Militias and the challenges of post-conflict peace: silencing the guns 9781350221390, 9781848135260, 9781848135277

Citation preview

Acknowledgements

First of all, we are grateful to the many people who shared their knowledge and ideas with us in Sudan, DRC, Afghanistan, Timor-Leste, London, Ottawa and Washington DC. Listing all the people who helped us along the way would be impossible. We are especially indebted to everyone we interviewed; thank you for your time, sincerity and honesty. Also, sincere gratitude to the numerous individuals at government agencies, international organizations, think tanks and NGOs who made our lives a bit easier during field work by providing us with muchneeded moral and logistical support. Thanks to all of our colleagues, peers and supporters who read various versions of the chapters and who provided critical insights. At the institutional level, we are grateful to the Norwegian Institute of International Affairs (NUPI) for commissioning this project. We are especially grateful to Andreas Vogt, Helene Revhaug and Stina Torjesen at NUPI for their support and encouragement. We also wish to thank the LSE and the American University in Cairo for providing additional financial support to complete the project. We are grateful to our editor, Ken Barlow, for shepherding the book to its conclusion. We would like to thank the anonymous reviewer for Zed Books for his or her enthusiasm for the text. Also thanks to the copy editor, Ewan Smith and his associates, for their editorial support. We would like to thank our families and friends who have supported and encouraged us along the way. CA, MT, MA London, Montréal, Bangkok

vi

Acronyms

AMF AMP ANA ANAP ANBP ANP ANSF AP3 CNDP

Afghan Militia Forces Alliance of Majority in Parliament Afghan National Army Afghan National Auxiliary Police Afghan New Beginnings Programme Afghan National Police Afghan National Security Forces Afghan Public Protection Programme Congrès national pour la défense du peuple; National Congress for the Defence of the People CNRT National Congress for Timorese Reconstruction CONADER Commission Nationale pour la Démobilisation et la Réinsertion; National Commission for Demobilization and Reintegration CPA Comprehensive Peace Agreement DDR disarmament, demobilization, reintegration DDRRR disarmament, demobilization, repatriation, reintegration or resettlement DIAG Disbandment of Illegal Armed Groups DRC Democratic Republic of Congo FALANTIL National Liberation Forces of an Independent East Timor FAR Forces armées rwandaises; Armed Forces of Rwanda FARC Revolutionary Armed Forces of Colombia FARDC Forces armées de la République Démocratique du Congo; Armed Force of the Democratic Republic of Congo FDLR Forces démocratiques pour la libération du Rwanda; Democratic Forces for the Liberation of Rwanda FDTL Timor-Leste Defence Force FOCA Force combattant abachunguzi; Force of Redeeming Fighters FRETILIN Revolutionary Front of Independent East Timor GoS Government of Sudan GoSS Government of South Sudan ICC International Criminal Court IDP internally displaced person vii

IDPPPH

Independent Directorate for the Protection of Public Properties and Highways by Tribal Support ISAF International Security Assistance Force ISF International Stabilization Force JIU Joint Integrated Units LDI Local Defence Initiative LRA Lord’s Resistance Army MDRP Multi-Country Demobilization and Reintegration Programme MLC Mouvement pour la libération du Congo; Movement for the Liberation of Congo MONUC United Nations Organization Mission in the Democratic Republic of Congo NATO North Atlantic Treaty Organization NCP National Congress Party NGO non-governmental organization NHDF Nile Hope and Development Forum PDF Pibor Defence Force PNTL National Police of Timor-Leste RCD-Goma Rassemblement congolais pour la démocratie–Goma; Rally for Congolese Democracy–Goma Renamo Mozambican National Resistance SAF Sudanese Armed Forces SMI Integrated Military Structure SPLA Sudan People’s Liberation Army SPLM Sudan People’s Liberation Movement SSDF South Sudan Defence Force SSR security sector reform UN United Nations UNDP United Nations Development Programme UNITA National Union for the Total Independence of Angola UNMIS United Nations Mission in the Sudan UNPOL UN Police

viii

Preface

Militias have been a feature of conflict for centuries. Although sometimes considered an archaic notion, the importance of militias to ­con­temporary war and peace is profound and prevalent throughout the world. While militias are most often associated with Africa and the governments of weak states, the US-led wars in Iraq and Afghan­istan have also seen militias play definitive roles, as highlighted most strikingly by the direct allying of the US military with localized Iraqi militias, the so-called Sons of Iraq. Overall, the diversity of locales, forms, motivations and relationships has meant that academia and the policy world have struggled to come to a clear understanding of what militias are, the roles that they play and why, and hence how best to respond to them. Given the prominence of militias in contemporary conflict and the international efforts to usher in peace and stability, there continues to be a driving need to further understand them. As such, the origin of this book was as a research project, ‘Reining in the real dogs of war: the demilitarization of armed groups and militias in African conflicts’, commissioned in 2006 by the Norwegian Institute of International ­Affairs (NUPI), and led by a team from the London School of Economics and Political Science, which sought to respond to this very need. The goal of the project was to push forward the idea that understanding the nature, motivations and conduct of militias is crucial to devising a strategic approach to demilitarizing these entities and, concurrently, fostering the conditions for long-term stability. This was achieved through a comprehensive study that actively theor­ ized and defined militias; provided an empirical mapping of the militias that emerged in post-conflict environments; assessed their relationship with internal and external actors; and developed and analysed strat­egies for demilitarization. Considering the diversity of militias and the challenges of responding to them, the project necessitated the development of a conceptual framework for deeper and more consistent analysis of militias. With this in mind, field research was undertaken in Sudan and the Democratic Republic of Congo (DRC), and using the process iteratively, a conceptual framework was developed and refined. Research was subsequently undertaken on Timor-Leste and Afghanistan. The ix

purpose of the book is to present the consolidated efforts of the original research project by describing the conceptual framework, showing its application through case studies, and concluding with broader suggestions, both for future academic research and policy formulation. Overall, the intention of the book is to make a serious contribution to a growing but still limited literature on militias and demilitarization. The key argument of the book, highlighted through empirical case studies, is that there are strong risks accompanying the imposition of security promotion interventions from above, particularly if they are divorced from the political, social and economic context in which such activities are inevitably embedded. The policy corollary to this is that internationally led forms of intervention – by definition, efforts led from above – are successful only insofar as they are able to frame their aims in terms which can translate into goals that are recognized by and meaningful to local actors. In this regard, integrating those significant factors which are derived from an understanding of local contexts and, concurrently, a convergence with local interests are key determinants of success. The case studies covered in this volume highlight the many contextual factors that shape these different undertakings: the alignment of regional interests and influences bearing on a disarmament, demobilization and reintegration (DDR) process supposedly taking place within a country; the political economy of war and peace distinctive to each conflict zone and cutting across formal borders; the unique characteristics of armed groups and their leadership; and the evolving global and normative environment in which DDR is conceived and carried out (Berdal and Ucko 2009). Through proactive fieldwork resulting in the case studies covered in this volume, it was revealed that the political, socio-economic and cultural circumstances in which DDR is undertaken are inextricably tied to the effectiveness of programme design and outcome. For example, this oversight is most obvious when the precise nature of the ‘security’ challenge, especially the character, motivation and operational mode of militias, is poorly understood by both national authorities and the international community of actors. Hence, this crucial element of understanding militias is central to our overarching argument. As well, ‘real’ qualitative, empirical differences in the historical, political, cultural and economic contexts exist among different ‘post-conflict’ states. Unfortunately, DDR programmes are too often approached in a reactive, ‘cookie-cutter’ fashion, and executed in a way that isolates the demilitarization process from the broader national processes linked to state consolidation, economic recovery, governance and politicization x

xi

Preface

of identity-based societal cleavages, as well as the alignment of regional interests and influences, the political economy of war, and the evolving global and normative environment in which DDR is conceptualized. As such, our argument, in line with other emerging thinkers on the matter, is that the political complexity of DDR must be recognized and placed at the centre of DDR programmes, and the continued application of universal DDR strategies will prove to be ineffective, if not dangerous, in these highly specific local contexts (see especially Muggah 2009; Berdal and Ucko 2009; and Torjesen 2009b). The outline of the book is as follows. Chapter 1 provides a general review of militias in contemporary world politics, showing their diversity in form and function. The chapter further articulates an introduction to demilitarization efforts aimed at militias. Chapter 2 is an elaboration of the conceptual framework that was developed by the authors to better analyse militias. This framework, used across the subsequent case ­studies, emphasizes the character, relationships, motivations and strategies of militias. As well, there is a discussion of the key issues that frame the local context, including historical, political and economic factors, international and regional dynamics, and the politicization of identity. This chapter is meant to be the common thread that binds the case studies, and allows for a much more nuanced comparison. The sequencing of the five subsequent case-study chapters reflects the idea of moving from a micro-level to a macro-level understanding of militias and the contextual challenges of demilitarization. This format highlights the diversity of militias between and within the case ­studies – ranging from paramilitary formations used as counter-insurgency forces, small localized village militias centred on self-defence, to mil­ itias that have spun off from failures within a state’s security sector and have subsequently sought to counter state power structures. As well as ­assorted objectives and structures, geographic diversity was also sought in the form of Afghanistan and Timor-Leste, to highlight militias as a phenomenon not limited to Africa. Applying a conceptual framework consistently to case studies allows us to identify key themes and lessons learned in the conclusion. The first three case studies focus on the analysis of individual militias in the ‘post-conflict’ context. Chapter 3 on the South Sudan Defence Force (SSDF) provides a discussion of the origins, key characteristics and operational mode of the militia, while also emphasizing the role of identity politics and relations with external and domestic actors in demilitarization efforts. In contrast to the state-sponsored SSDF, Chapter 4 explores small localized village militias, the so-called White

Army militias in Sudan. Local communities, the UN and indigenous NGOs all played significant roles in the demilitarization exercise, while contextual issues including the lack of economic redistribution and cross-border skirmishes remained as key challenges. Chapter 5 focuses on the ‘mutineers’ militia that emerged and flourished during the course of the 2006–08 political crisis in Timor-Leste. The chapter historicizes the evolution and prominence of this militia, while also discussing its impact on state-building efforts. From specific militias in the cases of Sudan and Timor-Leste, we move to examining various militias within a specific country. Chapter 6 explores the numerous militias that emerged in the DRC during the post-2003 period, placing special emphasis on: pro-government mi­l­ itias, the Mayi-Mayi; the Congrès national pour la défense du peuple (CNDP); and the Forces démocratiques pour la libération du Rwanda (FDLR). Contextual factors that had an impact on the efficacy of the demilitarization initiatives include a weak Congolese state and security apparatus, an ambitious and ineffective demilitarization programme which overlooked reintegration strategies, and identity politics at both the local and regional levels. Chapter 7 provides a historical understanding of the militias in Afghan­istan as well as a critical examination of demilitarization efforts since 2001. Furthermore, it concludes with an assessment of ongoing US efforts to ‘remilitarize’ local civilian actors as militias as part of its ­counter-insurgency campaign against the Taliban. Finally, the conclusion reinforces the notion that it is absolutely paramount for demilitar­ ization strategies to incorporate a rich understanding of militias and the local contexts where they operate. In addition, we discuss the implications of our findings for both the academic and policy communities, while also highlighting cross-cutting themes and the way forward. We should conclude by noting that as a longer-term research project this book presents in cumulative detail the work undertaken to date. In this regard, defining a conceptual framework for the improved analysis of militias has been the priority. Additionally, the application of such a framework to specific case studies has also been emphasized as a learning process. Needless to say, this of course still means there are wide openings for future analytical improvement. Perhaps the most obvious one, which this book provides insights and direction conducive to, is that of developing a detailed typology of militias allowing for more specificity of understanding, notably of the contexts in which they form and evolve, for better or worse. xii

1 | Introducing militias and demilitarization

One of the most serious problems confronting transitions from conflict to peace today is the role of militias. Whether armed by embattled governments in defence of their territory or fostered by external actors in the interests of greed or grievance, these non-statutory forces occupy an uncertain and deeply controversial position in the changing landscape of conflict. Linked variously to atrocities against civilians or international criminal elements, militias embody a new dimension in warfare that transcends the classic interstate and intra-state (government/guerrilla movement) disputes of the past. Part of what distinguishes them from more traditional combatants is their operational mode, especially their willingness to engage in violent tactics that defy international norms of conflict and their proclivity to embrace expediency in alliance-making. Moreover, as these militarized entities are prone to pursue conflict first and foremost in terms of local interests, militias are notoriously difficult to manage in the context of transitions from war to peace. As such, the conventional approaches to conflict management and resolution promulgated by the international community are singularly inadequate in addressing the issue of militias as well as the enduring effect that they have on post-conflict situations. In the first instance, unlike with most national armies or opposition guerrilla forces, local conditions can enable militias to resist formal demobilization and disarmament processes more readily. In part, this reflects the shifting balance of power within a given conflict, the a­ bsence of military command and control or distant territoriality from the central authorities that support these forces, and the influence of weak or even non-existent government administration over the conduct of war itself. Concurrently, the loosening of bonds between state instigators and other external supporters, often forged solely for reasons of mutual expediency, as conflicts drag on, is another aspect of militias which gives prominence to local factional politics. So too is the pull of commercial activity – sometimes initially embarked upon as a means of supporting military action while at other times nothing more than pure opportunism – which unleashes a dynamic of its own (the ‘greed’ cycle) which transcends any putative rationale for engaging in war. The 1

result is a highly complex conflict environment, subject to a multiplicity of sources that are as likely to be rooted in parochial concerns as they are in national issues or ethnicity, and one that is particularly immune to the established international templates aimed at interstate conflict resolution. The enduring impact of militias upon society is felt in a number of profound ways. In the first instance, the imprint upon peace negoti­ ations themselves – or even the failure to achieve more than ‘formal peace’ – is clear as militias are widely seen as potential ‘spoilers’ to any agreement. This can have the unintended effect of elevating the most disruptive armed groups to a standing whereby their participation is seen to be crucial to the success of any peace deal, and thus enhan­cing their bargaining power. Second, the long-term influence of militias can be seen in the flourishing arms trade, criminal activities such as ­narcotics, prostitution networks and illegal mining or timber trading that often sustained their role in the conflict and carry on after formal peace has been established. The impact of these two factors on the civilian population is clearly devastating, with low-level violence and mayhem continuing not only throughout peace negotiations but long afterwards, as has been demonstrated in a number of ‘post-conflict’ settings around the world, including Africa, the Middle East, Latin America and parts of Asia. Though mercenaries are often characterized as the ‘dogs of war’, surely it is these groups which have done more to earn that sobriquet. Transition from war to peace is, like all large-scale social change, complex and multifaceted. One crucial aspect of the transition is the demilitarization (which includes disarmament, demobilization, reinteg­ ration – DDR1) of former combatants. Through prodigious fieldwork and accompanying analysis, it becomes clearer that the efficacy of DDR is not only dependent on a thorough empirical analysis of the militias involved, but must also interface with a wide range of social, political and economic processes within which these armed actors are embedded. Understanding the nature, motivations and conduct of militias is crucial to devising a strategic approach to demilitarizing these entities and, concurrently, creating the conditions for long-term stability. In order to achieve this, a theoretical framework to address a number of key questions and issues was developed by the authors, aimed at identifying the motivations, social basis and operational mode of militia groups, as well as understanding their relationship with internal and external actors. Recognizing these contextual realities and integrating them into the substance of a DDR programme in a given post-conflict 2

Defining militias

Defining non-state actors chiefly by their independence from states and state authority would be extremely misleading, especially in the context of globalization. Increasingly, in both domestic and international politics, the theoretical purity of these supposedly ‘opposing’ idealized types is muddled by the complexities of practice ( Josselin and Wallace 2001: 2–5). However, for research manageability, non-state actors will be understood in terms of the definitional confines of those actors/­ movements that operate outside of the ‘formal’ state spheres (while also appreciating their discursive contexts and environments). As such, the focus will be on actors who are at least, ‘in principle, autonomous from the structure and machinery of the state’ (ibid.: 3). In particular, our focus will be on militias. Also, it should be noted that these non-state actors can exist at ‘state’/national, transnational/regionalized and individual levels (Davis 2009). In addition, the relations between the state and nonstate may be understood as ambiguous, clearly defined (encouraging or antipathetic) or indifferent, all existing along a continuum.2 These groups have a propensity to use violence for political and economic ends. In some contexts, they have a monopoly on the use of force/violence in a given territory or state. Also, the fact that these armed non-state actors operate outside the formal state sphere does not negate their potential to move into the formal state machinery and neither does it exclude interaction with units in the formal state system at any point of their operation. It is beyond the scope of this book to develop a comprehensive typo­ logy of armed non-state actors, but we feel it is imperative to discuss what terminology we use throughout our research. Our purpose here is not to create another nomenclature for the armed groups; instead we have used the ‘label’ of militias.3 Overall, like most concepts in international relations and security studies, that of militias is blurred at the edges. The word militia historically, notably in the context of American and British history, referred to a reserve body of citizens enrolled for military duty and called upon only in an emergency. However, in the contemporary period, the term has been used loosely in two consistent ways. First, it has been used to describe the private armed groups of 3

1 | Introducing militias and demilitarization

environment enhances the possibilities for the successful development of proactive strategies for demilitarization. At the heart of this process is a need to identify local security needs as understood by local actors in the context of their environment and ‘fix’ these into any post-conflict strategy. Failing to do so is likely to result – as events have too often shown – in a reversion to violence.

pro-regime strongmen and the paramilitary formations that organize in defence of the political order in a given country. Second, the term has also been used in connection with states where the central authority has been considerably weakened. In such cases, the armed formations established by warlords, tribal or regional strongmen, drug lords and the like are referred to as militias. For the purposes of this book, militias should be understood as a military force composed of civilians outside of a state’s formal military structure. The focus of this book is very much on understanding militias within a continuum of non-state armed groups in weak and failed states which resort to violence in pursuit of their objectives. These objectives can range from collective to private interests and can include challen­ ging the balance and structure of political and economic power as well as defending or controlling resources and territory for the benefit of a particular identity, including a state, or for the private accumulation needs of militia leaders. Critically, the line of continuity between a historical understanding of militias and that of contemporary militias is that they publicly explain and defend their armed presence by the need for ‘self-defence’ as understood by their own trajectory. While qualifying contemporary militias as non-state actors, it should be emphasized that the relationship between state and militia is still the definitive one. There are limited exceptions where ‘militias’ are still developed as official state institutions, and these should be understood more as historical anomalies, such as ongoing efforts in Venezuela to form ‘Bolivarian’ militias, and even as the foundation of a state’s national defence, such as the Swiss military in its militia structure. Rather, the vast majority of current militias should still be understood as non-state actors distinct from official state structures, even though their relationship with the state is imperative. Indeed, militias as paramilitary forces are a driving dynamic in contemporary international relations, especially in the context of weak and failed states, where militias provide an expediency of force for governments. Conversely, however, militias can also be understood in certain instances as insurgent forces as well. As described previously, the key point of militias is that they apply violence in pursuit of their respective objectives, including both challenging established state power structures and acting in the interests of particular identities, including those of a state. Historical examples of militias

As evidenced by the structure of the international system, much hope has been placed in the monopoly of ‘legitimate’ violence that states 4

5

1 | Introducing militias and demilitarization

are supposed to wield. However, the enduring reality has been that non-state armed groups, and specifically militias, continue to play a central role in ongoing global conflicts. Outside of the historical American and European experience with militias, which has dwindled, aside from the violence of the Balkans in the 1990s, the rest of the world has continued to see a very strong presence of militias and their centrality to conflicts. Indeed, it would be impossible to understand some of the most enduring conflicts of the past decades without appreciating the roles that militias have played in them. An examination of militias in the contemporary global context will be provided, after a brief discussion of the militias in the historical period. America’s history with militias goes as far back as the first European settlements, where settlers often benefited from no formal military presence and were largely left to defend themselves. However, militias mostly have a deep resonance in the USA because of their centrality to the American war for independence from British colonial rule. These ‘patriot militias’ were localized entities, drawing their commanders and militiamen from the communities where they were active under the premise of self-defence against British oppression. What is significant about this American understanding of militias was that their recruitment should be diverse and representative of local communities as well as focused on self-defence against imperialism.  Within this American context, what was important about militias as legitimate actors was that they were not ‘standing armies’ free to roam at will and freely detached from moderating relationships with local communities, i.e. the respective states of the nascent union. This meant that there was no ‘national militia’, since the understanding was very much that only individual states should form and maintain militias. Importantly, the militias were not subservient to the national military, but instead primarily subservient to local civilian control. Much of this sentiment stemmed from the resentment prior to independence of a heavy, domineering British military presence and lingering fears of a strong federal government assuming dictatorial powers over the states. Indeed, the Second Amendment of the US Constitution states this sentiment well: ‘A well regulated militia, being necessary to the security of a free state, the right of the people to keep and bear arms, shall not be infringed.’ However, this is not to say that militias were unequivocally supported as a perpetually benign force for good. A need was still felt to maintain minimal control over them – thus ‘well regulated’ – even by the federal government. Tellingly, the US president is not only ‘Commander in Chief’ of the national military, but also ‘of the

Militia of the several States’. Ultimately, with the gradual maturation of the USA under a stable federal government, the emphasis on state militias greatly waned. Militias as a national security priority were also heavily negated by the war of 1812. This was because it became evident that a strong national army and navy were imperative to battling the British once again, not as an insurgent force but as a fellow state. In contrast to the broad acceptance of militias as ‘positive’ actors by most Americans, at least in a historical sense, in other Western states there is a much more negative connotation to the term. Rather than as civilians acting in defence of their home communities, militias are instead perceived as predatory military organizations serving the nefar­ ious intentions of a state. Perhaps the best example of this was the French Militia, or Milice française, utilized by the Vichy regime against the French resistance movement during the Second World War. With strong support from Nazi Germany, these irregular forces were widely loathed for using torture and assassination against their fellow countrymen, as well as for their involvement in deporting French Jews to the Holocaust’s concentration camps. As a tool of the Vichy regime they were especially feared given that they were French and hence had an inherently better understanding of the resistance movement than did their German collaborators. While the French Militia was originally active only in Vichy France, as the war situation worsened for Germany and its French ally, they were also employed in occupied France, inclu­ ding Paris. Today the term milice in France still has exceedingly negative connotations, suitable as an invective. In contrast to France, and by example, the application of militias in Switzerland has been largely positive. With its long tradition as a neutral state, Switzerland bases its military on the ‘Swiss militia system’. Namely, there is mass conscription for every adult male, who is expected to serve in the military between the ages of twenty and thirty-four. This service is not continuous but rather undertaken in a series of progressive training programmes and short durations of active duty followed by inclusion in reserve forces. At any one time there are only approximately 140,000 active members of the Swiss military, but its overall size is approximately 220,000 men, with the reserves available for any national crisis. While there is a small professional corps within the military, approximating 5 per cent, what makes the Swiss militia system unique is the broad inclusion of the rotating masses of civilian ‘militia’ members. Tellingly, as part of what is essentially a national militia, all members of the Swiss military are required to keep weapons at their home and at the ready. From the strategic standpoint of a small neutral state, this 6

Contemporary examples of militias

The contemporary, post-Second World War period provides a plethora of examples of conflicts that involve a diversity of militias. Some of the most infamous militias of recent years have been the Janjaweed of Darfur, Sudan. The Janjaweed, roughly meaning ‘devils on horseback’, are militias composed of men from Arabic-speaking nomadic tribes best known for their involvement in the Darfur crisis starting in 2003. Their international notoriety derives from their central role in attacking nonArab, black Darfurian communities, mostly sedentary agricultural ones, and hence creating a massive displacement of several million civilians as well as 200,000 deaths (see Guha-Sapirm and Degomme 2005; de Waal 2007; and Flint 2009). The scale of the violence led some governments, notably that of the United States, to label it as genocide. Motivation for the violence undertaken by the Janjaweed ranges from simple land acquisition to the racist ideologies of Arab supremacy. In addition to their intensive attacks on civilians, the Janjaweed’s importance comes from their role as proxies for the Khartoum regime of Omar Hassan al-Bashir in its war against assorted Darfurian rebel groups, such as the Justice and Equality Movement and the Sudan Liberation Army. Along with several Janjaweed commanders, Bashir has been indicted, though not yet tried, for crimes against humanity and war crimes by the International Criminal Court for his alleged involvement in the Darfur violence. Beyond the extreme violence perpetuated by the Janjaweed in Darfur, it is vital to consider the militias involved in the 1994 Rwandan genocide. Following the assassination of Rwandan president Juvenal Habyarimana and Burundian president Cyprien Ntaryamira on 6 April 1994, the government, led by Hutu nationalist Théoneste Bagosora, initiated a genocide which lasted 100 days and resulted in the loss of 800,000 Tutsi and moderate Hutu lives and well over 1.8 million refugees (Prunier 1995: 312). The Rwandan national army, the Forces armées rwandaises (FAR), relied heavily on the well-organized Hutu militia known as the interahamwe (‘those who work/fight together’) and the impuzamugambi (‘those who have the same goal’), who began systematic killings of Tutsis and moderate Hutus. The militias had been organized 7

1 | Introducing militias and demilitarization

militia system is of vast utility as it negates the need to finance a large standing army, but conversely allows for a significant military force to be available should it be needed in a national emergency. Furthermore, the Swiss militia system has been enduring within the national cultural context as it places emphasis on mitigating gaps between the general public and the military.

months before the genocide began and were provided with government identification rosters with names and addresses of Tutsi families to facilitate the genocide, while the main Kigali radio station encouraged civilian Hutus to carry out massacres to eradicate the Tutsi population. By July 1994, the Rwandan Patriot Front (RPF) led by Paul Kagame had defeated the FAR and the extremist Hutu militias (known as génocidaires), many of whom fled to the Kivu provinces of Zaire/DRC and who still remain there. This massive relocation of people exported the Rwandan conflict to the eastern DRC, and the broader Hutu–Tutsi ‘exclusionary’ political calculus, in which accession to power by one group was automatically considered by the other as the precursor to their own destruction, took a grip of Congolese politics. This led not only to the downfall of the Mobutu and Kabila regimes, but the active involvement of Rwanda and Uganda in arming local Congolese ethnic militias to contain the threat posed by the génocidaires. The persistence of militias in the Congolese context is reflected by the conflict between the Forces démocratiques pour la libération du Rwanda (FDLR), a group consisting of former FAR, interahamwe and young Congolese men; the Mayi-Mayi, local defence groups; and Nkunda’s Congrès national pour la défense du peuple (CNDP), whose overarching ‘public’ purpose was to protect marginalized Tutsis. The Lord’s Resistance Army (LRA), a non-state armed group not easily labelled but perhaps best understood as a cultish militia, has wreaked extreme violence across central Africa for over twenty-five years.4 While it was initially concentrated in northern Uganda, over the years its area of operations has evolved considerably. During much of the 1990s and early 2000s it was active in southern Sudan, serving as a proxy force for Khartoum against the Sudan People’s Liberation Army (SPLA) insurgency. With the peace process that began in 2005 in Sudan, the LRA was increasingly pushed out of Sudan, prevented from returning to Uganda by a strong army presence, and instead started roving through north-eastern DRC, the Central African Republic and even appearing as far north as Darfur. With a thoroughly debauched ideology centred on violence, the LRA has continued to attack civilian populations wherever it has moved, causing mass casualties. The LRA’s leader, Joseph Kony, and his deputy have been charged with crimes against humanity and war crimes for their group’s long history of violence by the International Criminal Court, though they have yet to be apprehended and put on trial. Within the general context of anarchy that defines so much of central Africa, there has been a general phenomenon of localized self-defence militias sprouting up in the LRA’s wake. 8

9

1 | Introducing militias and demilitarization

Within the Middle East, militias have played a strong and enduring role in its recent history. Perhaps the best example of this is Hezbollah, which started out as a small Shiite Muslim militia in southern Lebanon. Catalysed by the Israeli occupation, which began in 1982 and ended in 2000, Hezbollah grew in size and scope to become an exceedingly wellarmed group as well as a major political party and social services provider for local communities. As a proxy force for Syria, and with extensive support from Iran’s Revolutionary Guards, Hezbollah has been able to acquire a military arsenal more befitting of a state. Notably, Hezbollah has an extensive array of rockets that allow it to target northern Israel, something highlighted by its 2006 war with Israel when northern Israeli cities were consistently barraged (International Crisis Group 2007a). This war, with no clearly defined victor, was exceptional as it pitted one of the world’s most sophisticated militaries against a non-state armed group. Even aside from Hezbollah, the importance of militias to Lebanon’s modern history has been profound. During the civil war from 1975 to 1990, militias from the assorted national communities, Sunni Muslim, Shiite Muslim, Maronites and Druze, as well as Palestinian refugee militias, all played driving roles in the conflict, most often working as proxies for neighbouring countries (see Ellis 1999 and Zahara 2002). The largest and most prominent formation of militias in the Middle East has occurred in Iraq since the US-led invasion in March 2003. Especially significant among the plethora of Iraqi militias have been the ‘Sunni Awakening Councils’, also known as the ‘Sons of Iraq’ militias. These militias, numbering from approximately 80,000 to 100,000 militiamen, began forming in late 2005. Composed primarily of members of Sunni communities and with support from the US military, they were primarily a response to extreme violence perpetuated by Shiite militants, such as the now disintegrated Mahdi Army militia, as well as al-Qaeda terrorists (International Crisis Group 2008a). Many of these Sunni militiamen had themselves been insurgents attacking US forces, but given the more existential threats faced by their communities in 2005 and 2006, they tactically allied themselves with the USA. With the establishment of the Sunni Awakening Councils, the dynamics of the insurgency in Iraq changed significantly as al-Qaeda was left increasingly isolated and overall levels of violence decreased. However, given their sheer numbers, the challenge of managing these militias has proved problematic. This has notably been because it has caused tensions with the Iraqi government, which has long publicly feared that the militias could potentially emerge in the future as a large, well-armed opposition. Furthermore, despite stated intentions by the US military that these

militias would gradually be subsumed into the formal security forces or be demobilized, the process has proved slow and the militias still exist largely free of governmental control. A number of other militias have been prominent to the recent history of Iraq. The Peshmarga is a large, well-organized militia operating in the Kurdish areas of northern Iraq. Within the wider regional context of Kurdish nationalism, Peshmarga militias have existed since the early twentieth century and have been involved in various Kurdish struggles, fighting for secessionism or autonomy. Inside Iraq, the Peshmarga for decades resisted the military forces of the Ba’athist regime of Saddam Hussein under the political leadership of the Kurdish Democratic Party and the Patriotic Union of Kurdistan. During the 2003 US invasion of Iraq, the Peshmarga allied themselves with the US military, furthering a relationship that stretched back to the first Gulf War in 1991. Since 2003, the Peshmarga, with forces estimated at a hundred thousand (see Beehner 2006; Assad 2008; Kakeyi 2010), have exerted official control over Kurdish areas and maintain autonomy from the national Iraqi military. Another critically important militia in Iraq’s recent history has been the Mahdi Army. This large militia was composed of members of the Shiite Muslim community and led by the radical cleric Muqtada al-Sadr. The Mahdi Army was the first Shiite group to attack US forces, and it played a central role in the manic sectarian violence that inflamed Iraq from 2004 to 2008. Following an extensive crackdown on its strongholds of Basra and Sadr City in Baghdad by US and Iraqi forces in mid-2008, the Mahdi Army announced it would disarm and pursue its goals primarily through the political process. This has largely transpired as the ‘Sadrist Movement’ emerged as a national political actor through the 2010 elections. Within Asia, perhaps the most prominent episode involving militias in recent history was the mass violence that unfolded in East Timor (now known as Timor-Leste) in 1999. Following the Timorese public’s rejection of proposed autonomy within Indonesia in a referendum, pro-Indonesian militias supported by the Indonesian military instigated violence across the territory in September 1999. This violence killed approximately 1,300 Timorese, forced over 250,000 to flee into neighbouring West Timor (US Department of State 2010), and resulted in much of the territory’s infrastructure being destroyed. As a result, an international peacekeeping force led by Australia was deployed and, following a number of confrontations with the peacekeepers, the militias were forced to retreat into West Timor themselves and the security situation stabilized. Throughout the Indonesian occupation, which began in 1975, the Indonesian army had 10

11

1 | Introducing militias and demilitarization

utilized Timorese militias as proxies to stifle secessionist sentiments, but especially so in the late 1990s as momentum for independence developed. There was strong involvement by militias in a number of notorious incidents of brutality towards civilians, such as the Liquiçá Church Massacre, in which dozens were killed. Elsewhere in Asia, militias have at times played strong roles in ­national politics and have even provoked changes to national governments. In the Solomon Islands, several militias representing assorted ethnic groups engaged in extended violence starting in the late 1990s. Much of the tension stemmed from the settlement of islanders from elsewhere on the archipelago upon the island of Guadalcanal. Here a militia was established, the Isatabu Freedom Movement, to foment intimidation against Malaitan settlers, who responded by forming their own militia, the Malaita Eagle Force. Within the context of escalating tensions between resident communities and their respective militias, further conflict erupted between these militias and the Solomon Islands government itself. Events escalated to the point where the prime minister, Barth­olomew Ulufa’alu, was briefly taken hostage by the Malaita Eagle Force and released only when he agreed to resign. Despite several peace agreements between the conflicted parties, v­iolence involving the militias continued, exacerbated by both rising crime and deteriorating governance as the political situation worsened. In response to the perpetuation of the crisis, an international peacekeeping force led by Australia and composed of 2,200 soldiers and police officers arrived in July 2003. In Southeast Asia, the Philippines has had the most endemic militia presence in recent times. The country has a long history of insurgency and sectarianism, and militias have been a common phenomenon, especially on the violence-plagued island of Mindanao. In response to multiple Muslim and communist insurgencies, the Philippines government, starting in the mid-1980s, encouraged militias to form and fight against insurgents, and the policy has continued to the present. Commonly referred to as ‘private armies’ within the Philippines, these militias were categorized either as ‘Citizens Armed Forces Geographical Units’ controlled by the military, ‘Civilian Volunteers Organizations’ controlled by the police, or as being without any ostensible control by the state but purely private affairs maintained by businessmen and politicians from powerful families. Regardless of the emphasis publicly placed on the rationale of community self-defence and support for state security forces, the militias have often entrenched themselves as violent tools for specific clans and political interests. The most notorious incident, though certainly not an isolated one, o ­ ccurred in November

2009 when fifty-seven civilians were massacred in the Maguindanao region of Mindanao. The violence was undertaken by a militia belonging to the Ampatuan clan in a dispute over upcoming elections with the rival clan of the Mangudadatus. The event gained notoriety because of extensive patronage links between the Ampatuan’s leading family and the national government of Gloria Macapagal-Arroyo. Another Asian country with a very strong presence of militias is Pakistan, and, as in the Philippines, they have been used by the government as proxies against strong insurgent movements. This has notably been in the country’s north-west, the Northwest Frontiers Province and the Federally Administered Tribal Areas, against the Pakistan Taliban and other militant groups increasingly involved in attacks against the Pakis­ tan government, particularly al-Qaeda terrorists. Starting in mid-2008 in the Bajaur Agency, the Pakistani military started to support lashkars, or tribal militias, from the resident Pashtun communities. In exchange for financial aid and small arms, these tribal militias have been expected to deny insurgents sanctuary in their home areas, particularly after the Pakistani army has conducted offensive ‘clearing’ operations in a particular area. The use of lashkars has been controversial, with many arguing that it is counterproductive in the longer term as many of these lashkars are in fact merely recirculated insurgents themselves, including Pakistan Taliban members, manoeuvring for short-term tactical advantages against local rivals. As such, critics have highlighted that efforts to form local ceasefires with militants masquerading as tribal militias have most often resulted in failure as the agreements have been routinely broken, the fighters involved simply returning to fight the government as well as continuing to harbour foreign terrorists. Lastly, critics have argued that facilitating lashkars simply adds yet more armed groups to an already exceedingly complex security situation, further compounding the challenges for civilians caught in the middle. In Latin America, militias have also been endemic, especially in the war-plagued Central America of the 1980s. More recently, the most in­ famous ones have been the right-wing paramilitary militias of Colombia, which came to prominence as the cocaine industry burgeoned in the 1990s, alongside the strengthening insurgency of the Revolutionary Armed Forces of Colombia (FARC). While much emphasis was publicly placed by these right-wing militias on their origins as local self-defence organizations provoked by the predations of leftist insurgents, most notably the FARC, they had strong links to the cocaine cartels, pro­ viding protection for their production areas and transport networks (Ortiz 2002). Additionally, it was long alleged that the Colombian 12

13

1 | Introducing militias and demilitarization

­ overnment had maintained supportive links to the paramilitary mil­ g itias, considering them as useful tools in their own campaign against leftist insurgents. The largest was the United Self-Defence Forces of Colombia (AUC), which was an umbrella organization formed in 1997 and claiming to represent up to 20,000 militiamen. Given its strong ties to the cocaine cartels and a number of large massacres of civilians, the AUC was labelled as a terrorist organization by both the USA and the EU. Beginning in the mid-2000s, the government implemented a comprehensive programme that was largely successful in demilitarizing the paramilitary militias, including the AUC, by offering amnesty and reintegration programmes to former combatants.5 However, the process has since been criticized for leaving gross human rights violations unpunished, and fears remain that many of the formal militias merely morphed into more nebulous criminal networks. In Venezuela, a unique application of militias has ­occurred recently. As part of his ‘Bolivarian revolution’, President Hugo Chávez in the mid2000s started to create a reserve force for the military by arming large numbers of civilians to form local militias. This national programme, known as the Bolivarian Militia, publicly claims to empower the general population to be able to resist foreign intervention, namely that expected from the USA. With 100,000 AK-47s purchased for its members, Chávez has argued that the militia is necessary to resist US ‘imperialism’, a ­ ggression from US-allied neighbour Colombia, and attempts by domestic ‘bourgeoisie’ to overthrow the government. Composed of such units as ‘worker and peasant battalions’, it is seen by its domestic critics, however, as a blatant attempt by Chávez to arm his own supporters from among the general public, thereby posing a serious challenge to Venezuela’s future as a democracy. Despite such concerns, militias as a national Venezuelan institution seem to be entrenched, so much so that 13 April, the day that Chávez returned to power following a short coup in 2002, is celebrated as a ‘Day of the Militias, the Armed People and the April Revolution’. Within Europe, the presence of militias in recent times has been largely non-existent with the unfortunate exceptions of the Balkans wars of the 1990s. During the Second World War militias were common throughout the continent but were especially prominent in Yugo­slavia. Notable were the Chetniks, Serb nationalist militias that collaborated with Nazi Germany and were responsible for extensive war crimes against Bosniaks, Croats and other ethnic minority groups. The Chetniks had existed since the beginning of the twentieth century, fighting during successive wars as royalists and Serb nationalists, but were banned after the Second

World War by the new government of Josip Tito, the partisans’ leader. When the fragmentation of Yugoslavia began in 1990, Serb nationalist militias, such as the Scorpions, Yellow Wasps, White Eagles and Arkan’s Tigers, quickly re-emerged as major armed actors, and were often under the control of Serbian nationalist political parties. These Serb militias often claimed lineage with the Chetniks and were extensively involved in major war crimes alongside more formal militaries, namely the Army of Republika Srpska, including the notorious massacre at Srebrenica. While Serbian militias were the most numerous and prominent in the wars, there were also nationalist militias from among the other ethnic groups involved in the extended conflict, such as the Croats’ Croatian Defence Forces and the Bosnians’ Patriotic League and Green Berets. Understanding demilitarization

