Civilian Protective Agency in Violent Settings: A Comparative Perspective 0192866710, 9780192866714

Citation preview

Civilian Protective Agency in Violent Settings

Civilian Protective Agency in Violent Settings A Comparative Perspective Edited by

J A N A K R AUSE JUA N M A SU L L O E M I LY PA D D O N R H O A D S AND J EN N I F ER W EL SH

Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © the several contributors 2023 The moral rights of the authors have been asserted All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2023935113 ISBN 978–0–19–286671–4 DOI: 10.1093/oso/9780192866714.001.0001 Printed and bound by CPI Group (UK) Ltd, Croydon, CR0 4YY Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

Preface Stathis N. Kalyvas

Notwithstanding the large military operations launched by superpowers during the past thirty years, from the United States invasion of Kuwait in 1991 and Iraq in 2003 to the more recent invasion of Ukraine by Russia in 2022, most contemporary warfare takes the form of civil conflict, erupting in postcolonial and very poor, typically ‘bottom-billion’ countries. The form and intensity of these conflicts vary considerably, yet three powerful assumptions recur about them—let us call them ‘the big three’. Civil wars are seen, first, as leading to widespread disorder akin to generalized chaos that defies logic; second, as producing massive and gratuitous violence that can also hardly be understood; and third, as reducing civilians to victims and little else. These assumptions are not completely inaccurate, yet they are not totally true either: they lack in detail and, especially, nuance. To wit, civil wars often lead to the creation of new forms of local order—sometimes forms of order that can be institutionally superior to those they replaced (as, for example, shown by Svensson and Smart in this volume); their violence tends to follow underlying logics and scripts that lend life a degree of predictability that might make it somewhat bearable and make possible systematic analysis; and despite their awful character, they still allow the people who suffer their impact some space for choice and, thus, agency. This latter dimension is the focus of this volume, and it has important implications for the way we understand and study conflict. The contributors to this volume approach the question of civilian agency with rich, field-based studies that provide empirical flesh for a novel concept: civilian protective agency. The term describes a broad set of actions carried out by individuals and communities seeking to cope with the manifold challenges of conflict. Civilians try, and sometimes succeed, in protecting themselves and their communities from the type of harm that would otherwise land on them. This concept is important because it challenges and forces us to rethink the ‘big three’: civil conflicts generate some forms of order, not just generalized chaos and disorder; produce violence that can be at times predictable; and offer civilians some reaction options that might mitigate their harmful impact. As it turns out, civilians can, and do, many things when faced with violent conflict. They can, of course, flee; but they might also mount resistance against armed actors; or they can just adapt to the situation they find themselves in from tacit everyday non-cooperation à la Weapons of the Weak to taking advantage of the situation to extract local or personal benefits. They can use violence or opt for a

vi

PREFACE

non-violent way of action—as well as many behaviours in between. These broad categories of action vary in terms of their actual form; they also vary across space and time, thus offering a broad canvas for investigation. It is possible to record these actions, in both rich descriptive ways and in a more abstract, measurable format. Once this research agenda develops, it will be possible to begin investigating the drivers of this variation in both subnational and cross-national ways, including the ways in which civilians realize the options they have, what they do with them, whether they are successful or not, and how their actions impact on the conflict itself but also its aftermath, for instance on social relations and social structures, as well as political loyalties, affiliations, and behaviour. In short, a rich and multifaceted research agenda lies ahead. The volume’s coverage is remarkably broad following at least three distinct criteria. First, its geographic focus is unusually expansive, covering several world regions: sub-Saharan Africa (Central African Republic, Mozambique, Democratic Republic of the Congo, Nigeria), the Middle East and North Africa (Syria, Iraq), Asia (Myanmar), the Americas (Mexico, El Salvador, Colombia), and even Europe (the Netherlands). Second, its coverage incorporates a large number of repertoires and dimensions of civilian action. Third, and perhaps most importantly, the editors and contributors try hard to apprehend the issue of civilian agency across different categories of political violence. Although most contributions focus on civil war settings, they extend beyond them, shedding light on inter-communal violence, organized criminal violence, and genocide among others. Lastly, some thoughts about the concept of protective civilian agency itself. There may be an inclination to assume that protection is necessarily and always a good thing for civilians and that civilian agency is also automatically a positive thing. However, this is not necessarily the case. First, protection is a term that can be deployed in a way that is threatening (as in ‘protection racket’), or at least in a way that reproduces and inflates malign behaviours (like those highlighted by Judith Verweijen in this volume). Second, there is nothing automatically benevolent about civilian agency. For one, civilians might support competing armed factions, thus wrecking local coexistence. For another, they make take up arms as full-time combatants (thus ceasing to be civilians) or act in a variety of nonfull-time combatant or non-combatant roles, with similar effects. Furthermore, they may provide information to combatants leading to an escalation of violence. Lastly, even when acting as an independent local militia or just local vigilantes, they might become embroiled in local militarized conflicts, either connected or unrelated to the surrounding conflicts, thus making things worse, as suggested by Corinna Jentzsch’s and Eduardo Moncada’s contributions. These are just but a few examples of situations in which civilian agency exercised with an initial eye toward protection broadly understood, may turn out to have counterproductive effects for the civilian population. This is just to highlight a fundamental, if often forgotten, fact about violent conflict: its transformational

PREFACE

vii

nature, its capacity not just to impact behaviour but to alter the identities of actors involved in it, civilian, combatant, and everything in between. Understanding the conditions under which civilian protection turns out to be beneficial for civilians is essential. Which is why we need more research like the rich, field-based work that is showcased in this volume.

Acknowledgements The editors of this book have all sought, in different ways, to understand how civilians act in situations of violence and conflict and how they relate to other important conflict players—locally, nationally, and internationally. The volume builds on several years of collaboration among us but also with our chapter authors. Some of the contributions grew out of an initial workshop held at the European University Institute in 2018, as part of the European Research Council-funded project, ‘The Individualization of War’. In addition, we hosted two workshops dedicated to the project—one in person at the Centre for International Peace and Security Studies at McGill University in January 2020, co-sponsored by the University of Amsterdam, and the other virtually in October 2020. We are indebted to the contributors for their active participation, excellent work, and unwavering dedication to the project during what has been a difficult and unusual time. In addition to the contributors to the volume, we are grateful to the following individuals who participated in the workshops and/or provided helpful input on the project: Oliver Kaplan, Jonny Thakkar, Dirk Druet, Megan Bradley, Adam Kochanski, Nicholas Salter, Rebecca Sutton, Miriam Bradley, Anastasia Shesterinina, Kate Pincock, Alexander Betts, Carla Suarez, and Betcy Jose. Stathis Kalyvas deserves special thanks for writing the preface and for inspiring a generation of scholars interested in civil war and political violence, many of whom are included in this volume. We would also like to acknowledge the ongoing support of our host institutions, University of Oslo, Leiden University, McGill University, and Swarthmore College as well as the financial support that this project received from the Folke Bernadotte Academy (Grant Agreement no 19-00186-1), the European Research Council Starting Grant (Agreement no 852816), the European Research Council Advanced Grant (Agreement no 340956), the Netherlands Institute for Advanced Study in the Humanities and Social Sciences (NIAS), Lang Center for Social and Civil Responsibility and Constance Hungerford Fund at Swarthmore College, and the Centre for International Peace and Security Studies at McGill University. Our work for this volume benefited from the help of several resourceful and brilliant research assistants: Ender McDuff, Lea Baro, Ava Shafiei, Zachary Zimmerman, Emma Klein, Addison Clifford, and Eaindray Taing. Special thanks go to Graydon

ACKNOWLEDGEMENTS

ix

Davidson who went above and beyond in helping to prepare the chapters for publication. Finally, we are grateful to Dominic Byatt at Oxford University Press for supporting this project as well as the peer reviewers for very helpful suggestions and feedback. Jana, Juan, Emily, and Jennifer

Contents List of Figures List of Tables List of Contributors

xii xiii xiv

Civilian protective agency: An introduction Jana Krause, Juan Masullo, and Emily Paddon Rhoads

1

PA RT O N E F O R M S O F C I V I L I A N P R O T E C T I V E A GEN C Y A C R O S S V I O L EN T SET T I N G S 1. Interregional networks and conventionalized gender roles in civilian resistance against genocide: Evidence from the Holocaust in the Low Countries Robert Braun and Kiran Stallone

25

2. The ties that bind: Civilian adaptation and social connectedness during the Syrian Civil War Kimberly Howe

44

3. Transformations in occult protection amid violence in the Central African Republic Louisa Lombard and Mobito Kozaga

63

4. Civil resistance against jihadists: Perceptions of, and resistance against, IS governance in Mosul Isak Svensson and Alanna Smart

80

5. The durability of resistance: Women’s high-risk mobilization for gender justice along the continuum of violence Julia Margaret Zulver 6. Protection as a spectrum: The different faces of protection in eastern Democratic Republic of the Congo Judith Verweijen

98

115

CONTENTS

xi

PA RT T W O V I O L EN T F O R M S O F C I V I L I A N P R O T E C T I V E A GEN C Y 7. Civilian violent mobilization and the intensity of civil war in Mozambique Corinna Jentzsch

135

8. Vigilantism as civilian protective agency: The case of autodefensas in Mexico Moshe Ben Hamo Yeger and Juan Masullo

152

9. Trajectories of civilian resistance to criminal victimization in El Salvador and Nigeria Eduardo Moncada

171

PA RT T H R EE T H E R O L E O F E X T ER N A L A C T O R S 10. Contingent civilians: Agency and action in mass atrocity contexts Zachariah Mampilly and Daniel Solomon 11. Civilian protection monitoring in war and ceasefire contexts: Evidence from Myanmar’s Kachin and Karen States Jana Krause 12. United Nations peacekeeping and civilian protective agency Emily Paddon Rhoads and Aditi Gorur The extraordinary actions of ordinary people: Concluding reflections on civilian protective agency Jennifer M. Welsh Index

191

210

227

244 261

List of Figures Figure 1.1. (a) Interregional rescue networks and (b) Jewish evasion in Holocaust Netherlands per municipality

29

Figure 1.2. The effect of increasing variables one standard deviation on Jewish survival with 90 and 93% confidence intervals

30

Figure 1.3. Interregional and intraregional ties by gender of the rescuer

31

Figure 1.4. The effect of increasing variables one standard deviation on Jewish survival with 90 and 95% confidence intervals

32

Figure 4.1. Perspectives on IS governance

84

Figure 4.2. Perceptions of IS members

85

Figure 4.3. Distribution of civil resistance actions and their specific actions

87

Figure 7.1. Typology of militias

139

Figure 7.2. Number of violent incidents in Zambézia province, 1976–94, based on national newspaper sources (N = 175) Source: Weinstein 2007

147

Figure 7.3. Number of battles between government and rebel forces in Zambézia province, 1978–93, based on provincial and district government reports (N = 262) Source: author’s coding, see appendix

147

Figure 7.4. Number of incidents of violence against civilians in Zambézia province, 1978–94, based on provincial and district government reports (N = 262) Source: author’s coding, see appendix

148

Figure 8.1. Number of vigilante groups per state, 2013–17

155

Figure 8.2. Expansion of vigilante groups per municipality 2013–17

156

Figure 8.3. Predicted probabilities of vigilante groups by cartel presence

163

Figure 8.4. Predicted probabilities of vigilante groups by homicide rate

163

Figure 9.1. Two pathways to predation in vigilante groups

176

List of Tables Table 8.1. Logistic regression of vigilante presence in Mexican municipalities, 2013–17 Table 10.1. Summary of case evidence about three explanations for civilian agency

161 200

List of Contributors Robert Braun is Assistant Professor of Sociology and Political Science at UC Berkeley. He studies intergroup relationships in times of social upheaval and is currently working on a book project that traces racist themes in children's stories throughout 20th century Central Europe. Aditi Gorur is a senior policy advisor at the United States Mission to the United Nations, and previously served as director of the Protecting Civilians in Conflict Program at the Stimson Center. Her research has included work on the protection of civilians by UN peacekeepers, civilian self-protection measures in conflict areas, and host-state consent for UN peacekeeping missions. She participated in this publication in her individual capacity; it reflects her personal views and does not represent the views of the U.S. government. Moshe Ben Hamo is a DPhil candidate in the Department of Politics and International Relations at the University of Oxford. Before joining Oxford, he studied and conducted research in Mexico (ITAM) and Switzerland (IHEID, and worked for the Mexican Ministry of Social Development. He specializes in the intersection between criminal violence and politics in Latin America. For his doctoral dissertation, he is exploring social responses to large-scale criminal violence in Mexico. His research is supported by Mexico’s National Council for Science and Technology (Conacyt), and Oxford’s Department of Politics and International Relations. He is currently a research fellow at the Center for U.S.-Mexican Studies (USMEX) at UC San Diego. Kimberly Howe is Assistant Research Professor at the Friedman School and Research Director of Conflict and Governance at the Feinstein International Center at Tufts University. She has conducted conflict-focused research in over a dozen countries and regularly provides consultation to governments, international organizations and NGOs on their humanitarian, development and stabilization polices and initiatives. Corinna Jentzsch is Assistant Professor (with tenure) of International Relations at the Institute of Political Science at Leiden University, and author of Violent Resistance: Militia Formation and Civil War in Mozambique (Cambridge University Press, 2022), which examines how and why community-based militias form and spread. Her research focuses on civilian collective action during civil war and conflict transformation and escalation. She is also broadly interested in social movements, political violence, peacekeeping, African politics, and fieldwork methods and ethics. She has conducted extensive fieldwork in southern Africa, including Mozambique, Zambia, and Malawi. Mobito Kozaga is a PhD candidate in Anthropology at the University of Bangui where he taught from 2013 to 2021. He holds two master’s degrees in Sociology and Anthropology. He has conducted several ethnographies in difficult terrain, in conflict areas in the Central

LIST OF CONTRIBUTORS

xv

African Republic, and notably in Bambari from 2014 to 2022. He is author of several articles on violence and difficult terrain. Jana Krause is Professor in the Department of Political Science at the University of Oslo. She is author of Resilient Communities: Non-Violence and Civilian Agency in Communal War (Cambridge University Press, 2018) and directs the research project Resilience Building: Social Resilience, Gendered Dynamics, and Local Peace in Protracted Conflicts (2020-2026), funded by the European Research Council. Louisa Lombard is Associate Professor of Anthropology at Yale University and a member of the editorial collective of HAU: The Journal of Ethnographic Theory. She is the author of State of Rebellion: Violence and Intervention in the Central African Republic and Hunting Game: Raiding Politics in the Central African Republic. She is currently researching how military peacekeepers confront ethical dilemmas in their work. Zachariah Mampilly is the Marxe Endowed Chair of International Affairs at the Marxe School of Public and International Affairs, CUNY, a member of the doctoral faculty in the Department of Political Science at the Graduate Center, CUNY, and co-founder of the Program on African Social Research. He is the author of Rebel Rulers: Insurgent Governance and Civilian Life during War and, with Adam Branch, Africa Uprising: Popular Protest and Political Change. He is the co-editor of Rebel Governance in Civil Wars with Ana Arjona and Nelson Kasfir; and Peacemaking: From Practice to Theory with Andrea Bartoli and Susan Allen Nan. He has held fellowships with the Institute for Advanced Study (New Jersey), the Open Society Foundation, the Harry Frank Guggenheim Foundation, the United States Holocaust Memorial Museum, and the Fulbright Program. Juan Masullo is an Assistant Professor (with tenure) at the Institute of Political Science at Leiden University. His research combines multiple methods, often involving extensive fieldwork, to study political and criminal violence. He is particularly interested in how individuals and communities respond to violence to protect themselves, and how experiences with violence shape political attitudes, including preferences for how to best deal with violence and crime. He has explored these issues in different countries affected by civil war and organized crime, including Colombia, Mexico, Brazil, and Italy. Eduardo Moncada is Associate Professor of Political Science at Barnard College, Columbia University. His research agenda focuses on the politics of crime and violence in Latin America, urban politics, and subnational analysis. He is the author of Resisting Extortion: Victims, Criminals, and States in Latin America (Cambridge University Press, 2021), Cities, Business and the Politics of Urban Violence in Latin America (Stanford University Press, 2016), and co-editor of Inside Countries: Subnational Analysis in Comparative Politics (Cambridge University Press, 2019). Moncada’s work has received support from the Harry Frank Guggenheim Foundation, the Ford Foundation, and the American Council of Learned Societies, among others. Emily Paddon Rhoads is Associate Professor of Political Science at Swarthmore College. She is author of Taking Sides in Peacekeeping: Impartiality and the Future of the United Nations (Oxford University Press, 2016) as well as several academic journal articles on civilian agency in armed conflict, peacekeeping, humanitarianism, and the United Nations.

xvi

LIST OF CONTRIBUTORS

She has conducted research in the Democratic Republic of the Congo, South Sudan, Central African Republic, Kenya, and Iraq. Her research has been supported by the European Research Council, Trudeau Foundation, Canadian Department of Foreign Affairs and International Trade, and the Swedish International Development Cooperation Agency. She is an elected fellow of the Rift Valley Institute and co-founder of the Oxford Central Africa Forum (OCAF). Alanna Smart has been a research assistant at the Department of Peace and Conflict Research, Uppsala University, Sweden, where she worked on projects related to armed conflict, civil resistance, peace processes, and Islamist armed groups. She holds an MSSc in Peace and Conflict Studies from Uppsala University and a BA in Politics from Durham University, UK, where she specialised in Middle Eastern politics, political Islam, and asymmetric armed conflicts. Her research interests also include rebel fragmentation, group preference change, and radicalisation/extremism. Prior to her work in academia, she worked for women’s rights and anti-corruption NGOs in Malaysia. She is currently working as an Embedded Intelligence Analyst for Sibylline, a strategic risk consultancy, based in London, UK. Daniel Solomon is a PhD candidate in the Department of Government at Georgetown University and Associate Research Fellow at the US Holocaust Memorial Museum's SimonSkjodt Center for the Prevention of Genocide. His research has been published in Comparative Politics, Genocide Studies and Prevention, and Qualitative and Multi-Method Research. Kiran Stallone is a PhD Candidate in the Sociology Department at UC Berkeley. Her dissertation examines gendered civilian agency in times of war, with a focus on Colombia. Isak Svensson is Professor at the Department of Peace and Conflict Research, Uppsala University, Sweden. His research focuses on religious dimensions of armed conflict, dynamics of nonviolent civil resistance, and international mediation in civil wars. His works appear in leading journals such as International Studies Quarterly, Journal of Peace Research, Journal of Conflict Resolution, and European Journal of International Relations. Svensson has been the Director of Research, The National Centre for Peace and Conflict Studies (NCPCS), University of Otago, New Zealand (2010–2012), and has served as the at-large Representative in the Governing Council, for the International Studies Association (ISA) in 2015–2016. His most recent book (co-authored) is Confronting the Caliphate: Civil resistance in Jihadist Proto-States (Oxford University Press, 2022). Judith Verweijen is an Assistant Professor in International Relations at the University of Groningen. She previously held positions at the Universities of Sheffield, Sussex and Ghent and at the Nordic Africa Institute. Judith’s research examines the interplay between violence, conflict, and armed mobilisation; micro-level civil–military and civil–rebel relations; and the militarisation of conflicts around natural resources. She focuses on eastern Democratic Republic of the Congo, where she has conducted extensive field research since 2010. She currently leads a research project on the drivers and effects of burning down houses as part of warfare in eastern DRC.

LIST OF CONTRIBUTORS

xvii

Jennifer M. Welsh is the Canada 150 Research Chair in Global Governance and Security at McGill University and Director of the Centre for International Peace and Security Studies. She was previously Chair in International Relations at the European University Institute and Professor in International Relations at the University of Oxford, where she co-founded the Oxford Institute for Ethics, Law and Armed Conflict. From 2013 to 2016, she served as the Special Adviser to the UN Secretary General, Ban Ki-Moon, on the Responsibility to Protect. She has published several books and articles on the ethics and politics of armed conflict, the ‘responsibility to protect’, humanitarian action and civilian protection, the UN Security Council, and Canadian foreign policy. Her 2016 book, The Return of History: Conflict, Geopolitics and Migration in the 21st Century, was based on her CBC Massey Lectures. Prof. Welsh is a member of the Internally Displaced Persons Protection Expert Group, established by UNHCR and the Global Protection Cluster, and sits on the Advisory Boards of the Global Centre for the Responsibility to Protect and the Auschwitz Institute for Peace and Reconciliation. In 2021, she was elected an International Honorary Member of the American Academy of Arts and Sciences and in 2022 became a Fellow of the Royal Society of Canada. Julia Zulver is a Marie Skłodowska-Curie Research Fellow at the University of Oxford and the Universidad Nacional Autónoma de México. Her research focuses on women’s high-risk leadership and mobilization in Latin America. Her book High-Risk Feminism in Colombia was published by Rutgers University Press in 2022, and in Spanish by Ediciones Uniandes in 2022.

Civilian protective agency An introduction Jana Krause, Juan Masullo, and Emily Paddon Rhoads

More than half the world’s population live in violent settings, such as civil wars, communal conflicts, cities plagued by gang violence, and entire areas governed by criminal organizations (United Nations 2020). Living exposed to diverse forms of violence, individuals and communities have found innovative—and sometimes counterintuitive—ways to protect themselves and others. A burgeoning body of research has developed to study the phenomenon of civilian agency and its implications for protection. Distinct research streams on civil war and rebel governance, communal violence and resilience, genocide and rescue behaviour, urban crime and criminal governance have all produced new insights about civilian agency and civilian self-protection. Yet knowledge on the topic remains highly fragmented, as it has developed within the confines of specialized subfields (e.g. micro-dynamics of civil war, politics of crime, peace studies, Holocaust studies), across several disciplines (e.g. political science, sociology, anthropology, history), and within specific area studies (e.g. Latin America for crime research, South Asia for election and communal violence research). The result is a compartmentalization of knowledge, raising questions about the comparability of findings and their scope conditions. The power of ordinary people to resist various forms of oppression and injustice as well as promote social and political change has been widely researched by students of revolutions, social movements, and, more recently, civil resistance. Although this work has stressed agency and people’s power, and much of it touches directly or indirectly on issues related to protection, a unified and explicit research agenda on civilian protective agency has yet to take shape. Further, academic scholarship is disconnected from major policy debates on the protection of civilians, conflict prevention, and peacebuilding. This is unfortunate, especially at a time when international organizations, non-governmental organizations (NGOs), and donor countries increasingly seek to integrate local protective agency into their programming given the follies of direct intervention. This volume aims to establish the study of civilian agency and its protective dimension across various violent settings as a systematic and unified field of

Jana Krause, Juan Masullo, and Emily Paddon Rhoads, Civilian protective agency. In: Civilian Protective Agency in Violent Settings. Edited by: Jana Krause, et al., Oxford University Press. © Jana Krause, Juan Masullo, and Emily Paddon Rhoads (2023). DOI: 10.1093/oso/9780192866714.003.0001

2

CIVILIAN PROTECTIVE AGENCY: AN INTRODUCTION

research. In doing so, it responds to calls by influential scholars for research that accounts for the similarities and differences of dynamics in the face of different expressions of political violence (Tilly 2003; Kalyvas 2019). As a starting point, we define civilian protective agency broadly as actions carried out by individuals and communities to protect themselves and/or others in violent settings. From here we delineate some important conceptual boundaries and motivate a set of core questions: How does civilian protective agency emerge in different violent settings? What forms does it take and do these differ according to the setting where they emerge? How does civilian protective agency impact the dynamics of violence in different settings? What are its consequences for conflict processes, governance arrangements, and sustainable peacebuilding? To engage with these questions, the volume brings together researchers spanning several social science disciplines and relying on a wide variety of research methods for data collection and analysis, including ethnography, interviewing, survey research, archival research, and statistical analysis. The contributors study civilian protective agency in different violent settings, including civil war, genocide, communal violence, and organized crime, and in various geographical world locations, from Syria to Mozambique, Sri Lanka to Mexico, Iraq to Colombia, and Western Europe. The volume offers conceptual foundations, new theoretical insights, and detailed empirics that advance our understanding of civilian protective agency and promote future research on the topic that is comparable, tractable, and cumulative. In what follows, we provide an overview of the origins of this diverse field of scholarship and practice, and we sketch the conceptual terrain. We define civilian protective agency, examine its constituent components (‘civilian’, ‘protection’, and ‘agency’), distinguish it from cognate terms, and situate the concept within the broader literature. We then review the various ways that civilian protective agency emerges as well as its consequences in terms of protection. Finally, we provide an overview of the volume.

1. Origins of a diverse field of scholarship Individuals and communities have always played an active role in their own survival and protection in the face of violence. And yet, with relatively few exceptions, their agency was long neglected by social scientists as well as the range of international organizations and NGOs that operate in violent settings. That has begun to change. In recent years scholars of civil war, building on the foundations laid by the anthropology of conflict and violence (e.g. Kriger 1991; Stoll 1993; Nordstrom 1997; Lubkemann 2008), have increasingly paid attention to civilian agency in conflict and post-conflict settings. This has been a remarkable turn from a previous generation of scholarship that either focused on macro-level processes

JANA KRAUSE, JUAN MASULLO, AND EMILY PADDON RHOADS

3

(e.g. civil war onset, duration, termination) in large-N cross-country analysis, or on meso-level dynamics with a dyadic understanding of conflict as a contest almost exclusively between non-state armed actors and the state. Both perspectives left little to no room for considering civilians as agents in their own right. While civilians have featured in studies of rebel recruitment and mobilization (e.g. Wood 2003; Weinstein 2007), Kalyvas established the centrality of civilian agency on the dynamics of civil war violence. He argued that armed actors and civilians jointly produce violence and showed that civilian actions (such as collaboration and denunciation) can shape both types and levels of civilian victimization (Kalyvas 2003, 2006). This insight was articulated in a model aimed at explaining violence against civilians and mostly stressed civilians’ ability to manipulate widespread violence for private purposes (Kalyvas 2006 p. 386). Yet, it implied that what civilians do has the power to alter violence and related war dynamics. Expanding on this insight, scholars have shown that civilian organization can limit levels of civil war violence (Kaplan 2017), influence the capacity of armed groups to control territory (Rubin 2019; Jackson 2021), and shape the nature and scope of rebel governance (Mampilly 2011; Arjona 2016; Breslawski 2020). Similar developments have occurred in related fields where scholars have analysed violent and nonviolent civilian agency in violent settings other than civil war. Scholars of communal conflict, for example, have showcased that civilian action can effectively prevent violence from emerging and escalating in areas surrounded by communal violence (Varshney 2003; Berenschot 2011; Carpenter 2012; J. Krause 2018; Dhattiwala 2019; Klaus 2020). Similarly, moving away from conventional understandings of episodes of mass violence as totalizing and unaccommodating, genocide researchers have established that even in extremely violent contexts such as the Holocaust and Rwandan Genocide, civilians retain some agency in the face of extermination campaigns. This work has examined the individual and collective self-protective actions of those persecuted, as well as the efforts of some to save others from harm despite enormous risks to themselves (Oliner 1992; Tec 1995; Fujii 2009; Maher 2010; Einwohner and Maher 2011; Monroe 2011; Luft 2015; Finkel 2017; Braun 2019). Looking beyond settings not formally designated as ‘armed conflict’, students of organized crime have also examined how civilians respond to different forms of criminal violence and governance (Fahlberg 2018; Ley et al. 2019; Osorio et al. 2021; Moncada 2022; Barnes forthcoming). Civilian agency has also gained attention within policy domains such as humanitarian relief, peacebuilding, and refugee protection, where various forms of ‘localization’, aimed at supporting individual and community agency, have become a priority for a host of international actors (Obrecht 2014; Gingerich and Cohen 2015; Wallace 2016; Julian 2020; Pincock et al. 2020; Kaplan 2021; McQuinn et al. 2021). With increasing recognition that violence in armed conflict represents only a fraction of violence worldwide, some of these same international actors have pursued similar activities in what the International Committee of the Red Cross

4

CIVILIAN PROTECTIVE AGENCY: AN INTRODUCTION

(ICRC) refers to as ‘situations other than war’ (Geneva Declaration 2015; Bradley 2020). Relatedly, peace researchers, following this ‘local turn’, have increasingly stressed how everyday civilian practices contribute to managing conflicts less violently and fostering peace locally (Mac Ginty 2014, 2021; Firchow 2018), as well as carving out pockets of peace in the midst of ongoing violence (Hancock and Mitchell 2007; Autesserre 2021).

2. Mapping the conceptual terrain We define civilian protective agency as actions carried out by individuals and communities to protect themselves and/or others in violent settings. We begin purposefully with a broad and inclusive definition that aims at identifying the concept’s four core attributes: the agent (civilians), the goal and action (protection), the beneficiary of the action (either the agents themselves or others), and the context in which the action takes place (violent settings). This definition builds on previous research that conceptualized protective civilian agency as tactical, strategic, and rescue agency, arguing that civilians’ time horizon and social knowledge shape protective action (J. Krause 2017). We provide a more comprehensive conceptualization of civilian protective agency that includes both nonviolent and violent practices and is based on a broader understanding of protection. In what follows, we further define the core constituent components of the concept (‘civilian’, ‘protection’, and ‘agency’), which together cover the four core attributes listed above. We discuss relevant demarcations and identify some of the empirical manifestations of civilian protective agency.

Who is a civilian? Defining the term ‘civilian’ has moral, legal, and political implications. As with most (if not all) existing definitions of civilian, empirical challenges are likely to emerge when it comes to distinguishing who is, and who is not, a civilian ‘on the ground’, given that in many contemporary violent settings frontlines are blurred, and armed actors hide amongst the population. Consistent with scholarship in conflict research, we adopt a definition based on International Humanitarian Law (IHL) that distinguishes between ‘combatant’ and ‘non-combatant’. Under Article 50 of Additional Protocol I, the civilian is defined in a negative manner as anyone who is not a combatant (Slim 2016; Williamson 2016), which implies that a civilian is not a recognizable member of an organized armed group with a clear command structure. To further distinguish combatants from civilians, civil war scholars have relied on the criterion of ‘full-time membership’ in state and non-state armed

JANA KRAUSE, JUAN MASULLO, AND EMILY PADDON RHOADS

5

organizations (Kalyvas 2006 p. 19; Steele 2009). Steele (2009 p. 421), for example, defines civilians as ‘individuals who do not participate in the military activities of any armed group, but who may be “part-time” affiliates or collaborators’. Related, conceptual and empirical work by scholars of different types of violent settings have shown that civilians (and civilian protective agency) can be both unarmed and armed (Finkel 2015, 2017; Jose and Medie 2015; J. Krause 2018; Bateson 2021; Masullo 2021; Schubiger 2021; Jentzsch 2022; Moncada 2022). We follow these two leads, allowing the concept of civilian protective agency to include both unarmed and armed individuals, so as to explore the protective role of part-time armed actors, such as community-initiated militias (Jentzsch within volume) and vigilante groups (Ben Hamo and Masullo within volume; Moncada within volume). Yet contributions in this volume (see chapter by Jentzsch) problematize this distinction further, noting that for militia members to be understood as armed civilians, the militia must be part-time and community-initiated (as opposed to state-created). While this definition makes for a largely inclusive conception of civilians, we exclude external protection actors, such as humanitarian actors or military peacekeepers. Several chapters in the volume account for these actors’ interactions with civilians and, in some cases, their support for civilian protective agency. They are not, however, the focus of the volume given their status as external actors. Civilian protective agency is about local actors exercising agency for the purposes of protection.

What is protection? According to the Oxford English Dictionary, protection is ‘[t]he action of protecting someone or something; the fact or condition of being protected; shelter, defence, or preservation from harm, danger, damage, etc.; guardianship, care; patronage’ (OED n.d.). As such, protection can be an action, a condition or even, as in the case of paid protection, a commodity. Later in this section we detail the forms of action that fall within ‘protection’. The key question here is then: protection from what? We adopt a broad conceptualization of protection to accommodate a wide range of threats and harms that give rise to protective action in violent settings. While the literature on wartime civilian agency has mostly focused on protection from physical violence by armed groups,¹ we follow the lead of scholars that recently ¹ For example, early work defined ‘civilian self-protection’ as strategies adopted by civilians to ‘avoid, mitigate or thwart violence by armed groups’ (Baines and Paddon 2012 p. 234). Similarly, a later key conceptual contribution referred to ‘actions taken to protect against immediate, direct threats to physical integrity imposed by belligerents or traditional protection actors’ (Jose and Medie 2015 p. 516).

6

CIVILIAN PROTECTIVE AGENCY: AN INTRODUCTION

have called for a more holistic concept of protection, recognizing that the indirect effects of conflict are amongst the most harmful to civilians (K. Krause 2016; Wise 2017). This involves attending to individual actors’ perceptions of their circumstances and the phenomena that affect their lives (Baines 2016; Paddon Rhoads and Sutton 2020). In doing so, harms that often cannot be attributed to the actions of armed groups and that do not derive directly from the conflict are included in understandings of protection (Nordstrom 1997; Carbonnier 2010; Gorur and Carstensen 2016). As such, alongside protection from direct and physical violence, the volume features research on protection from indirect and less physical threats, such as disruption in livelihoods and inadequate access to health care, food, shelter, and other basic necessities (Howe within volume), as well as spiritual threats and occult war practices (Lombard and Kozaga within volume). In widening the conceptual aperture, our approach is in step with feminist and humanitarian conceptions of protection as extending ‘beyond physical assistance to the protection of a human being in their fullness’ (Slim and Bonwick 2006 p. 23). According to the Inter-Agency Standing Committee (IASC) on humanitarian assistance, protection ‘encompasses all activities, aimed at obtaining full respect for the rights of the individual in accordance with the letter and spirit of the relevant bodies of law (i.e. human rights, humanitarian law, and refugee law)’ (ICRC 2001 p. 19).² While a broad conception captures a spectrum of threats and potential harms, it is not without challenges for scholars and practitioners alike. As Paddon Rhoads and Gorur (within volume; see also Paddon Rhoads and Sutton 2020) show, competing understandings of what constitutes a threat and how protection should be pursued exist within and across communities and in relation to international actors. As Krause underscores in her chapter on Myanmar, civilian protective agency is political agency and as such it is contested (Krause within volume).

What is ‘agency’ in civilian protective agency? A comprehensive understanding of agency is fundamental to establishing civilian protective agency as a distinct and thriving research field. We adopt an expansive conception of agency as the capacity to engage with and shape the violent settings and circumstances in which people find themselves. Before exploring the various forms that civilian protective agency can take on the ground, we briefly examine the rich sociological and anthropological literature on agency which informs our conception. We distinguish different theoretical perspectives ² ISAC is forum for coordination, policy development and decision-making involving the key UN and non-UN humanitarian partners.

JANA KRAUSE, JUAN MASULLO, AND EMILY PADDON RHOADS

7

and their implicit assumptions, teasing out the implications for analysing civilian protection in violent settings. Sociologists have referred to agency as the ‘existential capacity for exerting influence on our environments’ (Hitlin and Elder Jr 2007). A prominent conceptualization of agency refers to a temporally embedded process of social engagement— informed by the past but oriented towards the present and the future—shaped by the interplay of habit, imagination, and judgement (Emirbayer and Mische 1998). Agency is exerted differentially depending on the actor’s salient time horizon (Hitlin and Elder Jr 2007). Our conception stresses agency’s goal-seeking, projective, and purposive dimension³ because the disruptive context and dangers of violent settings represent circumstances where habit and pre-established routines may quickly fail to guide protective actions. Social psychologists theorize agency as rooted in an individual’s socialization and sense of self, including their perceived interconnectedness with others, moral beliefs, and worldviews. Early research into civilian agency during the Holocaust was informed by this work (Oliner 1992; Tec 1995; Monroe 1998, 2011). For example, to explain rescue behaviour during the Holocaust, Monroe (2011) argued that ‘social perspective’ determines ‘the menu of behavioural options’ actors see as available to themselves. In contrast, social anthropologists have proposed theorizing agency through the prism of ‘social navigation’, analysing how civilians navigate complex and often fast-changing environments (Utas 2005; Vigh 2009; Verweijen 2018). This approach emphasizes the fluidity of both individual agency and structural contexts over a more durable individual moral worldview and sense of self in guiding civilian action. Recent research has conceptualized civilian agency as fluid along a spectrum ranging from rescue behaviour to bystanding, evading, and participating in killing (Fujii 2009). Both the sociological and the anthropological perspectives create room for taking civilian meaning-making seriously. In Emirbayer and Mische’s (1998) conceptualization of agency’s projective dimension, projectivity ‘encompasses the imaginative generation by actors of possible future trajectories of action . . . in relation to [their] hopes, fears, and desires for the future’. In Vigh’s (2009) social navigation approach, ‘people invest a great deal of time in making sense of and predicting the movement of their social environment, in clarifying how they are able to adapt to and move in relation to oncoming change’. Meaning-making, imagination, and anticipation of potential threat scenarios can determine whether and how civilians effectively protect themselves and others in violent settings (J. Krause 2018), as chapters by Verweijen, Lombard and Kozaga, and Braun and Stallone within this volume highlight.

³ Note that this does not imply embracing a rational action-based conception of agency. Agency can be purposive and be driven by a multiplicity of considerations, of which rationality is only one.

8

CIVILIAN PROTECTIVE AGENCY: AN INTRODUCTION

Research on agency and intersectionality has received less attention in studies of civilian agency than in other fields (notable exceptions include works by Tec 2004; Bonkat 2014; J. Krause 2019; Zulver 2022). Contributors to our volume advance knowledge on one fundamental but often neglected aspect of social identity—gender—with chapters on women’s mobilization against violence (Zulver within volume), women’s rescue agency (Braun and Stallone within volume), and women’s contribution to civilian protection monitoring (Krause within volume). As the Conclusion notes, disaggregating civilian protective agency based on race, ethnicity, sexuality, age, socio-economic status, and (dis)ability are areas ripe for further investigation.

Forms of civilian protective agency across violent settings We encounter individual and collective civilian protective agency in various manifestations across different violent settings. In what follows, we distinguish four inter-related yet distinct forms of civilian protective agency: evasion, resistance, rescue, and adaptation. Evasion includes actions to escape the reach of armed actors to avoid immediate danger, recruitment, control, and/or persecution (Jose and Medie 2015). It can take the form of daily commuting, as in Uganda, at the height of the civil war, when thousands of children and youth walked to town centres and cities at night to avoid abduction by rebels (Baines and Paddon 2012). Yet, it can also involve longerterm flight, a more permanent departure from one’s home, akin to what Barter (2014) called ‘exit’ building on Hirschman’s ‘exit, voice and loyalty’ framework. As in various civil wars (Steele 2017; Schon 2020), flight was a survival strategy available to some ordinary Jews living in ghettos during the Holocaust (Finkel 2017). Resistance implies the refusal to act according to implicit and explicit demands by armed groups. This may include acts to oppose the imposition of certain ideologies and to deny, or at least mitigate, demands for material and non-material resources, such as information, food, shelter, and taxes. Across violent settings, civilians have resisted armed organizations to short-circuit spirals of civil war and communal violence (Kaplan 2017; J. Krause 2018), limit forms of rebel rule (Arjona 2016; Gowrinathan and Mampilly 2019; Jackson 2021), eschew protection rackets and criminal extortion (Moncada 2022), and undermine foreign occupation (Petersen 2001) and extermination campaigns (Maher 2010; Finkel 2017). While not every expression of resistance is necessarily aimed at protection, many documented instances of civilian resistance do involve the goal of protecting oneself and/or others. Forms of resistance with the clear objective of civilian

JANA KRAUSE, JUAN MASULLO, AND EMILY PADDON RHOADS

9

protection include the decision to designate areas off-limits to armed groups and establish peace zones (Hancock and Mitchell 2007; Mouly et al. 2015; Autesserre 2021), engagement with armed groups to negotiate norms of behaviour and terms of coexistence (Kaplan 2017; J. Krause 2018), as well as more specific and targeted efforts, such as the release of kidnapped community members or the creation of humanitarian zones or corridors. Moreover, protective resistance can take an armed form, such as when communities counter armed organizations with violence of their own in settings as diverse as civil wars, criminal conflicts, and genocide (Finkel 2015, 2017; Schubiger 2021; Jentzsch 2022). The distinction between evasion and resistance maps well onto existing categorizations that distinguish between survival and resistance (Schon 2020), nonengagement and engagement (Jose and Medie 2015), and victim’s and oppositional agency (Gowrinathan and Mampilly 2019). Protective agency can also take the form of rescue when people who are not directly targeted by armed groups make efforts to protect others living in their localities who have been singled out for persecution or extermination (Fujii 2009; Monroe 2011; Sémelin et al. 2014; Braun 2019). Rescue is distinct insofar as it captures actions aimed at saving others. Yet, it can also be seen as a form of resistance as it represents a challenge to an armed group’s ideology and orders. Rescue is a particularly risky form of action, as it commonly takes place in contexts of overwhelming military force such as mass persecution, extermination campaigns, and genocide. Lastly there is adaptation, a concept rooted in such diverse literatures as child development and climate change, which increasingly is incorporated into research on civilian agency in conflict settings (Howe within volume; J. Krause 2018). In complexity theory, adaptation is the ‘process whereby an organism fits itself to its environment’ (Holland 1995 p. 9). For civilians in conflict zones, adaptation is based on conflict perception, social knowledge, social learning, and ‘the capacity to imagine alternative futures’ (J. Krause 2018 p. 66). Adaptation often includes deliberately maintaining social identities that sustain hope and enable collective action. It can encompass a wide range of protective responses to the direct and indirect effects of violence, including changes in livelihood practices, education, and access to health care. As such, adaptation is a broad category that encompasses civilian protective actions that can enable and sustain evasion, resistance, and rescue agency.

3. Explaining civilian protective agency As the contributions to this volume attest, research explaining various dimensions of civilian agency across different violent settings has advanced considerably in recent years. Spanning distinct fields and disciplines, this research identifies

10

CIVILIAN PROTECTIVE AGENCY: AN INTRODUCTION

different factors and conditions that enable the emergence of civilian protective agency, as well as some of its consequences for the protection of civilians—what we refer to later in the chapter as the ‘protection dividend’.

Triggering and enabling civilian protective agency To explain why some individuals or communities engage in different forms of civilian protective agency, scholars explore both the forces that trigger people to protect themselves or others in their local contexts and the factors that provide them with the capacity to do so. It is in the context of violence that people engage in civilian protective agency. They do so to survive, to reduce existing levels of harm and in some cases to prevent violence from spreading into new areas. This deeply human impulse is visible across violent settings, including civil war, criminal conflict, communal violence, and mass atrocities. Yet the relationship between the dynamics of violence and civilian protective agency has been qualified in important ways. First, the relationship is not necessarily linear. For example, Ben Hamo and Masullo (within volume) show that armed mobilization against drug cartels in Mexico’s criminal conflict is more likely to take place when violence is neither too low nor too high. Second, the way people are targeted matters. Collective targeting, for example, triggered the emergence of civilian self-defence units during the Peruvian civil war (Schubiger 2021) and non-violent campaigns of civilian non-cooperation in the Colombian conflict (Masullo 2017). Similarly, selective targeting provided ordinary Jews with the skills necessary to mount sustained and organized antiNazi resistance during the Holocaust (Finkel 2015). Finally, violence does not necessarily have to be lethal to trigger civilian protective agency. Zulver (within volume) shows that non-lethal forms of gendered violence in the Colombian civil war incentivized women to mobilize resistance against armed groups (see also Kreft 2019). Similarly, criminal extortion and other non-lethal predatory practices have been a central trigger of community resistance to criminal organizations (Moncada 2022). Territorial control and competition between armed groups are important situational factors that form the incentives and disincentives that civilians consider when engaging in protective agency. Both armed and unarmed resistance to armed groups are more likely when local control of a territory is contested (Masullo 2017; Vu¨llers and Krtsch 2020; Jentzsch 2022). As Jentzsch (2022) shows for the Mozambican civil war, stalemates between competing armed actors can empower communities and lead them to create militias, as stalemates commonly increase violence and reduce armed actors’ capacity to reliably offer protection. Similarly, in contexts of criminal violence, the presence of multiple cartels can create incentives

JANA KRAUSE, JUAN MASULLO, AND EMILY PADDON RHOADS

11

for communities to establish vigilante groups as a self-protection strategy (Ben Hamo and Masullo within volume). Finally, the different levels of territorial control and types of violence exercised by the Nazis across Europe created divergent opportunities for both targets and non-targets to engage in protective agency (Balcells and Solomon 2020). These findings do not, however, preclude instances of civilian protective agency when a single armed group holds tight control of a territory. Excessively intrusive forms of rebel rule can also motivate resistance (Arjona 2015, 2016; Masullo 2017; Gowrinathan and Mampilly 2019; Mampilly 2011; Svensson and Smart within volume). Rather than protecting themselves from violence, civilians may seek to protect themselves from unwelcome systems of rebel governance, directly and indirectly challenging the rule of armed groups. While civilian perceptions of governance are crucial in motivating people to stand against rebel rule, Svensson and Smart (within volume) show that perceptions are an insufficient explanation. Residents of Islamic State (IS)-controlled areas in Iraq protested and engaged in acts of everyday resistance, despite holding favourable attitudes and assessments of IS governance. This highlights the importance of identity, as lack of identification with a group may push people towards resistance. Conflict conditions, including characteristics of armed groups, such as ideology (Sanı´n and Wood 2014), reliance on or sensitivity toward the civilian population (Weinstein 2007; Kaplan 2017; Jackson 2021), internal discipline and political education (Hoover Green 2018), and responsiveness to norms of restraint (Stanton 2016) all shape the space civilians have to engage armed groups to protect themselves or others (J. Krause and Kamler 2022). Beyond conflict conditions and situational factors, organizational capacity is needed to enable (some forms of ) protective action (Arjona 2016; Kaplan 2017; Masullo 2017; J. Krause and Kamler 2022). Mobilizing in the face of violence is challenging. Risks are high and benefits uncertain, armed actors closely monitor civilian behaviour, violence can weaken interpersonal trust and foment social divisions, and civilians often have very limited lead time to organize. Therefore, local community structures are crucial in enabling collective action. In particular, pre-existing communal and indigenous organizations, as well as previous experiences of collective action, have consistently been found as key facilitators of collective forms of protective agency (Varshney 2003; Arjona 2015, 2016; Kaplan 2017; Masullo 2017; Ley et al. 2019; Breslawski 2021). To understand why and how these factors matter, scholars underscore the importance of more specific mechanisms, such as leadership (Peterson 2001; J. Krause 2018; Masullo 2017) and social connectedness (Howe within volume; Schon 2020). As Howe demonstrates in her chapter on Syria, social connections have been critical in enabling both individuals and communities to engage in survival strategies such as flight and adaptation.

12

CIVILIAN PROTECTIVE AGENCY: AN INTRODUCTION

While these findings come mostly from civil and communal war settings, community structures also matter greatly for rescue in situations of mass persecution. For example, local religious minorities were more likely to protect potential victims of Nazi persecution than their majority counterparts, as they were better positioned to set up clandestine networks and they empathized more profoundly with the persecuted (Braun 2019). The gender composition of community structures is also significant for rescue. As evidenced by Braun and Stallone (within volume), during the Nazi occupation of the Netherlands, Dutch women were uniquely positioned to foster inter-regional rescue networks as they could strategically perform traditional feminine roles that allowed them to travel and carry out rescue work undetected.⁴ These findings have recently been complemented by studies of large-scale criminal violence and mass persecution that adopt a longer historical perspective to show that the enabling effect of community structures and past experiences of collective action can persist across generations and conflict periods. For example, experiences of armed mobilization dating back to the early 1900s, left legacies that facilitated the organization of armed self-defence groups amidst Mexico’s current criminal conflict (Sánchez-Talanquer 2018; Osorio et al. 2021). Similarly, the possibility of ordinary Jews resisting the Nazis during the Holocaust depended on organizational legacies left by pre-World War II political regimes and experiences of repression (Finkel 2015, 2017), as well as the readiness of non-Jews to help Jews evade persecution in the Low Countries hinged on the long-lasting impacts of the Reformation and Counter-Reformation (Braun 2018).

Consequences of civilian protective agency and the protection dividend What are the effects of the individual and collective actions that civilians undertake to protect themselves and/or others? There is some evidence that civilian protective agency can result in a positive protection dividend. However, as discussed in the Conclusion to the volume, further research is necessary to rigorously assess the impact of different civilian responses across diverse violent settings. In terms of violence, adaptation and community resilience have been shown to effectively prevent communal violence from emerging and from expanding into new areas (Carpenter 2012; J. Krause 2018). Similarly, local institutional mechanisms designed by civilians to protect themselves, such as peace committees, ⁴ In the context of civil war, concretely in the Sri Lankan case, ideological and cultural dimensions of rebellion have also created some space for certain women to challenge armed groups governing rules and practices (Gowrinathan and Mampilly 2019).

