either did right or did wrong. Yet, as this work has shown, genocide is infinitely more complex. If many génocidaires do not possess true genocidal intent what does this mean for criminal and moral responsibility? Who is responsible for genocide? Peter A. French distinguishes between the “morally weak” and the “morally wicked.” The wicked believe in the justness of their actions, while the weak merely acquiesce.29 Yet both categories are evident in genocide perpetration. The morally weak obey authority and peer pressure, knowing their actions are wrongful, while the morally wicked may be ideologues who subscribe to propaganda that the enemy group is an existential threat to their survival. The morally weak reduce their moral costs while the morally wicked reverse morality. Keeping in mind that both of these categories of individuals may be criminally culpable for genocide, where does the greatest moral culpability reside? People are often both weak and wicked, pitiful and powerful, doing what seems best for them. Even if a factual ignorance operates, no perpetrator can truly claim a moral ignorance, considering the universal prohibition on killing. Aristotle asserts that there are some acts where “it is impossible ever to do right in performing them: to perform them is always wrong.”30 Thus, a person who commits such acts is acting out of preference rather than a misperception of moral rightness. However, as our decision-­making model has shown, preferences may be partly determined by one’s society and social role. Despite the universal prohibition on killing, genocide is normative within the genocidal state, as well as within the killing group. The desire of individuals for coherence often drives their normative alignment, even where this involves killing. We might determine that the state is ultimately responsible for genocide, as genocide requires state power. Yet the state lacks an organic life of its own. Perhaps, then, primary responsibility rests with those who steer the ship of state – the (“morally wicked”) senior leaders who oversee the structures and messages of the apparatus. In Ruggiu, the Rwanda Tribunal decided that because “the Accused did not personally participate in the massacres and did use his pistol,” his responsibility was lesser.31 Such a finding is blind to the reality of genocide as a complex state crime with multiple levels of perpetration.

Conclusion: killing without consequence?   261 The concept of moral luck posits that the moral wrongfulness of individual acts may sometimes be entirely determined by circumstance. Consider the classic example of two equally responsible drunk drivers, both of whom swerve off the road. One is lucky enough not to injure anybody, while the other kills two pedestrians who happen to be walking at that point in the road.32 They both performed the same wrongful act, yet the individual whose actions caused death is likely to face greater moral and legal condemnation. In the case of genocide, perhaps a form of moral luck can operate when two individuals are both born without any pathological predispositions, yet one is unlucky to be born into a criminogenic state and exposed to significant pressure to participate in killing. But evil cannot only be located in consequences. In Jallad Khana (Bangladesh), the perpetrators brought one of my interview subjects to a latrine that was “full of blood,” and, while holding him over the pit, slit his throat.33 Nothing happened; the blade was too dull from repeated use. In the confusion that followed, the victim escaped. It would be absurd to contend that these perpetrators are not responsible for murder simply because of their moral luck. The capacity to follow moral values in the face of potential disapproval, discouragement, or ostracism evinces moral courage.34 The law often asks what an “average person” would do when facing a certain set of circumstances. We must consider the frightening possibility that the “average person” would commit genocide. How much moral courage can we reasonably demand from individuals? Yet we know that, in similar circumstances, some individuals will be rescuers and others perpetrators. Among the famous conformity and obedience studies, there were many individuals who resisted conformity: one-­third in the Milgram Experiment, two-­thirds in the Asch Experiment, and a third in the Zimbardo Prison Experiment.35 These are evils of conformity rather than evils of autonomy, and yet still many individuals will refuse to perpetrate. Many philosophers and advocates have argued that there is a duty to disobey orders contrary to the will of God (as St. Augustine put it) or “any law that degrades human personality” (as Martin Luther King, Jr. argued).36 Moreover, even the “morally determinative” context of the morally unlucky is itself produced by wrongful individual acts. Genocide stands as the ultimate manifestation of collective evil, yet the individuals committing genocidal acts are often not enduringly and immutably evil. The perpetration of genocide requires the depersonalization and denigration of the self. Genocide is grounded in the pre-­eminence of the group self, and, more importantly, where individuals do not maintain an empathetic awareness of the selves of others. Genocide cannot thrive where there is recognition of human dignity and shared humanity, and where the sinews that bind us remain visible and strong. Before annihilation can occur, victims must be dehumanized and demonized, whereby “their layers of goodness must be peeled away to reveal the badness at the ‘core’ of ‘their being’.”37 It can only occur in a narrowed, reductionist context that values or devalues others based on community membership. To prevent such dehumanization, we must also recognize the humanity of the perpetrators, and something of ourselves in them.

262   Conclusion: killing without consequence?

Notes   1 Singer, “The Concept of Evil.”   2 Interview R9 (Rwanda), Kigali Central Prison, July 2009.   3 Erikson, “On Pseudospeciation and Social Speciation,” 51.   4 An example of universal identity may be the Buddhist concept of Anatta (no-­self ).   5 Ignatieff, “The Danger of a World.”   6 Koonz, The Nazi Conscience, 4.   7 Koonz, 1.   8 Glover, Humanity: A Moral History, 26–27.   9 Glover, 20–21. 10 Kluger, “What Makes Us Moral,” 32. 11 United States House Committee on International Relations: Subcommittee on Africa, “Hearing on Rwanda’s Genocide.” 12 Koonz, 2. 13 Koonz, 277. This ethic of reciprocity is reflected in many religious traditions including the Christian “Golden Rule” (found in the Bible in Matthew 7:12 and Luke 6:31). It is also represented in Islam in the Haddith (sayings of the Prophet): “No one of you is a believer until he desires for his brother what he desires for himself;” Judaism, in the Talmud: “What is hateful to you, do not to your fellow men;” and Buddhism, the Udana-­Varga: “Hurt not others with that which pains yourself.” 14 Koonz, 5. 15 Freud, Civilization and its Discontents, 31. 16 Ignatieff, The Warrior’s Honour, 33. 17 Harff, “Empathy for Victims of Massive Human Rights Violations,” 8. 18 Harff, 15. 19 Sontag, Regarding the Pain of Others, 71. 20 Sontag, 118. 21 There is experimental evidence that empathy is moderated by physical distance. See: Miller, Cushman, and Hannikainen, “Bad Actions or Bad Outcomes?,” 585. 22 Glover, 28. 23 Sontag, 99–100. 24 Interview R58 (Rwanda), Rugerero TIG Camp, September 2009. 25 Powell, interviewed on New Books in Genocide Studies. 26 Interview C06 (Cambodia), Phnom Penh, November 2008. 27 Interview R26 (Rwanda), Gisenyi Prison, August 2009. 28 Sykes and Matza, “Techniques of Neutralization,” 231. 29 French, “Unchosen Evil and Moral Responsibility,” 30. 30 French, 36. 31 Prosecutor v. Georges Ruggiu, para. 78. 32 Williams and Nagel, “Moral Luck,” 15–135, 137–151. 33 Interview Ba1 (Bangladesh), Jallad Khana, September 2008. 34 Staub, The Psychology of Good and Evil, 8. 35 Baum, The Psychology of Genocide, 88. 36 Kelman and Hamilton, Crimes of Obedience, 70. 37 Weisband and Thomas, “The Biocorporeality of Evil,” 86.