A common response to address the challenges posed by militias has been demilitarization. Demilitarization is based on the notion that combatants are potential spoilers of the peace process, and dangerous ­elements to interject into a post-conflict environment. Combatants have the capacity to disrupt and undermine the peace process, either by returning to hostilities with their opponents or resorting to armed banditry (see Spear 2002; Berdal 1996; and Alden 2002). Therefore, recognizing that this volatility can be alleviated through targeted programmes, demilitarization is a process which not only focuses on structured disarm­ ament and demobilization of combatants, but also provides monetary incentives designed to facilitate reintegration into civilian life (see United Nations 2006c). Disarmament should achieve the removal of the means by which civil wars have been executed and lead to the creation of a stable environment, thus strengthening confidence- and securitybuilding among combatants, the government and the population. The form and extent of disarmament in a peace agreement need to be decided with reference to the norms of society under consideration, rather than applying a ‘standard model’. As well, emphasis must be placed on demobilization – as peace requires breaking the command-and-control structures under which rebel fighters operate, and making it difficult to return to organized armed rebellion. Reintegration and reinsertion activities are of paramount importance as they determine whether the demilitarization will be successful and effective in preventing future conflicts. Reintegration is the process by which former combatants acquire ‘civilian’ status and engage in sustainable employment and income-generating activities, thereby severing the dependency they had on military support networks during the conflict. Reintegration support 14

15

1 | Introducing militias and demilitarization

programmes contain elements such as: cash payments (in instalments); foodstuffs (or coupons); healthcare; civilian clothing; job placement; land distribution; credit schemes; training; counselling; etc. Although the current focus of DDR programmes has been on economic reintegration, effective reintegration must envision former combatants engaged in social and political reintegration activities as well. It is also important to keep in mind that with more integrated peace­ keeping operations, reintegration has involved security sector reform (SSR) initiatives, as former combatants are given the choice as to whether they wish to rejoin civilian life or to reintegrate into the national army. SSR is an especially important process, which has been one of the main ways of dealing with militias. For example, the incorporation of a major faction of the SSDF through an agreement with the SPLA was the start of the broader SSR process to rationalize the security sector of South Sudan in 2005. In the DRC, DDR processes were implemented simultaneously with SSR efforts. Even in the case of Afghanistan, the SSR process began by consolidating the militias of the Northern Alliance into the more formalized ‘Afghan Militia Forces’, which were then gradually transitioned into the Afghan National Army (ANA) and the Afghan National Police (ANP). As such, it must be asserted that any analysis of militias and demilitarization must emphasize SSR. Demilitarization is a central pillar of most large-scale peace operations, and its significance as an important precondition for post-conflict peace and stability has been internalized by most international actors, including the United Nations, regional bodies, bilateral donors, NGOs and states. At its core, demilitarization is meant to ensure that militias and other armed groups that were active during the conflict do not return to the battlefield or find other ways of undermining local and international efforts to build lasting peace; and to do so by reintegrating ex-combatants into the social, economic and political life of post-war society (Berdal and Ucko 2009). Much of the DDR literature generated mainly by UN agencies, governments and NGOs has been prescriptive and policy-oriented in character, focusing on those engaged in the delivery of DDR programmes and concerned with the identification of specific and transferable lessons from the experiences of numerous DDR processes since the mid-1990s. The literature highlights the ‘mechanics’ of DDR activities, including planning, organization, coordination and funding. The discussion of the political, economic and social context and process tends to be perfunctory (ibid.: 5). This is not to suggest that highlighting lessons learned is not important; rather, it is absolutely necessary in our understanding

of DDR, as the lessons learned from ‘successes’, such as Mozambique, El Salvador, Bosnia-Herzegovina and Liberia, and ‘problematic’ cases, such as Angola, Sri Lanka, Sierra Leone and Lebanon, reveal. The key concern is that the literature focuses on the technical and prescriptive aspects of DDR, with limited attention paid to fostering a more critical and comprehensive approach to demilitarization. It is important in any analysis of the DDR process to understand how this oversight finds its way into policy-making. The United ­Nations’ ‘Integrated DDR Standards’, although still concentrating on the nuts and bolts of DDR activities, do recognize the dangers of divorcing DDR programmes from their political context (in particular the need for ­social and economic reintegration), which is a valuable contribution in this regard. Edited volumes by Berdal/Ucko and Muggah take a critical ­approach on DDR and SSR. Berdal and Ucko argue that there are ‘real limitations and perhaps dangers, of applying universal DDR strategies in highly specific local conditions’, and that ‘this simple, seemingly straightforward conclusion should not detract from its fundamental importance’ (ibid.: 8; see also Torjesen and MacFarlane 2009). Muggah’s key critique focuses on questioning whether demilitarization and sec­ urity sector reform activities actually ‘work’ or not, as the outcomes have not necessarily been empirically tested or interrogated (Muggah 2009). Insights from the critical political economy perspective reveal that DDR fundamentally embodies the strategic and bureaucratic priorities of the security and development sectors. As such, it perpetuates the discourse and policy priorities of international donors and power-holding local elites. Similar to the project of ‘development’, DDR and SSR ‘can be seen as expressions of neo-liberal forms of power and governmentability’ (Colletta and Muggah 2009; see also Duffield 2007). In other words, DDR does not emerge spontaneously ‘from below’ but is part of a broader project of security – the legitimate control of force from above (Muggah 2009: 2). Demilitarization and related programmes are multidimensional processes, although they are frequently conceived narrowly as technical interventions by a constellation of international agencies and national governments. However, as both Berdal/Ucko and Muggah acknowledge, a critical discourse on demilitarization is emerging. Demilitarization in practice

The development of a basic framework for demilitarization came in response to the pressing need to tackle the challenges posed by the winding down of the Cold War in regional conflicts. Fitted around the evolving practices of the international community through the newly 16

17

1 | Introducing militias and demilitarization

empowered UN, the triad of disarmament, demobilization and reintegration became the core activities upon which this process was built. The use of UN peacekeeping troops as ceasefire monitors, the primary role assigned to them during the thirteen original peacekeeping operations between 1956 and 1988, was expanded to include joint civil–military observer teams overseeing the disarmament and demobilization process in pre-designated cantonment areas. UN police observers were later introduced to assist, along with the task of supporting the reinstatement of law and order, in the monitoring of compliance outside of the cantonment sites. Linking this ‘second-generation’ peacekeeping to provisions for general elections – which came to be the end point of UN peacekeeping operations – placed additional pressure on the UN to ensure that the disarmament and demobilization components were complete in advance of voting. While a small percentage of combatants were reintegrated into newly outfitted national armies, long-term concerns about creating conditions for stability, especially in the wake of the demobilization of tens of thousands of combatants into societies with few economic opportunities, compelled the international community to devise reinsertion programmes. This complex set of overlapping processes and programmes, organized and managed by a coalition of international actors and – formally or, in some cases, at least nominally – under UN auspices, gradually assumed the position of a template for addressing demilitarization in all post-conflict settings. The application of this comprehensive approach to DDR was, on the surface at least, initially seen to be a success. The UN Transition Assistance Group (UNTAG) in Namibia, the UN Angola Verification Mission I (UNAVEM I) and the UN Transitional Authority in Cambodia (UNTAC) in the late 1980s and early 1990s all seemed, despite some ominous signs, to run their course and produce relatively peaceful outcomes. Even the most optimistic interpretations of these missions had to recognize, however, that getting genuine compliance from parties to the conflict proved to be extremely difficult. The incentives for militarized organizations such as the Khmer Rouge, the National Union for the Total Independence of Angola (UNITA), the Mozambican National Resistance (Renamo) and others – as well as the governments in those countries – to engage in some form of recidivist behaviour were high, especially as they understood their bargaining position would be severely reduced once arms were removed and soldiers dispersed. It was only a matter of time before the tenuous fiction of UN control over DDR in one of the peacekeeping missions would dissolve into catastrophe.6 Though much is made of lessons learned through analysis of

s­ uccesses, in fact there is a cogent argument for saying that failure has been a more effective teacher. In the case of DDR, the near-collapse of the process in the wake of Khmer Rouge resistance to UNTAC pressure to participate in the process in Cambodia, and the brutal suppression of Swapo (South West Africa People’s Organization) infiltrations into Namibia under the helpless gaze of UNTAG, demonstrated the inherent weaknesses in the international community’s approach to DDR in postconflict settings. These severe shortcomings were to be exposed in full in Angola under UNAVEM II when the UNITA rebel movement, which had kept thousands of troops outside of the DDR process, responded to electoral failure in November 1991 by reigniting the civil war. One direct impact was the decision to extend the UN mission in Mozambique by another year and to redouble efforts to complete demobilization in advance of elections, as well as provide a well-funded reinsertion (then called reintegration) programme for ex-combatants. In the case of the 1992 UN mission to Somalia, despite efforts to introduce more robust forms of disarmament, no significant movement towards stability occurred, and eventually the international community abandoned its efforts in 1995. At the same time, the abject failure of the UN to react to the rearming of militias in Rwanda in early 1994 paved the way for genocide in that country. This, of course, contributed to the proliferation of militia across the eastern provinces of the DRC, and foreign intervention and instability continue to rock the region. Conclusion

The confluence of scholarly concerns and policy imperatives is evid­ ent in any study of militias and the process through which the inter­ national community hopes to subdue them. Our intention in producing this book is to contribute to that growing body of scholarly literature aimed at critiquing the DDR process. Our analysis of the empirical case studies that follow reveals the genuine risks accompanying the imposition of security-seeking interventions from above, particularly if they are divorced from the local political, social and economic context in which such activities are inevitably embedded. Accounting for and documenting the negative impact that rigid applications of scholarly approaches to managing the militias phenomenon has on attaining stability in postconflict situations is not enough. We believe that the guiding themes which frame our analysis of the empirical studies provide useful markers for scholars and policy-makers alike as to which realistic DDR inter­ ventions can be assembled and ultimately implemented. Therefore, in the next chapter we start out by developing a framework for analysis of 18

19

1 | Introducing militias and demilitarization

militias, the factors that drive their conduct and the role of local con­texts in setting their course of action. This scholarly review of the underlying principles which shape our understanding of militias and ultimately make their way into the policy process is a necessary corrective step to improving academic and policy responses to militias.

2 | Conceptualizing militias: a framework for analysis

Transition from war to peace is, like all large-scale social change, complex and multifaceted. As noted in the previous chapter, one crucial aspect of the transition is the demilitarization (which includes disarm­ ament, demobilization, reintegration – i.e. DDR1) of former combatants. The efficacy of demilitarization initiatives rests on two crucial pillars: understanding the motivations and character of the various militias involved; and recognizing that the contextual realities determine how the DDR programme will play out in a disaggregated post-conflict environment. Therefore, a rigid, ‘cookie-cutter’ approach to demilitarization (and security sector reform), developed and implemented mainly by the international community, is ineffective in meeting the myriad of challenges within the specific contexts of individual countries. As such, demilitarization initiatives must explicitly place these multilayered contexts at the heart of the approach itself. Failing to do so is likely to result – as events have too often shown – in a reversion to violence. This chapter aims to provide an interpretive framework for understanding and contextualizing militias. The framework draws upon the range of country-specific case studies examined in the following chapters, and will draw attention to the context and complexities of demilit­ arization efforts; concluding that the local context does indeed matter when devising strategies for demilitarization and security sector reform.2 As such, this framework of analysis provides a conceptual framework for understanding militias, as well as highlighting key political, economic and social contextual processes that texture the demilitarization (and overall peace) process. The conceptual framework will entail emphasis on the ‘empirical mapping’ of specific militias, which includes an analysis of militia motivations, organizational structure, strategies, relations with other actors, and the key issues framing the local context, such as the politicization of identity, the role of the international community and civil society organizations, exploitation of natural resources, and regional dynamics. In line with the emerging literature on contextualizing demilitarization, our analysis of militias operating in conflicts in DRC, Sudan, Timor-Leste and Afghanistan has led us to conclude that 20

From spoilers to militias

Any discussion of armed groups and militias must be prefaced by an acknowledgement of the work on ‘spoilers’ in peace processes. The spoiler concept developed as a descriptor of a selected set of actors within the peace process whose conduct was obstructive to its realization. Specifically, Steve Stedman coined the phrase in 1997 and defined spoilers as ‘leaders and parties who believe that peace emerging from negotiations threatens their power, world view and interests and use violence to undermine attempts to achieve it’ (Stedman 1997). The context within which the notion of spoilers arose was the nexus between postCold War efforts to resolve conflicts (Cambodia, Mozambique, Angola), which initially were thought to be relatively easy – turn off the tap of international support and they would disappear – and the rise of intrastate conflicts, especially those linked to state failure. Cases as varied as Angola (UNITA post-1992) and Somalia (1992) posed challenges to the established views on conflict management and resolution. He diagnosed a set of four problems associated with these actors which influence the degree of impact that they have on the peace process. These are position (inside/outside the peace process), number (of spoilers), type (limited, greedy and total) and locus (power base). Stedman developed a typology of motivation for potential spoilers and a concurrent set of approaches to managing them: ‘limited’, whose interests could be met by adjusting the peace process (inducement); ‘greedy’, who are opportunistic in their approach and can be addressed by acceding to their demands or adjusting aspects of the peace process (socialization); and ‘total’, whose aims are unalterable and therefore can only be marginalized or suppressed through an active international strategy of forceful action (coercion). His approach reflected discussions in academic literature emerging from the late 1980s onwards by William Zartman and others as the point at which conflict resolution is said to be ‘ripe’, experiencing a ‘hurting stalemate’ and thus able to move from a ‘zero-sum game’ approach to a ‘non-zero-sum game’ (Zartman 1989). Moreover, the academic literature on conflict had begun to shift to an emphasis on the overriding economic sources of sustained 21

2 | Conceptualizing militias

in order to develop and undertake effective demilitarization activities and foster long-term peace in a post-conflict environment it is essential to: i) understand the armed groups that can potentially destabilize (and ‘spoil’) the peace process; ii) recognize that the local context is paramount to devising effective demilitarization policies, and therefore must be essential to (and constitutive of) any peace initiative.

conflict, with the rise of the notion of Collier’s ‘greed and grievance’ school and others. It is against this broader context that the focus of analysts has in­ creasingly turned to outliers in the peace process. While parties to ­negotiations aimed at developing and implementing a peace agreement have received attention, increasingly the emphasis has shifted to under­standing local actors formally outside of the peace process. In this context, to a great extent the scholarly debate as to what is an ‘armed group’ or ‘militia’ has been refracted through three basic approaches, namely ideological, behaviourist and systems. In all these instances, underlying normative assumptions have coloured analytical work as well as introducing unintentional biases that have systematically distorted understanding of this phenomenon. In particular, each approach has sought to use its taxonomy as an evaluative tool in determining recognition for a given militia. The attendant implications for policy implementation are self-evident. The ideological approach has traditionally focused primarily on selfdeclared identity and aims of militias and their relationship to the dominant state actors, namely the United States and the Soviet Union, during the Cold War period. The link between the external environment and the identity of a given organization was a crucial determinant of access to support in the context of the ideological conflict. It was thus that from the late forties onwards armed groups sought to situate themselves in terms of this ideological paradigm as a strategy for mobilizing foreign support for local political struggles, or in Jean-François Bayart’s phrase, ‘extraversion’ (Bayart 2000). Though experience too often cast doubt upon the commitment of leaders and armed groups – from Ho Chi Minh’s spurned bid for American assistance against French colonialism to Jonas Savimbi’s shifting allegiances to China and the United States – to the Western or Eastern cause and consequently the efficacy of this interpretative approach, nonetheless ideology remains an important barometer (and has enjoyed a revival in the post-9/11 environment) in assessing and defining militias. One of the key dimensions of this support from the superpowers, often forgotten in contemporary studies of militias, was a tacit norm on their part to retain intact the existing structure of sovereign states. For this reason, militias and armed groups that articulated separatist agendas were actively shunned by leading states while predatory or criminally oriented groups without aspirations to political authority were regarded with suspicion and consigned to the outer realms of tactical opportunism by the superpowers. The collective surprise that has 22

23

2 | Conceptualizing militias

gripped academics at the apparent rise of these entities in the aftermath of the ending of the forty-year ideological conflict is a testimonial to their ignorance of this key tenet of bipolar conflict and the impact it had on shaping disputes, actors and outcomes (see Kaldor 2006). The current multipolar environment provides opportunities for the introduction of new discourses or tactical use of ideology – be it ethnic or religious in form – by various actors engaged in conflict. The behaviourist approach, which Paul Collier and Anke Hoeffler have been at the forefront of articulating, has posited a motivational orientation to interpreting militias. Economic drivers for conflict have been introduced into analysis of these groups and linked with ideational motivations, giving rise to a ‘greed and grievance’ school.3 Recognizing the interplay between economic sources of conflict, both in the form of relative deprivation and economic predation, the behaviourist school sought to analyse through their research which of these conditions for rebellion spurs on further rounds of exploitation and conflict (‘bootstrapping’) and under what circumstances. Their key finding is that ‘greed’ in the form of revenue possibilities coupled with organizational capacity provides the most significant indicators of rebellion within a given context (Collier and Hoeffler 1998: 144–5). Conversely, grievances, ethnic factionalism and other putative proximate causes are not found to be important in predicting civil war. This work, it should be noted, builds implicitly upon Ted Gurr’s political psychology of rational choice in assessing the causes of rebellion but attempts to frame the question of motivation in terms of group dynamics of choice as they relate to economic possibilities within a particular sovereign territory. This is discussed further in the theoretical framework. Gurr and others focus on grievances brought on by discrimination against ethnic minorities as producing political mobilization against the state, which can ultimately cascade into open rebellion (Gurr 1993; see also Gurr 1970). Significantly, the behaviourist approach does away with the Cold War-era position regarding the necessity of being a sovereign claimant as a sine qua non of international recognition, reflecting in part the changing global environment but also the demise of ideology as a moral compass for asserting legitimacy. In effect it ascribes no set criteria for recognition of a militia beyond that of success in having the capacity to maintain sustained presence based upon opportunistic exploitation of territorially based resources. By way of contrast, the systems approach places the emphasis on militias as social organizations and seeks to define and assess them in terms of their internal dynamics of control. Championed by P ­ ablo

Policzer, this interpretation focuses on relations between leaders and followers within organizations and uses the proxy of information feedback loops to determine the structure of decision-making and the degree of control held by the centre over members (Policzer 2002: 9–11). It defines group conduct in terms of the levels of ‘coercion’ that they invite, with leaders exercising control through a variety of strategies of internal and external monitoring of their agents. More so than with the behaviourists, the emphasis here is on internal group dynamics, and the underlying reason for adopting such an approach is that system theorists wish to find a way of addressing the changing international context of globalization. According to Policzer: ‘[M]any non-state armed groups operate like quasi-states, with sometimes complex domestic and international administrations and representations.’ He further notes that ‘leaders in many states have dismantled their administrations to the point where they have become almost indistinguishable from non-state armed groups’ (ibid.: 7). For this reason Policzer feels able to apply a matrix that consists of bureaucratic coercion, transparent coercion, blind coercion and hide-and-seek coercion equally to states and non-state actors alike. Political association by its very nature, for him, remains a deeply coercive process, one which democracies are as prone to pursue as dictatorships or armed groups such as the notorious RUF in Sierra Leone. Despite their differences, these three approaches share a desire to interpret armed groups and militias in terms of classic assessments of power and in relation to legitimized structures of authority. For all three, power is seen in an explicitly coercive framework which emphasizes the visible expressions of compulsion over persuasion. In part, this seems to reflect a common understanding of the local context of action as fundamentally Hobbesian and devoid of more subtle approaches or opportunities for action. There are schools of social sciences and African politics, especially contemporary work examining the socio­logy of conflict in non-Western societies, which have developed a richer appre­ciation of local dynamics informing actions by militias which it is important to incorporate in any analysis. With respect to authority structures, and in particular the relationship to a globalizing environment that is said to be diminishing the saliency of states, while the ideological approach is limited by its own history, the behaviourists and systems approaches attempt to incorporate these insights into their analysis. For example, the ideological school retains a preference for understanding militias in terms which echo the work in principal-agent theory whereby states’ seminal role in founding and/ 24

Framework of analysis

As part of a multilayered framework of analysis, we propose a model of conceptualizing militias. The framework is meant to find common ground between the traditional ‘military capabilities analysis’ approach employed by military commanders and intelligence specialists to understand an adversary’s style of war and the more dynamic approach that is anchored in the historical, anthropological and cultural narratives of modern non-state armed actors (Shultz and Dew 2006: ch. 2). This framework (see Table 2.1) will hopefully be instrumental in future IR theorizing, enabling empirical fieldwork to be conducted more consistently, as well as for drafting innovative policy formulation, especi­ ally demilitarization and security sector reform initiatives. The bedrock of analysis is an understanding that militias are lucid, rational players operating within the international system who can successfully politicize and instrumentalize identity cleavages, not only for the mobilization of fighters in particular and society in general, but also for the retention of their support base. Since this is the fundamental assumption underlying the model, a non-acceptance of this rationality undermines the entire model. The second level conceptualizes motivations of militias, which are loosely organized along a continuum of ‘greed and grievances’, and are fundamentally fluid, thereby having the ability to oscillate between both extremes, depending on the specificities of the case study and 25

2 | Conceptualizing militias

or fostering military organizations remains the most important guide to assessing actions undertaken by these entities (see Gould 2006). Though revived to an extent in the post-9/11 environment, which ascribes much of militia behaviour to ‘rogue states’, the place of global­ ization remains only partially considered. The behaviourist school, though it shuns the ideologically fuelled approach to militias in an effort to capture the dynamics introduced by globalization, remains fundamentally tied in a state-centric analysis that neglects the chan­ ging complexion of exogenous influences over action. (This is despite the fact that in later accounts of their work, Collier and Hoeffler allow for the influence of diasporas upon resurgent conflict.) The systems approach takes a different tack in that it explicitly does away with the domestic–international boundary and, concurrently, the distinction between states and non-state actors altogether. In so doing, it diminishes the importance of recognized structures of authority associated with sovereignty, which themselves continue to exercise a determining influence in shaping conflicts and their outcomes despite the globalization phenomena.

subsequent circumstances. The third level of analysis considers the tactics and strategies of militias, which rest on contextual uses of violence linked to the main motivations of the militia organization. This means that militias choose certain tactics and strategies to address strategic and expressive motivations, such as to garner community support, to display resolve to their adversary and the community, to recruit and retain fighters, to protect local communities, to occupy resource-rich areas, etc. The fourth key element of the framework focuses on the organization of the militias, which are structured along a continuum, ranging from the highly organized to the loosely, dispersedly organized. Finally, it is imperative to consider the robustness of the legitimacy and accountability of militias in the areas where they operate and the country at large; as well as unearthing militia relationships with external actors (including state supports and regional economic and political networks) and probing inter-non-state armed group relations (i.e. partnerships and alliances among these groups in a particular country). table 2.1  Multilayered framework of analysis Motivations • Lucid, rational actors • Oscillation between greed and grievances, with a focus on broader political and socio-economic contexts • Layers of motivation: combatants, leadership, society • Evolution of motivations hence understanding motivation on a continuum Strategies • Linked to motivations; choose certain strategies and tactics and tactics to address motivations • Linked to targets: certain strategies and tactics will be employed to express objectives and resolve to the adversary, community, regional/international actors Structure • Continuum: highly organized, hierarchical to dispersedly organized, networked • Evolution of structure and militia based on contextual factors (exogenous and endogenous to the organization) Other factors • Legitimacy and accountability: linked to motivations; militia’s identity; set of ideas, beliefs, values and narratives that bind the organization together • Relations with other actors (state supporters; non-state actors; regional actors/networks) • Regional dynamics • Role of the international community, civil society, local communities

Rationality and motivations  The framework of analysis first outlined in this chapter and then applied in the subsequent case studies of this 26

27

2 | Conceptualizing militias

book rests on one main assumption: militias are lucid, rational actors operating within the international system, often defying the prevailing norms of ‘international society’. Rationality must be understood as socially constructed and an actor’s interests and behaviour are subject to socializing norms and, as such, are fundamentally mutable. At the same time, the application of the traditional rational-choice approach to interpreting the conduct of militias is limited, as it does not ­account for bounded rationality, cognitive dissonance and other social and environ­mental considerations in understanding the behaviour of actors. A more dynamic reconfiguration of the rational-actor model utilized here enables us to construct a fuller understanding of militias that seem to function beyond the instrumental level of survival. In keeping with this insight, our country-specific case studies illustrate that motivations of militias are not only multifaceted but also change over time, based on the perceptions of the local context held by leaders, the organization and the community. This highlights the multilevel nature of inter­actions between various agents and the context; as well as enabling us to adopt an approach which emphasizes rationality as ‘bounded’ within specific frameworks of limits of information, cognitive shortcomings and culture.4 After a discussion of our understanding of rationality, it is imperative to understand motivations underlying militia behaviour. Although the greed and grievances thesis is mainly applied to understanding civil wars, this framework is useful in understanding the evolution and oscillation of motivations of militias. As discussed earlier, the nexus can basic­ally be understood as two ‘rational’ choice models. For the purposes of this book, it will be applied to the rationalization of militias. Much of the academic debate on the economic causes of post-Cold War armed conflict has become polarized by the competing ‘greed and grievances’ dichotomy, juxtaposing ‘loot-seeking’ with ‘justice-seeking’ rebellions, and more generally the significance of economic versus socio-political drivers of civil war. The greed argument, on the one hand, prominently articulated by Collier, posits that the ability to extract wealth through violence and war encourages armed/rebel groups to initiate and partake in civil conflict. This assertion is empirically substantiated by crosscountry econometric analysis and by game theory modelling of warlord competition, looting models, etc. (Collier and Hoeffler 2004; Berdal and Malone 2000; and de Soysa 2002). It should be noted that earlier versions of the greed approach focused on economic combatants, while later on emphasis was placed on the opportunity for or economic feasibility of organized violence (Ballentine and Nitzschke 2003: 3).

Proponents of the grievances thesis consider theories of conflict that explain group mobilization for rebellion in terms of inter-group or horizontal socio-economic or political inequalities (Stewart 2002: 107–9). Economic and political differentials among groups are thus of fundamental importance to group mobilization. Therefore, grievances are not necessarily socially constructed by rebelling insurgents, but are at the core of motivations in response t0 resource scarcity, maldistribution of resources or political exclusion, sometimes also using the crutch of politicized identity-based cleavages (Homer-Dixon 1999; Stewart 2002). Collier and Hoeffler found that an integrated greed-grievance model, with only one of the potential sources of objective grievance (i.e. ethnic dominance), added significantly to the explanatory power of the greed model (Collier and Hoeffler 2002). Keen, using a Foucauldian framework of analysis, asserts that the ‘functions of violence’ should include not only economic and political functions, but also ‘psychological functions’, which emphasizes not only the ‘subjectivity of the violent’, but also the ‘subjectivity of the “victims”’ (Keen 2002: 4). Without being pigeonholed into the rationality–irrationality debate, specifically in the context of rational economic and political functions versus irrational emotive functions, we argue that it is important to conceptualize moti­ vations along a continuum, and to emphasize again that motivations are not only fluid but also can evolve based on the contextual realities of a specific locale and period. It should be noted that the contexts in which militias find themselves are both caused and sustained by a complex and shifting interplay between political and economic factors; as such, while each approach has analytical utility, an accurate understanding of the broad political economy of conflict requires a holistic perspective incorporating both approaches, while also accounting for broader social, political and economic contexts. Within this perspective, it is imperative to unpack the many layers of militia motivations within the wider context of the specific country. Due to the ongoing violence taking place in many ‘postconflict’ states, such as the DRC, Sudan, Timor-Leste and Afghanistan, numerous groups emerged and flourished in this post-conflict disaggregated environment. In some cases, owing to the weak state and security apparatuses, the state (with the support of the international community) is willing to negotiate with any armed elements that could potentially be ‘spoilers’ to the ongoing peace process. Overall, the motivations of militias can be understood at three mutually constitutive and reinfor­ cing levels. At one level, militias are motivated by personal enrichment (at the 28

Strategies and tactics  Militias employ a number of strategies to achieve their particularistic motivations, ranging from attrition, exploiting nat­ ural resources, forced recruitment (including, although not restricted to, children), manipulations of identity, and partnerships based on expediency and mutual economic benefit. Owing to their relative asymmetric weakness, in purely military terms, most militias employ a largely nocturnal strategy, including being most active in remote areas where dense forests and mountain terrain hinder search-and-destroy military operations, frequently dispersing after launching attacks, and making small-scale assaults on weak army positions. As Kydd and Walter argue, the strategies employed by particular armed groups (and in our case, militias) depend on the target of ­persuasion (i.e. 29

2 | Conceptualizing militias

individual and leadership levels), and economic-related activities (at the organizational level), including occupying and controlling mining areas, keeping supply routes open for arms/ammunition and logistics, arms trafficking, tax collection and looting. However, there needs to be a division between the individual motivations of combatants and the leaders of the militia. Individual combatants find themselves in a disaggregated and hostile environment, where the austerity of the economic and social conditions provides these combatants with the incentive to engage in armed activity. The individual combatant receives many fewer of the economic benefits, as compared to high-ranking officers within an armed movement; which thereby forces individuals to exploit and loot local populations to supplement their income. Closely linked to personal enrichment and economic survival, a­ nother motivation for many militias is the prospect of providing security to their communities, although this is located within a dangerous identitybased trajectory. The politicization of identity and protection, even at the level of rhetoric, as a guiding motivation enables groups to garner local sympathies, and has proved to be an effective strategy for recruiting and retaining combatants. This thereby forms and cements a common and mutually binding identity, as well as creating an elusive ‘other’, to enable ongoing armed and economic activities. The third level, which is closely linked to both the other levels, involves taking advantage of the post-transition political climate and resisting the demilitarization process. The political transition and post-transition processes in many countries have emphasized political co-option, army integration and economic growth. As such, there is a low level of impunity, thereby making it easier for groups to emerge, engage in violent activities and finally be co-opted within the political and military structures.

the government, international community, local population), and the characteristic or message they wish to convey through their strategies (to exert power, show resolve, foster trustworthiness).5 The most significant strategy employed by militias is attrition against the government, inter­ national community and, in some cases, local communities. This strategy is meant to persuade the government and the international community that the group is strong and resolute enough to inflict serious costs, while also illustrating its prowess and drawing symbolic power. Additionally, this strategy can also be reinforced by intimidating and disorienting local populations by launching indiscriminate attacks against civilians; thereby enabling the militia not only to benefit economically, but also to gain social control over said communities. This also inflates the perceived power of the group, by creating a sense of fear and terror among the population, political leadership and international community. This strategy is employed directly to achieve specific political ambitions in the post-conflict context, whereby the greater the costs the militia is able to inflict, the more credible its threat to inflict future costs, and therefore the more likely it is that the government and the international community will yield to the militias’ demands and grant the required concessions (ibid.: 59–66). In addition, militias must acquire the necessary resources to finance training and logistical support for field operations, forge group co­hesion by recruiting and retaining members, and maintain a clandestine network of supporters. The main sources of income for many militias range from trafficking, pillaging local communities, systematic looting and the confiscation of crop yields to smuggling and the trading of foreign merchandise. This economic imperative is intertwined with regional conflict complexes that produce not only cross-border economic inter­ action but also population migration and transborder political and identity links, which add another layer of complexity (Pugh et al. 2004: 25). These regional war economies solidify transnational economic networks, regional arms and mercenary networks, and socio-political and identity-based networks. The access to sources of income is a necessary strategy to recruit and retain combatants, on many occasions relying on the tactical utility of politicized ethnicity. As such, the politicization of identity-based cleavages and the calls for protection of local communities from ‘the other’ are especially prevalent here, as many militias utilize strategic violence not only to meet certain tactical objectives but also to communicate the ‘underlying’ motivation of the organization – usually for the purposes of garnering support from the local community, displaying power and resolve to their adversary, and fostering group cohesion. 30

of militias range from highly hierarchical, centralized formal units to informal, network-like units. Militias that follow a hierarchical structure have a clearly defined top–down communication chain, a strict framework for the division of labour, and adhere to a strict chain of command. Key decisions are made by the core leadership of the organ­ ization. This structure is meant to emulate state-like military structures, where there are clear institutional structures and clear links of vertical communication within the organization. However, since most militias engage in activities to counter the asymmetry in their own power, many groups have established network-like organizational structures, which make it more difficult to externally dismantle them. A hub network structure in contrast is relatively similar to a hierarchical structure. The main difference between the two is the lack of a formal division of labour in a hub network structure to control organizational activities, which allows the group much more manoeuvrability and innovation. In contrast, chain and multichannel network structures emphasize that the chains of command and control are relinquished for the purpose of attaining organizational goals, such as regime change, policy change, social control and personal enrichment. While the chain network structure still ­retains well-structured mechanisms of communication, the multi­channel network allows for the free flow of communication and encourages individual behaviour (allowing local units and branches the autonomy to act) within the framework of the organization’s goals and objectives. This network-like structure is inspired by the deterritorialization of the new political order, and the decentralization and informalization of the organization not only makes it a dynamic, fastmoving entity but also emphasizes that the strategic behaviour of the organization itself relies on a process of vacillation between a territorial presence (as part of its activities) and a mode of disappearance and retreat (to regroup and create a further disorientation of its adversarial target) (Mishal and Rosenthal 2005). Militias, depending on the level of ‘institutionalization’, also form functionally specific sub-units, e.g. for military, intelligence, political, financial and social development. One key issue that requires further investigation is the financial structure and network, especially where militias acquire resources to support their activities, including the training of members. At times, many militias operate within nefarious networks, which mean that it can be difficult to trace where resources come from and in what quantities. Finally, it is imperative to understand the cohesiveness of the group, as militias do have the potential to suffer 31

2 | Conceptualizing militias

Militia structure  As our case studies reveal, organizational structures

from factionalism and disunity, which can undermine the functioning of the organization. Initial success in securing their objectives triggers a process of militia ‘institutionalization’ that extends beyond military and fiscal domains. Indeed, the development of organizational structures transforms these groups into forces to reckon with over the longer term, although it should be noted that institutionalization does not necessarily aim at, or result in, state-making. Regardless of their ultimate objective and of the final outcomes of their struggle, initially successful militias engage in a process of institutionalization designed to meet the financial and organizational requirements of a protracted conflict or the contextual demands of operating in a weak state. However, this institutionalization in itself does not secure their legitimization or their inclusion in peace negotiations, since there is always a normative judgement on whether a group is legitimate or not. In an effort to legitimize themselves in the eyes of other power contenders or international mediators, militias may likely seek legitimacy through emulating forms and procedures identified symbolically and substantively with state legitimacy. Militia institutions are likely to borrow the specifics of their institutional forms and procedures from the templates that they are most familiar with in their direct environment, most notably the structure of the government or other opponent that they are directly opposed to.

Legitimacy and accountability  Militias may follow a coherent ideology or a more ad hoc set of political beliefs and objectives that perform a number of crucial socio-political and psychological functions important to the effectiveness of the militia. Whatever form it takes, militias need a set of ideas, beliefs, values and narratives that bind the group together. For example, during the Cold War armed groups were drawn and committed to various left-wing and Marxist-Leninist ideologies. In the post-Cold War period, ethnic, cultural and religious ideologies have become dominant. Although militias can be motivated solely by financial and other strategic objectives, ideational factors form an important part of a militia’s identity. Aside from understanding the ideological, political, economic or other basis for the militia, it is important to assess the extent to which the organization is able to rationalize, justify and legitimize the actions undertaken by the group, especially the use of violence. The use of political rhetoric, specifically the politicization of identity-based cleavages, is used to create a social-psychological sense of unity, solidarity and commitment within the group which facilitates the ability of the 32

Relations with other actors  Militias often establish linkages with other state and non-state actors for a number of tactical and strategic reasons, especially acquiring resources (be they financial, military, intelligence or political). In order to balance the asymmetry in power between militias and the states (and other non-state actors) they challenge, they frequently seek the assistance and resources of others to level the playing field. As discussed earlier, since militias most often operate in overlapping clandestine networks, finding accurate information on the nature and extent of these linkages can sometimes prove to be difficult. However, when possible the key issues that should garner closer analy­ tical attention are: to what extent are militias reliant on the linkages with state and/or non-state actors, and at what costs to the group does this support continue? The relationship between the militias and state and non-state actors 33

2 | Conceptualizing militias

organization to recruit and retain fighters, but also to foster group cohesion. This ‘auto-propaganda’ also promotes a sense of collectivity and solidarity among the population (see Hoffman and McCormick 2004), thereby leading to the reinforcement of enabling social norms that ‘­allow’ violence to be used as a mechanism of achieving collective goals, including the security of a particular identity group. As well, the process of forging a militia identity involves using the concept of ‘othering’ in complex combination with the instrumentalization of the scapegoat mechanism, while also fostering a strategy of moral disengagement (see Bandura 1990). Few militias can exist or succeed without some form of support, especially from the local community – which is able not only to replenish their supplies and update their intelligence but also to provide safe havens when necessary as well as replacements and new recruits. This support is interlinked with and inseparable from the legitimacy of the militia organization; as violence undertaken without some political, ­social or economic purpose will generate little popular support. ­Although Mao’s analogy of the ‘fish’ (guerrillas) that would die without the ‘sea’ of popular support, and Che’s foco, or heart, of revolution, seem a bit clichéd in this discussion, it is important to highlight that both emphasize a strong link between armed groups and the population in order for the organization to have broad-based legitimacy, especially in the areas in which it operates. With legitimacy comes an element of accountability that binds militias to linking their ‘violent’ actions with achieving a justifiable political end to their overarching goals and objectives.

can be understood as mutually constitutive and dynamic, meaning that their interactions will in some sense have an effect on their perceptions of each other and the context in which they operate, as well as what motivates them to remain engaged in the relationship (Wendt 1987: 359–60). That being said, both the militias and the actor they interact with should be seen as being autonomous, as an entity, and in the simplest sense this means the ability to pursue particular interests (ranging from protecting an ethnic community to resource exploitation), using a range of political and military instruments (Thakur 2005: 194). Owing to the fact that militias tend to be weaker, in terms of military and political power, compared to other actors such as states, they exhibit a greater propensity to assert themselves in the local context. Therefore, shifting alliances are necessary for militias to accom­modate their own asymmetry and evolve within the broader environment (such as in peace processes and government military campaigns) in which they find themselves. Not only is this a tactic of survival, but if executed properly it can prove to be an effective strategy for achieving their broader objectives. Contextualizing demilitarization

Since the local context is of paramount importance not only for demilitarization and security sector reform strategies but overall peace in post-conflict environments, we felt it was imperative to discuss certain issues that are key to understanding local dynamics. Please note that we do not intend to provide an exhaustive list of contextual issues; rather, basing our exposition on fieldwork and case studies, we want to highlight major local circumstances that critically texture demilitarization. Our earlier section focused on an interpretative framework necessary to understand the unique characteristics of militias. This section will discuss the context in which these militias operate, as well as some of the overarching contextual realities facing post-conflict countries. Each country-specific case study will be textured by certain contextual factors, in addition to certain local dynamics which will no doubt have a significant impact on the efficacy of demilitarization initiatives.