JANA KRAUSE, JUAN MASULLO, AND EMILY PADDON RHOADS

13

have effectively prevented violence, dissuaded threats, and saved lives across civil conflicts (Kaplan 2017). In some contexts, communities have even managed to carve out ‘islands of peace’ amidst violent conflict (Hancock and Mitchell 2007; Mouly et al. 2015; Masullo 2017; Autesserre 2021). Moreover, civilian protection monitoring in conflict zones of Myanmar has helped mitigate the indirect but devastating long-term effects of violence against civilians by aiding civilian displacement and humanitarian support (J. Krause and Kamler 2022; Krause within volume). Temporary and permanent evasion has protected some individuals and groups from violence and recruitment in a range of settings, such as the Holocaust (Finkel 2017) and civil war (Baines and Paddon 2012; Schon 2020). Other forms of civilian protective agency have enabled entire communities to remain in place, avoiding displacement and loss of livelihoods (Krakowski 2017; Marston 2020). Relatedly, collective forms of protective agency have proved effective in limiting and/or shaping non-state governance. In the context of civil war, civilian resistance has limited the extent to which rebel governance interferes in civilian affairs (Arjona 2016) and allowed civilians to demand better, less violent, and more inclusive forms of governance (Breslawski 2020; Rubin 2019). This protective potential has also been found in contexts of criminal governance, where communities have managed to effectively resist extortion (Moncada 2022) and counter criminal governance in urban spaces (Lessing 2021). For example, in Rio de Janeiro, favela resident organizations have opposed criminal governance by effectively promoting social development and the integration of marginal areas into the political arena (Fahlberg 2018; Barnes 2021). While civilian protective agency may produce a protection dividend, it can also lead to more violence and deepen insecurity. Some forms of action, such as overtly refusing to cooperate with armed groups or hiding potential victims, can put civilians at higher risk of harm, while other forms can produce new violence and new protection challenges. For example, Verweijen (within volume) documents how civilians’ quest for immediate protection in the Democratic Republic of the Congo (DRC) entrenched the dominant position of armed actors in nonmilitary spheres of social life and sparked violent competition between protection providers. These transformations are not unique to civil war. Moncada (within volume) shows that the day-to-day challenges that vigilante groups face can over time lead them to engage in some of the same criminal behaviours they originally formed to stop—as has also been reported for various militia groups in civil war settings (Agbiboa 2021). These transformations underscore the importance of adopting a longer temporal perspective when assessing the consequences of civilian protective agency. As Jentzsch (within volume) shows for community-initiated militias in Mozambique, there are reasons to believe that a positive protection dividend may only be temporary, as armed actors can quickly adapt and learn how to respond to civilian protective strategies (see also Kaplan 2013).

14

CIVILIAN PROTECTIVE AGENCY: AN INTRODUCTION

International actors and civilian protective agency A growing body of scholarship examines the ways in which external actors— from humanitarians to human rights activists to peacekeepers—shape the emergence and trajectories of civilian protective agency through direct support for community organizations, the co-creation of new structures, and other types of engagement including training and network-building (Kaplan 2021; J. Krause and Kamler 2022; Paddon Rhoads and Gorur within volume). The mere presence of external actors with a protection mandate can alter how civilians think about their own protection, their risk calculations, and the actions they pursue (Paddon Rhoads 2016). In her chapter on Syria, Howe stresses the role that the international community can play in civilians’ local adaptation to chronic insecurity, identifying ways in which humanitarian and stabilization operations can both support and undermine civilian protective agency. Providing international support for civilian protective agency is not without challenges. Kaplan (2021) cautions that international peacebuilding and protection actors must guard against creating a ‘moral hazard’ by promoting ultimately risky civilian protection practices, while Krause and Kamler (2022) warn that supporting local protection mechanisms does not necessarily mean that international actors can ‘scale up’ local civilian protection. While international and local perceptions of effective protection practices may at times be misaligned, civilians often adapt protection knowledge provided by international actors to their reality on the ground, which allows them to mitigate risks to themselves and others (J. Krause and Kamler 2022; Krause within volume). Devolving decision-making authority and control to communities may also give rise to institutional risks for international actors, particularly where conceptions of protection and threat do not align, and/or the forms of civilian self-protection pursued contravene the values of the organization (Baines and Paddon 2012; Welsh et al. forthcoming). While support for civilian protective agency poses challenges for all international actors, there is variation in how they manifest across organizations. As Paddon Rhoads and Gorur show in their chapter, the distinct role of UN peacekeepers as armed protection actors and the UN’s perceived partiality in some contexts, makes supporting civilian protective agency particularly risky for the UN. Consequently, UN policy has not translated meaningfully into practice. While peacekeepers have significantly increased engagement with communities on protection, the authors find few instances of support for community-led protection efforts. Recognition of the diversity of approaches to supporting civilian protective agency as well as the inherent challenges will thus be critical as we look to a future in which international actors are likely to continue shifting away from ‘top-down’ protection approaches towards more localized forms of engagement that centre the agency of communities.

JANA KRAUSE, JUAN MASULLO, AND EMILY PADDON RHOADS

15

4. Overview of the book This volume is the first to examine civilian agency and protection across different types of violent settings. The book is organized into three parts. Part One features six chapters exploring distinct forms of civilian protective agency in different violent contexts. Contributors explore women’s rescue agency during the Holocaust in the Netherlands (Braun and Stallone), civilian networks and adaptation in the Syrian civil war (Howe), civilian adoption of occult practices for their own protection in the Central African Republic (Lombard and Kozaga), civilian resistance against Jihadist governance in the Iraqi conflict (Svensson and Smart), the durability of women’s resistance during the Colombian civil war (Zulver), and finally the commodification of civilian protective responses in the context of the DRC (Verweijen). Part Two shifts the attention from nonviolent action to exploring the ways in which protective agency can evolve into armed mobilization. Contributors analyse the relationship between community militias and the intensity of the Mozambican civil war (Jentzsch), self-defence groups in Mexico’s criminal conflict (Ben Hamo and Masullo), and trajectories of vigilantism in response to criminal victimization in Nigeria and El Salvador (Moncada). In Part Three, contributors establish vital linkages between the study of civilian agency and scholarship on international interventions, peacebuilding, and the protection of civilians. They consider the role of external actors in enabling, supporting, and sustaining civilian protective agency with chapters on civilian monitoring in Myanmar (Krause), mass atrocity prevention in Sri Lanka, South Sudan, and the DRC (Mampilly and Solomon), and UN peacekeeping across a range of contexts (Paddon Rhoads and Gorur). The Conclusion synthesizes the volume’s findings, reflects on the methodological and normative questions raised by the chapters, and identifies areas for future research and interdisciplinary collaboration. Our broad conceptualization of civilian protective agency facilitates the comparative analysis of the protection practices civilians adopt to not only save themselves and others from being killed but also to mitigate the manifold and devastating humanitarian consequences of violence across settings. When analysing civilian protective agency, we should never forget that civilians adopt such protection practices under utmost duress with high risks and costs to themselves, their families, and livelihoods. Consequently, civilian protective agency may not always be sustainable and it may expose individuals and communities to new protection threats and potential harms. Taken together, the contributions to this volume are a testament to the remarkable agency of civilians across violent settings. However, they should also serve as a cautionary reminder that people’s local protection efforts should never be taken for granted.

16

CIVILIAN PROTECTIVE AGENCY: AN INTRODUCTION

References Agbiboa, D. (2021). ‘The Precariousness of Protection: Civilian Defense Groups Countering Boko Haram in Northeastern Nigeria’, African Studies Review, 64/1: 192–216. Arjona, A. (2015). ‘Civilian Resistance to Rebel Governance’, in Arjona, A., Kasfir, N., and Mampilly, Z. C. (eds), Rebel Governance, pp. 180–202. Cambridge: Cambridge University Press. Arjona, A. (2016). Rebelocracy: Social Order in the Colombian Civil War. Cambridge: Cambridge University Press. Autesserre, S. (2021). On the Frontlines of Peace. New York: Oxford University Press. Baines, E. (2016). Buried in the Heart: Women, Complex Victimhood and the War in Northern Uganda. Cambridge Studies in Law and Society. Cambridge: Cambridge University Press. DOI: 10.1017/9781316480342 Baines, E., and Paddon E. (2012). ‘“This Is How We Survived”: Civilian Agency and Humanitarian Protection’, Security Dialogue, 43/3: 231–47. Balcells, L., and Solomon, D. (2020). ‘Violence, Resistance, and Rescue during the Holocaust’, Comparative Politics, 53/1: 161–80. DOI: 10.5129/001041520X15863824603010 Barnes, N. (2021). ‘The Logic of Criminal Territorial Control: Military Intervention in Rio de Janeiro’, Comparative Political Studies, 55/5: 789–831. DOI: 10.1177/00104140211036035 Barnes, N. (Forthcoming). Gang Governance in Rio de Janeiro. Cambridge University Press Barter, S. J. (2014). Civilian Strategy in Civil War: Insights from Indonesia, Thailand, and the Philippines. New York: Palgrave Macmillan. Bateson, R. (2021). ‘The Politics of Vigilantism’, Comparative Political Studies, 54/6: 923–55. DOI: 10.1177/0010414020957692 Berenschot, W. (2011). Riot Politics: Hindu-Muslim Violence and the Indian State. London: Hurst. Bonkat, L. (2014). ‘Survival Strategies of Market Women and Violent Conflicts in Jos, Nigeria’, Journal of Asia Pacific Studies, 3/3: 281–299. Bradley, M. (2020). ‘From Armed Conflict to Urban Violence: Transformations in the International Committee of the Red Cross, International Humanitarianism, and the Laws of War’, European Journal of International Relations, 26/4: 1061–83. DOI: 10.1177/1354066120908637 Braun, R. (2018). ‘Minorities and the Clandestine Collective Action Dilemma: The Secret Protection of Jews during the Holocaust’, American Journal of Sociology, 124/2: 263–308. DOI: 10.1086/699651 Braun, R. (2019). Protectors of Pluralism: Religious Minorities and the Rescue of Jews in the Low Countries during the Holocaust. New York: Cambridge University Press. Breslawski, J. (2020). ‘The Social Terrain of Rebel Held Territory’, Journal of Conflict Resolution, 65/2–3: 453–479. DOI: 10.1177/0022002720951857 Breslawski, J. (2021). ‘Keeping Armed Actors Out: The Protective Effect of Shuras in Afghanistan’, Journal of Global Security Studies, 6/1. DOI: 10.1093/jogss/ogz049 Carbonnier, G. (2010). ‘Official Development Assistance Once More Under Fire from Critics’, International Development Policy, 1: 137–142. DOI: 10.4000/poldev.141 Carpenter, A. C. (2012). ‘Havens in a Firestorm: Perspectives from Baghdad on Resilience to Sectarian Violence’, Civil Wars, 14/2: 182–204.

JANA KRAUSE, JUAN MASULLO, AND EMILY PADDON RHOADS

17

Dhattiwala, R. (2019). Keeping the Peace: Spatial Differences in Hindu-Muslim Violence in Gujarat in 2002. Cambridge: Cambridge University Press. Einwohner, R., and Maher, T. (2011). ‘Threat Assessment and Collective-Action Emergence: Death-Camp and Ghetto Resistance During the Holocaust’, Mobilization: An International Quarterly, 16/2: 127–46. Emirbayer, M., and Mische, A. (1998). ‘What is Agency?’, American Journal of Sociology, 103/4: 962–1023. Fahlberg, A. N. (2018). ‘Rethinking Favela Governance: Nonviolent Politics in Rio de Janeiro’s Gang Territories’, Politics & Society, 46/4: 1–28. Finkel, E. (2015). ‘The Phoenix Effect of State Repression: Jewish Resistance during the Holocaust’, American Political Science Review, 109/02: 339–53. DOI: 10.1017/S000305541500009X Finkel, E. (2017). Ordinary Jews: Choice and Survival During the Holocaust. Princeton, NJ: Princeton University Press. Firchow, P. (2018). Reclaiming Everyday Peace: Local Voices in Measurement and Evaluation after War. New York: Cambridge University Press. Fujii, L. A. (2009). Killing Neighbors: Webs of Violence in Rwanda. Ithaca, NY: Cornell University Press. Geneva Declaration. (2008). ‘Global Burden of Armed Violence’, Geneva: Geneva Declaration Secretariat. Gingerich, T. R., and Cohen, M. J. (2015). Turning the Humanitarian System on its Head: Saving Lives and Livelihoods by Strengthening Local Capacity and Shifting Leadership to Local Actors (Oxfam Research Reports). Oxfam Research Reports. Washington, DC: Oxfam. Gorur, A., and Carstensen, N. (2016). ‘Community Self-Protection’, in Willmot, H., Mamiya, R., Sheeran, S., and Weller, M. (eds), Protection of Civilians, pp. 409–27. Oxford: Oxford University Press. Gowrinathan, N., and Mampilly, Z. (2019). ‘Resistance and Repression under the Rule of Rebels: Women, Clergy, and Civilian Agency in LTTE Governed Sri Lanka’, Comparative Politics, 52/1: 1–20. DOI: 10.5129/001041519X156983520 40097 Green, A. H. (2018). The Commander’s Dilemma: Violence and Restraint in Wartime. Ithaca, NY: Cornell University Press. Hancock, L. E., and Mitchell, C. R. (eds). (2007). Zones of Peace. Bloomfield, CT: Kumarian Press. Hitlin, S., and Elder Jr, G. H. (2007). ‘Time, Self, and the Curiously Abstract Concept of Agency’, Sociological Theory, 25/2: 170–91. Holland, J. (1995). Hidden Order: How Adaptation Builds Complexity. New York: Addison-Wesley. ICRC. (2001). Strengthening Protection in War: A Search for Professional Standards. Geneva: International Committee of the Red Cross. Jackson, A. (2021). Negotiating Survival: Civilian-Insurgent Relations in Afghanistan. Oxford: Oxford University Press. Jentzsch, C. (2022). Violent Resistance: Militia Formation and Civil War in Mozambique. Cambridge: Cambridge University Press. Jose, B., and Medie, P. A. (2015). ‘Understanding Why and How Civilians Resort to Self-Protection in Armed Conflict’, International Studies Review, 17/4: 515–35. DOI: 10.1111/misr.12254

18

CIVILIAN PROTECTIVE AGENCY: AN INTRODUCTION

Julian, R. (2020). ‘The Transformative Impact of Unarmed Civilian Peacekeeping’, Global Society, 34/1: 99–111. Kalyvas, S. N. (2003). ‘The Ontology of “Political Violence”: Action and Identity in Civil Wars’, Perspectives on Politics, 1/3: 475–94. American Political Science Association, Cambridge University Press. Kalyvas, S. N. (2006). The Logic of Violence in Civil War. Cambridge: Cambridge University Press. Kalyvas, S. N. (2019). ‘The Landscape of Political Violence’. The Oxford Handbook of Terrorism. Oxford: Oxford University Press. DOI: 10.1093/oxfordhb/9780198732914.013.1 Kaplan, O. (2013). ‘Protecting Civilians in Civil War: The Institution of the ATCC in Colombia’, Journal of Peace Research, 50/3: 351–67. Kaplan, O. (2017). Resisting War. How Communities Protect Themselves. Cambridge: Cambridge University Press. Kaplan, O. (2021). ‘The International Committee of the Red Cross and Support for Civilian Self-Protection in Colombia’, International Interactions, 47/5: 898–927. DOI: 10.1080/03050629.2021.1945593 Klaus, K. (2020). Political Violence in Kenya: Land, Elections, and Claim-Making. Cambridge: Cambridge University Press. Krakowski, K. (2017). ‘Resisting Displacement amid Armed Conflict: CommunityLevel Conditions that Make People More Likely to Stay’, Journal of Peacebuilding & Development, 12/3: 68–84. DOI: 10.1080/15423166.2017.1370387 Krause, J. (2017). ‘Protective Civilian Agency in War: Emergence, Repertoires, Impact and Legacies’. Working Paper. University of Oslo. Krause, J. (2018). Resilient Communities: Non-violence and Civilian Agency in Communal War. Cambridge: Cambridge University Press. Krause, J. (2019). ‘Gender Dimensions of (Non)-Violence in Communal Conflict: The Case of Jos, Nigeria’, Comparative Political Studies, 52/10: 1466–99. Krause, J., and Kamler, E. (2022). ‘Ceasefires and Civilian Protection Monitoring in Myanmar’, Global Studies Quarterly, 2/1. Krause, K. (2016). ‘From Armed Conflict to Political Violence: Mapping & Explaining Conflict Trends’, Daedalus, 145/4: 113–26. Kreft, A.-K. (2019). ‘Responding to Sexual Violence: Women’s Mobilization in War’, Journal of Peace Research, 56/2: 220–33. DOI: 10.1177/0022343318800361 Kriger, N. J. (1991). Zimbabwe’s Guerrilla War: Peasant Voices. Cambridge: Cambridge University Press. Lessing, B. (2021). ‘Conceptualizing Criminal Governance’, Perspectives on Politics, 19/3: 854–73. Ley, S., Mattiace, S., and Trejo, G. (2019). ‘Indigenous Resistance to Criminal Governance: Why Regional Ethnic Autonomy Institutions Protect Communities from Narco Rule in Mexico’, Latin American Research Review, 54/1: 181–200. DOI: 10.25222/larr.377 Lubkemann, S. C. (2008). Culture in Chaos: An Anthropology of the Social Condition in War. Chicago: University of Chicago Press. Luft, A. (2015). ‘Toward a Dynamic Theory of Action at the Micro Level of Genocide: Killing, Desistance, and Saving in 1994 Rwanda’, Sociological Theory, 33/2: 148–72. CA: Los Angeles. Mac Ginty, R. (2014). ‘Everyday Peace: Bottom-Up and Local Agency in ConflictAffected Societies’, Security Dialogue, 45/6: 548–64.

JANA KRAUSE, JUAN MASULLO, AND EMILY PADDON RHOADS

19

Mac Ginty, R. (2021). Everyday Peace: How So-called Ordinary People Can Disrupt Violent Conflict. Oxford: Oxford University Press. McQuinn, B., Terry, F., Kaplan, O., and Gutiérrez-Sanin, F. (2021). ‘Introduction: Promoting Restraint in War’, International Interactions, 47: 795–824. Maher, T. V. (2010). ‘Threat, Resistance, and Collective Action: The Cases of Sobibór, Treblinka, and Auschwitz’, American Sociological Review, 75/2: 252–72. DOI: 10.1177/0003122410365305 Mampilly, Z. (2011). Rebel Rulers: Insurgent Governance and Civilian Life During War. Ithaca, NY: Cornell University Press. Marston, J. F. (2020). ‘Resisting Displacement: Leveraging Interpersonal Ties to Remain despite Criminal Violence in Medellı´n, Colombia’, Comparative Political Studies, 53/13: 1995–2028. DOI: 10.1177/ 0010414020912276 Masullo, J. (2017). A Theory of Civilian Noncooperation with Armed Groups. Civilian Agency and Self-protection in the Colombian Civil War (Doctoral Dissertation). Florence: European University Institute. Masullo, J. (2021). ‘Civilian Contention in Civil War. How Ideational Factors Shape Community Responses to Armed Groups’, Comparative Political Studies, 54/10: 1849–84. Moncada, E. (2022). Resisting Extortion: Victims, Criminals, and States in Latin America. Cambridge Studies in Comparative Politics. Cambridge: Cambridge University Press. Monroe, K. R. (1998). The Heart of Altruism. Princeton, NJ: Princeton University Press. Monroe, K. R. (2011). Ethics in an Age of Terror and Genocide: Identity and Moral Choice. Princeton, NJ: Princeton University Press. Mouly, C., Idler, A., and Garrido, M. B. (2015). ‘Zones of Peace in Colombia’s Borderland’, International Journal of Peace Studies, 20/1: 51–63. Nordstrom, C. (1997). A Different Kind of War Story. University of Pennsylvania Press. Obrecht, A. (2014). De-Internationalising Humanitarian Action: Rethinking the ‘Global–Local’ Relationship, Paris: Paris Humanitarian Affairs Think Tank and Institut de Relations Internationales et Strategiques. OED. (n.d.). ‘Protection: Oxford English Dictionary’. Oxford English Dictionary. Oliner, S. P. (1992). Altruistic Personality: Rescuers of Jews in Nazi Europe. New York: Free Press. Osorio, J., Schubiger, L. I., and Weintraub, M. (2021). ‘Legacies of Resistance: Mobilization against Organized Crime in Mexico’, Comparative Political Studies, 54/9: 1565–1596. DOI: 10.1177/0010414021989761 Paddon Rhoads, E. (2016). Taking Sides in Peacekeeping: Impartiality and the Future of the United Nations. Oxford: Oxford University Press. Paddon Rhoads, E., and Sutton, R. (2020). The (Self ) Protection of Civilians in South Sudan: Popular and Community Justice Practices’, African Affairs, 119/476: 370–94. DOI: 10.1093/afraf/adaa017 Petersen, R. D. (2001). Resistance and Rebellion. Lessons from Eastern Europe. Cambridge: Cambridge University Press. Peterson, R. (2001). ‘Production of Culture’, International Encyclopedia of the Social & Behavioral Sciences, 8: 328–32. Pincock, K., Betts, A., and Easton-Calabria, E. (2020). The Global Governed?: Refugees as Providers of Protection and Assistance. New York: Cambridge University Press.

20

CIVILIAN PROTECTIVE AGENCY: AN INTRODUCTION

Rubin, M. A. (2019). ‘Rebel Territorial Control and Civilian Collective Action in Civil War: Evidence from the Communist Insurgency in the Philippines’, Journal of Conflict Resolution, 64/2–3: 459–489. DOI: 10.1177/0022002719863844 Sánchez-Talanquer, M. (2018). ‘Legacies of Revolution: Popular Militias and the Rule of Law’, unpublished manuscript. Sanı´n, F. G., and Wood, E. J. (2014). ‘Ideology in Civil War: Instrumental Adoption and Beyond’, Journal of Peace Research, 51/2: 213–26. Schon, J. (2020). Surviving the War in Syria: Survival Strategies in a Time of Conflict. New York: Cambridge University Press. Schubiger, L. I. (2021). ‘State Violence and Wartime Civilian Agency: Evidence from Peru’, Journal of Politics, 83/4: 1565–96. Sémelin, J., Andrieu, C., and Gensburger, S. (2014). Resisting Genocide: The Multiple Forms of Rescue. Oxford: Oxford University Press. Slim, H. (2016). ‘Civilians, Distinction, and the Compassionate View of War’, in Willmot, H., Mamiya, R., Sheeran, S., and Weller, M. (eds), Protection of Civilians, pp. 11–28. Oxford: Oxford University Press. Slim, H., and Bonwick, A. (2006). Protection: An Alnap Guide for Humanitarian Agencies. London: Active Learning Network for Accountability/Overseas Development Initiative. Stanton, J. A. (2016). Violence and Restraint in Civil War: Civilian Targeting in the Shadow of International Law. Cambridge: Cambridge University Press. Steele, A. (2009). ‘Seeking Safety: Avoiding Displacement and Choosing Destinations in Civil Wars’, Journal of Peace Research, 46/3: 419–29. DOI: 10.1177/0022343309102660 Steele, A. (2017). Democracy and Displacement in Colombia’s Civil War. Ithaca, NY: Cornell University Press. Stoll, D. (1993). Between Two Armies in the Ixil Towns of Guatemala. New York: Columbia University Press. Tec, N. (1995). ‘A ltruism and the Holocaust’, Social Education, 59/6: 348–53. Tec, N. (2004). Resilience and Courage: Women, Men, and the Holocaust. New Haven, CT: Yale University Press. Tilly, C. (2003). The Politics of Collective Violence. Cambridge Studies in Contentious Politics. Cambridge: Cambridge University Press. DOI: 10.1017/CBO9780511819131 United Nations. (2020). “A New Era of Conflict and Violence”, UN75, https://www.un. org/en/un75/new-era-conflict-and-violence (accessed 11 April 2023). Utas, M. (2005). ‘Victimcy, Girlfriending, Soldiering: Tactic Agency in a Young Woman’s Social Navigation of the Liberian War Zone’, Anthropological Quarterly, 78/2: 403–30. Varshney, A. (2003). Ethnic Conflict and Civic Life: Hindus and Muslims in India. New Haven, CT: Yale University Press. Verweijen, J. (2018). ‘Civilian Resistance against the Military in Eastern DR Congo: A Combined Social Navigation and Structuration Approach’, Qualitative Sociology, 41/2: 281–301. Vigh, H. (2009). ‘Motion Squared: A Second Look at the Concept of Social Navigation’, Anthropological Theory, 9/4: 419–38. Vu¨llers, J., and Krtsch, R. (2020). ‘Raise Your Voices! Civilian Protest in Civil Wars’, Political Geography, 80: 1–12. DOI: 10.1016/j.polgeo.2020.102183

JANA KRAUSE, JUAN MASULLO, AND EMILY PADDON RHOADS

21

Wallace, M. S. (2016). Security Without Weapons: Rethinking Violence, Nonviolent Action, and Civilian Protection. New York: Routledge. Weinstein, J. M. (2007). Inside Rebellion: The Politics of Insurgent Violence. Cambridge: Cambridge University Press. Welsh, J., Paddon Rhoads, E., and Masullo, J. (forthcoming). ‘Risky Business: International Support for Civilian Self-Protection’. Williamson, J. A. (2016). ‘Protection of Civilians under International Humanitarian Law’, in Willmot, H., Mamiya, R., Sheeran, S., and Weller, M. (eds), Protection of Civilians, pp. 160–76. Oxford: Oxford University Press. Wise, P. H. (2017). ‘The Epidemiologic Challenge to the Conduct of Just War: Confronting Indirect Civilian Casualties of War’, Daedalus, 146/1: 139–54. Wood, E. J. (2003). Insurgent Collective Action and Civil War in El Salvador. Cambridge: Cambridge University Press. Zulver, J. M. (2022). High-Risk Feminism in Colombia: Women’s Mobilization in Violent Contexts. New Brunswick: Rutgers University Press.

PART ONE

F OR MS OF CI V I LI A N PR OTECTI V E AGENC Y ACR OSS V IOLEN T SET TI NGS

1 Interregional networks and conventionalized gender roles in civilian resistance against genocide Evidence from the Holocaust in the Low Countries Robert Braun and Kiran Stallone

Introduction This chapter focuses on the relational and gendered underpinnings of civilian protective agency, specifically the provision of help to victims of oppression by civilian non-victims. Contrary to other chapters in this volume, it looks at how certain civilians protect others—as opposed to how civilians engage in their own self-protection. We do this through a detailed study of Dutch women who rescued Jewish people threatened by mass persecution in the Netherlands during World War II. Although the editors of this book separate rescue and resistance into distinct categories, we see the two as intertwined. By transporting Jewish people across territories in order to facilitate their escape to places of safety (Semelin et al. 2014), Dutch women rescuers reinforced their non-cooperation and active ‘refusal to accept circumstances’ (Krause et al. within volume) by collectively opposing genocidal policies. By underscoring these cases, we show how private acts of rescue are often embedded in collective networks of resistance (Braun 2018). The clandestine nature of these rescue operations sets this form of resistance apart from other forms of resistance that involve public claim-making (e.g. Svensson and Smart within volume). The Dutch women discussed here carried out their rescue work through informal civilian networks they themselves created and that spanned the country. A recent wave of research has drawn attention to the notion that civilian resistance in general is most powerful when it is embedded in localized hubs of rebellion that are connected by strong supra-local ties between communities. While dense and localized networks (Krause et al. within volume) allow movements to overcome high-risk collective action problems by enabling recruitment,

Robert Braun and Kiran Stallone, Interregional networks and conventionalized gender roles in civilian resistance against genocide. In: Civilian Protective Agency in Violent Settings. Edited by: Jana Krause, et al., Oxford University Press. © Robert Braun and Kiran Stallone (2023). DOI: 10.1093/oso/9780192866714.003.0002

26

INTERREGIONAL NET WORKS AND GENDER ROLES

coordination (McAdam 1986; Sageman 2011; Della Porta 2013), and social sanctioning (Coleman 1988), they are not independently sufficient to successfully counter oppressors’ ideas and policies. Research on insurgency (Staniland 2014), terrorism (Shapiro 2013), clandestine networks (Braun 2018), resistance organizations (Finkel 2017), and social movements (Morris 1986) has revealed that powerful resistance often involves challenging much stronger opponents across wide swaths of territory, and that this requires the establishment of long-distance communication channels (Finkel 2017), the procurement of weapons, ammunition, illegal documents, and other supportive materials from disparate regions with unique access to resources (Parkinson 2013). It also entails the transportation of these resources across space (Shapiro 2013), the establishment of interregional networks of well-hidden shelters that enable movements to funnel persecuted individuals to places where they are less visible (Braun 2018), and the formation of interregional identities and coalitions that can anchor strong resistance for prolonged periods of time (Staniland 2014). Despite the importance of interregional ties for the organization of sustainable resistance campaigns, relatively little is known about how they are formed and maintained. Studies on civilian agency have noted an important gender dimension in the production of resistance and protection (Utas 2005; Baines 2016; Krause 2019; Hume and Wilding 2020; Paddon Rhoads and Sutton 2020). We develop this connection further by focusing on civilian protective agency under conditions of military occupation and genocide. Drawing on Goffman’s (1956) dramaturgical framework, we argue that whenever repressors adhere to traditional gender norms, women can play a more important role than men by maintaining long-distance connections between resistance operations in different local communities. Traditional beliefs constitute oppressors’ expectations through the activation of gendered norms, roles, and scripts of appropriate behaviour (Carpenter 2003). These scripts and roles set the stage for women resisters to act. In times of mass upheaval, we show that women can exploit gendered expectations of repressors by strategically playing up essentialist feminine roles and scripts (Connell 1987) that enhance security in general (Kandiyoti 2016) and conceal long-distance travel between localities in particular. Studying resistance is hard. Gaining direct access to contemporary resistance cells is nearly impossible because these cells need to reduce exposure to survive. Archival work on historical cases can alleviate this challenge because it allows us to study groups that are no longer under immediate threat, reducing the need for secrecy. Moreover, when political structures open and regimes change, former clandestine networks sometimes go public to gain recognition for their activities, opening up an array of archives and testimonies. Instead of focusing on contemporary cases, this study therefore examines a historical episode of clandestine collective action: the secret protection and rescue of Jews during the Holocaust. When German occupiers started deporting

ROBERT BRAUN AND KIRAN STALLONE

27

Jews from the Netherlands, groups of Dutch Gentiles who rejected Nazism began building secret organizations to facilitate the illegal shelter and migration of threatened neighbours, developing expansive interregional networks that traversed the country. The chapter utilizes an integrative multi-method analysis (Seawright 2016), combining a quantitative network analysis to establish patterns of rescue behaviour and in-depth qualitative methods to identify the mechanisms underlying the patterns outlined above. We proceed in two stages. First, based on German administrative records and extensive bodies of post-war testimonies, we present data on rescue networks and Jewish evasion per municipality. Statistical analysis of the data confirms that women played a key role in facilitating Jewish evasion by forming ties that bridged different localities. Second, to gain more insight into why this was the case, we focus on several prominent female rescuers with multiple interregional ties. This in-depth analysis of autobiographies, post-war testimonies, and secondary literature identifies three distinct roles that Dutch women performed to facilitate the formation of interregional resistance networks: (1) strategic motherhood and wedlock; (2) strategic self-devaluation, and (3) strategic coquetry. The performance of these three roles allowed women to fly under the radar as they funnelled Jews across the Netherlands and carried out other rescue work that required long-distance travel.

Interregional rescue networks, gender, and Jewish evasion Only 28,500 Jews evaded the frighteningly effective and efficient deportation campaign by the Nazis in the Netherlands. As crossing the Dutch border was extremely difficult, Jews who wanted to escape deportation had to find hiding places inside the country. To do so, they depended heavily on the aid of others, whether for sharing resources, providing shelter, or facilitating mobility between safe houses. Initially, Jews were often helped by neighbours, colleagues, family physicians, neighbourhood patrolmen, business relationships, former customers, or nearby friends. With the exception of some student networks, most early rescue activities remained geographically isolated and disjointed (Moore 1997). Over time, the need for interregional coordination became more important. Localized helpers realized that, despite their best intentions, they could not manage to shelter Jews for prolonged periods of time without outside help. By then, the extent to which rescuers could find help outside local communities determined whether they could continue rescue activities. Reliance on help elsewhere entailed risks and required clandestine activity. Sometimes Jews were both well-known and unpopular in their neighbourhoods. In such cases, the Nazis could have easily detected Jews if they stayed near their original homes. Relocation became crucial to retaining secrecy (Braun 2018). Despite their importance, interregional

28

INTERREGIONAL NET WORKS AND GENDER ROLES

connections were far from universal, and some rescuers were better able to negate fear and set up interregional relationships than others. The next section reveals that gender was paramount in this process. To investigate the relational underpinnings of rescue, we constructed two datasets. The first dataset provides georeferenced information on a large majority of Dutch rescue networks. Based on post-war testimonies collected by Yad Vashem, we geocoded the relations between rescuers active in the Netherlands during the Holocaust.¹ These testimonies from rescue workers were particularly useful for the reconstruction of the social structure of clandestine organization and activities, as the Yad Vashem entries almost always include brief descriptions of how different rescuers came into contact and began to work together. We marked rescuers as having a relationship if the testimonies mentioned that they funnelled Jews, illegal documents, food, or other resources to other rescue workers. This resulted in a database of around a thousand rescuers who were active in 260 localities with around 1900 interrelationships. Although most of these relationships were formed between rescuers active in the same locality, 41 per cent transcended town borders. These interregional ties are mapped in Figure 1.1a. The maps reveal that interregional networks covered the whole of the Netherlands, but were clustered in the western and more urban parts of the country. We paired the relational rescue network data with fine-grained data on Jewish evasion. Drawing on a combination of deportation lists, commemoration books and Nazi administrative records, we calculated the percentage of Jews who evaded deportation per municipality.² The database in Figure 1.1b provides information on 475 Dutch localities spread out over the country. To assess how important rescue networks were for evasion, we modelled the relationship between different network ties and evasion using ordinary least squares regression with robust standard errors while conditioning on three sets of control variables. The first set includes basic demographics such as logged population size, the proportion of the local population that was Jewish, and the number of rescue networks active in rescue municipalities. The second set taps the Nazis’ repressive capacity. Previous research has revealed that the local strength of the Nazis and the support they obtained from the local population varied greatly across the Netherlands. It seems plausible that repressive capacity and mobilization potential would influence both the emergence of rescue networks and deportation (Croes and Tammes 2004).

¹ Yad Vashem is a research institute and memorial that preserves information about the Holocaust and honours the memory of those who perished. The Yad Vashem Righteous Among the Nations Database is a collection of names and information about non-Jewish individuals who helped save Jews during the Holocaust. A public commission (supervised by a Supreme Court justice) examines all entries, and individual rescuers who are approved are awarded a medal in addition to public acknowledgement and documentation in the database. ² A detailed description of this data collection process can be found in Appendix A of (Braun 2020).

ROBERT BRAUN AND KIRAN STALLONE

29

SURVIVAL RATE JEWS 0 – 15% 15 – 35% 35 – 55% 55 – 85% 85 – 100% NO JEWS/MISSING

RESCUE NETWORK INTERREGIONAL TIE

(a) Interregional Rescure Networks.

(b) Jewish Evasion Rates Per Municipality.

Figure 1.1 (a) Interregional rescue networks and (b) Jewish evasion in Holocaust Netherlands per municipality.

Therefore, we controlled for the percentage of the population voting for the Dutch National Socialist Party before the German invasion, the percentage of the population working as Nazi informants or V-Manner, the local strength of the Nazi security apparatus in charge of Jewish deportations (the SIPO-SD), the strength of auxiliary police forces deployed to track down Jews, and the percentage of police officers who were members of the National Socialist Movement. All data were obtained from the archives of the Ministry of Justice in the Nationaal Archief in Den Hague.³ The third set of control variables captures local religious composition. Existing research has demonstrated that religious minorities are more likely to protect victims of mass persecution (Braun 2018). As the Netherlands was cut in half by the Reformation, this would suggest that Catholics were more likely to shelter Jews in the Protestant north and Protestants were more likely to protect Jews in the ³ The different measures of repressive capacity could not be combined into one index measure due to low reliability (alpha = .21). We did run separate models in which we entered the different items one by one instead of all at once. The results are identical to those presented in the main text. The effect of interregional ties did interact with repressive capacity. The impact of interregional ties was attenuated in regions with low repressive capacity. This suggests that these ties made it possible to funnel more resources and Jews when repression was less intense. However, these conditional effects were fairly weak.

30

INTERREGIONAL NET WORKS AND GENDER ROLES

Catholic south of the country. In addition, independent orthodox Protestant congregations that split away from the mainline Dutch Protestant Church because of dissatisfaction with modernist tendencies should also be overrepresented among rescue workers. We try to tap these minority effects by including the percentage of votes for the political party representing Orthodox Protestant interests and nonlinear measures of Roman Catholicism. The results of this analysis are depicted in Figure 1.2. In line with recent work that stresses the importance of interregional networks for civilian resistance (Morris 1986; Parkinson 2013; Shapiro 2013; Staniland 2014; Finkel 2017), the analysis revealed a strong correlation between Jewish evasion and the presence of ties that transcended local boundaries. Even if we control for the overall strength of rescue networks (as proxied by the number of all rescue network ties), the presence of interregional ties increased local evasion by almost 10 per cent, a considerable effect given that only 25 per cent of the entire Jewish population in the Netherlands survived the war. Next, we explored quantitative network data to determine gender differences in rescue ties. Figure 1.3 breaks down relationships between rescuers by gender. As is clear from the top left (Figure 1.3a), women rescuers were slightly less likely to have ties to other rescuers than men. The nature of those ties, however, varied greatly. Whereas men were almost twice as likely to have ties to nearby rescuers than women (Figure 1.3b), women were much more likely to have relationships with rescuers from different regions (Figure 1.3c). Ties involving women were almost

Interregional tie (0/1) # All Network ties % Jews Population (Logged) % Nazi votes 1937 % Nazi spies % SIPO-SD % Auxiliary police % Nazi police % Orthodox Prot. votes % Catholics % Catholics squared –10

–5

0 5 ∆ % Survival

10

15

Figure 1.2 The effect of increasing variables one standard deviation on Jewish survival with 90 and 93% confidence intervals.

ROBERT BRAUN AND KIRAN STALLONE Rescuers w. Only Internal Ties/All Rescuers

Rescuers w. Ties/All Rescuers

.8

.6

.4

.2

0

Female Rescuers (N=507)

.4

.3

.2

.1

0

Male Rescuers (N=412)

31

(a) All Ties By Gender

Female Rescuers (N=507)

Male Rescuers (N=412)

(b) Intraregional Ties By Gender

Interregional Ties/All Ties

Rescuers w. Interregional Ties/All Rescuers

.6 .4

.3

.2

.1

0

Female Rescuers (N=507)

Male Rescuers (N=412)

(c) Interregional Ties By Gender

.4

.2

0

Female Rescuers (N=862)

Male Rescuers (N=1,039)

(d) Interregional Ties / All Ties By Gender

Figure 1.3 Interregional and intraregional ties by gender of the rescuer.

twice as likely to transcend regional boundaries than ties that did not involve women (Figure 1.3d).⁴ There is also suggestive evidence that interregional ties involving women provided Jews more robust protection. This was revealed when we re-estimated the earlier described statistical model with one minor alteration. Instead of lumping all interregional ties together, we distinguished between ties involving women and ties that did not. A summary of the results is outlined in Figure 1.4. Overall, the effects are similar to those presented in Figure 1.3. However, one stark difference should jump out to the reader. Interregional ties involving women had a slightly stronger effect on local evasion rates than interregional ties between male rescuers. In sum, this section revealed that interregional networks increased Jewish evasion and were more likely to be produced by women than men. Furthermore, the protective effect of interregional networks produced by women was stronger than the protective effect of interregional networks produced by men. How can we explain this? The next sections provide an answer using in-depth case studies and dramaturgical theory. ⁴ The number of observations between Figures 1.3c and 1.3d differ because both male and female rescuers frequently initiated more than one tie. Hence, where Figure 1.3c looks at individual male and women rescuers as the unit of observation, Figure 1.3d takes ties initiated by male and female rescuers as a unit of observation.

32

INTERREGIONAL NET WORKS AND GENDER ROLES Interregional: Women

Interregional: No women # All network ties % Jews Population (Logged) % Nazi votes 1937 % Nazi spies % SIPO-SD % Auxiliary police % Nazi police % Orthodox Prot. votes % Catholics % Catholics squared –10

–5

0 5 ∆ % Survival

10

15

Figure 1.4 The effect of increasing variables one standard deviation on Jewish survival with 90 and 95% confidence intervals.

Interregional networks and the performance of gender roles Why were women overrepresented in interregional rescue networks? This question cannot be answered using statistical analysis. Instead, we utilized qualitative evidence to identify mechanisms that link gender to long-distance tie formation. We do so through an integrative mixed methods design. As a first step, we used our quantitative network data to systematize the case selection for qualitative analysis by selecting ‘on the line’ cases of women rescuers who saved large numbers of Jews and forged multiple interregional connections with rescuers elsewhere (Lieberman 2005). For these women rescuers we subsequently conducted a more in-depth study of post-war testimonies and secondary literature to inductively determine how and why they could forge and maintain connections that transcended regional boundaries. When compared to men, our analysis of qualitative evidence reveals that women could better conceal their work by strategically performing traditional feminine roles to obscure their true rescue motives. From the qualitative data, we pull out and theorize three main performances that the women rescuers used to support and protect themselves when travelling distances between provinces on rescue missions: (1) strategic motherhood and wedlock; (2) strategic selfdevaluation; and (3) strategic coquetry.

ROBERT BRAUN AND KIRAN STALLONE

33

To theorize these strategic performances, the work of Erving Goffman (1956) offers a useful starting point. Goffman draws on an analogy with the theatre to show that people are constantly engaged in what he calls ‘impression management’. Goffman distinguishes between people’s ‘frontstage’ and ‘backstage’ activities, writing that people curate ‘frontstage’ performances to present themselves as they wish to be seen by the world (the ‘audience’). Activities that occur ‘backstage’, on the other hand, are those the person carries out when there is no audience (Goffman 1956 pp. 66–86). Goffman notes that performers may present a particular frontstage image of themselves for their own benefit, or for that of others around them (Goffman 1956 pp. 10–11). If presenting a ‘false’ reality or perception of themselves, performers ‘must take care to enliven their performances with appropriate expressions, exclude from their performances expressions that might discredit the impression being fostered, and take care lest the audience impute unintended meanings’ (Goffman 1956 p. 44). Crafting a strategic performance requires appropriate knowledge of the norms, customs, and expectations in one’s social environment. Focusing on women’s strategies, Deniz Kandiyoti (2016 p. 274) writes that ‘[d]ifferent forms of patriarchy present women with distinct “rules of the game” and call for different strategies to maximize security and optimize life options with varying potential for active or passive resistance in the face of oppression’. The women rescuers (‘actors’) below had to understand the expectations of their Nazi ‘audience’ to appropriately craft their performances. What did the Nazi ‘audience’ expect from women? Research reveals Nazis had particularly conservative views and goals for women. The ideal Nazi woman was a wife and mother, and ‘Nazi leaders endorsed conservatives’ view that women’s “nature” was unsuited to academic study’ (Stephenson 2001 p. 70). Hitler announced that ‘[f ]uture motherhood must be the unalterable objective of girls’ education’ (Stephenson 2001 p. 70). Women were considered ‘emotional and instinctual’ as opposed to ‘rational and theoretical’, which were labels reserved for men (Stephenson 2001 p. 70). ‘Kinder, Ku¨che, und Kirche’ was a traditional German phrase used to describe women’s place in society.⁵ It restricted them to three conventionalized areas/duties: childrearing, the kitchen, and church, and became part of Nazi propaganda. Hitler invoked the Kinder, Ku¨che, und Kirche perspective by making a ‘call for women to return to hearth and home’ (Bridenthal 1973 p. 148). Aware of how Nazis wanted women to act, female rescuers performed the gendered roles and behaviours the Nazis considered appropriate, strategically using motherhood, marriage, coquetry, and self-devaluation to conceal their rescue

⁵ The Kinder, Ku¨che, und Kirche slogan and mentality was not a Nazi invention, and dates further back in history. Although Hitler drew on the phrase, the slogan was widely used during the Weimar Republic before Hitler came to power (Bridenthal 1973).

34

INTERREGIONAL NET WORKS AND GENDER ROLES

work. These behaviours and roles fit under what R.W. Connell calls ‘emphasized femininity’, a term she uses to discuss women’s conventionalized gender roles and behaviours. Emphasized femininity can include ‘the display of sociability rather than technical competence’, ‘fragility’, male ‘ego stroking’, ‘sexual receptivity’, ‘motherhood’, and the ‘acceptance of marriage and childcare’ (Connell 1987 p. 187). Connell (1987 p. 187) argues that ‘emphasized femininity’ is the parallel to hegemonic masculinity, and that it enables male domination. She also notes, importantly, that ‘[t]his kind of femininity is performed, and performed especially to men’.

Women rescuers in the Low Countries during World War II By playing up conventionalized gender roles and submissive feminine behaviours, women rescuers in the Netherlands enabled the continuation of male domination on the surface level. On a deeper level, however, they strategically used women’s subordinate behaviours to conceal important rescue work. Using supportive data including interviews, autobiographies, and archival entries from the Yad Vashem Righteous Among Nations Database, we take a closer look at some of these women. In order to travel beyond the narrow confines of their neighbourhoods for rescue work, the qualitative evidence demonstrates that the women exerted clandestine agency by tapping into and performing conventionalized gender roles. Through this process, they could go more easily ‘under the radar’, facing considerably fewer risks than men, who, as discussed below, the Nazis already viewed more suspiciously. Three distinct but interrelated performances emerged from our data: (1) strategic motherhood and wedlock; (2) strategic self-devaluation; and (3) strategic coquetry. From the data below, it becomes clear that the women rescuers were hardly conventional. Even though they strategically deployed gender norms for their rescue work, in their real lives they defied those very norms. They were university students, working professionals, and activists who risked their lives to oppose a dangerous dictatorship. Below, one of the women even refers to herself as a ‘tomboy’. Such backgrounds demonstrate that they were not the typical women of their decade who stayed at home to raise children and prepare supper for their husbands, and further support our argument that their behaviour was, in fact, calculated.