Works cited Baum, Stephen K. The Psychology of Genocide: Perpetrators, Bystanders, and Rescuers. London: Cambridge University Press, 2008. Erikson, Kai. “On Pseudospeciation and Social Speciation.” In War and Human Survival, edited by Charles B. Strozier and Michael Flynn. Lanham, USA: Rowman and Littlefield, 1996.

Conclusion: killing without consequence?   263 French, Peter A. “Unchosen Evil and Moral Responsibility.” In War Crimes and Collective Wrongdoing: A Reader, edited by Aleksandar Jokic. Oxford: Blackwell Publishers, 2001. Freud, Sigmund. Civilization and its Discontents. New York: Norton, 2005 (first published 1930). Glover, Jonathan. Humanity: A Moral History of the 20th Century. New Haven: Yale, 2001. Harff, Barbara. “Empathy for Victims of Massive Human Rights Violations and Support for Government Intervention: A Comparative Study of American and Australian Attitudes.” Political Psychology 8, no. 1 (March 1987). Ignatieff, Michael. “The Danger of a World without Enemies,” The New Republic, February 21 2009. Ignatieff, Michael. The Warrior’s Honour. New York: Henry Holt & Company, 1997. Kelman, Herbert C. and Lee Hamilton. Crimes of Obedience. New Haven, USA: Yale University Press, 1990. Kluger, Jeffrey. “What Makes Us Moral,” Time, December 3 2007. Koonz, Claudia. The Nazi Conscience. London: The Belknap Press of Harvard University Press, 2003. Miller, Ryan, Fiery A. Cushman, and Ivar A. Hannikainen. “Bad Actions or Bad Outcomes? Differentiating Affective Contributions to the Moral Condemnation of Harm.” Emotion 14, no. 3 (2014). Powell, Christopher. Interviewed on New Books in Genocide Studies (podcast), September 21 2013. Prosecutor v. Georges Ruggiu (Case No. ICTR-­97–31-I). Trial Chamber Judgement, June 1 2000. Singer, Marcus G. “The Concept of Evil.” Philosophy 79, no. 308 (April 2004). Sontag, Susan. Regarding the Pain of Others. New York: Picador, 2003. Staub, Ervin. The Psychology of Good and Evil: Why Children, Adults, and Groups Help and Harm Others. New York: Cambridge University Press, 2002. Sykes, Gresham M. and David Matza. “Techniques of Neutralization: A Theory of Delinquency.” American Sociological Review 22, no. 6 (December 1957). United States House Committee on International Relations: Subcommittee on Africa. “Hearing on Rwanda’s Genocide.” Washington Transcript Service, April 22 2004. Weisband, Edward and Courtney I. Thomas. “The Biocorporeality of Evil: A Taxonomy.” In Inside & Outside of the Law: Perspectives on Evil, Law, and the State, edited by Shubhankar Dam and Jonathan Hall. Oxfordshire, UK: Inter-­Disciplinary Press, 2009. Williams, B. A. O. and T. Nagel. “Moral Luck.” Proceedings of the Aristotelian Society 50 (Supplementary Volumes, 1976).

Index

Page numbers in italics denote tables, those in bold denote figures. accomplices 113 accountability 218, 231; see also responsibility adaptation 19–21 Adorno, Theodore 106 African Americans 72 Agamben, Giorgio 50, 65n30 age ranges 97–8 agency 46, 93, 98–100, 142, 153, 157–8, 177, 181 aggression 96, 101, 137, 147n64, 201, 228 Agnew, Robert 20, 201 Akazu patronage state 22; see also Rwandan Genocide Akcam, Taner 238 alcohol 137–8, 147n64, 156, 176, 178 alienation 28, 57 Allen, Francis 229 altruism 114–15, 174 Alvarez, Alexander 5, 6, 106 Amarondo (civilian security patrols, Rwanda) 23 amateur perpetrators see civilians Americas 217 Améry, Jean 18 amnesia 29, 217 Anders, Gunther 45 anger 137, 146n54 Angkar (regime, Cambodia) 53, 130, 181 animalization 73 animals 204 anomie (normlessness) in society 20–1, 23, 160 anonymity 11, 49, 125, 135 anti-miscegenation laws 52 Arendt, Hannah 138, 171, 253 Argentina 53

Arkan’s Tigers (paramilitary group, Bosnia) 95, 97, 106; see also paramilitaries Armenian Genocide 21, 23, 35, 37, 46, 96, 210–11, 224 army 55–6, 95, 99–100, 109 Arndt, Jamie 72–3 Aronson, Elliot 165 Aschenauer, Rudolf 184 attachment 109–10 Auftragstaktik (mission tactic, Nazi Germany) 61 Auschwitz-Birkenau concentration camp 73, 134, 162, 177–8, 202 Austria 21–2, 46 authoritarianism 25–6, 32, 45, 58, 63, 106 authority 48–9, 53, 124, 138, 176, 194, 230, 260; presence of figures of 133, 133–4; see also leaders; legitimate authorities autonomy 136, 152; denial of 181–2 avoidance behavior 156–7 Bach-Zelewski, Erich von dem 155 Balint, Jennifer 52, 244 Balkan Wars 69, 83, 87 banality 103, 137, 253 Bandura, Albert 45, 103, 172, 180, 231 Bangladesh 4, 11, 29, 199 barbaric enemy narrative 24, 78, 78, 83–4, 182 Barriga, Alvaro Q. 172 Baum, Stephen 104 Bauman, Zygmunt 20, 53, 195 Baumeister, Roy 155 Beccaria, Césare 245 Beck, Aaron 49, 172