Historical context  Without necessarily succumbing to the divisive relativist–universalist debate, we feel it is essential to recognize that ­although post-conflict environments share certain ‘thin’ characteristics, in essence a ‘thick’ (Walzer 1996) contextualized understanding reveals the nuances and underlying challenges to peace in a specific post-conflict society. In this regard, abandoning certain universalizing 34

Political context  The shape and form of political arrangements in a ‘post-conflict’ environment have a significant impact on how demilit­ arization will be undertaken and its outcomes. Demilitarization is an inherently political endeavour. Combatants who hold considerable power are often part of the formal political formation process and are allocated important positions within the state apparatus. In some cases, however, militias are excluded from the formal political process and in turn revert back to violence, thereby spoiling the overall peace efforts. Political co-option and power-sharing arrangements are prevalent in numerous cases; however, the prevalence of former combatants can undermine the legitimacy of the state. But the alternative option of political exclusion can bring forth increased violence and instability (Torjesen 2009b: 414; see also Sisk 2003). The government’s position on integration of former combatants, either in government, as part of the national army/police, or in terms of societal reintegration, can have an instrumental impact on the demilitarization process. As well, electoral politics is a crucial component that is often overlooked. Elections are extremely important as they are designed to end the transition from war to peace, to instil ‘democracy’ by creating robust political institutions and rules of competition, and to serve as symbolic end points for the involvement of the international community (Lyons 2002). However, all too often elections have proved to be risky because the competition for power through the ballot box sharpens 35

2 | Conceptualizing militias

categories, and appreciating the ‘true historicity’ and a longue durée perspective (Bayart 1993: 5) of the post-conflict state, focusing particularly on the societal and political logic underlying configurations of power and violence will not only provide for a much richer analysis of the militias and violent spoilers involved, but also provide a way forward for developing effective demilitarization and post-conflict peace-building strategies. As well, it is important to note that ‘historical consciousness is also necessary to escape the false dichotomy between “domestic” and “international” politics’ (Ferguson and Mansbach 1996: 24). This is especially significant when considering the colonial historical trajectories of many states, such as the DRC and Sudan, as well as the history of militias and violent actors, in the case of Afghanistan. It is important to note that although all post-conflict environments have a specific historicity, in some cases, such as the DRC and Afghanistan, it is much more embedded within the country’s political, economic and societal framework than in other cases, and therefore requires closer examination.

social d ­ ifferences – usually along the lines over which the war was fought. In many cases, this militarization of political parties results in providing electoral legitimacy to military-based authoritarian parties which continue to rely upon coercion, fear and chauvinism to remain in power. For example, the hurried search for ‘local’ Iraqi groups to be transformed into political parties and to form a new Iraqi government resulted in the co-option and elevation of various sectarian and militia-wielding elements, none of whom enjoyed much legitimacy or support among the Iraqi population (Ucko 2009). As such, there needs not only be a focus on the demilitarization of politics, i.e. the separation of political parties from militias, but also a focus on creating a broad coalition among competing elites (including militia ‘strongmen’) prior to elections that institutionalize the distribution of power and thereby reduce uncertainty. A broader issue related to this is state capacity, or more importantly the political will of the government and its legitimacy. In a post-conflict environment, the struggle for political power is contested by numerous actors, including militias. As such, the government’s resolve to not only foster legitimacy and good governance, but also to provide security, encourage a well-functioning economy, deliver basic social goods, and bring perpetrators of crimes to justice is fundamental. These tasks require state resolve and institutional capacity-building. This has significant implications for the demilitarization process, as a weak government is more likely to negotiate with militias that continue to operate in the post-conflict environment, and the national army is unable to protect civilians or combat new violent groups that may emerge in the security vacuum resulting from a weakened state. This is especially salient in the case of the DRC and Afghanistan. Therefore, effective demilitarization strategies must consider the political dimensions of the political arrangements that are the cornerstone of peace-building processes. Overall, the weakness of political governance can have a significant impact on demilitarization outcomes.

Economic context  Very closely linked to political governance and i­nstitution-building, the provision of tangible material benefits is crucial for the population to have confidence in the state’s ability to steer the peace and democratic transition processes. As such, there should not be a focus solely on deconstructing the previous institutions that may have served a previous regime’s interests, but also on addressing socio-economic inequalities and grievances, which in many circumstances are the roots of violence, through progressive and equitable 36

Politicization of identity-based cleavages  For our analysis, identity-based cleavages include ethnicity, communal, religious, regional/geographical and linguistic cleavages. As discussed earlier, militias often politicize identities for political and strategic objectives, and usually to garner the legitimacy of local populations. While we recognize the importance of the politicization of identity-based cleavages as a powerful device used by both militias and national governments in many instances, the pre­ occupation with identity undermines the complex interplay between the strategic politicization of identity, structural imbalances and turbulent regional contexts – all of which combine to generate a multilayered context in which demilitarization activities take place. We also employ a constructivist conceptualization of identity which does not take it as a given, but rather as being developed, sustained and transformed through the processes of interaction.6 The term itself originates in social psychology and refers to ‘images of individuality and distinctiveness (“selfhood”) held and projected by an actor and formed (and modified over time) through relations with significant “others”’ ( Jepperson et al. 1996: 59). As such, ‘the term (by convention) references mutually constructed and evolving images of self and other’ (ibid.: 59). In addition, identity should not be considered as an ‘essentialist’ attribute, but rather one of several identities; and as such they should 37

2 | Conceptualizing militias

economic development. As well, the government’s ability to manage the natural resources in the country is essential, as there is the possibility that resources could continue to be exploited for dubious political gain not only by the militias but also by the government. In this regard, the strengthening of the legal economy is important. Decreasing the impact of the war economy is paramount, as illicit activities provide the means of continuing armed conflict by militias. Finally, how funds for demilitarization are delivered must be examined with the overall context of demilitarization. One major concern has to be the robust disarmament and demobilization initiatives relative to reintegration activities, as well as the weak institutional capacity of national demilitarization agencies. For example, the Congolese national agency responsible for demilitarization, CONADER, was crippled with high levels of funding mismanagement and weak institutional capacity to help coordinate DDR programmes with MONUC, local NGOs, local governments, commun­ ities and former combatants. With budgetary limitations and numerous priorities, national governments must be cognizant of the various economic necessities of implementing an effective demilitarization programme among other post-conflict reconstruction priorities.

not be discussed outside the confines of their precise historical and geographical context, meaning that identities are a function of the circumstances under which they become salient (Chabal and Daloz 1999: 58–9). For example, under colonial rule in most African states, there was a prevalence of the ‘invention of ethnicity’ (ibid.: 58), meaning that ethnicity was constructed and instrumentalized to achieve the respective goals of the colonial administration. In the African context, specifically in the Great Lakes/Central African region, ethnicity formed the predominant component of Congolese, Rwandan, Burundian and Ugandan identities, usually broadly defined along the Tutsi–Hutu ethnic continuum. Mamdani asserts that rather than conceiving of identity as simply ‘invented’ or ‘imagined’, it should be discussed as the ‘making’ (Thompson 1963) of an identity, in the sense that identity serves as a means to achieve a particular political end (Mamdani 1996: 185). Interestingly, the politicization of identity exists in most states, as on the one hand identity has been forged in effect to ‘normalize a population by giving it a sense of unity’, and on the other hand it has been formed by exclusionary practices, which attempt to secure the ‘domestic identity process of spatial differentiation’.7 The notion of the politicization of identity-based cleavages has been the focal point for the literature on ‘identity politics’. Identity politics principally involves the claim to power on the basis of a particular identity, and as such identities are reinvented in the context of the corrosion of other sources of political legitimacy (Kaldor 2006: 77). In a post-conflict environment, this can be dangerous since both the state and militias utilize identity as a means to garner political accountability and legitimacy, which are essential to the consolidation of political power. For example, in the Kivus and Ituri regions of eastern DRC, the politicization of ethnicity has been used by local/provincial/national governments and armed groups alike to foster political legitimacy. Militias have also successfully used ethnicity to gain local community support, and as a means to recruit and retain fighters. The key identity-based conflict centres on the manipulation along dangerous Hutu–Tutsi and Bantu–Niolotic ethnic lines under an overarching rwandophone ethnic configuration, and on the debate on autochthony (conflict between ‘originaires’ and ‘non-originaires’) (Thakur 2008a).

Regional dynamics  Regional dynamics cannot be overstated within the context of demilitarization. First, the political economy of war ­emphasizes the prevalence of extensive transnational resources and arms trafficking networks (see Pugh et al. 2004). As well, many contempor­ary 38

The role of the international community  The role of the international community is a significant contextual issue when developing and executing demilitarization programmes. Depending on the nature of the external support, demilitarization efforts can be assisted with an infusion of financial resources, new ideas and the benefits of international personnel who bring their rich experiences to the country. However, the international community also brings with it technocratic, top-down, cookie-cutter approaches, as well as a short attention span with the application of ‘results-based’ indicators used as a ‘magic fix’ and a point of departure. The sometimes overly ambitious agenda, especially with regard to demilitarization, pays little attention to the local context as well as the institutional gaps among international organizations, donor agencies and the national government. The demilitarization project specifically, and the post-conflict reconstruction project in general, is marred by the international community’s inability not only to a­ ddress the root causes of the conflict but also to have a view of peace as the absence of war (negative peace). Post-conflict reconstruction, and especi­ ally demilitarization, is fundamentally about rebuilding and modifying the political, social and economic infrastructure to support the institutionalization of more stable and responsive systems to avert conflicts 39

2 | Conceptualizing militias

conflicts have had spillover effects in neighbouring countries in the form of refugee population movements and armed incursions, as in the case of DRC and Sudan, as well as forming part of a transnational ideological or political movement, such as al-Qaeda and the Taliban in Afghanistan. Being in a ‘bad neighborhood’ can also export negative consequences to other states (see Iqbal and Starr 2008); and therefore a regional perspective to conflict and peace must be emphasized. There is also increasing recognition that conflicts are fluid, and the ‘regional’ dimensions not only nuance our understanding but in some cases may reveal significant contrasts with other post-conflict environments (Shaw 2003: 487). Understanding the regional dynamics allows us to consider the various regional interests of various state and non-state actors at play, including actors who wish to enrich themselves in transnational economic networks, destabilize rival neighbours for political gain, or strengthen particular identity-based networks. Finally, the role of regional bodies, such as the African Union, the European Union, the Southern African Development Community (SADC), the Inter-Governmental Authority on Development (IGAD) and the Organization of American States (OAS), cannot be overlooked, as these bodies have vested interests in maintaining peace at the regional (and, in the case of the EU, the global) level.

(see Busumtwi-Sam 2004). Although the international community may control the form of the institutions they help build, such as the national army, without understanding the post-conflict contextual dynamics and the nature of militias and incorporating them into the overall national initiatives, the substance of the institutions will also leave them open to manipulation and decay in the long term. Again, governmental capacity to implement demilitarization programmes is of paramount importance (as discussed earlier); however, it is imperative to critically assess the role of the international community. Post-conflict decisions affect the lives of people, but there needs to be more emphasis placed on the international community understanding the impact of its own actions, in a critical and constructive manner. External actors are encouraged to make deeply interventionist decisions for other societies (sometimes in close coordination with a weak government), unbound by outside control or scrutiny, and unaffected by outcomes (see Uvin 2001). This lack of self-critique, ‘we-know-best’ attitude, and the overarching technocratic, universalized approach divorced from local contextual realities, has unfortunately been internalized in the normative thinking of the entities within the international community. This hegemonic DDR discourse leaves little for room for critique, change and bottom-up initiatives.

The role of civil society and local communities  Demilitarization efforts have mostly been focused at the national level, linking the state apparatus with international efforts from various actors, including the UN, bilateral agencies, regional bodies and international NGOs. Less attention has been paid to civil society and local community initiatives, which could perhaps provide vital insights to overcome the numerous gaps in the development and execution of effective demilitarization programmes. Civil society organizations and local communities, owing to their proximity to local-level actors at the end of the implementation chain, can play a key role in ensuring that peace is sustainable. As well, local communities have the advantage of access to local knowledge, which can inform a deeply contextualized understanding of the ­barriers to and opportunities for making peace a reality at the local levels. For example, in South Sudan, local communities and NGOs played a significant role in encouraging their youth (who were part of the White Army militias) to disarm peacefully and voluntarily, through various ‘sensitization exercises’ and ‘peace meetings’ (Arnold and Alden 2007). However, civil society should not be seen as a panacea for successful demilitarization or post-conflict reconstruction, as it can reinforce divisive policies based on patronage, especially those fostered by the 40

Conclusion: from the theoretical to the empirical

This chapter provides a multilayered framework to better understand militias, as well as highlighting key political, economic and social contextual processes that texture the demilitarization process. By reverting to and questioning the underlying principles which inform both the scholarship and policy work on militias, we have opened up the subject to a reconsideration of the sources of conduct. Rationality, albeit bounded by the narrow ideational and socio-economic factors that compose the setting in which militias arise and act, can be usefully applied to deepen our understanding of their behaviour. The following chapters contain empirical case studies that employ this conceptual framework, focusing on the empirical mapping of specific militias, and discussing the key issues that form the local environment within which local (and international) actors operate. By framing the examination of the militias in South Sudan, the eastern DRC, Timor-Leste and Afghanistan in terms which draw upon this set of theoretical propositions, we are able to draw out important aspects of the security situation that have a crucial bearing on peace and stability. Taken together, this analysis of the militias and the local contexts will allow for cross-country comparisons, as well as proactive policy suggestions and future theoretical modelling in the concluding chapter.

41

2 | Conceptualizing militias

war economy. As well, owing to the fact that ‘civil society’ and ‘local communities’ are most often identified in opposition to ‘failed state’ and ‘weak state’ by external actors, the dichotomy elevates local communities to an apolitical, virtuous position, while the ‘state’ is equated with being highly political, but institutionally weak and corrupt (see Lemarchand 1992: 187). As such, it is imperative to understand and appreciate that in addition to top-down implementation, there must be concurrent bottom-up processes aimed at constructing a new social contract and healing societal divisions (Prendergast and Plumb 2002).

3 | The South Sudan Defence Force

Sudan has endured a civil war in its south for most of its post-­ independence history. After twenty-one years of nearly unbroken fighting, the Comprehensive Peace Agreement (CPA) was signed between the Khartoum government and the rebels of the Sudan People’s Liberation Army (SPLA) in January 2005. The civil war in the south was a very complex one, involving a plethora of armed actors ranging from state armies, namely the Sudanese Armed Forces (SAF), to formal rebel movements, notably the SPLA. However, to a very significant degree it also involved militias, and indeed, a strong argument can be made that without understanding the role of militias it is impossible to appreciate Sudan’s recent history. Following decades of war, the demilitarization of the south, as envisioned by the CPA, was to be primarily through the application of DDR programmes and the separation of the opposition military forces, specifically the SPLA and the SAF. However, this official focus of the CPA obscured both the prominent role played by militias in the civil war in the south, and their continued presence there during the transition to peace. During the civil war, the most significant ‘other armed group’, to use the CPA’s terminology, was an amalgamation of militias loosely aligned under the umbrella of the ‘South Sudan Defence Force’ (SSDF). Entrenched ethnic and political disputes among southerners had long been unofficially exploited on a relatively small scale by the government in Khartoum. Through the signing of an agreement in Khartoum in 1997, however, the militias that were to form the SSDF became official government allies and, in this capacity, were comprehensively supported by Khartoum against the SPLA. This chapter will provide an em­pirical mapping of the SSDF through an analysis of its operational mode, moti­ vations and interests, as well as its relationship to external actors. The focus is on a specific time period and specific SSDF units in order to draw out these dynamics and trends from the post-CPA experience of South Sudan in dealing with militias. Hence, emphasis is placed on SSDF militias in two states, Upper Nile and Jonglei, as these were where they were most concentrated, and the period from January to August 2006, because this was the definitive period for the SSDF following the 42

The origins of the SSDF

The progression of the civil wars in Sudan and the relations of the armed groups that have fought them have been complicated and fractured. Sudan’s second civil war began in the late 1970s and was initiated by what was known as ‘Anyanya II’, a movement of veterans of the first civil war who were dissatisfied with Khartoum’s abrogation of the 1972 peace agreement and its guarantees of southern autonomy. By 1983, Anyanya II was joined in revolt against the central government in ­Khartoum by the SPLA, which was led by a defecting SAF officer named John Garang. During the 1980s, relations between these two groups soured, and ultimately competition between them led to the defeat of Anyanya II by the SPLA, and either the incorporation of its forces into the latter or their splintering into militias supported by the Khartoum government. In 1991, the SPLA itself underwent a major change when two of its senior commanders, Lam Akol and Riek Machar, defected from the leader­ship of John Garang over ideological differences concerning southern independence (O’Ballance 2000: 172). The two defecting ­leaders themselves separated in 1992, with Akol forming the SPLA-United and Machar eventually founding the South Sudan Independence Movement (SSIM). Additionally, the early and mid-1990s saw a proliferation of new armed groups in the south, as smaller factions broke away from these larger organizations or formed in their own right. As the 1990s progressed, most of these non-SPLA armed groups sought and received support from the Government of Sudan (GoS), so that they might be better able to protect their tribal homelands and/or pursue their own grievances with Garang’s SPLA forces. The relationship between these varied forces and the GoS was consolidated in 1997, through the signing of an agreement known as the Khartoum Agreement, and an addendum to it known as the Fashoda Agreement. Overall, the Khartoum Agreement stipulated that after an interim period of unspecified length, the south would be provided with a vote on self-determination, the long-held aspiration of those southerners in disagreement with Garang over the SPLA’s insistence on a united Sudan. The Khartoum Agreement also formalized the amalgamation of non-SPLA militias into an umbrella organization, the SSDF, led by Riek 43

3 | The South Sudan Defence Force

CPA’s signing. In particular, the chapter highlights the roles and effects that this loose alliance of militias had on prospects for demilitarization and sustainable peace in South Sudan in the immediate period after the CPA was signed.

Machar. For their part, the Khartoum Agreement left the GoS with the new opportunity to exploit the oilfields of the northern part of South Sudan, namely those in Unity state, since they were under the control of the SSDF. Even with its official standing as a partner in the Khartoum Agreement, the SSDF never had much internal coherence, as its various component militias were largely autonomous. This was primarily encouraged by the fact that SAF Military Intelligence supported individual commanders directly, which prevented the SSDF from forming into a coherent, unified organization (Institute for Security Studies 2004: 5). Furthermore, the fractured nature of the organization also resulted from the often conflicting individual aspirations of the leaders of its component militias, and the strong tribal and clan loyalties of the southern Sudanese (ibid.: 13). By 2002, Riek Machar had left the organization, which then came under the political leadership of Gatlauk Deng and the military command of Major General Paulino Matip. It is notable that even as ‘chief of staff’, Matip served ‘largely as a figurehead beyond the area of his immediate control’ in his home area in Unity state (ibid.: 6). Under the stipulations of the CPA, the SSDF was supposed to be subsumed within the SPLA or the SAF by January 2006. The CPA mandated that there could be no ‘third army’ in South Sudan after this deadline, yet the exact mechanisms for ensuring this were left decidedly ambiguous in the text. Under the CPA, the SPLM (which had been the political wing of the SPLA) was allowed to form the autonomous government of South Sudan until interim elections could be held prior to the referendum on independence. On 8 January 2006,1 an agreement was reached between the president of South Sudan, Salva Kiir, and Paulino Matip of the SSDF, entitled the Juba Declaration. Through the declaration, the majority of SSDF forces would come to join the SPLA, with Matip becoming the SPLA’s Deputy Commander in Chief. This was a major blow to the GoS, as the SSDF forces that had been securing the country’s major oilfields in Unity state on its behalf, under the command of Matip, were now a part of the SPLA. Nevertheless, following the Juba Declaration there was a reorganization of the SSDF, which reformed under the leadership of Major General Gordon Kong. Although the SSDF lost its strongest position in the south, namely in Unity state, and the majority of its troop strength to the SPLA, the organization maintained an armed presence in South Sudan, primarily in Jonglei and Upper Nile states but in other states as well. This was in contradiction to the dictates of the CPA. The usage by the GoS of allies, or what might be termed southerner 44

SSDF’s operational mode

‘Nobody can disarm us, we’re another army’ (SSDF soldier)2

The operational mode of the SSDF after the Juba Declaration was severely constricted by the implementation of the CPA. This was for several reasons. First, the SAF, the SSDF’s near-exclusive supplier of material support, had itself seen its access limited in the south, as SAF forces had to concentrate in CPA-defined ‘assembly points’, and had also been required to begin to redeploy to the north. Second, the SPLA for its part had been able to strengthen its presence in areas where the SSDF was previously able to maintain a consistent presence. Third, and most important, the Juba Declaration saw the majority of SSDF forces 45

3 | The South Sudan Defence Force

‘proxies’, has been a common tactic. As Emeric Rogier has noted, ‘a fundamental principle of Sudanese politics [has been] that northerners […] seek allies from “the other side” to fight their own-brother enemies, hence the formation of “cross-border” alliances’ (Rogier 2005: 12). The first such case in the second Sudanese civil war began in 1983, when some Anyanya II units aligned themselves as pro-government militias. Furthermore, within months of Garang’s defection from the SAF, it had supplied local tribal militias to fight the nascent SPLA. For instance, Murle militias were provided with weaponry to attack Dinka and Nuer communities, from which the SPLA was drawing support (Johnson 2004: 68–9). Overall, ‘Khartoum’s strategy of supplying southern militias and waging war by proxy’ began under the Nimeiri government and was continued ‘by all successive governments’ (Rogier 2005: 19). Considering that, the sad truth of the second Sudanese civil war in the south is that it was largely fought between southerners. Thus, the war has often been described as ‘civil wars within a civil war’, something that Khartoum governments have been adept at playing to their advantage. Khartoum’s ability to manipulate actors in the country’s periphery is proving a consistent trend. In the ongoing conflict in Darfur, Alex de Waal notes that Khartoum is once again resorting to ‘counterinsurgency on the cheap’ through its support for so-called ‘Janjawiid’ militias (de Waal 2004). According to de Waal, this tactic, tried and tested in the south, was used when Khartoum ‘sought out a local militia, provided it with supplies and armaments, and declared the area of operations an ethics-free zone’. As the conflict continues to rage in Darfur, it is no surprise to southerners that the roles of militias acting in tandem with Khartoum are at the centre of the country’s present civil war.

join the SPLA, notably some of the largest SSDF units that had been concentrated in Unity state under the command of Matip. The end result was to isolate the SSDF in pockets of Jonglei and Upper Nile states, with each commander controlling a small area of land surrounding one or two militia camps in their home districts. The UN estimated3 that in early and mid-2006, the major concentrations of SSDF forces in Upper Nile and Jonglei states were: 1 Gordon Kong in Nasir, commanding the ‘Al Nasir Forces’ with about 300 to 400 men; 2 Gabriel Tang in Pam Al Zaraf, commanding the ‘Pangak Peace Forces’ with about 1,000 men; 3 Ismail Konyi in Pibor, commanding the ‘Pibor Defence Forces’ with about 2,000 men; and 4 Thomas Mabior in Dolieb Hill, commanding the ‘Peace Forces ­Dolieb Hill’ with about 250 men. In addition to these pockets of SSDF forces, there were also concentrations in the Upper Nile’s capital city of Malakal. These forces were miscellany from the other units. For instance, there was a camp of at least one hundred SSDF soldiers who resided near the town’s airport, while others lived in the compounds of their commanders. It is notable that many of the commanders of the SSDF forces found outside Malakal tended to reside themselves in Malakal, while most of the top comman­ ders, such as Gordon Kong, resided almost permanently in Khartoum.4 In general, despite tensions between these pockets of SSDF forces and SPLA forces, there were relatively few violent clashes between them from January to August 2006. Fighting between SSDF forces and SPLA troops did take place in Longochuk county in north-eastern Upper Nile state on 16 April 2006, when the local SPLA attacked the local SSDF forces in retaliation for an assassination attempt the week before on a county commissioner who belonged to the Sudanese People’s Liberation Movement (SPLM).5 This attack killed a number of SSDF militiamen and forced the SSDF to retreat. Moreover, it was the first time that SPLA troops had attacked an SSDF force in Upper Nile state since the signing of the CPA. The only other significant confrontation was in mid-August in the Fangak area of western Upper Nile state, over disagreements surrounding the county commissioner’s post allotment for Fangak county. The tensions eventually led to fighting between SSDF forces commanded by Gabriel Tang-Ginye and the local SPLA forces. A pressing concern of SPLA authorities in 2006 was the fear that the SSDF was recruiting additional forces. They believed that this was being 46

47

3 | The South Sudan Defence Force

done for two main reasons.6 First, there was the pragmatic need for individual commanders to acquire and maintain sufficiently large forces to achieve what one SPLA general termed ‘big man power’.7 In order to be recognized by the GoS and SAF, militia commanders needed to have enough armed men backing them. Second, SSDF commanders, it was argued by the SPLA, were recruiting so that they could try to acquire new areas of control, or at least be in a position to do so in the future. This was primarily so that they might influence future elections and increase their political sway. However, it was not evident to the authors that the SSDF was actively trying to acquire new territories up to August 2006, despite SPLA allegations, and it was a consistent argument of the SSDF that it merely wanted to maintain the areas already under its control. Even with recruitment allegedly taking place, the overall size of SSDF forces was in any case significantly smaller than that of the SPLA in the area, which numbered in the tens of thousands. In addition to raw numbers of troops, the SPLA also maintained a significant advantage in firepower, as it had tanks, artillery and other heavy weaponry, whereas the SSDF was largely confined to small arms and a limited number of mortars and heavy machine guns. However, the SSDF remained a significant threat to the SPLA and the Government of South Sudan (GoSS), not because of its present number of ‘active duty’ members but rather because of its ability to rapidly recruit afresh, as well as call up ‘reservists’ should conflict resume in the south.8 Furthermore, while there were relatively few pockets of armed SSDF, its militiamen were very experienced veterans who were fairly well armed, even if they lacked heavier weapons. This level of experience and possession of suitable weaponry meant that the SSDF could not be dismissed as a fighting force.9 One issue of crucial importance to the dynamics of post-CPA South Sudan was the continued presence and movement of arms and ammunition. While the SSDF found itself increasingly isolated in pockets, it allegedly still received arms and ammunition from SAF, as it had for years.10 If proved true (even SSDF commanders have admitted that SAF was their major source of supplies), it was a serious violation of the CPA by the GoS.11 In addition, SSDF members also noted that they had significant stockpiles acquired from past battles, and that now as in the past it was possible to buy ammunition from soldiers and officers in the SPLA sympathetic to them or just plain corrupt.12 The primary means of transporting arms and ammunition was through the major SAF base in Malakal. Malakal was a convenient trans-shipment centre because of the SAF base, the presence of many SSDF militiamen (notably

commanders), river transport along the Nile and Sobat rivers, and access to roads to the more rural areas of Upper Nile state.13 There were also consistent reports of SAF using a white helicopter to deliver supplies to the SSDF pockets. SSDF motivations

The SSDF demanded four major concessions when it was negoti­ ating the Juba Declaration.14 The primary political demand was that the SSDF be assured ‘full political participation’ in the GoSS as well as in the Government of National Unity, which consists of a union of the SPLM and the ruling party in Khartoum, the National Congress Party (NCP), as per CPA stipulations. This demand practically meant that SSDF leaders would be accorded key positions at all levels of governance. The remaining three demands all concerned military concessions. First, it was requested that all SSDF soldiers and officers be allowed to join the SPLA at the same rank as they held within the SSDF. Second, the post-Juba Declaration’s ‘new SPLA’ had to be non-partisan and not controlled by the political party that had historically been attached to the SPLA, the SPLM. This meant that it had to be under the civilian control of the GoSS itself. Third, after the referendum authorized by the CPA for 2011, and assuming it to be for independence, the SPLA must then change its name to the ‘South Sudan Army’. While the SSDF was long dismissed by its opponents in the SPLA as mercenary ‘Arabs’ for the GoS, there were actually a number of thoughtful political objectives that were consistently adhered to by its members. Central to this was the long-held demand for southern selfdetermination, specifically leading to independence for the south, which went back to the SSDF’s origins in the Anyanya II movement. It has often bewildered foreign observers of Sudanese politics that the SPLA was officially fighting against the GoS even though it advocated a unified Sudan, while the Khartoum-allied SSDF was the southern military force conversely fighting for complete independence. This conundrum defined tensions between the SSDF and the SPLA, and has continued to provoke a contentious political atmosphere leading up to the CPA referendum on self-determination in 2011. At the Juba Declaration negotiations, an understanding was reached on the issue of self-determination, because the SPLM/A assured the SSDF negotiators that the referendum would take place no matter what obstacles were faced. Many ex-SSDF members who came over after the Juba Declaration noted that the CPA was in most ways very similar to the Khartoum Agreement of 1997, most significantly in that it guaranteed 48

49

3 | The South Sudan Defence Force

the south a vote on self-determination. Crucially, the CPA both set a time frame for the referendum and was recognized by the international community, attributes that the Khartoum Agreement lacked and which ultimately made the CPA the more viable of the two. Apart from ideological issues surrounding self-determination, the demand for replacement of the ‘SPLA’ with a ‘South Sudan Army’ was perhaps the most divisive one. At previous rounds of ‘south–south’ reconciliation meetings it had proved the stumbling block, and it is notable that Garang had consistently refused to compromise on this issue.15 As a result of Garang’s intransigence, SSDF leaders saw his untimely death as an opportunity to begin reconciliation efforts anew.16 And indeed, through the Juba Declaration a compromise was reached that satisfied the CPA requirements placed upon the SPLM/A, which stipulated that the SPLA be the army of the autonomous South Sudan in the interim period before the 2011 referendum, while also assuring the SSDF that if independence were to be selected in the referendum, the SPLA would cease to exist and would be replaced by a new entity known as the ‘South Sudan Army’. The political part of the Juba Declaration was considered a success by the SSDF members, who agreed to it because they did get a significant degree of political participation in the GoSS.17 This took the form of government positions totalling six ministers, six county commissioners and over twenty-five members of state assemblies. The military part seems to have been more problematic. The successes, as far as the SSDF negotiators were concerned, were that its members did officially transfer to the SPLA with the same ranks; the SPLA was no longer solely under the control of the SPLM political party but rather under the organs of the GoSS (though the degree of actual control is still ambiguous); and there was a formal integration procedure for incorporating the SSDF forces into the SPLA.18 The integration procedure was coordinated by an ‘old SPLA’ general and an ex-SSDF general (who could be considered ‘new SPLA’), who respectively chaired and co-chaired a ‘Military Technical Committee’. This committee submitted an integration report in late July 2006, reporting that formal integration had occurred but that challenges had been faced. The chair of the committee told the press at the time of the report’s submission that the primary obstacles to SSDF–SPLA integration had been ‘tribal clashes […] including a lot of man-made destruction organised by the Sudan Armed Forces who sent southern traitors’ to sabotage efforts by the SPLA to disarm armed civilians in Upper Nile and Jonglei states.19 These ‘tribal clashes’ referred to fighting between White Army

militias from Lou Nuer communities and the SPLA, to be discussed in the next chapter. The so-called ‘southern traitors’ were SSDF forces that did not follow the Juba Declaration, a topic to be covered later in the chapter. As part of the integration efforts, ex-SSDF militiamen were given US$300, five months’ worth of food, and free accommodation during the integration procedure. With the submission of the final integration report, ex-SSDF members received a regular salary as normal SPLA soldiers. Post-Juba Declaration motivations of ‘hold-out’ SSDF

‘Disarmament is not a solution. Anybody who tries to disarm us will destroy the peace.’  (SSDF soldier)20

One of the most oft-stated rationales of SSDF militiamen for not following Matip after the Juba Declaration was paranoia either that the referendum would not occur, or that it would be corrupted by the SPLM/A, which it was feared really would take the ‘making unity attrac­tive’ clause of the CPA very literally and push for a no vote in the 2011 referendum. One SSDF officer interviewed explained that it was necessary to maintain a military relationship with SAF, despite whatever reservations were held about continuing to ally with Khartoum, because it was felt an ‘insurance policy’ was required.21 This meant keeping the SSDF as a standing armed force until the 2011 referendum’s outcomes were known. How much of this was rhetoric, defending their continued existence in violation of the CPA and after the Juba Declaration, rather than sincere ideological belief, is difficult to ascertain. Nevertheless, SSDF members consistently argued that maintaining the SSDF as a fighting force would be necessary until independence was assured, preferably through the 2011 referendum but in the longer term if necessary.22 In this regard, it was argued that if the 2011 referendum were indeed for southern independence, then the SSDF would gladly cease to exist as it would have achieved its primary objective. Another common argument articulated against agreeing to the Juba Declaration was the more pragmatic one that there were simply not enough GoSS positions to share among SSDF leaders, so they simply did not feel impelled to join.23 The Juba Declaration agreed on the precise number of positions to be provided to SSDF members in the GoSS. This in itself caused much resentment among older SPLM/A members, as it obviously meant that they would be losing out on positions to make way for the concessions to the SSDF. There was likely also much tension created within the SSDF at the time over who would be given 50

51

3 | The South Sudan Defence Force

which positions and who would be left out. A demand frequently cited by the SSDF hold-outs was that if further reconciliation were to occur between the SSDF and SPLM/A, then additional positions would need to be shared in the GoSS by the SPLM/A. If this were to happen, then it would show that the ‘SPLA wants to share power, [and] then everything would be OK’.24 Furthermore, the availability of governmental positions probably played a key role in the individual rationales of SSDF leaders at the time, with those holding out weighing their options of acquiring a GoSS position in future rounds of south–south reconciliation talks. Moreover, the SSDF leaders who decided not to follow Matip into the SPLA after the Juba Declaration cited their unwillingness to concede on the issue of forming a ‘South Sudan Army’ to replace the SPLA as a primary motivation. They continued to demand that for the SSDF to reach a final reconciliation with the SPLA, a new southern army must be created, incorporating both the SPLA and SSDF forces, even before the 2011 referendum.25 The GoSS could not concede to this because it would have been a violation of the CPA, which stipulated that the SPLA is the army of the autonomous state of South Sudan prior to the 2011 referendum. Additionally, the unwillingness of hold-out SSDF militiamen to join the SPLA was driven by the reality that the SSDF was still in physical control of certain parts of South Sudan and maintained armed forces. In relation to this, many SSDF members still regarded the Khartoum Agreement as a legitimate agreement, recognizing the SSDF as the armed forces of the south, i.e. that they were in fact a government power of sorts.26 This was considered to be true at least by SSDF members for the areas under their control and, hence, the rhetorical question was asked concerning its subsuming into the SPLA: ‘So how can it give up its guns to another government?’27 In parallel with its continued political hopes that it indeed be treated as ‘another government’ in the south, the SSDF was intent on keeping the areas that it had under its control when the CPA was signed. Qualifying this intent was the expressed desire not to go on the offensive before the 2011 referendum, but still to resist any encroachment into its areas by the SPLA.28 Given the SSDF’s insistence both on maintaining itself as an armed group in control of the land it occupied at the time of the CPA’s signing, and not being an offensive organization, there were two main causes its members believed could provoke a return to war in the south.29 The first potential cause would be if the SPLA tried to forcefully disarm the SSDF. A speech by GoSS president Salva Kiir in July 2006 vowed to disarm the militias in an unspecified time frame if they did not join the

SPLA or SAF, as stipulated by the CPA.30 There appeared to be a genuine belief among SSDF rank-and-file militiamen that the CPA entitled the SSDF to continue to exist as an armed organization. One SSDF militiaman, speaking on behalf of a group of his comrades, explained this sentiment: ‘The CPA says the SSDF and SPLA should stay in their own areas and not attack anybody until the referendum, but now the SPLA is expanding everywhere.’31 Another SSDF militiaman expanded on the general theme, and ­argued that while they had some sense of being part of one country (i.e. ‘southerners’), they felt that the SPLA mistreated some southerners, namely that it unfairly targeted the SSDF. In his words, ‘the children of the one house are supposed to sit together’.32 The sincerity behind this belief of normal SSDF militiamen was quite plausible, considering there had been no broad, public campaign by the GoSS or the UN to disperse copies of the CPA to the general population, or to target SSDF soldiers with public awareness campaigns.33 Thus, it is likely that SSDF officers, who could be expected to understand the CPA, misinformed their soldiers of the CPA’s provisions, namely by falsely claiming that the SSDF was recognized and allowed to remain armed and in control of territory. The second primary concern of the SSDF that would provoke them to war was if the SPLA tried to relocate SSDF forces to the north. The issue was complicated by the fact that many SSDF officers held formal commissions in SAF. Many officer commissions were given out in 2004 by SAF, in the hope of solidifying relationships between SSDF comman­ ders and SAF when it became apparent that the CPA would be signed. Other SSDF officers, fewer in number, had been commissioned earlier. The primary benefit of joining SAF was to receive a steady salary. While official commissions may have been provided, these SSDF officers still considered themselves to be SSDF rather than SAF, and were fearful of being redeployed to the north since the CPA stipulated that the SAF must do so by around 2008. For this reason, SSDF members who were part of SAF sought to be part of the Joint Integrated Units ( JIU) stipulated by the CPA. The JIUs were units composed of both SPLA and SAF members and intended as a quasi-‘national army’ in the interim period before the 2011 referendum. Membership in them would have allowed the SSDF members to both stay in the south as part of SAF’s official JIU contingent and still receive a SAF salary. A last major motivation for SSDF members was their strong personal dislike of the SPLA’s long-time leader, John Garang. Two issues seem to have especially incensed SSDF militiamen about Garang. First, 52

The SSDF’s commercial and financial interests

The commercial and financial interests of the post-Juba Declaration SSDF can be considered on three levels: first, as commercial motivations for the entire organization, and second and third in the personal financial rationales of normal militiamen and officers. It is necessary to consider these motivations on separate levels as there are substantive differences between them. The principal commercial interests of the SSDF as an organization lay in the exploitation of the oil development opportunities in Upper Nile and Jonglei states. With the Juba Declaration, the SSDF lost control of Unity state, where the major oilfields of South Sudan are located. However, with their forces still spread around parts of Upper Nile and Jonglei states, the SSDF continued to maintain a presence in the second-largest oil-producing area of the south, known as the Melut Basin. In June 2006, the SSDF petitioned the GoS to recognize the contracts that the SSDF authorities had signed with a foreign oil exploration and management company.35 The GoS did not accept this petition, as to do so would have been a major violation of the CPA. However, the intention behind it was quite clear regarding some of the broader commercial motivations of the SSDF. Namely, it wanted the ability to sign contracts and exploit oil resources for the areas under its control. Clause 1 of the petition stated: 53

3 | The South Sudan Defence Force

many SSDF members cited the origins of their armed resistance to the Khartoum governments as going back to the Anyanya II movement in the mid-1970s, and thus considered themselves to be originators of the southerners’ armed struggle of the second civil war. Accordingly, it was all the more galling to them that Garang, who was originally an officer in SAF, formed the SPLA to fight both SAF and ‘secessionists’ from the south, namely Anyanya II. This led to a feeling that Garang personally corrupted the spirit of the original rebellion through his insistence that the SPLA seek a unified Sudan. Second, many SSDF members still bore grudges over the infighting that occurred both within the SPLA and between the SPLA and Anyanya II in the 1980s. During this time many of their friends and relatives were allegedly killed by Garang for opposing his iron grip over the SPLA.34 The forceful, divisive control that Garang allegedly exercised over the SPLA in the 1980s until it split in 1991 left SSDF members – many of whom were in the SPLA at the time – bitter towards him on a personal level and subsequently towards the organization he was so influential in shaping.