Strategic motherhood and wedlock First, the Dutch women rescuers strategically performed both motherhood and marriage to cover their work. Gender studies literature from non-war contexts

ROBERT BRAUN AND KIRAN STALLONE

35

offers related examples of the performance of wedlock and motherhood, although for different motives than those below. For instance, Anju Mary Paul looks at how women migrants presented themselves to their families in gendered ways as ‘dutiful daughters, caring mothers and/or supportive spouses, framing their desire to work overseas as an exclusive act of intrafamilial loyalty and self-sacrifice’ (Paul 2015 p. 183). Even though many of these women wanted to move so they could travel or pursue new careers, they knew to focus on conventionalized gender roles (wives, mothers, spouses) if they wished to successfully convince their conservative family members to approve of their moves abroad (Paul 2015 p. 283). In the wartime context, other examples underscore how women have strategically used gendered roles such as motherhood to mobilize for peace and protest war. During Argentina’s military dictatorship, the Madres de la Plaza de Mayo used motherhood while marching to discover the whereabouts of their disappeared children. In this instance, ‘motherhood empowered them to continue marching at a time when no public expression of dissent was allowed and to engage in other activities without risking being kidnapped, because mothers were not perceived as political subjects’ (Navarro 2001 p. 257). Scholars have made similar claims about the strategic use of motherhood in other violent settings, such as the Indonesian, Colombian, and Liberian conflicts (Medina 2013; Prasch 2015; Krause 2019). Although the women cited in this scholarship were actually mothers (many of the Dutch World War II women rescuers were not mothers and were instead performing motherhood), such work reveals that motherhood has been strategically emphasized for specific purposes elsewhere. Let us now look at the qualitative evidence from World War II rescue networks in the Netherlands to explore how women rescuers strategically performed conventionalized gender roles to protect Jews. First, we must note that in general, the gendered expectations of the time made women rescuers less likely to be viewed in a suspicious light while travelling with children (when compared to male rescuers). In an interview, Piet Meerburg—a leader of the Amsterdam Student Group, a rescue group created to help children escape from Amsterdam to Utrecht—confirmed this: ‘It was much more suspicious for a boy of twenty to travel with a child than a girl. It was absolutely a big difference. We went to fetch the children from the crèche together, but then the woman student accompanied them on the train to Friesland and Limburg’ (Dwork 1991 p. 53). Such rescue ‘[n]etworks assigned tasks according to skills, and most male network leaders assumed women were better suited to care for children. Jewish parents made the same association. They were more likely to trust the safekeeping of their children to a Christian woman than to a Christian man. . . . Network leaders knew that and assigned the job of prying Jewish children from their parents to women’ (Fogelman 2011 p. 249). Rebecca van Delft, a woman who was part of the Dutch NV (Naamloze Vennootschap or ‘anonymous venture’) rescue group and frequently travelled with Jewish children to Limburg, stated: ‘For young men it was very dangerous to do

36

INTERREGIONAL NET WORKS AND GENDER ROLES

such a thing by train, because German soldiers always asked young men to show their identification cards: young men should be working in factories in Germany. But young women they would not so easily suspect—and indeed, during the time I did the job I was never stopped by a German soldier’ (Dwork 1991 p. 41). Such data reveal that sending women to travel with children was a rescue network strategy based on the conclusion that men were more likely to be stopped and arrested than women. However, the women took further steps to protect themselves and the Jewish children they were funnelling between regions by actually performing motherhood. An de Waard studied law in Utrecht and was a member of the Utrecht Children’s Committee, a key rescue group. She was recruited by Anne Maclaine Pont, a fellow art history student (Flim 2018 p. 190).⁶ When she travelled by train with children, An ‘would turn her ring around to give the impression that she was a married woman traveling with her child’ (Flim 2018 p. 77). This is evidence of the strategic nature of women’s actions and the performance of both marriage and motherhood to deceive Nazis and other onlookers, satisfying the gendered expectations they held for women at that time. There are other related examples of women members of the Utrecht Children’s Committee who engaged in strategic motherhood while transporting Jewish children between cities. For instance, Alice Brunner often helped take Jewish children from Utrecht to Friesland and Limburg (Flim 2018 p. 180). A soldier once came into her train car to ask about a child. After Brunner claimed she was the mother, he complimented her for her ‘beautiful dark-skinned Aryan baby’ (van Rens 2013). Ankie Stork also travelled throughout the country to Nijverdal, Hellendoorn, and Ommen, and Hetty Voute established twelve different distribution offices around the country, and escorted children to safe houses (Yad Vashem 2020). Although we do not have detailed descriptions of how these last women behaved while travelling, it is likely they, too, were engaged in performance staging of some sort.

Strategic self-devaluation Strategic self-devaluation was deployed by women rescuers to get others to underestimate their potential. By emphasizing incompetence, they satisfied the gendered expectations Nazis held for women: passive and uneducated. The performance of incompetence served them to accomplish their goals. ⁶ While not all rescuers came from higher education backgrounds, many became involved in rescue work through their university networks. University student associations, unions, sororities, fraternities, and shared dormitories were popular places to recruit.

ROBERT BRAUN AND KIRAN STALLONE

37

In Weapons of the Weak, James Scott looks at ‘feigned ignorance’ as everyday resistance that peasants used to push back against restrictions posed by dominant classes (Scott 1985). While Scott looks at both men and women who perform feigned ignorance, other scholars have looked at the gendered performance of incompetence. For instance, Silvia Gherardi and Barbara Poggio examine the ‘tactical use of “female” stereotypes’ in male-dominant organizational workplace settings, and find that women may feign ignorance ‘to survive and, perhaps, get ahead in settings, which tended to be hostile towards them, or at any rate, treated them with suspicion’ (Gherardi and Poggio 2001 p. 254). In other words, performances of ignorance and inferiority were used as a risk-management strategy. The rescue work of Diet Eman powerfully demonstrates how effective this strategy could be in a wartime setting to reduce risks. Eman was a rescue worker from The Hague who trained in business and languages, and after the war worked as a head nurse at a hospital in Venezuela (Eman 1994 p. 389). During the war, Eman travelled throughout the country to distribute illegal documents and find safe hiding places for Jews. One day, she was travelling by train on an important mission back to The Hague, carrying blank IDs, ration cards, and resistance orders for Allied pilots. The Gestapo got on the train and started checking IDs. They required her to get off, walking her towards a tunnel: When we got into that tunnel and he [the soldier] started talking, I began to act really stupid. I thought to myself, Diet, your best chance is to be really dumb. You haven’t the slightest idea how you could have received a bad ID. So I said, ‘You know, sir, when you search those trains, what exactly are you looking for?’ I talked to him as if I didn’t have a clue as to what the whole business was about’ (Eman 1994 p. 188).

When taken to prison that same day and the Nazis seemed to have no evidence to hold her, Diet says: ‘I realized then that I had convinced him that I was really stupid’ (Eman 1994 p. 194). After her arrest, she notes she continued with the act: ‘I was playing the scared, stupid maid because I thought it was my only hope’ (Eman 1994 p. 247). To save herself, she worked hard to ‘convince them that [she] was stupid, uneducated, and very scared’ (Eman 1994 p. 277). She notes that this was challenging, and that her real character was markedly different: ‘I was always a tomboy’ (Eman 1994 p. 3). Hein, the man who would later become her fiancé, confirmed this in his journal: ‘Diet is a tomboy, and mischievous’ (Eman 1994 p. 9). By strategically creating an undervalued image of herself, Diet did not seem like a suspicious person to the Nazis. Her performance fit with Nazi views of women as uneducated and incompetent. Mieke Mees and Iet Van Dijk engaged in similar self-devaluation strategies. Mieke and Iet were rescuers from the Amsterdam Student Group, and ‘took away more children than anyone else’; Iet mostly travelled north to Friesland, and Mieke

38

INTERREGIONAL NET WORKS AND GENDER ROLES

went south to Limburg (Flim 2018 p. 160). Both women were medical students. They would pick up Jewish children from the Plantage Kerklaan neighbourhood in Amsterdam, where they would remove their stars and dress them in plain clothing (Flim 2018 p. 60).⁷ Iet was a ‘blonde, short, and exceptionally driven young woman’ (Flim 2018 p. 31). Carefully crafted performances made it possible for Iet to take many children across the country. Nina Mesdag, a source who knew Iet while she was a rescuer, stated: ‘Iet could look so incredibly stupid. And that saved her. I mean, she looked so stupid getting on the train—as if she didn’t have all her marbles . . . so they let her go’ (Flim 2018 p. 96). In a historical biography about Iet’s life, her nephew Jan confirmed his aunt’s strategic self-devaluation tactics, noting she avoided arrests in multiple risky situations because she ‘had a talent for looking really stupid’ (Van Dijk 2018 p. 84). In such situations, her nephew writes that she played up her image as an extremely naïve ‘stupid blonde’ (Van Dijk 2018 p. 168). Their wording—specifically that Iet could look stupid—suggests that Iet’s ‘stupid’ external appearance was deliberate. It implies that Iet, a high-achieving medical student, only looked as if she were stupid and without her marbles, when the reality was quite different.

Strategic coquetry The third strategy that comes through via our in-depth exploration of female rescuers is strategic coquetry. Watkins et al. (2013 p. 174) argue that ‘scholars have been reticent to talk about the use of sexuality as an instrumental tool of social influence that can be used to gain compliance, assets, favour, or any other desired end’. Addressing this gap, they focus on strategic sexual performances, which can include kissing, flirtation, winking, and banter, among others (Watkins et al. 2013 p. 175). When engaging with heterosexual men, women ‘may trade on subtle sexual performances, such as flirtation, that make men feel masculine, potent, and desirable. In exchange, men may offer women tangible assets, along with favours or compliments’ (Watkins et al. 2013 p. 177). This ties to the work of Edward Jones and Thane Pittman, who, elaborating on Goffman’s dramaturgical perspective, use the word ‘ingratiation’ to discuss strategies that an actor may use ‘to achieve the attribution of likeability’, which will then help the actor with a different hidden ulterior motive (Jones and Pittman 1982 p. 236). We are learning that such strategies also play out in war. For instance, Mats Utas discusses women’s ‘tactical agency’ in the Liberian wartime context, noting that women civilians carefully developed and maintained relationships with armed actors to ensure their own safety and protection (Utas 2005). ⁷ During the Holocaust, German authorities legally required all Jews to wear yellow stars (the Star of David) in order to isolate and differentiate them from the non-Jewish population.

ROBERT BRAUN AND KIRAN STALLONE

39

During the Nazi regime in the Netherlands, the story of Mieke Mees reveals how this worked in practice. Mieke Mees engaged in strategic coquetry and put forward a ‘frontstage’ image of herself while travelling on a mission to rescue Hanna van de Voort, a fellow member of the Amsterdam Student Group who was captured by the Nazis. In July 1944, Hanna was arrested by the Sicherheitspolizei, the German security police, and was interrogated in Eindhoven. In response to this arrest, Piet Meerburg, a leader of the Amsterdam Student Group, ‘sent the beautiful Mieke Mees, to try to obtain Hanna’s release’ (Yad Vashem 2020). Mieke was sent on this rescue mission specifically because of her looks. Once there, ‘she managed to arouse the interest of the Sipo commander and made a date to go out with him on the condition that he release “that silly woman from Tienray”’ (Yad Vashem 2020). While speaking with the commander, Mieke ‘made it clear that she was rather put out about the imprisonment of a naive, simple woman from Tienray who had nothing to do with anything. As a result, Hanna was released. Mieke did not show up for her date with the SD officer’ (Flim 2018 p. 126). In this instance, Mees strategically used her external appearance and an ‘ingratiation’ performance strategy to get the commander to make an exchange. She also employed a gendered narrative, referring to her rescue colleague Hanna as a ‘silly woman’ to convince the commander that her rescue colleague was not who the Nazis thought she was. Like Mieke Mees, Margot Lawson also clearly describes strategic flirtation during the Nazi occupation of the Netherlands. Margot grew up in a wealthy family, was a highly educated aspiring diplomat, and described herself as ‘very selfconfident’ (Monroe 2013 p. 62). During World War II, she moved to Amsterdam to protest Hitler’s policies, obtained Dutch citizenship, and began sheltering and protecting Jews (Monroe 2013 p. 16). Margot discusses a relationship she had with a Gestapo commander who was a high ranking general, noting that this relationship was strategic. When interviewed about whether she was ‘friendly with this man all during the war’, Margot responds: ‘Oh, was I! Sure, because I had to. . . . Oh, yes. We had an affair! I had to. I had to’. The interviewer proceeds: ‘How were you able to do that?’ Margot responds: ‘I was very young, and I just …. [Margot shrugged.] I saved a lot of people, a lot of people’ (Monroe 2013 p. 72). This quote reveals Margot’s awareness of the sacrifice she was making to protect Jewish people from harm by strategically involving herself with the commander.

Conclusion Directly after World War II, documentation and analysis of resistance predominately looked at acts that were ‘public and political’ in nature, a focus that ‘reserve[d] the honor of being a resister primarily to men’ (Dwork 1991 p. 254). From academic work on more recent conflicts, we are beginning to learn about

40

INTERREGIONAL NET WORKS AND GENDER ROLES

the multiple and varied roles that women can play in resistance during conflicts and wars. Such scholarship highlights that women contribute to broader resistance work by behaving in a low-profile manner in order to achieve crucial results (Utas 2005; Parkinson 2013; Viterna 2013; Krause 2019; Hume and Wilding 2020). The data above show that by strategically tapping into conventionalized gender roles, women are in a unique position to exert protective agency in violent settings. Using a mixed methods approach, this chapter focused on the historical World War II context to examine the clandestine ways in which Dutch women exerted protective agency. Through the calculated performance of gendered roles, Dutch women anchored national resistance movements by maintaining social bridges between remote localities to enable the rescue and protection of Jews. In other words, through strategic roleplaying, they facilitated the long-distance transportation of goods and people. This chapter showed how a dramaturgical perspective on gender and civilian resistance enhances our understanding of what part women play in resistance movements. Even when confronted with much stronger opponents who have the power to impose norms of appropriate behaviour, women can facilitate resistance by strategically performing roles that enhance the security of rescue movements as a whole. In the context of war and militarization: (1) women are less likely to be seen as political agents and combatants and feminized roles allow for clandestine activity because they attract less suspicion; and (2) women can draw on existing skills and knowledge from non-war settings in order to strategically deploy emphasized femininity to exert agency and enhance security during war. Existing research emphasizes that rescuers must ascertain a certain cultural grammar to engage in altruistic resistance (Monroe 2013), and this chapter similarly indicates that mastery of different cultural scripts is needed to successfully pull off challenging resistance work. Second, the chapter draws attention to the unique ways in which gender shapes the structure of social networks in general and resistance networks in particular. By performing traditional feminine roles in a way that obscured the true motives behind long-distance travel, women were well-positioned to produce important social bridges between different localities. With the evidence presented, this chapter reveals that scholars of civilian resistance should, in addition to the cultural-normative dimension of gender, also focus on its relational underpinnings and networked effects. More generally, the project draws attention to the importance of local civilian networks in protecting victims of violence. While these networks are instrumental for humanitarian assistance because they have large local constituencies on the ground that are well-positioned to respond promptly, they seldom get the same amount of scholarly attention as foreign interventions and the work of external organizations. However, as both the chapters by Paddon Rhoads and Gorur as well

ROBERT BRAUN AND KIRAN STALLONE

41

as Howe point out in this volume, the two interact in producing civilian protective agency (Paddon Rhoads and Gorur within volume; Howe within volume). Together, the findings from this chapter also make a broader contribution to existing work on civilian agency during conflict. They underscore the importance of looking at women’s agency ‘within the context in which women are able to act’, even if such behaviours are ‘small acts and quiet politics’ that would otherwise be overlooked (Hume and Wilding 2020 pp. 3–4). The performances we analyse here were not necessarily visible to the average observer, but they were crucial for the continuation of collective resistance. Moving forward, we must continue to hone our understanding of the complexities of gendered agency in wartime by identifying the unique and easily missed ways in which women act in war.

References Baines, E. (2016). Buried in the Heart: Women, Complex Victimhood and the War in Northern Uganda. Cambridge Studies in Law and Society. Cambridge: Cambridge University Press. DOI: 10.1017/9781316480342 Braun, R. (2018). ‘Minorities and the Clandestine Collective Action Dilemma: The Secret Protection of Jews during the Holocaust’, American Journal of Sociology, 124/2: 263–308. The University of Chicago Press. DOI: 10.1086/699651 Braun, R. (2020). ‘Religious Dominance and Empathy’, Theory and Society, 1–29. Springer. Bridenthal, R. (1973). ‘Beyond Kinder, Ku¨che, Kirche: Weimar Women at Work’, Central European History, 6/2: 148–66. DOI: 10.1017/S0008938900015934 Carpenter, R. C. (2003). ‘“Women and Children First”: Gender, Norms, and Humanitarian Evacuation in the Balkans 1991–95’, International Organization, 661–94. JSTOR. Coleman, J. S. (1988). ‘Social Capital in the Creation of Human Capital’, American Journal of Sociology, 94: S95–120. University of Chicago Press. Connell, R. W. (1987). Gender and Power. Stanford: Stanford University Press. Croes, M., and Tammes, P. (2004). Gif laten wij niet voortbestaan. Nijmegen: Aksant. Della Porta, D. (2013). Clandestine Political Violence. Cambridge, UK: Cambridge University Press. Dwork, D. (1991). Children with a Star: Jewish Youth in Nazi Europe. New Haven, Connecticut: Yale University Press. Eman, D. (1994). Things We Couldn’t Say. Grand Rapids: Eerdmans Publishing Company. Finkel, E. (2017). Ordinary Jews: Choice and Survival During the Holocaust. Princeton, NJ: Princeton University Press. Flim, B.-J. (2018). Saving the Children: History of the Organized Effort to Rescue Jewish Children in the Netherlands, 1942–1945. Potomac, Maryland: CDL Press. Fogelman, E. (2011). Conscience and Courage: Rescuers of Jews during the Holocaust. New York, NY: Doubleday Anchor. Gherardi, S., and Poggio, B. (2001). ‘Creating and Recreating Gender Order in Organizations’, Journal of World Business, 36/3: 245–59. Elsevier.

42

INTERREGIONAL NET WORKS AND GENDER ROLES

Goffman, E. (1956). ‘The Presentation of Self in Everyday Life’, Social Sciences Research Centre, 5. Hume, M., and Wilding, P. (2020). ‘Beyond Agency and Passivity: Situating a Gendered Articulation of Urban Violence in Brazil and El Salvador’, Urban Studies, 57/2: 249–66. SAGE Publications Ltd. DOI: 10.1177/0042098019829391 Jones, E. E., and Pittman, T. S. (1982). ‘Toward a General Theory of Strategic SelfPresentation’, Psychological Perspectives on the Self, 1/1: 231–62. Kandiyoti, D. (2016). ‘Bargaining with Patriarchy’, Gender & Society. Sage Publications. DOI: 10.1177/089124388002003004 Krause, J. (2019). ‘Gender Dimensions of (non) Violence in Communal Conflict: The Case of Jos, Nigeria’, Comparative Political Studies, 52/10: 1466–99. Lieberman, E. (2005). ‘Nested Analysis as a Mixed-Method Strategy for Comparative Research’, The American Political Science Review, 99/3: 435–52. McAdam, D. (1986). ‘Recruitment to High-Risk Activism: The Case of Freedom Summer’, American Journal of Sociology, 92/1: 64–90. University of Chicago Press. Medina, R. M. (2013). ‘La politización de la maternidad ante la impunidad en Colombia: el caso de las madres de Soacha’, Revista Internacional de Pensamiento Polı´tico, 8: 41–53. Monroe, K. R. (2013). The Hand of Compassion: Portraits of Moral Choice During the Holocaust. Princeton, NJ: Princeton University Press. Moore, B. (1997). Victims and Survivors: The Nazi Persecution of the Jews in the Netherlands, 1940–1945. London: Arnold. Morris, A. D. (1986). The Origins of the Civil Rights Movement. New York, NY: Free Press. Navarro, M. (2001). ‘The Personal is Political: Las Madres de Plaza de Mayo’, pp. 241–58. In Power and Popular Protest: Latin American Social Movements, Updated and Expanded Edition, edited by Susan Eckstein and M. A. G. Merino. Berkeley: University of California Press. Paddon Rhoads, E., and Sutton, R. (2020). ‘The (Self ) Protection of Civilians in South Sudan: Popular and Community Justice Practices’, African Affairs, 119/476: 370–94. Parkinson, S. E. (2013). ‘Organizing Rebellion: Rethinking High-Risk Mobilization and Social Networks in War’, American Political Science Review, 107/3: 418–32. JSTOR. Paul, A. M. (2015). ‘Negotiating Migration, Performing Gender’, Social Forces, 94/1: 271–93. DOI: 10.1093/sf/sov049 Prasch, A. M. (2015). ‘Maternal Bodies in Militant Protest: Leymah Gbowee and the Rhetorical Agency of African Motherhood’, Women’s Studies in Communication, 38/2: 187–205. Taylor & Francis. Sageman, M. (2011). Leaderless Jihad: Terror Networks in the Twenty-First Century. Philadelphia, PA: University of Pennsylvania Press. Scott, J. C. (1985). Weapons of the Weak: Everyday Forms of Peasant Resistance. New Haven: Yale University Press. Seawright, J. (2016). Multi-Method Social Science: Combining Qualitative and Quantitative Tools. Cambridge, England: Cambridge University Press. Semelin, J., Andrieu, C., and Gensburger, S. (eds). (2014). Resisting Genocide: The Multiple Forms of Rescue. Oxford, England: Oxford University Press. DOI: 10.1093/acprof:oso/9780199333493.001.0001 Shapiro, J. N. (2013). The Terrorist’s Dilemma: Managing Violent Covert Organizations. Princeton, NJ: Princeton University Press.

ROBERT BRAUN AND KIRAN STALLONE

43

Staniland, P. (2014). Networks of Rebellion: Explaining Insurgent Cohesion and Collapse. Ithaca, NY: Cornell University Press. Stephenson, J. (2001). Women in Nazi Germany. Abingdon, England: Routledge. Utas, M. (2005). ‘Victimcy, Girlfriending, Soldiering: Tactic Agency in a Young Woman’s Social Navigation of the Liberian War Zone’ Anthropological Quarterly, 78/2: 403–30. Van Dijk, J. (2018). Iet. Utrecht, Netherlands: Pharos Uitgevers. van Rens, A. (2013). Vervolgd in Limburg: Joden en Sinti in Nederlands-Limburg tijdens de Tweede Wereldoorlog., Vol. 76. Hilversum, Netherlands: Uitgeverij Verloren. Viterna, J. (2013). Women in War: The Micro-Processes of Mobilization in El Salvador. Oxford, England: Oxford University Press. Watkins, M. B., Smith, A. N., and Aquino, K. (2013). ‘The Use and Consequences of Strategic Sexual Performances’, Academy of Management Perspectives, 27/3: 173–86. Academy of Management Briarcliff Manor, NY. Yad Vashem. (2020). ‘Amsterdam Student Group’. Yad Vashem: The World Holocaust Remembrance Center. Retrieved 11 November 2020, from

2 The ties that bind Civilian adaptation and social connectedness during the Syrian Civil War Kimberly Howe

Introduction The civil war in Syria, which began in 2011, has been characterized as one of the most deadly and destructive conflicts of our time. Conflict-related violence has disproportionately and brutally affected civilians, with staggering humanitarian consequences. While the Government of Syria (GoS) has regained military primacy as of this writing, violence against civilians and widespread displacement continues (OCHA 2021a). In 2021, nearly 90 per cent of the population were living below the poverty line, with over 13 million in need of humanitarian assistance (OCHA 2021b). During the most violent years of the war, Syrians living in opposition-controlled areas were subject to targeted and indiscriminate attacks by Syrian, Russian, Turkish, and Iranian forces, with the GoS being responsible for the largest proportion of killings (Syrian Network for Human Rights 2019; Violations Documentation Center in Syria 2020). Civilians have also been directly targeted by members of the opposition or caught in the crossfire, as various armed groups vied for territorial control and engaged in counter-terror operations. The conflict—directly and indirectly—has prevented civilians from accessing food, water, health care, education, livelihoods, and other basic needs. And international actors, either through lack of political will, lack of access, or inadequate resources, have been unable to protect Syrians from violence or provide them with adequate humanitarian support (Niland et al. 2015; Howe 2016). While the agency of individuals, families, and communities is severely circumscribed in such a context, civilians in Syria have been able to adapt in various ways. However, our understanding of civilian protective agency during the Syria civil war is limited to date. In recent years, there has been an increasing focus on understanding civilian protective agency, or the strategies used and actions taken by civilians— individually and communally—to manage the vagaries of civil war. In particular, scholars have studied the wide range of relationships that civilians forge with

Kimberly Howe, The ties that bind. In: Civilian Protective Agency in Violent Settings. Edited by: Jana Krause, et al., Oxford University Press. © Kimberly Howe (2023). DOI: 10.1093/oso/9780192866714.003.0003

KIMBERLY HOWE

45

armed groups in order to keep themselves safe (Baines and Paddon 2012; Jose and Medie 2015; Arjona 2016; Kaplan 2017; Masullo 2021). Relatedly, research has elucidated the individual and community factors that prevent conflict escalation and support processes of non-violence (Krause 2019; Carpenter 2012). The Syria case, however, underscores the necessity of broadening the concept of protection in its traditional form—recognizing that civilians face multiple threats during war that span beyond violence—including exposure to crime, meeting their basic needs, and accessing livelihoods, among others. This contribution follows the work of Paddon Rhoads and Sutton (2020) in South Sudan, where they consider the full range of civilian responses to harms emanating from conflict—whether direct or indirect. This inquiry examines the lived experiences of Syrians and identifies a wide range of adaptive strategies that individuals, families, and communities employ to protect themselves against violence perpetrated by conflict actors, as well as harms stemming from the deprivation of basic needs. I draw on the work of Jana Krause (2018 p. 8) to conceptualize adaptation to conflict, which is conceived as ‘social knowledge and learning, self-organization, anticipation, and imagination’—a set of facilitative factors that enable civilians to adapt to adversity including war. While each of these factors are found to be deeply embedded in the protective strategies developed by Syrian civilians, this study also revealed the central importance of social connectedness for adaptation and survival. Social connectedness can be defined as the ‘sum of people’s linkages including social networks on which they can draw; the extent and strength of those networks and their ability to mobilize resources; the nature of obligation that such networks carry; and the reciprocity presumed in terms of collective risk and mutual support’ (Humphrey et al. 2019 p. 9).¹ In the following pages, I delineate how Syrian civilians engaged² in individual and spontaneous action such as short-term strategic movement and displacement to manage threats of violence. Through social networks, information and knowledge were shared in order to decrease risk, secure safe passage, and erect protected spaces. Within communities, volunteer networks organized alert systems against oncoming aerial and ground attacks, created search and rescue units, and built underground hospitals and schools. Each of these initiatives relied on social connectedness, communications systems, and continuous learning to support adaptive processes within an extremely lethal environment. Such findings link with the recent scholarship on the importance of social connectedness for coping and resilience in fragile settings. For example, Schon’s (2020) work in Syria revealed that civilian survival strategies were largely determined by social ¹ These conceptualizations have been borrowed from Daniel Maxwell et al. (2016). ² I have opted to use the past tense in this chapter given that data were collected between 2013 and 2020. However, given the ongoing violence and material deprivation in Syria, it is likely that many of these strategies are utilized today.

46

CIVILIAN ADAPTATION AND SOCIAL CONNECTEDNESS

networks and resource endowments. In a very different context, Maxwell and Majid (2016) show the decisive role that social connectedness played in surviving the 2011 Somalia famine. In South Sudan and Uganda, a series of studies show how social connectedness allows conflict-affected households to access economic and non-material support, and how it influences decisions about marriage and displacement (Kim et al. 2020; Stites and Humphrey 2020). Beyond protection from violence, this study also reveals the conditions under which certain households were able to adapt their livelihoods during the conflict, and thus benefit from more disposable income and greater food security. Among other characteristics, these households were significantly more likely to be located in ‘inclusive’ communities—that is, communities that provided leadership opportunities for women and youth and encouraged their participation in civil society. Inclusive communities are also marked by girls’ access to higher education, and the presence of organizations concerned with women’s issues. This type of gender and youth inclusion in wartime supports livelihood adaptation by improving freedom of movement, expanding access to income generation opportunities, deepening social networks, and acquiring information and knowledge related to conflict dynamics that can improve self-protection. Such insights dovetail with existing research that underscores the importance of gender inclusion for positive peace outcomes (Ellerby 2013; Chenoweth 2019; Krause 2019) and the prevention of violence (Arjona 2016; Kaplan 2017; Krause 2018). These findings also connect to scholarship calling attention to the gendered dimensions of adaptation and social connectedness (Fordham 2004; Fulu 2007; Humphrey et al. 2019; Krause 2019; Kim et al. 2020). This chapter likewise analyses how and why local humanitarian action developed in Syria in the way it did. The trajectory of Syrian civilian humanitarian response over time was both a function of extreme local need and a shift to a remote management model by the international community. Local humanitarians rolled out a successful response because they were well-connected to their communities, built complex information systems, and shared knowledge— all of which facilitated their ability to maintain access in an extremely volatile conflict space. Such findings complement the existing body of literature on the localization of humanitarian action and civil society development in conflict (Gingerich and Cohen 2015; Wall and Hedlund 2016; Haddad and Svoboda 2017; Hernández Cárdenas and Tello Méndez 2017; Pereda 2017; Howe and Stites 2019). A range of implications can be drawn from these findings. I directly address the international humanitarian community, which has either ignored or acted counterproductively towards the life-saving protective actions taken by Syrians during the conflict. Given the importance of social connectedness for adaptation, a natural conclusion is to invest in a range of communication systems at the household, community, and organizational levels. Secondly, the centrality of age and gender

KIMBERLY HOWE

47

inclusion for successful livelihood adaptation suggests that international interveners should prioritize inclusive programming when supporting civil society and governance programmes. Lastly, I urge a strategic rethink of how international humanitarian organizations partner with Syrian non-governmental organizations (SNGOs).

Researching Syria This chapter pulls from a mosaic of studies conducted between 2013 and 2020, where I have worked independently, in conjunction with colleagues, or in support of humanitarian and stabilization organizations seeking to improve programming inside Syria. Inquiries were grounded in qualitative and quantitative methodologies, cross-sectional and longitudinal in design, and conducted both remotely and in person. While each study was designed around its own distinct set of research questions,³ I have returned to the original data sets afresh. Here, data have been reanalysed to draw comparisons across studies, to highlight distinctions between them, and cull insights related specifically to civilian adaptation. The findings presented here are based on a reanalysis of data collected from the following sources: (1) self-administered surveys with a network of Syrian researchers located in five governorates focusing on the civilian protective and adaptive strategies observed in their own communities⁴; (2) surveys and qualitative interviews with households in Syria to understand resilience during conflict, sponsored by an international NGO (Howe et al. 2018; Howe and Patros 2020)⁵; (3) this same organization later commissioned a mixed-methods study to understand the development of Syrian humanitarian civil society and the impact of their support on these organizations (Howe and Sheely 2019);⁶ (4) Longitudinal qualitative research with members of Syrian and international organizations involved in cross-border assistance between Turkey and Syria (Howe and Stites 2019);⁷ and (5) an in-depth multi-year inquiry about the rise of Syrian rebel governance ³ The main publications arising from each study are listed in subsequent citations. ⁴ Thirty-seven self-administered surveys were collected across seventeen opposition-controlled communities in 2016. Funded by the Norwegian Ministry of Foreign Affairs. ⁵ Data collected in 2017. Surveys conducted with 350 community key informants in 115 communities, and 1,168 households in 124 communities. Eighty-two qualitative interviews conducted in ten communities within the quantitative sample. Study locations in opposition-controlled areas in north, south-central, and northeast Syria. ⁶ Data collected in 2019. Qualitative interviews with sixty-eight representatives of Syrian civil society, some organizations followed over six months. Online survey with 900 respondents of civil society members inside Syria or working on the Syria conflict. ⁷ Qualitative interviews with representatives of forty-six organizations operating in oppositioncontrolled Syria. Ten organizations served as longitudinal organizational case studies over twelve months. Data collection between 2013 and 2014.

48

CIVILIAN ADAPTATION AND SOCIAL CONNECTEDNESS

and the influence of Western support to these bodies (Mukhopadhyay and Howe forthcoming).⁸⁹ The majority of data were collected in opposition-controlled Syria, which includes Kurdish-majority areas, besieged communities, and locations controlled by factions of the Free Syria Army, Al-Qaeda affiliates (Jabhat al Nusra, later Hay’at Tahrir al-Sham), and other armed Islamic groups. Data were also collected through interviews with members of Syrian civil society located in Iraq, Jordan, Lebanon, and Turkey, who were active in these opposition-controlled areas. Data were collected in communities acutely impacted by conflict, including at times of active violence and deprivation. Findings presented here do not include information collected in geographic areas controlled by the GoS or the so-called Islamic State (ISIS).¹⁰ As a result, what is presented may not be inclusive or representative of civilian adaptive practices in all locations. Additionally, as findings draw on a diverse set of data sources, information is not always symmetrical between communities or households included in this study, thus limiting generalizability across the country.

The ties that bind: Civilian adaptation and social connectedness Individual and spontaneous action for self-protection Syrians living through the war have devised a complex set of adaptive strategies to manage continuous threats of violence, whether perpetrated by the Assad government and its allies, armed groups affiliated with the opposition, ISIS, or criminal gangs. Each possibility for violence engendered a differentiated protective strategical response, whether coming from aerial bombardments, assassinations, abductions, car bombs, shelling, or unexploded ordnances (UXOs). Strategic movement was the most frequently cited planned action for managing this wide range of violent threats, regardless of the perpetrator. Families restricted their own movement by pulling their children from school, and avoiding mosques, markets, and hospitals, all of which were regularly targeted through air attacks. Faced with bombing, individuals and households temporarily moved away from urban centres to agricultural spaces, to larger, more sturdy buildings, or to abandoned buildings. Many households described their reliance on underground shelters. At times, such space was limited to a single family, but more often, it was ⁸ Cross sectional and longitudinal interviews with members of Syrian civil society and international interveners from 2013 to 2020. ⁹ Across these studies, each research participant was involved in a detailed informed consent process, and no deception was used. ¹⁰ However, some data were collected from communities that had previously been under ISIS control or government control.

KIMBERLY HOWE

49

shared across families if not neighbourhoods. These secure spaces were a common meeting point—not just in pursuit of physical safety, but a venue for social solidarity. New connections and mutually supportive networks were formed here, and in some cases, the existing social structures of a community were reconfigured for mutual benefit. These spaces also became locales for sharing knowledge and information about war dynamics. As so eloquently described by this male youth from a besieged suburb of Damascus: People have shown their true colours . . . As a result of bombings, you are forced to enter homes—homes of people who you could’ve been fighting with a year earlier—but this person has a basement. He opens his house to you . . . when you enter his house, you find 20–30 even 40 people . . . despite everything—and all the difficulties—you feel happy and proud of people coming together.¹¹

Households also used more permanent forms of displacement when their communities became sites of intra-opposition contestation, when livelihood opportunities were too sparse for survival, or where alternative locations held the promise of safety. This finding falls in line with Schon’s (2020) work on the Syrian conflict, which demonstrates that migration is a main survival strategy for civilians, a decision based on both motivation and opportunity. Rates of internal displacement in Syria remain the highest in the world (Internal Displacement Monitoring Center 2020). In 2018, seven years into the war, a study with a representative sample found that the average number of times a household displaced was 3.7 since 2011 (Howe et al. 2018). Displacement was both planned and unplanned, but two common themes arose from interviews. Households made decisions about movement based on the location of friends and relatives. In addition, individuals and households which decided not to leave created an anchor for others to stay, as expressed by this adult male from South Central Syria: ‘When I see my brother or my neighbour, I feel reassured. However, if I see him packing his bags and leaving, I’d think twice. Him staying is more or less a motivation for me to stay put and cope’.¹² For local travel, calculations about safe movement were made in conjunction with extended families, neighbours, and friends. Information about checkpoints— their location, who was in charge, and what they were demanding for passage— was readily diffused through social media and WhatsApp, in real time. If checkpoints needed to be crossed, some families sent out women and elderly men, while opting to keep their conscriptable-aged sons at home, on the assumption that women and the elderly were less vulnerable. Other families decided that movement was secure if one was willing to bribe those manning checkpoints. Syrian ¹¹ Interview with male youth south-central Syria, 2017. ¹² Interview with adult male south-central Syria, 2017.

50

CIVILIAN ADAPTATION AND SOCIAL CONNECTEDNESS

men and women alike described that they generally avoided travel by foot, particularly at night, and preferred to move in small groups of three to five, rather than individually. In addition, civilians described a heightened vigilance based on social learning, often transmitted across communities. A poignant example was provided by this thirty-year-old male from Daraa: ‘We have developed an eye for plastic cans that appear to be colourful and out of place. I don’t know what is that skill exactly, but we can smell a trap now. Even a five-year-old knows not to kick any random object on the sidewalk’.¹³

Community organization and action against violence In addition to households and families, communities organized to protect themselves against the threat of conflict-related violence. The most institutionalized initiatives were found in communities that were regularly bombed by the Syrian regime and allies—including locations that were besieged and many opposition strongholds in the north. Such initiatives relied on the existing skills of individuals and/or their willingness to engage in dangerous activities. An organization’s ability to be effective was dependent on a shared and continuous flow of knowledge and information, as well as the ability to adjust and be nimble in the face of shifting threats. Furthermore, each activity was embedded within social networks that were underpinned by trust. Civil Defense Units (CDUs) were cited as a community-level organization instrumental in providing civilian protection against violence. Known to the outside world as the White Helmets, CDUs consisted of a network of 3,000 volunteers, mostly self-organized, often locally supported, and periodically funded by external donors.¹⁴ CDUs were active in search and rescue efforts, particularly in areas of heavy bombardment. CDU volunteers were considered enemies of the state, and were regularly targeted through ‘double-tap’ bombings, whereby the regime and its allies would drop a single bomb on a community, wait for rescue workers to arrive, and then target them with a second bomb.¹⁵ Half of the 252 volunteers killed during rescue operations lost their lives in double-tap bombings (The White Helmets 2020b). Despite the immense human loss and ongoing danger, the organization itself and the individuals within it continued their operations, demonstrating an ability to rapidly learn and adapt to the violent strategies of conflict actors. ¹³ Self-administered questionnaire #24. ¹⁴ An overview of their work can be found on their website: https://www.syriacivildefence.org/en/ our-work/an-overview/; or https://www.whitehelmets.org/en/. ¹⁵ Two hundred and fifty-two White Helmets have been killed while working since 2013, and more than half have died as part of a double-tap strategy. See The White Helmets (2020a).

KIMBERLY HOWE

51

In my own interviews across opposition-controlled Syria, CDUs were named as the single organization that consistently saved lives following bombardments and shelling. These teams were primarily composed of male youth, although women were involved in emergency first aid. Certain locations, such as Aleppo City, had all female CDUs to assist female victims of violence. CDUs were also effective in projecting an international persona based on bravery and strict adherence to principles of neutrality. Media campaigns assisted the organization by securing essential fundraising. Information drives about the White Helmets also alerted the world to the severity of the conflict for Syrian civilians (and the regime’s implication therein)—culminating in their 2016 nomination for the Nobel Peace Prize. The organization’s capacity to attract benefactors and cultivate a positive international reputation is related to their strong communication strategy as well as their local and international social networks. In addition to CDUs, early warning systems were created to notify communities of incoming bombardments or attacks. ‘Tower alert systems’, or ‘spotter networks’ were formed on top of makeshift lookouts, and guards provided warnings through walk-talkies, sirens, social media, and radio stations.¹⁶ These early warning systems were reportedly organized by men who had been defectors of the army, were current members of armed groups, or regular civilians. Defectors brought forward their military knowledge and former experiences to these contemporary posts. Some communities also constructed underground schools and medical facilities to continue basic services under the shower of regular, if not continuous, bombardment. Unlike spotter networks, these activities were heavily staffed by women who served in senior administrative roles, and as teachers, doctors, nurses, midwives, secretaries, and cleaners. Common to women in these positions, was a sense that they were a part of a greater revolutionary project, as embodied in the words of this teacher, ‘fighting illiteracy is the most effective way to fight Assad’.¹⁷ These initiatives also provided an avenue for women to join the larger opposition movement—described as an ‘outlet for females to give to the revolution’.¹⁸ The implication of gender inclusion in such roles will be discussed in subsequent sections.

Livelihood adaptation to meet basic needs Seven years into the conflict, the majority of Syrian households had lost their main income source as a result of the conflict, but a sizeable sub-group—one third—had restarted their pre-war income activity or had begun a new income generation ¹⁶ These strategies are also described in Center for Civilians in Conflict (2016). ¹⁷ Female from Daraa self-administered questionnaire #8. ¹⁸ Female from Idlib self-administered questionnaire #1.

52

CIVILIAN ADAPTATION AND SOCIAL CONNECTEDNESS

activity. The following section relies on data collected across 124 communities in opposition-controlled Syria (commonly referred to as north, northwest, and south-central Syria).¹⁹ Households that had been able to adapt their livelihoods in this way were significantly better off in terms of food security, and had more disposable income available to them.²⁰ Counterintuitively, they did not have higher pre-war endowments, meaning livelihood adapters did not have higher levels of education or wealth than other households before the war began. The following paragraphs outline the characteristics common to these households, and how this process was facilitated by access to social knowledge, learning, and the practice of anticipation, as well as reliance on social connectedness. Livelihood adapters share a cluster of common characteristics—they had significantly better access to markets, more opportunities for loans, and higher rates of social contact through analogue and digital communications. They also received remittances more regularly than households that had not adapted their livelihoods.²¹ In addition, livelihood adapters had specific age and gender characteristics and were more likely to live in inclusive communities. As will be demonstrated, each of these facilitative factors are embedded explicitly or implicitly in social connectedness. Markets and finance: Early in the conflict, the formal banking system collapsed in opposition-held Syria, requiring those who stayed to rely on informal borrowing and loan schemes. While loan sharks existed, Syrians reported that they more often borrowed from family, friends, and neighbours when in need. For lenders, they purposively supported people in their social circles because, as they explained, a base of underlying trust served as a guarantee for the loan. If the friend, family, or neighbour defaulted, the lender was persuaded that he or she had good reason.²² Across the sample of more that 1,100 respondents, less than a quarter of households received remittances from abroad, although this rate was higher than pre-conflict levels.²³ However, households that adapted their livelihoods were more likely to receive remittances than other households. Remittances in the case of Syria, as well as elsewhere, are a likely indication of the quality of social and familial networks—a phenomenon studied in other contexts (Maxwell and Majid 2016; Bryant 2019; Stites and Humphrey 2020). Market strength is also related to social knowledge, learning, and anticipation as ‘local businesses faced with crisis rely on their networks to gather market information; exchange

¹⁹ Sample size 1,168 households. ²⁰ For more details on this quantitative finding and related regression models, see Howe et al. (2018). ²¹ Such differences are statistically significant. For more details on this quantitative finding and related regression models, see Howe et al. (2018) ²² Interviews with money lenders, opposition-controlled Syria, 2017. ²³ Rates of remittance receival were higher in south-central Syria than northeast or north Syria.

KIMBERLY HOWE

53

goods, services and financing, and anticipate shifts in the operating environment’ (Humphrey et al. 2019 p. 8). Households that adapted their livelihoods were significantly more likely to live in areas with stronger markets. Communications systems: The importance of communication systems for selfprotection against violence has been made apparent in the previous section, but these systems were also necessary for livelihood adaptation. Social media platforms and related phone and internet use were described as critical resources, allowing many of the livelihood adapters to find work. Social media diffused advertisements for work and an internet connection allowed candidates to apply for positions. Male and female youth described how communication with friends and family provided them both with the encouragement to start new livelihood projects and, at times, led eventually to the necessary capital for such startups. Internet connection in some parts of the country allowed young Syrians to continue their studies via online platforms. Social media was also described as a vehicle to communicate to the world about the violence and suffering endured, and served as an outlet to connect people active in the opposition.²⁴ Social connectedness also supported the psychological well-being of Syrian civilians, particularly their ability to cope and maintain hope in the face of loss and ongoing conflict. Young adults spoke about an increased need for friendships, as well as a deepening and expansion of friendships arising from a shared experience of war. Social media facilitated communication, which for many was described as ‘constant’, and connected individuals separated by distance as well as those next door. Interviews with Syrian youth demonstrated the role of friends as a central coping strategy, a cornerstone of their identities, and a main contributor to their overall well-being.²⁵ A thread running through many interviews was well- reflected in the voice of this young man: ‘What encourages you to stay strong and what encourages self-preservation is mostly friends’.²⁶ These findings connect to Krause’s (2018) work, which demonstrates that inclusive social identity supports social cohesion and the maintenance of hope, a core aspect of social resilience.

Gender, age, and community characteristics Syrian households that adapted their livelihoods embodied specific gender and age characteristics. While pre-war income earners were generally adult males, seven years into the war, households that had at least one youth (male or female) ²⁴ Interviews with new income earners, north, northwest, and south-central Syria, 2017. ²⁵ Interviews with male and female youth, north, northwest and south-central Syria, 2017. ²⁶ Interview with adult male, south-central Syria, 2017.

54

CIVILIAN ADAPTATION AND SOCIAL CONNECTEDNESS

working were significantly better off in terms of food security and disposable income.²⁷ The fabric of the community also mattered. Livelihood adapters were statistically more likely to live in ‘inclusive’ communities—that is, communities where (1) women and youth could participate or lead governance bodies and civil society organizations; (2) there were opportunities for women and girls to access higher education; and (3) there were organizations run by women or focused on women’s issues.²⁸ The positive relationship between gender-inclusive political participation and positive outcomes, such as promoting nonviolence and supporting successful peace processes, has been well-established by other scholars (Chenoweth 2019; Krause 2019). For the Syrian case, there are several pathways that connect inclusion to livelihood adaptation. First, inclusive communities permitted greater freedom of movement for women and girls, which increased their engagement in paid work, and thus lead to an improvement in household income. Second, broader participation in higher education, civil society, and governance expanded the social networks of Syrian women and youth. These linkages multiplied avenues for sharing information, generating knowledge, and accessing resources and support. Interviews throughout opposition-controlled Syria highlighted that social connections based on quality—those accumulated through participation in education, work, civil service, or civil society—provided individuals and households with increased information about security, opportunities for income generation, access to informal loans, and forums for the exchange of ideas. Social connections also provided opportunities to participate in the revolution, and the moral support necessary to mitigate some of the loss and trauma of living through war. At the same time, no participant suggested there were negative repercussions for being socially connected. Moreover, interviews with young Syrians described how work and engagement in the broader community provided them with bumps in self-esteem—a positive process that extended to their families. This female youth described a space that she was able to access because of the war—a theme that was reflected in other interviews with male and female youth: Before I used to say, ‘I wish I were a boy’ because they have more options in life. Now, however, there is no difference between men and women. Thank God I am a woman! If I were a man, I’d probably be serving in the army, or have migrated out of the country. Now I am optimistic about life, and I have become productive. And I bring a sum of money home every month . . . this feeling is very beautiful.²⁹

²⁷ The findings presented in this paragraph are found through regression analysis of surveys conducted with 1,168 households in 124 communities in north, northeast, and south-central Syria in 2017. Regressions controlled for the number of income earners in a household. ²⁸ For more details on this quantitative finding and related regression models, see Howe et al. (2018). ²⁹ Interview with female youth, northern Syria, 2017.

KIMBERLY HOWE

55

Interviews with Syrians—male and female, youth and adults—reflected a perception that the war has expanded women’s roles, a pattern that has been documented by scholars in other locations (Wood 2008; Howe and Uvin 2009; Nı´ Aoláin et al. 2011; Berry 2018; Lazarev 2019; Braun and Stallone within volume ). Study participants indicated that even where women had been present in public life prior to the conflict, the war permitted them to take on more powerful roles and command more respect. Examples included women who had become judges, members of local governance bodies, and leaders of civil society organizations. This discourse was more pronounced in, but not exclusive to, northeast Syria, where communities are dominated by Syrians of Kurdish background. Previously you might have only seen one or two women or young women from well to do or better-connected families involved in politics. Now you see so many young women involved in politics. In making decisions both within the household and outside . . . they don’t hesitate to go to a public meeting and speak up.³⁰

Not all communities, however, saw this type of expansion. Even beyond the repressive grip of ISIS, some communities became more conservative, effectively closing off possibilities for women and girls to move freely, work, attend university, sit on local councils, or take leadership roles in civil society. The most common explanation for such a reversal was the outmigration of more liberal and educated segments of communities, which allowed the space for more conservative elements of society to dominate. Another explanation relates to protection. Citing attacks against girls and women based on their gender, influential family members and community leaders circumscribed the movement of females.

The rise of Syrian humanitarianism With limited help from the outside, Syrians living in opposition-controlled areas organized individually and collectively to provide for the basic needs of their communities. Humanitarian work became one of the most common forms of civil society participation during the conflict. The following paragraphs will trace how the success of SNGOs was contingent on their ability to adapt to a rapidly changing context through engagement with various forms of social learning, accessing knowledge, action taken in anticipation of events, and reliance on social connectedness. Data from this section arise from longitudinal case studies of a variety SNGOs and interviews with staff and leaders of such organizations.³¹ ³⁰ Interview in northeast Syria, 2017. ³¹ Interviews conducted between 2013 and 2020, these included organizations operating in north, south-central, and northeast Syria, as well as besieged areas.