Index   265 beliefs 30 Bengalis 29, 51 Bhavnani, Ravi 153 biophysical factors 107–8 biophysical responses 137 bodies/corpses 34, 209, 211 boredom 106 Bosniaks (Bosnian Muslims) 72, 77, 194, 245 Bosnian Genocide 11; context 46, 56–7; perpetration 127, 137, 154, 159, 161, 194, 196; perpetrators 95, 98, 105–6, 116, 175; post-genocide 218, 239–42, 245; pre-genocide 19–21, 25, 29, 37; propaganda 77; see also Bosniaks; Omarska; Serbs bounded rationality theory 145n17 Bowlby, John 105 Bradfisch, Otto 181 Brandt, Karl 35 Brannigan, Augustine 62, 63 Brehm, Hollie Nyseth 62, 98 broken window theory 21 Browder, C. G. 104 Browning, Christopher 6, 11, 97, 154, 156, 179 brutality 56 brutalization thesis 34–5 bullying 104 bureaucracy 53–4, 99, 199 Burma 51 Burundian Genocide 11; context 46, 57; perpetration 124, 151, 156, 159, 199, 203, 205, 209–10; perpetrators 96, 98, 107, 109, 174; pre-genocide 22, 24, 29, 33; propaganda 70, 76, 79; see also CNDD-FDD; Sans Échec bystanders 113–15, 127, 145, 227, 258, 260 Cambodian Genocide 4, 11; context 46, 51, 63; perpetration 127, 130–1, 196, 199, 206; perpetrators 80, 98, 101, 130, 181; post-genocide 226–7, 244–5; pregenocide 19, 20, 36, 38; propaganda 73, 80; see also Angkar; chhlop; Khmer Rouge; Pol Pot; Tuol Sleng cannibalism 160–1 Catholic Church 35 Ceceka (“be silent”, Rwanda) 184 Chalk, Frank 4 chhlop (intelligence cadres, Cambodia) 19 children 52, 97–9, 105, 138, 156, 180, 196, 205

choice 2, 63–4, 96, 101, 117, 176; see also decision-making; free will churches 35 Cilician Massacre 37 citizenship 51 civic nationalism 222 civilians 54, 58–9, 80, 99, 100, 131, 146n28, 152; see also bystanders claim of inevitability 179 claim of inner opposition 181 claim of normality 178–9 claim of relative acceptability 179–80 Clarke, R. V. G. 107 clothing 73 CNDD-FDD (National Council for the Defense of Democracy, Burundi) 29, 96, 159, 162–3 co-action effect 178 “cockroaches” euphemism 28, 36, 73, 81, 82; see also dehumanization coercion 31, 45, 47–8, 100, 110, 131, 142, 154, 164, 176, 198, 244 cognitive dissonance 53–4, 110, 112, 141–2, 172, 254 cognitive opening 71–2 collective action problem 139 collective intent/motive 241 collective memory 223–4 collective rehabilitation 228–9 collective trauma 226–7 Collins, Randall 59, 154 colonial era 23, 198; post- 62 compartmentalization 54, 135, 155, 177, 202 compassion 258 compassion killing 180, 184 competition 159 complicity 178, 243 compromised perpetration 102–3, 110–11 concealment 128, 184, 209–11, 217 concentration camps 46, 59–60, 98, 104–5, 128, 137, 156, 160, 175, 177, 205, 209; see also Auschwitz-Birkenau concentration camp confessions 225 conformist perpetration 102–3, 109–10, 140 conformity 49–50, 64n24, 111, 135, 154, 182, 261 confrontation tension 154–5 conscience 256–9 conspiracy narratives 81 conspiracy as a form of criminal responsibility 238

266   Index conventional values 45–6 conventionalization 52, 95, 194 Copes, Heith 183 coping mechanisms 200–2 Cornish, Derek Blaikie 107 corvée (“communal labor”, Rwanda) 62, 157 courts 194, 220–2, 236–8 creativity 159 crime 20–1, 93–4, 109 crime coercive/facilitative systems 61 crimes against humanity 4, 21, 45, 171, 227, 238–9, 243, 244 criminal behaviour 60–1 criminal history 95–6 criminalization of social groups 27–8, 36, 52 criminals, convicted 95–6, 102, 106, 110 criminology 5–6, 105–6, 137, 171 Croatia 77, 84, 241 Croats 95 crowd psychology 181–2 Dallaire, Roméo 257 Damant, Froukje 17–18, 113 Darfur Genocide 6, 96 Darley, John 114, 187n71 deception 155, 170, 225–6 decision-making 128–44, 129; direct and indirect perpetrators 142–4; immediate margin of discretion 133, 133–8; moral rules 141–2; norm bandwagons and tipping points 144–5; risks and incentives 138–41; role margin of discretion 131–2; social margin of discretion 130–1 decision-priming 129, 129 dehumanization 28, 34, 71–3, 82, 130, 137, 152, 174–6, 183, 261; see also “cockroaches” euphemism demarginalization 221 Democratic Republic of Congo 98 denationalization 51 denial 61, 135, 224, 226, 230; systems of 183–4, 209–11 denial of autonomy 181–2 denial of genocide 201 denial of humanity see dehumanization denial of injury 177–8 denial of responsibility 176–7 denial of the victim 174–5 depression 105–6 derealization 178, 202, 226 Des Forges, Alison 83, 140, 154, 195