The Government of Sudan hereby recognizes the right of self-­ determination by South Sudan Defence Forces. As such, the GoS will recognize any and all contracts and agreements signed and executed by the appropriate authorities of the SSDF for the areas of Southern Sudan that the SSDF exercises control.

The second clause further explains the SSDF’s purported legal basis: ‘These rights of the SSDF are expressly stated in the peace agreements signed between the GoS and SSDF, namely the Khartoum and Fashoda Peace Agreements of 1997.’ This attempt by the SSDF to induce the GoS to renew its commitment to the 1997 agreements built upon earlier threats to other oil companies, specifically White Nile Oil and Total, barring them from exploring and drilling for oil in a concession area known as Block B1. A SSDF press release dated 5 March 2006 encouraged these oil com­ panies to voluntarily leave Block B1 as ‘failure to comply with this notice could have serious consequences similar to the current situation in the oil rich Delta region of Nigeria’.36 This represented both an actual threat – in the Nigerian case, disruption of the oil industry by militias came to threaten the country’s whole economy – and an interesting case of ‘learning’ from other militias. The same SSDF press release also argued that the foreign oil firm it intended to partner with had signed contracts guaranteeing that the SSDF would ‘secure the territories to allow the company to operate in a safe and secure environment’. This would include ‘protecting the area from outside forces, including the SPLA and GoS forces, and other foreign companies that do not have agreements with the SSDF’. The SSDF’s aggressive public attempts to assert some ‘sovereign rights’ over territories that it ostensibly controlled, and its relationships with foreign oil interests, indicated its broader two-pronged strategy, which combined commercial interests with political ones. The first part of this strategy was that, as a military force, it claimed the right to exclude the GoS and the GoSS from encroaching on to its territories. The second prong of the strategy, political in nature, was to continue to publicly claim the sanctity of the Khartoum Agreement as legitimizing its continued existence as a political and governmental entity in South Sudan, with the legal right to sign and manage oil contracts, despite the provisions of the CPA, which forbid such third parties from doing so.

Financial interests  It was common for SSDF members to assert that their support from SAF was not monetary in form. This seems, publicly 54

55

3 | The South Sudan Defence Force

at least, to have been a matter of pride, since to accept cash would have implied that they were ‘mercenaries’ for SAF.37 Nonetheless, it was widely known that SSDF officers, notably commanders, did receive cash payments from SAF.38 Ex-SSDF officers have explained that they consistently received cash payments while allied with SAF. It was up to them to use this money to support the militiamen under their command. In addition to cash, other forms of payment for continued support were provided to officers, notably the provision of houses and cars to senior SSDF commanders in Khartoum, and to a lesser extent in the major towns in the south, such as Malakal. One of the most blatant examples of SSDF members receiving cash payments from SAF was during the retribution attacks that were undertaken by SSDF forces against Shilluk areas in Upper Nile state after one commander, Lam Akol, defected from the SSDF back to the SPLM/A in late 2003. Many of the predominantly Nuer SSDF militiamen who participated in attacks in Shilluk areas in early 2004 were paid in Sudanese dinar notes by SAF. This was especially apparent because 2,000-dinar notes were used, and as they were a brand-new issue, they had never been seen in South Sudan before. The only people who had them in­ itially were the Nuer militiamen spending their newly acquired notes in Malakal after being paid by the resident SAF Military Intelligence officers. The notes hence acquired the local nickname of ‘Nuer money’. That SSDF commanders received financial payments directly from SAF also seems to have helped convince some of the hold-out SSDF officers to remain behind after the Juba Declaration.39 When the Juba Declaration was announced, there was some mutinying among SSDF forces as parts of units refused to join the SPLA, and rallied behind previously secondary commanders to remain as SSDF units.40 The r­ ationale of these secondary officers, newly promoted to commanders after their predecessors joined the SPLA, was that they would then be in a position to directly receive payments from SAF. Overall, those SSDF commanders who were able to maintain consistent relations with SAF, and hence receive a steady cash income of their own, had achieved the most prominence because of it.41 A good example of this was the Murle SSDF commander Ismail Konyi, who has had a very long-term relationship with SAF going back to the 1980s. Konyi had a very loyal following among his militiamen, mainly because he was consistently able to support them.42 However, the Murle militiamen of the SSDF appear to have been the exception, and while other SSDF commanders did receive cash payments from SAF, their militiamen do not appear to have received regular cash payments themselves.43

The SSDF and the role of identity politics

South Sudan has a very diverse population dominated by the two largest tribes, the Nuer and the Dinka. These are followed in size by a range of other tribes, including the Fertit, Murle, Anuak, Maban, Shilluk and a large number of even smaller tribes loosely grouped together as ‘Equatorians’. It is beyond the scope of this chapter to detail the role that ethnicity has played in national and regional politics, but it is important to note a few crucial aspects of identity politics as they specifically relate to the SSDF. The first dynamic of note was the perception held by members of the SSDF of the alleged Dinka bias within the SPLA. There has long been tension between the Dinka and Nuer, whose historical relationships were largely defined by each raiding the other’s cattle. The tension between them translated into the post-independence politics of the south. The SPLA was led largely by Dinkas, while the SSDF had mostly Nuer leadership. The perception existed, despite the fact that Dinka and Nuer are present in significant numbers in both organizations, that each had its ethnic base in either one of the tribes. One of the most common sentiments heard from SSDF members was that they objected to the ‘Dinkanization’ of the south, and believed the SPLA to be a partisan ‘tribal militia’ serving the purposes of the Dinka.44 The strong ethnic divide between the SPLA, which does still have a predominantly Dinka senior officer corps, and the SSDF, which had a primarily Nuer base, was widely known. This ethnic rivalry between the two organizations also carried political overtones in terms of which tribe should ‘naturally’ lead the south. Many Nuer SSDF members claimed that the notion that the Dinka are the largest tribe in the south, and hence have a claim to southern leadership, is a conspiratorial lie. This insistence on the Nuer being sidelined by a Dinka conspiracy was a common, and seemingly increasing, refrain of the SSDF’s many Nuer members. This could be seen in the SSDF’s press releases claiming that the ‘Nuer nation’ was being exploited, especially since its territory was where the oilfields were located yet it arguably did not share in the oil revenues sufficiently.45 The second identity dynamic that was crucial for the SSDF relates to the notion of being ‘southerners’. It was very upsetting to SSDF members that they were considered ‘southern sell-outs’ by the SPLA. The argument that they were simple proxy militias for the GoS was especially offensive because of its connotations that they were subservient to the GoS and hence working towards its goals. Despite their partnership with the GoS through the Khartoum Agreement, the SSDF members 56

SSDF relationships with external actors

This section of the chapter details the relationships of the SSDF with external actors, by which are meant those that come from or functionally operate outside South Sudan.

Government of Sudan (GoS)  The GoS had a long history of ‘divide and rule’ tactics for fighting its civil war in the south through allied mil­ itias. Even prior to the Khartoum Agreement, the GoS reached out to non-SPLA armed forces, such as the individual tribal militias before they united under the SSDF umbrella, by supplying them with arms, ammunition and other logistical needs. With the signing of the CPA, and even with the subsequent defection of the majority of the SSDF to the SPLA after the Juba Declaration, the GoS was still widely accused of continuing its support for the SSDF (UNMIS Mediation Briefs). For instance, the Special Representative of the UN secretary-general, Jan 57

3 | The South Sudan Defence Force

saw themselves as being ‘true southerners’, as they were seeking full independence for the south, while they conversely argued that the SPLA was the actual ‘sell-out’ to southerners because it had long fought for a unified Sudan. Especially embittering to many SSDF was to be called ‘Arabs’ by their southern opponents. As one SSDF soldier remarked: ‘the SPLA tries to discredit us by calling us “Arabs”’.46 The question of ‘southerner’ versus ‘Arab’ was one of heartfelt emotion and subsequently of strong political value. In addition to the driving nature of the Dinka–Nuer relationship, owing to their relative populations in South Sudan, there were other important issues of identity politics. Other smaller tribes have long felt the need to protect themselves from the dominance and aggression of the more numerous Dinka and Nuer. An example of this was the formation of the Equatorian Defence Force (EDF) in 1995, which was largely formed to protect Equatorians against abuses committed by the SPLA. Furthermore, the Murle have long sought to isolate and protect themselves from their larger neighbours. Their tribal militia, the Pibor Defence Force (PDF) of Ismail Konyi, had a strong identity centred on notions of defending the Murle people. It is interesting to note that while members of the PDF consider Ismail Konyi to be an SSDF general (as well as in SAF), he was most importantly a commander of the Murle.47 Overall, identity politics played a very significant role in the internal relations of the SSDF, as well as in its relations with other actors in South Sudan’s complex politics, something that will be discussed further in the next chapter.

Pronk, in reference to SAF support for the SSDF, said: ‘Nobody has the freedom to continue as a third army. If they continue fighting then they are just warlords […] I understand some forces are still giving them assistance and I made an appeal to them not to do so anymore.’48 The post-CPA support from the GoS to the SSDF came in the form of cash payments to individual commanders, transfers of arms and ammunition, and provision of transportation resources and basic food­stuffs. SAF also maintained a barracks for the SSDF in Khartoum called Lokondo, for which it paid the rent and provided food.49 The GoS, through SAF Military Intelligence (MI), had never wanted the SSDF to coalesce into a cohesive organization whereby it might more easily turn into a threat to the GoS, and thus purposely kept it as a loose umbrella organization with little real command and control over its component militias (Institute for Security Studies 2004: 5). For example, while the SSDF ostensibly had a unified command, Khartoum always bypassed this to liaise and interact with individual SSDF commanders and units in the south, and hence fragmented the SSDF by maintaining individual logistical and financial arrangements.50 As the SSDF’s militias were largely scattered in pockets around South Sudan, supplying their logistical needs was much harder to accomplish after the signing of the CPA, because of the increased SPLA presence. SAF never entirely denied that it maintained logistical support for SSDF militias.51 However, it did deny that it provided arms and ammunition, and insisted instead that it only provided foodstuffs because it was in the process of integrating SSDF forces into the SAF or waiting for them to be demobilized by the GoSS. To some extent this was plausible, given that there were multiple pockets of SSDF forces maintaining individual relations with SAF MI. For instance, SSDF forces interviewed in Ketbek, Upper Nile state, noted, somewhat bitterly, that SAF was not presently providing them with arms and ammunition.52 However, other SSDF units, such as those of Gabriel Tang in Fangak, Thomas Mabior in Dolieb Hill, or the SSDF units in Adar, likely did still receive significant arms and ammunition supplies from SAF. As mentioned previously, the SSDF unit best supported by SAF was the PDF of Ismail Konyi.53 Frequent official complaints were made by the SPLA to this effect, and SAF officially and consistently denied the allegations but did concede that Ismail Konyi, the PDF commander, received orders and maintained an intensive relationship with SAF commanders based in Juba.54 That the SSDF had no significant source of support other than through SAF was widely acknowledged. The SSDF admitted as much 58

Oil companies  As mentioned previously, the SSDF had already signed a contract with one foreign company in 2004 to explore for and manage oil in SSDF-controlled areas. Through its press releases the SSDF notified concerned parties that it would permit only SSDF-approved companies to develop oil resources in SSDF-controlled areas. A lot of this can be dismissed as pure rhetoric, since the SSDF controlled only small isolated pockets of territory, not areas comprising entire ‘blocks’ of oil concessions. Exceptionally, however, there was a significant presence of SSDF forces around the major oil-producing town of Adar, which pumps out of the Melut Basin of northern Upper Nile state. GoSS officials there accused the resident oil company, Petrodar, of being supported by the local SSDF forces surrounding Adar.58 There was no formal relationship between the SSDF and Petrodar, apparently only an informal, local ‘understanding’, but the mere presence of the SSDF forces prevented the SPLM county commissioner from maintaining a GoSS presence. Officials of the ruling party in Khartoum, the NCP, 59

3 | The South Sudan Defence Force

itself.55 Compared with armed groups in some other African countries, the SSDF had no access to natural resources – such as diamonds, lumber or gold – that could be sold to raise money. Sometimes individual militiamen sold cattle to buy arms or ammunition, but this was not enough to sustain the SSDF more broadly. The SSDF had a longer-term interest in oil production, but no actual source of income from oil exploitation. As one local political commentator noted, the pockets of SSDF units truly were ‘satellites of SAF […] [and] would collapse without SAF support’.56 The rationale for the GoS, through SAF, to support the SSDF was twofold. First, it wanted to create general instability in the south, so there was little to no development to use as a political argument and leverage against the SPLM. Second, the GoS wanted to create instability so that it could argue for the redeployment of SAF forces out of the south to be delayed. And indeed, SAF, as of 9 July 2006, was supposed to have withdrawn half of its forces to the north, but had withdrawn only 35 per cent.57 It officially explained that it was behind schedule because it was worried about the continued presence of ‘other armed groups’, namely SSDF units, in the south, and felt its presence was needed (conveniently) for security reasons. It was notable that among some SSDF militiamen there was a feeling that SAF was giving them barely enough support to maintain themselves, essentially at a subsistence level, in the belief that if a war were to start again in the south, the remnant SSDF forces could quickly be expanded to fight the SPLA. In the meantime, the political and financial costs of this support would be mitigated.

l­ iaised directly with the oil company, bypassing the GoSS/SPLM officials completely. The overall significance of the Adar SSDF was that it was still in control of an actual oil-processing area, compared with other pockets of SSDF that did not directly control major oil-pumping areas. Adar was also a CPA-sanctified assembly point for SAF, so it was easy for them to resupply and maintain SSDF forces there. SAF could also easily send SSDF forces that had previously been based in Khartoum to Adar, since it is near the north–south border. Relationships with other domestic actors

Intra-SSDF relations  Even with official consolidation through the Khartoum Agreement, the SSDF never had much internal coherence because of SAF manipulation and the fractured nature of southern politics. As a result, SSDF units had little day-to-day contact with one another. This was especially the case after the signing of the CPA and the conclusion of the Juba Declaration, as the leftover SSDF units had been isolated into pockets of localized control. What interaction did occur between units was mostly between very senior officers in the larger cities, notably Malakal and Khartoum. The exception to this was the SSDF militiamen within Malakal. Many units from around Upper Nile and Jonglei states had sent groups of troops to Malakal periodically in the past, since Malakal was an SAF garrison town during the civil war. After the war, and with subsequent new arrivals from the field, there were a good number of militiamen from the various SSDF militias in the town. Even in Malakal, however, the relationship was largely fraternal and friendly rather than regimented. Individual militiamen were under the direct control of their respective commanders and did not respond to an overarching ‘SSDF chain of command’ or interact with other units in a notably coordinated manner. The SPLA  The SPLA and SSDF had long been adversaries, with the SPLA frequently accusing the SSDF of being simple proxies of SAF. This overarching viewpoint of the SPLA concerning the SSDF was articulated in an internal SPLA communiqué from early 2006, which noted that the GoS’s ruling party, the NCP, had changed from ‘confronting us directly into [confronting us through] a proxy war’.59 The communiqué further stated: Our partners [the NCP in the Government of National Unity] are still holding the other armed groups and using them to fight the Government of South Sudan (GoSS). And the main aim in this is to make the South ungovernable […] Their hopes and whims are to make the Government 60

Given the SPLA’s concern over the NCP’s intention to foster instability in the south through its continued support of the SSDF, it implemented a three-pronged strategy for dealing with instability relating to the SSDF. The emphasis of this strategy was on isolating and minimizing the SSDF’s presence in the south. The first facet of the SPLA’s post-CPA security strategy was to begin to rationalize the SPLA and broaden its inclusiveness. The most important aspect of this was the Juba Declaration, which brought the bulk of the SSDF into the SPLA, starting in January 2006. It was the foundation of the GoSS’s overall security strategy to absorb as much of the SSDF as possible into the SPLA, and then rationalize the entire force into a modern, professional army, rather than as a force composed of what were still essentially localized, segregated militias employing guerrilla tactics. The second facet was to isolate the SSDF into small pockets but otherwise to avoid direct military confrontations. It was notable that up to August 2006 there were relatively few open battles between the SPLA and the SSDF, with the only significant ones being in north-central Upper Nile state after an assassination attempt by the SSDF on a county commissioner and a fight between the forces of SSDF commander Gab­ riel Tang and the local SPLA in the Old Fangak area of north-western Jonglei state. The SPLA maintained that it did not have the authority or mandate to forcefully disarm the SSDF, which it claimed should be done peacefully by the GoSS’s DDRR Commission, which had been very slow to form.60 Apparently, the SPLA never had any plans to forcefully disarm the SSDF.61 Pagan Amum, secretary-general of SPLM, articulated this point: ‘According to the CPA, they should be disarmed by force [now, since they missed the January 2006 deadline], but that is not our intention at this stage.’ This view was echoed by GoSS president Kiir, who said force might be used to disarm SSDF units if they did not voluntarily join SAF or the SPLA, or resisted DDRR efforts by the GoSS DDRR Commission. Overall, the SPLA seemed to be resisting the temptation to militarily engage the SSDF pockets, because it was not worth the tensions it would provoke within the SPLA with ex-SSDF members (namely Matip) and subsequent frictions with Khartoum over the CPA’s implementation. Most importantly, there were options other than military force available to the SPLA for confronting the continued presence of the SSDF. The 61

3 | The South Sudan Defence Force

of South Sudan failed in administering the Southerners, so that at the end of the day, the CPA get destroyed before the six years interim period.

rhetoric of the SPLA did, however, come with the caveat that it might eventually act with force if the SSDF continued to resist joining the SPLA or SAF, or refused the GoSS DDR Commission’s programming once it was established and running.62 Rather than confront them militarily, the SPLA strategy was to induce individual SSDF commanders to defect. The SPLA had been active in reaching out to specific commanders in order to ‘poach’ them for the SPLA. An example of this was the SPLA’s attempts to recruit the PDF commander Ismail Konyi, who was offered a seat in the GoSS parliament on an SPLM ticket, but by August 2006 had yet to decide whether to join the SPLM.63 While there were some successes, mostly among more junior officers, the vast bulk of SSDF soldiers resisted SPLA overtures after the Juba Declaration. Furthermore, the SPLA also sought to avoid direct military confrontations but still hinder the SSDF by attempting to close down their supply lines from SAF. One of the major hopes of the SPLA was that when SAF was forced through the CPA provisions to withdraw its forces to the north (other than its share of the JIUs), the SSDF units would disappear because they were so dependent on SAF. This was a key reason why the SPLA avoided direct military confrontations with SSDF units. The third facet of the SPLA’s post-CPA strategy was to disarm major concentrations of heavily armed civilians in the south, notably the village militias of the so-called White Army in Upper Nile and Jonglei states, and major disarmament exercises were undertaken in early and mid-2006, as will be discussed in the next chapter. The goal of this effort was to prevent the SSDF from agitating more broadly against the GoSS, for instance by using the White Army militias as proxies. A major worry of the SPLA during the disarmament exercises was the ability of White Army militias to acquire ammunition, either from their informal, fraternal relationships with individual SSDF commanders supplied by SAF, or by going to Ethiopia to buy ammunition from Ethiopian rebels or well-armed communities.64 A major challenge of the disarmament of the White Army militias in fact was the catalysing role that individual SSDF militiamen played in encouraging them to resist the SPLA’s disarmament exercises. Conclusion

This review of the SSDF has highlighted key traits and dynamics of a militia outside the formal peace process in Sudan. Following the CPA, and catalysed by the Juba Declaration, the SSDF was left isolated in small pockets in the south, notably in Jonglei and Upper Nile states. Despite 62

63

3 | The South Sudan Defence Force

this, it remained a significant armed force and consistently maintained an intent to stay as a fighting force until southern independence was assured, preferably through the 2011 referendum but in the longer term if necessary. The unwillingness of hold-out SSDF members to join the SPLA was driven largely by the reality that the SSDF was still in physical control of certain parts of South Sudan, albeit small and isolated, centred on individual commanders with strong, long-felt animosities towards the SPLA. With SAF’s continued military and logistical support, the SSDF retained a belligerent attitude towards the SPLA and the GoSS more broadly well into 2006. In response, the SPLA implemented a multifaceted approach to counter the SSDF. The initial focus had been to reach out to the SSDF to integrate it into the SPLA, as stipulated by the CPA. This was done successfully to a large extent through the Juba Declaration. Second was the attempt to further isolate the SSDF into small pockets, but otherwise to try to avoid direct military confrontations. The SPLA attempted to do this by blocking supplies to the pockets as well as trying to induce individual commanders to defect, thereby gradually eroding the SSDF-controlled areas in the south. In a notable coup, the SPLA finally managed to persuade Ismail Konyi to join it in October 2006. Overall, the strategies applied by the SPLA against the SSDF in 2005 and 2006 proved successful, having left it marginalized and widely unable to agitate against the GoSS. Largely as a result, the SSDF officially ceased to exist in June 2007 when its military wing collapsed and its political wing transitioned into the new South Sudan Democratic Front, a national political party. However, despite the SSDF’s official demise, several remnant militias continued to maintain small, localized presences, notably those of Gordon Kong and Gabriel Tang-Ginya in ­Upper Nile state (Small Arms Survey 2008: 3). These small remnants have largely been left untouched by the SPLA, which has been content to leave them to wither further. However, as the implementation of the CPA climaxes around the 2011 referendum on independence, the ‘SSDF’ still has the potential to play the role of major spoiler to peace and reconstruction in the south. The major reason for this is that, despite being a major achievement, the Juba Declaration that brought the bulk of the SSDF into the SPLA has never been fully implemented. The integration of the SSDF militias into the SPLA, while officially undertaken, has in fact remained limited in scope, leaving the old militia units largely detached from the larger SPLA. As recently as late 2009, for instance, General Paulino Matip was

complaining that his SSDF veterans were being sidelined by the broader SPLA (International Crisis Group 2009a: 14). Additionally, within the communities where the SSDF had been strong, the underlying resistance histories and ethnic tensions that were so encouraging to the SSDF’s initial formation and sustainment have largely been left lingering. Given these challenges, and within the context of a national conflict that has stretched out over decades, indeed between generations, the danger is that those leaders and communities that had been central to the SSDF will simply re-emerge during a future crisis, be that a return to north–south war or internal conflict within a newly independent South Sudan. The case of the SSDF highlights opportunities and challenges for responding to large, state-sponsored militias. On the one hand, the GoSS and the SPLA showed that political negotiation and reconciliation in the form of the Juba Declaration can be immensely powerful tools for responding to militias. That the vast bulk of the SSDF were quickly removed as a major threat to peace following the signing of the CPA was an undeniable achievement. Furthermore, the patience of the SPLA in isolating and marginalizing the remnant SSDF pockets deserves due commendation. While many would have preferred to directly attack such outstanding spoilers, the tactic of letting these small pockets gradually wither away proved to be a more effective response ultimately as it did not disturb the broader efforts of reconciliation started through the Juba Declaration. On the other hand, as is so often the case, the slow and limited reform of the security sector has left the SSDF veterans a potential medium for agitation should there be a return to war. Lastly, while there has been an effort by many southerners to pursue calm in the run-up to the 2011 referendum, there is no guarantee that such calm can be maintained should the south vote for independence. Considering South Sudan’s long and tortuous history of internal conflict, the vestiges of the SSDF could easily rise again.

64

4 | The White Army militias of South Sudan

As the previous chapter highlighted, militias maintained and manipulated by the state to counter a strong insurgency have played a very prominent role in Sudan’s extended civil war, as well as the peace process started in 2005. However, in addition to state-sponsored militias such as the SSDF, small localized village militias have also played an exceptionally important role in the dynamics of the conflict in South Sudan as well as the subsequent peace process. In this regard, an important grouping of militias in South Sudan was the so-called ‘White Army’ militias. These militias were groupings of armed civilians, mostly youth, which during the course of the long civil war, notably during the 1990s, coalesced into village-level formations. The presence of numerous White Army militias created a significant degree of instability in Upper Nile and Jonglei states. This continued even after the signing of the CPA, and their militant presence was one of the major security challenges for the nascent autonomous GoSS. The purpose of this chapter is to provide an empirical mapping of the White Army militias in South Sudan and to highlight policy considerations for the demilitarization of similar localized, youth-driven militias more broadly, based on the lessons learned from South Sudan’s experiences in 2006. Specifically, the chapter focuses on the disarmament exercise of White Army militias in Upper Nile and Jonglei states from January to August 2006, primarily by the SPLA but with parallel efforts by the UN and NGOs. Overall, the demilitarization of the White Army militias of South Sudan on the one hand highlights the complexities inherent in managing a disarmament process aimed at militias and, on the other hand, illustrates the possibilities for immediate successes inherent in a well-considered campaign. However, as the security situation in South Sudan has at times been worse than that in Darfur since the CPA was signed, the lesson of White Army militias is that while specific militias are fairly easy to disarm, achieving broader peace and stability is much harder.

65

Overview of White Army militias

White Army history and presence in South Sudan ‘The White Army is not real army and aren’t proper soldiers. Anybody with a gun can claim to be ‘White Army’ soldier. They are just local p ­ eople who have guns and it’s just name given to gunmen in ­villages.’  (Lou Nuer elder)1

As described above, the ‘White Army’ was not a single, coherent armed force.2 Rather it was a generic name given to bands of armed civilians, primarily male youth, who resided mostly in Jonglei and Upper Nile states in South Sudan.3 White Army militiamen did not serve as full-time, dedicated soldiers but rather intermittently in collectives of armed civilians, who undertook cattle raids against their neighbours, occasionally banditry, and generally carried on with their everyday lives armed and militant. Often much of the fighting was actually between the various constituent groups that made up what are broadly known as White Army militias. However, at other times they did temporarily unite to fight larger, commonly identified enemies. The armed youth in question come from Dinka as well as Gawaar and Lou Nuer communities. The UN estimated that at the beginning of 2006 there were ‘20–30,000 mainly youth’ members of the White Army militias who were ‘loosely organized under the leadership of chiefs’.4 Violence associated with the White Army militias had traditionally been at its peak during the dry season, when cattle are congregated along the Sobat and Nile rivers, thus leading the various militias into greater contact with one another as well as other communities. Historically, the seasonal movements of people with their cattle were relatively less violent, but over decades of war – leading to high levels of armaments accumulation and the near-complete dissolution of governmental ­authorities – they became more so. The reason why the White Army militias were taken seriously and considered in the immediate post-conflict reconstruction policies of South Sudan was that their presence severely hindered the ability of the nascent, exceptionally weak GoSS to exert effective governmental control over the south, namely in Jonglei and Upper Nile states. While there was no single army, the presence of many bands of militiamen who operated largely without restraint in their home areas meant that they needed to be disarmed to allow for effective governance to be implemented. As will be discussed in greater detail below, attempts by the SPLA to disarm White Army militias, specifically from the Lou Nuer community in the first half of 2006, resulted in a great deal of violence. 66

History of disarmament exercises undertaken from January to July 2006  In response to the presence of White Army militias, an exercise was begun by the SPLA in December 2005 to disarm them. ‘Peace meetings’ were held with the respective communities by the SPLA, to encourage the peaceful disarmament of the White Army militias, something that was overall agreed upon by the parties. However, in January 2006 a SPLA disarmament contingent that arrived in western Jonglei state was ambushed by White Army militias, resulting in the deaths of up to three hundred SPLA soldiers, forcing the SPLA to temporarily retreat.5 While the SPLA momentarily withdrew from western and central Jonglei state, it proceeded with disarmament exercises in Upper Nile state, which occurred relatively uneventfully. Intermittent fighting between the SPLA and White Army militias continued for the next several months and peaked again in central Jong­lei state in March and May when a unit of 3,000 SPLA undertook forceful disarmament in the Yuai area.6 During this forceful disarm­ ament exercise, White Army militias and SPLA forces were engaged in extensive combat. Furthermore, extensive pillaging of local communities was undertaken by White Army militias during the mayhem and the SPLA selectively burnt hundreds of huts that belonged to White Army members who were resisting them. The fighting eventually stopped after the militias suffered significant losses, ran out of ammunition, and subsequently retreated northward.7 Following the violence in western and then central Jonglei state, efforts were made to ensure a peaceful disarmament exercise in eastern Jonglei state, notably in and around the town of Akobo. During July 2006, a peaceful disarmament exercise was undertaken there by the SPLA and its partners. The county commissioner of Akobo county had given a time frame of 15–30 July for armed civilians in the county to voluntarily disarm, after which it was publicized that they could be forcefully disarmed by the SPLA.8 The commissioner had been keen that instead of the large force of SPLA soldiers, such as that used in central Jonglei, the SPLA should undertake a voluntary disarmament using smaller dispersed forces. It was hoped this would lessen tensions with local communities and prevent the SPLA presence from becoming a burden on already food-insecure local communities. The disarmament exercise was undertaken through the manning 67

4 | The White Army militias

This resistance to disarmament was conducted in a fairly coherent manner, enough to inflict significant casualties upon the SPLA numbering well into the hundreds.

of a dozen or so ‘disarmament posts’ where groups of approximately ten SPLA soldiers, supported by ‘deputized’ armed civilians, collected guns that were brought in by civilians in the vicinity. The UN provided logistical and technical support through their UNDP DDR Unit, while UNMIS Military Observers and Force Protection soldiers provided a general presence and monitoring of the situation. Furthermore, the UNDP promised that those who turned over weapons would later receive compensation for each gun in the form of a packet of goods, such as fishing and mosquito nets and other small items. More than a thousand guns were collected during the exercise.9 Overall, it can be summarized that a key goal of the SPLA’s post-CPA security strategy had been to disarm major concentrations of heavily armed civilians in the south, namely the White Army militias in Upper Nile and Jonglei states. The SPLA has never been deluded into thinking its initial disarmament exercises could collect absolutely all the firearms, or even most for that matter, held by civilians in Jonglei and Upper Nile states. Rather, the strategy had been to end attacks by large formations of White Army militias, such as the attacks by 800–1,000 militiamen on Ulang and other Sobat corridor communities in March and April 2006.10 Following the aforementioned major disarmament exercises, it was hoped by the SPLA that a majority of civilians would have been disarmed, and thus there would only be small remnant groups of White Army militia of perhaps ten to fifteen people. These smaller groups could then be disarmed gradually through SPLA bush patrols. The SPLA argued at the time that because of the problem of cattle raiding it would maintain a security presence in villages, but after complete disarmament – e.g. of the Murle in southern Jonglei state – all SPLA troops would go back to their barracks.11 In 2006 this had been the SPLA’s strategy in any case, and the balance of the chapter will review its efficacy, the role of other actors in it, and continuing challenges for engendering peace and stability in its aftermath. Emphasis will be placed on contrasting the SPLA’s forceful disarmament exercise from January to May 2006 in western and central Jonglei state with the voluntary disarmament exercises in Akobo county in eastern Jonglei state in July 2006. An empirical mapping of White Army militias

White Army militias’ motivation  In order to fully understand the White Army militias, it is necessary to undertake some ‘cognitive mapping’ of its members, starting with a review of their motivation. An analysis of the motivations behind the White Army militias must be broken 68

Rationale for owning a gun and belonging to a White Army militia ‘If you have a gun, you believe you can get something. If you don’t have a gun, you don’t feel like you have anything at all.’  (White Army youth)12

A starting point for understanding how the White Army militias came into existence is a consideration of the general anarchy that has existed for decades in South Sudan. The civil wars fought in the south have almost entirely prevented the presence of effective governance. When combined with the proliferation of small arms and the fractured nature of the tribal systems of the south, an environment was fostered that was very conducive to the development of small, localized militias. One exWhite Army member disarmed in Jonglei state described the rationale for its existence: ‘The White Army only existed because there was no government.’13 With no presence of a coherent government, and the constant flux of assorted other larger armed groups operating in the south, namely the SSDF and SPLA, many youth felt free to join their local White Army militia as a form of self-defence.14 This argument by White Army youth cannot be discounted as the need for self-protection by local populations from the larger rebel movements, such as the SPLA, which developed a reputation for harassing and exploiting civilians, and from other local militias and armed groups such as bandits, was a very significant one. In addition to this environment of anarchy and the need for selfprotection, the creation of local militias had decidedly commercial rationales as well. Cattle raiding had long been an important aspect of the socio-economic systems of South Sudan. Ownership of a firearm and membership in a militia greatly facilitated the ability of otherwise exceedingly marginalized youth to undertake cattle raids. The existence of general anarchy, compounded by a breakdown in traditional conflict resolution mechanisms, left what one southerner called a ‘war culture’ in which people ‘use guns to enrich themselves’, mostly through cattle raiding.15 This ability to generate some income was significant given that there were few other economic outlets available to youth in the war-torn south. Furthermore, the acquisition of cattle is significant culturally in the south to Nuer society – for instance, cattle are central to dowries, which must be supplied by the groom in marriage ceremonies. The fact that the civil war in the south went on for so long meant 69

4 | The White Army militias

down into two levels – first, the impetus for owning a gun and joining a White Army militia; and second, the rationale for resisting disarmament exercises through violence.

that an entire generation knew little other than a wartime existence, so that the ‘only things in the minds of the youth is a war culture and a gun’.16 The sense of frustration and isolation among rural youths was also very strong. Ownership of a gun and membership in a local militia was very empowering to youths who otherwise felt that their options were limited.17 Adding to this sense of empowerment that being armed gave White Army youth was the belief that through their membership in militias and ownership of firearms, they were guaranteed recognition in war as well as a post-conflict environment that otherwise would have ignored them completely.

Rationale for resisting disarmament ‘You can’t collect my gun when you are my enemy.’  (White Army youth)18

That the resistance by White Army militias was so violent and broad was a surprise to the SPLA, which experienced significant casualties when it began its disarmament exercise, precisely because it wasn’t expecting so much resistance. There were three factors behind the strong resistance: (i) a fear and even hatred of the SPLA; (ii) civilians in Jonglei and Upper Nile states worrying about a need for self-protection generally; and (iii) a dislike of the very idea of being disarmed. As regards the first consideration, despite the fact that it was the largest rebel movement in the region, the SPLA was not well regarded in much of the south. This was especially so in the ‘Nuer belt’ of Upper Nile and Jonglei states, where much of the White Army militias were resident. The Lou Nuer, who formed the bulk of the White Army militias, were also very strongly represented among other anti-SPLA actors in the south, notably the SSDF described in the previous chapter. The result of this was that Lou generally were still very wary of the SPLA, because it had been fighting with significant numbers of them for many years. This continuing wariness of disarmament was fundamentally a matter of trust – many White Army militiamen felt that they simply could not trust the SPLA to be the ‘ultimate provider of security’ in South Sudan.19 This overarching mistrust and dislike of the SPLA was compounded by the presence of several SSDF commanders who agitated against the SPLA and coordinated much of the White Army militias’ resistance in western and central Jonglei.20 As for the second reason, the resistance by the White Army militias was so intense because of a fear that disarmament would leave them exposed to attacks by opportunistic neighbours. Cattle raiding was still common throughout the south and was often a violent undertaking. 70

71

4 | The White Army militias

The Lou Nuer areas, where the White Army militias put up the most significant resistance, are surrounded by neighbours who have a history of raiding the Lou Nuer for cattle, as well as of general pillaging. The Lou have traditionally had animosities with the Dinka and since the early 1990s have had a pronounced conflict with the Jikany Nuer. Perhaps most of all, the Lou were very fearful of their aggressive Murle neighbours to the south. As one Lou elder exclaimed, ‘The Lou have a dangerous life.’21 The resistance to the disarmament by White Army militiamen was strongest where it was felt that to lose their weapons to an organization they didn’t trust, i.e. the SPLA, was going to be greatly compounded by being newly defenceless and surrounded by enemy neighbours who still had their own weapons. The third reason for resistance to disarmament by the White Army militias in central Jonglei was a mixture of personal considerations on the part of its members. Many southerners were still fearful of the peace agreement falling apart, and as most had been carrying guns for a long time, they did not want to give them up quite yet since they were still anxious about a peaceful future.22 This fear was compounded by their strong dislike of the SPLA, since they did not know how the SPLA would actually treat them. Another facet of the dislike of disarmament felt by the youth was that they firmly believed that they were due financial compensation for their weapons since these had been mostly acquired with their personal resources. For instance, an AK-47 assault rifle had likely cost anywhere from five to seven cows, while a PKM machine gun could cost from ten to twelve cows.23 A further facet was the youth’s fear of making a living without a gun, which was necessary for cattle raiding since cattle, as mentioned previously, were essential to pay wedding dowries and acted as a form of hard currency in the war-devastated south. While many Lou Nuer feared being raided by their neighbours, they were as fearful that they themselves would be incapable of raiding should they be disarmed, one of the main reasons they joined the militias in the first place. Lastly, it was articulated by many youth and observers that owning a gun was ‘much more exciting’ than not owning one. As one youth explained, the ‘main cause of all [our] problems is idleness’.24 Owing to a sense of frustration and isolation, many White Army youth wanted to keep their guns because to lose them would mean an even more marginalized, idle life out in rural villages. It is worth concluding the section with a brief description not just of why armed resistance was undertaken but also of what the White Army youths’ hopes were for a better future. This is best illustrated by

a youth who had been disarmed voluntarily in Akobo.25 He noted: ‘We handed over our guns because we want peace. If there is peace there will be development. If people remain with their guns, there can be no development and it can cause problems, like death.’ The youth of the region felt that ‘development means education’.26 Ultimately, the youth wanted to ‘normalize’ their lives, and despite the fact that they had been conditioned by decades of war, there was a broad understanding of the benefits of education and of other lifestyles being possible beyond the violence of war.27 These sentiments were encouraged by the return of IDPs and refugees who had been exposed to new ideas, namely education and cash economies. These sentiments were also encouraged by the pressing reality of death by extreme violence. The youth’s blunt language – e.g. ‘problems, like death’ – highlights an ­interest in exchanging an extreme of life with guns with a hope for a more peaceful, moderated future. The hesitancy, however, came from a fear of the SPLA, their neighbours, and general anxieties about disarmament exercises overwhelming that hope. White Army militias’ operational mode and identity politics

Operational mode  The operational mode of the White Army militias centred on their nature and existence as relatively small, usually auto­ nomous, village militias following clan and lineal lines. Much of the violence involving them was actually between the assorted militias themselves, but there was the possibility of uniting into larger configurations during significant confrontations against common enemies.28 The best example of this was the fighting in May 2006 against the SPLA around Yuai in central Jonglei. There were allegations that White Army militias received supplies from SAF,29 but while this is plausible on a very limited scale, it was more likely that whatever support White Army militias were getting, it was coming through a couple of SSDF commanders, also on a fairly limited scale.30 The resources that the White Army militias used in their resistance to the SPLA’s disarmament exercise were most generally stock on hand, and what could be secured through battle victories and the occasional sale of cattle for armaments. The relationship between the SSDF and White Army militias was an important one, even though it can be considered informal, largely tactical and driven by a common dislike of the SPLA. Both White Army militiamen and SSDF members interviewed noted that there were some kinship affiliations, but that the SSDF and White Army militias never had any systematic, formal relationships.31 However, during the confrontations between the SPLA and White Army militias there was a 72

Identity politics  Identity politics for White Army militias followed similar lines to the broader ethnic divides of South Sudan. In general this meant that the Nuer and Dinka cleavages and historical animosities were important. Among Nuer members of White Army militias, the perception of the SPLA as a partisan Dinka organization was a prevalent and strongly felt one. For the Lou Nuer, it was believed rather conspiratorially that this explained why the GoSS wished to disarm them before their Dinka neighbours farther to the west or their Murle neighbours to the south. The logic behind this belief was that ‘the Dinka SPLA’ wanted to further strengthen their position in the south by collecting all of the arms held by Nuers, who would then be further weakened by attacks from their still-armed neighbours. Conversely, localized ethnic loyalties seem to have played a positive role in some SPLA and White Army interactions. For instance, there were reports of the ‘recycling’ of weapons back to ethnic kin and family by individual SPLA soldiers after they had been collected during the various disarmament exercises.34 Key themes arising from the disarmament of White Army militias in July 2006

It is important to highlight key themes from the voluntary, peaceful disarmament exercise that took place in eastern Jonglei state in July 2006, contrasting with the initial violent efforts at disarmament by the SPLA in early 2006. While disarmament efforts in early 2006 were under­taken by the SPLA acting on its own, the July 2006 efforts were still driven by the SPLA but importantly involved many other actors, notably the UN. This section will firstly highlight some of the actors involved in eastern Jonglei, their roles and the beneficial aspects of them.