56

CIVILIAN ADAPTATION AND SOCIAL CONNECTEDNESS

Humanitarianism in Syria began with the onset of the war.³² Initially composed of makeshift initiatives but becoming increasingly organized with time, individuals—males and females, youth and adults—worked as volunteers based on the skills they possessed. As with collective activities described above, humanitarians generally emerged from their natal communities and enjoyed positive social standing. They were trustworthy in terms of resource management, and their loyalties were clear—in other words, they were not informants for the regime. In Aleppo city, housewives accepted food donations and cooked in groups—soup-kitchen style—for people in the hardest-hit neighbourhoods. Engineers volunteered to repair water and electricity lines. A Syrian dentist told me how he recuperated dental instruments from abandoned clinics on the regime side of the city to create a clinic for dental emergencies on the opposition side. Medical students set up emergency medical points in neighbourhoods that had been bombed earlier in the day.³³ These volunteer networks grew over time, and organizations became more professionalized. As time passed, and the West was unable to provide meaningful political or military support for their cause, many former political activists, desperate to support their communities, became humanitarians. Syrian humanitarianism arose from two predominant sources of demand. First was the extreme humanitarian need present in civilian communities. Secondly in 2013, the international humanitarian community largely withdrew their direct operations from Syria following a spate of kidnappings and murders of NGO staff. The switch of the international humanitarian system to a remote management model led to their dependence on SNGOs to provide humanitarian assistance to most of opposition-controlled Syria (Haddad and Svoboda 2017; Howe and Stites 2019). The formation of SNGOs, and their ability to operate in such a rapidly changing and extremely dangerous environment, proved to be dependent on social connectedness and continuous adaptation based on social knowledge and learning. SNGOs were required to understand the intimate needs of their communities to provide the appropriate type of assistance. They were also dependent on continuous reliable information in order to understand the minute risks involved in distribution and movement. Effective SNGOs shipped goods across battle lines (cross-line and cross-border), negotiated access with armed groups, and largely managed to protect their staff from kidnappings and targeted killings. One SNGO director I interviewed showed me a map of Syria, marked with fifty checkpoints, and explained how they moved materials from the Turkish border to a besieged ³² The Syrian regime had repressed the development of Syrian civil society, aside from the Syrian Arab Red Crescent (SARC) which was viewed as an ancillary institution of the government. During the conflict, the SARC did not operate in opposition-controlled areas. ³³ Interviews with medical students and activists, Gaziantep, Turkey and Aleppo Province, Syria 2013–16.

KIMBERLY HOWE

57

community in Damascus.³⁴ Another described the minute negotiations needed to convince Kurdish YPG fighters to allow flour to move into a district under their military control.³⁵ A third explained how they were able to protect humanitarian assistance from being appropriated by an Al-Qaeda affiliate, through a combination of public shaming, tough negotiation skills, and guarantees of equal distribution of aid to the community.³⁶ SNGOs managed to maintain their operations because they learned the language of international donors, and how to model and retrofit themselves according to donor requirements. This process of knowledge exchange and learning was manifest in the SNGO’s ability to write proposals for international audiences, produce organigrams, recite humanitarian principles, justify financial systems, and design monitoring and evaluation plans, among others. Interviews with senior representatives of SNGOs revealed how the intense requirements of upward accountability to donors had negatively impacted the local humanitarian response in a myriad of ways (explained in the next section). However, this type of organizational isomorphism—a survival process described elsewhere (DiMaggio and Powell 1983; Barnett 2005)—is a type of adaptation that allowed at least some Syrian SNGOs to remain afloat over time, and to provide humanitarian assistance to conflict-affected communities.

Conclusions and implications for the international community This chapter sheds light on how civilians living through the Syrian civil war developed and utilized strategies of protective action in order to mitigate continuous threats from a range of violent actors, and how individuals and households meet their basic needs. Interviews with Syrian civilians showed that protection was conceived as a concept that at once embodied, but moved beyond, the threat of violence,³⁷ to include food security, livelihoods, and access to water and medicine, as well as psychological strength to manage the ongoing traumas of war. Through a multi-year research process relying on a variety of methods and data sources, the centrality of knowledge, learning, anticipation, and social connectedness to facilitate adaptation has come to light. These conditions are fundamental for supporting civilian protective agency during war. They enabled community initiatives to be effective in developing and supporting early alert systems, search and rescue networks, or the creation of safe locations to provide medical care and education. These factors also supported the development of SNGOs to provide effective ³⁴ Interview with local non-governmental organization (LNGO) representative, Gaziantep, Turkey, 2014. Also cited in: (Howe and Stites 2019) ³⁵ Interview with LNGO representative, Gaziantep, Turkey, 2014. ³⁶ Interview LNGO Representative, opposition-controlled Syria, 2019. ³⁷ As also suggested by (Paddon Rhoads and Sutton 2020).

58

CIVILIAN ADAPTATION AND SOCIAL CONNECTEDNESS

humanitarian relief, and households to adapt their livelihoods. When households and communities are more inclusive in terms of female and youth participation in livelihoods activities, civil society, higher education, and formal leadership roles, households have increased access to knowledge, and they deepen their social connections, which in turn, supports livelihood adaptation. These findings, taken together, add to the scholarship on gender in conflict, international humanitarian intervention, as well as social resilience during conflict. How should the international community—given its limited military and humanitarian involvement in Syria—interpret and translate these findings into policies and programmes for Syria and beyond? While the following insights are specific to the Syria case, remote management models and so-called investment in local partners (e.g. the localization agenda) are two trends that are becoming more common in humanitarian crises around the world. As such, many of the points summarized below can be considered for other crises. First, a premium should be given to supporting communications systems, as they have been central to maintaining social connectedness, receiving and transferring knowledge, and supporting learning and anticipation in Syria. Social media, internet and other forms of communication were essential to CDUs, spotter networks, and households that sought safe passage locally or further afield. Access to digital and analogue communications supported civilians in locating new jobs, planning livelihood projects, continuing their education, receiving remittances, and securing small informal loans. Communication outlets have also supported the emotional wellbeing of individuals by connecting them with families and friends, both near and far. Moreover, SNGOs could not function without access to social media, WhatsApp, internet, and phone connection. The international community could support communications in the form of infrastructure (satellite or other forms of transmission) to communities, or by providing households with internet and phone credit, or cash transfers. In addition, SNGOs could receive direct support from their donors for communications—a core operating expense that is generally not considered ‘allowable’. These interventions are relatively low cost, but could provide extensive benefit to the adaptive and protective processes of conflict-affected people. The international community could also prioritize gender and youth inclusion as part of their support for civil society and governance programmes. Western actors, particularly the United States and European states, have engaged in a range of so-called stabilization activities that have sought to support the development of civil society and local governance structures. However, this has been done without a consistent eye towards gender and age inclusiveness, despite the positive knockon effects for individuals, households, and communities. Lastly, the current model of humanitarian engagement with Syrian civil society overlooks and, at times, weakens the adaptative processes that SNGOs rely on for their successful operations. Much of the international assistance coming to

KIMBERLY HOWE

59

SNGOs has been pre-determined, and yet communities are adept at identifying and prioritizing their needs. SNGO representatives have repeatedly called attention to the mismatch between the needs identified locally and the assistance on offer. As a result, Syrian organizations report that their credibility is weakened, and that they suffer reputational damage. Top-down humanitarian engagement thus erodes the social connections that are so essential to effective SNGO operations. The requirement for SNGOs to scale up in the form of mini-International non-governmental organizations (INGOs) in order to receive funding from international donors, places an enormous organizational, financial, and operational strain on these organizations. In this process, they are required to shift their attention and resources away from the communities they serve, towards their donors—in essence, moving from a model of downward accountability to upwards accountability (Howe and Stites 2019). Rather than focus on the security of staff, the safe passage of goods, or rolling out emergency assistance following chemical attacks and bombings, they are required to spend hours in training (for each donor) to comply with donor requirements, and to build an organization as if they were sitting in Geneva instead of Aleppo. While there are very good reasons for such procedures, this process privileges international knowledge over local grassroots expertise, a phenomenon replicated throughout the world (Autessere 2014; Campbell 2018). This model of working weakens the capacity of SNGOs to engage in the vital adaptive processes of knowledge, learning, and anticipation and to strengthen their social connections—all shown in this chapter to be ingredients of effective civilian self-protection. The international community could learn much from local organizations about how they: acquire knowledge and skills; (try to) remain accountable to the people they serve; manage chronic risk; remain nimble; and how they adapt, cope, and survive through years of protracted conflict.

References Arjona, A. (2016). Rebelocracy: Social Order in the Colombian Civil War. New York: Cambridge University Press. Autessere, S. (2014). Peaceland. Cambridge: Cambridge University Press. Baines, E., and Paddon, E. (2012). ‘“This is How We Survived”: Civilian Agency and Humanitarian Protection’, Security Dialogue, 43/3: 231–47. Barnett, M. (2005). ‘Humanitarianism Transformed’, Perspectives in Politics, 3/4: 723– 40. Berry, M. (2018). War, Women and Power: From Violence to Mobilization in Rwanda and Bosnia-Herzegovina. Cambridge: Cambridge University Press. Bryant, J. (2019). Remittances in Humanitarian Crisis. Overseas Development Institute (ODI). Retrieved from Campbell, S. (2018). Global Governance and Local Peace: Accountability and Performance in International Peacebuilding. Cambridge: Cambridge University Press.

60

CIVILIAN ADAPTATION AND SOCIAL CONNECTEDNESS

Carpenter, A. C. (2012). ‘Havens in a Firestorm: Perspectives from Baghdad on Resilience to Sectarian Violence’, Civil Wars, 14/2: 182–204. Center for Civilians in Conflict. (2016). Waiting for No One: Civilian Survival Strategies in Syria. Retrieved from Chenoweth, E. (2019). Women’s Participation and the Fate of Nonviolent Campaigns: A Report on the Women in Resistance (WIRE) Data Set. One Earth Future Foundation. Retrieved from DiMaggio, P. J., and Powell, W. W. (1983). ‘The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in Organizational Fields’, American Sociological Review, 48/2: 147–60. Ellerby, K. (2013). ‘Gendered Security? The Complexities of Women’s Inclusion in Peace Processes’, International Interactions, 39/4: 435–60. Fordham, M. (2004). ‘Gendering Vulnerability Analysis: Towards a More Nuanced Approach’, in Bankoff, G., Frerks, G., and Hilhorst, T. (eds), Mapping Vulnerability: Disasters, Development and People, pp. 174–82.: London; Sterling, VA: Earthscan Publications. Fulu, E. (2007). ‘Gender, Vulnerability, and the Experts: Responding to the Maldives Tsunami’, Development and Change, 38/5: 843–64. Gingerich, T., and Cohen, M. (2015). Turning the Humanitarian System on its Head (Research Report). Oxfam. Retrieved from Haddad, S., and Svoboda, E. (2017). What’s the Magic Word? Humanitarian Access and Local Organizations in Syria (Working Paper). Humanitarian Policy Group at ODI. Retrieved from Hernández Cárdenas, A. M., and Tello Méndez, N. G. (2017). ‘Self-Care as a Political Strategy’, Sur-International Journal on Human Rights, 14/26: 181–91. Howe, K. (2016). No End in Sight—A Case Study of Humanitarian Action and the Syria Conflict. The Feinstein International Center at Tufts University, Kings College and ODI. Retrieved from Howe, K., and Patros, T. (2020). Household Resilience During Conflict: Qualitative Comparative Analysis for the Case of Syria. Mercy Corps. Retrieved from

Howe, K., and Sheely, R. (2019). Investing in Syrian Humanitarian Action: Performance Evaluation Report. Mercy Corps. Howe, K., and Stites, E. (2019). ‘Partners Under Pressure: Humanitarian Action for the Syrian Crisis’, Disasters, 43/1: 3–23. Howe, K., and Uvin, P. (2009). ‘“I Want to Marry a Dynamic Girl” Changing Gender Expectations in Burundi’, in Life after Violence: A People’s Story of Burundi. London: Zed Books. Howe, K., Krystalli, R., Krishnan, V., Kurtz, J., and Marcanas, R. (2018). The Wages of War: Learning from How Syrians have Adapted their Livelihoods through Seven Years of Conflict. Mercy Corps. Retrieved from Humphrey, A., Krishnan, V., and Krystalli, R. (2019). The Currency of Connections: Why Local Support Systems are Integral to Helping People Recover in South Sudan.

KIMBERLY HOWE

61

Washington, DC: Mercy Corps. Retrieved from Internal Displacement Monitoring Center. (2020). Syria. Retrieved September 21, 2020, from Jose, B., and Medie, P. (2015). ‘Understanding Why and How Civilians Resort to SelfProtection in Armed Conflict’, International Studies Review, 17: 1–25. Kaplan, O. (2017). Resisting War: How Civilians Protect Themselves. Cambridge: Cambridge University Press. Kim, J., Humphrey, A., Marshak, A., Gathuoy, N. M., and Krishnan, V. (2020). The Currency of Connections: Why do Social Connections Matter for Household Resilience in South Sudan. Washington, DC: Mercy Corps. Retrieved from Krause, J. (2018). Resilient Communities: Non-violence and Civilian Agency in Communal War. Cambridge: Cambridge University Press. Krause, J. (2019). ‘Gender Dimensions of (Non) Violence in Communal Conflict: The Case of Jos, Nigeria’, Comparative Political Studies, 52/10: 1466–99. Lazarev, E. (2019). ‘Laws in Conflict: Legacies of War, Gender, and Legal Pluralism in Chechnya’, World Politics, 71/4: 667–709. Masullo, J. (2021). ‘Refusing to Cooperate with Armed Groups. Civilian Agency and Civilian Noncooperation in Armed Conflicts’, International Studies Review, 23/3: 887–913. Maxwell, Dan, and Majid, N. (2016). Famine in Somalia: Competing Imperatives, Collective Failures, 2011—2012. Oxford: Oxford University Press. Maxwell, Daniel, Majid, N., Adan, G., Abdirahaman, K., and Kim, J. J. (2016). ‘Facing Famine: Somali Experiences in the Famine of 2011’, Food Policy, 85: 63–73. Mukhopadhyay, D., and Howe, K. (forthcoming). Good Rebel Governance: Revolutionary Politics and Western Intervention in Syria. New York: Cambridge University Press. Nı´ Aoláin, F., Haynes, D. F., and Cahn, N. (2011). On the Frontlines: Gender, War and the Post-conflict Process. Oxford: Oxford University Press. Niland, N., Polastro, R., Donini, A., and Lee, A. (2015). Independent Whole of System Review of Protection in the Context of Humanitarian Action. Norwegian Refugee Council. Retrieved from OCHA. (2021a). Syrian Arab Republic: Developments in Northwest Syrian and Ras Al Ain—Tell Abiad (Situation Report No. 32). United Nations. OCHA. (2021b). Humanitarian Needs Overview: Syrian Arab Republic (Humanitarian Programme Cycle). United Nations. Paddon Rhoads, E., and Sutton, R. (2020). ‘The (Self ) Protection of Civilians in South Sudan: Popular and Community Justice Practices’, African Affairs, 119/476: 370–94. Pereda, C. P. (2017). ‘Resilience in Times of Repression’, Sur International Journal on Human Rights, 14/26: 141–49. Schon, J. (2020). Surviving the War in Syria. Cambridge: Cambridge University Press. Stites, E., and Humphrey, A. (2020). The Currency of Connections: The Role of Social Connectedness among South Sudanese in West Nile, Uganda. Washington, DC: Mercy Corps. Retrieved from Syrian Network for Human Rights. (2019). Civilian Death Toll. Retrieved from

62

CIVILIAN ADAPTATION AND SOCIAL CONNECTEDNESS

Violations Documentation Center in Syria. (2020). Monthly Statistical Report on Casualties in Syria. Retrieved from Wall, I., and Hedlund, K. (2016). Localisation and Locally-Led Crisis Response: A Literature Review. L2GP. Retrieved from https://www.local2global.info/wp-content/ uploads/L2GP_SDC_Lit_Review_LocallyLed_June_2016_final.pdf. The White Helmets. (2020a). The White Helmets. Retrieved 20 December 2020, from

The White Helmets. (2020b). Retrieved 20 December 2020, from Wood, E. J. (2008). ‘The Social Processes of Civil War: The Wartime Transformation of Social Networks’, Annual Review of Political Science, 11: 539–61.

3 Transformations in occult protection amid violence in the Central African Republic Louisa Lombard and Mobito Kozaga

Introduction During widespread violent upheaval, the foundations of social life—who can be trusted, the rituals of everyday interactions—are supplanted by a ‘cosmos of doubt’ (Krohn-Hansen 1994 p. 375), and people use magical thinking and magical action to recover agency (Girard 1977; Hoffman 2011). One of the most striking recent examples of magic at war has been the bulletproofing techniques combatants develop and use, like the mystically-infused ronko jackets worn by fighters in Sierra Leone’s war (1991–2002), or any of a range of other treatments, including vaccination, charms, fetishes, and closely-observed rituals. People doing immersive research with armed groups have encountered these magical components to combatants’ prosecution of war and shown how fighters call on invisible forces to help themselves survive and vanquish foes during war (Lan 1985; Richards 2009; Hoffman 2011). The ways that civilians develop and enact magical modes of protection have been less explored (though see Masullo and Jentzsch 2019). Accounts of fighters’ magical military practice have argued that these practices are creative, often effective, bring legitimacy, and are a marker of social distinction to those practising them. But studying occult practice during war without being situated in or alongside a group that is invested in imbuing its practices with legitimacy presents a rather different picture, one permeated by a changing mix of ardour, ambivalence, fear, and secrecy. In the sporadic yet persistent war in Central African Republic (CAR) since 2013, more people have been developing, providing, and seeking out occult services in order to protect themselves than during normal times. We give some examples of the effects of the wartime popularization of occult protection strategies. We do this in order to make a case for caution in generalizing about the status and effects of occult practice and expertise during war, since so much depends on perspective, and even one individual’s perspective can change over time. Magical thinking might be inherent to dealing with war and the perception of being surrounded by risk

Louisa Lombard and Mobito Kozaga, Transformations in occult protection amid violence in the Central African Republic. In: Civilian Protective Agency in Violent Settings. Edited by: Jana Krause, et al., Oxford University Press. © Louisa Lombard and Mobito Kozaga (2023). DOI: 10.1093/oso/9780192866714.003.0004

64

TRANSFORMATIONS IN OCCULT PROTECTION AMID VIOLENCE

(Girard 1977; Richards 2009), but to attempt to extract a coherent theory of it is to miss the essence of both the tragedies and creativity of war.

Harnessing the invisible world during war Occult or transcendent forces, magic, witchcraft: all these terms and more could be mobilized to refer to the broad realm we discuss in this chapter. And yet despite the proliferation of terms, none faithfully captures the domain we are trying to describe. This is because it is easy for people who do not see those kinds of forces as operative in their lives to see them therefore as the product of, and evidence of, other people’s backwardness. Do they really believe that? Many ask this of accounts of, for instance, the genital-stealing panics that swept across West and Central Africa beginning in the 1990s (see Bonhomme 2009). In doing so, they forget that all of us are in the grip of ‘invisible ontologies’—Kwame Anthony Appiah’s (1992) term for all those things we understand to be causal forces in the world that are not verifiable by the senses. Occult practices have a deep history in warfare. Prophet-led wars and rebellions have occurred across many centuries and continents (Wenzel 2010). In Africa, spirit mediums, healers, and other practitioners of the craft of putting invisible forces to work for people were leaders in the struggle against European conquest and colonialism. Those Africans whom Europeans identified as ‘political leaders’ had been co-opted for indirect rule. Their counterparts—prophets and mediums—appeared ‘religious’ to the colonizers and so were left out of official authority structures. They often led the charge for resistance and offered a path to a future wellbeing far more attractive to many people than that provided by the colonizers (Feierman 1985). In struggles for independence, rebels understood spirit mediums and others who could channel immanent spiritual forces as comrades-in-arms whose expertise was necessary. For instance, freedom fighters in Zimbabwe had an official ideology of radical, modern, black socialism. But the fighters paired the official stance with reliance on the support of their ancestors, whom the spirit mediums could help keep on their side. The spirit mediums developed new rituals to channel the ancestors’ aid for the guerrillas (Lan 1985). This ‘remarkable “co-operation” between ancestors and their descendants, the dead and the living, the present and the past’ (Lan 1985 p. xviii) was one of the main forms the guerrillas’ resistance to white rule took. More recent studies of rebels and the invisible world also emphasize the creativity of ritual practice. Ritual ‘is a sociological way of forming hypotheses about collective interests’, noted Paul Richards (2009 p. 502) in a discussion of the war in Sierra Leone (1991–2002). ‘Magic is the bubbling cauldron of practices from which both the means–ends rationality of craft production, and the collective and contractual commitments of organized, institutionalized religion can be decanted (Mauss 1936). War is,

LOUISA LOMBARD AND MOBITO KOZ AGA

65

par excellence, the sphere of magic because it is a sphere of improvisatory and unregulated technique. The cult of war is the most extreme of all such cults, and the most extremely magical’ (Richards 2009 p. 506).

Many Sierra Leonean fighters wore a ronko jacket. This garment was made of a special, thick cloth that had been treated and impregnated with mystical forces that made the person wearing it bulletproof. Outsiders seeking to test the truth of that claim to bulletproofness would take the jackets and shoot them. When the bullet destroyed the garment, they were confirmed in their suspicion that Sierra Leoneans’ faith in the ronko jackets was a misguided, backward belief. But they misunderstood: the jackets worked not just because of their material properties but because of the social rites that were attached to them, rites that helped people forge a collective understanding of the world in the midst of the chaos of war (Richards 2009 pp. 505–6)—rites which the jacket-testers, of course, had no knowledge of or participation in. Danny Hoffman has further studied bulletproof fighters in the Mano River Region and argues their magic is a ‘military technology’. Saying that magic is a military technology implies that it is the purview of armed actors. That was indeed the case in Zimbabwe, according to Lan; it was the guerrillas who were most invested in innovative ritual to channel the ancestors’ power toward the guerrillas’ new ends. But it does not seem to be the case in a general sense. After all, wars like those in the Mano River Region in the 1990s and early 2000s sowed chaos not just for fighters, but for nearly everyone. Non-fighters, too, had to figure out how to save themselves and their communities in the midst of the moral panic of war, when ‘collective concern is heightened, emotional energies are stirred but, because the world of ritual and the world of mundane collective practices has become disordered, there is no balancing feedback from the material world’ (Richards 2009 p. 503). Inventing magical practice, for instance in order to better channel ancestors’ spirits, is both the response to the moral panic and among the factors perpetuating it. Such has been the case in the simmering violent conflicts in the CAR. Civilians there have tried to channel invisible forces in order to protect themselves. Similar to the case Jentzsch (within volume) discusses, in CAR it can be difficult to differentiate between armed group members and civilians. There has been a spectrum of involvement in fighting, from people who avoid any participation to people who are full-time members of fighting forces. We focus on non-Muslim peasants in the central CAR region around Bambari. Our interlocutors generally partake in the autochthony grievances espoused by the movement known as Anti-Balaka/Balaka/Balakism, whose adherents include standing armed groups, punctually-operationalized local defence forces, and people without any direct role in armed groups. We focus on civilians, and in local understanding that encompasses both the local defence actors and the people without any armed group connection. Self-described local-defence actors might at times strike offensively too, but in local understanding they are civilians, because they

66

TRANSFORMATIONS IN OCCULT PROTECTION AMID VIOLENCE

do not give up their peasant/family responsibilities except for certain periods of time (such as for occult initiations). They explain much of the increase in the ranks of the adepts of occult action.

Methods The material in this chapter stems from ethnographic research conducted by both authors over more than ten years. Ethnographic research requires extended periods of ‘being there’ so that what people say is part of an experiential context, allowing for assessment not just of stated positions but also the differences between what people say and what they do, and the non-linguistic, meaningful aspects to the ways people interact and learn from each other, and how their practices and self-understanding change over time. It is difficult to sum this up in terms of a specific number of interviews. Mobito Kozaga worked in Bambari, the focus of this article, from 2014 to 2022. Louisa Lombard has visited Kozaga there for research on two occasions, in 2015 and 2017, and also conducted research elsewhere in the country during this period. The article also draws on the expertise of organic intellectuals in the region, such as a pastor who has undertaken several interviews and developed relationships with people who turned to occult means during the war.

Histories of occult practice in CAR Two dynamics in CAR’s difficult history are particularly relevant in the context of the problems of this chapter: first, that violence, while not constant, has been repeated, and intermittently quite intense; and second, that people have felt their interventions in the spiritual realm to be not as effective as they once were, and have developed new ritual practice to respond to the exigencies they face. Currently, there are several discrete yet interconnected modes of occult practice among Equatorial Africans. One is likundu (sorcery), another is the protective treatments offered by travelling marabouts and anganga (healers, in a broad sense), and a third is the practices of organized religion. If ever there was a period of easy coherence in the realm of magical practice (doubtful), there is not now. Instead, these modes are interwoven in awkward ways. The durability of the resulting tapestry is frequently tested, especially now that more people are getting involved. Central CAR, the region of our research, was a zone of both raiding and refuge over the nineteenth century, when slave-raiding and colonization both swept into the area. Because of this upheaval, people with roots from throughout the broader sub-region now call the area home. People speak of the most populous ‘ethnic group’ as Banda; that is in fact a term that came to be used over the course of the nineteenth century and is a cross

LOUISA LOMBARD AND MOBITO KOZ AGA

67

between an ethnic marker and a proletarian one. Prior to the colonial era, growing up as a Banda person required going through a painful initiation around puberty. During this initiation, youth would be beaten by their elders and were expected to endure it silently. The beatings communicated to the initiate that he or she belonged (you are one of us) and that belonging carried responsibilities, such as deferring to the elders (Giles-Vernick 1996). At the same time, initiates began to gain knowledge of and experience with the occult realm that the elders oversaw. The initiation system, and its vision of the proper order of individual and collective life, came under strain as Christian missionaries and their schools became more widespread over the later colonial period. Youth turned to teachers and priests with the expectation they could help them access a good life. Rather than working in elders’ fields, they worked in their teachers’ (Giles-Vernick 1996). Around independence (1960), the first generation of Central African clerics came of age. Most prominent in Bambari was one who became known as Ngoutide. Ngoutide was hailed as a prophet, and he advocated the total destruction of the old ways. He was so successful that today there exist no known material artefacts related to the old Banda ways in the Bambari region (Mayneri 2014). The pathways through which everyone gained some familiarity with the workings of the occult and other realms of power and danger, chiefly initiation and the gendered secret knowledge that people continued to learn over the course of their lives, became overgrown. Occult knowledge became the domain of specialists, and became suspect. People still turned to anganga (healers) in times of need, but did so quietly. Anganga who become renowned for their skill stand as beacons of aid in the midst of daily realities that many Central Africans experience as stormy. One anganga in Paoua in northwestern CAR in 2011, for instance, said that he had treated about 900 people in the last year, including 357 in the previous quarter alone (World Bank Group 2012). Good anganga can figure out why a person is suffering and propose a course of treatment, usually involving a variety of plants that have been specially prepared. They learn which plants work because spirits guide them toward the right ones. As a corollary, anganga know more about the workings of anti-social, illness-inducing magic than laypeople, and could use it themselves, making them vulnerable to accusations. Their position has also been vulnerable because the official Christian position (across Catholic, Protestant, and Evangelical denominations) has been that their work is satanic. Anganga’s historical role ‘countering sorcery and healing its effects’ (Douglas 1999 p. 180) has not ended, but they now bump into Christian competitors who see them as nefarious imposters. In addition to anganga, who can be settled or itinerant, there are travelling marabouts who peddle occult protections of various kinds. Marabouts (sometimes called Madébo, a Central Africanization of marabout) are a work-defined branch of the Peulh, a diverse mega-ethnicity whose members live from West Africa to

68

TRANSFORMATIONS IN OCCULT PROTECTION AMID VIOLENCE

South Sudan. Marabouts started selling their medicinal and occult wares in CAR several decades ago. Over the last fifteen years, people in neighbouring South Sudan and the Democratic Republic of the Congo have targeted them for expulsion, arguing that they collaborate with armed insurgents and are otherwise social undesirables. Those attitudes are present in CAR too, and Peulh have been among the Anti-Balakas’ targets for violence, dispossession, and expulsion. But Peulh marabouts have also been crucial sources of occult knowledge, particularly for self-armouring, among both Anti-Balaka and Anti-Balaka-adjacent people. And while some Central Africans purchased marabouts’ services before the crisis, our conversations with Central Africans suggest the trade has boomed since.

Crisis in CAR Political violence in CAR has generally been on the increase over the last twentyfive years. Beginning in the 1990s, army mutinies and criminality made the capital, Bangui, a dangerous place and occasioned the deployment of a first international peacekeeping mission (by now, there have been a dozen such missions). Since then, rebellion and international intervention have been more or less constant, although their geographic dispersion changes and is never total, such that certain areas can experience quotidian peasant life while others are in torment—a good case of what Lubkemann (2005) referred to as ‘fragmented war’. The intensity of both violence and intervention markedly increased beginning in 2013, when a coalition of men-in-arms and rebels calling themselves Seleka (alliance in Sango, CAR’s lingua franca) claimed power in Bangui. Seleka included many armed men who Central Africans identified as foreigners (Chadian, Sudanese), and they ruled over towns in ways Central Africans experienced as brutal, and as granting favour to their Muslim co-religionists. People rose up in protest, forming militias that came to be known as the Anti-Balaka. The Anti-Balaka were not a group, but rather the mostly-male social product of an upswell of popular anger. Some were organized and stood at the ready for self-defence, while others joined mostly in moments of ‘forward panic’ (when a dangerous enemy suddenly appears in an attackable form) (Collins 2009). They targeted not just Seleka but any Muslim they found. The worst of the violence occurred in 2013–2014, when a massive UN multidimensional peacekeeping mission (about 13,000 strong) was also deployed.¹ But while a peace agreement was signed in 2019, violence continues. The number of displaced people has remained fairly constant, at more than a quarter of the population (UNHCR 2021). In 2021 there was a marked uptick in violence, both ¹ For fuller accounts of this period, see (Mayneri 2014; Lombard 2016; Marchal 2016; Glawion and Vries 2018).

LOUISA LOMBARD AND MOBITO KOZ AGA

69

by armed groups and by Central African troops operating together with Russian trainer/mercenaries, all of whom ‘demonstrated complete disregard for international humanitarian law’ (UN Panel of Experts Established pursuant to Security Council 2021 p. 12). Nearly every day there is an incident of kidnapping, killing, detention under brutal conditions, and/or armed robbery. Some towns have been mostly spared, while others endure recurring incidents. Bambari and the surrounding area have been especially affected by violence, because it has been the main base of one of the most powerful armed groups, the Union of Central African Patriots (UPC, from French), which many Central Africans see as comprised of illegitimate foreigners. Even before the ‘crisis’, as the post-2013 period has become known, Central Africans did not feel they lived in a coherent moral-juridical/secure order. State organs, when not withered to an aching memory, included employees who were either irrelevant or brutal and predatory. Some people, principally certain chefs de village, cultivated mediation and negotiation skills, and it was to them people turned when they had disputes to resolve. But overall people have felt that norms and expectations are under immense strain, and that no one has been strong enough to impose a new order. Cultural distinctions and ways of doing things, from agriculture to marriage, have been disrupted, with grave consequences. People are exposed to violence and have sought out and developed means of armouring themselves, of adding a layer of sacred, occult protection beyond that provided by their everyday body.

Making your body that of a man American gun rights proponents try to attract people to their side with questions like ‘Are you someone who protects others? Or are you someone who needs others to protect you?’ Many men in rural, central CAR have been asking each other versions of these questions. The self-defence mobilizations in many villages around Bambari, such as Mingala, Satema, Yangbassi, and Morouba, have been one result.² In addition to obtaining guns, men joining these mobilizations have sought to ‘sala tere ti mo koli’, or ‘make your body into a man’. When they speak of making themselves men, they mean travelling to a specialist, usually in a remote village, to undertake a specific change to their bodies, namely one that makes their bodies impervious to physical aggression and gives them greater capacity to strike ² The boundary between armed groups and local actors who feel threatened and work together to defend themselves is not completely firm, but there are key differences. For the less organized actors, the heads of neighbourhoods or villages play a key role in mobilizing able-bodied men from their localities. Village chiefs try to play their role in protecting their inhabitants by relying on the human and material resources of their locality. The inhabitants of towns such as Mingala, Satema, Yangbassi, and Morouba have no relationship with the political leaders and are often surprised by the appalling violence are those who resort to occult practices to protect themselves against aggression.

70

TRANSFORMATIONS IN OCCULT PROTECTION AMID VIOLENCE

back against harm. To make one’s body a man is to make oneself invulnerable to any kind of aggression. Others use the term ‘going back to the source’—that is, returning to old ways Christian missionaries robbed them of. This desire, and the practices marshalled to fulfil it, hearken to the initiations that used to structure people’s lives but that had been sapped of their power and authority over the last decades. The pain endured during initiation hardened the body and let initiates know they had the toughness to truly belong; thanks to the ritual and secret aspect of the process, the beating and the pain were beyond-the-human and occult and the effects were greater than those of a quotidian fight. In the crisis period, the means people have used to develop more-than-human toughness have been varied and drawn on eclectic expertise, including that developed by anganga, elders, and marabouts. The elders of certain villages became known as the purveyors of especially powerful protective power. Men would travel long distances to obtain their cures, which they also referred to as ‘taking other forces’ (‘prendre d’autres puissances’). One consequence of the war has therefore been that people have turned back to elders they had previously overlooked, finding in them important sources of protection and strength. Another has been an increase in the markets for and purveyors of magical protection. As the phrasing ‘make your body a man’ suggests, gender is an important dividing line around occult practice and protection. Women seek out occult protection as well, but men are more frequently the seekers of forms that require them leaving home to obtain. Many women say that men abandon them during fighting, leaving them alone to look after the children when things get dangerous. Men say they took the necessary step of seeking out the skills and knowledge that would protect themselves and their families. The time frame matters: some men leave during an attack, but when they do return feel better girded to fight back. The taking of other forces includes vaccinations to make the vaccinated person not feel pain and/or be impervious to knives/machetes and bullets, or even to make it so that if the person is severed limb from limb the pieces of his body reassemble and he comes back to life; metamorphosis (changing a person—an enemy—into a non-human form); and other means of foiling and weakening threatening forces. The potions might be prepared with fire on the roof of a straw house and yet the house would not burn down. The materials they use toward these ends are many. A pastor who counsels people who entered into occult practices for the purposes of protection and self-strengthening but have decided to leave behind those ways and instead seek the protection of the church reported that the repentant generally make two deposits of occult materials. The first consists of those items they carry on their body, usually hidden, and so therefore limited in size and quantity. These can include braided pieces of tree bark, oil, nails from small carnivores, and miniature gris-gris. The second deposit might include larger gris-gris, fetishes, masks, animal horns, such as from antelopes, carnivore claws, skulls of taboo animals, feathers from taboo birds, the enemy’s spent cartridges that have been ensorcelled,

LOUISA LOMBARD AND MOBITO KOZ AGA

71

and potions known as ‘négligeances’ that counter-act a moment of lapsed attention when it comes to the rules of black magic. These materials are activated through the knowledge and rituals of the people developing and using them, and de-activated once presented in the light of day outside their proper context. Another aspect that constitutes a mixture of religious belief with mysticism is the fact that the ‘waranga’, gris-gris often used much more by the Seleka (Muslims) and to some extent by the Christians, contain in the package inscriptions of Koranic passages and powders containing bullets of different weapons. Given this protection, the death of an Anti-Balaka, a Seleka, or a soldier at the front is interpreted as the consequence of unresolved family or social problems. The person who fell must have had problems with his wife or one of their relatives that they did not resolve before deployment to the field. Or the victim stole or took the property of some people by force. Their misfortune is explained as a function of their breach of moral codes about proper sociality, in terms of family and property relations.

A former Balaka repents For a more detailed account of occult practice during war, we turn to the narrative of a former Balaka who left their ranks in early 2015, as he explained to a pastor in Bambari at the time:³ When the Seleka came into the city [Bangui, the capital], they sowed terror, mistreating and killing and pillaging. As for me, in Bangui the Seleka killed my father and they took our family house. I left Bangui to come to Bambari and I tried, since I value my life, to go to my relatives in the countryside thirty kilometres from town. At that time, the Mbororo Peulhs were acting as masters, killing, pillaging, and burning houses and grain storage. It was at that moment that I decided to become Balaka. I went to someone who was already Balaka and he brought me to a madébo who was going to make me strong. To do this we had to go into the bush. I had to bring some money, a chicken, and some bark. There are several ways to armor someone so that he becomes Balaka—it depends on the madébo. There is a mainline Balaka fetish, but the madébos finish it with the occult powers from each region or tribe. In front of the other initiates he begins saying incantations. He vaccinated me in various places. He made me take potions—on my ³ This account is the narrative of a former self-defence Balaka, as heard and re-created by the pastor to whom he confessed his occult engagements. It reflects the pastor’s epistemology, in which God and Jesus stand in opposition to the Balaka ‘demons’ (other terms that could have been used include ‘spirits’—known in Sango as ‘toro’). The account stands as a view onto the kinds of practices and the worldview of the participants. In particular, the importance of occult armouring as a mode of protection and defence comes through quite strongly. But trying to assess the effectiveness of these tactics toward the ends of collective protection and defence is much trickier, particularly since even this key actor in that process was regretful and uncertain about what he had actually done.

72

TRANSFORMATIONS IN OCCULT PROTECTION AMID VIOLENCE forearm, he put some powder, hitting with his hand until this substance penetrated the skin. At that point a first test became possible. I saw a strongman come out of one corner, a machete in his hand. I was told to put my arm on the trunk of a tree. This strongman could have amputated my arm with one chop of the machete. But surprisingly, no matter how many times he hit me, there was no effect, not even a trace of pain. So that was the first test. We went on to the second, that of invisibility and bulletproofing. Based on how that went he told me that in battle I could disappear in front of the enemy and reappear behind him. Or the enemy won’t see you and instead will see a tree trunk but meanwhile you’re right there, seeing him. The mysterious demon of invisibility transforms his sight. At this stage, you are called to be still and silent until the enemy turns and then you can attack him from the back. At this stage they taught me the passwords that enable the technique of ‘three steps disappeared’. After that stage, it’s the initiation into invulnerability. This practice lets you come back to life after you have been killed and cut into pieces. You can come back to life after the enemy has abandoned your corpse. One of the elements of this step is for example that you bring a hen, as I was told to do. After an incantation said over the chicken, she needs to lay six eggs each night, and these eggs need to be cooked and you eat them there. You do this for six days, because the chicken has gotten the power to lay six eggs each night for six days. At the end of these six days, the chicken is killed, and cooked with certain potions and you eat it. You become impenetrable.

At this point the pastor inserted several explanatory notes into the account:

It is important to note that for all these processes of armoring, there are rules that must be respected. It is this aspect that explains the case of Balaka who fall in combat. … I also forgot to say that after the first step of the machete test they place you five or ten meters from the barrel of a gun. They shoot at you but without effect, all the bullets fall to the ground in front of you, or are deviated, depending on the kind of armoring your madébo uses.

During the war, new people entered into the realm of occult practice. They did so out of a mixture of interests and emotions: they sought to protect themselves and their loved ones, in part by effectively countering those they saw as their enemies and aggressors. In the process, new ‘hierarchies’, as the pastor in Bambari put it, have emerged in the world of occult practice. Hierarchy makes relations sound stable, though, and in fact people’s positions can shift as they move in and out of occult practice and networks. There are both passers-through of the invisible world, as well as people who become true adepts, and even adepts might later renounce their occult work and status. Those who renounced needed to

LOUISA LOMBARD AND MOBITO KOZ AGA

73

definitively leave the occult ways behind them, and one way to do so was to confess their activities to a pastor. The pastor noted that initiates were diverse in their level of knowledge and commitment. He explained: It’s important to note that there is a hierarchy in the world of the Balaka. There are simple initiates who are fighters and there are bosses who hold the sacred power . . . . I have always insisted, and they would tell me that they are from a hierarchy and they fight. Give me an example of this hierarchy. They say that some are small sorcerers more than the others—some don’t eat human flesh, others do. Some are males, some are females. Some are defenders, others destroyers. In short.

What the pastor describes is different from during normal times, when participation in the occult is more circumscribed and includes primarily professionals like the anganga and people in particular precarious relationships and social positions. (Witchcraft accusations tend to be within families and fall heavily against older women and unsupported children.) During war, younger men became newly involved, and to a greater degree than would otherwise have been the case. This was not unheard of; the young men’s participation can be seen as an off-key echo of initiation practices from the period before Christian missionaries and their churches became widespread. But the world has changed a lot since then, and the secret, occult knowledge is now less a matter of knowing the right way to live and more one of specific tools that protect a person in the context of war. The pastor then continued recreating the former Balaka’s account: I never entered into the caste of the trainers, because what they do is really a big mystery. Pastor, I killed, I violated the enemy to get vengeance. I did not want to be subjugated by Muslims. I am telling you the truth, pastor, now that I decided to leave that circle and come back to God, the way of thinking of the militiamen I knew, no one is willing to be subjugated by these Chadians and Sudanese. We had lived in peace with them, and they were the ones who taught us the culture of violence . . . . I am a totally lost soul. Help me, servant of God. I know that you will ask me to bring all my gris-gris and fetishes, yes, no problem, but what about the substances that entered into my blood? I am a very unhappy person.

I am a very unhappy person. The former initiate’s unhappiness owed to his having practised and sown devilish arts. While he was among the other Balaka, the initiations and other magical practices drew those gathered in the scrublands together with the shared purpose and collusion that comes from holding important, powerful secrets. It felt like a form of civilian protective agency: taking action to become strong, to become men, and to take on all comers. And perhaps it was that. After all, this young man returned home in good health. He might not have done so had

74

TRANSFORMATIONS IN OCCULT PROTECTION AMID VIOLENCE

he stayed put. While he was among the initiates, he likely would have agreed that their shared practices made them more confident and motivated in the context of hardship they were enduring, and that the group of men together in the bush were able to share information and became close in ways that made them feel more in control and safer. The initiates took positions in new social hierarchies that could help them organize themselves and work together in a time of uncertainty, fear, and hardship. But later, once the threat was not so intense, the initiate regretted what he had done. He worried about the effect it would have for him in the future, since the occult substances had entered his body and become part of him. Occult civilian protective agency, then, has some of the same observable social knowledge (Krause within volume) and social networking effects as non-occult protective action, but its transgressive and dangerous sides underscore the need to consider those effects in multiple time horizons. People’s attitudes toward occult work are frequently contradictory, and change, as do their assessments of its necessity and effects. While a person might see it as necessary, and even exciting, to participate in practices attuned to the invisible world, that view can change over time, with the person coming to renounce that period in his life and see it as an embarrassment or a regret. Sometimes people’s perspectives change because they witness an instance of failed sorcery, other times because of a personal epiphany, and yet other times because of shifting personal relationships (falling out of friendship, meeting someone new, one’s spouse’s feelings).

Travelling salespeople and the occult in town In his article ‘Dressed to Kill’, Paul Richards (2009) includes a tantalizing aside: when he reconnected with male friends who had fought in the Sierra Leonean bush, they complained about the false initiators they had encountered. It is a rare admission that entrepreneurs of spiritual practice, too, include both hucksters and failures. And they do not just approach combatants, or would-be combatants, with their services and skills. They also approach regular people who otherwise see themselves as having no truck with the occult, except that in desperate times doing so can seem necessary. For instance: gris-gris, which are prepared by specialists. Many describing the appearance of armed men in Africa have noted that combatants wear gris-gris— amulets or charms worn around the neck that protect the bearer. The gris-gris usually contains Koranic verses that have been packed very tightly into a neat bundle, usually together with some other charged substances: oil from a second tree that is growing out of a larger tree (that is, a tree that has sprouts of a different kind of tree growing from it), certain plants that have been dried and pulverized. In Ndele, a town in northern CAR where Lombard lived for several months in 2009–10, Central Africans claimed that gris-gris were a Chadian innovation that

LOUISA LOMBARD AND MOBITO KOZ AGA

75

had only recently taken root in CAR, and that they were something for men-inarms. But by 2017, in the central region of CAR, such as around Bambari, many people, including non-fighters, were purchasing and wearing gris-gris. Bulletproofing falls into the same category of occult military technology that has been taken up, including in new ways, by people who are not members of armed groups and who are trying to keep themselves safe in the midst of rampant violence. Itinerant Madébos will visit villages offering their services in making people impervious to bullets, usually through what is known as a vaccination. That is, a mixture of substances is inserted under the skin of the person being treated. In addition to the vaccination, there is usually a list of prohibitions and taboos that the person must uphold for the vaccination to work. Abstaining from sexual intercourse is a frequent injunction. Bulletproofing in this sense is not like Kevlar; rather than one clear causal mechanism there are many factors that go into making it work, and those factors are both material and social.⁴ Even despite the tendency for confirming rituals to explain why bulletproofing does not always save those who have been thus vaccinated, there are still cases that people recognize as pure failures in the technology, whose purveyors then fall from grace and become charlatans. In 2017, in a village just over twenty kilometres outside of Bambari, a travelling madébo arrived and said he could make people bulletproof. People were intrigued and asked for proof. The madébo worked all night to prepare the substances that would make for an effective treatment. There was some discussion about who the substance should be tested on. If it was tested on a man, how would they know for sure he had abided by all the abstentions and other stipulations? So they agreed that the test subject would be a child, someone who was innocent of the world of adults and adult-specific taboos. The child was thus vaccinated. The following morning at dawn they had the child sit in front of the madébo, gave the madébo a gun, and told him to test the medicine. He shot, hit the child, and the child died. Everyone in the village immediately turned against the false madébo. He was, in the words of one witness, ‘dépiécé’—torn limb from limb. We will never know what the madébo was thinking, if his practices had previously worked and now failed or if he was simply a venal fraudster. But the incident shows that we cannot fully set aside the question of the effects of witchcraft as witchcraft—since sometimes it has the hoped-for occult effects, and other times it does not. People who understand the world to be shaped by occult forces and action, too, differentiate between charlatans and skilled practitioners, but their ability to know what works and what does not is severely tested by the new needs and new protection tactics that people try out when they feel profoundly

⁴ Although to be fair to Central Africans, in the case of industrialized military armour too, people killed or injured are often described as having worn it incorrectly, rather than seeing their maiming as evidence that the technology is flawed.

76

TRANSFORMATIONS IN OCCULT PROTECTION AMID VIOLENCE

threatened. We have already discussed why the observable effects of occult action are tricky to generalize about. The magical effects are even less accessible to social science methods,⁵ though they are very important to people participating in them. And many people are doing just that. Kozaga bumped into the widespread yet secretive realm of civilian occult practice while in Bakala, a small town north of Bambari. Kozaga did a favour for a young man whom he did not know, who lived there. As a thank you, the young man proposed to give Kozaga an object that he should swallow. This object, once it entered into his body, would make him invulnerable. The young man proposed that this would be a good trait given that Kozaga ‘traverses conflict zones’. Here was an opportunity for Kozaga to experiment with this occult source of resilience. But it also raised crucial questions around ethics and epistemology. Would participating in this experimentation really let Kozaga better understand the stakes of the bulletproofing and anti-fragility practices? Maybe. But Kozaga declined the offer, because he deemed that it crossed a line beyond the participant-observation philosophy of ethnographic research into complicity that would mark him and effectively put him on one side in a way he sought to avoid, both in his research and when wearing his other professional hat as a humanitarian. The young man replied that he understood Kozaga’s reasons for reluctance. He said: ‘I’m here. You can come and see me later when you decide you would like to have it.’ He did not want to show Kozaga what the substance was. That would remain secret until Kozaga agreed to ingest it. Those who partake of each other’s occult substances, then, must trust each other sufficiently to take that risky step, which is useful social knowledge in a context of violence. But he did explain the substance’s properties: it would permit him to disappear and then later reappear when an enemy was attacking. It was a men’s Balaka potion of the kind developed for and by initiates in the bush, but now appearing in town as well. On another occasion, in a small town near Bambari, Kozaga purchased a gris-gris. The practice of wearing gris-gris has roots in Islamic traditions more prevalent in the regions north and west of CAR, but it has become widespread for people to use them in CAR as well over the last twenty years or so. A gris-gris is a tiny, tightly-wrapped package, and Kozaga was curious what it consisted of and how it was made. He began to carefully unwrap the one he had obtained. A friend who happened to be nearby saw him and cautioned that he not do so, since grisgris can be powerful and are not to be meddled with once they are made. But in this case Kozaga was not concerned about the risk, and was not concerned that people would see him as complicit or otherwise taking sides by learning about

⁵ One issue is that learning about occult practice can compromise the researcher’s position, particularly if the researcher lives long-term in that place, because, even as it has become more widespread, occult practice is seen as dangerous and morally complicated.

LOUISA LOMBARD AND MOBITO KOZ AGA

77

occult practice. He found the gris-gris to contain what people always said it contained: carefully-folded pieces of paper with verses from the Koran written on them. Kozaga found it interesting to be able to confirm what people said about the material components of gris-gris. But it was also the case that the dissected gris-gris was as lively as any cadaver. The force it had while in use was gone.