desensitization see habituation/learning by doing deterrence 245–8 deterrence theory 245–6 deviancy 17–18, 26–7, 36, 99, 102, 116, 142, 171, 218 direct perpetrators 100–1, 142–4 discipline 165 discourses 78, 78, 78–87 disgust 100, 139, 150, 155 disobedience 48, 130–1, 133–4 distance/distancing 117, 136, 160, 178 doctors 178, 180, 182, 202 doubling, theory of 178, 202, 230 doubt 112 drugs, use of 137, 202 Drumbl, Mark 244 dual-track approach 210–11 duress 176, 184 duty 49, 132, 172, 179 East Pakistan see Bangladesh; Bengalis; Pakistan Eastern Front 55, 65n64, 97, 137–8, 204, 207, 210 economic crises 21–3, 38n15 education 110 Ehlers, Erst-Boje 131, 200 Eichmann Trial 156, 253 Einsatzgruppen (death squads, Nazi Germany) 1, 48, 54–5, 65n53, 99, 127, 131, 135, 138, 155–6, 181, 184, 200, 204–5, 227; see also Holocaust elites 22, 26, 30, 33, 36–8, 62, 100, 108, 150–3, 192–3, 219–20 emergency powers/states of emergency 50 emotions 75, 137–8 empathy 45, 105, 114, 158, 258 empowerment 205 Enabling Act (1933, Nazi Germany) 50 enjoyment 205 entrepreneurs/entrepreneurial perpetration 20, 25, 100–2, 101, 108–9, 140, 143, 151, 244 escalation 36–7, 58, 191–4 escape 162 ethics 256, 259 Ethiopia 4 ethnic norms 31–2 ethnicity 23, 31, 73, 151–3, 158, 174 euphemisms 73–4, 157–8, 177, 183, 231 euphoria 205 euthanasia killings 178, 182, 184, 187n37 evil 1–2, 103, 179, 253, 259–61

Index   267 existential fear 18, 23, 33 existential threat 18–19, 21, 23–8, 33, 37, 46 exposure 124, 126, 129 external controls 116 external costs 115 external deviance 63 external justification 170 extreme violence 160–1, 231 extremism 192–3, 221 Êzidî 162, 199 Falthauser, Valentin 184 fascism 79, 106 fatalism 230, 260 fear 18–21, 23, 29–33, 84–5, 127, 154, 165 Fein, Helen 35 female perpetrators 12, 46, 47, 97–8, 113, 132 Festinger, Leon 172 feudalism 62–3 fictions 194 “Final Solution, the” 157; see also Holocaust fleeing 156–8 “foot in the door” phenomenon 58 football fan clubs/football supporters 97, 106 forced marriage 162 forced perpetration 95 foreign enemy narrative 78, 83, 182 foreignness 72–3, 83 forward panic 58–9 Frank, Hans 181 free will 2, 230, 259–60 freedom 32 free-riders 85, 139, 153 Freikorps (volunteer units, Nazi Germany) 107 French, Peter A. 260 French Revolution 72 Freud, Sigmund 181, 258 Fromm, Erich 32 frustration 136–7 Fujii, Lee Ann 6, 9, 135, 151 fundamental attribution error 72 Gacaca (Rwandan court) 184 gangs 57–8 gassings 46, 155, 160 gender 12, 98, 113, 132, 228 genetics 107–8 génocidaires (perpetrators, Rwanda) 2, 63,

94, 103–4, 107, 110, 127, 137, 154, 172, 175, 178, 195, 231, 247, 256, 260 genocidal state 127–8, 133, 138, 142, 179, 210, 223, 226, 245–6, 249, 259–60; context of 19–23; moral disengagement of 44–6; as organizational culture 60–1; reforming 219–22 genocide convention see United Nations (UN) Convention on the Prevention and Punishment of the Crime of Genocide genocide 1–6; adjusting to 191; “completed”/“interrupted” 198–9; comprehending 253–6; contexts of 46–7; crime of 3–5; criminology of 8–12; denial of 201; of 23–6; enigma of causation 5–6; forgetting 217–19; genesis of 17–18; “hunting” as 157–8; individual responses; 111–16; interrupted 217; interviews 12, 12–13, 13; momentum 211; organization of 52–4, 151, 153; organizing 52–4; participatory 46–7; progression of 191–9; prosecuting 237–45; spaces of 32–7; specialized 46–7; theory on 7–8; see also genocidal state; killing; perpetration; perpetrators; violence Germany see Nazi Germany ghettos 46, 59 Gibbs, John C. 172 Gikongoro, Rwanda 75, 193 Gisenyi, Rwanda 22, 192 Glazar, Richard 60 Glover, Jonathon 106 goal gradient theory 199 Goebbels, Joseph 201–2 Goldhagen, Daniel 154 Green, Penny 160 Greiser, Arthur 69 Gross, Jan 134–5 Grossman, Dave 154, 157 group dynamics 134–6 group membership 29–30, 49, 72 group norms 154, 185 group radicalization 28–31 group self 207, 261 group settings 174 group solidarity 144 group values 158 group-binding narratives 78, 85–7 groupism 49, 61, 110 guardianship 126 Guardiens de la Paix (Tutsi paramilitary group, Burundi) 151 Guatemala 224

268   Index guhiga (“hunting”, Rwanda) 157–8 guilt 103, 205, 228, 244, 249 Gujarat pogrom 59 gurakora (“work”, Rwanda) 157 Gurtner, Franz 50 guterwa (civic defenses, Rwanda) 193 “Gypsy hunts” 73 habituation/learning by doing 93, 95, 127, 157, 160, 191, 197, 200, 203, 203–5, 211 Habyarimana, Juvénal 21, 22, 27, 31, 36, 38, 83, 137, 193 Hagan, John 6 Hamidian Massacre 37 Hamilton, Lee 129 Haritos-Fatouros, Mika 106 harm 43–5 Hartl, Albert 131, 200 hate crimes 108, 257 hate propaganda 70, 74, 76, 78, 152 hate speech 62, 76, 152, 221 Hatzfeld, Jean 11 Herero Genocide 46 Hess, Rudolph 85 Hesse, Heinrich 180 Heydrich, Reinhard 162 hidden enemy narrative 78, 80, 182 hierarchy 52–3, 56, 63, 106, 231 high-propensity individuals 140–1, 157 Hilberg, Raul 54, 177 Himmler, Heinrich 143, 162, 195 Hinton, Alex 11, 159 Hirschi, Travis 109–10, 201 history narratives 25, 34, 79, 182 Hitler, Adolf 35, 50, 85, 105–6, 119n80, 162, 181 “Hitler Directives” 171 Hoffer, Eric 72, 183 Hola, Barbora 103 Holocaust 19, 21, 35, 45–8, 54–5, 65n64, 94–9, 104, 113–14, 127, 134, 137, 146n28, 157, 160, 161, 175, 180, 192, 194–6, 199, 204, 209–10; see also Auschwitz-Birkenau concentration camp; Einsatzgruppen; “Final Solution, the”; Jews; Nazi Germany; Nuremberg Control Council Number Ten trials, post-; Nuremberg Trials Hood, Bruce 49 Horak, Eugen 134 Horn, Otto 156 Horowitz, Donald L. 126 hostile framing theory 172