Role of local communities ‘Disarmament means a lot. It is peace itself. The armed youth were the main problem, causing all the violence, destruction and hostility. Before 73

4 | The White Army militias

limited presence of SSDF commanders agitating the armed youth to resist dis­armament. Although the SSDF did not participate as organized units with the White Army militias, the presence of smaller groups of experienced SSDF soldiers led by a few notable SSDF commanders seems to have had a major impact on the extent of resistance that was witnessed.32 Of crucial importance was the role these SSDF commanders played in briefly uniting the disparate militias of the White Army to fight what was perceived as their common enemy, the SPLA.33

disarmament we couldn’t gather or visit because we were afraid some violence could happen. Guns were like a disease going around.’  (Akobo Women’s Association leader)35

The role of the local communities in the peaceful disarmament exercise for White Army youth in Akobo county in eastern Jonglei was a very profound one. For the rural communities in general, there was both frustration about the prevalence of weapons-bearing youth and fear of a potentially violent clash subsequently with the SPLA, which meant families and neighbourhoods were keen to see their youth turn over firearms voluntarily. Thus, there were two dynamics pushing for peaceful disarmament: a very real fear of forceful disarmament by the SPLA and the traditional dynamics of communalism prevalent among the Lou Nuer.36 Together, these two forces catalysed a peaceful disarmament exercise in Akobo county, where the youth were actively encouraged to voluntarily disarm by their family and neighbours.37 The communities’ fear of being ‘collectively punished’ (by default) through forceful disarmament by the SPLA was exceptionally strong. In central Jonglei, the fighting between the White Army militias and the SPLA and the resultant mayhem had seen up to a thousand huts burned, up to five hundred youths killed, and tens of thousands of cattle rustled.38 Furthermore, as a longer-term result of the violence, communities faced food shortages as the violence had seen food stores destroyed, livestock stolen or scattered, and crops left unplanted. While the violence had occurred mostly in March and May, the repercussions in terms of hunger for the broader community in central Jonglei were felt most direly during the lean season in July and August.39 There is a general debate in post-conflict recovery circles as to whether to offer incentives for disarmament to communities or to individuals (see, for example, Kingma 2000). In the South Sudan context, specifically in mid-2006, the community approach was the more suitable one because the government, and to a lesser extent the international community in the form of the UN, could hold the whole community accountable. Simply put, if the SPLA came into an area, then the whole community blamed their youth for provoking the confrontation, which meant the community pressured their youth into voluntary disarmament to pre-empt such a confrontation.40 Conversely, if an individual approach were to have been used, there would have been little correlation between the actions of individuals, be they good or bad, and developments for the community as a whole. In addition to these two dynamics, there were some basic demands 74

Role of government  The need to disarm White Army militias was a strong one given the inability of the GoSS, even with the support of the UN and donors, to implement basic governance because the presence of so many armed civilians was debilitating. As the county commissioner of Akobo county noted: ‘Nobody could carry out their administrative duties because there were too many arms; people wouldn’t listen to the government!’44 Additionally, rampant homicide within the Lou Nuer community was resulting from unmitigated banditry, cattle raiding and blood feuding. The deputy governor of Jonglei state said that the government was forced to take proactive action to disarm civilians since among the Lou Nuer alone it was estimated that ‘at least 700 Lou Nuer 75

4 | The White Army militias

of community members, which largely mirrored those of the White Army youth. The two most common ones were security provision once the disarmament had been undertaken, and also that there should be ‘universal disarmament’.41 In an anarchic region where there was both a deep-felt mistrust of the government army (i.e. the SPLA) and a fear of attacks by still-armed tribal neighbours, there was a basic interest in personal physical security. If these fears of the communities could have been alleviated, then their openness to participation in demilitarization programming would have been all the greater. This is largely because when previously isolated communities say they want development and engagement, it really means that they are willing to change by accepting and committing to a peace process in a ‘new’ South Sudan, grounded in a sense that their community is actually part of a larger country, not just an isolated group in a state with no real government.42 Some caution regarding participation by local communities must be observed. A major complaint frequently cited by the SPLA has been their difficulty working with ‘traditional leadership’. SPLA officers noted rather cynically that they were often told to work with ‘the chiefs’ by the UN and NGOs, but the difficulty of this was that they couldn’t rely on Lou Nuer traditional chiefs because it was alleged that they were the ones instigating much of the violence. Under Lou Nuer culture, which has no lineal or even ‘elder’ leadership, White Army members were free to choose their own leaders, i.e. ‘chiefs’, which made working through them difficult. To work through the chiefs, so the SPLA argued, would only encourage them to agitate against the disarmament exercise and hide weapons too.43 In any case, the lesson highlighted by the SPLA was that foreigners need to be careful about simplistic understandings of ‘traditional leadership’, the ‘chiefs’ and their ostensibly positive mediating role in relations with local communities and militias.

[were] killed by infighting’ in 2005 (which means after the CPA was signed and ‘peace’ had begun).45 The government played two basic roles in eastern Jonglei in July 2006. One was brokering agreements between the SPLA, local communities and White Army militias, and the second was facilitating the actual disarmament process. In terms of mediation, the Akobo county government played a key role in mitigating the anxieties of the armed youth. Such youth noted that in their minds a key difference between Akobo county and those farther to the west where violence had erupted was the role of the county commissioners.46 According to White Army youth, in Akobo county the commissioner explained exactly what was going to happen (disarmament protocols and interim security provision steps) whereas in Uror county (the county to the west of Akobo in central Jonglei) ‘the commissioner [there] said lots of things that were confusing’.47 In terms of the role that local government could play in the immediate follow-up to voluntary disarmament, the Akobo county administration had three priorities: the need to re-establish the traditional Lou Nuer compensation method for homicides in order to limit them; the need to bring notably bad criminals to justice for past crimes; and the need to provide basic community services.48 Of primary interest was the need to reinstate the compensation system for homicides, whereby the guilty person would compensate the victim’s family with an agreed number of cattle, thereby preventing ‘blood feuding’ between the victim’s and guilty party’s clans and sub-clans, as well as punishing the guilty person financially.

Role of the UN  The role of the UN in the voluntary disarmament in eastern Jonglei in July 2006 was very conducive to its ostensible success. Whereas the UN had been unable to participate actively in the disarm­ ament exercise in central Jonglei in early 2006, which notably turned violent and coercive, it was able to help ensure that the exercise in eastern Jonglei was peaceful. The UN’s official mandate in South Sudan does not cover civilian disarmament; rather it is for demobilization of redundant SPLA soldiers after a registration exercise. However, the Akobo county commissioner, after seeing the violence in the counties to the west of his, wanted to avoid forceful disarmament by the SPLA, and contacted the UN for help.49 The UN’s role was largely to maintain a presence of Military Observers (‘UNMOs’) and a ‘Long Range Patrol’ of force protection troops in Akobo during the exercise. Additionally, UNDP provided logistical support and technical advice for the exercise, such as 76

Role of NGOs  The role of NGOs, notably indigenous ones, was very significant in the voluntary disarmament exercise in Akobo county. Their role centred primarily on arranging mediation between communities and with the local and state governments as well as with the SPLA. Secondarily, NGOs resident in Akobo county saw themselves as playing a key role in post-disarmament outreach to the newly disarmed youth of the White Army militias. The Nile Hope and Development Forum (NHDF), an indigenous NGO, was instrumental in arranging ‘peace meetings’ in Jonglei state.53 The first peace meetings were held in August 2005 for inter-Akobo reconciliation; the second in November 2005 to create dialogue and reconciliation between the Jikany Nuer of Nasir and Akobo’s Lou Nuer; and the third meeting was held in July 2006 between the Murle and Akobo’s Lou Nuer to agree on voluntary disarmament and reconciliation. At the peace meetings participants agreed to encourage reconciliation between communities through dialogue and that community leaders would mobilize 77

4 | The White Army militias

how to register guns and manage collection sites. It was notable that the UNMIS/UNDP efforts were significantly furthered through the expertise of a few individual staff members who had extensive knowledge of South Sudan affairs, built over years of experience in the south. For instance, some UNMIS/UNDP officials spoke local languages and nearly all had at least several years’ experience of working in emergency situations with local political actors and groups in the south. Such experience and expertise are priceless, and the institutional success of the UN is dependent on them. It is doubtful that a general UN response without the thoughtful input of expert individuals with field experience could have achieved nearly as much. The most beneficial aspect of the UN’s participation in the voluntary disarmament exercise was its mere presence, which acted to pre-empt much of the potential for violence. Ex-White Army youth explained that they felt the presence of foreign military staff and aid workers signalled that changes were really taking place in South Sudan after the signing of the CPA.50 Further to these hopes there was the impact that the UN presence had on the broader community. A sign of this was that commercial trade improved immediately upon completion of the disarmament exercise.51 Traders who arrived in Akobo from downriver after the disarmament exercises’ completion explained that they had decided to come precisely because they had heard that there was a UN presence and that the disarmament exercise had taken place peacefully as compared to events in other neighbouring counties.52

their communities to disarm peacefully, with the youths giving up their guns voluntarily. One of the most important aspects of the peace meetings was to encourage dialogue between various communities so that they could ‘see that they are all “southerners”’.54 The war had stopped interaction between the region’s communities and isolated them from one another aside from negative, violent contact. For instance, reconciliation between the Murle and Lou communities was encouraged and participants were supposed to take news of the reconciliation efforts back to their communities.55 Lastly, the local NGOs were careful to work with traditional problem-solving mechanisms utilizing local leaders. As one NGO staff member articulated, ‘Chiefs can do the job, but we can help facilitate it.’56 Another interesting aspect of the peace meetings was the blunt advice that the local NGOs proffered to community representatives – namely, that they faced a clear choice between forced disarmament by the SPLA and voluntary disarmament, with participants being warned that ‘they had seen what had happened to the neighbouring county’.57 This pragmatic advice was then sent back to the communities through ‘community mobilizers’ to explain to White Army youth and other community members the dichotomous decision they faced. This ‘sensitization’ of armed youth to the pros and cons of disarmament options was crucial to Akobo county’s voluntary disarmament exercise. In the neighbouring counties to the west, outreach to the armed youth in order to explain to them their options was impossible owing to security concerns which limited NGOs’ access, and the fact that the youth here had a negative perspective on potential outcomes for non-voluntary disarmament.58 The second role that NGOs played was to encourage youth to engage in new activities after the disarmament exercise, something felt to be crucial to longer-term peace. An indigenous NGO staff member explained regarding newly disarmed youth, ‘If [we] have disarmament but nothing else, they will just sit on the roadside.’59 Follow-up measures being implemented centred on providing training to disarmed youth in basic and civic education and vocational skills.60 The goal of this outreach and training is ‘to give them some basic skills to achieve some sustainable development and to get their minds off guns and cows’.61 The leader of an Akobo youth association further noted that ‘The ones facing the problems and the ones being disarmed are the youth […] so it [the training] is very important to us.’62 The NGOs’ outreach to disarmed youth was mirrored somewhat by the action of the state and central governments, which prioritized offering options to youth to join the SPLA or nascent police forces.63 78

This section addresses some of the notable challenges to the dis­ armament exercise and subsequent efforts to stabilize the situation in Upper Nile and Jonglei states. It is important to qualify the short-term successes of the voluntary disarmament exercises by considering some of the longer-term challenges that lingered following them.

Prevalence and corruption of cultural norms of retribution and cattle raiding  While many arguments can be made for the merits of collecting firearms, holding ‘peace meetings’ and undertaking follow-up activities such as youth training schemes, underlying ‘peace’ in what are seminomadic pastoral communities, such as the Nuer, are some deeper cultural attributes. These attributes played a significant role in fomenting violent activities in the region, one defined by its historical anarchy and a strong presence of everyday violence. One of the strongest of these was the ‘normality’ of cattle raiding in South Sudan as a key facet of the socio-economic system there, and related to it, a propensity to engage in blood feuds between clans and sub-clans after homicides and other transgressions occur. However, after decades of war, these cultural attri­butes and practices, which were integrated through ascribed rituals and norms of exchange within the socio-economic systems of the seminomadic, pastoralist tribes of the south, have intensified significantly. First, the rate and level of cattle raiding, as well as the degree of violence that firearms permit, helped to accelerate and intensify a practice that previous to the civil war was undertaken at lower levels of frequency and violence. The various larger armed groups, such as the SPLA and SSDF, also began to engage in cattle raiding, thereby increasing both its scope and prevalence in South Sudan. Overall, cattle raiding can be seen to have reached a level that completely exceeded traditional mechanisms to keep it in check, such as peace agreements negotiated by elders, especially during times of drought. Second, blood feuds were an accepted practice among the Nuer historically but implemented only in very limited circumstances when other means of conflict resolution had failed. One Nuer elder explained, ‘Even before guns, in Nuer tradition, there is retribution, but if you kill somebody else, a compensation arrangement can be made.’64 This ‘compensation arrangement’ then prevents the tensions from escala­ ting into a violent blood feud. The escalation in the number of blood feuds between Nuer groups saw this cultural practice distorted by the prevalence of firearms, which weakened normal conflict resolution means among the Nuer, namely the cattle compensation scheme. As 79

4 | The White Army militias

Outstanding challenges to disarmament exercises in South Sudan

the civil war progressed, and with it the massive population displacements, increases in the prevalence of firearms, the weakening of older cultural norms of behaviour and the introduction of new ones that were conditioned out of firearm-driven violence, the cattle compensation scheme gradually stopped working in rural communities in Upper Nile and Jonglei states. As a result, following transgressions, violence between groups started to proceed immediately to blood feuds defined solely by revenge killings. The simple collection of firearms was unlikely to quell cattle raiding or blood feuding by itself; largely because it is near impossible to immedi­ately, or ever, collect absolutely all of the firearms in places that have become so saturated with them as South Sudan. These cultural factors require follow-up in the form of the reassertion of traditional conflict resolution mechanisms, i.e. the cattle compensation system, and the introduction of more modern forms, such as functioning police and legal systems. A basic disarmament exercise was a first step, however, because, as one local highlighted, before disarmament, ‘civilians were more powerfully armed than the government’, and in many cases the SPLA as well.65 It was hoped, as articulated by another southerner, that now that an initial disarmament had occurred, ‘the government will be more powerful than the civilians so it can restart the compensation system’.66 Lastly, the result of escalations in cattle raiding and blood feuding also served to limit a very simple, universally applicable problem-solving method, namely dialogue between aggrieved parties, because commun­ ities and individuals stopped interacting with one another aside from through violence. As the Akobo county commissioner explained of the situation resulting from such a prevalence of weapons: ‘People are not sharing common things like marriages, dances, grazing. Each clan now grazes their cattle alone. Usually people moved together but because of guns they stopped. No single clan hasn’t fought with another clan. All have fought.’67 When normal, friendly interaction became largely impossible between neighbours, it was only natural that suspicion and hatred should develop and violence ensue.

Ethiopia–Sudan ‘frontier’ issues  The Ethiopia–Sudan border has historically been the setting for a significant level of cross-border interchange linked to conflicts within both states, and this phenomenon was especi­ ally prevalent in eastern Upper Nile and Jonglei states. The neighbouring Ethiopian region of Gambella has long had a high prevalence of 80

The efficacy of the SPLA’s ‘rolling wave’ of disarmament ‘Even now [we] worry about the Murle. The people of Akobo county are worried […] worried because the Murle are still armed. Protection should come from the SPLA.’  (Akobo county commissioner)73

When it was decided to implement a disarmament exercise for armed civilians in the south for late 2005 and 2006, the SPLA chose a strategy that could be termed a ‘rolling wave’. While many parties had called for ‘universal disarmament’, such an exercise was impossible. The SPLA did not have enough troops or the logistics to support a disarmament 81

4 | The White Army militias

firearms and several armed groups operating in it. A persistent concern of those implementing the disarmament exercises was connected to the likelihood of small arms flowing back into South Sudan after the disarm­ament exercise or of neighbouring groups living in Ethiopia raiding newly disarmed communities in South Sudan.68 Implementing disarmament on one side of a firearms-laden border is all the more challenging when the other side is not concomitantly undertaking such programming as well. It is noteworthy that one of the disarmament posts for the aforementioned July exercise, officially for Akobo county in South Sudan, was actually physically located in Ethiopia, demonstrating the great fluidity of the border region.69 The near-total lack of Ethiopian border patrol posts was a major concern of Akobo county government officials, as was a less-than-desired level of coordination between Sudanese and Ethiopian officials over border control issues.70 Interestingly, while the border region still had a weak control cap­ acity, there had been coordination between the SPLA and the Ethiopian army to jointly attack Ethiopian Anuak rebels and other armed groups in the Gambella region of Ethiopia in February 2006.71 Parallel to this, it was believed that Ethiopian army forces had been allowed access to limited pieces of South Sudanese territory in order to counter the forces of the Gambellan Peoples Liberation Movement (GPLM) and the Oromo Liberation Front (OLF). In exchange for this access, it was believed that the SPLA was given freer rein to control Lou Nuer areas in western Gambella state. It is notable that the one area of Akobo county not disarmed during the July 2006 exercise was the area in its south-eastern corner around the town of Pochalla. There were some Ethiopian rebels in that area, namely the Anuak GPLM, and there was a fear among UN staff and county officials that there could be a future problem of armed Anuaks attacking recently disarmed citizens in other areas of the county.72

exercise for all of South Sudan at one time. Thus, it was decided to ‘roll’ across South Sudan, disarming one group before proceeding to the next. As the ‘wave’ of disarmament exercises proceeded, the SPLA would then provide security to the newly disarmed groups while their neighbours were disarmed as well. In theory it all sounds quite simple and straightforward, yet the challenges of implementation proved considerable and bear important lessons for future disarmament exercises. First, many communities felt that the ‘rolling wave’ strategy was ­really just a way for the SPLA to retaliate against individual communities, namely the Lou Nuer, with whom they had fought frequently during the civil war. As one Lou elder explained, ‘The Lou can accept the disarm­ ament if others are being disarmed too.’74 However, if this was seen as not happening, the resentment of many Nuer for the SPLA as a ‘Dinka militia’ was only exacerbated, as often seems to have been the case. Second, for the civilians experiencing the SPLA’s disarmament exercise, their doubts weren’t so much that the SPLA couldn’t disarm them, but rather as to whether there would be any protection for them afterwards. And in fact, Lou Nuer fears of attacks against freshly disarmed communities by Murle coming from the south and to a lesser extent by Jikany Nuer from the north were widely realized. Murle raided Lou Nuer communities in both central and eastern Jonglei after disarm­ ament exercises.75 These Murle attacks were mostly cattle raiding and petty banditry against Lou villages. The ability of the SPLA to provide blanket security for disarmed communities was decidedly lacking in the immediate aftermath of the exercises.76 However, while the SPLA was not able to provide enough troops for comprehensive security in the aftermath of the disarmament exercise, communities still felt trapped between wanting a larger SPLA force to provide security for them and being afraid of the SPLA in general, as well as fearing that a larger force would ‘eat off’ communities that were already food insecure. The large SPLA contingent of 3,000 troops that had forcefully disarmed central Jonglei in May 2006 stayed in the area afterwards and supplied itself with foodstuffs from the local community, further compounding their food insecurity and the locals’ resentment of the SPLA.77 Furthermore, while the SPLA has instituted a comprehensive upgrade and reform effort, the development of police forces has been greatly hindered in South Sudan. The CPA is not very explicit on issues regarding police forces and there is a great deal of ambiguity over jurisdiction issues with the national police (i.e. those under the control of Khartoum) and the development of local police forces. As a result of this, as well as rural South Sudan never having had much in the way of a civil police 82

Conclusion

A review of the history and efficacy of the disarmament exercises of the White Army militias in South Sudan from January to August 2006 highlights many policy considerations for the demilitarization of similar militias elsewhere. A key goal of the GoSS’s and SPLA’s post-CPA security strategy had been to quickly disarm major concentrations of heavily armed civilians in the south, namely the White Army militias in Upper Nile and Jonglei states. The SPLA never naively believed that its initial 83

4 | The White Army militias

presence, Upper Nile and Jonglei states had little to no police presence even a year and a half after the CPA’s signing. There was certainly an overarching lack of police capacity to guarantee the security of a civilian population threatened by still-well-armed groups in 2006. The challenge of protecting newly disarmed communities from stillarmed neighbours was especially pressing considering the still-armed presence of the Murle to the south of the Lou Nuer. This consideration had a dichotomous effect on disarmament dynamics in southern Jonglei state. First, it was widely expected that the Murle would be disarmed after the Nuer and that there was a good chance that armed resistance would be undertaken by armed Murle civilians in a repeat of the violence in central Jonglei.78 Compounding this fear was the likelihood that armed Murle youth would be encouraged to resist disarmament by SSDF components, primarily the Pibor Defence Force (PDF) militia of Ismail Konyi.79 The rationale behind this potential agitation by SSDF components was to prevent the SPLA from establishing itself in the area around Pibor, the stronghold of the PDF. Counteracting this agitation was the call by other Murle actors, namely prominent chiefs, that the community undergo a peaceful disarmament exercise, as had been accomplished in Akobo county, rather than experience forceful disarmament by the SPLA.80 In order to help this along, it was suggested by them that UNMIS establish a presence in Pibor and other Murle communities to counter the influence of Ismail Konyi and his PDF component of the SSDF. Encouraging this perception within parts of the Murle community that they should opt for peaceful disarmament was the fact that the Lou Nuer had been disarmed and that this in itself relieved a great deal of the pressure to remain armed themselves. This de-escalation of the communities’ ‘arms race’ meant that voluntary disarmament of the Murle could potentially be easier to achieve as part of a rolling wave of disarmament. As a Murle chief articulated: ‘We mostly had guns because SAF gave them out in the past but now if everybody gives up guns, everybody can be safe.’81

disarmament exercises could collect absolutely all of the firearms held by civilians. Rather it was hoped that, following the major disarm­ament exercises, the White Army militias as a major cause of insecurity would be greatly weakened. This was largely achieved as the White Army mil­ itias effectively ceased to exist as powerful armed groups, capable of broadly wreaking havoc on the security situation and preventing the implementation of basic governance. At the heart of this limited success was a calibrated demilitarization campaign based on influence, incentives and coercion, coupled with a publicity operation by a recognized authority that provided information on disarmament that was sufficiently persuasive to produce the desired effect. A comparison of the SPLA’s forceful disarmament exercise from January to May 2006 in western and central Jonglei state with the voluntary disarmament exercise in eastern Jonglei state in July 2006 shows important lessons learned. While these processes were both driven by the SPLA, the voluntary exercise in eastern Jonglei also involved many other actors who played roles conducive to its peaceful unfolding. In contrast, the disarmament exercise that took place from January to May 2006 in western and central Jonglei state was coercive when the SPLA effectively acted on its own. In eastern Jonglei the role of local communities in encouraging their youth to disarm peacefully was crucial and was encouraged by a fear of the ‘collective punishment’ of a coercive SPLA exercise as well as cultural proclivities towards communalism. Furthermore, the simple presence of UNMIS and UNDP staff during the exercise was very pacifying in a generally volatile environment, while the local county government helped to facilitate mediation between the SPLA and White Army youth as well as coordinate the collection of arms. Lastly, local NGOs played a crucial role in ‘sensitizing’ armed youth to the benefits of a peaceful, voluntary exercise, as well as instigating useful ‘peace meetings’ between various actors necessary for a successful exercise. However, while the White Army militias as a specific phenomenon of Upper Nile and Jonglei states have widely ceased to exist since 2006, the level of violence in their home areas has actually increased. Indeed, within several years of the CPA’s signing, the level of violence in South Sudan had actually eclipsed that in Darfur, and in 2009 alone some 2,500 people were killed and over 350,000 internally displaced.82 This violence was mostly tribal conflict, especially in Upper Nile and Jonglei states, between Nuer, Murle and Dinka communities. The Lou Nuer, whose youth were strongly represented in the White Army militias, have been especially involved in the ongoing violence as they have fought 84

85

4 | The White Army militias

at assorted times with virtually all of their neighbours. For example, fighting with the Murle in March 2009 left over 750 people dead in the Akobo area, where the relatively peaceful disarmament exercises had occurred in 2006, while in May 2009 over seventy people died in fighting with the Jikany Nuer just north of Akobo. Although the individual disarmament actions of 2006 went a long way towards removing the White Army militias, that large-scale violence continues makes it clear that the wider efforts at demilitarization in South Sudan, and especially in Upper Nile and Jonglei states, have hardly been conclusive. A number of limited disarmament efforts similar to those of July 2006 have been repeated at different times and locales in South Sudan over the last several years, and much emphasis has been placed on community mediation and voluntary disarmament. However, as the Inter­ national Crisis Group noted, ‘despite much time, effort and resources, these initiatives have not produced sustainable peace’ (International Crisis Group 2009a: 7). In 2006, there had been much hope placed within the GoSS on a ‘rolling wave’ of disarmament given that ‘universal dis­ armament’ was unfeasible. While a rolling wave may have been relatively effective at dealing with specific groups such as the White Army militias in the limited locales of Jonglei and Upper Nile states, it has proved ineffective for the wide-scale disarmament of the general public in South Sudan. At its simplest this has been because of the ongoing weakness of the state and a comprehensive failure to enforce a monopoly on violence following individual disarmament efforts. Without such imposition, violence between communities has been largely left unchecked. As they feared, communities that were successfully disarmed, such as the Lou Nuer in 2006, were indeed left open to attack by neighbours such as the Murle, Dinka and Jikany Nuer. Resulting from that, and understandably, the Lou Nuer rearmed themselves in 2007 and 2008 (ibid.: 3), albeit not in the specific form of White Army militias. Other communities in South Sudan have most often been justifiably hesitant to turn in weapons, given their lack of faith in the SPLA and GoSS to provide them with physical security. In the years since the initial disarmament exercises of 2006, the tendency of local communities in areas where the White Army militias had been present has been to continue to seek security through their own means rather than hoping for the state to provide it. For example, in 2010 Jonglei state was still without even a single police station and the quality of the South Sudan Police Service has remained ‘abysmal’ (ibid.: 19). The SPLA for its part has been unable to comprehensively step into the security void and continues to suffer from its own internal

divisions, such as the limited integration of ex-SSDF militias, and a preoccupation with the SAF in the run-up to the 2011 independence referendum. For its part, UNMIS has not maintained a consistent presence in rural areas, such as it did in Akobo in 2006, and has not placed emphasis on protecting civilians in extremis, preferring instead to leave the response to violent tribal incidents to the SPLA rather than directly intervening. Indeed, the most pressing lesson learned from the White Army militias is that while it is relatively easy to disarm specific groups, the bigger challenge is getting state security forces and/or international peacekeepers to impose a monopoly on violence and to rightly protect communities that have been disarmed, either through their own volition or by force. Until that happens, there will continue to be examples of local communities forming their own militias.

86

5 | The mutineers of Timor-Leste

While the previous chapters have highlighted militias within the context of weak states suffering from a large plethora of armed actors, this chapter will instead articulate how a weak state initially without multitudes of armed groups saw its nascent, seemingly progressing security sector implode and spin off new armed opposition in the form of militias. Critically, the chapter highlights how a militia can act as a medium for broader political tensions within a state and hence how the specific dynamics and contexts of the situation prove even harder for conventional responses to be effective. Despite many fears that TimorLeste’s post-independence progress was most likely to be threatened by the Indonesian-aligned militias that wrought exceptional destruction there in 1999, the biggest threat to Timor-Leste’s rationalizaton has actually proved to be tortuously internal. Namely, a ‘mutineer’ militia arose from tensions within the Timor-Leste Defence Force (FDTL) and from April 2006 to April 2008 Timor-Leste experienced significant violence. This crisis in turn sparked a broader political upheaval in the country leading to the fall of one government and extended social unrest. The uniqueness of the mutineers as a group left them largely outside of the UN mission’s and international peacekeepers’ mandates.1 More significantly, the turmoil they caused overwhelmed the nascent institutions of the Timorese state, leaving the country mired in an environment of fear and mistrust. A history of the 2006–08 crisis

In August 1999, following a referendum rejecting autonomy within Indonesia, extreme violence perpetuated by militias backed by the Indo­ nesian military provoked a massive outpouring of refugees from the annexed Indonesian territory most commonly known as ‘East Timor’, a former Portuguese colony. Subsequently, the UN and a coalition of regional military forces led by Australia intervened to restore order and prepare the territory for eventual statehood, which was achieved for the Democratic Republic of Timor-Leste in 2002 (Steele 2002). The international intervention continued after independence, notably through UN missions, and had civilian aid components undertaking judicial 87

reform, democratic governance and socio-economic development, as well as military components emphasizing security sector development.2 Widely hailed for several years as a clear success story of state-building by the UN because of relative political stability and democratic governance, this judgement on Timor-Leste’s progress soured in April and May 2006 when widespread violence erupted in the capital, Dili, following protests by recently fired FDTL soldiers complaining of discrimination (Nevins 2007: 163). The protesting soldiers were known locally as ‘petitioners’ because they petitioned FDTL commanders and President Xanana Gusmão to rectify alleged inequities within the armed forces. In the years immediately after independence, Timor-Leste’s security forces had experienced increasing tensions centred upon what Simon Philpott termed an ‘­ethnic/territorial element’ (Philpott 2006: 151). This was a simmering discontent mirroring broader social tensions within Timor-Leste between ‘easterners’ (lorosae), who mostly speak Fataluki or Makasse, versus ‘westerners’ (loromonu), who are mostly Tetum speakers. The peti­ tioners were mostly younger soldiers originating from western TimorLeste who complained of systematic discrimination by an FDTL officer corps dominated by former National Liberation Forces of an Independent East Timor (FALANTIL) rebels, the resistance movement that had fought against Indonesian occupation and whose members were largely from the eastern part of the country. The complaints focused upon regional distinctions being used to deny younger westerner soldiers adequate opportunities for promotion, as well as resentment that they were unable to deliver their salaries to family members residing in the western half of the state from FDTL bases located in the eastern half. In January 2006, a group of 159 FDTL soldiers led by Lieutenant Gastão Salsinha presented a petition to Gusmão detailing complaints against FDTL commanders and asking for corrective action to be taken. Gusmão instructed the commander of the FDTL, Brigadier General Taur Matak Ruak, to handle the issue decisively. But instead the situation deteriorated as more disenchanted FDTL soldiers congregated in Dili to protest against the alleged discrimination. By March, the number of petitioners had risen to 593, and after they refused a final order by Ruak to return to their barracks, they were summarily dismissed en masse. This was a significant act: the total size of the FDTL was only 1,400 soldiers. On 24 April the petitioners began a new protest in front of the main government building in Dili, with the crowd attracting large numbers of agitated youths. As the protests continued, the petitioners’ complaints 88

89

5 | The mutineers of Timor-Leste

of discrimination expanded to call for the government of Prime Minister Mari Alkatiri to be replaced, under the claim that his officials encouraged the FDTL’s institutional discrimination. By 28 April the situation had spiralled out of control as youths and petitioners attacked the main government building. This quickly escalated into street violence spreading around Dili. Alkatiri subsequently ordered the FDTL to enforce order, prompting even more street fighting. Violence continued into the next day, including significant numbers of arson attacks. As a result of the violence, the petitioners dispersed, with many going to districts south of Dili. Following the protests by the petitioners and the spike of violence, the situation took a further turn for the worse on 3 May when additional turbulence erupted within the FDTL. The head of the FDTL Military ­Police, Major Alfredo Reinado, seventeen other FDTL soldiers and several members of the National Police of Timor-Leste (PNTL) special units all deserted with their arms, the group that this chapter terms ‘mutineers’. Reinado argued that his mutineers’ armed presence in the form of a selfdefence militia was necessary to ensure the safety of the petitioners, his fellow westerners, from violence by easterners in the FDTL. On 23 May a firefight erupted between Reinado’s mutineers and an FDTL patrol, leaving one FDTL soldier and two of Reinado’s followers dead. By this point the role of the PNTL had become critical. A provocative incident had occurred on 8 May in Gleno, where many of the petitioners had congregated after the violence in late April. A large protest against the April incidents degenerated into violence. Easterner police officers providing security for a local government official trying to placate the protesters had been disarmed by a westerner PNTL commander and subsequently assaulted by the crowd, leaving one officer dead. Throughout May the political and security situation deteriorated further, with some westerner police officers deserting to join the petitioners. Events worsened greatly when some police officers joined a 24 May attack on FDTL headquarters on the outskirts of Dili. The attack was undertaken by a mix of police officers and armed civilians and ex-soldiers and left one FDTL officer dead. The tensions between the PNTL and the FDTL climaxed on the 25th, when FDTL soldiers attacked PNTL officers at their headquarters in Dili. Following negotiations, the police officers inside were to have been ­allowed to retreat, unarmed, to the nearby UN compound. However, while being escorted out of the headquarters by UN police, the unarmed PNTL officers were fired upon by FDTL soldiers; eight police officers were killed. At that point, Dili erupted into wide-scale violence between

police and soldiers, while large groups of youth, mostly westerners, attacked easterner neighbourhoods. The government requested international military assistance on 24 May and Australian Defence Force soldiers began arriving the next day as the violence peaked, to form what would become the International Stabilization Force (ISF), eventually including New Zealand Defence Force personnel as well.3 The UN confirmed after a special investigation that the April and May violence left thirty-eight people dead and approximately 150,000 internally displaced people (IDPs), mostly easterners within Dili and its vicinity (United ­Nations 2006a: 42). While the street fighting diminished, the political confrontation esca­lated as Timor-Leste’s main political actors swapped blame for the violence. The tensions between Gusmão and Alkatiri were especially damaging. Alkatiri charged that political factions supporting Gusmão manipulated the situation, encouraging the petitioners in order to discredit Alkatiri and topple his government. On 19 June, Gusmão sent a letter to Alkatiri asking him to resign or face dismissal over allegations that he had participated in the illegal distribution of police weapons to civilians during the unrest. Gusmão followed this with an impassioned speech criticizing Alkatiri’s leadership of the ruling FRETILIN party, and threatened to resign in protest. Alkatiri succumbed to the pressure on 26 June and was replaced by José Ramos-Horta as prime minister. Despite the influx of over one thousand ISF soldiers and a renewed UN mandate strengthening the presence of UN Police (UNPOL), the security situation remained tense. More people died between August 2006 and February 2007 than during the initial violence of April and May 2006 (International Crisis Group 2008c: 3). An exceptionally large IDP population, chronic youth violence on the streets, and arson a­ ttacks centred on property disputes were all conducive to creating an extended highly volatile environment in Dili. Crucially, Reinado and his mutineers also refused to surrender or disarm, which exacerbated a dire political and security situation characterized by a breakdown in law and order following the near-collapse of the PNTL. On 26 July, Reinado and fourteen mutineers were arrested on weapons, murder and desertion charges by Portuguese paramilitary police. However, they escaped from jail on 30 August in controversial circumstances and retreated into the mountains south of Dili. In late February 2007, the security situation worsened when Reinado raided a border police post and stole armaments. In response and at the behest of the government, Australian ISF soldiers raided Reinado’s camp on 3 March in a failed attempt to capture him, although it did leave five of his mutineers dead. The raid sparked 90

91

5 | The mutineers of Timor-Leste

an intense round of street violence in Dili in support of Reinado, forcing Ramos-Horta to call off the ISF’s efforts and push for dialogue instead. Subsequently, in tandem with ongoing government efforts at negotia­ ting with the petitioners, an extended dialogue attempted to achieve peaceful reconciliation between the government and the mutineers. This left Reinado generally free to move around and to give interviews to the international press, because he was not being actively pursued by ISF soldiers, UN police or Timorese security forces. Despite the ongoing instability, multiple successful elections were held in 2007. Ramos-Horta won the presidential polls on 9 May as an independent, despite a strong challenge from the FRETILIN candidate, Francisco ‘Lu-Olo’ Guterres. On 30 June, parliamentary elections were held, bringing Gusmão to power, this time as prime minister. Gusmão led a newly formed party, the National Congress for Timorese Reconstruction (CNRT), the lead party of a coalition government called the Alliance of Majority in Parliament (AMP). The AMP consisted as well of the Democratic Party, the Social Democratic Party and the Timorese Social Democratic Association. The significance of the 2007 elections was that FRETILIN, which had ruled with a strong parliamentary majority for five years, was comprehensively replaced by Gusmão and his political allies. The AMP parties were largely united around opposition to FRETILIN, a theme they focused upon during the election. In a step that significantly raised political tensions after the election, FRETILIN publicly challenged the legality of Gusmão’s CNRT forming the government, citing a different interpretation of the constitution and claiming that FRETILIN had the right to challenge because it had won more seats than any other party. Significantly, the parliamentary elections produced results paralleling the ‘geographic divide’ that dominated the April/May 2006 violence, with FRETILIN strong in the eastern half of the country while the CNRT and its coalition partners, notably the DP, did well in Dili and the western half (Cotton 2007: 464). Despite the successful elections of 2007, violence again flared in Dili in early 2008. On 11 February, Reinado’s mutineers attacked RamosHorta at his home and Gusmão as he travelled in a convoy (International Crisis Group 2008b: 1). While Gusmão was left unscathed despite his car being fired upon, Ramos-Horta was shot three times but lived, following emergency treatment in Australia.4 The attacks did, however, leave Reinado and one of his followers dead. The exact nature of the attacks is still disputed, yet what is unchallenged is that Reinado’s mutineers were directly involved in a shooting incident that ultimately left the president

seriously wounded.5 After the attacks, a state of siege was declared by the government, allowing the FDTL to lead a joint command with the PNTL to maintain civil order through martial law and to try to capture the remaining mutineers, who had disappeared into the mountains. Ramos-Horta returned to Dili on 17 April and soon after the mutineers peacefully surrendered and were imprisoned pending the conclusion of criminal investigations of the attacks. Empirical mapping

From this brief history of the turbulent events of April 2006 to April 2008 it is possible to empirically analyse the mutineers, the group paramount to the crisis, as well as the petitioners, given their strong connection to the initial development and ultimate outcomes of the mutineers. It is important to note that while the two groups were in fact ‘linked problems’, there is still a need to separate the petitioners from Reinado’s mutineers because their actions and purposes were substantively different (International Crisis Group 2008c: 19). This empirical mapping of the mutineers will provide an analysis of their motivations, the government’s responses to them, and the involvement of the international community. To begin with, the motivations of the petitioners were complex, evolving as the national political situation changed. Ostensibly, the crisis started as a fairly mundane ‘labour dispute’ involving junior officers, namely westerners, who felt they were being systematically discrim­ inated against by easterner FDTL ‘management’.6 After the initial petition claims went unresolved during January and February 2006, the situation festered and eventually escalated into a national crisis when FDTL troops were tasked with returning order to Dili’s streets, which provoked overt confrontation with the protesting petitioners. From then on, what may have initially begun as an internal FDTL labour dispute assumed national prominence as the petitioners began to view themselves, and be seen by significant numbers of Timorese, as representing a ‘justice movement’ aiming to rectify historical animosities stemming from the resistance era.7 The petitioners’ movement stemmed largely from debates over who fought harder during the resistance era, with many FALANTIL veterans from the east, established as officers in the FDTL, believing that westerners were generally compliant with Indonesian rule. A national sense of division was strengthened by a 23 March 2006 speech by Gusmão. In it he argued that the FDTL’s senior officers were wrong to lay off the petitioners on 16 March, thereby raising tensions with Alkatiri, who had 92