Conclusion Our experience has been that it is close to impossible to be impartial when it comes to seeking knowledge about civilian occult protective agency during war. Perhaps this is one reason why the most detailed treatments of occult practice during war have had their starting point from within an armed group—like David Lan in Zimbabwe and Danny Hoffman in Sierra Leone and Liberia. In these cases, there was a group the ethnographer was working closely with and coming close to assimilating into. Although the ethnographer might not see himself as taking a side, there was an effective side-taking in that he was working with one group much more closely than others. For researchers who are trying to canvas perspectives a bit more widely, or who are working with civilians whose attitudes about what kinds of practices are licit and good and which destructive are diverse, the endeavour of ascertaining how occult practice is part of civilians’ protective agency can be fraught. We say this not to discourage, but to show why occult action and magical thinking are worthy of serious consideration, rather than prior dismissal on grounds of their falsity. Appeals to invisible forces, and attempts to influence and channel them, are widespread during war, and are made by fighters and civilians alike. Sometimes those appeals work. Sometimes they do not. Only sometimes is it possible to identify what will fall into which category in advance. That is part of why the young man who confessed to the pastor, previously such an eager adherent, would later describe himself as a ‘totally lost soul’. The madébo who was torn limb from limb was thoroughly lost as well. During the war in CAR more Central Africans became directly involved in occult practice than they had before, and in many cases they did so for the specific aim of protecting themselves and their fellows. But the effects of their having done so, both for themselves and their communities, are hard to parse. Even amid other forms of violence, occult agency often counts as particularly transgressive, because it entails spurning everyday ways of comporting oneself, such as consorting with taboo animals and plants. This underscores the need for caution and care when making arguments about the status and effects of civilians’ using magic to protect themselves. Occult practice, from relatively benign and untrained methods to those that require extensive apprenticeship, is incredibly pervasive, and often essential to people’s experiences of taking action to protect themselves.

78

TRANSFORMATIONS IN OCCULT PROTECTION AMID VIOLENCE

References Appiah, K. A. (1992). In My Father’s House: Africa in the Philosophy of Culture. Oxford: Oxford University Press. Bonhomme, J. (2009). Les voleurs de sexe. Anthropologie d’une rumeur africaine. Paris: Le Seuil. Collins, R. (2009). Violence: A Micro-Sociological Theory. Princeton: Princeton University Press. Douglas, M. (1999). ‘Sorcery Accusations Unleashed: The Lele Revisited. 1987’, Africa: Journal of the International African Institute, 69/2: 177–93. Feierman, S. (1985). ‘Struggles for Control: The Social Roots of Health and Healing in Modern Africa’, African Studies Review, 28/2–3: 73–147. Giles-Vernick, T. (1996). ‘Na Lege Ti Guiriri (on the Road of History): Mapping Out the Past and Present in M’bres Region, Central African Republic’, Ethnohistory, 43/2: 245. Girard, R. (1977). Violence and the Sacred. Baltimore: Johns Hopkins University Press. Glawion, T., and Vries, L. de. (2018). ‘Ruptures Revoked: Why the Central African Republic’s Unprecedented Crisis Has Not Altered Deep-Seated Patterns of Governance’, The Journal of Modern African Studies, 56/3: 421–42. Cambridge University Press. DOI: 10.1017/S0022278X18000307 Hoffman, D. (2011). The War Machines: Young Men and Violence in Sierra Leone and Liberia. Durham, NC: Duke University Press. Krohn-Hansen, C. (1994). ‘The Anthropology of Violent Interaction’, Journal of Anthropological Research, 50/4: 367–81. [University of New Mexico, University of Chicago Press]. Lan, D. (1985). Guns & Rain: Guerrillas & Spirit Mediums in Zimbabwe. Oakland: University of California Press. Lombard, L. (2016). State of Rebellion: Violence and Intervention in the Central African Republic. London: Zed Books Ltd. Lubkemann, S. C. (2005). ‘Migratory Coping in Wartime Mozambique: An Anthropology of Violence and Displacement in “Fragmented Wars”’, Journal of Peace Research, 42/4: 493. Marchal, R. (2016). Brève histoire d’une transition singulière, p. 73. Réseau des organisations de la société civile de Centrafrique pour la gouvernance et le développement. Retrieved 27 January 2022, from Masullo, J., and Jentzsch, C. (2019). ‘Violent or Non-Violent Action? Civilian Resistance and Tactical Choice in Colombia’s and Mozambique’s Civil Wars’. Presented at the APSA Annual Meeting, Washington, DC. Mauss, M. (1936). ‘Les techniques du corps’, Journal de Psychologie, 32: 3–4. Mayneri, A. C. (2014). Sorcellerie et prophétisme en Centrafrique: l’imaginaire de la dépossession en pays banda. Paris: Karthala Editions. Richards, P. (2009). ‘Dressed to Kill: Clothing as Technology of the Body in the Civil War in Sierra Leone’, Journal of Material Culture, 14/4: 495–512. SAGE Publications Ltd. DOI: 10.1177/1359183509345950 UNHCR. (2021). ‘Operational Data Portal: Central African Republic’. Retrieved 12 January 2022, from

LOUISA LOMBARD AND MOBITO KOZ AGA

79

UN Panel of Experts Established pursuant to Security Council. (2021). ‘Letter Dated 25 June 2021 from the Panel of Experts on the Central African Republic Extended Pursuant to Resolution 2536 (2020) Addressed to the President of the Security Council’, UN. Wenzel, J. (2010). Bulletproof: Afterlives of Anticolonial Prophecy in South Africa and Beyond. Chicago, IL: University of Chicago Press. World Bank Group. (2012). Understanding Access to Justice and Conflict Resolution at the Local Level in the Central African Republic. Washington, DC: World Bank. Retrieved 28 January 2022, from

4 Civil resistance against jihadists Perceptions of, and resistance against, IS governance in Mosul Isak Svensson and Alanna Smart

Introduction In June 2014, on the grounds of the Great al-Nuri Mosque in the centre of the old city of Mosul, Abu Bakr al-Baghdadi proclaimed the dawn of the ‘Caliphate’ in Iraq and Syria (BBC 2017). Mere weeks earlier, a small contingent of 200 Islamic State (IS) fighters had arrived at the city gates to little opposition. By the time 400 IS reinforcements followed, the 25,000-strong Iraqi Security Forces that guarded the city had abandoned their posts and fled (Hoekstra 2020). Leaving weapons, ammunition, and vehicles in their wake, the IS fighters quickly established control, marking the beginning of sweeping changes to the political, social, and economic life of the city. Some Mosul residents welcomed IS and supported their rule, with videos published on social media showing many jubilantly cheering in the streets alongside pro-IS marches (IOM 2019). As a predominantly Sunni city—approximately 80 per cent (UN Habitat 2016)—many had felt hopelessly marginalized by Nouri alMaliki’s sectarian policies and had suffered at the hands of the predominantly Shia security forces. Consequently, many hoped IS would re-empower the Sunni community and believed their announcements that they were there to liberate Mosul from oppression and corruption. Many others fled—engaged in an exit strategy that allowed them to evade the violent and repressive grasp of IS. Others resisted violently—they took up arms and fought back. This was especially the case for those belonging to religious and ethnic minorities, including Christians, Kurds, Yazidis, and Turkomans. Civilians often exercise significant agency when faced with turbulent change and violent contexts, engaging in strategies to protect themselves, or others, from violence (Krause 2018), maintain their autonomy (Kaplan 2017; Masullo 2021), and preserve their ways of life (Arjona 2015, 2016). These coping mechanisms can play out through individual agency, such as fleeing from violence (e.g. Finkel 2017) or hiding individuals from persecution or genocide (Oliner and Oliner 1988;

Isak Svensson and Alanna Smart, Civil resistance against jihadists. In: Civilian Protective Agency in Violent Settings. Edited by: Jana Krause, et al., Oxford University Press. © Isak Svensson and Alanna Smart (2023). DOI: 10.1093/oso/9780192866714.003.0005

ISAK SVENSSON AND AL ANNA SMART

81

Braun and Stallone within volume), or through collective agency, such as forming zones of peace (Hancock 2007) or community-initiated militias and vigilante movements (Jentzsch within volume; Moncada within volume). In this chapter, we are interested in a particular form of protective civilian agency used by those who remained in Mosul, living under IS rule: civil resistance. While civilian resistance refers to the refusal to act according to implicit and explicit demands by armed groups and may include violent resistance (Krause et al. within volume), civil resistance is a subset of resistance practices that is carried out non-violently. In the case of Mosul, it includes concrete behavioural acts that are not in line with the dictates, rules, laws, and moral codes implemented by IS. Despite IS’s repressive policies, and an initial outpouring of civilian support, civil resistance under IS was remarkably common. Far from being only passive victims, enthusiastic collaborators, or engaging in violent resistance, Mosul residents displayed several non-violent adaptive strategies through which they managed their relations with IS. Whilst some of these actions served to protect themselves and others from violence, most civil resistance actions witnessed in Mosul served a different purpose—protection from occupation and external interference. Civil resistance is a strategy through which civilians can preserve their autonomy and protect their ways of life, including their traditions, norms, and values from those who wish to destroy them. Consistent with this view, Arjona (2015, 2016) finds that rebel groups which enact extreme interventions into local affairs are more likely to trigger widespread and ‘full’ resistance. As we explore here, whilst perceptions of the quality of IS governance were generally positive, the extreme interventions of IS, an occupying force with large numbers of foreign fighters, into all areas of civilian life, triggered widespread civil resistance. Whereas research on civil resistance against IS has been scarce (although note Aarseth 2018; Svensson and Finnbogason 2021), there is a significant body of research on the dynamics and outcomes of civil resistance more broadly (Sharp 1973; Schock 2004; Chenoweth and Stephan 2011; Nepstad 2011; Vinthagen 2015). Civil resistance builds on the idea that power is constructed through different social support structures—and that it is these support structures that subsequently enable a powerholder to be in place. Civil resistance undermines the power-structure by carrying out actions that shift the social basis of power. Using survey data of 1,024 participants who lived under IS rule in Mosul, gathered between October and November 2018, this chapter explores how civilian agency manifests under Islamist regimes.¹ We discuss how people resisted, who resisted, and why they chose to do so. In doing so, we highlight the perseverance of this form of civilian agency in even the most repressive and brutal context.

¹ The survey methodology and a comprehensive analysis of the result is described in Svensson et al. (2022).

82

PERCEPTIONS OF, AND RESISTANCE AGAINST, IS GOVERNANCE

Methodology Our chapter relies on the testimonies of 1,024 residents of Mosul, surveyed between October and November 2018—a year after IS’s retreat from the city. All participants lived under IS rule in Mosul between June 2014 and July 2017. The survey sampled 41 districts in West Mosul, with 20–25 individuals per district. All participants were aged 18 years or older, with 52.5 per cent women and 47.5 per cent men. West Mosul was chosen for the stratified sample of populated neighbourhoods based on six potential confounding variables: area (size), ethnicity, highways, houses, pre-2014 security, and sleeper cells. For example, larger areas may systematically see more military activity than smaller neighbourhoods, as might areas with strategically important highways and major roads. The data for all six variables was taken from an ethnographic map of the city, a recent neighbourhood GIS file, and a UN Habitat Profile of the City. There were two key differences between West and East Mosul. West Mosul’s constituency was predominantly Arab or ArabMixed, with no substantial Kurdish or Turkoman populations. West Mosul also had more IS safe houses prior to 2014, compared to other areas of the city. Given the greater presence of IS safe houses, and that IS received its support from a predominantly Sunni Arab constituency, it is reasonable to assume that West Mosul would have generally been more supportive of IS than other parts of the city, and that this may be reflected in the survey answers. All behavioural studies, particularly those reliant on self-reporting, carry the risk that individuals may misrepresent events and behaviour. Because IS suffered defeat, respondents may be inclined to diminish the extent of their support or exaggerate the extent of their resistance. Incentives to misrepresent behaviour are further compounded by the occurrence of retributive violence against alleged IS supporters since IS’s fall. Despite this, there is reason to suggest the responses are trustworthy. First, participants were assured of the strictest anonymity and confidentiality, and the survey was completed via digital devices, precluding in-person interactions. This reduced the likelihood of social desirability bias and concerns regarding retributive risks. Second, for each specific civil resistance action registered, generally only a minority of the sample reported engagement. If there were strong incentives to inflate one’s resistance, we would expect to see higher levels of self-reporting across all or most actions. Third, participants registered positive opinions about IS in some respects and negative opinions in others. If there were significant social desirability bias present, we would expect to see negative responses to IS in all or most aspects. Consequently, whilst we cannot eliminate the possibility that some respondents may have overstated their resistance towards IS, we do believe that the responses were generally reasonable and trustworthy and should reveal an accurate estimation of respondents’ perceptions of IS governance and their experiences of civil resistance.

ISAK SVENSSON AND AL ANNA SMART

83

How did civilians in Mosul perceive IS governance? Driven by their particular form of Salafi-jihadist ideology and their quest to build a twenty-first-century caliphate, IS sought to dominate the political, economic, and social lives of Mosul’s citizens (Gerges 2017). Borrowing from Arjona’s (2015, 2016) terminology, IS established a fully fledged ‘rebelocracy’ that sought full intervention in all civilian affairs. The new foundations for governance in Mosul, and the obligations of its citizens, were laid out in the Document of the City, published shortly after IS’s arrival in June 2014 (Revkin 2016a). They overhauled the governance systems, creating new diwans (ministries), including new policing and judicial systems, and departments for the provision of public goods (UN Habitat 2016; Abdul-Ahad 2018). They managed the economy, and codified new norms and practices that strictly regulated the public and private lives of Mosul’s citizens. Strict dress codes were implemented, with women required to wear allblack, multi-layered abayas, complete with a niqab and gloves (Mahmood 2015a). Meanwhile, men were forced to grow their beards (Mahmood 2015b). Codes of conduct, including rigid segregation of the sexes, were implemented in all public spaces (Michlig et al. 2019).

Perceptions of IS governance To understand how IS governance was perceived, the questions were divided into six categories: (1) safety and security; (2) crime and corruption; (3) distribution of wealth and resources; (4) employment and economic opportunities; (5) social trust among residents; and (6) morality in society. Specifically, they were asked whether they perceived an improvement or deterioration in relation to the former Iraqi governance, using a 5-point scale from ‘much worse’ to ‘much better’.² The results, presented in Figure 4.1, show an astounding consensus that IS governance was much better than its predecessor on all dimensions surveyed. The areas of safety and crime were perceived as most improved, with 90 per cent and 87 per cent stating that they had improved significantly under IS rule; 75 per cent also believed that both morality and distribution were much better under IS. Lastly, 72 per cent and 74 per cent felt that governance of the economy and interpersonal trust were much better. For some areas, these results are unsurprising. IS increased policing, overhauled the justice systems, and established civilian rights for the protection of self and property (Lister 2014; Revkin 2016b), establishing law and order in a ² To gauge perceptions of governance, we posed the following question: ‘do you think Daesh made the following better or worse?’ with regards to six categories, ‘safety and security’, ‘crime and corruption’, ‘distribution of wealth and resources’, ‘employment and economic opportunities’, ‘social trust among residents of Mosul’, and ‘morality in society’.

84

PERCEPTIONS OF, AND RESISTANCE AGAINST, IS GOVERNANCE 100 90

Percentage of participants

80 70 60 50 40 30 20 10 0

Safety

Much worse

Crime

Distribution

Somewhat worse

Economy

About the same

Trust

Somewhat better

Morality Much better

Figure 4.1. Perspectives on IS governance.

city that had been marred by human rights abuses and violence (HRW 2013a, 2013b, 2014; Collard 2014). They projected the image of an uncorrupted ruler and established divisions which investigated complaints against their members and punished those found guilty of indiscipline and misconduct (Revkin and McCants 2015; Revkin 2016a, 2016b; Callimachi 2018). They embarked on social reforms and welfare programmes, including free health care, soup kitchens, and public transport subsidies (Lister 2014; Mahmood 2015a). Crucial infrastructure, such as roads and footpaths, and public services, such as waste removal and street cleaning, were vastly improved (Lister 2014; Robinson et al. 2017). Heavily influenced by Saddam Hussein’s Faith Campaign of the 1990s, Mosul was significantly more conservative than much of Iraq (Al-Aqeedi 2016). Consequently, some IS regulations, such as the banning of alcohol, might not have appeared as restrictive but rather as moral, in line with the city’s long-standing conservative values. In other categories, the results are more puzzling. Perceptions of social trust improved markedly despite IS’s use of informants. However, it is possible that this is linked to broader changes in demographics as minorities fled the city, as social trust tends to be higher amongst ethno-religious groups. The impact of IS on the economy of Mosul is also disputed, with some claiming the economy almost collapsed (UN Habitat 2016; Lafta et al. 2018; Michlig et al. 2019), whilst others suggest the group had only a moderate effect with increased economic activity in some areas, such as marketplace activity (Robinson et al. 2017). Overall, however, perceptions of IS were generally favourable across all surveyed dimensions of governance. It should be noted that the survey did not include those who fled. Many who were deeply opposed to IS’s state-formation project, such as those from religious and ethnic minorities, were already exiled when the

ISAK SVENSSON AND AL ANNA SMART

85

survey was implemented. For those who remained, however, IS governance was seen as quite effective.

Perception of IS members We also asked respondents about their perceptions of IS members.³ As demonstrated in Figure 4.2, most respondents had ‘extremely negative’ attitudes towards IS members, and very few had extremely positive attitudes. Overall, negative perceptions clearly dominate. This is unsurprising, as the group’s fighters were broadly feared for their barbarous treatment of those who opposed them. But beyond their savagery, it is likely that IS’s composition also affected civilian perceptions. IS’s ranks were diverse, having received unprecedented numbers of foreign fighters who had travelled to commit their lives to jihad. This influx was encouraged by IS, with many high-ranking positions, including al-Wali (Governor) of Mosul, given to foreigners (UN Habitat 2016; Mironova 2019). Across Iraq and Syria, an estimated 40,000 individuals from eighty countries travelled to live under IS, with at least 5,000 originating from Europe (Cook and Vale 2018; Micheron 2020). In 2016, US officials estimated approximately 1,000

Extremely positive Very positive Somewhat positive Somewhat negative Very negative Extremely negative 0

20 40 60 80 100 120 Percentage of participants (for each type of IS member) Foreign: Outside Middle East

Middle East

Iraq

140

Mosul

Figure 4.2. Perceptions of IS members. ³ To gauge participants’ perceptions of repression, the following question was posed during the survey: ‘How do you feel towards different types of Daesh members?’, with regards to the following categories, ‘Daesh members from Mosul originally’, those from ‘other parts of Iraq’, the ‘Middle East’, and ‘from outside of the Middle East’.

86

PERCEPTIONS OF, AND RESISTANCE AGAINST, IS GOVERNANCE

of the 3,000–5,000 fighters in Mosul were foreigners (Seldin 2016). Many of these foreign recruits had no links to, and little knowledge of, the local people and culture. Whereas local recruits often joined IS because of anger at Sunnimarginalization or financial hardship (Tawfeeq and Carter 2014; Mironova 2019), foreign recruits were deeply committed to radical religious narratives and the IS project (Dawson and Amarasingam 2017). This is crucial, as it likely impacted the extent to which locals identified with IS. Studying the impact of jihadist foreign fighters in Chechnya and Syria, Rich and Conduit (2015) found that an influx of radical foreigners, who had little solidarity with the local people and their grievances, negatively affected cohesion and the domestic perceptions of armed groups, in part through the creation and exacerbation of ideological cleavages. Furthermore, whilst the responses suggest that both foreign and local fighters were viewed negatively, anecdotal evidence suggests that civilian–IS relations differed between foreign recruits and their local counterparts. For instance, locals were reportedly particularly fearful that their daughters would be forced to marry foreigners, causing many families to flee (Lafta et al. 2018). The survey responses on perceptions of governance and IS members reveal an interesting observation: whilst governance was generally perceived to be high quality, many residents held negative views of IS forces. This suggests that civil resistance was not related to the quality and efficiency, or lack thereof, of governance, but rather was related to the rejection of IS and its members, more broadly.

Civil resistance against IS in Mosul To understand civil resistance to IS, nonviolent resistance can be classified into three categories: public actions of protest, non-cooperation, and everyday resistance. Public actions of protest are expressions of civil resistance that occur in public settings. These include demonstrations or sit-ins that openly highlight participants’ grievances against authorities. As highly visible acts of resistance, they can be accompanied by significant risks in repressive societies. Non-cooperation, on the other hand, refers to the curtailment and withholding of social, political, or economic cooperation. Non-cooperative actions might include tax avoidance or the halting of production, thus weakening the resource extraction capacities of the (proto-)state. The third category is everyday forms of resistance. First elucidated in Scott’s (1985) seminal book Weapons of the Weak, everyday resistance refers to challenges to authority that do not include open confrontations with authority or the norms it espouses. Rather, it includes daily challenges as manifested in the private lives of citizens, such as defying IS’s prayer mandates, listening to non-religious music, and practising forbidden sports.

ISAK SVENSSON AND AL ANNA SMART

87

Not all acts of resistance are necessarily performed as a way of publicly revealing preferences or non-consent towards the incumbent rulers. Nor are they necessarily ‘political’, in a stricter sense of the word. Many acts of civil resistance are personal decisions about how to carry out one’s daily life in a context where daily life is highly regulated. It is the lived experiences of disobedience. In societies where leaders are demanding the complete transformation of social and individual lives, in this case according to divine regulations (as stipulated by Sharia law), any act of non-compliance can be seen as a form of resistance—even when the motivations for such resistance can vary considerably. The civil resistance actions included within the survey of Mosul residents are elucidated in Figure 4.3. Beginning with Gene Sharp’s (1973) seminal list of 198 methods of non-violent action, viable resistance behaviours were chosen based on discussions with the local team and thoughtful consideration of local culture, to ensure the concept of ‘civil resistance’ was appropriately applied and captured within the Iraqi context. Overall, the results suggest that everyday resistance and non-cooperation were the most common forms of civil resistance under IS, with 83 per cent and 78 per cent engaging in some level of these actions, respectively. Far fewer—27 per cent of the sample—stated they had engaged in some level of public actions of protest. This disparity in engagement between highly visible actions (public protests) and less visible actions (non-cooperation and everyday resistance) is unsurprising,

Percentage of participants

90 80 70 60 50 40 30 20 10 0

Public Showing slogans/banners; Directly confronting IS members; Occupying public spaces; Participating in protests/ demonstrations.

Noncooperation Quitting school/university; Not going to work; Refusing to cooperate with IS institutions; Refusing to participate in public events; Avoiding paying taxes; Stopping to produce; Using alternative markets; Resigning from job.

Everyday resistance Not praying regularly; Working slowly;Listening to nonreligious music; Attending the funerals of IS victims; Playing musical instruments; Smoking/ drinking alcohol; practising forbidden sports; Delaying compliance with IS orders.

Figure 4.3. Distribution of civil resistance actions and their specific actions.

88

PERCEPTIONS OF, AND RESISTANCE AGAINST, IS GOVERNANCE

given the often-brutal reprisals that accompanied challenging IS authority. However, even relatively private forms of civil resistance carried significant risk to participants.

Everyday resistance Following their march into the city, IS quickly established control over the religious institutions, dictating the content of Friday prayers, appointing pro-IS imams, and executing imams who refused to pledge their allegiance (Zarocostas 2014). IS also expected residents to attend prayers regularly. Even so, 34 per cent of survey participants said they neglected to pray regularly, as was mandated by IS ordinance. According to one resident, many soon stopped attending the Mosque altogether, to evade being pressured into giving an oath of allegiance (BBC 2015). Other common acts of everyday resistance included the deliberate slowing of workplace productivity (22 per cent), listening to non-religious music (20 per cent), and attending the funerals of those killed by IS (19 per cent). Under IS, only music ‘without instruments and incantations that were not quite singing, with successions of voices praising the group that gave it some sort of rhythm’ were allowed and were played by the official IS radio station—al-Bayan (Mu¨ller and Castelier 2017). However, many continued to listen to banned music and radio stations. This resistance was not only private, as one resident remembered: ‘When you got in a taxi, if you did not have a long beard, long hair, or military dress, and if you did not seem too suspicious, then the driver would often play forbidden music on the BBC, a forbidden radio station’ (Mu¨ller and Castelier 2017). A new radio network—Radio al-Ghad—was also established by residents who had fled the city to Kurdish Peshmerga-held territory (Stevens 2018). Alongside playing banned music, the radio also began to broadcast live calls from within the city. The calls predominantly came from women, likely a result of their relegation to the private sphere that meant women’s voices were less likely to be recognized than men’s.⁴ IS members began to phone-in themselves, attempting to challenge the presenters on air. Soon, the radio station began allocating them timeslots and pressed them on issues of governance, such as public service provision (Stevens 2018). The least common acts of everyday resistance, with under 5 per cent participating, were delaying compliance with IS instruction and watching/practising forbidden sports. IS had banned sports, in particular football, as they were seen as western, un-Islamic imports that were incongruent with Sharia law. As with other ⁴ We can draw interesting comparisons with Braun and Stallone (within volume) who found that performative femininity and expected gender roles allowed women to play unique roles in resistance networks during the Holocaust, whereas here, women in Mosul were able to participate in resistance in unique ways because of the expectations that women would remain in the private sphere.

ISAK SVENSSON AND AL ANNA SMART

89

regulations, IS employed harsh punishments against those who were caught breaking them. For example, in March 2015, IS militants in Mosul released a decree that anyone caught watching the El Clásico, a match between Real Madrid and FC Barcelona, would receive eighty lashes (Almousa 2015). Engaging in acts of everyday resistance entailed considerable risks. The Diwan al-Hisbah (Department of Accountability), responsible for ensuring compliance with established rules and doling out punishments for transgressions, were notorious for their brutality (Al-Aqeedi 2016). Aided by informants, the group patrolled the streets fining or flogging those who were caught smoking, drinking, or listening to music (HRW 2016). Death penalties were also meted out for everyday offences with seemingly casual indifference. In 2016, two young men were shot dead outside a Mosque for failing to attend Friday Prayers (Withnall 2016). When considering the risks, the significance of participation in everyday acts of resistance becomes apparent. According to a report by Niqash, some locals viewed their smoking as a direct challenge to IS rule (Shamdeen 2015). Such sentiment was not isolated. During interviews conducted by Vale (2020 p. 15), one woman recalled that before escaping from IS territory, her sister had applied makeup, against IS codes, in order to ‘demonstrate its lack of control over her face and body’. Another woman recalled taking a ‘selfie’ with a friend in Mosul’s centre, using a banned mobile phone and without a mahram (mandatory male escort), explaining ‘it was not just taking photos for fun, but we intended to bother them because we really hated them and tried to do anything against them’, meanwhile understanding the risks, saying ‘we could have lost our lives from that’ (Vale 2020 p. 15). These accounts reveal that even the most seemingly mundane of everyday actions were often imbued with symbolic resistance, intended to prevent what they perceived as overbearing and unwarranted external interventions on their private lives, and were carried out with the knowledge of the huge risks they faced.

Non-cooperation The most prevalent form of non-cooperation was absenteeism from education, with over a third reporting quitting school/university as a deliberate act of resistance against IS. This is corroborated by Aarseth (2018) and Lafta et al. (2018) who found that attendance dropped dramatically following the fall of Mosul. This was in sharp contradiction of an IS ordinance published in November 2014, that mandated all children attend school from the ages of 6 to 15 years and dictated significant changes to the curricula (Arvisais and Guidère 2020). Schools were viewed as breeding grounds of potential recruits, and IS members sought to recruit children with promises of material goods and by encouraging them to be informants. The new curriculum focused on Sharia education and military training, including fighting and the use of arms (Hawramy and Shaheen 2015; IID 2016). Textbooks

90

PERCEPTIONS OF, AND RESISTANCE AGAINST, IS GOVERNANCE

mirrored these new priorities, often containing illustrations of children carrying weapons and references to extreme violence. For example, a maths book might ask how many ‘unbelievers’ could be killed by a suicide bomber. Many families withdrew their children from school, to protect them from recruitment (HRW 2019) and from IS’s violent ideology, believing it to be immoral, un-Islamic, and at odds with the pre-existing values and norms of Mosul (Aarseth 2018; Michlig et al. 2019). Reportedly, it was frequently teachers who advised students they trusted to stop attending. In higher education, purging of ‘unislamic’ studies such as philosophy, political science, and law caused many to view IS education as antiknowledge and anti-intellectual. Those who warned others or refused to send their children, or to attend themselves, were aware of the risks. As one teacher explained, ‘our opposition to Daesh was secret, it was a closed circle between the teachers, the pupils, the supervisor and the headmaster. . . . If that information was passed on to Daesh, they would have beheaded us immediately’ (Aarseth 2018 p. 53). Despite this, many continued this form of resistance throughout all three years under IS, as is evidenced by the difficulties reintegrating students back into education following IS’s fall (Fox 2019). Marginally fewer respondents engaged in non-cooperation through absenteeism from work or refusing to cooperate with IS institutions (33 per cent and 25 per cent respectively). As with students, many teachers avoided attending work, further affecting the ability of IS to enforce its mandated education. That IS saw these actions as a threat to their project is clear, as they would post warnings threatening to execute teachers who failed to comply, along with the names of those who had been absent. According to reports, after being identified, teachers would frequently return to work for a short amount of time, before stopping attending again. Many teachers refused to cooperate with the curriculum mandates, and several were subsequently executed (GCPEA 2018). One such example is Ashwaq al-Naimi, a primary school teacher who was publicly executed in September 2014, and has since been held up on social media and local news as a ‘martyr and icon of Mosul’s resistance’ (Aarseth 2018 p. 57). Other forms of non-cooperation included refusing to participate in public events and avoiding taxes, each reported by about 20 per cent of the respondents. Taxation represented a substantial portion of IS income, estimated to be greater than oil extraction, kidnapping, looting, and foreign donations (Heißner et al. 2017). Mosul in particular was a major source of revenue, with IS records showing that the majority of its revenue came from protection rackets, extorting commercial, reconstruction, and oil sectors (Johnston and Bahney 2014). The taxes imposed by IS were significantly higher than those imposed by the Iraqi government, and constituted a considerable burden (UN Habitat 2016). Zakat (mandatory charity) absorbed as much as 10 per cent of citizens’ earnings— defended by the group as a necessity borne from being a ‘nation in a time of war’ (US Homeland Security Comnittee 2016 p. 10). Overall, greater compliance with

ISAK SVENSSON AND AL ANNA SMART

91

tax regulations over other forms of non-cooperation is understandable, given the severe punishments for noncompliance. Avoiding utilities taxes was considered theft and was punished by dismemberment of one’s hand (Abdul-Ahad 2018). Refusal to pay Sharia-mandated zakat was considered apostasy and therefore punishable by death (Revkin 2016a). Regardless, approximately a fifth of respondents reported tax avoidance. Whilst avoiding tax was an act of resistance, it is likely that the indirect consequences of IS rule, including falling incomes and increased costs of living (Hawramy and Shaheen 2015; Michlig et al. 2019), made tax avoidance a necessary strategy for some residents to ensure they had adequate finances for basic commodities.

Public acts of protest The most popular public act of protest, reported by 11 per cent of the sample, was showing slogans and/or banners. The most prominent example of this was the painting of the letter ‘M’ across walls throughout the city. Reportedly standing for several meanings including Muqawam (resistance), Mucaarad (opposition), and Muwajah (confrontation), the letter was adopted by numerous resistance groups and anti-IS individuals to display unity and defiance against IS rule (Damon et al. 2016). Citizens also daubed slogans including ‘ISIS are dogs’ and ‘we fought for the people’s choice’ (Abdulrazaq and Stansfield 2016; Fordham 2016), and posters of Saddam Hussein were hung around the city, angering IS fighters (Abdulrazaq and Stansfield 2016). Surprisingly, given the risks entailed, 8 per cent of respondents said they directly confronted IS members. According to interviews by Aarseth (2018), students engaged in direct confrontations with IS members over conduct in schools, sometimes citing frustrations regarding inconsistent rules that reportedly changed from day to day. In one witness report, a student recounted an instance in which female students had confronted the hisbah on campus, stating that they had difficulties performing their studies due to the enforcement of the niqab and gloves, reportedly arguing it ‘was not right, we have our rights, and they should not force people to behave in a way they did not want’ (Aarseth 2018 p. 57). Rather than punishing the students, the hisbah only told them to continue focusing on their studies. Occupying public spaces and holding demonstrations also occurred under IS, but rarely. Less than 6 per cent and 5 per cent of respondents engaged in each, respectively. Whilst this shows that public actions of protest against IS were rare, it highlights that some individuals did engage in public defiance. For example, Vale’s (2020) study of women’s resistance to IS found notable examples of public protests, including a woman who staged a sit-in protest outside of a Mosul church where her husband was imprisoned, refusing to move until the hisbah brought

92

PERCEPTIONS OF, AND RESISTANCE AGAINST, IS GOVERNANCE

her husband to her. In 2014, Mosul’s residents protested the planned destruction of the city’s al-Habda minaret, effectively blocking IS and forcing their retreat on the issue until the minaret’s eventual demolition in 2017 (Isakhan and Meskell 2019).

Understanding the relationship between IS governance and civil resistance In this chapter, we have shown that civil resistance strategies were varied, with the prevalence of resistance amongst Mosul’s residents suggesting that IS’s project was (at least partly) resisted by many. By illustrating the myriad of strategies civilians used to engage in resistance—public protests, non-cooperation, and everyday resistance—this chapter has called attention to the remarkable prevalence and perseverance of civilian agency, in one of the world’s most brutal, repressive, and violent contexts. Moreover, this chapter has tried to demonstrate that understanding Mosul’s widespread civil resistance as a consequence of governance is, by itself, insufficient in terms of an explanation. This is because the perception of governance does not appear to be a significant driver of resistance—people of Mosul appreciated several aspects of IS governance and yet engaged in various forms of civil resistance. Consequently, to understand why the people of Mosul resisted IS, it is not sufficient to take governance alone into account. Consequently, it is important to study civilian identification with the IS project. The overwhelmingly negative perceptions of IS members can help to explain why so many civilians resisted. Whilst IS was partly successful at governing, the group failed in enacting key components of successful governance: co-identification and popular legitimacy. As an occupying force pursuing extreme interference in civilian affairs, their attempts to establish a new unifying culture, built on radical interpretations of Islam, was incongruent with the city’s established identity and culture. This was likely exacerbated by IS’s reliance on foreign fighters, who lacked social and cultural links with the local people and were therefore unable to tap into legitimacy building narratives frequently founded in nationalist sentiment or shared cultural norms. In failing to adequately engage popular beliefs and posit themselves as representatives of their constituents, IS failed to strengthen identification with their constituents and thus did not achieve the comprehensive legitimacy necessary to produce consent from those they ruled. Attempts by IS to establish new norms were met with both individual and collective resistance. Thus, beyond serving as a means to protect themselves or others from violence, civil resistance in Mosul also served as a strategy through which civilians attempted to preserve their ways of life in the face of unwanted and extreme interference.

ISAK SVENSSON AND AL ANNA SMART

93

Acknowledgements This chapter relied on a survey, for which we acknowledge collaboration with Jonathan Hall and Erik Skoog, and appreciate earlier research assistance by Daniel Finnbogason, Dino Krause, and Luı´s Martı´nez Lorenzo. We appreciate the financial support for this research from the Marianne and Marcus Wallenberg Stiftelse (no. MMW 2013.0025) and Stiftelsen Riksbankens Jubileumsfond (NHS14-1701:1).

References Aarseth, M. B. (2018). ‘Resistance in the Caliphate’s Classrooms: Mosul Civilians vs IS’, Middle East Policy, 25/1: 46–63. DOI: 10.1111/mepo.12324 Abdul-Ahad, G. (2018). ‘The Bureaucracy of Evil: How Islamic State Ran a City’. The Guardian. Abdulrazaq, T., and Stansfield, G. (2016). ‘The Enemy Within: ISIS and the Conquest of Mosul’, Middle East Journal, 70/4: 525–42. Middle East Institute. Al-Aqeedi, R. (2016). Hisba in Mosul: Systematic Oppression in the Name of Virtue (Occasional Paper). George Washington Program on Extremism, pp. 1–13. Washington, DC: George Washington University. Almousa, H. (2015). ‘Isis Militants Threatened to Give 80 Lashes to Anybody Caught Watching El Clasico Because Game “Is Product of Decadent West”’. Raqqa is Being Slaughtered Silently. Retrieved 30 September 2020, from Arjona, A. (2015). ‘Civilian Resistance to Rebel Governance’, in Arjona A., Kasfir N., and Mampilly Z. (eds), Rebel Governance in Civil War, pp. 180–202. Cambridge: Cambridge University Press. DOI: 10.1017/CBO9781316182468.009 Arjona, A. (2016). Rebelocracy: Social Order in the Colombian Civil War. Cambridge Studies in Comparative Politics. Cambridge: Cambridge University Press. DOI: 10.1017/9781316421925 Arvisais, O., and Guidère, M. (2020). ‘Education in Conflict: How Islamic State Established Its Curriculum’, Journal of Curriculum Studies, 52/4: 498–515. Routledge. DOI: 10.1080/00220272.2020.1759694 BBC. (2015). ‘Inside Mosul: What’s life like under Islamic State?’ BBC News. BBC. (2017). ‘Battle for Mosul: IS “blows up” al-Nuri mosque’. BBC News. Callimachi, R. (2018). ‘The Case of the Purloined Poultry: How ISIS Prosecuted Petty Crime’. The New York Times. Chenoweth, E., and Stephan, M. J. (2011). Why Civil Resistance Works: The Strategic Logic of Nonviolent Conflict. New York: Columbia University Press. Collard, R. (2014). ‘Life in Mosul Gets Back to Normal, Even With ISIS in Control’. Time. Retrieved 14 December 2021, from Cook, J., and Vale, G. (2018). From Daesh to ‘Diaspora’: Tracing the Women and Minors of Islamic State, p. 76. London: International Centre for the Study of Radicalisation (ICSR), Kings College London.

94

PERCEPTIONS OF, AND RESISTANCE AGAINST, IS GOVERNANCE

Damon, A., Alkhshali, H., and Laine, B. (2016). ‘M for Muqawama: Secret Resistance Fighting Isis’. CNN. Retrieved 14 December 2021, from Dawson, L. L., and Amarasingam, A. (2017). ‘Talking to Foreign Fighters: Insights into the Motivations for Hijrah to Syria and Iraq’, Studies in Conflict & Terrorism, 40/3: 191–210. Routledge. DOI: 10.1080/1057610X.2016.1274216 Finkel, E. (2017). Ordinary Jews. Princeton, NJ: Princeton University Press. Fordham, A. (2016). ‘A s Assault On Mosul Looms, Resistance Grows—And ISIS Grows More Brutal’. NPR. Fox, T. (2019). ‘After Isil, Children Try to Catch up with School in Mosul’. Al Jazeera. Retrieved 30 September 2020, from GCPEA. (2018). Education Under Attack 2018. Global Coalition to Prevent Education from Attack. Retrieved 16 May 2022, from Gerges, F. A. (2017). ISIS: A History, revised edition. Princeton, NJ: Princeton University Press. Hancock, L. (2007). Zones of Peace. Bloomfield, CT: Lynne Rienner. Hawramy, F., and Shaheen, K. (2015). ‘Life Under Isis in Raqqa and Mosul: “We’re Living in a Giant Prison”’. The Guardian. Heißner, S., Neumann, P. R., Holland-McCowan, J., and Basra, R. (2017). Caliphate in Decline: An Estimate of Islamic State’s Financial Fortunes, pp. 1–20. London: The International Centre for the Study of Radicalisation (ICSR), King’s College London. Hoekstra, Q. (2020). ‘How Mosul Fell: The Role of Coup-Proofing in the 2014 Partial Collapse of the Iraqi Security Forces’, International Politics, 57/4: 684–703. DOI: 10.1057/s41311-019-00199-3 Human Rights Watch (HRW). (2013a). ‘Iraq: Investigate Fatal Police Shootings in Mosul’. Human Rights Watch. Retrieved 30 September 2020, from Human Rights Watch (HRW). (2013b). ‘Iraq: Abusive Commander Linked to Mosul Killings’. Human Rights Watch. Retrieved 30 September 2020, from Human Rights Watch (HRW). (2014). ‘Iraq: Security Force Abuses Compound Violence’. Human Rights Watch. Retrieved 30 September 2020, from Human Rights Watch (HRW). (2016). “We Feel We Are Cursed”: Life under ISIS in Sirte, Libya. Human Rights Watch. Retrieved 30 September 2020, from Human Rights Watch (HRW). (2019). “Everyone Must Confess”: Abuses against Children Suspected of ISIS Affiliation in Iraq. Human Rights Watch. Retrieved 17 January 2021, from International Organisation for Migration (IOM). (2019). West Mosul: Perceptions on Return and Reintegration Among Stayees, IDPs and Returnees, pp. 1–27. Baghdad: The UN Migration Agency–Iraq Mission. Iraqi Institution for Development (IID). (2016). Education in Mosul Under the Islamic State (isis) 2015–2016. Erbil: Iraqi Institution for Development.

ISAK SVENSSON AND AL ANNA SMART

95

Isakhan, B., and Meskell, L. (2019). ‘Unesco’s Project to “Revive the Spirit of Mosul”: Iraqi and Syrian Opinion on Heritage Reconstruction after the Islamic State’, International Journal of Heritage Studies, 25/11: 1189–204. Routledge. DOI: 10.1080/13527258.2019.1578988 Johnston, P. B., and Bahney, B. (2014). ‘Opinion | Hitting ISIS Where It Hurts’. The New York Times. Kaplan, O. (2017). Resisting War: How Communities Protect Themselves. Cambridge: Cambridge University Press. DOI: 10.1017/9781316671887 Krause, J. (2018). Resilient Communities: Non-Violence and Civilian Agency in Communal War. Cambridge: Cambridge University Press. DOI: 10.1017/9781108675079 Lafta, R., Cetorelli, V., and Burnham, G. (2018). ‘Living in Mosul during the Time of Isis and the Military Liberation: Results from a 40-Cluster Household Survey’, Conflict and Health, 12/1: 31. DOI: 10.1186/s13031-018-0167-8 Lister, C. (2014). Profiling the Islamic State (Analysis Paper No. 13). Brookings Doha Center, pp. 1–57. Washington: The Brookings Institution. Retrieved from Mahmood, M. (2015a). ‘Life in Mosul One Year on: “Isis with All Its Brutality Is more Honest than the Shia Government”’. The Guardian. Mahmood, M. (2015b). ‘Double-layered veils and despair … women describe life under Isis’. The Guardian. Masullo, J. (2021). ‘Refusing to Cooperate with Armed Groups Civilian Agency and Civilian Noncooperation in Armed Conflicts’, International Studies Review, 23/3: 887–913. DOI: 10.1093/isr/viaa090 Micheron, H. (2020). ‘IS “Advisers”: How Online Jihadists Guided European Volunteers to Syria’. Global Network on Extremism and Terrorism. Retrieved 17 January 2021, from Michlig, G. J., Lafta, R., Al-Nuaimi, M., and Burnham, G. (2019). ‘Providing Healthcare Under Isis: A Qualitative Analysis of Healthcare Worker Experiences in Mosul, Iraq Between June 2014 and June 2017’, Global Public Health, 14/10: 1414–27. DOI: 10.1080/17441692.2019.1609061 Mironova, V. (2019). From Freedom Fighters to Jihadists: Human Resources of Non State Armed Groups. Causes and Consequences of Terrorism. New York: Oxford University Press. DOI: 10.1093/oso/9780190939755.001.0001 Mu¨ller, Q., and Castelier, S. (2017). ‘After Years of Silence under IS, Music Is Back in Mosul’. Middle East Eye. Retrieved 30 September 2021, from Nepstad, S. E. (2011). Nonviolent Revolutions: Civil Resistance in the Late 20th Century. Oxford; New York: Oxford University Press. Oliner, S. P., and Oliner, P. M. (1988). The Altruistic Personality: Rescuers of Jews in Nazi Europe, pp. xxv, 419. New York: Free Press. Revkin, M. (2016a). Does the Islamic State Have a “Social Contract”? Evidence from Iraq and Syria (Working Paper No. 9). The Program on Governance and Local Development, pp. 1–37. Gothenburg: University of Gothenburg. Retrieved from

Revkin, M. (2016b). The Legal Foundations of the Islamic State (Analysis Paper No. 23). The Brookings Project on U.S. Relations with the Islamic World, pp. 1–45.

96

PERCEPTIONS OF, AND RESISTANCE AGAINST, IS GOVERNANCE

Washington: The Brookings Institution. Retrieved from Revkin, M., and McCants, W. (2015). ‘Experts Weigh in: Is Isis Good at Governing?’. Brookings. Retrieved 14 December 2021, from Rich, B., and Conduit, D. (2015). ‘The Impact of Jihadist Foreign Fighters on Indigenous Secular-Nationalist Causes: Contrasting Chechnya and Syria’, Studies in Conflict & Terrorism, 38/2: 113–31. Routledge. DOI: 10.1080/1057610X.2014.979605 Robinson, E., Egel, D., Johnston, P. B., Mann, S., Rothenberg, A. D., and Stebbins, D. (2017). When the Islamic State Comes to Town: The Economic Impact of Islamic State Governance in Iraq and Syria. Santa Monica: RAND Corporation. Retrieved 30 September 2020, from Schock, K. (2004). Unarmed Insurrections: People Power Movements in Nondemocracies. Minneapolis: University of Minnesota Press. Scott, J. C. (1985). Weapons of the Weak: Everyday Forms of Peasant Resistance. New Haven: Yale University Press. Seldin, J. (2016). ‘Islamic State Fighters Give Ground Around Mosul But Are Likely to Dig in | Voice of America—English’. VOA News. Retrieved 17 January 2021, from Shamdeen, N. (2015). ‘Iraqi Smokers Light up to Challenge Extremists’ Rule around Mosul’. Niqash. Retrieved 30 September 2020, from Sharp, G. (1973). The Politics of Nonviolent Action. New York: Sargent Publisher. Stevens, M. (2018). ‘Blood between Us’, The RUSI Journal, 163/1: 4–10. Routledge. DOI: 10.1080/03071847.2018.1449929 Svensson, I., Finnbogason, D., Krause, D., Lorenzo, L. M., and Hawach, N. (2022). Confronting the Caliphate: Civil Resistance against Jihadist Proto-States. New York: Oxford University Press. DOI: 10.1093/oso/9780197605608.001.0001 Svensson, Isak, and Finnbogason, D. (2021). ‘Confronting the Caliphate? Explaining Civil Resistance in Jihadist Proto-States’, European Journal of International Relations, 27/2: 572–95. SAGE Publications Ltd. DOI: 10.1177/1354066120976790 Tawfeeq, M., and Carter, C. J. (2014). ‘Isis Recruiting on the Rise in Iraq, Officials Say’. CNN. Retrieved 17 January 2021, from UN Habitat. (2016). City Profile of Mosul, Iraq: Multi-Sector Assessment of a City Under Siege, pp. 1–100. Erbil: United Nations Human Settlements Programme in Iraq (UN-Habitat). Retrieved 30 September 2020, from UNIC. (2014). UN Condemns Mass Executions in Iraq, Urges Leaders to Prevent Sectarian Reprisal. Cairo: United Nations Information Centre in Cairo. Retrieved 30 September 2020, from US Homeland Security Comnittee. (2016). Cash to Chaos: Dismantling ISIS’ Financial Infrastructure. House Homeland Security Committee Majority Staff Report. Washington: US Homeland Security Committee. Vale, G. (2020). ‘Defying Rules. Defying Gender?: Women’s Resistance to Islamic State’, Studies in Conflict & Terrorism, 1–24. Routledge. DOI: 10.1080/1057610X.2020.1816680

ISAK SVENSSON AND AL ANNA SMART

97

Vinthagen, S. (2015). A Theory of Nonviolent Action: How Civil Resistance Works. London: Zed Books. DOI: 10.5040/9781350251212 Withnall, A. (2016). ‘Isis “Beheads Teenage Boys for Listening to Pop Music and Missing Friday Prayers”’. The Independent. Retrieved 30 September 2020, from

Zarocostas, J. (2014). ‘Imam of Mosul Mosque Executed Before “Caliph” Gave Sermon’. The Sydney Morning Herald. Retrieved September 30, 2020, from

5 The durability of resistance Women’s high-risk mobilization for gender justice along the continuum of violence Julia Margaret Zulver

Introduction [We pay] homage to all of the women who ‘wove life’ resisting and defending their rights in Putumayo . . . We women speak and resist . . . because despite everything that happened, we are still alive, we are women of peace, women of harmony, women luchadoras [ fighters], women who say YES to peace and NO to violence.¹ The opening quote to this chapter highlights how a group of women in southern Colombia frame their grassroots mobilization—as a struggle for rights, a defence of their lives, and a resistance against violence. During times of intense conflict, women in different parts of Colombia undertook actions to keep themselves safe, including mobilizing for gender justice. As this chapter will argue, women’s mobilization in contexts of conflict can be considered an expression of civilian protective agency, because it serves as a way for them to illuminate— and actively resist—the gendered violence they face at the hands of armed actors. In Colombia, mobilization came with a high risk of retribution. Violence, including displacement, torture, rape, murder, kidnapping, disappearance, and threats, were widely used as a tool of territorial and social control in the conflict. The official Victims’ Registry documents over 9 million victims of the armed conflict, of whom 50.3 per cent are women (Unidad para las Vı´ctimas 2020). The Registry includes over 34,000 cases of sexual violence. While both men and women experienced violence during the armed conflict, the violence directed at women was additionally coloured by hegemonic gender norms and resulting power dynamics. Speaking about the paramilitary incursion of the early 2000s into the Montes de Marı´a in the north of the country, ¹ Written on a poster in the Alianza office in Mocoa, Putumayo. Photo on file with author.