hostis humani (enemies of humanity) 26 human rights 50, 101, 128, 259 humanity gap (social distance) 28–31, 78, 83, 176 humiliation 160 humor 159 Hutus 22, 25, 28–9, 36, 47, 72, 81, 85–6, 96–7, 103, 135, 151–2, 154, 184, 219, 246–7; “Ten Commandments of the Hutu” 85–6, 143, 147n80 Ibyitso (accomplices, Rwanda) 80 identity 28–9, 45, 140, 157, 171–2, 178, 230, 256–7, 259 ideologues 100–1, 140 ideology 73–5, 95–6, 106, 136, 142–3, 153–4, 162, 172, 183, 185, 200–1, 209 igitero (“to wage war”, Rwanda) 201 Ignatieff, Michael 160, 258 Immediate margin of discretion 133, 133–8 imperialism 74–5, 220 impunity 125, 126, 128 Impuzamugambi (paramilitary group, Rwanda) 99; see also paramilitaries incentives 138–41, 154; financial 108, 139 incitement 3, 25, 47, 70–1, 74, 98, 108, 129, 221 indigenous peoples 198–9, 217 individual radicalization 31–2 individual recidivism 112, 199–209, 203 individual rehabilitation 229–31 individual trauma 227–8 individual will 145 individualism 61 Indonesian Genocide 46, 218, 224 inevitability, claims of 179 infiltration narratives 23, 80–2 informers 113 Ingando program (Rwanda) 10, 230 in-groups 20–1, 24, 28, 33, 38, 46, 49, 71–2, 74, 78, 80, 84–5, 87, 110–11, 130, 139, 154, 158, 185–6, 224 inhibitions to killing 100, 137, 154–6, 157, 200, 201, 202, 206 injury, denial of 177–8 Inkotanyi (RPF fighters, Rwanda) 81–3, 84 inner opposition, claims of 181 institutionalization 45, 95 intent 238–43, 259–61 Interahamwe (paramilitary group, Rwanda) 23, 35, 56–7, 63, 76, 97, 99–100, 111, 115, 132, 151, 180, 193–4; see also paramilitaries

Index   269 intergenerational transmission 6, 25 internal controls 116 internal costs 115 internal deviance 63 internalization 113 International Covenant on Civil and Political Rights (ICCPR) 248 International Criminal Court (ICC) 4, 161, 220–1, 243, 246 International Criminal Tribunal for Rwanda (ICTR) 4, 229, 239, 243, 260 International Criminal Tribunal for the Former Yugoslavia (ICTY) 4, 229, 239, 245, 249 intervention 114–15, 199, 217 intoxication 137–8, 147n64, 156, 176–7 inyenzi (“cockroaches”, Rwanda) 73 Iraq 12, 56, 221 ishyano (abominable act of killing, Rwanda) 19 Islamic State 12, 162, 199, 221 isolation 127–8, 185, 188n93 Jackson, Robert 247 Jallad Khana (Bangladesh) 261 Janjawiid (perpetrators, Darfur) 96 Japan 155 Jelisić, Goran 196, 242 Jessee, Erin 19 “Jewish question, the” 162 “Jewish World Powers” 80–1 Jews 17, 21–3, 46, 59, 72, 75, 79–83, 134, 137, 143, 155, 162, 171, 179–80, 198, 201–2, 209–10, 241; see also “Final Solution, the”; Holocaust joint criminal enterprise 178, 194, 238, 243, 246 Jonassohn, Kurt 4 Josting, Ervin 180 Józefów massacre 134, 156, 179–80 justice 4, 236–50; prosecuting genocide 237–45; prosecution as prevention 245–8; transitional 218–21, 225, 232n3, 236, 248–9 justification 74, 204 just-world thinking 35–6, 175 Kamenets-Podolsk 138 Kang Kek Iew (Duch) 131 Kangura magazine (Rwanda) 77, 79, 87, 175 Kapos (prisoner foremen, Nazi Germany) 96 Karama Church, Rwanda 211

Kayishema, Clément 196 Kelman, Herbert C. 129 Kenya 220–1 Kenyatta, Uhuru 220–1 Kershaw, Ian 48, 113 Ketsch, Karl 201 Khmer Rouge 19, 20, 28, 34, 36, 38, 51, 53–4, 63, 73–4, 82–3, 97–8, 108, 127, 130–1, 162, 231, 244, 259–60; see also Cambodian Genocide; Pol Pot; Tuol Sleng Kibungo, Rwanda 193 Kibuye, Rwanda 196 Kigali, Rwanda 22, 193 killing 150–65; boundaries of violence 158–63; from context to 150; inhibition 154–6; options 156–8; in organizations 54–5; prohibition of 19–20; as “work” 63, 157, 177; see also perpetration; perpetrators; prohibition of killing; violence King, Martin Luther, Jr. 248, 261 Klehr, Josef 206 Kluger, Jeffrey 257 Kony, Joseph 165 Kosovo 22 Kremer, Johann Paul 202 Kren, George M. 105 Kretschmer, Karl 203 Krumbach, Alfred 135 Kuper, Leo 36 Kvocka, Miroslav 249 kwicha kirazira (“killing is taboo”, Rwanda) 19 Lachs, John 54 Langerbein, Helmut 99 Latané, Bibb 114, 187n71 Latvia 65n53, 99 law 134, 184, 248, 260–1 Le Bon, Gustav 181 leaders 100–1, 108, 143–4, 153, 244; see also authority; perpetrators, high-level Lebensraum (expansion concept, Nazi Germany) 198 ledger metaphor 201 legal ordering 237–8 legal systems 51–2 legitimacy 35, 70, 81, 175 legitimate authorities 18, 70, 116, 130, 133–4, 138, 140, 175, 182, 186, 197, 244 Lerner, M. J. 35 Lewin, Kurt 199