93

5 | The mutineers of Timor-Leste

authorized the decision. Gusmão was heavily criticized for the speech because it entrenched in the public discourse the notion of ‘easterners versus westerners’ by stating that many easterner FALANTIL veterans, now in the FDTL, thought other Timorese ‘from Manatuto to Oecusse’ were ‘militias’ children’, i.e. westerners (International Crisis Group 2006a: 7). The overall effect of the speech, even if unintentional, was to undermine the leadership of Alkatiri and encourage divisions between easterners and westerners nationally. While the petitioners’ dilemma may be described as a labour dispute that escalated into a national political crisis, the petitioners were also at times decidedly opportunistic. This is especially so in the case of their leader, Salsinha, who had been accused of smuggling sandalwood before the petition movement started and was likely looking for an opportunity to avoid being demoted and marginalized within the FDTL. Overall, many Timor-Leste observers feel that the petitioners themselves were largely a transient movement, subject to political manipulation, and hence their significance primarily came from the national tensions they manifested and the political conflict that was ultimately provoked. 8 The assorted governments’ responses to the petitioner movement were erratic. The Alkatiri government took a strict approach by supporting dismissal of the petitioners and the deployment of the FDTL in Dili in April 2006. As an interim government before national elections, the Ramos-Horta administration did not adopt a very strong approach to resolving the petitioners’ issues other than by encouraging dialogue. After assuming power in August 2007, the Gusmão government regarded resolving the petitioners’ complaints as a national priority. The primary challenge for the Gusmão government was that what started as fairly simple claims of job discrimination became attached to broader political tensions, particularly those involving Reinado’s mutineers, the IDP situation in Dili and security sector reform.9 A primary goal of the Gusmão government was to keep the petitioners separate from Reinado’s armed, and hence more dangerous, mutineers, a difficult task considering that Reinado proclaimed himself ‘com­mander’ of the petitioners in November 2007 and enjoyed broad support from them.10 In its attempt to separate Reinado (as well as Salsinha after he declared himself Reinado’s deputy) from the vast majority of petitioners who remained unarmed and interested primarily in returning to the FDTL, the Gusmão government called for the petitioners to gather at barracks in Dili to allow for dialogue and negotiations to take place. The endeavour had varying results. Calls for petitioners to congregate in November 2007 were an embarrassing failure when only eighteen

showed up to talk to the government. Many, estimated in the hundreds, instead congregated at the mutineers’ camp to join widely publicized parades.11 The Gusmão government called for another try at negotiation and cantonment in early February 2008. While the turnout was initially slow, following the 11 February attacks, petitioners grouped en masse at the Dili site, with approximately 670 turning up. Those attacks seemed to have catalysed the cantonment of petitioners because of a fear that the FDTL might attack them if they resisted cantonment or that the Gusmão government would cancel all possibilities for dialogue. Once they were encamped at the Dili site, the government felt obliged to categorize the petitioners as follows: (1) original signatories to the petition; (2) adherents to the petition, i.e. those who left the FDTL to participate in the Dili protests; (3) pre-crisis FDTL absentees; and (4) post-crisis absentees.12 The government was frustrated that those ‘petitioners’ cantoning themselves included many FDTL absentees not directly involved in the initiating crisis from January to May 2006; officials termed them ‘petitioners-plus’.13 While the government had originally wanted to deal only with the 593 actual petitioners, it decided to include the post- and pre-crisis absentees because of worries that some were joining Reinado’s mutineers.14 Following cantonment, the Gusmão government offered the petitioners a questionnaire to ascertain their future needs and catalogue complaints against the FDTL and the government. Based on the answers, the government then determined to offer two possibilities, an economic benefits package or the chance to reapply to join the FDTL.15 Crucially, the second option required the petitioners to go through the regular application process for FDTL recruitment. They would either be rejected or accepted under the normal guidelines. If they were readmitted they would be able to assume the ranks they had at the time they left the FDTL. The questionnaires were a frustrating development for the petitioners, who had expected a group negotiation with the Gusmão government and return to the FDTL en masse, a specific outcome that approximately 80 per cent of the petitioners demanded.16 The petitioners soon felt trapped by their inability to obtain collective dialogue, once it became apparent that the open hostility of senior FDTL officers to their readmission meant that it was increasingly unlikely to happen.17 For its part, the government felt the need to offer some possibility of a return to the FDTL but was still hesitant to entirely disregard the original decision of FDTL commanders.18 The cantonment of the petitioners was concluded by August 2008 with them returning peacefully to their home 94

95

5 | The mutineers of Timor-Leste

districts. They universally opted for the financial package offered by the government with none reapplying to the FDTL because of the conditions placed upon that option.19 Petitioners who left the FDTL after 31 January 2006 received US$8,000 while those who had left before received three years’ salary plus US$1,500. As for the actions of Reinado’s mutineers, the key difference between them and the petitioners was that they had left their barracks in May 2006 armed, whereas the petitioners were never armed. The one, albeit crucial, exception was Salsinha and several of his more militant petitioner colleagues who participated in the February 2008 attacks. Prior to that incident Salsinha had been a petitioner, neither armed nor having fought FDTL troops alongside the mutineers. Indeed, aside from their general political agitation against the government, Reinado’s militia group should be deemed specifically mutinous because of their firefight with the FDTL in May 2006, the February 2007 raid on the police post, and their armed attacks on the national leadership in February 2008. Reinado’s rebellion against authority had begun with calls for Alkatiri to resign and his initial declaration of support for Gusmão’s leadership. As mentioned previously, the major argument by Reinado for having left armed was that after the FDTL had been deployed against the petitioners on 28 April 2006 by the Alkatiri government, there was a need to protect the petitioners from easterners in the FDTL.20 Even after the resignation of Alkatiri, Reinado argued for the need to stay armed to provide an independent guarantor that the Dili government would not become violent against the public and to protest about the lack of prosecution of Alkatiri officials for the spring 2006 violence. The petitioners as a group managed to maintain themselves as a political and social movement after the initiating violence of 2006. But Reinado’s mutineers became ever more regarded as a violent movement intent not just on rectifying specific claims of discrimination in the armed forces. Ramos-Horta stated that he believed Reinado had eventually deluded himself into believing he could assume power by assassinating the country’s top leaders: ‘[Reinado believed] killing me and Gusmão would give him power and make him the strongman of the country.’21 Even before the February 2008 attacks, the UN recommended that Reinado and many of his followers be charged with assorted crimes relating to the attack on the FDTL patrol in May 2006 (United Nations 2006a: 48). As his stand-off continued into 2007, Reinado became increasingly critical of Gusmão, declaring his leadership a primary cause of TimorLeste’s political unrest.22 However, through the February 2008 attacks

Reinado’s approach extended to an outright attack by the mutineers on the state’s leadership. The transition from a relatively benign revolt to attempted murder confused Reinado’s followers in the general public, especially since he had encouraged them in 2007 to vote for both Ramos-Horta and for parties allied with Gusmão’s attempt to wrest parliament from FRETILIN. It appears that during the course of 2007 Reinado increasingly felt that Gusmão had betrayed him by diverging from Ramos-Horta’s attempts at a negotiated settlement. Reinado put great effort into presenting himself as a petitioner, even while he refused to disarm. Being a commissioned officer, well spoken and charismatic, he was widely respected by the petitioners, who saw him as a man who could stand up to governments in Dili. Reinado told a parade of petitioners on 22 November 2007 that he would ‘lead my soldiers down to Dili’ if the petitioners were not reinstated in the FDTL (International Crisis Group 2008c: 19). Given Reinado’s popularity with significant numbers of the Timorese populace, especially in Dili neighbourhoods such as Taibessi and Comoro (as evidenced by the rioting that occurred in early March 2007 when Australian ISF soldiers tried to capture him), the government had trouble devising strategies to deal with him and his mutineers. The erratic government responses, oscillating from high-level dialogue meetings to outright confrontation, were compounded by approaches that at times diverged quite substantially between Ramos-Horta and Gusmão. As president, Ramos-Horta actively encouraged dialogue to get Reinado and his followers to surrender peacefully. This included using a Swiss non-governmental organization (NGO), the Centre for Humanitarian Dialogue, to facilitate a meeting with Reinado in August 2007. Ramos-Horta even gave Reinado a safe-conduct badge, preventing his arrest, in order to encourage a peaceful solution.23 Periodic dialogue meetings between Reinado and Ramos-Horta continued until 13 Jan­ uary 2008, when they last met. For his part, Gusmão as prime minister, frustrated at having been in the weaker post of the presidency, took an equally assertive, though contrasting, approach by seeking to detach the mutineers from the petitioners through cantonment and being less open to dialogue. The international response to both petitioners and mutineers was limited because neither group fell clearly under the mandates of either the UN or ISF missions. The UN had no mandate to involve itself directly with the FDTL, its security role being limited to UNPOL’s ‘executive policing’ imperative in partnership with the PNTL.24 For its part, the ISF also lacked a mandate to directly involve itself with the FDTL’s 96

Significance for state-building

International efforts to solidify the state of Timor-Leste since independence in 2002 were severely hampered by the 2006–08 crisis. By themselves, neither the mutineers nor the petitioners were particularly significant. Reinado’s mutineers numbered only several dozen at most, while the petitioners never took up arms (aside from Salsinha and a few of his more militant colleagues towards the end of the crisis). However, both groups had profound ramifications for Timor-Leste’s stability for two major reasons: as a medium of agitation for the major political parties and because they catalysed a damaging lag between nation-building – consolidating a national identity – and state-building – solidifying the institutions of government. The more immediate utility of both the mutineers and petitioners was as tools for Timorese political parties to agitate with. What began as an internal FDTL labour dispute quickly escalated as it was opportunistically seized upon by major political actors with a keen interest in bringing down the Alkatiri government.27 Tensions between TimorLeste’s political leaders reach back three decades over ideological and personal divisions within the resistance movement (Richmond and Franks 2008: 189). The personal animosities between Gusmão and Alkatiri had long been an open secret in Dili but spiralled immediately after independence when Alkatiri’s FRETILIN drafted a constitution that 97

5 | The mutineers of Timor-Leste

management.25 Rather, the ISF’s purpose was to ‘ensure the promotion of long-lasting stability’ by providing a stabilizing presence of foreign troops in the country.26 Given these mandate constraints, the UN mission and ISF, as well as major donor states, mostly played behind-the-scenes roles, encouraging the government to seek peaceful solutions to situations that were essentially the FDTL’s internal disputes. What actions were undertaken by international actors towards either the mutineers or the petitioners had to be undertaken at the explicit request of the Timorese government. Assistance was requested at several points, for instance when Reinado and his compatriots escaped from jail in August 2006 and UNPOL was tasked by the government to arrest him. Timorese courts issued an arrest warrant but UNPOL asked the ISF to handle the actual arrest because the group was armed. The ISF was also requested to attack the mutineers’ base south of Dili after the latter’s raid on the police border post in February 2007. Otherwise, UNPOL and the ISF were largely limited to providing riot control during times of street violence. UNPOL also helped to provide security at the petitioners’ cantonment in Dili after February 2008.

created a weak presidency for Gusmão (Gorjão 2002: 326). Compounding tensions was international engagement with Timorese political actors that at times was decidedly unbalanced, namely a tendency by early UN missions to over-rely on Gusmão while expressing fears of Timor-Leste becoming a one-party state dominated by FRETILIN (Chesterman 2002: 66–9). Alkatiri and FRETILIN hence have long expressed suspicions of an international bias in support for Gusmão. By early 2006, these political tensions had worsened. Many TimorLeste observers felt that some of Gusmão’s political allies saw a chance to encourage public agitation against the Alkatiri government, given its ongoing dispute with the petitioners. The International Crisis Group, for instance, argued that ‘Gusmão’s supporters, if not Gusmão himself, appear to have decided to use the petitioners in that [political] s­ truggle’ (International Crisis Group 2006a: 8). Conspicuously participating in the April 2006 protests were individuals who would later become prominent members of Gusmão’s CNRT party. One political analyst further argued that there were ‘clear indications that opposition parties hijacked those protests to discredit and destabilize the government’ (Horta 2006). This manipulation was highlighted by the fact that immediately before the protests the Alkatiri government had announced with petitioner l­ eaders that an amiable solution was imminent, only for the petitioners to subsequently announce during the protests that they wanted parliament to be dissolved. By the time the second round of petitioners’ protests started in April 2006, many observers suspected that the political dynamics were ‘less about discrimination in the army and more about political control of the country’ (International Crisis Group 2006a: 8). Gusmão himself in his 23 March 2006 speech was openly critical of Alkatiri’s authorization to FDTL commanders to sack the petitioners, offering credibility to their complaints of discrimination. This political manoeuvring escalated when Gusmão publicly threatened in late June 2006 to resign unless Alkatiri did, citing the latter’s handling of the petitioner and mutineer developments. Alkatiri himself long maintained that the unrest of 2006, including the provocation of the petitioners and the armed actions of the mutineers, was a concerted plot instigated by opposition politicians to topple FRETILIN (ibid.: 11). When the crisis deepened in April and May, the petitioners and Reinado’s mutineers were in fact primarily agitating for the removal of the Alkatiri government and openly declaring support for Gusmão. Yet Reinado eventually condemned Gusmão as well, even sensationally claiming in January 2008 that Gusmão was the ‘mastermind’ of the 98

99

5 | The mutineers of Timor-Leste

2006 violence – including his mutineers’ actions.28 However, the UN investigation into the violence explicitly cleared Gusmão of directing the mutineers, noting that he ‘did not order or authorize the armed group of men under the command of Major Reinado to carry out criminal actions’ (United Nations 2006a: 30). Still, even if Gusmão did not personally encourage armed revolt against the Alkatiri government in 2006, he did not do much to prevent the turbulent situation from overwhelming it politically. Political agitation using the medium of petitioners and mutineers continued even after the April/May 2006 violence. The ongoing presence of the mutineers in the mountains and the lack of a resolution to the petitioners’ complaints provided fodder for the political opponents of FRETILIN to campaign on during the presidential and parliamentary elections of 2007. Gusmão’s CNRT party was successful, in coalition with the smaller parties that ultimately formed the AMP, in critiquing FRETILIN’s handling of the 2006 crisis as being unresolved, hence requiring other political parties to assume power.29 However, the most blatant manipulation of the situation came on 26 April 2007, when Ramos-Horta called off the military search for Reinado as part of an election-season deal for Democratic Party support (International Crisis Group 2007b: 7). The Democratic Party had its regional support base in Timor-Leste’s western half, where Reinado’s mutineers were encamped and enjoyed much public support, making their continued freedom popular with voters there. On the whole, there was very little political will among Dili’s political elites to resolve the major problems of the mutineers and petitioners in 2006 and 2007. Upon assuming power, however, the Gusmão government faced the need to respond to the ongoing presence of the petitioners and mutineers. Consequently, the administration itself became susceptible to charges that it was not resolving the country’s problems, as evinced by the lack of progress with either group. Officials began to feel that, despite their efforts at resolving the problems of national division, FRETILIN was seeking to use the petitioners and mutineers for agitation.30 As the opposition, FRETILIN members softened their rhetoric against both the petitioners and Reinado and complained instead of Gusmão’s inability to respond to their complaints. Even after the February 2008 attacks, political manipulation of the continued presence of the mutineers and petitioners carried on. There were oft-repeated charges by FRETILIN, and suspicions expressed by international observers as well, that the Gusmão government and the FDTL command were using the state of siege to establish the FDTL’s

institutional primacy over the PNTL.31 This was important in political terms because the FDTL historically had stronger ties to Gusmão than to FRETILIN (Chopra 2002: 982). Additionally, the FDTL had long felt marginalized by the UN and ISF, with General Ruak arguing that ‘I can do what they [the international forces] are doing, my soldiers can do this’ (International Crisis Group 2008c: 17). While it is arguable that the reserve shown by the FDTL towards apprehending the remaining mutineers was a wise decision in February and March 2008, it was also politically opportune. In this way the Gusmão government allowed the FDTL to conclusively assert its primacy over the PNTL in terms of national security: the state of siege constitutionally gave the FDTL command decisions over the PNTL through a so-called Joint Command, which was in effect between February and May 2008 in response to the February attacks. The relationship had been a simmering source of discontent because the former felt the latter had historically been given more support by international donors and the Alkatiri government. This had effectively sidelined the FDTL nationally and contributed to its institutional resentment of the PNTL (Hood 2006: 73). The FDTL’s desire to assert its prominence nationally came from both senior commanders and Gusmão, even as the 2006 crisis left the armed forces divided as an institution and a wellspring of national tension. Hence the state of siege, extended by the mutineers’ remaining at large, allowed the FDTL to rejuvenate as a national institution while becoming ‘reconciled’ with the PNTL. The second major ramification of the petitioners and mutineers for the state-building project was to catalyse the lag between efforts to entrench the new state and its institutions and the need to solidify a Timorese nation. As a UN head of mission once argued, ‘This is a country doing […] both nation-building and state-building at the same time.’32 The challenge was that what national identity existed in a state of numerous ethnic groups and languages was essentially founded upon anti-colonial nationalism, namely the idea that the ‘unifying discourses of blood, soil, and shared suffering form the basis of East Timorese national identity’ (Philpott 2006: 136–7). While this form of nationalism may have been useful during the resistance era as a consolidating dynamic, interpreting that resistance history since independence has proved to be a divisive exercise (Leach 2006: 222–37). Indeed, it was downright counterproductive, as easterner– westerner tensions proved highly susceptible to political manipulation centred on contested understandings of resistance history. This left the country prone to a damaging identity debate over who contributed more 100

101

5 | The mutineers of Timor-Leste

to achieving independence and hence who should be paramount within an independent Timorese state. The crisis of 2006–08, propelled as it was by tensions within the FDTL, pushed this identity discourse into becoming a contentious national debate open to political manipulation. Essentially, Timor-Leste suffered through a profound national identity problem manifested through the 2006–08 turmoil.33 This was especially so for the FDTL, given its place within Timorese society as a ‘living monument’ to the resistance era and the FALINTIL rebel movement in particular (in homage to that history, the FDTL’s full name is officially the FALINTIL-FDTL).34 In that sense, it should come as no surprise that the explosive ingredients of the national crisis emerged from identity disputes within the FDTL. Both the mutineers and the petitioners stemmed from the FDTL’s failings as an institution to improve its coherence and overcome historical animosities rooted in its own origins as a resistance movement against foreign occupation. Since its creation in 2001 the FDTL has proved susceptible to divisions (International Crisis Group 2008c: 70–72). Its 1st Brigade, founded in February 2001 from FALINTIL members, was 90 per cent easterners, so efforts were made to recruit westerners.35 By 2005, the FDTL was indeed fairly balanced, but many of the recruits were junior in rank to older, veteran easterners. The tensions within the FDTL between easterners dominating the officer corps and younger westerners recruited post-2001 were not easily managed. Thus, the easterner–westerner tension was not so much an ‘ethnic’ issue as a set of debates on regionalism and identity, conflated with the resistance experience. This was apparent because many FALANTIL veterans serving in the FDTL officer corps displayed a superiority complex towards westerners that derived from a belief that independence was mostly achieved by easterners.36 Even though tensions went back several decades, they assumed prominence when the petitioners alleged that this dynamic had become institutionalized within the FDTL through systematic discrimination. Another important source of tension, one that exploded violently in May 2006, was negative opinions held by older FALANTIL veterans serving in the FDTL about PNTL officers because many had served as police for the Indonesians. This was a sentiment that had developed because the UN’s initial recruitment efforts for the PNTL did see significant numbers of former Indonesian police hired (Martin and Mayer-Rieckh 2005: 112). Ultimately these tensions relating to legitimacy in the context of national identity formation, given often diverging interpretations of resistance history, overwhelmed the consolidation of the FDTL and subsequently the PNTL, by April and May 2006.

The development of the petitioners as a movement on a national scale exemplified the weakness of institution-building in Timor-Leste, what the UN special commission on the crisis lamented as the ‘frailty of State institutions’ (United Nations 2006a: 2). Even with a significant international presence since 1999, the security sector of Timor-Leste has remained chronically undeveloped (Hood 2006: 61). The Inter­national Crisis Group, for instance, argued that the UN has been unable to lay ‘adequate foundations’ for the security sector (International Crisis Group 2008c: 5). This was tragically highlighted by the reality of police and soldiers fighting one another on the streets of Dili in 2006. Had the security sector been better developed, the FDTL by itself should have been able to resolve a fairly basic grievance within its ranks, specifically the labour complaints of the original 159 petitioners. Yet the discord within the FDTL over historical animosities surrounding national identity ultimately proved a ‘challenge too big, too early’ for the state of Timor-Leste.37 The broader national debate over what it meant to be ‘Timorese’ in an independent era was negatively compounded by an environment of direly weak state institutions, notably security forces that did not enjoy a unifying national identity. This was especially so when societal antagonisms stemmed more from personal issues rather than ideational ones, viz. interpretations of resistance-era history versus national political aspirations. The pensioners and mutineers were the starkest example of violent conflicts arising over personal legitimacy in a culture that prioritized resistance to national identity formation. Lastly, it is important to qualify Timor-Leste’s fragility as a state vis-à-vis the 2006–08 crisis (Cotton 2007: 455–70). Most importantly, Timor-Leste always had a government in place and never suffered any sort of violent, sustained revolt by insurgents. Yet the chronic weakness of the security sector since 2002 has in fact meant that minor and arguably manageable incidents escalated to cause national crises.38 The key requirement for state-building is that solid institutions, especially those of the security sector, are capable of surviving individual governments, whatever their respective strengths and weaknesses.39 In a highly contested political space where actors were deeply divided, especially FRETILIN versus Gusmão and his allies, this weakness showed itself clearly and propelled Timor-Leste into a disastrous situation: nearly half the state’s defence forces left their posts (the petitioners) and subsequently were sacked by their commanders. Others (the mutineers) deserted with their arms and eventually attacked the national leadership, leaving the president nearly dead. 102

Timor-Leste has enjoyed significant amounts of international support stretching back to the initial UN mission in 1999. However, even with this support the country’s development and stability have been stymied, particularly by the 2006–08 crisis, through the presence of the mutineers and the government responses to them. From a policy perspective, the mutineers highlighted the limits of what international actors, be they the UN or major regional powers such as Australia, can achieve within the context of an indigenous political situation that is in manic flux. Only limited policy tools were available, since much of the turmoil was political. That is, it resisted foreign, ‘programmable’ activities capable of quickly responding to contentious Timorese political issues such as identity formation. However, there are still two important lessons that can be garnered from Timor-Leste’s experiences with the mutineers. The first lesson for the international community is how important mandates are to resolving issues regarding militias. Specifically, this means that comprehensive security sector mandates for the UN and/ or international peacekeepers are crucial. While UNPOL could monitor and influence the PNTL extensively, there was no comparable oversight of the FDTL because the UN had no mandate for extensive involvement. After the initiating violence of 2006, UNPOL received a renewed mandate for restoring and maintaining order and for building up the PNTL, yet the UN more broadly had no strengthened mandate to engage the FDTL more deeply.40 For its part, the ISF security forces that arrived after the initiating 2006 violence were limited by their security ‘watcher’ role of merely maintaining a calming presence. The requirement that the government explicitly request support in ­order for the ISF or UN to undertake specific coercive actions, especi­ ally in relation to the mutineers’ armed presence, was exacerbated by dam­aging Timorese political dynamics, specifically the political parties agitating against one another. Both the petitioners and mutineers, once they had manifested themselves through the April and May 2006 violence, left the UN and ISF unable to act decisively because of their limiting mandates. This relegated both entities to playing weak secondary roles relative to the government and the FDTL, essentially using their ‘good offices’ to encourage peaceful solutions.41 Ultimately, the mutineers as well as the petitioners were representative of the crises of the FDTL. If the cohesiveness and development of the FDTL had been more effectively strengthened, starting in 2001, it is likely that the 2006–08 crisis would never have happened. The reality was that the international community, through the UN missions, was never 103

5 | The mutineers of Timor-Leste

Conclusion

able to engage the FDTL in a way that permitted significant influence over the armed forces’ development, and hence their cohesiveness. This contrasted markedly with the stronger international emphasis that was placed on the PNTL. These dynamics highlight the fact that peacekeeping mission mandates that engage one major part of the security sector, the police force, while neglecting another, the army, are likely to provoke significant problems in the longer term. Despite the need for emphasis on rationalizing the FDTL, reform of the entire security sector has lagged. As the International Crisis Group has argued, the Timor-Leste government ‘has taken few steps to a­ ddress the problems in the security sector which led to the 2006 crisis’ (International Crisis Group 2009c: 1). This has meant that personal relationships and key leaders within the political elite continue to dominate the functioning of the army and police with limited institutionalization. Dampening tensions and resolving disputes since 2006 have continued to rely on personal dynamics, namely ‘political quick fixes based on personalities’ rather than institutional protocols and mechanisms (International Crisis Group 2009b: 1). More worrying, following the disintegration of the PNTL in April and May 2006, the FDTL had been able to broaden its own role in internal security provision, especially through its dominance of the Joint Command after the attacks in February 2008. A strengthened national role for the FDTL continues in 2010, but the army has not been subject to even minimal reform efforts by the international community, this despite the fact that the mutineers came out of the FDTL rather than the PNTL. Given the previous point, the challenge for longer-term national security is that the FDTL’s own stability continues to be rooted in the charisma and stature of former rebels and resistance heroes like Gusmão and the FDTL commander, Taur Matan Ruak. This may suffice presently, but hardly bodes well for when new generations assume national leadership. Compounding the lack of reform to the FDTL is the fact that the transformation of the PNTL, while at least attempted by the inter­ national community, has been limited since 2006. Notably, efforts to improve the PNTL have been slowed owing to continuing ambiguous and poorly implemented mandates for UNPOL. Its evolution from an executive policing function sparked by the April/May 2006 violence to instead playing a more secondary, advisory task as the security situation stabilized after April 2008 has been ambiguous and tortured (ibid.: 4–6). Additionally, with a significant role played by Gusmão, the distinction between the police and the army has grown even hazier, with the PNTL being promoted with a ‘paramilitary style of policing’ to the detriment 104

105

5 | The mutineers of Timor-Leste

of an emphasis on community policing (ibid.: 1). Considering that so much of the violence starting in 2006 was located within communities, this is arguably an unfortunate policy decision. A second major lesson from Timor-Leste’s experience with the ­mutineers is that in the context of limited international mandates, ­timing is crucial for taking decisive action. The Timorese crisis of 2006–08 showed that all too often mundane situations were allowed to fester, and hence consistently escalated into worse problems involving small groups such as the mutineers and petitioners. The best example of this was the handling of the initial FDTL petition in January 2006. Without a strong response by either FDTL commanders or the Alkatiri government in January and February, by April the dispute had expanded into protests that involved nearly six hundred petitioners and sparked a national crisis. Another example was the escape from jail of Reinado and fourteen of his mutineers in August. Arrest warrants were issued but his continued evasion of capture provoked inconsistent policy responses from the government and an apparent unwillingness by both the UN and major states, notably Australia, to force the issue. The lesson to be learned is that quick, targeted interventions and/ or pressure at crucial moments need to be applied by the international community on host governments. In Timor-Leste, the best oppor­tunities were the initial petition in January 2006 (to keep the issue as an in­ ternal, bureaucratic labour dispute) and the prison escape of Reinado in August (to prevent the government from procrastinating in its a ­ ttempt to capture the armed mutineers). In the context of a small state under­ going intense political and social flux, the damage that can be done by a few influential individuals and groups can be profound. The mutineers proved that disastrously. That quick and decisive action can be taken was seen in the positive international response to the February 2008 attacks. The UN quickly evacuated the wounded Ramos-Horta to Australia and inserted a calming presence of UNPOL officers on the streets. Australia deployed additional ISF soldiers immediately. For its part, the Gusmão government also responded quickly to prevent the street violence that had been so devastating in 2006 from recurring, by implementing a strong PNTL presence on the streets supporting UNPOL. Civil-society actors were also quick to respond. For example, the Catholic Church immediately called in gang leaders from turbulent neighbourhoods to urge them to maintain calm in their communities. We may conclude by highlighting the fact that Timor-Leste is still a state with much potential to achieve stability, despite pessimism provoked by the recent past. In an attempt to keep the situation in

perspective, Ramos-Horta argued in 2007 of the ongoing unrest, ‘There is no civil war or bombs bursting on the streets […] these are just growing pains of a young country.’42 That the incidents of February 2008 did not spiral violently out of control as happened in April and May 2006 is the best sign that Timor-Leste is in fact stabilizing: no governments were toppled and neither was there wide-scale street violence. Furthermore, it is notable that after the cantonment of the petitioners in February 2008 and the surrender of the mutineers in April, the political tensions surrounding them calmed and there has been no further violence directly related to them. In a significant step towards entrenching the rule of law, following a seven-month trial, twenty-three of the mutineers were convicted of crimes relating to the attacks on Gusmão and Ramos-Horta in March 2010. Salsinha was sentenced to nearly eleven years in jail and others involved received sentences of up to sixteen years. The presence of the petitioners and the resultant national tensions have also been largely resolved and have certainly ceased to be a cause of widespread instability. Most petitioners took advantage of the government’s financial incentives to cease agitating against the government and have returned to normal civilian life. Importantly, with the subsidence of the petitioners as a movement and the peaceful trial of the mutineers, the broader national tensions between ‘easterners’ and ‘westerners’ have also waned. This is not to say that such tensions no longer exist, simply that since early 2008 they have not escalated into violence. Despite that, President Ramos-Horta would surely be the first to conclude that Timor-Leste still has a long way to go towards stability that can be sustained without direct international support. Hence, he has called for the UN to stay until 2012, giving Timor-Leste ‘the time and space’ to further stabilize.43 Whatever international presence is ultimately implemented during the balance of that time, key to that support must be a comprehensive effort at consolidating the security sector so that relatively small groups, such as the mutineers, cannot again provoke national crises.

106

6 | Militias in the eastern Democratic Republic of Congo

The DRC remains a fragile ‘post-conflict’ state, as violence and political instability continue to persist in the country, especially in the eastern regions. The country experienced a number of political crises beginning with the overthrow of former president Mobutu, and culminating in the devastating civil war from 1998 to 2003 that drew in the involvement of over ten African countries and resulted in significant human costs. The DRC conflict resulted in the deaths of approximately 5.4 million Congolese between 1998 and 2008, half of them children; 1.3 million internally displaced persons (IDPs) and half a million Congolese refugees in 2007 alone;1 and the systematic use of sexual violence against women by both the members of the Congolese army and militias and armed groups. Although the conflict in the DRC came to an official end in 2003, the task of post-conflict reconstruction has proved to be much more difficult. However, despite numerous coup attempts, sporadic violence and the repeated delay of general elections, a fragile peace has gradually taken hold in parts of the country. A notable exception to this has been the eastern parts of the country, specifically Ituri and the Kivus, where numerous armed groups and militias took part in the national demilitarization process, against a backdrop of increased tensions and violence perpetrated by the rise of dissident militias. This chapter seeks to provide an analysis of the demilitarization process in the eastern DRC, while placing greater emphasis on continued efforts to demilitarize dissident militias that emerged in the post-2003 period. Through fieldwork conducted in the Kivus and Ituri,2 this chapter argues that a textured and nuanced understanding of the countryspecific context must be central to any demilitarization process, and a cookie-cutter, technocratic approach to demilitarization is in­effective and in the long run unsuccessful. This conclusion is in line with the overarching perspective of the book, and the realities on the ground revealed that the demilitarization process in the DRC was marred by a number of challenges based on certain contextual realities. While these factors are not necessarily unique to the DRC, as these challenges can be found in most post-conflict states, the specificities and nuances of 107

the Congolese context make this case study unique, allowing for future comparative analysis with other post-conflict case studies. The most significant challenge facing the demilitarization process has been the rise (and fall) of numerous dissident militias, especially in Ituri and the Kivus.3 Since the post-2003 period, numerous militias have surfaced, not only challenging the demilitarization process, but also undermining overall post-conflict efforts. This chapter will discuss the implications of three armed groups and militias that have been especially active in the post-2003 period: i) pro-government militias, the Mayi-Mayi; ii) the broad rebel movement, known as the Congrès national pour la défense du peuple (CNDP); and iii) the Forces démocratiques pour la libération du Rwanda (FDLR). The transitional justice process, specifically the issue of providing amnesty to commanders and the extension of deadlines for armed groups to join the demilitarization process, has played a crucial role in the resistance of certain armed groups to demilitarization. As well, in the broader context of post-conflict reconstruction, the implementation of a demilitarization programme while renegade militias were still operating had the poten­ tial to undermine the overall peace process. Another crucial factor that undermines the demilitarization process centres on the weak Congolese state and security apparatus. The country’s historical trajectory of violence and authoritarianism created a contextual backdrop which not only increased the likelihood of conflict, but also made the task of building strong state institutions an arduous one. This, fused with localized conflicts over land and natural resources and the role of regional actors, has contributed to the weakness of the Congolese state. In addition, the implementation of the demilitarization process has been problematic. The disarmament, demobilization, reintegration (DDR) programme had been poorly funded (by the World Bank), while the Congolese coordinating bodies that oversaw the demilitarization process have been institutionally weak and hampered by political problems. In addition, the demilitarization process was poorly thought out, with an overemphasis on the technical and logistical aspects of disarming and demobilizing former combatants, with reintegration remaining a secondary consideration. The demilitarization process was further confounded by the overly ambitious agenda of the UN mission, Mission des Nations Unies en République Démocratique du Congo (MONUC) and the international community, which focused on simultaneously implementing demilitarization and security sector reform (SSR) processes, resulting in a poorly organized, poorly trained, underpaid and at times criminalized national army (the Forces armées de la République 108

Demilitarization efforts

Briefly, demilitarization is the process that encompasses the dis­ armament and demobilization of combatants, and their integration into society. Similar to the situation in a number of post-conflict states such as Uganda, Timor-Leste and Afghanistan, in the DRC the DDR process was interfused with SSR, as combatants were given the choice of reintegrating into civilian life or joining the Congolese national army, the FARDC. This combined ‘dual-track’ (or tronc commun) process linked DDR with army integration. Both DDR and SSR activities involve the security sector (military and police), and both activities have overlapping institutions responsible for their management, including MONUC and the Congolese army. Also, addressing the needs of former combatants is directly linked to opportunities to reform the security sector, especially in order to achieve longer-term security and development.4 Here it is important to clearly define SSR, which in the broader sense tends to refer to the processes of reforming the army, police and other security services, within a reformed judicial system (Bryden 2007: 2). Demilitarization efforts are now supervised by the Ministry of ­National Defence, Demobilization and Ex-servicemen, and managed by the Programme Implementation Unit (PIU), which is attached to the office of the minister in charge of demobilization. The DDR activities are guided by the Programme National pour le Désarmement, la Démobil­ isation et la Réinsertion (PNDDR), and financed by various international agencies, including the World Bank, UNDP, African Development Bank, and donor agencies. The PIU replaced the Commission Nationale pour la Démobilisation et la Réinsertion (CONADER), owing to serious problems with coordination with other governmental agencies. From an early stage, there was a lack of clarity and vision between the three institutions leading the national demilitarization programme: the CIDDR (Comité Inter-Ministériel chargé de la Conception et de l’Orientation en matière de Désarmement, Démobilisation et ­Réinsertion), which 109

6 | Militias in the eastern DRC

Démocratique du Congo, FARDC). Finally, factors including the politicization of ethnicity, the role of natural resources and the role of external actors supporting various armed groups will also be briefly examined throughout the chapter, so as to provide a holistic understanding of the contextual factors that challenge the demilitarization process. The chapter will first provide a brief introduction to the demilitarization process as well as the post-conflict environment; followed by a critical examination of the contextualized factors that challenged the demilitarization process in the DRC.

developed policies, supervised the implementation of DDR activities; CONADER, which was responsible for the overall coordination, management and implementation of DDR; and CGFDR (Comité de Gestion des Fonds de Désarmement, Démobilisation et Réinsertion), which was created to provide financial management and procurement support. As such, there was widespread competition among the three agencies, especially between CONADER and CGFDR, thereby leading to delays in the procurement process, scheduling and implementation of the DDR programme. Although the CGFDR was abolished in May 2005 and its responsibilities transferred to CONADER, the weak institutional cap­ acity of and allegations of corruption within CONADER undermined the demilitarization efforts. The World Bank and donor community provided CONADER with a joint contribution of US$200 million for the implementation of the DDR programme, amounting to a unit cost of just over US$1,200 per person.5 The SSR process was coordinated by the FARDC, and major international partners involved include Belgium, Angola, South Africa, the EU and France; although support for SSR continues to be confronted by significant coordination and coherence issues (Hoebeke et al. 2008). Those former combatants who chose reintegration into the FARDC become the responsibility of the Structure Militaire d’Intégration/Integrated Military Structure (SMI), which, along with the FARDC (and MONUC for logistical support), transported the former combatants to the centres de brassage. The brassage process runs for about forty-five days of ‘military training’, after which the former combatant is officially integrated into the newly unified army and is posted to a FARDC brigade, usually different from his place of residence. The main function of the brassage process is to break the chains of command of the different warring groups. However, armed groups and militias such as the CNDP and the Mayi-Mayi had demanded that former combatants who are integrated into the national army remain in the region where they operate. (Hence the introduction of the mixage process. Following the 2006 conflicts in North Kivu, the FARDC made a difficult compromise with Nkunada’s CNDP in 2007 by ‘mixing’ brigades loyal to Nkunda with FARDC brigades; a practice that was not supported by the international community and MONUC.) Mixage differs from brassage as combatants will not go to a brassage centre but will be ‘mixed’ in their own area and deployed there. This issue will be discussed later when examining the various armed groups still operating in the eastern regions of the country. The precursor to the national DDR programme was the DDR programme in Ituri, known as the Disarmament and Community Reinser110

111

6 | Militias in the eastern DRC

tion Programme (DCR), launched in September 2004. Between 2004 and 2006, the DCR programme implemented by CONADER and UNDP disarmed and demobilized 15,811 adult combatants; and has, since 2006, been formalized as part of the national DDR programme (Bouta 2005). The programme aimed to disarm and reintegrate into civilian life the various Ituri armed groups and militias.6 Although many senior commanders of the militias and armed groups were determined to maintain their power and protect their economic interests, there was an overarching favourable reaction from the combatants, as many of them were interested in the community integration component, which focused on small income-generating activities and vocational training. However, despite initial successes of community sensitization and voluntary disarmament, the community reintegration phase of the DCR programme was problematic owing to financial management, disbursement delays and lack of effective community reintegration projects (Amnesty International 2007: 21–3). This led to the resurgence of violence, and some combatants rejoined militias and armed groups, including the Union des patriotes congolais (UPC), the Front de résistance populaire de l’Ituri (FRPI) and the Mouvement révolutionnaire du Congo (MRC). However, a number of the leaders of Ituri militias decided to partake in the DDR/ SSR process, many of them initially garnering high-ranking military portfolios – only later being handed over to the International Criminal Court (ICC).7 Unable to learn from the lessons of the DCR programme, the national demilitarization programme faced similar challenges. As of 2008, there were eighteen integrated brigades, composed of former combatants as well as members of the former Congolese army (Forces armées congolaises, FAC); and the composition of the brigades is determined by the dominance of certain armed groups. For example, the 5th brigade had a number of former combatants from the Mouve­ment pour la libération du Congo (MLC) and the Rassemblement congolais pour la démocratie-Goma (RCD-Goma), as it was originally stationed in North Kivu, but later relocated to Kananga, Kasai-­Occidental. Generally, in the rest of the country, the brigades have a greater deal of cohesiveness. However, breaking up the different armed groups and militias in the eastern region has proved to be a difficult task, as different factions remain suspicious of each other, and therefore co­hesion remains ‘fragile’ (United Nations 2009: 2). It should be noted that a number of integrated battalions, i.e. those that have gone through the brassage process, are being trained by various bilateral donors, in­cluding France, Belgium, the EU and the USA. Owing to this, integrated battalions receive varied levels of military training

and strategic ethos. This is confounded by the fact that numerous battalions are not yet integrated owing to the halting of the DDR programme, resulting in a weak and uncoordinated national army. There has been piecemeal support by various bilateral donors, including Belgium, France, the USA, the European Union and South Africa, to provide army reform programmes, mainly to assist with the training of brigades or pro­viding salaries to the soldiers. In the short run, they have been effectively able to train and pay soldiers; however, in the long term the problem of sustainability in paying wages, and ­differing military training/ethos, could be detrimental to the national army. Those who opted for civilian reinsertion were given US$110, which is meant to be used for transport for returning back home and supporting basic needs. After that, for the next twelve months, the former combatant receives $25 to meet basic living needs and to ‘hold them over’ until a reintegration programme has been established in their community.8 As of February 2007, CONADER had disarmed and demobilized 115,000 adult combatants and 54,000 children associated with armed groups and forces (enfants associés aux forces et aux groupes armées, EAFGA).9 Of the adult combatants, approximately 45,000 have been reinserted,10 and the remaining 70,000 await reintegration programmes in their communities. Most of the children have been returned home to their families, with the support of UNICEF.11 There are an additional 85,000 adult combatants, and an estimated 15,000–20,000 children, currently waiting to go through the DDR process, owing to the World Bank’s termination of support for CONADER.12 In addition, owing to the presence of foreign armed groups and militias in the eastern DRC, the DDR process operated parallel to the disarmament, demobilization, repatriation, reintegration or resettlement (DDRRR) process for foreign armed groups, coordinated mainly by MONUC, guided by the Multi-Country Demobilization and Reintegration Programme (MDRP) spearheaded by the World Bank and UNDP. With regard to DDRRR, as of February 2007 MONUC had repatriated a total of 14,100 soldiers (and dependants) from foreign armed groups, mainly from Rwanda, Uganda and Burundi, since 2003.13 However, the most notorious foreign armed group, the FDLR, was unable to be entirely terminated by the DDRRR programme, and continues to remain a central security concern for the eastern regions of the country. It is now the target of sustained cooperative efforts between the Congolese and Rwandan armies, with the support of MONUC.