Julia Margaret Zulver, The durability of resistance. In: Civilian Protective Agency in Violent Settings. Edited by: Jana Krause, et al., Oxford University Press. © Julia Margaret Zulver (2023). DOI: 10.1093/oso/9780192866714.003.0006

JULIA MARGARET ZULVER

99

the Centro Nacional de Memoria Histórica (CNMH) notes: ‘women and girls— dehumanized, dispossessed of their own bodies—became territories to be colonized by all-powerful masculinities who declared themselves the victors through [these acts of ] brutality’ (Centro Nacional de Memoria Histórica 2011). Meanwhile, the CNMH notes of paramilitaries in the south of the country: ‘Violence aimed at breaking supposed ties to subversives, punishing relations with guerrilla fighters, or emitting demoralizing messages to the enemy, marked the lives and bodies of the women of El Placer’ (Centro Nacional de Memoria Histórica 2012). In both parts of the country, daring to challenge the social order imposed by armed groups was dangerous work. However, as this chapter explains, women’s groups at the grassroots level did mobilize, not only to meet practical needs, but eventually in the strategic pursuit of gender justice.² Studies of civilian agency in conflict settings have set out to disprove the trope that in the face of violence, individuals and communities are powerless (see the literature review in the volume’s introduction). Indeed, scholars have documented the multi-faceted ways in which people are able to resist, negotiate, and create pockets of peace in otherwise violent settings. And yet, as Krause (2019 p. 1470) argues the ‘gender dimension of civilian agency and nonviolence has not been systematically analyzed’. Indeed, providing evidence on this ‘often neglected’ aspect is one of this volume’s stated goals (Krause et al. within volume). In parallel, other streams of scholarship increasingly challenge narratives that paint women as weak, vulnerable, or powerless actors in need of (masculinist) protection during conflict (Moser and Clark 2001; Carpenter 2005; Baines 2015; Goetz and Jenkins 2016). Studies of women as insurgents (Kampwirth 2006; Viterna 2013; Trisko Darden et al. 2019) or even as actors committing wartime violence (Sjoberg and Gentry 2015; Sjoberg 2016) continue to break with essentializing victim–perpetrator binaries. We know that women experience conflict differently; we also know that they find different ways to resist. Scholarly and policy analysis often appears to take mobilization for granted. However, there is nothing ‘automatic’ about women’s mobilization, as mothers, peacebuilders, or otherwise. Women assess their options and make calculated decisions about the risks and hardships they are willing to endure. To add to the conversation about gendered dimensions of civilian protective agency, this chapter examines women’s high-risk collective action, including the pursuit of gender justice. It asks two questions: (1) Why do women engage in high-risk collective action, despite the risk of violence? (2) Why do some women’s organizations persist while others do not? To answer these two research questions about the emergence and duration of mobilization, the chapter traces the actions of two women’s organizations: The ² The practical vs strategic gender interest distinction comes from the work of Maxine Molyneux (1985).

100

WOMEN’S MOBILIZ ATION FOR GENDER JUSTICE

Alianza de Mujeres Tejedoras de Vida (Women’s Alliance: Weavers of Life) and the Liga de Mujeres Desplazadas (League of Displaced Women). Both organizations engaged in high-risk collective action³ during some of the most violent moments in the Colombia conflict. They also took advantage of a decrease in violence to actively and publicly support peace when the Government signed accords with the Marxist-Leninist Fuerzas Armadas Revolucionarias de Colombia (FARC) rebels in 2016. In the current moment of ‘precarious peace’ (Maher and Thomson 2018), however, Alianza is once again engaging in high-risk mobilization, while Liga is losing mobilizational momentum. These cases provide valuable comparative evidence by contrasting organizations operating in similar conflict contexts, but with different mobilizational practices over time. This chapter analyses the emergence of gendered civilian agency and the durability of women’s resistance during different iterations of conflict. Beyond the war/peace dichotomy, this chapter focuses on women’s agency more comprehensively—beyond their stereotypical roles as peacebuilders or mothers— whose experiences with violent conflict catalyse them to mobilize for gender justice more broadly, even when such action exposes them to threats, violence, and even death. The analysis is based on original qualitative fieldwork carried out in 2015, 2016, and 2018 in Bolı´var, and in 2018 and 2019 in Putumayo. I draw on interviews with leaders and everyday members of women’s organizations, as well as local state authorities and international humanitarian organization staff. I also engaged in participant observation and ‘hanging out’ with both the Liga and Alianza, including travelling with members, and attending marches, rallies, and internal meetings. Historical details were gleaned from both interviews and existing publications, including CNMH reports.

Women’s high-risk mobilization in Colombia The Colombian conflict has taken centre stage in studies of civilian agency (see Masullo 2015; Arjona 2016; Kaplan 2017; Justino 2019). For the most part, these studies have not, however, investigated the uniquely gendered dimensions of civilian agency.⁴ Research focusing on women’s participation in the Colombian conflict have, in large part, centred on women’s roles in bringing about peace (see Rojas 2009; Paarlberg-Kvam 2019) or the impact of conflict on women’s post-conflict empowerment (Justino et al. 2012). As in other conflicts, women experienced the Colombian conflict in different ways compared to their male counterparts (Meertens and Stoller 2001; Meertens ³ In this chapter I use ‘high-risk mobilization’ and ‘high-risk collective action’ synonymously. ⁴ An important exception is the work of Villareal and Ríos (2006) which focuses on women’s nonviolent resistance in Chocó, Nariño, and Cauca.

JULIA MARGARET ZULVER

101

2010; Centro Nacional de Memoria Histórica 2017). An intersectional lens further shows how race and ethnicity also factored into these experiences (see Marciales Montenegro 2015; Acosta et al. 2018; Zulver 2021a). We know that conflict opens different opportunities for women, in terms of engagement in the political sphere, civil society, or at the grassroots level (Tripp 2015; Berry 2018; Webster et al. 2019). Indeed, studies posit that certain women’s experiences with (gendered) violence in Colombia served to catalyse mobilization, not only to redress this violence, but also for gender justice more broadly (Domingo et al. 2015; Kreft 2019; Zulver 2022). Colombia is currently undergoing a reorganization of its violent landscape. Despite the demobilization of the FARC in 2016, other groups have entered the power vacuums left in the spaces once controlled by the guerrillas. Various criminal elements and armed groups vie for control of territory and illicit economies, leading to what Maher and Thomson call a ‘precarious peace’ (Maher and Thomson 2018). Mass displacement has once again surged (UNHCR 2019), and according to interviewees, women who ‘seek to the advance the rights of women’ are targeted with violence (Zulver 2021b). Feminist scholars talk about the gendered continuum of violence, and question whether the end of hostilities actually constitutes ‘peace’ for women (Cockburn 2004; Berry and Rana 2019). Similarly, I too query the utility of separating out war/peace in Colombia, particularly as it relates to women’s experiences of grassroots activism. With that said, I have argued elsewhere (Zulver 2021b) that the uptick in gendered violence represents more than just war as normal; it constitutes what Berry refers to as ‘patriarchal backlash’ (Berry 2017). That is, the violence that some women experience in contemporary Colombia goes beyond previous efforts to use women’s bodies for territorial domination. It constitutes retribution for supporting the peace process, for creating ties with state institutions, and for transgressing gender norms by becoming empowered. This chapter explains women’s high-risk collective action during and after the Colombian conflict.⁵ While the women in question did not necessarily engage in what Masullo (2015) describes as ‘staying put’ by creating peace territories, they defied what may have been expected of them, particularly given the risks of (gendered) violence, and formed organizations that sought to provide for women’s material (food, shelter) and non-material (identity, belonging, meaning) needs. The ‘high-risk’ nature of the mobilization comes both from the context in which they mobilize (i.e. conflict zones featuring high levels of violence), and from the transgression of gender norms that acting publicly in a machista society implies. In this sense, it is both the setting and collective action that bring about high-risk exposure, even though women’s collective action is also about protecting themselves. This further factors into an assessment of the domain of losses. Simply ⁵ Using the 2016 Peace Agreements as a temporal bound of the conflict.

102

WOMEN’S MOBILIZ ATION FOR GENDER JUSTICE

existing as a woman in a territory dominated by militarized masculinities that actively dictate women’s behaviour means that non-participation in an organization does not guarantee safety, and that everyday life is characterized by gendered expressions of violence (see Cockburn 2004). Thus, when women resist, to use the language of civil resistance studies, they become ‘agents of change’, resisting the gendered social order (including hegemonic gender roles) imposed by violent actors (see Svensson and Smart within volume; Braun and Stallone within volume). When it comes to understanding why women engage in high-risk collective action, a typical cost–benefit explanation is insufficient. As Loveman (1998) notes in her study of high-risk collective action in the Southern Cone, it is difficult to assess the ‘grammar of rational calculation’ when the cost is calculated as a probability of death or extreme physical harm. I suggest that we can better understand women’s motivations if we engage with prospect theory, which tells us that ‘if actors perceive themselves to be in a losing situation they are inclined toward risks; if they see a winning situation around them, they prefer caution’ (Kahneman and Tversky 1979; Kahneman 1992; Weyland 2004). Women operating in a ‘domain of losses’ (i.e. a conflict context where they are unable to control external violence) value the incremental material and non-material benefits of participation in a grassroots organization over inaction (which, importantly, does not actually guarantee safety). While this explanation might be useful in understanding high-risk collective action broadly (and not just as it applies to women), I argue that part of the calculation to participate in a women’s organization (as opposed to a non-gendered initiative) is a function of both the gendered nature of violence to which women are exposed, and the gendered material and non-material benefits women can accrue through participation. As a woman in Bolı´var told me: When we started to talk, we realised that each of us wasn’t alone. This was something that was crucial, fundamental. ‘I’ll take your hand, let’s go!’ It was a process of organizational strengthening. From this moment, we started to work—we didn’t stop. We haven’t stopped. (interview in Zulver 2022)⁶

Kreft (2019) analyses the relationship between ‘gender-differential outcomes and gender-specific violence in conflict’. She uses Tilly’s (1978) understanding of mobilization around threats to theorize that women mobilize in response to the ‘threat that [conflict-related sexual violence] poses to them as women . . . regardless of whether they personally are victims’ because it is understood ‘in distinctly gendered terms and as a violent manifestation of societal gender inequalities’ ⁶ Interview dated 3 July 2015.

JULIA MARGARET ZULVER

103

(Tilly 1978 p. 222). Women living in conflict zones are at risk simply because they are women. When they engage in collective action, they elevate their levels of risk, insofar as they are actively, and not only passively, challenging the imposed social order. At the same time, however, organizing collectively can also elevate their potential protection under a ‘safety in numbers’ logic. The women’s organizations discussed in the remainder of this chapter had the opportunity to join existing mixed sex organizations but chose instead to form their own groups. Not only did they mobilize against their specifically gendered experiences of conflict and resist the social order imposed by paramilitaries, but they also evolved to mobilize for women’s rights more broadly.

La Alianza de Mujeres Tejedoras de Vida in Putumayo In the south of Colombia, the Putumayo province stretches along the border with Ecuador and Peru. It has strategic value for armed groups; its rivers, geography, and climate make it a prime location to cultivate, process, and transport coca and cocaine (Tate 2015). Historically, Putumayo has been contested by armed groups, meaning that at various moments the territory has become a war zone, impacting the daily lives of its inhabitants. During much of the 1990s, the FARC had control of Putumayo; they regulated business, imparted justice, and intervened in day-to-day life. In 1999, however, the right-wing Autodefensas Unidas de Colombia (AUC) paramilitaries violently moved into the territory, terrorizing citizens they accused of being left-wing guerrilla sympathizers (Cancimance López 2014; Centro Nacional de Memoria Histórica 2012). As Tate (2017) notes, the situation was made more complicated by the complicity of Colombia’s armed forces with paramilitary activities. The years of paramilitary control had clear and damaging effects on women’s lives. As the CNMH documents, rape, torture, sexual violence, and murder were used symbolically by the paramilitaries to terrorize the population (Centro Nacional de Memoria Histórica 2012). This gendered violence was sometimes enacted to punish women accused of being FARC sympathizers, but also to punish and humiliate men who were unable to intervene when their female relatives were publicly raped and abused by paramilitaries (Centro Nacional de Memoria Histórica 2012 p. 168). Paramilitaries established different codes of dress and behaviour for ‘decent’ vs ‘indecent’ women; those deemed ‘indecent’ were brutally punished. Another tactic involved disappearing women when they engaged in social and community leadership, thus becoming ‘an obstacle for armed groups to act’ (Alianza de Mujeres Tejedoras de Vida del Putumayo 2009). Despite such horrendous violence and the risks of death or corporal punishment, certain women resisted. One of these women was Fátima Muriel, an educator who travelled widely throughout the province. Initially interested in

104

WOMEN’S MOBILIZ ATION FOR GENDER JUSTICE

orphaned children whose parents had been murdered, her attention shifted to the small organizations of women that existed in the towns around the province. Muriel and some of her colleagues decided to unite them together under one umbrella organization: the Alianza de Mujeres ‘Tejedoras de Vida’ del Putumayo. In the early days, the associations within the Alianza did not participate in large scale mobilization to confront the paramilitaries. Their actions were subtle and localized, aimed at creating social fabric amongst women through the provision of emotional and psychological support, as well as educational workshops about rights and the documentation of abuses. An interviewee from El Tigre who helped form a local women’s association related that much of what she and her compañeras did at the time was visit each other’s houses, both in the town and the countryside, to talk to women and ‘rid them of their sadness’ and to unite women who were otherwise ‘totally alone’.⁷ The interviewee was clear that doing so was a high-risk activity, and that the paramilitaries closely monitored their meetings. Over time, the associations began to engage in more public expressions of collective action. In 2004, they held a 500-woman strong march in Mocoa to protest the assassination of an indigenous governor who had vocally opposed the FARC’s control of her territory (Centro Nacional de Memoria Histórica 2012). Muriel also spoke to me about travelling to the military bases of different armed groups (including the FARC, the paramilitaries, or the armed forces) to demand that they release children they recruited, returning them home to their families. A report written by the Alianza outlines the strategies undertaken by organizations acting with a ‘collective voice’. These included: Denouncing displacements before the Personerı´a, denouncing before the Defensorı´a del Pueblo, directing claims to competent authorities . . ., staying at the margins of the conflict and not engaging with the ideas, actors, or economies of armed groups, participating in peace marches . . . and simply organizing. (Alianza de Mujeres Tejedoras de Vida del Putumayo 2009)⁸

When the Alianza held its first retreat in 2003, the women adopted three central themes: ‘women, human rights and armed conflict; women’s history and political participation; and women and social and economic empowerment’ (Tate 2017). We can see that from an initial desire to meet women’s practical needs (providing immediate support in a context of gendered violence), the Alianza transitioned to tackle more strategic gender interests, including projects to promote women’s political and economic empowerment. Moreover, they began to challenge the conditions that give rise to gendered experiences of conflict themselves, including by ⁷ Interview in El Tigre, Putumayo, 18 November 2018. ⁸ These actions appear to fit within Masullo’s (2020) definition of civilian non-cooperation.

JULIA MARGARET ZULVER

105

openly discussing the realities of women’s experiences in Putumayo at the national and international levels (see Tate 2013). Over time, however, Alianza’s mobilization and cohesion has left them vulnerable to ongoing violence in a territory where armed groups continue to seek territorial control. A Colombian academic who has worked in the province for many years told me: ‘Putumayo is an emblematic [province] of women’s mobilization, they rendered visible what happened [during the conflict] and they lobbied hard for peace. This mobilization strengthened the women, but it also exposed them’ (in Zulver 2021b). According to a government employee in Bajo Putumayo, Alianza is: [T]he strongest [organization] in the region and they are punished as an example to others. They have made connections with the institutions and they were the first civil society group to really speak up during the peace. They were the biggest and strongest [group] to take advantage of the political opportunity . . . so they became a target.⁹

La Liga de Mujeres Desplazadas in Bolı´var Colombia’s Caribbean coast has served as a theatre of war throughout the country’s armed conflict. From 1997 to 2005, the AUC paramilitaries engaged in a ‘route of terror’, fighting guerrillas and using massacres, public torture, displacement, sexual violence, and threats as part of a plan to achieve social control (Centro Nacional de Memoria Histórica 2011). The violence that women suffered during paramilitary rule of the Caribbean region is a reflection of the ‘violent repertoires and protocols of social regulation . . . promoted by the commanders of [the paramilitaries] . . . to achieve its goals in the Caribbean’ (Centro Nacional de Memoria Histórica 2011). El Pozón is a slum outside of the city of Cartagena. In the late 1990s and early 2000s, it was a receptor zone for those families who had been displaced from around the Caribbean coast (Lemaitre et al. 2014). Women, many of whom were recently widowed and had suffered sexual violence, arrived with their children and only what they could carry with them. The neighbourhood had little police or institutional presence, leading to ongoing security risks. Armed groups made life dangerous for women: threats, physical attacks, and sexual violence were common. Amid this danger, however, the Liga de Mujeres Desplazadas (the League of Displaced Women) was formed. Initially, a small group of women came together to share food in an olla común (communal kitchen) and later to raise money to support an ill neighbour. When Patricia Guerrero, a lawyer, arrived in 1998, she ⁹ Interview with government employee, Bajo Putumayo, 11 September 2019.

106

WOMEN’S MOBILIZ ATION FOR GENDER JUSTICE

met this informal group and began to talk to them about the ways in which their experiences of conflict and displacement were shaped by their gender. Over time the group began to grow. Eventually, the Liga’s goals expanded beyond creating social cohesion and a shared identity for women victims of the armed conflict. In 2003, Guerrero was able to secure funding for a project that would build houses for the women of the Liga, promising the security of home ownership that would take them out of the precarity of the slum. By 2006, the women were able to move into the Ciudad de las Mujeres, the City of Women, in the municipality of Turbaco, an area that was paramilitary controlled. Such actions were not without peril, however. While in El Pozón, women faced threats, sexual violence, stalking, and the recruitment of their children, and ‘with more social and political empowerment came more threats of sexual violence’ (Lemaitre et al. 2014). During the construction of the City of Women, women were threatened, unmarked cars would circle the neighbourhood, and eventually, one of the women’s husbands was killed while guarding the building site overnight. Turbaco has historically held geostrategic importance, given its proximity to the Port of Cartagena, which is an important site of drug trafficking. Armed groups actively aimed to repress any expressions of community resistance, particularly when this took place in a territory over which they wanted complete control. Insecurity increased in the 2000s, particularly after the official demobilization of paramilitaries through the Justice and Peace Law of 2005. Demobilized young men returned to Turbaco, where the Liga—including the physical City of Women— was one of the most visible organizations (Lemaitre et al. 2014). In an interview, Guerrero notes that the empowered women of the Liga threatened paramilitary aspirations for both territorial and social dominance:

A community that demands rights and makes itself visible is more difficult to intimidate and displace. So a threat becomes a tool for whichever armed actor . . . who start to lose control over their territories. I think that in the case of the League of Displaced Women, it was because we were carrying out this resettlement in an area controlled by paramilitarism. We are women who dare to put a stick in the wheel and obstruct organized crime. (interview in Nobel Women’s Initiative 2016)

Despite the violent contexts in which they lived—their domain of losses—the women of the Liga persisted in their mobilizing activities. Today, the City of Women remains a symbol of the Liga’s pursuit for gender justice, which both recognizes the gendered impacts of conflict on women and makes a statement about women’s rights to housing, security, and gender justice more broadly.

JULIA MARGARET ZULVER

107

Gendered civilian agency over time The Alianza provides an important example of how the resistance that women learned during moments of conflict can be retooled and refocused when violence returns (see McAdam et al. 2001). Violence has returned to Putumayo, with a noticeable uptick since late 2018, according to interviews with members of the organization and local state employees. Certain women have been threatened; others have displaced themselves due to ongoing threats, including toward their children. In a September 2019 interview, Muriel was clear: ‘Things are going back to how they were before’ (in Zulver 2021b). In the face of this violence, the Alianza has not ceased its activities in the region. Instead, the women continue to participate throughout the province, but with a lower profile. The Alianza’s self-protection strategies include not wearing identifying clothing, not travelling at night, and not publicly announcing their activities until the last minute (Zulver 2021b). Nonetheless, they continue to engage in economic empowerment initiatives for women, hold workshops with lawyers so that women can denounce the violence they suffered during the conflict, manage an anti-gender violence campaign, and paint truth murals that feature the names of women killed in the province. The Liga, on the other hand, does not seem to be drawing on its repertoires of action in the same way. Although during fieldwork in 2016 most of the women who held leadership roles within the organization maintained the rhetoric that they will continue to luchar (struggle), ethnographic observation in 2016 and again in 2018 did not necessarily confirm that they are doing so with the same spirit or organizational strength as in previous years. In 2016, one woman confided in me that she does not like going to events anymore, but that she felt obligated to attend. On a bus ride around the Montes de Marı´a, another leader told me that she sees some of the women in the Liga as ‘apathetic’ now; those who got their houses no longer want to participate.¹⁰ Throughout fieldwork, Guerrero expressed concern to me that the women are not adopting the necessary leadership positions to move the Liga into the future. I attended an executive meeting where there was clear tension between those present, specifically around issues regarding inertia and ongoing mobilization. Why do we see these differences between two women’s organizations born out of similar experiences with conflict, at a similar point in time, with a similar leadership structure? The remainder of this chapter will engage with two potential explanations for why the Alianza endures while the Liga suffers mobilizational fatigue.

¹⁰ Field Notes, 26 October 2016.

108

WOMEN’S MOBILIZ ATION FOR GENDER JUSTICE

Fluctuating domain of losses One potential explanation for this variation could be related to a fluctuating domain of losses, whereby the durability of mobilization is dependent on women continuing to operate in what they see as a losing situation. In Putumayo, for example, women of the Alianza have been stalked, threatened, and disappeared recently (i.e. since 2018). Although it is clear that social leaders around the country are targeted because of their activities (Prem et al. 2018), this violence takes on a gendered dimension when it consists of a patriarchal backlash against women for their empowerment (Berry 2017; Zulver 2021b). Moreover, nonparticipation does not guarantee safety. By October 2019, 13 women—not part of the Alianza—were victims of feminicide in the province (see Conexión Putumayo 2019). In the City of Women, on the other hand, it is not clear that violence is as intense as in previous years. During fieldwork, women residents gave the impression that in recent years, violence has reduced. One woman told me that she feels safe living in the City but feels insecurity because of a lack of employment and health care. Another told me that she likes living in the neighbourhood because it is so tranquilo (calm).¹¹ To draw on a different, illustrative example, during a separate fieldwork trip to a town in Catatumbo that experiences intense violence at the hands of clashing armed groups,¹² I met with women who have come together to form an association of recyclers. During an interview in April 2019, they were open about the ways in which the experience of being part of the association has been empowering. They spoke about how they have learned to resist and confront machismo and unfair gender dynamics in their families and in their community. What struck me as interesting, however, was the line that the women drew between acceptable and unacceptable forms of mobilization, despite being clear that any mobilization incurred exposure to risk. One of the members suggested they hold a march against the killings taking place in the neighbourhood at the hands of warring non-state armed groups. The leader of the organization was categorical in her reply: this would put them at too high a risk.

¹¹ There is an important distinction between the chapter of the Liga in Turbaco, and the chapters in San Jacinto and Carmen de Bolívar (Montes de María). During fieldwork in 2018, I was told that members of the Liga in Carmen de Bolívar have been stalked and threatened and were granted a security scheme by the National Protection Unit. Moreover, there is evidence that some threats (not Liga-specific) are linked to mobilization around women’s rights (Arredondo 2019; Redacción Nacional 2020). ¹² This part of the country, along the Venezuelan border, also features high coca production and both historical and ongoing legacies of violence. This research was part of the CONPEACE project, funded by Global Affairs Canada.

JULIA MARGARET ZULVER

109

This nuanced line became even finer: the women explained to me that they had reached a point where they cannot continue to develop public discourse that promotes women’s rights; pushing the conversation any further would ‘threaten the social order of the neighbourhood’ and put them at a higher risk of violence. I left this meeting without a clear understanding about the exact location of the justifiable/unjustifiable binary. When it comes to participating in high-risk collective action, however, the boundary clearly does exist. The example in Catatumbo shows the upper limit of women’s mobilization; this association was not willing to engage in certain activities because they were too dangerous. Conversely, perhaps one of the reasons that the women of the Liga are not mobilizing with the same intensity as in previous years is because they no longer consider that their environment constitutes a domain of losses. To be clear, I do not want to suggest that the City of Women is a safe place. With that said, if the Liga’s security situation in the City is such that non-participation in the organization is in fact a guarantee of relative safety, then exposure through public activism may no longer be a justifiable course of action. Whilst this chapter does not provide a comprehensive answer to where the limits of mobilization lie, it is worth noting that women’s high-risk collective action is not infinite. When the risk of violence becomes too high, women begin to strategically plan the activities that they will engage in; when the risk of generalized violence is low (but the risk of violence for participation in a public-facing organization continues), high-risk collective action is not a desirable course of action.

Confidence in ongoing success An alternative explanation for the durability, or lack thereof, of women’s high-risk mobilization could be linked to the confidence in ongoing successes. As mentioned above, part of the decision about why it is ‘worth it’ to participate in high-risk collective action is premised on the perceived ability to obtain material and non-material benefits.¹³ Perhaps when the material benefits run out, and continue to remain elusive over time, the non-material benefits of group membership (identity, meaning, belonging) alone no longer justify ongoing participation. On one hand, the Alianza has had recent experiences with success, including its ability to obtain financial support for projects, like the internationally funded Peace Mediators diploma. A walk through their office in Mocoa shows the multiple donors who have funded their work, including the Swiss-Norwegian Cooperation ¹³ See also Wood’s (2003 p. 235) concept of ‘pleasure in agency’: ‘that positive effect associated with self-determination, autonomy, self-esteem, efficacy, and pride that come from the successful assertion of intention’.

110

WOMEN’S MOBILIZ ATION FOR GENDER JUSTICE

Fund, Mercy Corps, United Nations High Commissioner for Refugees (UNHCR), and United Sates Agency for International Development (USAID). They have longstanding relationships with international figures who have helped them secure funding for initiatives such as the truth murals and economic empowerment projects. Their Facebook page highlights the many activities they engage in, including marches, performances, a candle-lit vigil for murdered compañeras, and meetings with members of local government. On the other hand, since the construction of the City of Women in 2006, many women in the Liga do not feel that they have been successful in obtaining material benefits, despite ongoing mobilization. During fieldwork in 2015 and 2016, many women spoke about the resentment they feel. Although they have housing, they do not have income and therefore cannot pay for gas or electricity. Their biggest concerns were not about safety, but rather lack of access to food, health care, and education for their children. Another major concern is linked to reparations (both individual and collective), to which the women are entitled through the 2011 Victims’ Law. Although the collective reparations process has advanced since I first met the Liga, when I last brought it up in 2018, the women continued to bemoan the stagnation of the process. While this situation is not unique to the Liga, it is an ongoing source of frustration. We can see that the Alianza has been able to secure ongoing resources with which to finance its activities. The Liga, however, has not had the same experience, leading members to feel frustrated with the lack of material benefits to membership, and perhaps therefore demotivated to participate in ongoing mobilization activities.¹⁴ In terms of understanding the durability of mobilization, then, the lack of pay-off (a confidence in their growing strength)—particularly in a context where the threat of violence is not as imminent as in the past—explains why the Liga has lost mobilizational momentum over time.

Conclusion This chapter has analysed two cases of women’s grassroots organizations in Colombia to engage with the gendered elements of high-risk collective action in the face of the violent imposition of a gendered social order, and to assess the limits to high-risk mobilization over time. Following Sandvik (2018), Kreft (2019), and Zulver’s (2022) work on how women sometimes mobilize because of the (gendered) violence to which they are exposed, this chapter has used the ¹⁴ After finalizing this manuscript in 2022, I was informed that the Liga is moving forward with their collective reparations programme and is building a restaurant-gallery in Turbaco. Future research will shed light on how this impacts their ongoing mobilizational morale.

JULIA MARGARET ZULVER

111

experiences of the Liga and Alianza to show why and how women participated in high-risk collective action during some of the darkest days of Colombia’s armed conflict. It also explored issues of continuity and durability of mobilization over time. In the introduction to this volume, the editors reflect on how civilian protective agency impacts individuals differently, with gender as one important variable that requires more systematic investigation. This chapter has expanded beyond essentializing narratives of women in conflict and post-conflict settings as mothers, or as possessing an inherent peacefulness that predisposes them to post-conflict reconstruction (see also El-Bushra 2007; Ni Aolain et al. 2011). It has shown that women are capable of mobilizing, during and after conflict, for women’s rights and gender justice. In doing so, it has added to a burgeoning conversation about the ways in which civilian protective agency is gendered. These findings have further implications for the Women, Peace and Security agenda. Despite the various United Nations Security Council resolutions that acknowledge and promote women’s unique role in post-conflict contexts, twenty years later women engaging in activism and peacebuilding continue to face grave risks. Women participating in fragile spaces shaped by logics of militarized masculinities require guarantees of access to basic services, as well as safety and protection mechanisms (GNWP 2020 p. 18). Empirical examples like those presented in this chapter provide localized, grounded, and gendered evidence of how, when, and why women choose to mobilize for self-protection. They can thus help inform the actions governments and the international community must take in order to support women’s high-risk collective action.

References Acosta, M., Castañeda, A., Garcı´a, D., Hernández, F., Muelas, D., and Santamaria, A. (2018). ‘The Colombian Transitional Process: Comparative Perspectives on Violence against Indigenous Women’, International Journal of Transitional Justice, 12: 108–25. Alianza de Mujeres Tejedoras de Vida del Putumayo. (2009). Caracterización de la Situación de las Mujeres del Putumayo (2007–2008). Mocoa. Arjona, A. (2016). Rebelocracy: Social Order in the Colombian Civil War. New York: Cambridge University Press. Arredondo, J. (2019). ‘The Slow Death of Colombia’s Peace Movement’. The Atlantic. Baines, E. (2015). ‘“Today, I want to Speak Out the Truth”: Victim Agency, Responsibility, and Transitional Justice’, International Political Sociology, 9/4: 1–17. Berry, M. E. (2017). ‘Barriers to Women’s Progress after Atrocity: Evidence from Rwanda and Bosnia-Herzegovina’, Gender and Society, 31/6: 830–53. Berry, M. E. (2018). War, Women, and Power: From Violence to Mobilization in Rwanda and Bosnia-Herzegovina. Cambridge: Cambridge University Press.

112

WOMEN’S MOBILIZ ATION FOR GENDER JUSTICE

Berry, M. E., and Rana, T. R. (2019). ‘What Prevents Peace? Women and Peacebuilding in Bosnia and Nepal’, Peace and Change, 44/3: 321–49. Cancimance López, A. (2014). Echar Raı´ces en Medio del Conflicto Armado: Resistencias cotidianas de colonos en Putumayo. Universidad Nacional de Colombia. https:// repositorio.unal.edu.co/handle/unal/52046. Carpenter, C. (2005). ‘“Women, Children, and Other Vulnerable Groups”: Gender, Strategic Frames and the Protection of Civilians as a Transnational Issue’, International Studies Quarterly, 49/2: 295–334. Centro Nacional de Memoria Histórica. (2011). Mujeres y Guerra: Vı´ctimas y Resistentes en el Caribe Colombiano. Bogotá. Centro Nacional de Memoria Histórica. (2012). El Placer: Mujeres, Coca, y Guerra en el Bajo Putumayo. Bogotá. Centro Nacional de Memoria Histórica. (2017). La Guerra Inscrita en el Cuerpo: Informe Nacional de Violencia Sexual en el Conflicto Armado. Bogotá. Cockburn, C. (2004). ‘The Continuum of Violence: A Gender Perspective on War and Peace’, in Giles, W. and Hyndman, J. (eds), Sites of Violence: Gender and Conflict Zones, pp. 24–44. Berkeley: University of California Press. Conexión Putumayo. (2019). ‘Defensorı´a del Pueblo rechazó casos de feminicidios en el Putumayo’. Conexión Putumayo. Domingo, P., Rocha Menocal, A., and Hinestroza, V. (2015). Progress Despite Adversity: Women’s Empowerment and Conflict in Colombia. London. https://odi.org/ en/publications/progress-despite-adversity-womens-empowerment-and-conflictin-colombia/. El-Bushra, J. (2007). ‘Feminism, Gender, and Women’s Peace Activism’, Development and Change, 38/1: 131–47. DOI: 10.1111/j.1467-7660.2007.00406.x GNWP. (2020). Equality for Women, Peace for All: 2020 Annual Report. Global Network of Women Peacebuilders. https://gnwp.org/wp-content/uploads/2020GNWP-Annual-Report-web.pdf. Goetz, A. M., and Jenkins, R. (2016). ‘Agency and Accountability: Promoting Women’s Participation in Peacebuilding’, Feminist Economics, 22/1: 211–36. Justino, P. (2019). ‘Civilian Action in Conflict Settings: The Case of Colombia’, IDS Bulletin, 50/3: 95–113. Justino, P., Mitchell, R., and Mu¨ller, C. (2012). Quantifying the Impact of Women’s Participation in Post-Conflict Economic Recovery (No. 131). Brighton: Households in Conflict Network (HiCN). Kahneman, D. (1992). ‘Advances in Prospect Theory: Cumulative Representation of Uncertainty’, Journal of Risk and Uncertainty, 5/4: 297–323. Kahneman, D., and Tversky, A. (1979). ‘Prospect Theory: An Analysis of Decision under Risk’, Econometrica, 47/2: 263–92. Kampwirth, K. (2006). Women and the Guerrilla Movements: Nicaragua, El Salvador, Ciapas. Athens, OH: Centre for International Studies, Ohio University. Kaplan, O. (2017). Resisting War: How Communities Protect Themselves. Cambridge: Cambridge University Press. Krause, J. (2019). ‘Gender Dimensions of (Non)Violence in Communal Conflict: The Case of Jos, Nigeria’, Comparative Political Studies, 52/10: 1466–99. Kreft, A.-K. (2019). ‘Responding to Sexual Violence: Women’s Mobilization in War’, Journal of Peace Research, 56/2: 220–33.

JULIA MARGARET ZULVER

113

Lemaitre, J., Sandvik, K. B., López, E. S., Mosquera, J. P., Vargas Gómez, J., and Guerrero, P. (2014). Sueño de vida Digna: La Liga de las Mujeres Desplazadas. Bogotá: Universidad de los Andes. Loveman, M. (1998). ‘High-Risk Collective Action: Defending Human Rights in Chile, Uruguay, and Argentina’, American Journal of Sociology, 104/2: 477–525. DOI: 10.1086/210045 McAdam, D., Tarrow, S., and Tilly, C. (2001). Dynamics of Contention. Cambridge: Cambridge University Press. Maher, D., and Thomson, A. (2018). ‘A Precarious Peace? The Threat of Paramilitary Violence to the Peace Process in Colombia’, Third World Quarterly, 39/11: 2142–72. Marciales Montenegro, C. X. (2015). ‘Violencia sexual en el conflicto armado colombiano: racismo estructural y violencia basada en género’, VIA IURIS, 19: 69–90. Masullo, J. (2015). The Power of Staying Put: Nonviolent Resistance against Armed Groups in Colombia. Washington, DC: ICNC Press. Masullo, J. (2020). ‘Civilian Contention in Civil War: How Ideational Factors Shape Community Responses to Armed Groups’, Comparative Political Studies, 544/10: 1849–84. SAGE Publications Sage CA: Los Angeles, CA. Meertens, D. (2010). ‘Forced Displacement and Women’s Security in Colombia’, Disasters, 34/2: 147–64. Meertens, D., and Stoller, R. (2001). ‘Facing Destruction, Rebuilding Life: Gender and the Internally Displaced in Colombia’, Latin American Perspectives, 28/1: 132–48. Molyneux, M. (1985). ‘Mobilisation without Emancipation? Women’s Interests, the State, and Revolution in Nicaragua’, Feminist Studies, 11/2: 227–54. Moser, C., and Clark, F. (2001). ‘Introduction’, in Moser C. and Clark F. (eds), Victims, Perpetrators or Actors? Gender, Armed Conflict and Political Violence. Zed Books: London. Ni Aolain, F., Haynes, D. F., and Cahn, N. (2011). On the Frontlines: Gender, War, and the Post-Conflict Process. Oxford: Oxford University Press. DOI: 10.1093/acprof:oso/9780195396645.001.0001 Nobel Women’s Initiative. (2016). ‘Meet Patricia Guerrero, Colombia’. Nobel Women’s Initiative. Paarlberg-Kvam, K. (2019). ‘What’s to Come is More Complication: Feminist Visions of Peace in Colombia’, International Feminist Journal of Politics, 21/2: 194–223. Prem, M., Rivera, A., Romero, D., and Vargas, J. F. (2018). Killing Social Leaders for Territorial Control: The Unintended Consequences of Peace (No. 0019). LACEA Working Paper Series. Redacción Nacional. (2020). ‘Integrantes de la Red de Mujeres de Magdalena denuncian amenazas de muerte’. El Espectador. Magdalena. Rojas, C. (2009). ‘Women and Peacebuilding in Colombia: Resistance to War, Creativity for Peace’, in Bouvier, V. (ed.), Colombia: Building Peace in a Time of War. Washington, DC: United States Institute of Peace. https://lib.ugent.be/en/catalog/ rug01:001674186. Sandvik, K. B. (2018). ‘Gendering Violent Pluralism: Women’s Political Organising in Latin America’, Third World Thematics: A TWQ Journal, 1–16. Routledge. DOI: 10.1080/23802014.2018.1477527 Sjoberg, L. (2016). Women as Wartime Rapists: Beyond Sensation and Stereotyping. New York: New York University Press. Sjoberg, L., and Gentry, C. (2015). Beyond Mothers, Monsters, Whores: Women’s Violence in Global Politics. London: Zed Books.

114

WOMEN’S MOBILIZ ATION FOR GENDER JUSTICE

Tate, W. (2013). ‘Proxy Citizenship and Transnational Advocacy: Colombian Activists from Putumayo to Washington, DC’, American Ethnologist, 40/1: 55–70. Tate, W. (2015). Drugs, Thugs, and Diplomats: U.S. Policymaking in Colombia Stanford, CA: Stanford University Press. Tate, W. (2017). ‘Post-Accord Putumayo’, The Journal of Latin American and Caribbean Anthropology, 22/1: 164–73. Tilly, C. (1978). From Mobilization to Revolution. New York: Random House. Tripp, A. M. (2015). Women and Power in Postconflict Africa. Cambridge: Cambridge University Press. Trisko Darden, J., Henshaw, A., and Szekely, O. (2019). Insurgent Women: Female Combatants in Civil Wars. Washington, DC: Georgetown University Press. UNHCR. (2019). ‘Desplazamientos Masivos: Enero a julio 2019’. UNHCR. Unidad para las Vı´ctimas. (2020). ‘Registro Unico de Victimas’. Retrieved 30 August 2019, from Villareal, N., and Rı´os, M. A. (2006). Cartografia de la Esperanza: Iniciativas de Resistencia Pacı´fica desde las Mujeres. Bogotá: International Peace Information Service/Corporación Ecomujer. Viterna, J. (2013). Women in War: The Micro-processes of Mobilization in El Salvador. London: Oxford University Press. Webster, K., Chen, C., and Beardsley, K. (2019). ‘Conflict, Peace, and the Evolution of Women’s Empowerment’, International Organization, 73: 255–89. Weyland, K. (2004). The Politics of Market Reform in Fragile Democracies: Argentina, Brazil, Peru, and Venezuela. Princeton, NJ: Princeton University Press. Wood, E. J. (2003). Insurgent Collective Action and Civil War in El Salvador. New York: Cambridge University Press. Zulver, J. M. (2021a). ‘A sociación de Mujeres Afro por la Paz: Feminism with the Body and Face of a Woman’, Latin American Perspectives, 48: 0094582X2110207. DOI: 10.1177/0094582X211020742 Zulver, J. M. (2021b). ‘The Endurance of Women’s Mobilization during “Patriarchal Backlash”: A Case from Colombia’s Reconfiguring Armed Conflict’, International Feminist Journal of Politics, 23/3: 440–62. Routledge. DOI: 10.1080/14616742.2021.1901061 Zulver, J. M. (2022). High-Risk Feminism in Colombia: Women’s Mobilization in Violent Contexts. New Brunswick: Rutgers University Press.

6 Protection as a spectrum The different faces of protection in eastern Democratic Republic of the Congo Judith Verweijen

Introduction The Ruzizi Plain is a small strip of land in South Kivu province of eastern Democratic Republic of the Congo (DRC) that hosts a multitude of armed groups (Verweijen et al. 2020). These include so called ‘Mai-Mai groups’, a common designation for armed groups that employ discourses of community self defence against putative foreigners (Verweijen 2015a). Most Mai-Mai groups in the Ruzizi Plain demand a financial contribution from Fuliiru (an ethnic group) cattle owners which is calculated per head of cattle. This contribution was instituted in consultation with customary chiefs and other local authorities at a time when the Bafuliiru experienced tensions with the Barundi, another ethnic group. Given that, at that time, the Mai-Mai were seen to defend the Fuliiru community against Barundi militias, cattle owners were eager to pay this contribution. Not only did it provide them with security and help protect their cattle against potential theft, they also felt that by paying the Mai-Mai, they were supporting the cause of their community. However, when the conflict abated after 2015 and Barundi militias disappeared, this financial contribution was increasingly seen as extortion. Yet cattle owners continued to pay, as those who refused often saw their cattle looted by the very Mai-Mai that were supposed to protect them. As a local leader explained: ‘They [militias] have only personal interests today . . . In the past they used to protect us. But now they only bother people’¹ (see also Verweijen et al. 2020). The story of the Mai-Mai in the Ruzizi Plain illustrates the multifaceted and shifting nature of protection in eastern DRC, in particular, the protection that civilians seek from armed groups and forces to shield themselves against rampant insecurity. At times, this protection is solicited on a voluntary basis, as people deem certain armed actors well placed to ensure their physical and other forms of security. Yet in other cases, armed actors impose themselves on civilians, forcing them to pay protection fees that are experienced as extortion. In addition, some forms ¹ Interview with Fuliiru community leader, Sange, February 2017.

Judith Verweijen, Protection as a spectrum. In: Civilian Protective Agency in Violent Settings. Edited by: Jana Krause, et al., Oxford University Press. © Judith Verweijen (2023). DOI: 10.1093/oso/9780192866714.003.0007

116

THE DIFFERENT FACES OF PROTECTION IN E ASTERN DRCONGO

of protection assume the nature of a public good from which the population at large benefits. However, other forms have a more private character, implying the benefits are limited to specific networks, groups (such as ethnic groups), or individuals. These more private forms generally take place within the framework of patronage relations or assume the character of a commodity, implying the protection provider gets a direct monetary return. In sum, protection in eastern DRC assumes different forms that are informed by distinct logics. These different forms and related logics cannot always be neatly delineated, as they sometimes overlap. Moreover, as the example of cattle owners in the Ruzizi Plain shows, they may shift over time. The literature on civilian protective agency that zooms in on civilians’ relations with armed actors has tended to pay only limited attention to the more private dimensions of protection, including how protection can constitute a commodity or be part of patronage relations. An important reason for this is the strong focus on community-level structures, the collective action of communities, and protection outcomes for communities as a whole (e.g. Kaplan 2013; Mouly et al. 2015; Krause 2018; Rubin 2020; Masullo 2021). This focus is echoed in the literature on rebel governance, parts of which look at how civilians’ engagement with armed groups shapes these groups’ governance, which is generally conceptualized in terms of the provision of public goods (e.g. Mampilly 2012; Arjona 2014; Kasfir 2015). Most of these studies therefore overlook forms of private protection, which overlap and intersect with more publicly oriented forms of (security) governance. Another reason why attention to private protection has been limited is the relatively narrow focus on protection against physical harm that marked the early literature on civilian protective agency (see volume introduction). As this chapter shows, within private security arrangements, the provision of physical protection goes hand in hand with protection against, for instance, being busted by state actors when engaging in illegal activities or being harmed by someone holding grudges against you. To understand why civilians solicit protection from armed actors, it is important to take this entire range of protective practices into consideration. Scholarship on patronage and organized crime offers useful insights for grasping how private forms of protection work. Through my engagement with these two bodies of literature, I make three inter-related points. First, I argue that the complex entanglement of protection as a public good, protection as a commodity, and protection as patronage render it pertinent to conceptualize protection as located on a spectrum between these ideal types, ranging from more public to more private and from more voluntary to more coerced. Second, distinguishing different forms and logics of protection—even when they blur and overlap—allows for a deeper comprehension of civilian agency. Contrary to more public forms of protection, patronage and commodified protection entail higher levels of coercion that circumscribe—but do not negate—civilians’ room for manoeuvre. Third, paying attention to the prevalence of private protection is useful for gauging the

JUDITH VERWEIJEN

117

longer-term effects of civilian protection on the social order. In the case of eastern DRC, these effects are largely negative and set vicious cycles in motion that lead to an ever-greater demand for private protection while eroding both civilian and military public authority. The rest of this chapter proceeds as follows. In the next section, I briefly outline the methods utilized. Subsequently, I describe the key features of protection in eastern DRC, drawing on the literatures on patronage and private protection. The following section provides an in-depth empirical examination of the three main forms of protection provided by armed actors in eastern DRC. I will then examine the longer-term effects of the institutionalization of private protection on eastern DRC’s social order. The last section reflects on what the case of eastern DRC can tell us about protection in other violent settings and points to future directions of research.

Methods This chapter draws on data collected during twenty-eight months of qualitative fieldwork in eastern DRC’s Kivu provinces between 2010 and 2019 for a range of different research projects on the relations between civilians and armed actors, including non-state armed groups and the Armed Forces of the DRC (FARDC). These research projects examined the following aspects: (1) civilians’ perceptions of armed actors, including when they were seen as legitimate; (2) civilians’ agency vis-à-vis armed actors, including civilian initiatives to protect themselves against harm and solicit other forms of protection and favours; (3) a wide range of arenas of civilian–military interaction (e.g. revenue generation, dispute settlement, security, local governance); (4) the factors shaping armed actors’ behaviour towards civilians, including violence. Research methods included both individual and group interviews, informal conversations and observations, for instance, at roadblocks, border posts, mining sites, markets, beaches, and harbours.