270   Index Liau, Albert K. 172, 230 Lifton, Robert Jay 106, 207, 230 Lithuania 65n53, 161 local participation 152, 192–4 Łódź ghetto (Poland) 59 looting 95, 109, 139 Lord’s Resistance Army (LRA, Uganda) 48–9, 75, 97–8, 158, 161, 163–5, 204, 211, 227 low-level perpetrators 26, 43, 49, 99–100, 108, 139, 226, 248 low-propensity individuals 140–1 loyalty 30, 32, 56, 58, 63, 103, 106, 109–10, 132, 135, 152, 164, 174, 182 Lukić, Milan 127 lustration 221 McDoom, Omar Shahabudin 62, 97, 137 malingering see deception Mann, Michael 11, 97–8, 107 marginal perpetrators 95–6 marginalization 21, 27, 32, 35–6, 45, 51, 56–9 martial law 50 Maruna, Shadd 183, 229 masculinity 48, 58, 99, 106 Mass, Peter 159 mass media 75–7, 83–5, 175 mass participation 46, 54–5, 59 massacres 22, 25, 35, 37, 156–7, 202, 205, 211 Mattner, Walter 204 Matza, David 171, 176, 231, 260 May, Larry 241 Mégret, Frédéric 249 memory 223–6; collective 223–4; individual 224–6 men 57–8, 97–8, 107, 132, 205 mental conditions 104, 107–8 Mentz, Willi 180 mercy killing 180 Merton, Robert 20 Milgram, Stanley 48, 158 militias see paramilitaries Mills, Judson 165 Milosevic, Slobodan 33–4, 77, 79, 85, 105, 245 Minsk 143 Mironko, Charles 193 Mladić, Ratko 25, 37, 245 mob violence 59, 192 mobilization 61–3, 99, 125, 139, 255 mobilizing perpetrators 150–4 Mobius, Kurt 47

Moll, Otto 177–8 momentum, genocidal 198–9, 211 Mommsen, Hans 195 moral community 17, 24, 36–7, 43, 45, 71, 114, 170, 175, 218, 221, 224, 231, 249, 254–9 moral conflict 103, 141, 254 moral consistency 141 moral context 2, 62–3, 111, 116, 124, 130, 132, 135, 170, 185, 220, 253–4 moral cost 157, 183 moral disengagement 44–6, 47, 94–5, 101–4, 108, 110, 117, 137, 142, 158–9, 200 moral doubt 158 moral habits 47, 93, 101, 116, 127, 131, 152, 204, 254 moral hazards 139 moral hierarchies 231 moral identity 45, 140, 157, 256–7, 259 moral luck 261 moral neutralization 50, 87, 95, 115, 131, 134, 141, 142, 170–82, 200–1, 207, 225, 228, 230, 254; neutralization-drift theory 171–4, 173; techniques 170–82, 173, 201, 225, 245 moral neutralization-drift theory 171–4, 173 moral norms 70, 94, 111, 130, 132, 142, 257, 259 moral principles 174, 257 moral risk 139, 141–2 moral rules 71, 93, 102, 116, 130–1, 136, 141–2 moral rupture 26–7, 46, 52, 182, 218–19, 254 morally weak/wicked 260 motivation 99–100 MRND (Mouvement républicain national pour la démocratie et le development, Rwanda) 30, 62, 74, 77, 151–2, 193 Muslims 37, 51, 72, 77, 81, 84, 95, 127, 152, 186n7, 194, 196, 241 Myanmar see Burma myths/mythohistories 25, 71 Namibian Genocide see Herero Genocide narcissism 104, 159, 161 narratives, enemy 70, 72, 74, 78, 78–80, 83–4, 87, 182–3 nationalism 34, 227 Nazi Germany: context 45, 46–7, 50–4, 59–61, 63; perpetration 130, 138, 146n28, 155–6, 161–2; perpetrator rationalization 171, 177–8, 180, 182,

Index   271 184, 187n37; perpetrators 94–5, 97–8, 106–7, 110, 113; post-genocide 219, 221, 227, 229, 257; pre-genocide 21–3, 28; propaganda 69, 75, 81, 83; sustaining genocide 192, 198, 200–2, 207, 209 Nazi Party 20, 27–8, 34, 50, 98 “necessary evil”, killing as 94–5, 100, 199, 202 necessity, subjective 134 Needleman, Carolyn 61 Needleman, Martin L. 61 Nelson, Douglas 142 Netherlands 69–70, 80–2 neutralization see moral neutralization Ngeze, Hassan 175 Neitzel, Sonke 45, 231 non-participation 110, 129, 138–40, 153–4, 163, 174, 255 norm bandwagons 18, 144–5 normality, claim of 94–5, 178–9 normative shifts 71–3, 111 normativity 141 normlessness (anomie) in society 20–1, 23, 160 norms facilitating violence 47–51; conformity 49–50; obedience 47–9; urgency 50–1 NSB (Dutch Nazi Party) 79, 82 Nuon Chea (“Brother Number Two”, Khmer Rouge) 54, 101; see also Cambodian Genocide Nuremberg Control Council Number Ten trials, post- 243 Nuremberg Trials 247 nurses 178 Nyiramasuhuko, Pauline 226 obedience 47–9, 56, 63, 64n18, 111, 131, 134, 138, 152–4, 176, 260–1 Ohlendorf, Otto 135, 160 Omarska 137; see also Bosnian Genocide Ongwen Trial 48–9, 161; see also Lord’s Resistance Army; Uganda Oppenheimer, Joshua 218 opportunity space 45, 125 ostracism 7, 31, 46, 85, 112, 134–5, 139–40, 154, 225, 231, 261 “other” 46, 70, 72, 98 Ottoman Turks 21, 23, 37, 210–11; see also Armenian Genocide out-groups 18, 21, 25–30, 33, 38, 49, 72–8, 80, 83–5, 135, 158; see also victim groups

pain-blood-death complex 205 Pakistan 224 paralysis 127 paramilitaries 56–8, 95–7, 99–100, 106, 107, 152, 154, 185, 194 Parker, Patricia 6 passivity 94–5, 113–14, 124, 127–8 pathological perpetration 102–3; biophysical factors 107–8; psychological factors 103–7 patriarchy 96, 98, 132 patronage 22, 33, 63 payments 161–2 peasants 83, 134 perpetration 124–5; decreasing cost of 203, 203–6; forced 163–5; ideological 102, 108, 143–4; mass violence and individual will 145; matrix 129, 129; opportunity 125–8, 126; pathways to 101–11; see also decision-making perpetrator narratives 10–11, 224–5; evolution of 223–31; reproduction of 245 perpetrators 2, 5, 93–117, 259–60; coalitions of 99–101; controls 116; direct 100–1, 142–4; distance and propensity 117; effects of killing on 199–200; genocidal context 45–6, 48–50; highlevel 74, 99–100, 139, 226, 229, 248; identity 207–8; indirect 142–4; instrumental 103, 143–4; mobilizing 150–4; opportunity 125–8, 126; profiling 94–9; propaganda 74; R22, Hutu Male, Mbazi Sector, “Ignace” 196–8; reluctant 100; as a term 4–5; trauma 200, 226–8; variable individual responses to genocide 111–16; see also female perpetrators; leaders personality disorders 104 Petri, Erna 98–9, 241 Plato 256–7 Plavsic, Biljana 20–1 pogroms 59, 99, 134, 192, 199 Pol Pot 143; see also Cambodian Genocide; Khmer Rouge; Tuol Sleng Poland 4, 69, 134, 156, 171, 202, 248 polarization 33–4 police 55–6, 72, 100, 131, 135 politics 28, 33, 37, 59, 70, 83–4, 114; crises 21–3, 38n15 Politika 79 positive reinforcement 206 Post-Traumatic Stress Disorder (PTSD) 213n55, 227–8