112

Although the Congolese government, with the support of MONUC and other donors, has made significant efforts to demilitarize Congolese and foreign armed groups,14 numerous groups such as the FDLR, Mayi-Mayi and CNDP have continued to operate in and occupy resourcerich areas of the Kivus and Katanga15 with virtual impunity. Based on extensive field research in the eastern DRC (Kivus and Ituri), this section will provide a discussion of the character, objectives and context of these militias. This analysis utilizes the framework of analysis for conceptualizing militias discussed in Chapter 2. Since the post-2006 election period, the violence perpetuated by the CNDP, the FDLR and the Mayi-Mayi in the Kivus remains the major factor that challenges the demilitarization process, and undermines the overall peace process. In order to address the rise of these militias, the Nairobi and Goma agreements of November 2007 and January 2008 made way for the launch of the Amani Pact (also known as the Pact d’engagement). The agreements were signed between the government and twenty-two rebel groups, and addressed the issue of demilitarizing the signatories and the repatriation of internally displaced persons. Although the CNDP, led at the time by renegade general Laurent Nkunda, promised to undergo the DDR process and to transform itself into a political party, violence ensued between the CNDP and the FARDC, MONUC and Mayi-Mayi between August and December 2008. The CNDP is motivated to protect the Kinya-rwandan Tutsi popu­ lation in the Kivus against the FDLR, as well as other ‘anti-Tutsi’ groups such as the Mayi-Mayi. The organization, led by Nkunda, mainly ­operated in North Kivu; however, he was instrumental in the capture of Bukavu (South Kivu) in 2004, in collaboration with General Jules Mutebusi. Other armed collectives including the Group of 47 and the Moramvia Group operate clandestinely in South Kivu (Thakur 2008a: 59). However, North Kivu has remained the domain of Nkunda, especi­ ally in the post-2006 period. In this regard, he has been able to mobilize well over four thousand combatants (directly under his control) to support his military operations.16 In addition, Nkunda has commanded significant support from the 81st and 83rd brigades in the FARDC, most of whom were formerly loyal to the dismantled RCD-Goma, totalling 6,000 soldiers – all of whom have refused to partake in the army integration process.17 The co-option of rogue FARDC brigades has led to uneasy but lucrative partnerships between Nkunda and the FARDC, including opportunitistic senior offi­cers, such as Captain Akatomba of the 118th brigade of the FARDC in North Kivu. Akatomba has asserted 113

6 | Militias in the eastern DRC

Rise of militias in the post-2003 period: an empirical mapping

that he and his soldiers would never join the army integration process because the Tutsi population will not be properly defended under an integrated army.18 In January 2009, Nkunda was arrested by Rwandan authorities in Rwanda territory, in response to his resistance to the joint Congolese– Rwandan military operation targeted at the FLDR. Nkunda’s arrest is being viewed as part of a military cooperation agreement between the two governments, in which the arrest of Nkunda has allowed Rwanda to send Rwandan troops into the DRC.19 The detention took place just a couple of weeks after a splinter faction of the CNDP, led by Bosco Ntaganda,20 agreed to partake in the DDR process as well as assist with the joint military operation against the FDLR. With the splintering of the CNDP, there have been new prospects for peace in the region, culminating in the March 2009 agreement between the government and the CNDP. Under the 2009 agreement, former CNDP combatants would be integrated into the FARDC and the police. However, there are major concerns about the CNDP’s terrible record of human right violations; and that those individuals responsible for serious abuses against civilians must be brought to justice, rather than being rewarded for their previous behaviour. The refusal of the Congolese government to hand over Bosco Ntaganda to the ICC for war crimes committed in Ituri reinforces this concern about human rights abuses being ignored for political expediency.21 The Mayi-Mayi are pro-government militias that have historically fought alongside the government during the two Congolese wars, usually as part of a local defence force, against rebel movements and foreign armies. The Mayi-Mayi were treated as an independent entity and participated in the Inter-Congolese Dialogue in 2002, and as such were given positions in the National Assembly, provincial administration and national army. However, owing to their alliance with the government, they never formed a strong political organization of their own, and eventually splintered into several groups (International Crisis Group 2006b). Therefore, in the post-transition period, with the rebel threat contained, the Mayi-Mayi militias have taken on a life of their own and are now one of the most significant security threats in the country.22 In the post2004 period, there were a number of Mayi-Mayi groups operating in North Kivu (one led by Colonel Yemeni Jackson, one by twelve-year-old Colonel Baraka, and one operating in the ­Vurondo region), with a collective strength of 1,600–2,300 combatants. Although the groups led by Jackson and Baraka have been dismantled since February 2007, small bands of approximately three to five hundred combatants continue 114

115

6 | Militias in the eastern DRC

to avoid DDR and engage in personal enrichment activities (Thakur 2008a: 57). However, the Mayi-Mayi group operating in the Vurondo area is composed of individuals who are interested only in personal enrichment. With no clear leader and a collective strength of 600–800 soldiers, the group is involved in pillaging, stealing and raping local populations. They have no intention of demilitarizing, and remain as criminalized, armed bands in North Kivu.23 South Kivu is also home to well over seven different Mayi-Mayi groups, with a collective strength of 3,500–4,500 combatants. Those groups that have undergone the DDR process, such as armed collectives led by local warlords Colonel Dunia, Colonel Kalala Ruhara, Colonel 106 and Colonel Abdul, demanded that the commanders and former combatants operate in their local ‘areas of origin’ (ibid.: 58). In July 2007, and most recently from August to September 2008, many other bands of Mayi-Mayi came together to form the superstructure of the Coalition of Congolese Patriotic Resistance (PARECO) (under General La Fountaine), and provide support to the FARDC in its fight against the CNDP. Despite the presence of other foreign armed groups, such as Ugandan militias – Allied Democratic Forces-National Army for the Liberation of Uganda (ADF-NALU) and the Lord’s Resistance Army (LRA) – and Burundian militias – Front de libération nationale (FLN) and Conseil national pour la défense de la démocratie-Forces pour la défense de la démocratie (CNDD-FDD) – the Rwandan militia FDLR remains the largest foreign armed group in the eastern DRC. The FDLR was formed in 2000 as a political movement, and is composed of ex-FAR (Forces armées rwandaises) and interahamwe elements that fled to the eastern DRC during the 1994 genocide. This group has a radical pro-Hutu revisionist view of the Rwandan genocide, and aims to organize military operations against the Rwandan government (International Crisis Group 2005a). The armed wing of FDLR, known as FOCA (Force combattant abachunguzi), plays a significant role in the decision-making process of the FDLR. The FDLR’s estimated strength in 2007 was about 8,000 combatants in the Kivus (5,000 in North Kivu, and 3,000 in South Kivu),24 although currently it is approximately 5,000.25 As mentioned above, joint Congolese–Rwandan military operations are being undertaken to crush the FDLR in the eastern regions of the country.26 MONUC has been unable to dismantle the FDLR within the DDRRR process, under the auspices of the MDRP, owing primarily to the elusive and clandestine nature of the FDLR, as well as the recruitment and co-option of local communities within the organization and the internalization of their ideology (usually through coercion). It would be imprudent to firmly

assert that the local communities ‘protect’ FDLR elements, as the experience of the local population with the FLDR varies from area to area.27 In January 2010, the FARDC and MONUC jointly launched Amani Leo (translates from Kiswahili as ‘peace today/now’), which aims to regain control of mining territories in North and South Kivu from rebels, especially the FDLR; while also ensuring security for the local population.28 The objectives of militias operating in the eastern DRC can be understood at three mutually constitutive and reinforcing levels. The first and overarching level is emphasis on personal enrichment (at the individual and leadership levels), and ‘criminal’ economic-related activities (at the organizational level), including occupying and controlling mining areas, keeping supply routes open for arms/ammunition and logistics, arms trafficking, tax collection, and looting. Both the FDLR and CNDP are propelled by this factor. However, there needs to be a division between the individual motivations of combatants and those of the leaders of the armed groups and militias. Individual combatants find themselves in a disaggregated and hostile environment, where the presence of state ­authority is a hollow shell propped up by the international community and MONUC. In addition, the austerity of the economic and social con­ ditions in the eastern regions further provides individual combatants with the incentive to engage in armed activity. The individual combatants receive far less of the economic benefits, as compared to high-ranking officers within an armed movement; which thereby forces individuals to exploit and loot local populations to supplement their income. However, many are not willing to challenge leadership and make themselves a potential target should they leave the militia or attempt to splinter. One former combatant commented that ‘the AK means power and life’, assert­ing that beyond merely providing for themselves and their fam­ ilies, power and life symbolized having a livelihood, a sense of self-worth, and control over their destiny and life.29 However, not to romanticize individual motivations, within the rubric of personal enrichment there are clear elements of economic exploitation. In addition, it is important to note that the contextual reality of the inability of the state to assert itself in the east, especially with regard to social development, redistribution of resources and power, and provision of security, allows this economic predation to feed into these armed militias’ objectives. Closely linked to personal enrichment and economic survival is the motive for many militia groups of providing security to their commun­ ities, albeit one following a dangerous ethnicized trajectory. The use of ethnicity and protection, even at the level of rhetoric, enables armed groups to elicit local sympathies, and has proved to be an effec­tive 116

117

6 | Militias in the eastern DRC

strategy for recruiting and retaining combatants. This thereby forms and cements a common and mutually binding identity, as well as creating an elusive ‘other’, to enable ongoing armed and economic activities. For example, the Mayi-Mayi militias have historically been a local defence force30 recruited and organized along ethnic lines, and worked very closely with the Congolese army, usually in opposition to groups such as the CNDP in North Kivu. The key focus of the Mayi-Mayi militias has been on protecting land and defending local populations against ongoing violence by various armed groups and militias, as well as protecting ‘Congolese’ soil against foreign incursion. The CNDP’s rhetoric of protecting the ethnic Tutsi community in the Kivus, who felt politically and ethnically targeted by non-Tutsi communities and armed militias such as the Mayi-Mayi militias and FDLR, as well as certain battalions in the FARDC, also reinforces this view. The FDLR, on the other hand, demands that the Congolese government organize an ‘inter-Rwandan dialogue’, with the aim of establishing a power-sharing political arrangement in Rwanda. President Kagame has long refused to participate in such negotiations, and as such FDLR has restated its aim of gaining political power in Rwanda militarily. Although the FDLR manifesto stresses reconciliation, political power-sharing and development, its discourse is heavily tainted by a purely ethnic interpretation of Rwandan history and politics, thereby leading to the Congolese government and international community refusing such political demands.31 The organizational structure of militias in eastern DRC ranges from highly centralized formal units to informal, disparate units. The Mayi-Mayi militias are mostly localized and informal structures, and their numbers range from groups of 100 to over 1,000. According to ­MONUC officials, most of the Mayi-Mayi are unarmed, except for prim­ itive weaponry. However, owing to the prevalence of a significant arms trafficking network, they can easily gain access to small arms; therefore a disgruntled group of Mayi-Mayi with access to small arms could become a ‘nuisance’ for ongoing and future post-conflict activities.32 The leadership and soldiers are Congolese, and usually from local villages.33 In contrast, the CNDP and FDLR are highly organized, in comparison to the Mayi-Mayi. The CNDP was previously under the direct command of General Nkunda (with the splinter CNDP group led now by Ntaganda), and the organization has a clear chain of command which operates along the local community and clandestine networks. The organization receives some level of financial and political support from the l­ocal ‘Tutsi’ population and leadership, as well as moral support based on a natural sense of ethnic identification and survival. In addition,

Nkunda has strategically co-opted various elements of the FARDC, formerly RCD-Goma soldiers, to align with the objectives and activities of these armed collectives. The armed wing of the FDLR, known as FOCA, observes a much stricter military structure, with highly mobile battalions. Some of the FOCA combatants are Congolese and under the age of eighteen, most of whom were either kidnapped or coerced into joining the organization; and the leadership is predominantly Rwandan or ‘Hutu’-Congolese.34 These militias employ a number of strategies to achieve their res­ pective objectives, ranging from attrition, exploiting natural resources (especi­ ally CNDP and the FLDR35), forced recruitment (including, although not restricted to, children), manipulations of identity, and partnerships based on expediency and mutual economic benefit. Owing to their relative asymmetric weakness, in purely military terms, most militias employ a largely nocturnal strategy, including being most ­active in remote areas where dense forests and mountain terrain hinder search-and-destroy military operations,36 frequently dispersing after the attack, and making small-scale assaults on weak army positions. As well, combatants in the FDLR receive ideological indoctrination, mainly to reinforce the dangerous Hutu–Tutsi ethnic divide. The FDLR especially control their troops through terror and discipline – there have been many reported cases of summary executions of FDLR deserters.37 In addition, the filling of the ‘security’ vacuum by various militias, such as the CNDP, FDLR and Mayi-Mayi groups in the eastern regions, whether at the level of rhetoric or political strategy, is interfused within a dangerous politicization of ethnicity. Briefly, the ongoing violence in the Kivus has a politically charged ‘ethnicized’ face, finding its roots in the second Congolese war, in which the dominant role of Rwanda and Uganda inflated the local, historically contained conflict over land tenure issues and resources into a regionalized ethnic war (see Human Rights Watch 2001). In the Kivus, the Kinya-rwanda speakers of Tutsi ethnic origin, referred to as Banyarwanda in North Kivu and Banyamulenge in South Kivu, were placed at the centre of the second Congolese war and eventually ‘victimized’ for the political interests of local and regional actors. Their identities continue to be manipulated along dangerous Hutu–Tutsi and Bantu–Niolotic ethnic lines along the broader Great Lakes rwandophone ethnic configuration, despite centuries of mutual coexistence and intermarriage among the many ethnic groups in the Kivus. The CNDP base their concern on the exceptionalism of the Kinyarwanda Tutsi community in the Kivus, and the fact that they continue to be targeted by the FDLR and the Mayi-Mayi. Overlapping with this 118

The weak Congolese state

After a fragile and difficult political transition from 2003 to 2006, the delayed elections of July–October 2006 (the first multiparty elections to be held since 1960) brought promise of security, state accountability and justice, and prospects for a transition to long-lasting peace. Despite some post-election violence which led to the exile of former vice-president and opposition leader Jean-Pierre Bemba,38 the rest of the country has remained ‘relatively’ peaceful under the leadership of President Joseph Kabila, except for the Kivus and Ituri regions. The renewed violence in the Kivus following the elections highlights the larger question of whether the DRC can adequately address the fundamental challenges of governance and security, two pillars that are essential to a stable, viable state. The overarching contextual reality affecting the demilitarization 119

6 | Militias in the eastern DRC

imprecise yet powerful identity polarization, the Mayi-Mayi have also internalized the discourse on autochthony/allochthony, which essentially reinforces the ‘local’/‘stranger’ and ‘originaires’/‘non-originaires’ dualities. The ‘local’/‘stranger’ duality of autochthony/allochthony expresses itself in the DRC through rumours, political tracts and speeches, and draws energy from imprecise overlaps with other powerful and pre-existing identity polarities ( Jackson 2006). As such, owing to their strong claim as ‘authentic’ Congolese, or ‘autochthones’, the MayiMayi in Kivus have reconstructed their identity and are generally antiKinyarwanda, anti-Tutsi and anti-Rwanda. Therefore, the key focus of the Mayi-Mayi has been on defending and protecting ‘Congolese’ soil/ interests against domestic incursions supported by foreign actors. This politicization of ethnicity has detrimental effects on the demilitarization process, especially in providing a dangerous but potent narrative to justify ongoing violence. This discussion of the various militias operating in the eastern DRC is intended to illustrate that the presence of violent actors not only under­mines the demilitarization process, but could potentially overturn the gains achieved in the peace process. In addition, the insistence on pushing ahead with the demilitarization process by both the Congolese government and the international community, without necessarily ensuring a peaceful environment and filling the security vacuum in the eastern regions (with a robust national security apparatus), created insurmountable challenges for the successful implementation of the demilitarization process. The overly ambitious demilitarization agenda, reinforced adamantly by the international community, will be discussed later.

­ rocess is the lack of a robust Congolese state and security appar­atuses. p In addition to the state and security ‘problematique’, other issues inclu­ ding a historical trajectory of violence and authoritarianism, localized conflicts over land and resources, especially in the eastern regions, and the role of the regional actors have all contributed to the overall weakness of the Congolese state. This has in turn fuelled the inability of the state to address the challenges of demilitarization in an effective manner. The country’s historical trajectory of violence and authoritarianism has created a contextual backdrop which increased the likelihood of conflict, and made the task of building strong state institutions an arduous task. This chapter does not necessarily examine the conflict and the process of demilitarization through a historical lens, but fully acknowledges the importance of history in shaping the political and conflict dynamics in the country. The Congolese state has a long history of state collapse and conflict; and since independence, state capacity and the ability to project authority have remained limited. The colonial history was not only brutal but also left the country ill prepared for independence, while the Mobutu regime saw the reinforcement of patrimonialism at the expense of weak state institutions, including governance and military apparatuses. Indeed, one of the fundamental causes of DRC’s two wars was the erosion and collapse of state institutions. Mobutu’s government of thirty-two years had severely undermined the army, administration, parliament and courts and looted the state’s financial resources. The new government formed under President Laurent Kabila in 1997 seemed to develop many similarities to its predecessor. State institutions remained weak and corrupt; while in eastern DRC, rebel movements endured, relying on the extraction of natural resources and taxes and establishing parallel administrative structures. According to Prunier, the Congolese state is suffering ‘internal political paralysis’, as the country continues to face significant challenges; however, there appears to be ‘no long-range strategy’ to address them. This may be due to the fact that the current government is not unified, but rather is a ‘coagulation of groups opera­ ting out of completely mercenary interests […] or outdated ideological strands’ (Prunier 2009: 315). The local and regional dynamics that further encouraged the rise of armed groups must also be considered briefly. From the early 1990s onwards, local competition for access to economic resources, especially in the eastern DRC, was linked to a larger process of state collapse following the overthrow of the Mobutu regime and due to the regional dynamics of the conflict (the Rwandan genocide, refugee influxes and 120

121

6 | Militias in the eastern DRC

the role of Rwanda and Uganda). Although local competition between communities remained limited initially, the arrival of Rwandan Hutu refugees following the 1994 genocide led to an intermingling of local conflict dynamics with a regional struggle for power. This localized competition coupled with external interests (especially neighbouring Rwanda and Uganda) has led to the rise of various armed groups, inclu­ding Laurent Kabila’s Alliance des forces démocratiques pour la libération du Congo (ADLF), the Rassemblement congolais pour la démocratie (RCD), the Mouvement pour la libération du Congo (MLC) and most recently the CNDP (United Nations 2008: 15–18). The country’s en­ormous natural resource wealth and high level of poverty have made the eastern regions the site for local competition among political ­leaders, armed groups and economic opportunists. With the lack of a strong national government that is able to provide security and development for the population, society becomes fractured, relying on identity-based cleavages as a means to garner political and economic benefits – a reality that has been exploited by various armed groups. The Global and Inclusive Agreement, signed by the warring factions, the political opposition and civil society in South Africa in December 2002, changed the country’s power structure. Without undermining the DRC’s rich historical trajectory, it appears that the Congolese state was finally on a path towards state transformation. The focus of the trans­ ition and post-transition governments (both under President Joseph Kabila) was to encourage government power-sharing, accountability and social development – ideas that were reinforced in the 2006 constitution, which set up checks and balances to prevent abuses of power and corruption. However, parliament, auditing bodies and judicial courts remain weak, owing to a lack of human and financial resources to perform their functions. It is unfair to assume that the Congolese state would be able to provide all these services instantly; but in light of the demilitarization efforts, this state weakness provides a backdrop to understanding some of the institutional roadblocks hampering the effective implementation of DDR. Owing to the weakness of the state and military apparatus, dissident armed groups such as the FDLR, Mayi-Mayi militias, CNDP and others have risen and continue to destabilize the eastern regions – a factor examined in the previous section. As well, the national government was unable to secure firm deadlines for armed groups and militias to partake in the demilitarization process. Recognizing this, militias have since 2004 continued to exploit this governmental weakness by trying to fill the security vacuum in the eastern regions and ­providing

­ hysical and economic ‘security’ to the local populations, as well as p engaging in violent and privatized economic activities. However, it must be acknow­ledged that the government has had some success in disarming prominent military commanders from armed groups operating in Ituri, including Thomas Lubanga of the Union des patriotes congolais, and Mathieu Ngojolo of the Mouvement révolutionnaire congolais. Both leaders received amnesty under domestic law and were ‘integrated’ within the FARDC; however, they were handed over to the International Criminal Court to face charges of war crimes. Although the government must be commended for its efforts to promote transitional justice, such a Janus-faced position can lead to a level of mistrust which can potentially derail ongoing demilitarization efforts. For example, CNDP’s break away from the Amani Pact and resumption of fighting in mid-2008 is clearly indicative of this mistrust. Ambitious demilitarization (and security sector reform) agenda

The UN Integrated Disarmament, Demobilization and Reintegration Standards (IDDRS) framework stresses that DDR should not be pressed if the security situation is not ripe (United Nations 2006c: Section 3.10); however, there seems to be an organizational drive within MONUC to implement the DDR programme in the DRC. Many observers in the inter­ national community assert that the presence of so many small armed groups and militias in the eastern regions of the country should not be seen as posing a significant threat to security since these groups are not able to orchestrate a fully fledged rebellion against the government. However, given the history of violence in the country and region, coupled with the chronically weak Congolese state and security apparatuses, it should be recognized that these dissident armed groups do in fact pose a serious security challenge to the DDR process and to peace in general. According to a MONUC official, these groups have a significant ‘capacity for nuisance’,39 and with hostilities and rivalries between certain groups due to varying underlying motivations, and also the inherent cooperation among militias, especially for mutually beneficial personal enrichment, there is a high probability of continued violence in the future. For example, Prunier comments that the presence of the FDLR is ‘a kind of permanent irritant in the relations’ between various communities, including between the Banyamulenge and Kinyarwanda speakers, and between the so-called ‘originaires’ (natives) and ‘non-originaires’ groups (Prunier 2009: 322). The dominant role of the UN mission in the DRC remains a central concern, as it raises issues of dependency and lack of Congolese 122

123

6 | Militias in the eastern DRC

government ownership over the post-conflict reconstruction process. MONUC was established in November 1999, after the signing of the Lusaka Ceasefire Agreement in July 1999, and since then has played a significant role in helping the country with post-conflict reconstruction. Most recently, MONUC’s mandate was extended to 30 June 2010, and it was decided that from 1 July 2010 the mission would be referred to as Mission de l’Organisation des Nations Unies pour la stabilisation en République démocratique du Congo (MONUSCO). MONUC was the largest UN peace mission to date, with over 16,500 military personnel and over 2,500 civilian staff, with an annual budget of over $1.2 billion.40 In this regard, it is important to recognize MONUC’s significant ­efforts in stabilizing the country, including helping to organize the 2006 elections, helping to stabilize Ituri (and the Kivus), achieving ­effective DDRRR of foreign armed groups, and attempting to support the arms embargo. However, it should be noted that the hasty push for elections, especially by the international community spearheaded by MONUC, may have helped fuel ethnic tensions and marginalized e­ thnic minor­ities, thereby making the existence of militias more probable (see International Crisis Group 2007c). In addition, the demilitarization strategy is part of an overly ambitious agenda. MONUC continues to face overwhelming operational and logistical challenges, including an illplanned demilitarization programme as well as the problematic fusion of the demilitarization process with security sector reform, which has resulted in an overall weak Congolese army. This issue is further confounded by an unfortunate paradox: lack of sufficient financial, military and capacity-development support from the international community; however, there is a reinforcement of the non-negotiable nature of the international community’s stance on achieving post-conflict ‘success’ in the DRC. The overemphasis on disarmament and demobilization by both MONUC and CONADER resulted in the dwindling of the overall DDR budget, leaving limited funds for the implementation of reintegration programmes – or rather, reintegration was a ‘second thought’.41 As the numbers quoted earlier reveal, a significant number of combatants are still waiting to enter the demilitarization process, while an overwhelming number of former combatants who have been demobilized and disarmed are waiting to be reintegrated back into civilian life. Most former combatants are disillusioned by the reintegration process, as they had been promised a significant level of support in their communities, either in the form of training or tools/supplies for economic activity. The former combatants awaiting reintegration programmes

currently partake in whatever limited income-generating activities are available, such as subsistence farming, etc.; and in some cases do absolutely nothing.42 The ‘idleness’ of these former combatants is extremely dangerous, as they are susceptible to reverting to violence as a means of economic survival. The reintegration programmes in many communities in the eastern DRC have been ill conceived and poorly funded. CONADER, the national agency responsible for demilitarization, was crippled with high levels of funding mismanagement and weak institutional capacity to help coordinate programmes with MONUC, local NGOs, local government, communities and former combatants. Currently, only 40 per cent of former combatants have undergone training as part of the reintegration programme. As of 2007, owing to lack of funding (due to corruption within CONADER and the FARDC), most former combatants are forced to undergo military reintegration, and the option of reintegration back into civilian life was no longer available. In addition, there are a limited number of NGOs that have the institutional capacity to undertake such reintegration programmes.43 Reintegration strategies, specifically civilian reinsertion, have focused mainly on immediate support for basic living needs, as the provision of the monthly stipend for one year (only) illustrates. However, there has been no focus on community integration programmes, with the provision of training, education and tools/ supplies for economic activities. Also, the lack of funding for and financial mismanagement of CONADER activities clearly illustrate that a key feature of the DDR operation centres on an overly ambitious agenda, and technocratic ‘results-based’ indicators, with limited attention paid to the local context or to the institutional capacities of UN and government agencies that are meant to undertake this process. An illustration of this is the World Bank’s allocation of $1,200 per ‘former combatant’ for the demilitarization process. The lack of institutional capacity and the World Bank’s focus on ‘ownership’ rather than the ‘capacity-building’ of CONADER illustrate that the international community’s demilitarization efforts have been based on quantifying results rather than adopting a holistic approach to the demilitarization process. According to some Congolese observers, MONUC serves a ‘morally righteous agenda’, without having effective agents and means to implement the agenda.44 In the words of a Congolese government political adviser, ‘MONUC is a big monster without a head’, meaning that the international community is still unclear and ambiguous about its role in the DRC, and lacks the political will and capacity to address the numerous challenges the country faces.45 124

125

6 | Militias in the eastern DRC

In addition, MONUC and the international community are determined to ‘make the transition from conflict to post-conflict work […] as that goal is intrinsically non-negotiable’.46 This can be illustrated by the international community’s push for elections as early as 2004 (but they were brought forward to 2006 owing to logistical concerns). The international community also continues to push for military co-option (including with alleged ‘war criminals’ and warlords) within the DDR and SSR processes, which comes into direct conflict with issues of trans­ itional justice, as there is a significant level of impunity in exchange for partaking in the demilitarization process. However, as the DSRSG for MONUC articulated, ‘we [i.e. MONUC] are buying peace [via impunity], and fundamentally undertaking a state-building process’.47 He further goes on to say that in time issues of transitional and restorative justice will be addressed, as the first task in a post-conflict environment must be to ensure security, vis-à-vis the demilitarization and security sector reform processes.48 Despite an incomplete DDR process, the Congolese government and MONUC pushed ahead with SSR by providing training to the ‘­re­integ­rated’ army and police. Those former combatants who choose reintegration into the FARDC become the responsibility of the Structure Militaire d’Intégration (SMI), which, along with the FARDC (and MONUC for logistical support), transported the former combatants to the centres de brassage. There are a number of problems with the brassage process. The main function of the process was to break the chains of command of the different warring groups. However, there is the likelihood of continued mistrust between the former combatants of different factions according to their ‘ethnic’ origins or political affiliations. The centres de brassage lack adequate shelter, basic equipment, food, water, medicine and electricity; and many combatants are not paid and therefore resort to pillaging nearby populations (Onana and Taylor 2008: 510). As well, since brassage is simply the process of mixing groups together, the Congolese army is superficially unified. The creation of a professional army under the control of the state is meant to happen as part of SSR, and not necessarily DDR. Although the processes of DDR and army reform were linked in theory and design, as discussed earlier, the funding of both has remained segregated, thereby directly undermining the tronc commun process (ibid.: 511). Support for the DDR programme has been rather substantial as compared to that for army reform; which is most likely due to the increasing demands on donors to ensure electoral preparations. As such, since the Congolese military hierarchy lacked sufficient

funds to undertake army reform, the army remained unprofessional and poorly equipped. The connection between DDR and SSR is extremely important, and unfortunately, in a disaggregated environment, the linkages have been poorly translated into reality. Numerous integrated battalions of the FARDC have been trained by various donors, including the French, Belgium, the EU and the USA, as well as MONUC, each receiving varied levels of military training and exposed to a range of different strategic cultures and ethos. As such, SSR has been piecemeal and ineffective. This is further complicated by the halting of the DDR programme, and the presence of numerous ‘non-integrated’ battalions has resulted in a weak and uncoordinated national army. In addition, the army administration is plagued with high levels of corruption. The failure to pay FARDC salaries and provide rations has unfortunately led to increased criminality and violence against civilians. These clandestine activities are a means of supplementing the soldiers’ monthly salaries (less than $25 a month, which translates to less than $1 a day). Also, many of these soldiers continued to operate with virtual impunity. According to a MONUC official, the FARDC are the ‘most prominent crooks’ in the Kivus49 since they are provided with an army uniform and a weapon, which gives them unprecedented authority and ‘legitimacy’ to pursue economic activities, mainly for personal survival and benefit. In addition, owing to the co-option of numerous military leaders in the FARDC, the ‘standard’ military structure with a larger percentage of soldiers than military commanders (represented by a ‘triangle’) is basically reversed (represented by an ‘inverted triangle’).50 This has led to over 70–75 per cent of the national army being composed of (and overrun by) commanders, and barely 25–30 per cent of soldiers. For example, in 2007, the composition of the 14th brigade of the FARDC in North Kivu consisted of 950 generals (in a brigade of 3,484 soldiers).51 Therefore, the focus of SSR should not only be on strengthening the ­national army by providing professional training and modern weapons, but also on combating corruption and creating financial incentives (proper pay, housing, food) for soldiers not to engage in nefarious economic activities. As well, SSR includes establishing rule of law and building the legitimacy of the national army – issues which still need to be addressed. As the FARDC stands now, it is structurally incapable of containing the FDLR and CNDP, and countering the current security challenges in the eastern regions of the country.

126

This chapter has argued that a number of interconnected contextual realities have posed significant challenges to demilitarization in the eastern DRC. The most significant has been the weak state and security apparatuses, which has allowed the rise of dissident armed groups and militias, all of whom have used violence as a means to achieve their objectives and created instability in the country and region, thereby undermining the overall ‘post-conflict’ process. The prevalence of violent actors in any post-conflict environment has the ability to reverse the gains achieved during the ‘peace’ process. This factor is not only crucial in the practical sense, but also challenges the foundations of how to define peace and ‘post-conflict’. Are conceptualizations of peace that focus entirely on the absence of war entirely sufficient, or should there be a focus on ensuring freedom from fear (for civilian populations) and an end to impunity? Most observers would agree that an effective demilitarization strategy must move beyond the absence of war and focus on long-term durable peace. Linked closely to this factor has been the paradoxical position of the international community on the demilitarization process, and overall post-conflict strategy. The ambitious mandate to combine demilitarization with SSR has proved to be problematic, owing to a poorly funded and institutionally weak national coordinating agency. The lack of focus on capacity-building at the national, provincial and local levels in the DRC clearly reveals the inter­national community’s failure to understand local realities. The focus on robust technocratic objectives, such as disarming and demobil­ izing combatants without an effective (and properly funded) reintegration programme, undermines the overall demilitarization process. In addition, the political instrumentalization of ethnicity, natural resources exploitation, and the role of regional actors are all crucial factors in understanding the challenges facing the demilitarization process in the country. Although these factors are not necessarily unique to the DRC, it is important to realize that these local contextual factors must be examined in order to understand the effectiveness of the demilitarization process. After which, more holistic and dynamic demilitarization strategies can be devised, taking into account lessons learned from country-specific studies.