The spectrum of protection in eastern DRC Civilians in the DRC are exposed to multiple forms of insecurity, which overlap and intersect. They are at constant risk of ambushes, robberies, and armed attacks, not only by armed groups but also by soldiers of the national armed forces, who sometimes engage in banditry and looting. A member of a civil society organization in Fizi said: Often, there are coalitions of coupeurs de route (those committing ambushes). The FARDC provides the arms, and after that, the sponsors are going to look for

118

THE DIFFERENT FACES OF PROTECTION IN E ASTERN DRCONGO

guys to commit these acts. They let others do that, so they are not suspected . . . They play a game, even in town, there are armed robberies, but this is always in complicity with soldiers.²

Furthermore, security and intelligence services arbitrarily arrest people and torture and harass them. Indeed, the state apparatus is a profound source of insecurity in the DRC. It is riddled with competing power networks which seek to extract wealth from citizens. In the words of one man: ‘The state in the Congo, it’s tracasser pour mieux regner [rule-by-harassment]’.³ Tracasseries (forms of harassment) come in many different forms, including outright intimidation, exaggerated user fees, and seeking out offences or alleging non-respect for often ambiguous and badly known procedures and regulations (Trefon 2009; Eriksson Baaz and Olsson 2011; Verweijen 2019; Thill 2021). Generating income is another source of insecurity for civilians. To start with, opportunities to generate livelihoods are limited and the object of fierce competition. In addition, property rights are ill enforced, debts are often not repaid, and arrangements to share profits not respected. Economic competitors may also engage in foul play, such as asking state agencies to bother their opponents by imposing inflated ‘taxes’ or withdrawing their licence (Verweijen 2013, 2015b; Raeymaekers 2014). Such behaviour creates grudges and disputes, which are generally difficult to settle in eastern DRC. Whether related to land, access to natural resources, local governance, family affairs, or past violence, disputes are omnipresent and often continue to linger. An important reason for this is that civilian dispute regulation mechanisms are weak, as the involved authorities lack the capacity for enforcement and have at times limited legitimacy. Several factors explain this weakness, including the variable legitimacy of customary chiefs, the general erosion of the authority of civilian leaders to the benefit of armed actors, and the nature of the Congolese state apparatus (Vlassenroot and Raeymaekers 2004; Verweijen 2015c; Vlassenroot et al. 2016). Because of the difficulty of resolving conflicts, sometimes combined with a desire for revenge or to inflict harm on opponents, some people try to settle disputes and grudges in a violent manner. The result is the ever-present possibility of violent settlement, which constitutes an important source of insecurity that those embroiled in disputes try to protect themselves against (Eriksson Baaz and Verweijen 2014; Verweijen 2019). Reflecting the multitude of sources and forms of insecurity, civilians see the concept of security (usalama) in multi-faceted terms, as encompassing at the same time physical, economic, social, and political security (Verweijen 2015b). Protection needs are similarly multifaceted and include both protection against physical insecurity and a wide range of services that are commonly associated ² Interview with member of civil society organization, Baraka, November 2010. ³ Interview with human rights activist, Uvira, February 2017.

JUDITH VERWEIJEN

119

with private protection. Varese (2014 p. x) identifies these services as including ‘protection against extortion, protection against theft and police harassment, protection of property rights, protection in relation to credit obtained informally and the retrieval of loans, protection of thieves, and the settlement of a variety of disputes’. As can be gleaned from this list, private protection services importantly overlap with core governance functions exercised by public authorities. Indeed, the demand for private protection rises where these governance functions are not adequately provided by the state (Gambetta 1993). Armed actors in eastern DRC provide this entire range of protection services to civilians in different forms. At times they provide protection as a common (nonexcludable) good, implying everyone can enjoy the benefits (cf. Samuelson 1954). For instance, when the Congolese army patrols at night to prevent rebel attacks, everyone living in the area is protected. However, protection can also be provided as an excludable good, implying that only certain people or groups of people can enjoy the benefits (Samuelson 1954). This type of protection assumes two forms: First, it can be a commodity, implying its provision has the character of a transaction involving a specific, well-delineated service for a specified price. One example is minibus operators in the Ruzizi Plain, who pay a fixed monthly contribution to armed groups uniquely to not have their buses ambushed. Second, protection may be a dimension of a social relation that is predominantly patronage-based, but often also involves other social ties, such ethnic affiliation or economic relations. One example is gold traders in Misisi, an artisanal gold mining town in rural South Kivu. These traders cultivate patronage relations with army officers deployed in the area to protect themselves against rampant banditry but also to do business with them (Verweijen 2015b). Another example is Mai-Mai groups who are mobilized along ethnic lines and provide security to members of the same ethnic group, sometimes against militias issuing from ethnic groups they are in tension with. As further discussed below, these different forms of protection can be provided by the same armed actors, sometimes simultaneously, and they may also blur or morph over time, for instance, assuming an increasingly private and coercive character. One reason for this is that private protection is an inherently ‘ambiguous’ commodity (Gambetta 1993 p. 29). Those providing it may at the same time be the very source of the dangers against which they offer protection, in which case protection becomes more coercive. Protection as patronage similarly has many ambiguous features, oscillating between persuasion and coercion. Patron–client ties are reciprocal but highly asymmetric relations that are generally considered to have two overlapping dimensions. The first is a dyadic, personal relationship between a patron and a client, which involves the granting of protection in exchange for political and other forms of support, loyalty, and information (Erdmann and Engel 2006). In eastern DRC, a key component of such protection is what is generally called trafic d’influence or influence peddling among higher-placed officials and other political

120

THE DIFFERENT FACES OF PROTECTION IN E ASTERN DRCONGO

and economic elites, for instance, to obtain jobs, permits, tax exemptions, and preferential administrative treatment. Trafic d’influence can be an important form of physical protection as it may prevent clients from falling victim to state predation. However, connections with officials also constitute a formidable instrument of coercion that patrons can use to push their clients to render them particular services. The second dimension of patronage concerns a hierarchical relationship between, on the one hand, a patron and, on the other hand, a network of clients which involves the exchange of collective goods for (political) support (Erdmann and Engel 2006). This includes security and symbolic goods such as ‘representation’, implying the embodiment of collective forms of identification and furthering collective interests (Chabal and Daloz 1999). In the DRC, this collective dimension of patronage often involves ethnic ties, whereby particular armed patrons—including armed group commanders—are seen to represent and protect certain ethnic groups. Where ethnic identities are polarized, there tends to be a fair amount of social pressure on people to support the armed actors deemed to protect their identity group (Verweijen 2015a). This adds an extra dimension of coercion to patronage relations with armed actors, which bear an inherent risk of (ultimately) turning violent, even when initially undertaken on a voluntary basis.

The different faces of protection Even though they sometimes overlap, protection as a public good, protection as transaction, and protection as part of patronage are inscribed in different logics and entail different interactions between protector and protected. In the following subsections, these differences are illustrated with empirical examples.

Protection as a transaction Protection as a transaction is particularly salient in illegal economic sectors, where people pay either the armed forces or armed groups in order not to be caught by officials trying to enforce the law. Whilst at first sight, this form of protection may appear very different from shielding civilians against harm, its ultimate purpose can be similar: ensuring survival. Across eastern DRC, hundreds of thousands of people depend for their income on direct or indirect involvement in illegal economic sectors, such as illegal logging and charcoal production, poaching, illegal fishing, the cultivation and trade in cannabis, and the trade in prohibited locally brewed or imported wines and liquors (Verweijen 2013, 2015b). In a context marked by deep poverty, losing that income can be a matter of life and death.

JUDITH VERWEIJEN

121

One area where the protection of illegal activities happens at a very large scale is Virunga National Park, located in North Kivu province. Since this is a protected area, it is strictly forbidden to cultivate, hunt, and log trees. Yet a range of armed groups and FARDC units facilitate these ‘guerrilla livelihoods’ activities in exchange for protection fees (Verweijen and Marijnen 2018). For instance, near Kiwanja, farmers have to pay the Rwandan rebel group Democratic Forces for the Liberation of Rwanda (FDLR) to farm their fields, which are partly located in the park. They pay a designated amount per patch of cultivated land and have to provide one basin of beans every time they pass an FDLR position in harvest time, receiving a so called jeton (token) upon payment. Given that the park guards of the Congolese Institute for Nature Conservation (ICCN) do not patrol in the areas occupied by the FDLR, people can cultivate there without being apprehended. The practice of paying a jeton is purely transactional; it concerns a oneoff interaction, which does not necessarily involve wider relations. Protection emerges here, therefore, mostly as a commodity. While the practice is imposed and resented, people do get a return. Moreover, because of its routinized nature, this payment system generally generates limited anxiety among farmers. They know exactly what to pay, when and how, and that as long as they pay, they will not be bothered. One woman commented: ‘We don’t like the FDLR but we suffer more from the park guards, because they arrest us and destroy our fields’ [in areas outside the control of the FDLR].⁴ This quote illustrates that even when imposed, protection arrangements can become appreciated when they are seen to provide a useful service, which somewhat lessens the degree of coercion involved. Inversely, private protection arrangements that were once initiated voluntarily may become increasingly coercive. In an isolated mountain area in South Kivu, most parts of which are not accessible by road, ambulant traders walk from market to market on foot. On the most insecure trajectories of their journey, they are accompanied by FARDC soldiers whom they pay between 400 and 800 Francs congolais (Congolese Francs) ($0.5–0.75) depending on the route. While this arrangement had started on a more or less voluntary basis, having been negotiated with the army by representatives of the association of ambulant traders, over time, it has become compulsory. As one trader exclaimed: ‘Refusing to pay? . . . It is impossible! [haiwezekane!]’.⁵ Yet even where protection is imposed, civilians may still exercise agency by negotiating the terms of coercion. In 2011, another FARDC brigade arrived in the area and tried to raise the price for accompanying traders on the Kalingi–Kitumba trajectory from 800FC to 1,500FC. In the course of the ensuing negotiations, representatives of the association of ambulant traders rejected the proposed increase, eventually settling on a compromise of a rise of 200FC only (Verweijen 2015b). ⁴ Interview with group of female farmers, Kiwanja, January 2019. ⁵ Interview with ambulant trader, Mikalati, January 2011.

122

THE DIFFERENT FACES OF PROTECTION IN E ASTERN DRCONGO

Protection as patronage Within their zones of operation, FARDC and armed group officers often develop patronage relations with local leaders and elites, including customary chiefs, administrative authorities, and businesspeople. These relations do not only enhance local elites’ physical safety, but also reinforce their economic and political security, facilitating income-generation and governance activities. Moreover, armed actor protection can help local leaders gain the day within the many conflicts around local authority that persist in eastern DRC, such as customary succession conflicts, disputes over administrative boundaries, and quarrels between customary and administrative authorities (Verweijen 2016a; Hoffmann et al. 2020; Verweijen et al. 2020). In addition, armed patrons can engage in trafic d’influence (influence peddling) to shape administrative, security, and judicial decision-making processes, such as negotiations surrounding the amount of taxes to be paid, job hires, and arrests and prosecutions (Verweijen 2015b). As one man commented: ‘Everything is negotiable in the Congo. And if you do not have a parapluie (patronage/protection) you can end on the losing side’.⁶ How protection works within the framework of patronage is well-illustrated by the warlord Bede (also known as Obedi or Bedy) Rusagara, who led an armed group in the Ruzizi Plain and adjacent hills between 2011 and his death in 2015. Bede managed to build up a large network among politicians and businesspeople at both the national and the provincial level. His contacts included the governor of South Kivu province, who was allegedly instrumental in Bede’s release from prison in 2011. This extensive network did not only shield him and his collaborators against being apprehended or prosecuted, it also allowed him to intervene in all sorts of disputes and administrative processes (Verweijen 2016a). People appealed to Bede to obtain jobs in the local administration or influence the outcome of land conflicts by putting pressure on the department of land affairs in Uvira. As one man testified: ‘My plot of land was sold illegally, and the land title was annulled, even though it had been registered. Then Bede called me to say he had fixed it’.⁷ Bede also provided jobs in his own businesses, in particular to Bafuliiru from his village of origins, Mutarule, where he was constructing a huge, multi-storey house (a status symbol in eastern DRC). These actions earned him a high level of popularity in parts of the Ruzizi Plain, in particular around Mutarule. As one observer testified: The people of Mutarule regretted his death, when someone was looking for a job, he gave them access to positions. Among the young people here are some who had jobs as drivers in his minibuses. Bede had a say in NGOs, in public ⁶ Interview with businessman, Uvira, April 2014. ⁷ Interview with member of civil society organization, Bukavu, May 2015.

JUDITH VERWEIJEN

123

administration. . . . And as he circulated in the area, the rate of banditry began to decrease.⁸

Another said: ‘The population gave him power and said “you are a General”. When he controlled the area here, there was a bit of stability in the zone’.⁹ Such testimonies, however, came uniquely from Bafuliiru sources, as the Barundi and other groups living in the Plain felt threatened by Bede’s group (Verweijen 2016a). The case of Bede illustrates how the provision of physical security is but one element within the provision of a wider array of protection services. In addition, it shows how individual and collective patronage can overlap, as armed actors simultaneously protect ethnic groups as a whole and provide personal benefits to members of those groups. Civilians make use of these overlaps when leveraging patronage relations with armed actors, which is a crucial aspect of civilian protective agency in eastern DRC. Seizing on the reciprocity of patron–client ties, civilians try to shape armed actors’ provision of physical security not only to their own benefit, but also to that of wider groups, communities, or areas. A clear example is the head teacher at a school in a remote area of Fizi territory in South Kivu, who had developed patronage relations with a commander of the armed group Biloze Bishambuke. He used these relations not only to enhance his own safety but also to demobilize child soldiers fighting in the group’s ranks. Another example is the tax collector of Itombwe sector, who is a relative of Mai-Mai commander Ebuela Mtetezi. Allegedly, he used these personal relations to ensure the Mai-Mai would not hinder any humanitarian activities in their zone of influence, ensuring that the entire population there could access these services.

Protection as a public good It is not unusual for armed patrons-cum-private protectors to also provide forms of public security or protection as a non-excludable good. Despite being involved in business activities and private protection, and being at times at the root of insecurity, the FARDC also engage in military activities that bring a measure of security. They conduct patrols, push armed groups out of certain areas, are present along roads where bandits and armed groups lay ambushes and provide escorts to public authorities in insecure zones (Verweijen 2015b). In Lubero, a woman commented: ‘We want more FARDC deployment. Without the military we feel insecure here as the rebels are nearby’.¹⁰ Civilians, in particular local leaders, often try to shape the military’s public security practices, both by motivating them to undertake certain military activities and ⁸ Interview with local leader, Mutarule, May 2015. ⁹ Interview with member of civil society organization, Mutarule, February 2017. ¹⁰ Interview with farmer, Lubero, April 2010.

124

THE DIFFERENT FACES OF PROTECTION IN E ASTERN DRCONGO

by trying to curb abuses. For example, in Lubero, the FARDC went to patrol in a certain area after local leaders had drawn the military’s attention to the precarious situation there during a so called conseil de securité (local security meeting). In shaping the military’s public security practices, civilians often draw on and try to leverage patronage relations, using personal contacts within the military (Verweijen 2015b). For instance, certain businesspeople from the Fizi chapter of the Federation of Congolese Businesses (FEC) used their personal connections in the FARDC, specifically a brigade commander they did business with, to reduce the number of roadblocks in the area, which crippled trade and inflated prices. Even while motivated by self-interested reasons, this initiative was to the benefit of all road users, showing how public and private forms of protection can entwine. Whilst the armed forces sometimes provide protection as a public good then, this does tend to come at a price. Like armed groups, the FARDC lives off the population’s back. They expect households and village chiefs to provide in-kind contributions, including manioc roots, manioc or maize flour, palm oil, chickens, and goats. In addition, they demand effort de guerre (war effort) contributions, which are payments in cash or in-kind (e.g. fuel, batteries), from individual economic operators (e.g. shops, restaurants), associations of economic operators (e.g. the fishermen’s committee) and sometimes civil society organizations. Both the military and armed groups also ‘tax’ at roadblocks, mining sites, beaches, harbours, and markets. These forms of ‘taxation’ come on top of the protection payments on illegal activities discussed above (Verweijen 2013, 2015b; Hoffmann et al. 2016). These various contributions are generally not voluntary, although there may be room for negotiation. For instance, when passing roadblocks, civilians may try to lower the amount to be paid or demand an exemption by feigning importance, by emphasizing personal relations with influential commanders, or by claiming a social status that allows for exemptions, such as belonging to a religious or humanitarian organization (Schouten et al. 2021). In most cases, however, not paying is simply not an option. Moreover, civilians fear that non-payment will render them vulnerable to targeted predation, as armed actors remember who refused to cooperate with them and will take revenge in the future. Where public security providers simultaneously act as private protectors, those enjoying the benefits of their public security practices may still become victim of their predation. This blurs the boundaries between protection as a public good and private protection. Many motards (motor-taxi drivers) in eastern DRC hesitate to resist payment to the FARDC at roadblocks as they fear they will otherwise get into trouble. In particular, they fear getting ambushed by bandits who are in collaboration with the military or to fall victim to soldiers’ tracasseries, for instance, by being asked to show certain paperwork that they do not have. Shopkeepers and other economic operators would also rather pay effort de guerre, anticipating negative consequences for their business if they do not, even without explicit intimidation.

JUDITH VERWEIJEN

125

As one of them commented: ‘If you do not pay, they do not forget’.¹¹ A trader explained that he generally gave generously for the effort de guerre—more than some of his colleagues—as soldiers were among his most important clients and as this could ensure their help in the future, for instance, when attacked by bandits.¹² Hence, civilians are also paying financial contributions to armed actors to avoid future trouble and ensure future help, whether provided in the form of public or private protection. These examples show that paying financial contributions to armed actors can be motivated by both fear and opportunism, which further illustrates the meandering boundaries between coercion and persuasion within protection arrangements with armed actors (Verweijen 2018). The cases also highlight that civilian protective agency in eastern DRC has different time horizons: protection is not always about immediate benefits and shielding civilians against immediate danger; it is in many cases (also) about constructing relations that will help avert future misfortune, following the approximate logic of an insurance policy (Verweijen 2018). Given that this insurance-type of thinking often informs the construction of patronage relations, where reciprocity tends to be deferred rather than immediate, excluding more private forms of protection from the analysis of civilian protective agency risks providing only a partial insight into how such agency plays out at various time scales.

The long-term effects of the institutionalization of private protection As illustrated above, many civilians in eastern DRC resort to private protection arrangements with armed actors or are forced into them. The resulting prevalence of private forms of protection has several wider effects on eastern DRC’s social order. On the one hand, private protection has a useful function. It enlarges the options for civilians’ protection in the face of deficient and at times abusive state security services. In addition, the protection of illegal activities ensures the livelihoods of a considerable part of the population. Patronage/protection relations also provide civilians with a certain measure of leverage over armed actors, including to provide security as a public or private good or mitigate abuses. On the other hand, the institutionalization of private protection arrangements with armed actors engenders significant negative effects in the long term, the four most important of which are discussed in the following. The first negative effect, which is both a cause and a consequence of the omnipresence of private protection, is fostering structural distrust towards the ¹¹ Interview with shopkeeper, Lusuku, December 2011. ¹² Interview with trader, Luberizi, November 2011.

126

THE DIFFERENT FACES OF PROTECTION IN E ASTERN DRCONGO

state security services. The fact that military personnel are involved in protecting illegal economic activities and in providing private protection services undermines the military’s public security activities and enhances their engagement in abuses. FARDC officers may misuse their public position to create revenuegeneration opportunities for their patronage networks or can order or engage in violence at the behest of their clients, for instance, to physically harm or even kill the latter’s opponents. Furthermore, where client networks follow identity lines, patronage practices will lead officers’ impartiality in relation to inter-community conflicts to be cast into doubt (Verweijen 2015b; Verweijen and Brabant 2017). In addition, systematic engagement in economic activities and meddling in civilian affairs diverts attention, time, and resources from the armed forces’ public security mission. Inadequate public security provision and abusive behaviour by the military, in turn, reinforce civilians’ demand for private protection arrangements, either with FARDC officers or non-state armed actors (Verweijen 2016b). Secondly, the institutionalization of private protection leads to the militarization of personal and collective conflicts, which structurally undermines civilian authority. To start with, the presence of armed protection providers creates incentives for conflict parties and competitors to settle their disputes, grudges, and jealousies by force, or simply to harm their opponents (Kalyvas 2006). Such ‘settlement’ of conflicts, however, is fundamentally unstable, as those who are disadvantaged feel aggrieved and will seek future occasions to undo the settlement or take revenge (Eriksson Baaz and Verweijen 2014). Furthermore, the presence of armed groups, such as the Mai-Mai, who claim to protect particular ethnic groups, contributes to the militarization of inter-community disputes. It prompts other ethnic groups to support armed groups that protect their own community, lest they end up in a weaker position. The presence of armed groups linked to specific ethnic communities, in turn, causes incidents and tensions to turn violent earlier, as it aggravates distrust and facilitates revenge actions. As a consequence, peaceful conflict resolution becomes more difficult (Verweijen 2016b; Verweijen and Brabant 2017). The systematic involvement of armed actors in the settlement of personal and collective conflicts also undermines the authority of civilian leaders, in particular their capacity to address these conflicts. As a result, it becomes even more attractive for civilians to solicit armed actors to regulate the disputes they are involved in. A third effect of the institutionalization of private protection is the entrenchment of illegal economic activity. Due to the absence of official regulatory frameworks and ill-defined property rights in illegal markets, trust in enforcement tends to be low while the risks of involvement are high. Protection is therefore a crucial enabler of illegal activity. Furthermore, the greater the offer of protection, the greater the incentives to engage in such activity, since it becomes at once less risky and easier (Gambetta 1993). In this way, the omnipresence of state and

JUDITH VERWEIJEN

127

non-state armed actors in eastern DRC fosters the proliferation of illegal incomegeneration activities. Moreover, as people become dependent on illegal sectors for their income, the demand for protection from armed actors increases, which allows these actors to reinforce their wealth and power position. One form of illegal activity that has soared over the past years is violent banditry. Ambushes, kidnappings, armed robberies, and contract killings are omnipresent in both urban and rural areas. Consequently, violent banditry has become an even bigger source of insecurity for civilians than clashes between armed groups and forces, fuelling a further demand for protection from armed actors (Verweijen 2019). Fourth, private protection arrangements with armed actors become selfenforcing due to three interlocking mechanisms: the institutionalization of distrust, comparative disadvantage, and normalization. When being protected by armed actors, people have incentives to cut dishonest deals, as they know that those who are disadvantaged will be wary of seeking redress. Consequently, people enjoying armed actor protection may sooner sell goods of lesser quality than promised, tip off bandits to rob suppliers or buyers, or not repay debts. A high number of dishonest deals in a given place will cause the (potentially) disadvantaged to similarly solicit protection from armed actors, which creates incentives to also start engaging in dishonest practices. As a result, distrust becomes selfperpetuating, or in Gambetta’s (1993 p. 27) words ‘endogenous’, leading incentive structures to become skewed against honesty. This dynamic also contributes to turning private protection schemes into a collective action problem: those not participating are at a comparative disadvantage and risk losing out if they do not act likewise (Verweijen 2016b). Both the entrenchment of distrust and the risk of comparative disadvantage lead to a third mechanism: the normalization of militarized protection schemes. When they are engaged in at a large scale and over a long period of time, people start considering private protection arrangements with armed actors as ‘normal’ and ‘taken-for-granted’, even though they are at times ethically dubious (Verweijen 2013, 2016b). Given the multi-faceted insecurity to which people in eastern DRC are exposed on a day-to-day basis, and the deficient workings of the state security services, it is not surprising that soliciting forms of private protection comes across as a logical and appealing solution. This is especially the case where people already have close contacts with armed group or FARDC officers, who are often co-villagers, co-ethnics, business partners, or family members. In sum, while for individual civilians soliciting private protection from armed actors is a logical way to protect themselves against insecurity, in the long term, this practice ends up reproducing various drivers of insecurity while also entrenching the demand for private protection. These negative effects are not the result of conscious decisions, but rather stem from the fact that practices that make sense to individuals have unfavourable outcomes at the collective level (Verweijen 2013, 2018).

128

THE DIFFERENT FACES OF PROTECTION IN E ASTERN DRCONGO

Conclusions: Looking beyond the DRC This chapter has demonstrated that protection in eastern DRC is located on a spectrum between protection as a public good, protection as a commodity, and protection as individual or collective patronage. Each of these forms have different underlying logics and come with distinct modes and degrees of coercion. To adequately grasp civilian protective agency in war settings, including how private connections are mobilized for collective security outcomes, it is crucial to consider the entire range of protection options that are available to civilians and where these are located on the spectrum. Yet the literature on civilian protective agency has so far paid limited attention to the ways in which protection as a commodity and as patronage may blur and overlap with other forms of protection. Overlooking private forms of protection, however, may lead to an inadequate understanding of the different time horizons involved in civilian protective agency, in particular the importance of forward-looking practices, such as constructing patronage relations. It may also obscure the longer-term effects of civilian protection strategies on the social orders in which they unfold, which in turn shape what protection options are available to civilians and the incentives for choosing them. It cannot be assumed, however, that private protection and its blurring with more public forms is as salient in other contexts as in eastern DRC. First, in other contexts, the nature of the state apparatus may differ. For instance, the military may not be implicated in (the protection of ) illegal revenue generation to the same extent. In addition, the workings of the state apparatus may be less profoundly shaped by patronage networks. Second, in other contexts, armed groups and the armed forces may not have the same level of social proximity to the population, which facilitates forging private protection arrangements. Among the over 120 armed groups active in eastern DRC today, many operate in a relatively small area concentrated around the main commander’s zone of origins, with most recruits being (originally) from the same area. As a result, armed group members tend to be relatives, co-villagers, former classmates, or ex-neighbours from the population. Military personnel too may develop high degrees of familiarity with civilians owing to deficient rotation policies and the fact that soldiers often live intermingled with the population (Verweijen 2015a, 2016b). Third, eastern DRC stands out for the protracted nature of violent conflict, which has been ongoing since the mid-1990s. The longevity of violence has contributed to the deep social embedding of armed actors and has provided them with a pronounced role in economics, politics, and local governance. This has normalized protection arrangements with armed actors, while also enabling these actors to deliver a wide range of protection services (Verweijen 2013, 2015b, 2016b). Whilst the observations in this chapter may not be fully applicable elsewhere, they do raise several questions around civilian protective agency that merit further research in other contexts. How do patron–client ties between armed actors

JUDITH VERWEIJEN

129

and civilians promote or hamper civilian protective agency? How do protection as commodity, protection as a public good, and protection as individual or collective patronage mutually shape one another and when and why do civilians solicit one form rather than another? And finally, when are protection arrangements with armed actors entered into voluntarily by civilians and when are they imposed, and how does the balance of coercion and persuasion shift over time? Gaining an insight into these questions is crucial for developing a better understanding of how civilians in violent settings interact with armed actors to protect themselves against variegated forms of insecurity.

References Arjona, A. (2014). ‘Wartime Institutions: A Research Agenda’, Journal of Conflict Resolution, 58/8: 1360–89. SAGE Publications Inc. DOI: 10.1177/0022002714547904 Chabal, P., and Daloz, J.-P. (1999). Africa Works: Disorder as Political Instrument. London: Bloomington: International African Institute in association with James Currey, Oxford; Indiana University Press. Erdmann, G., and Engel, U. (2006). ‘Neopatrimonialism Revisited—Beyond a CatchAll Concept’, SSRN Electronic Journal. DOI: 10.2139/ssrn.909183 Eriksson Baaz, M., and Olsson, O. (2011). ‘Feeding the Horse: Unofficial Economic Activities within the Police Force in the Democratic Republic of the Congo’, African Security, 4/4: 223–41. Routledge. DOI: 10.1080/19392206.2011.628629 Eriksson Baaz, M., and Verweijen, J. (2014). ‘Arbiters with Guns: The Ambiguity of Military Involvement in Civilian Disputes in the DR Congo’, Third World Quarterly, 35/5: 803–20. Routledge. DOI: 10.1080/01436597.2014.921431 Gambetta, D. (1993). The Sicilian Mafia: The Business of Private Protection. Cambridge, MA: Harvard University Press. Hoffmann, K., Vlassenroot, K., and Marchais, G. (2016). ‘Taxation, Stateness and Armed Groups: Public Authority and Resource Extraction in Eastern Congo’, Development and Change, 47/6: 1434–56. International Institute of Social Studies. Hoffmann, K., Vlassenroot, K., and Mudinga, E. (2020). ‘Courses au Pouvoir: The Struggle over Customary Capital in the Eastern Democratic Republic of Congo’, Journal of Eastern African Studies, 14/1: 125–44. Routledge. DOI: 10.1080/17531055.2019.1711321 Kalyvas, S. N. (2006). The Logic of Violence in Civil War. New York: Cambridge University Press. Kaplan, O. (2013). ‘Nudging Armed Groups: How Civilians Transmit Norms of Protection’, Stability: International Journal of Security and Development, 2/3. Kasfir, N. (2015). ‘Rebel Governance—Constructing a Field of Inquiry: Definitions, Scope, Patterns, Order, Causes’, in Arjona A., Kasfir N., and Mampilly Z. (eds), Rebel Governance in Civil War, pp. 21–46. Cambridge Cambridge University Press. DOI: 10.1017/CBO9781316182468.00:2 Krause, J. (2018). Resilient Communities: Non-violence and Civilian Agency in Communal War. Cambridge: Cambridge University Press. Mampilly, Z. (2012). Rebel Rulers: Insurgent Governance and Civilian Life during War. Ithaca, NY: Cornell University Press.

130

THE DIFFERENT FACES OF PROTECTION IN E ASTERN DRCONGO

Masullo, J. (2021). ‘Civilian Contention in Civil War. How Ideational Factors Shape Community Responses to Armed Groups’, Comparative Political Studies, 54/10: 1849–84. Mouly, C., Idler, A., and Garrido, B. (2015). ‘Zones of Peace in Colombia’s Borderland’, International Journal of Peace Studies, 20/1: 51–63. Raeymaekers, T. (2014). Violent Capitalism and Hybrid Identity in the Eastern Congo: Power to the Margins. Cambridge: Cambridge University Press. DOI: 10.1017/CBO9781139976862 Rubin, M. A. (2020). ‘Rebel Territorial Control and Civilian Collective Action in Civil War: Evidence from the Communist Insurgency in the Philippines’, Journal of Conflict Resolution, 64/2–3: 459–89. SAGE Publications Inc. DOI: 10.1177/0022002719863844 Samuelson, P. A. (1954). ‘The Pure Theory of Public Expenditure’, The Review of Economics and Statistics, 36/4: 387–9. The MIT Press. DOI: 10.2307/1925895 Schouten, P., Verweijen, J., Murairi, J., and Batundi, S. (2021). ‘Paths of Authority, Roads of Resistance: Ambiguous Rural Infrastructure and Slippery Stabilization in Eastern DR Congo’, Geoforum. DOI: 10.1016/j.geoforum.2021.09.017 Thill, M. (2021). Congo Cop: Performing the State in Central Africa (Doctoral Dissertation). Ghent University. Retrieved 20 February 2022, from Trefon, T. (2009). ‘Public Service Provision in a Failed State: Looking Beyond Predation in the Democratic Republic of Congo’, Review of African Political Economy, 36/119: 9–21. Routledge. DOI: 10.1080/03056240902863587 Varese, F. (2014). ‘Protection and Extortion’. The Oxford Handbook of Organized Crime. Retrieved 20 February 2022, from . DOI: 10.1093/oxfordhb/9780199730445.013.020 Verweijen, J. (2013). ‘Military Business and the Business of the Military in the Kivus’, Review of African Political Economy, 40/135: 67–82. Routledge. DOI: 10.1080/03056244.2012.761602 Verweijen, J. (2015a). ‘From Autochthony to Violence? Discursive and Coercive Social Practices of the Mai-Mai in Fizi, Eastern DR Congo’, African Studies Review, 58: 157–80. DOI: 10.1017/asr.2015.42 Verweijen, J. (2015b). The Ambiguity of Militarization: The Complex Interaction Between the Congolese Armed Forces and Civilians in the Kivu Provinces, Eastern DR Congo (Doctoral Dissertation). Utrecht University. Verweijen, J. (2015c). ‘The Disconcerting Popularity of Popular In/justice in the Fizi/ Uvira Region, Eastern Democratic Republic of the Congo’, International Journal of Minority and Group Rights, 22/3: 335–59. Brill Academic Publishers. DOI: 10.1163/15718115-02203003 Verweijen, J. (2016a). A Microcosm of Militarization: Conflict, Governance, and Armed Mobilization in Uvira, South Kivu. London: Rift Valley Institute. Verweijen, J. (2016b). Stable Instability: Political Settlements and Armed Groups in the Congo. London: Rift Valley Institute. Verweijen, J. (2018). ‘Civilian Resistance Against the Military in Eastern DR Congo: A Combined Social Navigation and Structuration Approach’, Qualitative Sociology, 41/2: 281–301. Springer. Verweijen, J. (2019). Violent Cities, Violent Society. Analysing Urban Violence in the Eastern Congo. London: Rift Valley Institute.

JUDITH VERWEIJEN

131

Verweijen, J., and Brabant, J. (2017). ‘Cows and Guns. Cattle-Related Conflict and Armed Violence in Fizi and Itombwe, Eastern DR Congo’, The Journal of Modern African Studies, 55: 1–27. DOI: 10.1017/S0022278X16000823 Verweijen, J., and Marijnen, E. (2018). ‘The Counterinsurgency/Conservation Nexus: Guerrilla Livelihoods and the Dynamics of Conflict and Violence in the Virunga National Park, Democratic Republic of the Congo’, Journal of Peasant Studies, 45/2: 300–20. Taylor & Francis. DOI: 10.1080/03066150.2016.1203307 Verweijen, J., Twaibu, J., Dunia Abedi, O., and Ndisanze Ntababarwa, A. (2020). ‘The Ruzizi Plain. A Crossroads of Conflict and Violence’. Monograph, Ghent: Governance in Conflict Network. Retrieved 20 February 2022, from Vlassenroot, K., Mudinga, E., and Hoffmann, K. (2016). Contesting Authority: Armed Rebellion and Military Fragmentation in Walikale and Kalehe, North and South Kivu. London: Rift Valley Institute. Vlassenroot, K., and Raeymaekers, T. (2004). Conflict and Social Transformation in Eastern DR Congo. Ghent: Academia Press.

PART T WO

V IOLEN T F OR MS OF CI V I LI A N PR OTECTI V E AGENC Y

7 Civilian violent mobilization and the intensity of civil war in Mozambique Corinna Jentzsch

Introduction ‘Civilian agency’ is most often understood as nonviolent interventions in conflict settings by unarmed actors less powerful than their armed (and violent) counterparts.¹ These unarmed ‘civilians’ are distinguished from members of armed groups and ‘combatants’; civilians’ vulnerable, unarmed status makes them worthy of protection. However, intra-state wars without clear frontlines make it difficult to identify and protect civilians. The fact that rebels tend to hide among civilians has raised concern that those ‘caught between the crossfire’ have no easy escape route from armed conflict (Jose and Medie 2015). Yet, even though they may not be formal members of armed groups, civilians often play important supporting roles in providing information, food, and other resources to armed groups, making civilians crucial enablers of armed group activities and also of the infliction of violence on other civilians (Wood 2003; Kalyvas 2006). As a consequence, the deliberate targeting of civilians has become a strategy of war, to deter the provision of crucial support to the enemy (Kalyvas 2006; Ferris 2011 pp. 246–7), which again raises the issue of civilian protection (Jose and Medie 2016). Whilst civilians often find themselves at the heart of armed conflict, they do not receive sufficient protection from armed groups, including state armed forces. Hence, they develop ‘civilian protective agency’ (see volume introduction). They engage in ‘civilian self-protection’ to defend themselves against ‘immediate, direct threats to their physical integrity’ imposed by armed groups (Jose and Medie 2015, 2016). The activities that people affected by war develop to protect themselves can range from nonviolent to violent and unarmed to armed.² Much of the focus of ¹ There are some exceptions to this where authors have conceptualized civilian agency as violent action, see Schubiger (2021), Masullo (2021a), Jentzsch (2022), and the chapters by Masullo and Ben Hamo and Moncada within volume. ² There is a difference between violent/nonviolent and armed/unarmed self-protection, as civilians can make use of violence without making use of weapons. The main difference I make in this article is between unarmed and armed, as I assume that a group that arms itself underlies different organizational constraints and possibilities than a group that does not have access to weapons.

Corinna Jentzsch, Civilian violent mobilization and the intensity of civil war in Mozambique. In: Civilian Protective Agency in Violent Settings. Edited by: Jana Krause, et al., Oxford University Press. © Corinna Jentzsch (2023). DOI: 10.1093/oso/9780192866714.003.0008

136

CIVILIAN VIOLENT MOBILIZ ATION

the research agenda has been on nonviolent and unarmed means of protection. Much of the scholarly and practitioner interest in the power of nonviolent and unarmed civilian protection and resistance aims to explain the success of nonviolent action in bringing about social and political change (Chenoweth and Stephan 2011; Arjona 2016; Kaplan 2017; Krause 2018; Masullo 2021b). Such success is counterintuitive as those who pursue nonviolent action are faced with much more powerful counterparts. The focus on nonviolent and unarmed forms of protection overlooks, however, that civilians often take up arms for protection, with important implications for how war evolves and ends. This chapter therefore focuses on armed forms of civilian protective agency—the emergence of armed self-defence forces in order to protect local residents from the consequences of war. I seek to make three contributions. First, I discuss in what ways armed civilian self-protection further complicates the distinction between civilians and combatants. I argue that a certain type of armed self-protection strategy, the formation of community-initiated part-time militias, can be considered part of civilian protective agency as this type of militia corresponds with the conceptualization of civilians in conflict studies. Secondly, I analyse the conditions under which armed civilian self-protection actually protects civilians. I discuss the theoretical expectations for the consequences of the emergence of armed self-protection strategies and apply them to a case study of armed self-protection in Mozambique during the civil war from 1976 to 1992. I make use of a data set of violent events that I compiled from archival sources and interview material I collected during fieldwork in Mozambique to show the trends of the evolution of violence during the war. Next, I conduct a careful case study of armed self-protection in a particular district in central Mozambique. Overall, I show that in Mozambique, certain conditions facilitated civilian protection, but the protection effect was temporary, as the rebels quickly learned how to respond to the new armed challenge.

Armed self-protection and the concept of the ‘civilian’ In the wake of a focus on micro-dynamics in civil war, much conflict research has focused on civilians as pivotal actors in war, whilst at the same time emphasizing the difficulty of distinguishing between civilians and combatants. Recognizing that armed groups require popular support to achieve military objectives such as territorial control, Kalyvas (2006, 2008) shows how violence shapes civilians’ willingness to provide such support. Concepts such as ‘civilian agency’, ‘civilian self-protection’, and ‘civilian protective agency’ aim to demonstrate the power that those without a gun can have over those who do have access to weapons. Such power lies in the access to crucial knowledge and information, strong leadership,

CORINNA JENTZSCH

137

and the sheer number of people able to act collectively (Kaplan 2017; Krause 2018). However, civil war scholars also emphasize the difficulty of distinguishing between civilians and combatants. This calls into question the persistent use of the concept of the ‘civilian’ and the fact that much theorizing relies on it. ‘Civilian’ in ‘civilian self-protection’, for example, refers to the idea that these strategies are ‘primarily selected and employed by civilians’ (Jose and Medie 2016); the concept of ‘civilian protective agency’ developed by the editors in the introduction to this volume is used in a similar way. Can we speak then of ‘armed civilian self-protection’, or does this further blur the conceptual line between civilians and combatants? Do civilians, once they arm themselves, turn into combatants? Scholars of civilian agency and civilian selfprotection include armed forms of self-protection in their typologies of civilian agency. In fact, Kaplan (2017 p. 57) lists six strategies that communities adopt to protect themselves, the last of which is ‘local-based armed resistance against external armed groups to protect residents’. The question remains, however, whether including armed forms of resistance in such a list actually obscures the distinction between combatant and civilian even further. It helps to analyse the distinctions made in international law and conflict studies to evaluate this question. At the core of international humanitarian law is the distinction between combatants and civilians used to identify legitimate military targets in war and protect civilians. As discussed in the introduction to this volume, the definition of ‘civilians’ in international humanitarian law is a negative one: according to Article 50 of the Additional Protocol I to the Geneva Conventions from 1977, civilians are those individuals who are not combatants, meaning that they are not recognizable members of organized armed groups with a clear command structure. In civil wars, this distinction is difficult, as armed groups often do not use uniforms, or they follow a decentralized command structure, which makes their group difficult to be recognized as an army with command and control. Another criterium offered by the Additional Protocols is therefore that civilians lose their civilian status when they directly participate in hostilities (Additional Protocol I Articles 45.1, 51.3; Additional Protocol II Article 13.3). Although important, the term ‘direct participation in hostilities’ lacks a clear definition, which led the International Committee of the Red Cross (ICRC) to provide interpretive guidance. The ICRC guidance emphasizes that the term means participation in specific acts that intend to cause harm in support of a party to the conflict, rather than an activity linked to one’s ‘status, function, or affiliation’ (Melzer 2009 p. 44).³ Even before the publication of that guidance, codes of conduct for armed forces developed similar, additional criteria to identify ³ This interpretive guidance is not uncontested, as subsequent debate showed. See Goodman and Jinks (2009).

138

CIVILIAN VIOLENT MOBILIZ ATION

legitimate military targets. Some Rules of Engagement of the United States (US) military, for example, explicitly allow for the (lethal) targeting of any individual that shows ‘hostile intent’ by participating in or being complicit in a violent incident against US or allied forces, or against civilians that are in need of protection.⁴ However, these criteria are the focus of much debate and are regularly reviewed and refined. This is true not only for US military operations, but also, for example, for Dutch military operations such as the one in Afghanistan (Boddens Hosang 2009). In conflict studies, the distinction between civilians and combatants has followed a different logic. Most fundamentally, civilians’ crucial role in supporting armed groups has been recognized. Arjona (2016 p. 25) defines civilians as ‘persons who live in a given local territory, do not participate in hostilities, and are not full-time members of any state or non-state armed organization’. In her work on civilian displacement, Steele (2017 p. 18) defines civilians as ‘individuals who do not participate in the military activities of any armed group but who may be “parttime” affiliates or collaborators’. Both Arjona and Steele emphasize that civilians do not directly take part in violence, which makes them, if they do support armed groups in some other way, ‘part-time’ members. Full-time members are those who do participate in hostilities, thus combatants. However, in wars that are fought on a local level, within communities, that distinction is more difficult to make. In communal wars, for example, Krause (2018 p. 22) refers to members of thugs and gangs as ‘armed civilians’ in contrast to ‘unarmed civilians’. This implies that civilians do not lose their civilian status when carrying weapons as any membership in armed groups for them remains temporary. What appears to be a (somewhat implicit) definition of being a civilian in recent research on armed conflict is the spontaneous, fleeting, and temporary character of participation in organized violence. It is this sense of temporary membership in armed groups that we can define as making a civilian. As long as community residents do not permanently join existing armed groups and established conflict actors, they do not lose their characteristics as civilians, as they will still engage in their day-to-day lives to support their livelihoods and their families. This is in contrast to the ICRC interpretation of who is a civilian, as civilians lose their protection status when engaging in acts that amount to direct participation of hostilities. This means that what I consider here as civilian self-protection might not amount to a protected activity under international humanitarian law, which implies that community-initiated militias may be seen as legitimate military targets in armed conflict, for the time they engage in these activities.

⁴ US Army, Center for Army Lessons Learned, Rules of Engagement Vignettes Handbook, No. 11-26, May 2011, https://info.publicintelligence.net/USArmy-ROE-Vignettes.pdf.

CORINNA JENTZSCH

139

Conceptualizing armed civilian self-protection Building on the discussion above, I conceptualize armed civilian self-protection as the formation of a specific type of militia. I conceive of militias as ‘armed organizations that exist outside of the formal security apparatus of the state; they emerge as “counter-movements” against insurgents either on the initiative of community residents or state representatives’ (Jentzsch 2022 p. 2; cf. Jentzsch et al. 2015). Not all militias can be considered a form of civilian self-protection, however. I distinguish between ‘community-initiated’ and ‘state-initiated’ militias as well as between full-time and part-time forces (see Figure 7.1). Given the definitions discussed above, only community-initiated part-time militias should be defined as a civilian self-protection strategy. Communityinitiated militias emerge from grassroots processes of collective action, while state-initiated militias form when states decide to make civilians part of their counterinsurgency strategy. Community-initiated militias may be co-opted by states at later stages in the war, while state-initiated militias may become more independent over time. I further distinguish between part-time and full-time forces to capture the level of professionalization of militias and limit the category of ‘civilian’ militias. For example, full-time forces stay in barracks, are on duty day and night, and have continuous access to weapons and sometimes even uniforms. They may go on missions outside the village they are stationed in, which implies that they conduct both offensive and defensive operations. Part-time forces are more likely to reside in their own houses and, for example, be on duty during the night whilst going about their regular activities during the day. Those forces may have limited access to weapons, and no access to uniforms. Their activities commonly include patrolling, intelligence gathering, policing tasks such as arrests, and defence in case of an attack. The two dimensions result in the typology shown in Figure 7.1. State-initiated full-time militias are professionalized and state-directed, such as the Janjaweed in Sudan’s Darfur region (Flint 2009). State-initiated part-time militias are illustrated by the Guatemalan civil patrols who patrolled villages by night, but otherwise went about their daily routines (Bateson 2017; Remijnse 2002). Community-initiated

Level of professionalization High

Low

State

State-initiated full-time militias

State-initiated part-time militias

Community

Community-initiated full-time militias

Community-initiated part-time militias

Initiator

Figure 7.1 Typology of militias.

140

CIVILIAN VIOLENT MOBILIZ ATION

militias rarely emerge as full-time forces, but rather evolve into those. For example, the Kamajor militias during the civil war in Sierra Leone represent communityinitiated part-time militias before they evolved into a full-fledged army, and community-initiated full-time forces after they replaced the state army (Hoffman 2011). I consider community-initiated part-time militias as an armed civilian selfprotection strategy and as one form of protective civilian agency, as it emphasizes the grassroots character of armed mobilization and the temporary status of militia members.

Consequences of armed civilian self-protection What are the consequences of the formation of community-initiated part-time militias for civilian protection? Kaplan (2017 p. 51) contends that ‘local-based armed resistance’ is the most contentious self-protection strategy, which implies that employing this strategy could lead to further escalation of the conflict and an increase in violence against civilians. In the following, I outline potential mechanisms by which armed self-protection may affect protection outcomes such as forms of direct, physical violence, but also indirect forms of violence such as forced displacement. Militias can help end the war through two main mechanisms. Due to their local character, militias can provide crucial local knowledge and intelligence to effectively target insurgents, limit attacks against civilians, and separate insurgents from their civilian resource base (Peic 2014 p. 163). As a consequence, in states in which governments collaborate with militias, they have a better chance of defeating insurgents (Peic 2014). In addition to providing crucial information, militia deployment multiplies armed forces in cases where the state army has a lack of recruits, facilitating a more effective state response to the insurgents (Carey et al. 2013). I term these two mechanisms the intelligence provision and the force multiplicator mechanisms. However, militias may also contribute to violence against civilians. First, they may trigger additional violence perpetrated by the conflict’s armed groups. Armed resistance against insurgents can pose significant costs on the armed group, which raises their interest in militarily defeating the militias, attracting even more violence upon civilians. Indeed, by eliminating the civilian support base for insurgents, armed groups may retaliate and indiscriminately target civilians (Clayton and Thomson 2014; Clayton 2016). With less information on those who feed information to the other side, rebels are inclined to target civilians more indiscriminately (Kalyvas 2006). I term this the retaliation mechanism. Second, militias may become violent actors themselves, which contributes to the increase in the length and lethality of civil wars (Starn 1995; Hoffman 2011;

CORINNA JENTZSCH

141

Mitchell et al. 2014; Aliyev 2020a). Militias can (further) fragment armed groups and develop a vested interest in continued fighting, which hampers negotiated settlements (Stedman 1997; Staniland 2015). They might develop material interests in benefiting from the spoils of war (Aliyev 2020b), or intensify ethnic divides in cases in which the incumbent and the militia are part of the same ethnic group (Abbs et al. 2020). More generally, militias may polarize the conflict, as the multiplication of armed groups provides opportunities for community residents to settle local conflicts (Clayton and Thomson 2016). These dynamics may lead militias to engage in violence against civilians. I call this the fragmentation mechanism. Thirdly, collaboration between rebels and militias or state forces and militias may contribute to an increase in violence against civilians. The incumbent regime, for example, can co-opt militias and outsource violence to them. The state’s main advantage of outsourcing violence is that it can avoid being held accountable for human rights violations perpetrated by forces in a loose alliance with the state (Carey et al. 2013; Mitchell et al. 2014). In some cases, militia violence takes place alongside state violence (Cohen and Nordås 2015). I call this the co-optation mechanism. Overall, militias empower civilians, but in a way that can hamper conflict resolution and further violent conflict. But why do communities then choose for armed self-protection? As I show in the following, the peaceful effects of self-defence are often short-term, and therefore temptatious but also temporary.