272   Index posturing 110, 156–8 poverty 96 power 18, 20–3, 25–6, 28, 31–7, 44, 45, 48, 53, 59, 135, 162, 236; see also state power pre-colonial era 62 prisons 59–60 privacy 60 private opinions 69 private sphere 58 privatization 113–14 progression 191–9; end of genocide 198–9 prohibition of killing 19–20, 138, 155, 177, 173, 182, 200, 260 propaganda 20, 25, 27, 31, 36, 45, 47, 48, 53, 69–87, 128–9, 136–8, 144, 153, 170, 174, 182–3, 192, 196–8; dehumanization dynamic 71–3; functions of 70–1; genocidal discourses 78, 78, 78–87; genocidal ideology and 73–5; mechanisms of 75–7 propensity 93–4, 102, 110, 117, 129, 157; to rescue 114–16 property 51, 109, 162 prosecution 237–48; attribution of responsibility 243–4; genocidal intent 238–43; legal ordering 237–8; as prevention 245–8; reproduction of perpetrator narratives 245 Protocols of the Elders of Zion, The (Nazi propaganda) 75 proxies 192 proximity 97, 101–2, 117 Prunier, Gérard 96 psychological dissociation 202 psychology 49, 58, 72, 103–7, 135–6, 155; crowd 181–2 psychopathy 96, 104–5 public meetings 152–3 public opinion 69, 144, 192 public/private spheres 32–7, 46, 52 ‘purity’ 74 race 28, 71, 82, 83, 132 racism 221 radicalization 26–32; group 28–31; individual 31–2; state 27–8 radio 76–7, 80, 85, 175 Radio Rwanda 77 Ranatovic, Zeljko (Arkan) 106 randomness (genocidal serendipity) 112, 151 rape 139–40, 163 Rappoport, Leon 105

rational-choice theory 138 rationality 128–9, 143, 158 rationalization 76, 170, 200–2, 217; see also moral neutralization Rauschenback, Mina 74 Ravensbrück concentration camp 209; see also Holocaust reason 75 Reasoning Criminal Model 107 recidivism 112, 199–209, 203 recruitment 153, 165 Red Star Football Fan Club 97, 106 red-outs (memory loss) 137, 226 refugees 23 rehabilitation 228–31, 249 Reichsgau Wartheland (Nazi Germany) 63 Reichstag, burning of the (1933) 27 relative acceptability, claims of 179–80 relativism 257–8 remains see bodies/corpses remorse 112, 241, 229, 230, 240, 241–2 remote violence 155 rescue 154 rescuers 114–16, 195–6, 261 residential schools 52 resistance 113, 126, 127, 156–8, 227 responsibility 142, 177–8, 226, 242; attribution of 243–4; denial of 176–7 restraint 36–7 retreatism 114 revenge 19, 78–9, 161, 199 Ricoeur, Paul 9 riots 59, 109, 126 risks 138–41, 154 Rohingya minority (Burma) 51 role margin of discretion 131–2, 226 role models 106 roles 111–14, 112–13, 182 routine activity theory 126 RTLM (Radio-Télévision Libre des Milles Collines, Rwanda) 76–7, 80, 175, 194 Ruggiu, George 196 Ruhengeri, Rwanda 22 Russia 23 Rwanda Patriotic Front (RPF) 22, 28, 34, 37, 62, 73–5, 81–3, 108–9, 158 Rwandan civil war (1991) 87 Rwandan Genocide 1, 10, 12, 12, 13; context 43, 46–7, 49, 51–2, 54–6, 59, 62; genocidal killing 151–4, 156–9, 161–2, 164; organization and mobilization of genocide 192–3, 195–6, 199–202, 204–9, 211; perpetrator decision-making 124, 126–8, 130,

Index   273 132–7, 133, 139–41, 143–4; perpetrator rationalization 173–82, 184; perpetrators 94–100, 103, 108–10, 114–15; postgenocide 218–19, 225–7, 229–31, 244–7, 249, 256–7, 260; pre-genocide 19–24, 27–30, 33–8; propaganda 69, 71, 73–7, 79–86; see also génocidaires; Hutus; Tutsis Rymond-Richmond, Wenona 6 sadism 95, 104–5, 140, 159 sadistic shift 208–9, 255 Sans Échec (paramilitary group, Burundi) 56–7, 96, 109, 203; see also paramilitaries Sarajevo, Bosnia 259 Scalia, Damien 74 scapegoating 29–30, 71–3 Schlink, Bernhard 2 Schmitt, Carl 257 Schutzstaffel (SS, Nazi Germany) 28, 55, 95–9, 105–7, 110, 132, 134, 137, 155–6, 180, 200–1, 203, 206, 209–10; see also Einsatzgruppen secondary transgression hypothesis 206 Security Prison 21 (S-21, Cambodia) 98 self-control 94, 102, 116, 136–7, 141–2 self-deception 226 “self-defense”, genocide as 19, 23, 28, 33–4, 85, 136, 184 self-definition 158–9, 183 self-esteem 104, 106, 114 self-fulfilling prophesy 73, 128 self-image 142, 173, 185 self-justification 200–2 self-objectification 61, 124, 142, 181 self-pity 100 self-serving distortions 172 self-worth 114 selfhood 31–3, 71, 139, 230–1, 261 Sémelin, Jacques 160 sender-receiver model 70, 75 sentencing 112, 241, 244 Serbs 20–2, 25, 33–4, 37, 56, 69, 72, 77, 79, 83–5, 95, 106, 116, 154, 175, 220 Sereny, Gitta 11 Šešelj, Vojislav 220, 245 sexual violence 11, 98, 162–3, 165 shame 134, 174, 226, 228, 249 Shawcross, Arthur 105 Sherman, Lawrence W. 247 shootings 46, 99, 138 Silberberg, Eugene 142 Simmons, C. H. 35