Armed self-protection during the civil war in Mozambique I analyse armed civilian self-protection in the form of community-initiated parttime militias in the context of the civil war in Mozambique, which took place from 1976 to 1992 between the party in power, Frelimo, and the rebel movement Renamo.⁵ The Mozambican war was foremost a war about the people (Cahen et al. 2018; Jentzsch 2022), which led the civilian population to develop protective agency. Gaining control over the population through violence and forced resettlement was an important strategy of war, as the population provided important ⁵ Renamo stands for ‘Resistência Nacional Moçambicana’ (National Mozambican Resistance) and Frelimo stands for ‘Frente de Libertação de Moçambique’ (Mozambique Liberation Front). The party Frelimo was the successor of the main liberation movement before Mozambique’s independence in 1975. For an introduction into the war, see Vines (1991) and Finnegan (1992). Overall, the war cost an estimated 1 million lives and displaced almost 5 million people, both as a consequence of fighting and war-induced famines and diseases, and took a heavy toll on infrastructure, with 60 per cent of primary schools and 40 per cent of health clinics destroyed in 1992 (Hanlon 1996 pp. 15–16). The data for this chapter come from interviews and archival documents collected during 13 months of fieldwork in Mozambique in 2011–12, 2016, and 2019. See Jentzsch (2022) for details on methods and data collection.

142

CIVILIAN VIOLENT MOBILIZ ATION

resources such as intelligence, food for the fighters, and labour for the transport of supplies. Civilians developed different strategies of self-protection, ranging from individual to collectively organized ones. Fleeing war-affected areas and hiding in the bush, often from the late afternoon until the next morning, was a frequent response to the violence. As the war dragged on, such activities became more organized. Groups of young men were sent to patrol at night, and the government provided several thousand residents with limited training and knives and spears to defend communities (Lauriciano 1986). When it became clear, however, that the government and the rebels were stuck in a military stalemate in 1988, a more organized and armed self-protection strategy emerged. The military stalemate had led to heavy fighting and violence against civilians, without providing either side the necessary leverage to defeat the other and leaving the population without much protection (Jentzsch 2022 ch. 4). A famine in the central region worsened the humanitarian situation. As a last resort, a militia formed in the Nampula and Zambézia provinces in northern Mozambique. Naparama, as it was called, formed as a grassroots initiative and recruited peasants who would go about their work during the day, but join the group as parttime fighters at night. It was an armed movement led by a traditional healer who succeeded in increasing the number of his followers from an estimated 400 in July 1990 (Maier 1990a) to an estimated 4,000–6,000 members across both provinces (Jentzsch 2022). The group gained control over two-thirds of the northern territory within a short amount of time, becoming ‘one of the most important military and political factors in contemporary Mozambique’ (Wilson 1992 p. 561). The militia had a central command that conducted its own operations and mobilized units at the district level, which operated independently. The units were unified by the belief in a potion that supposedly rendered fighters immune to bullets. The Naparama leader Manuel António formed units in Zambézia after he, as he claimed, had discovered a powerful medicine to protect fighters during battle. Every new recruit underwent an initiation ritual during which they were ‘vaccinated’ with the medicine. To ensure the medicine’s effectiveness, members had to obey strict rules, at home and on the battlefield, such as avoiding certain foods and only using spears and knives in combat (armas brancas, ‘white weapons’). Deaths during battle were usually attributed to violations of these rules. I first conduct a case study of the formation of these militias in a district in central Mozambique that was affected by the war to understand how and when community-initiated part-time militias contribute to civilian protection. I refine and explore further the mechanisms identified above, identifying the conditions under which the intelligence gathering and force multiplier mechanisms dominate. In a second step, I look at the conflict from a regional and national perspective to demonstrate the (temporal) limits of civilian protection through armed self-protection.

CORINNA JENTZSCH

143

How and when armed self-protection succeeds ‘It was the Naparama who ended the war’, an elder told me in a village in northern Mozambique.⁶ For the peasants who experienced the war in rural Mozambique, the community-initiated militia provided much-needed relief. A community resident in the Nicoadala district argued that Naparama delivered on its promises: ‘When [Manuel António] Parama came, he said that no one will ever [have to] flee again from their home. And indeed, we were able to sleep in our homes’.⁷ People perceived the return to their homes as a precondition for securing their livelihoods, which meant that protection from direct and indirect forms of violence was a precondition for protection from hunger.⁸ In analysing the civilian response to civil war violence, I focus on a district in the Zambézia province immensely affected by the war, Lugela. Lugela was a district of 107,000 people (according to the census of 1980) about 240 km from the provincial capital Quelimane and 58 km from Mocuba, an important town with military headquarters along the main north–south Highway EN1. The war came to the district early due to the proximity to Malawi, the mountainous terrain with thick forests, and the state-owned tea factories that became the first targets of rebel violence.⁹ Initially, local chiefs, promised that they would regain the power that Frelimo had taken from them at independence, overwhelmingly supported Renamo.¹⁰ However, popular support decreased when Renamo began expanding its military presence in the rural areas, recruiting youths from occupied areas and targeting civilians indiscriminately.¹¹ Renamo units attacked and occupied the district town in March 1986 for several months, and then four more times until 1988. Whenever Frelimo regained control of the town, rebels constantly attacked, and control of surrounding areas changed frequently.¹² Whenever the army was able to organize convoys with supplies from Mocuba, Renamo returned and assaulted the convoys to steal the supplies.¹³ This chaotic situation had severe impacts on civilian protection. People constantly had to flee from violence and search for food, which contributed to their

⁶ Interview with male civilian (m10), Nahipa, Mecubúri, Nampula, 26 October 2011. ⁷ Interview with civilian (f1), Nicoadala, Zambézia, 14 September 2011. ⁸ Interview with former Naparama combatant (Nm1), Nicoadala, Zambézia, 9 September 2011. ⁹ 11 República Popular de Moçambique, Província da Zambézia, Gabinete do Administrador do Distrito do Lugela, Relatório mensal de 15 Agosto a 15 de Setembro de 1982, 14 September 1982 (AGZ, Quelimane); Interview with community leaders (Gr-L), Lugela, Tacuane, Zambézia, 14 June 2012. ¹⁰ Interview with civilian (m36), Lugela, Zambézia, 18 June 2012. ¹¹ Interview with community leaders (Gr-L), Tacuane, Lugela, Zambézia, 14 June 2012. ¹² Interview with local government official (Gm15), Quelimane, Zambézia, 22 February 2012. ¹³ Interviews with local government officials (Gm15), Quelimane, Zambézia, 22 February 2012; (Gm27), Lugela, Zambézia, 19 June 2012.

144

CIVILIAN VIOLENT MOBILIZ ATION

‘nomadic existence’.¹⁴ Only one military battalion—half of which consisted of newly trained local forces—protected the district.¹⁵ Moreover, since the army did not have any supplies, it had to live off the population and often did so by force. The administration’s work was made difficult by the fact that part of it was stationed in a camp for the displaced close to Mocuba (Namagoa). Naparama emerged precisely in that camp for the displaced of Lugela district, in Namagoa, presumably in late 1990, as a part-time community activity. Naparama leader António, coming from Mocuba, asked the local administration for permission to mobilize followers, and neighbourhood secretaries helped to mobilize volunteers to patrol the camp at nighttime.¹⁶ Shortly thereafter, the district administration sent the militia to Lugela to help facilitate the return of the displaced from Namagoa. The force reached a size of 240 men and quickly became part of the local security apparatus.¹⁷ However, the militia retained its part-time, and thus civilian, character as recruits stayed in huts at the militia’s headquarters but went to their homes during the day when not on missions.¹⁸ The militia managed to help the administration keep control over the district town and protect civilians from violence and displacement. The mechanisms that shaped Naparama’s impact on civilian protection were intelligence gathering and force multiplier, more than retaliation, fragmentation, or co-optation. Even though they became part of the local security apparatus, Lugela’s Naparama managed to retain a certain level of autonomy, which ensured that they remained accountable to local residents who sought protection. The militia was able to avoid fragmentation and retaliation, mainly, as I argue, due to its positive relationship to the local administration and the army. The district administration recognized that it had to rely on communityinitiated militias as a last resort to protect civilians, leading a government official who appeared resigned to state: ‘Even if Naparama had not appeared, the people would have asked the military for firearms’.¹⁹ The most important protection effort was ‘recuperating’ civilians who had been abducted by Renamo and resettled in rebel-controlled areas. The militia’s success in returning abducted family members to protected areas generated extensive support for Naparama, from both the

¹⁴ República Popular de Moçambique, Província da Zambézia, Gabinete do Administrador do Distrito de Lugela, Relatório trimestral (Julho a Setembro/87), 1 October 1987 (AGZ, Quelimane); República Popular de Moçambique, Província da Zambézia, Gabinete do Administrador do Distrito de Lugela, Relatório, período de 1 a 30 de Junho, ano de 1988, 15 July 1988 (AGZ, Quelimane). ¹⁵ Interviews with former Frelimo combatant (Fm24), Lugela, Zambézia, 12 June 2012; and local government official (Gm15), Quelimane, Zambézia, 22 February 2012. ¹⁶ Interviews with former militiamen (Gr-Mm1), Lugela, Zambézia, 9 June 2012; and former Naparama commander (Nm46), Lugela, Zambézia, 10 June 2012. ¹⁷ Interview with former Naparama commander (Nm46), Lugela, Zambézia, 10 June 2012. ¹⁸ Interviews with former Naparama combatant (Nm49), Tacuane, Lugela, Zambézia, 15 June 2012; and community leaders (Gr-Lm4), Tacuane, Lugela, Zambézia, 14 June 2012. ¹⁹ Interview with local government official (Gm27), Lugela, Zambézia, 19 June 2012.

CORINNA JENTZSCH

145

population and local administration.²⁰ The fact that the militia was led by a highlyregarded, Frelimo-loyal community leader, Namudavale, also contributed to the trust in the militia. Namudavale, a former combatant of the liberation struggle, was secretary of the local Frelimo party structures that had mobilized people after independence and held administrative positions in the Namagoa camp for the displaced. The trust also found expression in the fact that, when Naparama’s strength seemed to wane after its leader António had died in 1991, the local administration supplied the militia with firearms.²¹ A relationship based on mutual trust also evolved between Naparama and the armed forces. When I asked Namudavale about allegations that the military stole relief aid during the war, he demonstrated great loyalty by strongly denying any such corruption existed within the armed forces. He also emphasized that Naparama and the army conducted joint operations, during which each group remained independent and respected each other.²² Former Frelimo combatants confirmed the positive relationship between Naparama and the army in Lugela, stating that they considered each other ‘friends’ who were ‘fighting for the same purpose’.²³ This affirmation by Frelimo soldiers of their close collaboration with Naparama is remarkable, as in other districts and on the provincial level, Frelimo representatives usually downplayed their reliance on Naparama. Officially, Frelimo could not concede that the party compromised its commitment to a socialist ideology that is hostile to traditional forms of power (Dinerman 2006). In sum, trust between the district administration, militia, and army prevented Naparama from becoming an additional armed group in the conflict that could have contributed to fragmentation, competing interests, and violence. The community-initiated militia grew quickly in size, was integrated into the local security apparatus, and played a vital part in civilian protection.

The limits of armed self-protection Trusting relations between militias and local administrative and security institutions can avoid fragmentation and co-optation, but as I show with regional data on violent events, rebel retaliation is more difficult to prevent. Qualitative and quantitative evidence from interviews, journalistic accounts, and archives shows that Naparama succeeded in pacifying areas and protecting civilians in many areas across the Nampula and Zambézia provinces. But this success was temporary, which demonstrates the limits of armed civilian protective agency. ²⁰ Interview with local government official (Gm27), Lugela, Zambézia, 19 June 2012. ²¹ Interview with former Frelimo combatant (Fm24), Lugela, Zambézia, 12 June 2012. ²² Interview with former Naparama commander (Nm46), Lugela, Zambézia, 10 June 2012. See also Interview with former Naparama combatant (Nm49), Tacuane, Lugela, Zambézia, 15 June 2012. ²³ Interview with former Frelimo combatant (Fm24), Lugela, Zambézia, 12 June 2012.

146

CIVILIAN VIOLENT MOBILIZ ATION

In the memory of those affected, but also in journalistic reports of the time, Naparama was described as a force that had a surprising and important impact on the war across the Zambézia and Nampula provinces, which allowed people to return home and take up work on their fields (Maier 1990a, 1990b). However, after a while, Naparama activities also contributed to violence against civilians (Wilson 1992 p. 574; Nordstrom 1997). After its initial successes and Renamo’s retreat, the militia became the main target for Renamo operations, increasing the number of attacks on civilians and even triggering Renamo to form an anti-Naparama force, a militia of their own (Jentzsch 2017). Moreover, Naparama became increasingly involved in violence against civilians themselves, looting goods and deliberately killing civilians they suspected of working with Renamo, in particular during operations that sought to recuperate people to government-held areas. These dynamics can be illustrated with evidence from government reports that I collected in the archives in Zambézia’s provincial capital Quelimane.²⁴ In 1985 and 1986, Renamo expanded and occupied many district capitals, which were liberated during the government’s counteroffensive in 1987–88 (Jentzsch 2022 ch. 4). The rebels were not able to re-establish control over all the previously occupied district capitals after that counteroffensive and Naparama’s emergence, but they continued to stage attacks against these areas from rural rebel bases. Moreover, data of violent events from newspaper and archival government reports provide a picture of a decline in violence in Zambézia in 1991 after Naparama was well established (see Figures 7.2, 7.3, and 7.4). However, there is an increase in violence in 1992, which could be explained by reference to several developments. First, Naparama’s main leader in Zambézia, Manuel António, was killed in December 1991, presumably by Renamo forces. After the death of their leader, many Naparama units disintegrated and ended their activities, which probably gave Renamo more space to operate. Secondly, the peace process advanced significantly in 1991 and 1992, which triggered last efforts by Renamo to occupy districts before the signing of the ceasefire agreement on 4 October 1992. Part of the peace process was the planning of multiparty elections, and Renamo knew that it would fare better if it continued to have influence in certain areas beyond the signing of the peace agreement.

²⁴ I coded events from my collection of archival documents that include reports about district-level political, economic, and social developments and district radio messages to the provincial government about special occurrences. From the archival reports from Zambézia province and with a research assistant, I created a data set of violent events which includes over 1,300 events from 1974–94. Each event is categorized as a distinct type of violence so that the analysis can differentiate between violence against civilian and military targets and demonstrate patterns of violence across space and time. The data set focuses on the following actors: government armed forces, rebel forces, state-initiated militias, and community-initiated militias. The typology of events the data set includes builds on ACLED, which focuses on changes in territorial control (Raleigh et al. 2010). Battles can result in changes in territorial control or allow a conflict party to remain in control. The data set also codes violence against civilians as a separate category.

CORINNA JENTZSCH

147

100 90 80 70 60 50 40 30 20 10 0

1976 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994

Figure 7.2 Number of violent incidents in Zambézia province, 1976–94, based on national newspaper sources (N = 175). Source: Weinstein 2007.

100 90 80 70 60 50 40 30 20 10 0

1978 1979 1980 1981 1982 1983 1985 1986 1987 1988 1989 1990 1991 1992 1993

Figure 7.3 Number of battles between government and rebel forces in Zambézia province, 1978–93, based on provincial and district government reports (N = 262). Source: author’s coding, see appendix.

The data have to be read with caution though, as the year with the most violence is also the year with the most documents in my document collection. But it does reflect the historiography of the war, which shows that Renamo fought until the end to be well prepared for national multiparty elections (Do Rosário 2009). This means that while armed civilian self-protection can have a positive

148

CIVILIAN VIOLENT MOBILIZ ATION 100 90 80 70 60 50 40 30 20 10 0

1978 1979 1981 1982 1984 1984 198519861987198819891990 1991 19921993 1994

Figure 7.4 Number of incidents of violence against civilians in Zambézia province, 1978–94, based on provincial and district government reports (N = 262). Source: author’s coding, see appendix.

effect at the local level, its effect remains temporary and dependent on national peace processes that may influence incentives to (re-)escalate the conflict.

Conclusion This chapter explores the ways in which armed civilian self-protection can contribute to the protection of civilians in wartime and makes three major contributions. First, I argue that civilian protective agency is not only nonviolent. Armed forms of civilian self-protection do not contradict the concept of civilian selfprotection when recognizing the spontaneous, fleeting, and temporary character of civilian participation in organized violence for the purpose of protection. I show that a particular type of militia, community-initiated part-time militias, can be considered a form of civilian protective agency, which is in line with the definition of a civilian in this volume’s introduction—someone who is not a recognized member of an organized armed group. Secondly, I show that militias can have various effects on civilian protection depending on which mechanisms dominate the context in which militias act. When militias contribute to the gathering of local intelligence and multiply forces, then they contribute to civilian protection. When, instead, rebels retaliate, armed groups fragment, or the state co-opts militias, then militias contribute to more violence against civilians. Analysing the emergence of community-initiated part-time forces in a particular district in Mozambique— grassroots armed groups that engaged in patrolling, collecting intelligence, and

CORINNA JENTZSCH

149

limited military operations—I find that trusting relations between the state, army, and militia elites prevented the militia from becoming an additional actor in the conflict or being completely co-opted by the state. Thus, the militia was able to engage in civilian protection. Thirdly, regional data on violence from Mozambique demonstrates that rebel retaliation against militias and their supporters is difficult to prevent over the long term, and so, while militias can have an initial peaceful effect, their long-term effect might be to contribute to more violence overall. We see here that local war and peace is embedded in national war and peace, and if developments at the national level incentivize armed groups to escalate the conflict, they challenge local protection efforts (whether armed or unarmed). Thus, while fragmentation can be averted when militia leaders have close relations with state and army, retaliation by rebels is more difficult to avert and depends on national peace and conflict resolution processes.

References Abbs, L., Clayton, G., and Thomson, A. (2020). ‘The Ties that Bind: Ethnicity, Progovernment Militia, and the Dynamics of Violence in Civil War’, Journal of Conflict Resolution, 64/5: 903–32. Aliyev, H. (2020a). ‘Why Are some Civil Wars More Lethal than Others? The Effect of Pro-Regime Proxies on Conflict Lethality’, Political Studies, 68/3: 749–67. Aliyev, H. (2020b). ‘Pro-Regime Militias and Civil War Duration’, Terrorism and Political Violence, 32/3: 630–50. Arjona, A. (2016). Rebelocracy. Social Order in the Colombian Civil War. New York: Cambridge University Press. Bateson, R. (2017). ‘The Socialization of Civilians and Militia Members: Evidence from Guatemala’, Journal of Peace Research, 54/5: 634–47. Boddens Hosang, H. J. F. R. (2009). ‘A andachtspunten in de ISAF ROE vanuit het strategisch-juridische kader’, Militair Rechtelijk Tijdschrift, 5: 219–26. Cahen, M., Morier-Genoud, E., andDo Rosário, D. (2018). ‘Introduction: The Civil War in Mozambique: A History Still to Be Written’, in Cahen M., Morier-Genoud E., and Do Rosário D. (eds), The War Within. New Perspectives on the Civil War in Mozambique, 1976–1992, International Security, pp. 1–14. Woodbridge, Suffolk: James Currey. Carey, S. C., Mitchell, N. J., and Lowe, W. (2013). ‘States, the Security Sector and the Monopoly of Violence: A New Database on Pro-Government Militias’, Journal of Peace Research, 50/2: 249–58. Chenoweth, E., and Stephan, M. J. (2011). Why Civil Resistance Works: The Strategic Logic of Nonviolent Conflict. New York: Columbia University Press. Clayton, G. and Thomson, A. (2016). ‘Civilianizing Civil Conflict: Civilian Defense Militias and the Logic of Violence in Intra-State Conflict’, International Studies Quarterly, 60/3: 499–510. Clayton, G., and Thomson, A. (2014). ‘The Enemy of My Enemy Is My Friend … the Dynamics of Self-Defense Forces in Irregular War: The Case of the Sons of Iraq’, Studies in Conflict & Terrorism, 37/11: 920–35.

150

CIVILIAN VIOLENT MOBILIZ ATION

Cohen, D. K., and Nordås, R. (2015). ‘Do States Delegate Shameful Violence to Militias? Patterns of Sexual Violence in Recent Armed Conflicts’, Journal of Conflict Resolution, 59/5: 877–98. Dinerman, A. (2006). Revolution, Counter-Revolution and Revisionism in Post-Colonial Africa. The Case of Mozambique, 1975–1994. London/New York: Routledge. Do Rosário, D. (2009). Les mairies des «autres»: une analyse politique, socio-historique et culturelle des trajectoires locales. Le cas d’A ngoche, de l’Île de Moçambique et de Nacala Porto. Université Montesquieu/Bordeaux IV, Bordeaux. Ferris, E. G. (2011). The Politics of Protection. The Limits of Humanitarian Action. Washington, DC: Brookings Institution Press. Finnegan, W. (1992). A Complicated War: The Harrowing of Mozambique. Berkeley: University of California Press. Flint, J. (2009). Beyond “Janjaweed”. Understanding the Militias of Darfur. Geneva: Small Arms Survey, Graduate Institute of International and Development Studies. Goodman, R., and Jinks, D. (2009). ‘The ICRC Interpretive Guidance on the Notion of Direct Participation in Hostilities Under International Humanitarian Law: An Introduction to the Forum’, NYU Journal of International Law & Politics, 42: 637. Hanlon, J. (1996). Peace Without Profit: How the IMF Blocks Rebuilding in Mozambique. Oxford: James Currey. Hoffman, D. (2011). The War Machines: Young Men and Violence in Sierra Leone and Liberia. Durham, NC: Duke University Press. Jentzsch, C. (2017). ‘Auxiliary Armed Forces and Innovations in Security Governance in Mozambique’s Civil War’, Civil Wars, 19/3: 325–47. Jentzsch, C. (2022). Violent Resistance. Militia Formation and Civil War in Mozambique. New York: Cambridge University Press. Jentzsch, C., Kalyvas, S. N., and Schubiger, L. I. (2015). ‘Militias in Civil Wars’, Journal of Conflict Resolution, 59/5: 755–69. Jose, B., and Medie, P. A. (2015). ‘Understanding Why and How Civilians Resort to Self-Protection in Armed Conflict’, International Studies Review, 17/4: 515–35. Jose, B., and Medie, P. A. (2016). ‘Civilian Self-Protection and Civilian Targeting in Armed Conflicts: Who Protects Civilians’, Oxford Research Encyclopedia of Politics. Kalyvas, S. N. (2006). The Logic of Violence in Civil War. New York: Cambridge University Press. Kalyvas, S. N. (2008). ‘Promises and Pitfalls of an Emerging Research Program: The Microdynamics of Civil War’, in Kalyvas S. N., Shapiro I., and Masoud T. (eds), Order, Conflict and Violence, pp. 397–421. New York: Cambridge University Press. Kaplan, O. R. (2017). Resisting War. How Communities Protect Themselves. New York: Cambridge University Press. Krause, J. (2018). Resilient Communities. Non-Violence and Civilian Agency in Communal War. New York: Cambridge University Press. Lauriciano, G. (1986). ‘Resistência popular cresce na Zambézia. Dez mil pessoas armadas com zagaias’, Notı´cias, 22 November. Maier, K. (1990a). ‘Renamo Flee at Sight of Rag-Tag Army’. Independent, 27 July: 12. Maier, K. (1990b). ‘Triumph of Spears Over Guns Brings Refugees Home’. Independent, 23 February. Masullo, J. (2021a). ‘Refusing to Cooperate with Armed Groups. Civilian Agency and Civilian Noncooperation in Armed Conflicts’, International Studies Review, 23/3: 887–913.

CORINNA JENTZSCH

151

Masullo, J. (2021b). ‘Civilian Contention in Civil War. How Ideational Factors Shape Community Responses to Armed Groups’, Comparative Political Studies, 54/10: 1849–84. Melzer, N. (2009). Interpretive Guidance on the Notion of Direct Participation in Hostilities Under International Humanitarian Law. Geneva: International Committee of the Red Cross. Mitchell, N. J., Carey, S., and Butler, C. (2014). ‘The Impact of Pro-Government Militias on Human Rights Violations’, International Interactions, 40/5: 812–36. Nordstrom, C. (1997). A Different Kind of War Story. Philadelphia: University of Pennsylvania Press. Peic, G. (2014). ‘Civilian Defense Forces, State Capacity, and Government Victory in Counterinsurgency Wars’, Studies in Conflict & Terrorism, 37/2: 162–84. Raleigh, C., Linke, A., Hegre, H., and Karlsen, J. (2010). ‘Introducing ACLED: An Armed Conflict Location and Event Dataset Special Data Feature’, Journal of Peace Research, 47/5: 651–60. Remijnse, S. (2002). Memories of Violence. Civil Patrols and the Legacy of Conflict in Joyabaj, Guatemala. Amsterdam: Rozenberg Publishers. Schubiger, L. I. (2021). ‘State Violence and Wartime Civilian Agency: Evidence from Peru’, Journal of Politics, 83/4. Staniland, P. (2015). ‘Militias, Ideology, and the State’, Journal of Conflict Resolution, 59/5: 770–93. Starn, O. (1995). ‘To Revolt against the Revolution: War and Resistance in Peru’s Andes’, Cultural Anthropology, 10/4: 547–80. Stedman, S. J. (1997). ‘Spoiler Problems in Peace Processes’, International Security, 22/2: 5–53. Steele, A. (2017). Democracy and Displacement in Colombia’s Civil War. Ithaca, NY: Cornell University Press. Vines, A. (1991). Renamo: Terrorism in Mozambique. London: Centre for Southern African Studies, University of York in association with James Currey. Weinstein, J. M. (2007). Inside Rebellion: The Politics of Insurgent Violence. New York: Cambridge University Press. Wilson, K. B. (1992). ‘Cults of Violence and Counter-Violence in Mozambique’, Journal of Southern African Studies, 18/3: 527–82. Wood, E. J. (2003). Insurgent Collective Action and Civil War in El Salvador. New York: Cambridge University Press.

8 Vigilantism as civilian protective agency The case of autodefensas in Mexico Moshe Ben Hamo Yeger and Juan Masullo

Introduction Santiago Sochiapan, home to approximately 2,500 people, is a municipality in the state of Veracruz in Eastern Mexico. Known for its religious festivities, Santiago Sochiapan recently made the news for entirely different reasons. Organized crime had made life nearly impossible there. Residents could not reach hospitals, go to work, take their kids to school, or even do their daily grocery shopping without fear of being robbed, extorted, kidnaped, or killed. On several occasions, victims turned to local authorities, including the prosecutor’s office, only to find an unresponsive government. They decided to take matters into their own hands and create an armed self-defence force to protect their community. Various criminal organizations fight for control of Santiago Sochiapan. Los Zetas, one of the deadliest drug cartels in the country, used to reign in this territory. Yet, in recent years, the Cartel Jalisco Nueva Generación (CJNG) and smaller affiliated gangs challenged Los Zetas’ dominance. Amid this confrontation and with no trust in local authorities, residents organized to patrol the area. Today, Santiago has a self-defence force comprised of hundreds of farmers, cattle ranchers, and housewives, armed with everything from sticks and machetes to AR-15 assault rifles. Moreover, this self-defence force forms part of a more extensive network of communities across the state: the People United Against Crime (PUCD, or Pueblos Unidos Contra la Delincuencia). This is only one of many similar stories of communities organizing and arming themselves to counter criminal violence in contemporary Mexico. Since 2006, fuelled by an aggressive military campaign launched by the federal government to fight drug cartels, the country has experienced unprecedented levels of criminal violence. Criminal organizations have fragmented into smaller, yet highly predatory groups. They have increased their presence across the country and diversified their illicit activities from drug trafficking to various expressions of criminal violence, including extortion, kidnapping, human trafficking, and homicides.

Moshe Ben Hamo Yeger and Juan Masullo, Vigilantism as civilian protective agency. In: Civilian Protective Agency in Violent Settings. Edited by: Jana Krause, et al., Oxford University Press. © Moshe Ben Hamo Yeger and Juan Masullo (2023). DOI: 10.1093/oso/9780192866714.003.0009

MOSHE BEN HAMO YEGER AND JUAN MASULLO

153

Many ordinary civilians have mobilized in the face of increased insecurity. Many of their responses have been nonviolent, including mass demonstrations, the formation of civic associations, and the creation of neighbourhood watch groups. Yet, as in other violent contexts explored in this volume (Jentzsch within volume; Moncada within volume), ordinary civilians have also opted for armed mobilization. This chapter focuses on armed self-defence organizations, such as the force that recently emerged in Santiago Sochiapan, typically known as vigilante groups.¹ In line with the volume’s core conceptual framework (Chapter 1), we examine vigilantism as a potential form of civilian protective agency in the context of criminal wars. Is vigilantism a response to the threat posed by criminal organizations? Do vigilante groups operate where violence is the most intense? Is vigilantism a response to all types of criminal violence? To answer these questions, we map the spatial distribution of the phenomenon across the country in the last decade and examine its association with patterns of criminal violence. This chapter aims to break descriptive ground by providing a more detailed and comprehensive overview of how the incidence of vigilante groups is linked to different expressions of the country’s criminal war. Our findings support the claim—made by many self-defence groups themselves—that vigilantism is a protection-seeking response to criminal violence. Yet, they provide important nuance that deepens our understanding of the dynamics of vigilantism as civilian protective agency. First, we find that vigilante groups do tend to operate in municipalities where major drug cartels are present. However, we further show that the number of cartels matters: the likelihood of vigilante presence is substantially higher in the face of multiple drug cartels. Second, while vigilante groups tend to operate in the face of criminal violence, they do so only when and where levels of violence are neither too high nor too low. Finally, we find that not all types of criminal violence seem to matter equally. While there is consistent and strong evidence that homicides are associated with the incidence of vigilantism, kidnappings and extortion are not (or are only weakly) associated with it. Overall, our findings suggest that vigilantism in Mexico is an expression of civilian protective agency vis-à-vis drug cartels and some forms of criminal violence. A better understanding of the forces that shape the incidence of vigilantism is crucial for understanding Mexico’s ongoing criminal war and politics more generally. With the rapid expansion of vigilantism in 2013 and its continued presence,

¹ We follow Phillips’ (2017 p. 1360) definition of vigilante groups as ‘sustained associations of private citizens voluntarily seeking to illegally control crime or other social infractions in a planned, premeditated way, involving force or the threat of force’. In this chapter we do not engage in a conceptual discussion around important questions such when a vigilante group becomes a criminal group or the basis on which we can conceptualize vigilantes as civilian. Doing so goes beyond the scope of our chapter. Conceptual boundaries have been discussed in detail elsewhere (e.g. Johnston 1996; Moncada 2017; Bateson 2021), including in this volume (see Introduction and Jentzsch’s chapter).

154

THE CASE OF AUTODEFENSAS IN ME XICO

vigilante groups have established themselves as ‘third actors’ (Jentzsch et al. 2015; Jentzsch 2022 ch. 2) in the country’s drug war. In places such as Santiago Sochiapan, these groups enjoy considerable legitimacy (which state security forces often lack), closely cooperate with communities, and are effective providers of public security. In other places, however, they have been co-opted by opportunism and by criminal groups, morphing sometimes into criminal organizations themselves. This has further undermined the rule of law and, on occasions, led to increased violence and insecurity. Having a clear map of the spatial distribution of vigilante incidence in the country is a crucial step towards identifying where communities are being left to their own devices, where civilian protective agency is improving security conditions, and where organized responses are feeding organized crime and spurring more violence. A comprehensive approach to restore public security and the rule of law in Mexico must factor in variation in the presence of vigilantism, the concrete activities they engage in and the effects they have on the communities where they operate.

Overview of vigilantism in Mexico Civilian armed mobilization in response to violence and insecurity is not a new phenomenon in Mexico. Historically, ordinary Mexicans have taken up arms to protect their communities from various external threats. Documented examples date back to the early 1800s, when vigilante groups were created during the civil strife that followed independence from Spain (Vanderwood 1992). Armed responses were also common after the armed phase of the Mexican Revolution (1910–17), when ordinary civilians organized into peasant militias known as agraristas to occupy land, while landlords resisted by forming paramilitary groups known as guardias blancas (Sánchez-Talanquer 2018). Further, one of the most illustrative examples comes from the Cristero Rebellion (1926–29), when citizens took up arms in central and western Mexico to defend themselves from antiCatholic policies imposed by the central government (Meyer 1976; Osorio et al. 2018). Moreover, the use of policı´as comunitarias (communal policing) is a tradition that many indigenous communities have practised for decades, especially in the states of Guerrero and Chiapas. Vigilantism has also been a present feature of contemporary Mexico. For example, in the 1990s, indigenous communities, regulated by customary law, organized violence of their own in response to common and petty crime (Gómez Durán 2012; Rea 2012). However, the phenomenon took a whole new form in the context of the ‘war on drugs’ launched by former president Felipe Calderón in 2006, as ordinary civilians increasingly turned to forms of community self-defence. Although there is evidence of vigilante presence and activity from the early years

MOSHE BEN HAMO YEGER AND JUAN MASULLO

155

of this war, 2013 marked a turning point with a sharp increase in the number of vigilante groups in various parts of the country. While vigilantism was initially concentrated in the neighbouring Pacific states of Guerrero and Michoacán where Calderón first launched the ‘war on drugs’, by the end of 2017, more than 200 vigilante groups had emerged in at least 26 of the country’s 32 states. Although 2013 and 2014 represent the peak years of vigilantism in Mexico, the phenomenon has persisted and even regrown since then. While Guerrero and Michoacán still host a larger share of vigilanta groups, vigilantism has spread geographically throughout large areas of the country. As Figures 8.1 and 8.2 illustrate, the presence of vigilante groups expanded from 99 municipalities in 14 states in 2013 to 214 municipalities in 26 states in 2017. By the end of 2017, vigilante organizations were present in almost 10 per cent of all Mexican municipalities. The specific processes that lead to vigilante emergence vary from town to town and group to group, as do their internal structures and modus operandi, and their relationship with local governments, security forces, and communities. In some instances, their ranks are formed by individuals carrying machetes and wearing no uniforms, while others act more like an organized army, with uniformed members armed with shotguns and automatic rifles. The activities they engage in also

State

Guerrero Michoacán Chiapas Veracruz Hidalgo Oaxaca Tlaxcala Puebla México Tamaulipas Sonora Yucatán Guanajuato Tabasco Morelos Jalisco San Luis Potosí Chihuahua Sinaloa Baja California Sur Quintana Roo Nuevo León Durango Coahuila Ciudad de México Baja California 0

20 40 Number of vigilante groups

Figure 8.1 Number of vigilante groups per state, 2013–17.

60

156

THE CASE OF AUTODEFENSAS IN ME XICO

Municipalities with vigilante groups (2013)

No Yes

Municipalities with vigilante groups (2017)

No Yes

Figure 8.2 Expansion of vigilante groups per municipality 2013–17.

vary widely, from interrogating suspects and handing them over to the authorities, to setting up checkpoints and barricades on major roads to stop criminals, to arresting suspects, engaging in direct firefights, and carrying out executions.² Despite this variation, the common narrative in Mexico is that these groups are formed to protect communities from criminal violence and to counter drug cartels. This narrative is promoted by vigilantes themselves. For example, in one of the first reported cases of the recent wave of vigilantism, a group in the small city of La Ruana in Michoacán led by local lime grower Hipolito Mora, declared its ² For a useful conceptual distinction between forms of vigilantism and an explanation of the different pathways to each of these forms, see Moncada (2022 ch. 2).

MOSHE BEN HAMO YEGER AND JUAN MASULLO

157

raison d’être explicitly as defending residents from the Knights Templar, one of the country’s strongest drug cartels (Wolff 2020). Similarly, and even more explicitly, in the capital of the state of Tabasco in southeast Mexico, the network of which the Santiago self-defence group came to form part, was created under the name of People United Against Crime (PUCD) with the declared aim of protecting residents from Los Zetas cartel (Navarro 2020). Time and again, vigilante leaders declare to the media that their primary goal is to protect ordinary people from criminal violence and restore peace and order in their communities.³ Yet, some of these groups have been co-opted by drug cartels and used as fronts to fight rivals, while others have evolved into criminal organizations themselves. The Familia Michoacana, for example, was originally formed as a vigilante group in the Tierra Caliente region of Michoacán with the mission of fighting Los Zetas cartel. However, after succeeding in driving out the cartel, they expanded into other states and became a drug cartel, engaging in various criminal activities besides drug trafficking, such as kidnapping, extortion, and racketeering (Asfura-Heim and Espach 2013 p. 144). Moreover, while some vigilante groups have managed to remain operative for long periods of time, others have dissolved shortly after their rise. In some instances, the state has successfully succeeded in disarming and disbanding them either by force or by negotiating their incorporation into local security forces (Navarro 2020). The different trajectories and outcomes of vigilante groups are a relevant and intriguing subject (e.g. Moncada within volume). However, in what follows, we focus on whether and how their presence across the Mexican territory is associated with the dynamics of the criminal war, which is crucial for an understanding of the phenomenon as an expression of civilian protective agency.

Research on vigilantism The phenomenon of vigilantism has recently attracted wide interest in the social sciences (Bateson 2021; Cohen et al. 2023). In addition to valuable efforts of conceptual clarification (Moncada 2017; Bateson 2021), scholars have theoretically and empirically explored a range of issues related to vigilantism: variation in the form it takes (Moncada 2022); the use of vigilante justice (e.g. Goldstein 2003; Handy 2004; Jung and Cohen 2020; Kloppe-Santamarı´a 2020); its relationship with political order, state formation, and democratization (Obert and Mattiacci ³ See, for example, ‘Comunidad de Hidalgo arma su grupo de autodefensa’ El Universal, available at: https://archivo.eluniversal.com.mx/estados/2015/hidalgo-autodefensa-1081252.html; ‘Conforman grupos de autodefensa por inseguridad en Coatzacoalcos’, Quadratin, available at: https:// veracruz.quadratin.com.mx/Conforman-grupo-de-autodefensa-por-inseguridad-en-Coatzacoalcos/; and ‘A hora en Nayarit srugen grupos de autodefensa ciudadana’, AZ Noticias, Available at: https:// aznoticias.mx/index.php/NAYARIT-MOVIL/7671-CREARAN-NUEVO-GRUPO-DE-AUTO DEFENSA-CIUDADANA-EN-NAYARIT

158

THE CASE OF AUTODEFENSAS IN ME XICO

2018; Sánchez-Talanquer 2018; Smith 2019), and public and police support for vigilantism (e.g. Tankebe 2011; Haas et al. 2012; Nivette 2016; Zizumbo-Colunga 2017, 2019), among others. In terms of geographical coverage, this research has been equally wide-ranging: from pre-civil war United States to post-Apartheid South Africa and post-civil war Guatemala (Brown 1976; Smith 2015; Bateson 2017). Mexico has featured prominently in this literature. There are rich case studies on specific instances of vigilantism in the country (e.g. Guerra Manzo 2015; Fuentes Dı´az and Fini 2018; Guerra 2018; Wolff 2020), as well as valuable efforts to identify larger patterns and single out some of the main determinants of the phenomenon. Studies examining vigilante emergence in Mexico have gone beyond traditional explanations based on issues of state weaknesses and effectiveness,⁴ identifying some unusual suspects. Scholars have, for example, identified how economic factors (Phillips 2017), migration (Ley et al. 2019a; Pérez-Armendáriz and Duquette-Rury 2019), and long-term historical legacies (Osorio et al. 2018; Sánchez-Talanquer 2018; Wolff 2020) have played a role in the emergence of the recent wave of vigilante groups in Mexico. While these studies shed new light on the determinants of vigilantism, their main focus has not been its association with the dynamics of criminal violence. Consequently, we still lack a general (and up-to-date)⁵ examination of vigilantism in Mexico as a response to the dynamics of organized crime in the country. This is surprising, given widespread narratives about vigilantism being a response to increased criminal violence and the fact that most vigilante groups in the country manifestly express being a response to drug cartels’ presence and violence. In this chapter, we take on the task of empirically examining this relationship. Identifying whether and how vigilante groups are indeed a response to criminal violence in Mexico is a first step towards unpacking its character as a form of civilian protective agency. Before presenting our main results, in the next section we briefly describe the data and methods that we used in our empirical examination.

Data and methods To examine the association between Mexico’s criminal context and the incidence of vigilantism we employ logistic regression models with random effects and robust standard errors suited for longitudinal data. The use of random effects is appropriate in this case because most independent variables in the model change ⁴ For classic works on vigilantism, see Rosenbaum and Sederburg (1974) and Abrahams (1987). ⁵ Most existing studies, both large-N and case studies, focus on the years of the outbreak (2013–14), without exploring how the phenomenon has evolved in the subsequent years.

MOSHE BEN HAMO YEGER AND JUAN MASULLO

159

slowly over time and the timeframe under study is relatively short (Wooldridge 2001). Our dependent variable tracks the evolution of vigilante incidence over a fiveyear period between 2013 and 2017. Across all models, it is a binary measure of instances of vigilantism at the municipal level per year, taking the value of one if vigilantes operated in a given municipality-year, and zero otherwise. An instance of vigilantism is an event signalling the presence of a vigilante organization in a municipality that is salient enough to capture the attention of local or national media outlets. These events range from the formation of new groups to disputes with government authorities and armed confrontations with criminal organizations. The data come from news reports from national and local media outlets gathered by Lantia Consultores—a leading research consultancy that studies security-related issues in Mexico.⁶ These data allowed us to discern over 300 instances of vigilantism between 2013 and 2017, which occurred in 214 different municipalities across the country. This empirical base, covering a total of 12,280 municipality–year observations over a five-year period, constitutes a development over previous studies (e.g. Phillips 2017; Ley et al. 2019a; Osorio et al. 2021), which either focus on a single year (2013) or capture fewer instances of vigilantism. Our data record locally rooted groups that came to fruition and that have remained autonomous. We set aside groups with alleged ties to criminal organizations. To exclude such groups, we rely on the information recorded in the media reports linked to each instance. For example, Proceso, a national news outlet, reported the formation of a new vigilante group in Cuetzalán del Progreso, Guerrero in 2014. However, the report also mentions that the main promoter of the group, a rancher from a neighbouring municipality, had been signalled by state authorities for his connection to Guerrero Unidos, a criminal organization.⁷ In this case, as in several equivalent ones, we excluded this instance from our dataset. As we rely solely on information provided by the press, we recognize that our data might well include groups that have been infiltrated by criminal groups after their emergence. Yet, given our focus on whether instances of vigilantism are associated with different expressions of criminal violence, rather than the trajectories of vigilante groups, this issue is less of a concern for our analysis. We use different measures to capture criminal violence. Our primary measure is homicide rates, coming from the Mexican Secretariat of Public Security (SNSP) and the National Institute of Statistics and Geography (INEGI). We use both the homicide rate in the previous year and its squared term to check for both linear and non-linear relationships. Homicides were selected as the main indicator of ⁶ We thank Eduardo Guerrero, director of Lantia Consultores, for sharing the data with us. ⁷ ‘Ganadero vinculado al narco impulsa grupo de autodefensa en Guerrero’, Proceso, available at: https://www.proceso.com.mx/nacional/2014/4/14/ganadero-vinculado-al-narco-impulsa-grupode-autodefensa-en-guerrero-131438.html

160

THE CASE OF AUTODEFENSAS IN ME XICO

criminal violence for two main reasons. First, homicides are probably the most reliable measure of crime in Mexico, as they are likely to suffer less underreporting relative to other forms of criminal violence. Second, theoretically speaking, such a serious and impactful form of violence is more likely to trigger risky civilian selfprotection responses. In addition to lethal violence, we test for associations with non-lethal expressions of criminal violence, such as violent assaults, kidnapping, and extortion. These data come from the SNSP. To test whether instances of vigilantism are associated with the presence of drug cartels, we relied on a dataset compiled by Sobrino (2020), which tracks the presence of nine major Drug Trafficking Organizations in all Mexican municipalities between 1990 and 2017. This dataset was built by scraping Google News Mexico and using natural language processing, covering almost 900 local, national, and international media outlets reporting in Spanish. To the best of our knowledge, this constitutes the most comprehensive dataset on cartel presence in Mexico to date.⁸ Our models control for a wide set of standard and theoretically informed variables, such as municipality size (a log of the total population), socio-economic conditions (poverty, percentage of households that received remittances, migration intensity, and municipal Gini coefficient), state capacity (ministerios públicos per 100,000 inhabitants, GDP per capita), level of urbanization, and ethnic composition. Data for these variables come from Mexico’s main statistical offices, most notably the National Council for the Evaluation of Social Development Policy (CONEVAL) and the INEGI, and Mexico’s central bank (Banxico). These controls help attenuate the downsides of using random effects in our models. Finally, we control for both spatial and temporal autocorrelation by adding variables that measure whether there have been incidents of vigilantism in contiguous municipalities in the past year and whether incidents have been recorded in the same municipality in the past year.

Vigilantism as a response to criminal violence Our first consistent finding is that vigilante presence is more likely in those areas where some of the major drug cartels are present. Across all model specifications the presence of a major cartel is positively associated with a higher probability of observing a vigilante instance—even if magnitudes vary. This association is statistically significant at conventional levels (p < 0.05; see Table 8.1). Moving from a ⁸ The nine cartels included in the dataset are: CJNC, Familia Michoacana, Cartel de Sinaloa, Cartel de Tijuana, Cartel del Golfo, Los Zetas, Beltran Leyva, Templarios, Cartel del Golfo. We thank Fernanda Sobrino for sharing her data on cartel presence.

MOSHE BEN HAMO YEGER AND JUAN MASULLO

161

Table 8.1 Logistic regression of vigilante presence in Mexican municipalities, 2013–17. DV: vigilante

Model (1)

Model (2)

Model (3)

Model (4)

Homicide t-1 ×1000

2.943∗∗∗ (5.41)

2.747∗∗∗ (5.11)

2.953∗∗∗ (5.42)

2.929∗∗∗ (5.38)

Homicide t-1 ×1000 (sq)

−1.728∗∗∗ (−3.61)

−1.591∗∗∗ (−3.39)

−1.741∗∗∗ (−3.63)

−1.724∗∗∗ (−3.61)

One cartel

0.119 (0.52)

0.0957 (0.42)

0.118 (0.52)

0.108 (0.47)

Two or more cartels

0.467∗ (2.14)

0.440∗ (2.05)

0.463∗ (2.13)

0.466∗ (2.14)

Indigenous %

−0.00301 (−1.02)

−0.00112 (−0.38)

−0.00341 (−1.14)

−0.00335 (−1.11)

Poverty

0.0193∗∗ (3.23)

0.0204∗∗∗ (3.33)

0.0197∗∗∗ (3.29)

0.0187∗∗ (3.11)

Gini

10.72∗∗∗ (6.44)

10.41∗∗∗ (6.34)

10.75∗∗∗ (6.46)

10.76∗∗∗ (6.49)

Urbanization

−5.643∗∗∗ (−5.04)

−5.190∗∗∗ (−4.70)

−5.654∗∗∗ (−5.04)

−5.668∗∗∗ (−5.05)

GDP per capita

0.00747 (1.19)

0.00958 (1.55)

0.00709 (1.12)

Public ministries

Remittances

−5.099 (−1.04) −0.00640 (−0.54)

Remittances (sq)

0.09620∗∗ (2.66) −0.00460∗∗

Migration

Extortion

−1.113 (−0.80)

−0.999 (−0.74)

−0.0358 (−0.94)

−0.0387 (−1.01)

−1.161 (−0.83)

−1.047 (−0.76) Continued

162

THE CASE OF AUTODEFENSAS IN ME XICO

Table 8.1 Continued DV: vigilante

Model (1)

Model (2)

Model (3)

Model (4)

Population Log

1.375∗∗∗ (7.93)

1.419∗∗∗ (8.05)

1.365∗∗∗ (8.03)

1.411∗∗∗ (7.87)

Temporal lag

1.109∗∗∗ (4.13)

1.116∗∗∗ (4.18)

1.109∗∗∗ (4.12)

1.102∗∗∗ (4.13)

Spatial lag

0.615∗∗∗ (3.72)

0.622∗∗∗ (3.79)

0.610∗∗∗ (3.69)

0.608∗∗∗ (3.68)

_cons

−11.80∗∗∗ (−7.18)

−12.58∗∗∗ (−7.52)

−11.72∗∗∗ (−7.12)

−11.82∗∗∗ (−7.11)

lnsig2u

0.0476 (0.11)

−0.0409 (−0.09)

0.0533 (0.13)

0.0627 (0.15)

N

12,241

12,241

12,241

12,241

t statistics in parentheses ∗ p