situational action theory 8, 93–4, 116, 124, 129 slavery 54, 162 “Slavs” 198 “sleepers” 105 slogans 75 Smeulers, Alette 102, 103 Sobibór death camp 128, 160, 177 Social Bond theory 109–10 social bonding 152 social context 10 social distance see humanity gap social margin of discretion 130–1 social ostracism 154 social relations 113–14 social ruptures 217 socialization 56, 105 sociology 59 Soldaten (Neitzel and Welzer) 45, 231 Solomon, Richard 205 Sontag, Susan 258 South Africa 36, 218 Soviet Union 127 Speer, Albert 54 Spiegel, David 208 Sprayer, Ludwig 182 Srebrenica massacre 127, 152, 222, 239–40, 257 Staerklé, Christian 74 Stalin, Josef 104 Stangl, Franz 128, 160, 177 state, the 6, 37, 50–2, 259; radicalization 27–8; see also genocidal state; state power state power 7, 62–3, 87, 93, 101–2, 125, 179, 222, 227, 249, 254–5, 260 Staub, Ervin 106 Steiner, John 105 stereotyping 72 Sternberg, Karin 78 Sternberg, Robert J. 78 stigmatization 47, 78, 78, 78–85 Stoetzel, Jean 69 Story-Based Theory of Hate 78 strain 21–3, 38n15, 45–6, 59, 96, 136 strain theory 19–21, 38n12 Straus, Scott 6, 11, 47, 62, 89n52, 97, 152 stress 200 Sturmabteilung (SA, Nazi Germany) 28, 107 subcultural narratives 184–5 subcultural theory 172 substance abuse 202 Sunstein, Cass 18

274   Index survival discourses 26–8, 201 Sutherland, Edwin 106 Svensson, Robert 93–4 Sykes, Gresham 171, 176, 231, 260 tabula rasa 198, 223 Tajfel, Henri 36 target determination 126, 161 Tätigkeits-und Lagebericht (publication, Nazi Germany) 200 Taubner, Max 209–10 technicalization 72, 155 technique, killing 155, 157 television 77 “Ten Commandments of the Hutu” 85–6, 143, 147n80 tension 63–4, 95, 154–8, 172 Terra Nullius (uninhabited land) 198–9 theatre, genocide as 135, 160 theft 108, 139, 162 threat 18–19, 21, 23–8, 33, 37, 46, 136 tipping points 36, 144–5 Todorović, Stevan 241 torture 60, 96, 106, 162, 175, 209 training 56, 157, 193, 196–8 transcendence 32 transgressive communities/subcultures 57–8, 132, 135, 142, 184–5, 247 transitional justice 218–21, 225, 232n3, 236, 248–9 trauma 200, 226–8 Trbić, Milorad 239–42 Treblinka concentration camp 60, 128, 137, 156, 160, 177, 180; see also “Final Solution, the”; Holocaust Treiber, Kyle H. 43 trials 223–4, 237, 244 tribunals 218, 220, 229–30, 249 Tuol Sleng (prison, Cambodia) 60, 73, 130–1, 159, 175, 206; see also Cambodian Genocide Turks 23, 37, 72, 224 “Tutsiness” 151–2 Tutsis 22–5, 27–9, 34, 36, 47, 70, 72–6, 79–83, 85, 110–11, 135, 151–2, 154, 158, 175–6, 196, 199, 201, 211, 218–19, 246–7; see also Hutus; Rwandan Genocide Uganda 4, 11, 97–8, 160–1, 163–5, 204; see also Ongwen Trial Ukraine 65n53 umuganda (work parties, Rwanda) 62–3, 157, 193

unemployment 96, 107, 110 United Nations (UN) Convention on the Prevention and Punishment of the Crime of Genocide 3–4 United Nations General Assembly Resolution 96 (I), “The Crime of Genocide” 238–9 United States of America (USA) 56, 72 universal values 103 urgency 50–1, 154 Ustase (fascist group) 25, 77, 79 utilitarianism 34, 74 vanguard perpetrators 100, 104 Verwimp, Philip 98, 109, 144 victim groups 17–21, 24–36, 44–7, 51–2, 59, 70, 73, 110–11, 114, 130, 137, 160, 220–2, 226, 240, 245, 248, 250; categorization of 17, 26, 36, 46, 75; see also out-groups victim inferiority 128 victim vulnerability 25, 124–8, 126 victim-blaming 136 victimization 101, 103, 110–11, 127, 136; reversal of 84, 136, 175 victims, denial of 174–5 Vietnam War 105, 178 Vietnamese (Yuon) 83; see also Cambodian Genocide violence 17–19, 21, 23, 25–9, 34–6, 45–7, 62, 71, 95–6, 132, 135, 145, 255; escalation of 191–4; extreme acts of 160–1, 231; structures facilitating 51–60 violence amateurs see civilians violence specialists/professionals 54, 55–8, 100–1, 131, 152, 157 Voltaire 69 von Reichenau, Walther 209 vulnerability, victim 25, 124–8, 126 war 23, 35, 144, 201 war crimes 21 War Crimes Chamber of the Court of Bosnia 239–41 Ward, Tony 160 Weber, Max 143 Welzer, Harald 45, 231 White Eagles (paramilitary group, Bosnia) 57, 95; see also paramilitaries Wiesel, Elie 5 Wikström, Per-Olof H. 43, 93–4 William, Schabas 21 witnessing 135

Index   275 women 12, 46, 47, 85–6, 97–8, 113, 132, 162–3, 205, 211 World War I 21, 23, 107; post- 106 World War II 25, 79–82, 134–5, 155, 157, 171; Eastern Front 55, 65n64, 97, 137–8, 204, 207, 210; post- 171, 228; see also Holocaust; Nazi Germany xenophobia 29, 72

Yazidi see Êzidî young men 57, 96–8, 107 Yugoslav Army (JNA) 99 Yugoslavia (former) 22, 55, 74–5, 199; see also Bosnian Genocide Yuon (Vietnamese) 83 Zimbardo, Philip 103–4, 181