127

6 | Militias in the eastern DRC

Conclusion

7 | Afghanistan’s long and ongoing experience with militias

There is perhaps no other state that has had such an intensive, extended involvement with militias as Afghanistan. Indeed, without an understanding of the participation of militias in Afghanistan’s wars, it is impossible to truly appreciate Afghanistan’s recent tortured history. Militias were central to the Soviet war in the 1980s and the civil war of the 1990s, and have played a consistently prominent role since the fall of the Taliban government in November 2001. This chapter will provide a brief history of Afghan militias during the Soviet and civil wars followed by a description and contextual analysis of efforts since 2001 by the international community and the Afghan government at demilitar­ izing militias. This demilitarization centred on fairly conventional DDR programming implemented by the UN but suffered from the ongoing weakness of the Afghan state and the entrenchment of a ‘warlord class’ within the ruling political elite. In conclusion, the chapter will examine efforts in the past several years by the US military and the Afghan government to use local community militias in its counter-insurgency campaign against the Taliban, what could be deemed a ‘remilitarization’ strategy. Afghan militias in the 1980s and 1990s

Militias were prominent in the evolution of the Soviet Union’s extended intervention in Afghanistan, which began in December 1979 when Soviet forces removed President Hafizullah Amin from power and installed Babrak Karmal. The Soviet war can be characterized as an extended insurgency composed of seven major mujahedin groups which fought to topple Moscow-backed communist regimes in Kabul. In their response to the mujahedin insurgent groups, the usage of militias as proxy forces by the Kabul governments, with backing from Moscow, increased during the course of the war. There were an estimated 10,000 militiamen in 1981, 20,000 in 1984 and up to 70,000 by 1990 (Giustozzi 2000: 205). Initially, Kabul and its Soviet allies most often used localized militias that were focused on preventing insurgents from accessing and influencing their home areas. The best example of these local militias 128

129

7 | Afghanistan’s experience with militias

were ‘Self-Defence Groups’, which were ‘non-ideological, part-time units strictly limited to the protection of work places and villages’ (ibid.: 200). Other similar efforts advanced by Kabul were Milishia-i Sahard, or Border Militia, and the Ghund-i Qawmi, or Tribal Regiments. However, while much of the militia effort initially focused on protecting local areas from mujahedin influence, as the war progressed the emphasis evolved into directly buying the loyalty of mujahedin groups as part of the broader Soviet counter-insurgency strategy. Often muja­ hedin were persuaded to defect and persuaded to join the localized militias. By 1989 there were an estimated 50,000–60,000 ex-mujahedin who were serving in militias allied with the government (ibid.: 205). Buying the support of mujahedin groups was not cheap for the government, however, and, as Giustozzi notes, ‘a large part of the money printed after 1986 went to militias’ (ibid.: 209). This dynamic of buying off ex-mujahedin and subsuming them into the local militias was a risky strategy since there was also a decidedly opportunistic propensity of local commanders to switch loyalties, often at the inducement of money. What relationships lasted were often more dependent on the exigencies of patronage than any deeper loyalty (O’Ballance 1993: 124). A further dynamic as the war evolved was for Kabul to utilize mil­ itias for more than just maintaining a localized armed presence; rather they became directly involved in the ‘active fighting’ elsewhere in the country from their home locations (Giustozzi 2000: 214). By the time the ­Soviets left, several militias had become especially prominent, fighting as large formations outside of their home areas. The most notable one was Abdul Rashid Dostum’s Jowzyani militia, which for a time was fighting in Kandahar province, on the other side of the country from its home base, and had reached a size of 20,000 men. This was possible given how the militias had morphed from being smaller armed formations into large, essentially division-strength structured units with heavy weaponry. The Najibullah regime became especially dependent on militias as the Soviets planned for and ultimately withdrew, and given the extreme dysfunction and dwindling of the national army. By 1990 the number of militiamen that Najibullah relied upon was between 60,000 and 70,000 men, twice the size of the army (Rubin 2002: 161). Like much of the communist regime’s leadership, Najibullah was a Pashtun, the majority ethnic group of Afghanistan. As the Soviets were withdrawing and doubts increased about the survival of Najibullah’s communist regime, tensions escalated among the ruling Pashtun elite. Subsequently, Najibullah increasingly placed an emphasis on garnering support for himself from non-Pashtun militias, especially in the north

and north-east (ibid.: 109). This was ultimately counterproductive as the militias, notably Dostum’s, became unwilling to follow the orders of Najibullah, both because he rather belatedly tried to re-establish Pashtun dominance of the military (Ahady 1995: 625) and because he simply lost the ability to pay them adequately (Giustozzi 2000: 238). While the Soviets had attempted to buffer the Najibullah regime following their withdrawal through massive aid provision, this eventually dwindled following the political tensions within the Soviet Union itself. Without access to Soviet aid, Najibullah could not maintain his militia supporters and they withdrew their support (Rubin 2002: 265). The Najibullah regime quickly collapsed as both mujahedin parties and former militia allies overwhelmed it in April 1992. With the overthrow of Najibullah came the conclusion of the communist era in Afghanistan. What followed was an extended civil war, lasting throughout the 1990s, as the plethora of large armed groups, including both the seven major insurgents groups and the large militias that had supported Najibullah, sought to assume and maintain control of the government. The ex-communist militias were especially powerful as they had come to control large stockpiles of Soviet armaments as Najibullah’s regime collapsed; this was especially true for Dostum and left him in a strong position as the civil war started. With Cold War imperatives removed, American and Russian interests in Afghan­ istan largely abated. However, given the void in power caused by the superpowers’ withdrawal, regional powers became increasingly involved in supporting assorted armed factions, such as Pakistan’s support of Hizbi Islami, Iran’s support for Hizbi Wahdat, and Uzbekistan’s aid to Dostum’s large militia. The extent of the power held by the commanders of these competing armed groups, evolving constantly in alliances of convenience, was immense. Giustozzi has argued that the main cause of ‘the crisis of the Afghan state’ in the early 1990s was in fact the power accumulated by these commanders, which resembled ‘that of a feudal lord’ (Giustozzi 2000: 243). However, while the 1990s were indeed characterized by a plethora of immensely strong warlords competing against one another through large militias, there were also significant numbers of small, localized militias. Even by 1989, a Soviet estimate suggested that there were probably 3,000 village militias that were not operating on behalf of any larger actors but were merely seeking to protect their home areas (ibid.: 218). The desire for localized self-defence in the wider context of violent anarchy saw the emergence of the most notable militia of the 1990s, the Taliban. The Taliban originated from a local militia of dis130

Demilitarization efforts since 2001

The US overthrow of the Taliban regime during October and November 2001 was achieved through small numbers of US Special Forces partnering with the so-called Northern Alliance, an amalgamation of primarily non-Pashtun armed groups, and anti-Taliban leaders from the Pashtun community, such as Hamid Karzai. Avoiding the need for large numbers of US ground forces did, however, leave the USA dependent on local armed proxies, notably the large formations of the Northern Alliance. In the immediate aftermath of the Taliban’s overthrow, these forces were used extensively by the USA in eastern Afghanistan in operations searching for remnants of Taliban and al-Qaeda militants and providing security for US bases. In 2002 the Northern Alliance’s component forces were formed into the Afghan Militia Forces (AMF), a quasi-military structure of the Ministry of Defence but still largely under the control of the respective warlords, such as Abdul Rashid Dostum, Mohammad Atta and Qasim Fahim. AMF commanders were most often given formal positions in the new Ministry of Defence and monthly stipends were provided to AMF soldiers. This process was widely criticized, both domestically and internationally, as it was seen as legitimizing the warlords and militias who were central to the violence of the 1990s. Furthermore, it was argued that the presence of the AMF ultimately served to undermine the new state’s institutions by entrenching warlords and militias into them at their very founding. A particular individual prone to such critiques was Dostum, the wily commander of the largest militia to serve Najibullah. He had eventually joined forces with the Northern Alliance, which was led by Ahmad Massoud, although his militia force had over 40,000 men, three times as many as Masoud’s Jamiat-e Islami had (Rubin 2002: 270). With the AMF militias the strongest Afghan armed forces in the country, 131

7 | Afghanistan’s experience with militias

gruntled ex-mujahedin, the ‘Mullahs’ Front’, which had been formed in the late 1980s (Lyon 2009: 180). The originating motive for the militia was resentment at the general mayhem that existed after the Soviets left, most of it blamed on the proliferation of large militias and the dom­ inance of warlords. With a penchant for hanging opponents from tank ­barrels, the Taliban established themselves in the autumn of 1994 when they achieved a series of victories against local tribal militias and local warlords who had been brutalizing south-eastern Afghanistan (Edwards 1998: 723). By 1996 the Taliban had grown immensely in size and had taken control of Kabul, establishing a government that lasted until it was overthrown by the USA in late 2001.

priority was given to implementing a disarmament, demobilization, reintegration (DDR) programme to incorporate them fully into the Afghan state’s official security organs or to return them fully to civilian life. The DDR effort began in October 2003 and was under the management of the UNDP through a programme called the Afghan New Beginnings Programme (ANBP).1 The ANBP officially targeted 100,000 former combatants with the main emphasis on the AMF militiamen. A routine DDR process was utilized: participants turned in weapons and received some food and clothing, followed by opportunities for vocational training or the option of joining the Afghan National Army (ANA) or the Afghan National Police (ANP). As mentioned, many of the mid-level AMF commanders were promoted into government ministries while many of the most prominent AMF commanders had already been given senior positions in the government, such as ministries and governorships. This was considered compensation of a sort and has proved to be controversial over the longer term as it was seen to embed men with dubious backgrounds in government (Gossman 2009: 2). By 2006, 62,326 former combatants had been through the DDR process and there were also 24,536 female relatives of combatants who received education and income opportunities. The total cost of the programme was approximately $100 million, mostly financed by Japan. The initial DDR effort, through the ANBP, focused very heavily on the large militias of the AMF. However, there were still copious numbers of smaller militias, what were termed ‘illegal armed groups’, and there was concern that these smaller militias were involved in drug trafficking and crime along the major transport routes. Often these local militias protected the property and interests of local power brokers, such as former commanders, politicians and businessmen. Hence, a secondary DDR process was started called the Disbandment of Illegal Armed Groups (DIAG), which aimed to collect the weaponry of the smaller, localized militias left over from the ANBP process. DIAG started in the run-up to the parliamentary elections in 2005 and lasted until 2007, during which time 1,600 local illegal armed groups were targeted. Through DIAG 5,700 weapons were collected and through the concomitant efforts of ISAF, US and ANSF forces, an additional 14,000 weapons were gathered from local militias (ibid.: 20). The years immediately following the downfall of the Taliban, up until approximately mid-2006, were relatively peaceful as the insurgency was confined to small pockets in the south and east. During this period, rather than a driving emphasis on counter-insurgency, international efforts focused on humanitarian aid provision and the capacity-building of the 132

133

7 | Afghanistan’s experience with militias

nascent government of Hamid Karzai. Significantly, with the Taliban insurgency limited in scope, the international community, through the UN and its ANBP and DIAG programmes, could proceed with demilitarization efforts relatively unhindered. Overall, the DDR processes of the mid-2000s can be considered to have achieved some very positive results. Considering the plethora of large armed groups present after the downfall of the Taliban, notably the component forces of the Northern Alliance, those that were later to constitute the AMF, it was an achievement of the ANBP to have disbanded these large formations by 2005, and to have made a comprehensive collection of heavy weaponry. Furthermore, the DIAG programme did make significant strides towards disbanding many of the smaller, localized militias that dominated much of rural Afghanistan, as well as collecting significant quantities of small arms. The overarching challenge of ‘state-building’ in Afghanistan has been how to demilitarize the warlords of the 1990s, including those who helped topple the Taliban, while also endeavouring to create a new government post-2001. Use of the Northern Alliance as the primary means to overthrow the Taliban, and a willingness to engage its component warlords through the Bonn process, ultimately entrenched them at the heart of the current government, especially in the security and intelligence apparatuses. Hence, the strongest critiques of the international community’s demilitarization efforts were that they placed ‘spoilers’, in the form of former AMF commanders, in the Afghan government, arguably contributing to longer-term, ongoing insecurity, corruption and lawlessness in Afghanistan.2 While large AMF militias may have dis­appeared, the former commanders were still able to dominate the political life of Afghanistan. This was notably true of the ANA and ANP, which were heavily dominated by the Northern Alliance veterans because their leaders, such as Fahim and Dostum, came to dominate the Ministry of Defence and the Ministry of Interior from their very inception. By way of further example, the DIAG process had specifically been intended to vet candidates, starting with the parliamentary elections of 2005, in order to force commanders to disband their local militias and hence help cleanse the political system of warlordism. Yet only a handful of candidates were actually disqualified, which meant that many parliamentary candidates still assumed office and did not actually have to disband their local militias. Overall, the extensive involvement of militia commanders in the DDR processes left them with a mere ‘façade of respectability’, which allowed the international community to present its technical successes at DDR without resolving core issues (Giustozzi 2009: 85). Following

the manic violence of the 1990s, which had been largely caused by the inability of most of the same insurgent and militia leaders to come to an understanding following the collapse of the Najibullah regime, an international hope, and assuredly one of normal Afghans, was that the dominance of these warlords over Afghan’s future would be finally ended. However, that the same warlords who were central to the extreme misery of the civil war in the 1990s have been able to manipulate demilitarization processes to benefit their own power consolidation through the formal offices of governance has been a crippling challenge to state-building efforts in Afghanistan and the broader peace process. Subsequently, with many former warlords and their followers legitimized through government posts, ‘strong correlations between past abuses and current abuse of power’ have remained a persistent dynamic of the Afghan state (Gossman 2009: 32). Indeed, that a ‘warlord class’ arose, one that was largely independent of both governmental and traditional checks and balances, is the most often cited critique of demilitarization efforts in Afghanistan in the immediate aftermath of the Taliban (Bhatia and Sedra 2008: 16). The strategies of the USA must be given special emphasis in these developments, considering its centrality to the toppling of the Taliban and subsequent state-building efforts. Balancing state-building and counter-terrorism imperatives has been the biggest challenge strategic­ ally for the USA, especially in the early years of the intervention. While the USA may have supported the official demilitarization processes, the ANBP and DIAG, it also allowed former warlords to assume dominance of much of the state apparatus even while their large militias in the form of the AMF were gradually disbanded. Furthermore, even while the demilitarization processes were ongoing, the US military relied on arming and financing local leaders to undertake much of the search for leftover al-Qaeda terrorists, with the result that ‘a new generation of US-backed “warlords” was added to traditional ones’ (Saikal 2006: 531). A continuing willingness of the USA both to endorse national demilitarization efforts and to use militias as local proxies was one that had mixed results: it met the immediate security goals of the USA but challenged longer-term security within Afghanistan itself. Ultimately, as Ahmed Rashid starkly noted, while the USA may have ended the Taliban regime and expelled al-Qaeda from Afghanistan, ‘they had brought back another evil: warlords’ (Rashid 2008: 131). This occurred both at the local level through the presence and use of militias and, arguably more damning, at the national level through the institutionalization of warlords in the government. 134

135

7 | Afghanistan’s experience with militias

A willingness to allow Afghanistan’s strongmen, i.e. its ‘warlord class’, to continue to play a dominant role was largely because the USA, as well as the international community more broadly through the UN, wished to keep the foreign presence small. The international community’s i­ nitial strategy, mostly through the offices of the UN, focused on a ‘light footprint approach’ that was meant to allow the Afghans to drive the process of political reconciliation and post-conflict reconstruction and to assume greater ownership of that (Gossman 2009: 10). Concomitantly, the early presence of Western troops was relatively limited. The USA had a small military presence of approximately 10,000 troops while the International Security Assistance Force (ISAF), composed of some US forces but mostly of NATO and assorted other US allies, was also generally less than 10,000 strong. Without significant numbers of troops available to establish a wide, strong presence to hunt down the remnants of the Taliban and al-Qaeda, other possibilities, such as continuing to use militia proxies, were chosen. Furthermore, without a strong international presence, accommodation with and engagement of the Northern Alliance’s warlords were more necessary than if a stronger initial presence had been implemented. This was particularly true of forming the new government. Persuading strong warlords to dismantle their large militias required ‘compensating’ them somehow, the ­chosen method being to allow them to assume key positions in the new government. In sum it can be deemed that initial efforts were defined by a considerable amount of pragmatism. While some short- to near-term success should be claimed for the demilitarization of the large AMF formations, ­support of the warlords has still been a ‘double-edged sword’ because of their ongoing dominance of the Afghan government ( Jones 2006: 115). In sum, over the past thirty years there have been numerous attempts to create a viable state in Afghanistan. The Soviet-backed regimes of Amin, Karmal and Najibullah between 1979 and 1992 fell with the demise of the Soviet Union. This was followed by the mujahedin’s rule under Burhanuddin Rabbani between 1992 and 1996, which in the end had little international support. Finally, the Taliban regime grabbed the state and imposed their own brand of political and religious governance between 1996 and 2001; only to be toppled by the United States and its allies as part of the ‘global war on terror’. Despite the various attempts to establish a viable state in Afghanistan, the overarching context has remained the same: a weak state apparatus is inexorably linked to war; the inability of leaders, including the current Karazi government, to ‘build stable coalitions of disparate actors around the universalizing

state-building project’ (Angstrom 2008); and the role of international or regional actors, especially in striking deals with political elites and local warlords. As can be seen in both the Soviet and American attempts, ‘due to the peculiarities of warfare in Afghanistan – large armies are not necessarily translated into political power – it seems that war regenerates the tribal system of local patronage’ (ibid.: 390). The remilitarization of Afghanistan?

By mid-2005 and early 2006 the Taliban, at this time often termed ‘neo-Taliban’, were able to greatly intensify their insurgency against the Karzai government and his Western backers.3 As the insurgency strengthened much of the international presence, specifically that of the USA and ISAF, shifted in focus heavily towards counter-insurgency e­ fforts. During the course of 2007 and 2008 the insurgency expanded even more rapidly, and concomitantly the presence of Western soldiers began to increase as a shift was made away from the ‘light footprint ­approach’ towards a broad military campaign active across the entire country. Despite an expanded Western military effort the Taliban insurgency was able to grow stronger. In addition to the expansion and boldness of the Taliban’s attacks, the two major pillars of the US and ISAF co­alition’s counter-insurgency strategy, a strengthened Afghan government more competent to administer the country and Afghan security forces able to bear the primary burden of the counter-insurgency effort, were simply not materializing. Even with years of capacity-building efforts by the USA and ISAF, the ANA and ANP remained incapable of assuming a significant portion of the security portfolio, lacking competence and the numbers necessary to do so. Furthermore, the Afghan government for its part has consistently been plagued by crippling corruption, an inability to provide basic social services, and an ever widening legitimacy gap with the public, highlighted most strikingly by the dubious presidential election of August 2009.

The attractiveness of militias  As the security situation continually worsened (and by 2007/08 even the immediate provinces around Kabul were consistently violent), there was broad consideration by the USA of supporting local militias even though the DDR processes had gone so far towards demilitarization in Afghanistan. The momentum for this interest in militias was especially pronounced given the use of militias in Iraq, the so-called ‘Sons of Iraq’ programme, by the US military, which was widely considered a definitive success in calming the horrendous security situation there in 2007. Echoing many of the arguments found 136

Do they [Afghans] believe they can protect their own community in the way that Afghanistan has done for all of these centuries? What we’re trying to do is not arm them or disarm them in this regard, but strengthen the community in such a way that it is more self-reliant, and it can resist the infiltration, and the intimidation, and the night letters, and the beheadings […]4

A special interest of US defence officials were locales where a resistance had endogenously sprung up against the Taliban. Defense Secretary Robert M. Gates furthered this argument when he noted, ‘At the end of the day the only solution in Afghanistan is to work with the tribes and the provincial leaders in terms of trying to create a backlash […] against the Taliban’ (Bruno 2008). However, there was still much resistance to the idea from within the US military and diplomatic establishment, in addition to scepticism from the Karzai government. The NATO commander in Afghanistan in early 2008, US General Dan McNeil, responded that there was a need for extreme caution: ‘What we should not do is take actions that will reintroduce militias of the former power brokers’ (ibid.). In other words, while many warlords may have assumed assorted posts in the Afghan government, the DDR processes had been successful at disbanding 137

7 | Afghanistan’s experience with militias

in Iraq, a general theme of the consideration of militia efforts by US officials in Afghanistan was to emphasize the need to provide better security for the rural communities that bore the brunt of the intensifying insurgency. An assumption of the formal demilitarization efforts was that the Afghan state would fill the security void left through disarm­ ament, but this widely failed to happen, especially in rural communities, where the ANA and ANP were never able to establish a strong presence. With this in mind, the argument began to be formulated that there was a need to augment Afghan forces from within local communities themselves. Prime Minister Gordon Brown of the UK was actually the first to talk publicly about the idea in December 2007 when he argued that the British military should ‘increase our support for community defence initiatives, where local volunteers are recruited to defend homes and families modelled on traditional Afghan “arbakai”’ (UK Prime Minister’s Office 2007). Arbakai, prominent in Pashtun communities, can be understood as ‘a tribal based community policing system grounded in volunteer grassroots initiatives’ (Tariq 2008: 3). Echoing Brown’s arbakai theme, in early 2008 US ambassador William Wood commented on ideas circulating about creating ‘civil defence forces’:

the large AMF militias and there was no desire for such types of large militias to be re-created. Additionally, there were attempts to dampen enthusiasm for simplistic hopes of Iraqi successes being quickly em­ ulated in Afghanistan. As the US commander General David McKiernan highlighted in late 2008, ‘What I find in Afghanistan […] is a degree of complexity in the tribal system which is much greater than what I found in Iraq years ago’ (ibid.). Simply put, while Iraq had hierarchical tribal structures with a relatively strong degree of coherence, the tribal structures of Afghanistan, specifically of the Pashtuns, were more dispersed, lacking the type of cohesion that would allow for similar engagement approaches. The most consistent critique of suggestions to utilize local militias was that it was contrary to the whole state-building project. Strengthening the ANA and ANP had long been the central pillar of the US ‘exit strategy’, and arguments were made that a reversion to utilizing militias, of whatever form and however limited in scope, would be a step backwards.

US militia efforts  These were all very profound concerns about and critiques of suggestions that the USA and its allied militaries should engage local militias. However, given the pressures of a worsening security situation, the exigencies of utilizing militias were too strong for the US military to completely resist. One of the strongest arguments was that with the ANA and ANP lacking in abilities and numbers, and even with the addition of 30,000 US troops in late 2009, there was still a need to engage local militias to prevent the security situation from getting even worse. President Obama’s revised Afghanistan strategy of autumn 2009 did not increase the proposed numbers of ANA and ANP significantly, at approximately 240,000 soldiers and police in total. Many commentators, such as Max Boot, have long argued that the Afghan security forces, even assuming they are competent, are ‘not in the proper ratio with the population which needs to be protected’.5 Ultimately, as Afghanistan has demonstrated time and again, when the central government is weak, militias will be utilized by it in one form or another, and an occupying power will most likely collude to support that engagement. By 2008, with the Afghan security forces still lacking in capacity and the Taliban insurgency rapidly increasing in intensity and scope, the US military began to undertake several systematic efforts at utilizing local community militias – most often referred to generically as ‘community defence forces’ – to deny the insurgency the ability to control rural areas. The focus of these efforts was very much on what Bhatia termed ‘community militias’, which in his typology are small, localized militias 138

139

7 | Afghanistan’s experience with militias

that focus on immediate self-defence and contrast strongly with the other two types found in Afghanistan: 1) political-military organizations organized around tanzims, the mujahedin insurgent groups centred on political parties such as Hezb-i-Islami, and 2) warlords and strongmen, such as Ismael Khan and Abdul Rashid Dostum (Bhatia and Sedra 2008: 73). The USA was interested in utilizing localized community militias because they were felt to be a good counter to the Taliban insurgency as it was taking form in rural areas. As Stephen Biddle highlighted, while the Taliban is composed of some full-time foreign fighters, many of its local forces are actually ‘indigenous Afghan seasonal militias (amateurs who fought when the agricultural calendar permitted and whose tactical proficiency was very limited)’ (Biddle 2005/06: 167). In this sense, the community militias were a very direct counter to the Taliban’s own usage of local fighters. A strong emphasis on local self-defence was a pragmatic attempt to justify the new efforts within the broader context of a state that does not have a monopoly on violence and hence is unable to ensure security for its people, especially in rural areas. In defence of the idea of using community militias, an argument was made that such usage was merely recognition that Afghanistan has always had a weak central government and hence local communities have always needed to be involved to some extent in their own protection. In this regard, it was contended that they should not be viewed simplistically as totally undermining the central government, or, as Seth Jones argued, ‘That shouldn’t be a deal breaker; that’s just the way Afghanistan has worked historically’ (Bruno 2008). Nevertheless, the official long-term goal is still to have Afghan security forces with a sovereign monopoly on violence. In response to the critique, however, a key need to establish strong enough linkages between whatever militias might be formed and their host communities with the Afghan government was highlighted. Towards this end, utilizing community councils and meetings was emphasized, and programmes such as the Afghan Social Outreach Program (ASOP) were set up to help with the formation of district-level shuras (community councils) that could then help with the management of local militia forces. Any new efforts were also supposed to make the linkages stronger through the use of vetting procedures by local leaders and local shuras. As Secretary of Defense Robert Gates articulated to local Afghans, ‘We’re going to help you, but you have to agree among yourselves and empower certain leaders to work with you, in coordination with the national government so you’re not creating some parallel structure.’6 The first large, official programme instigated by the US military was

the Afghan Public Protection Programme, often referred to as AP3, which began officially in early 2009. Overall, the intent of the programme was less to form a ‘tribal militia’ and more of an effort to build a ‘community watch’ system.7 In this regard the AP3 was meant to be an ‘auxiliary police’ effort whereby locals would be allowed to support local ANP. As the Afghan defence minister, General Abdul Rahim Wardak, announced, ‘[…] they are not going to be like militias’, as the participants would wear uniforms, have clear chains of command, and most importantly receive government oversight from the Ministry of Interior and supervision from local ANP (Council on Foreign Relations 2009). Participants received a salary starting at $100 and going to not more than $250 (i.e. no more than ANP salaries), and were also provided with a weapon, uniform and ammunition. The main purpose of AP3 members was defined as securing their local area by bolstering the ANP through guarding sections of road and government buildings. The pilot programme for the AP3 started in early 2009 in Wardak province when approximately 243 men were recruited in pilot districts to work in support of the ANP. Pilot participants attended a three-week training course provided by US Special Forces. Furthermore, participants were vetted by local ANP as well as local elders and religious leaders. The AP3 effort succeeded the Afghan National Auxiliary Police (ANAP) programme, which was also meant to provide support to the ANP through the use of local auxiliary forces. The focus of the ANAP was southern Afghanistan, and members were recruited locally, provided with basic training, and armed. The ANAP had started in 2006 but closed in 2008 because its membership was riddled with drug addicts and petty criminals and ended up worsening local tensions. Critics of the programme argued it was a failure because it did not have strong linkages to local tribal structures in Pashtun areas. For its part, the AP3, piloted in Wardak province, immediately south of Kabul, struggled to recruit locals from the Pashtun majority, mostly succeeding only in recruiting from the Hazara minority.8 This imbalance created local tensions and was one of the major reasons why the pilot effort was delayed in being rolled out to other provinces. The second large programme, started in mid-2009, was termed the Local Defence Initiative (LDI), although it was initially referred to as the Community Defence Initiative. As the name suggests, and in parallel with the AP3 programme’s intent, it was ‘designed to assist the local population to provide their own security with defensive “neighbourhood watch”-type programs’, as stated in a NATO press release.9 The LDI was formulated to support communities with localized militias 140

141

7 | Afghanistan’s experience with militias

which had asked for assistance in their own resistance efforts against Taliban insurgents. As a US military official explained, ‘The idea is to get people to take responsibility for their own security. In many places they are already doing that.’10 In this regard, unlike the AP3, it was not meant to provide a quasi-official corps of police auxiliaries, paying and organizing them, but rather was meant to encourage a community’s own resistance by providing ‘collective benefits to the community through aid and development’.11 The US military’s initiating LDI efforts focused on Special Forces teams channelling assistance to those locales where local movements led by tribal elders were acting to resist Taliban influences, such as in Paktia and Nangahar provinces. The specific roles of LDI militias were limited to protecting villages and infrastructure, such as roads and bridges, in their home areas. The relationship of the Afghan government to the LDI was not as clear as that for AP3. While some senior Afghan government officials have commented in support of its rationale, the official involvement is ambiguous. For instance, the Afghan interior minister, Hanif Atmar, explained the local dynamics underpinning the LDI idea: ‘What we are talking about is a local, spontaneous and indigenous response to the Taliban. The Afghans are saying, “We are willing and determined and capable to defend our country; just give us the resources.”’12 Whatever potential role the Afghan government might play, the LDI was an idea that was given momentum by General Stanley McChrystal when he assumed control of US and NATO forces in mid-2009. According to one of the general’s advisers, ‘McChrystal was always quite dismissive’ about AP3, because he considered it too slow and requiring too many resources.13 The LDI was meant to allow the US military to more directly engage local communities without needing to navigate the extended bureaucracies and general inefficiencies of the Afghan government. By example, in mid-2009, in Ahcin district, in Nangargar province, US Special Forces started providing assistance to the Shinwari tribe to resist the Taliban after a fight had developed between them and the insurgents.14 In January 2010 the USA extended even more support to the Shinwari tribe, providing US$1 million in aid for local development projects, and in exchange 170 elders signed a pact to keep the Taliban out of their area. The US aid money, as a Shinwari elder explained, is being provided directly to locals because among the local government ‘probably 95 percent of them are corrupt’.15 With the replacement of General McChrystal by General David Pet­ raeus in early July 2010 even greater impetus was given to util­izing localized militias as a counter-insurgency tool, i.e. as ‘community defence

forces’. Petraeus had been the commander in Iraq during the ‘surge’ of US troops there in 2007, which had involved the heavy partnering of American troops with Sunni militias, widely considered a success within the US military. Petraeus sought to emulate his past efforts in Afghanistan with the evolution of the AP3 programme, as well as smaller individual efforts by US Special Forces, into the ­Local Police Force programme.16 This effort is meant to expand previous militia efforts systematically across the country but keep the emphasis on allowing local communities to play a limited self-defence role. Unlike the AP3 ­programme, however, it was meant to be composed of smaller militia units operating only in their own communities, and was intended to have a stronger connection to the national government, specifically the interior ministry. These changes were intended to alleviate two primary concerns of Afghans. The first was felt by the Kabul government, and especially President Karzai; namely that it was losing control of armed actors in the country, and hence the new effort was very much a ‘­policing’ one under government oversight. From the perspective of Afghan communities, it was hoped that local residents would embrace and support the militias more strongly if they were formed only from their own communities and kept small and strictly localized. Local Police Force militia members are armed and allowed to patrol their home communities, achieving what the US military hopes will be a ‘community watch on steroids’.17

Afghan government militia efforts  While the US military, through the LDI and AP3 programmes, and now the Local Police Force, increased its own engagement of local militias following the escalation of the Taliban insurgency, the Afghan government had itself been pursuing its own militia strategies for years. At its simplest, in the context of a weak central government without even rudimentary control over the security situation, the Afghan state has a history of relying on local militias to make up for its own lack of monopoly on force. Or, as M ­ ichael ­Bhatia has argued, the modern Afghan state since 2001, starting with the AMF, has continued a ‘long tradition of state endorsement of militias’ (Bhatia and Sedra 2008: 28). The Afghan government, which is a very centralized system with provincial governments merely being extensions of Kabul, has at times allowed militias to be kept by local government officials. For example, in 2004 there were an estimated sixty-one ‘­arbakai’ groups, small localized Pashtun militias, in Paktya province, which were encouraged and supported by the provincial government, mostly to help provide security during the electoral processes (ibid.: 142

143

7 | Afghanistan’s experience with militias

218). These groups had between 30 and 400 members and were limited to their home areas. Another example of a militia being sanctioned by the government was the 800-strong Kandak Auxiliary Uruzgan, led by a former ANP commander, which was allowed to provide security along the transport routes linking Uruzgan to northern Kandahar province. It was under the control of the Uruzgan governor, Jan Mohammad Khan. The Afghan government has also at times encouraged the development of local militias that have not had any direct, official linkages to the government but have been in strategic locales. By further example, Karzai in June 2006 expressed support for arbakai militias directly along the border with Pakistan in south-east Afghanistan. As described previously, in mid-2006 the Karzai government undertook to use several ‘community militias’ in Helmand, paying militiamen $200 per month and providing them with basic training (ibid.: 17). This 2006 government initiative, the ANAP, was meant to be an auxiliary police programme. A further dynamic of the Afghan government’s involvement with militias is that many of the AMF militias, or components thereof, have morphed over the years into private security companies. The companies were officially sanctioned by the Afghan government and focused mostly on the provision of security guards for everything from road construction, guest houses for international organizations, to trucking companies, and even NATO bases. In the early years of the intervention, US Special Forces often hired local militias to guard their remote bases, a trend that has continued. For example, the German army uses a private security company to guard its base at Feyzabad, while a French army base in Tagab, Kapisa, does the same. COMPASS Security is the largest security company and was approved by the government up to a level of 20,000 employees, mostly for providing security to foreign installations and for road convoys. Many of its guards are Tajiks from Panshjir and were former members of the Northern Alliance and subsequently the AMF. As is common for militias, the critique of Kabul’s sanctioning of militias, in whatever form, was based on the fear that it was a reversal of the DDR processes and ultimately served only to both revive local disputes and warlordism and undermine the central government. Nonetheless, the Afghan government increased its interest notably over the course of late 2008 and 2009. In early 2009 the Independent Directorate for the Protection of Public Properties and Highways by Tribal Support (IDPPPH) was established by Karzai. The IDPPPH was led by Aref Khan Nurzai, a close confidant of the president, and its stated purpose was to recruit and manage local militias, through the government security forces but with the help of local community shuras, along the national

ring-route highway and for other needs of the central government. Nurzai argued publicly that his efforts were meant to emulate the successes of local militias in Iraq: ‘We are trying to recreate the Awakening of Iraq here in Afghanistan.’18 For instance, Nurzai enlisted 12,500 militiamen in twenty-two provinces for the period of the August 2009 elections to help provide security around polling stations.19 As an ANP general noted of the usage of local militias in Paktika province at the time, ‘I don’t have enough police to protect the people. It’s a gamble but I hope the arbakai can help.’20 At the same time as the IDPPPH was formed, there was other government support to militias as the insurgency increased in scale. Most notably, as the Taliban became active in northern Afghanistan in the summer of 2009, local militias, especially in Kunduz province but also in Badghis, Faryab and Takhar, rose to resist them. Often they coalesced around former AMF commanders, at times with the encouragement of local police commanders. One of the motivations for these local militias, which have mostly formed in Tajik and Uzbek communities, has been to resist the growing strength of Taliban cells operating out of local minority Pashtun ‘settlement’ areas. There were reports that there were up to seven hundred of these militias by early 2010, encouraged by the district governments and supported by local ANP units.

Critiques of ongoing militia efforts  The US military’s recent attempts to utilize local militias have been highly criticized. As public awareness of the LDI and AP3 programmes and now the Local Police Force has increased, several major critiques have been raised from assorted quarters. The most damning one has been that Afghanistan and the international community have already invested a huge amount of resources and political capital in the demilitarization of militias, precisely because they are sources of instability, and subsequently that it is staggeringly counterproductive to now be forming new ones, i.e. to be undertaking remilitarization. As one Afghan human rights activist argued of the LDI programme, its effort is a ‘source of insecurity itself’ and most likely will only ‘produce a new round of warlords who would fight for resources and positions for a long time’.21 There have been critiques suggesting that rearming some communities and not others in the context of a still largely anarchic countryside is most likely to provoke fighting between communities. At its simplest, the Afghan ‘security dilemma’ is one whereby ‘the forces used to protect a community easily become involved in offensive operations against neighbouring communities’ (Bhatia and Sedra 2008: 30). Whatever short-term benefits that may be 144

145

7 | Afghanistan’s experience with militias

derived from fighting the insurgency are arguably not worth more than the longer-term costs of adding new armed groups to the dynamics of insecurity. A second major critique has been that the US military, and even its Special Forces, simply does not have sufficient grasp of the complexity of the local socio-political environment to be in a position to make wise decisions about forming community militias. A failure to fully comprehend local realities was highlighted by the US military’s effort to partner with the Shinwari tribe in Nangarhar. By March 2010, following disputes over access to US aid provided through their deal with the US military, extensive violence had broken out within the tribe, between the Ali Sher Khel and Shublai sub-tribes, leaving thirteen people dead.22 Additionally, the term ‘tribal militia’ or arbakai is bandied about endlessly but the nuances behind it are not widely understood. Important in this regard is that it really only applies to the Pashtuns, and even within that large community there has been so much social fracturing after three decades of war that many communities simply lack the cohesion to manage their own militias. Indeed, the mountainous far south-east of Afghanistan is one of the few areas of the country that has not at one time or another seen the entrenchment of warlords and the replacement of tribal structures (Tariq 2008: 2). Important in this regard is that arbakai forces are dependent on the decisions of community jirgas (council meetings). The jirgas represent the local community, from which the arbakai’s members come, and this provides the arbakai forces with a ‘downward accountability mechanism’ which makes them substantively different from other militias (ibid.: 8). Attempts to use arbakai forces that do not make use of representative jirgas are more likely to fail – witness what was tried in Tagab district in Kapisa province versus the more successful example of Afghan government efforts in Kunar province in 2004, which engaged the arbakai only indirectly through the local jirga, to which financial support was provided (ibid.: 9). The third major critique has been internal to the US government. As the LDI programme was gaining traction within the Department of Defense in late 2009 fears were being raised within the Department of State that not enough formal oversight linkages with the Afghan government were being created. As a US embassy spokesman argued, ‘We are committed to doing this right, and that means taking the time for the Afghan government and people to decide on whether and how to proceed.’23 The LDI was strongly pushed by the then US commander, General Stanley McChrystal, who has a Special Forces background and was central to much of the ‘irregular warfare’ aspects of the US ­campaign in Iraq, i.e.

the ‘Sons of Iraq’ programme. His desire to have more flexibility to use Special Forces to directly engage local communities that have shown signs of resistance has faced opposition from the US embassy in Kabul. The embassy has been more insistent that the Afghan government drive the process, as the AP3 efforts and the more recent Local Police Force initiative allowed for. Subsequently, the Department of State was slow to approve financing for LDI efforts, but the US military still pushed ahead with its own financial resources. The spat between the State and Defense departments over the LDI starkly highlighted the gaps between imperatives for state-building versus waging counter-insurgency operations within the US government and the prominent role that militias can play in that internal discourse. With these criticisms in mind, it is worth considering the variety of local perspectives on the formation of community militias in 2009 and 2010. While it is impossible to generalize too much about both public perceptions and individual motives, especially in a country where mass surveying in the countryside is difficult, some general themes can nonetheless be drawn out. The first need is to articulate the context of rural Afghanistan presently, which is defined by direly poor economic conditions and very real possibilities for routine, everyday violence. Decades of war and the lack of a government monopoly on violence have rendered the countryside a Machiavellian survivalist playground, something only compounded by the upsurge of the Taliban insurgency since 2005. In this environment local communities are pressed to focus on their immediate future and negotiating between the presence of large armed groups – Western armies, Afghan security forces and assorted insurgent groups. In this context, locals are ‘concerned mostly with providing for their immediate future’.24 A massive sense of disempower­ ment is common, or, as one local commented on feeling marginal to security developments, ‘All I’ve got is the turban on my head and a piece of bread.’25 In the ongoing anarchy of rural Afghanistan, local communities face a ‘situation of terrifying uncertainty’, with many armed groups around them, all threatening them for compliance of some sort, and with little control over outcomes directly affecting themselves (Kilcullen 2009: 67). Given this environment of relative anarchy, relationships and perceptions of power are critical to determining motives and ultimately decisions taken. Local communities, feeling highly vulnerable to the ebb and flow of the assorted armed groups passing through their home areas, most often feel compelled to make opportunistic decisions for sheer survival, namely ‘the people feel the need to both support the 146

Conclusion

Militias have long been present in Afghanistan, and they played central roles in the wars that have ravaged the country since 1979. Hence, a major crux of the challenge for stabilizing Afghanistan is how to contend with militias and the warlords who have driven them. Many different approaches have been taken towards them, but if there has been one overarching theme it is that of engagement, because the central government has always been weak and militias have always provided an expediency of force to attempt to make up for that. The Soviets and their allies in the Afghan communist governments sought to use militias in 147

7 | Afghanistan’s experience with militias

government and the enemy, depending on who is in the area at any given time’.26 Simply put, feeling disempowered means that local communities will naturally gravitate towards those larger armed groups that are both present and perceived as having the greater degree of power over their lives. Collaboration with either side is hence driven by a very rational desire to have at least some minimal degree of order and predictability and hence community or personal security amid the broader anarchy. If some degree of reactive engagement is ultimately necessary, simple ‘fence-sitting’ and not choosing sides being a luxury not afforded, it may instead be likely that opportunistic relationships will be forged. For instance, many traditional leaders, known as maleks in Afghan villages, have seen their long-standing authority circumscribed by the Taliban, who more often empower religious leaders, mullahs, and hence they look for support from the Afghan government or the US military, which allows them to reassert themselves in rural communities (ibid.: 108). Conversely, David Kilcullen has argued that many of the insurgents are actually ‘accidental guerillas’, locals who do not share the insurgents’ rigid ideology but feel compelled to join them in order to resist what they perceive as foreign interference in their affairs, or feel alienated by the national government (ibid.). This opportunistic engagement of the insurgents allows local communities to act upon wide disdain for the Karzai government and an ongoing Western military presence, notably in Pashtun areas. Lastly, when the presence of insurgents attracts Afghan and Western security forces, and vice versa, with fighting naturally en­ suing, a strong motive arises among local communities to keep both sides out of their home areas. As an Afghan elder near Kunduz rationalized about the need for local militias, which were formed there in late 2009, ‘Both the Taliban and international forces were killing us, this was too much. So we picked up our guns and forced the Taliban out of our village; now we are living a peaceful life.’27

assorted ways and put a great deal of emphasis on co-opting insurgents into joining allied militias. Amid the violent anarchy of the 1990s the Taliban movement started out as a small community militia focused on local self-defence and garnered much of its momentum by juxtaposing itself against the large, predatory militias of the warlords fighting for control of the country. Following the fall of the Taliban, the single biggest challenge was how to engage the component forces of the Northern Alliance with the need to build a new, sovereign government. Engagement was allowed for, and while the large militias, eventually formed into the AMF, were themselves disbanded through the DDR processes, the warlords who had led them were widely entrenched in the nascent government. As Ahmed Rashid has argued, the US decision to embrace warlords such as Dostum and Fahim ‘was the most fatal mistake’ (Rashid 2008: 135). This was mostly because it called into question the broader intents of demilitarization as the new government was left with a legitimacy gap with much of the Afghan public because of the widely held loathing of the men responsible for the extremely violent 1990s. Additionally, while the DDR processes largely unfolded during a period of relative calm, with the upsurge in the Taliban insurgency starting in 2005 and 2006 there have been multiple efforts, some formal and others informal, by both the Afghan government and the US military to utilize militias. While there has been no attempt to re-form the large, semi-structured militias of the old warlords, the new efforts have focused on engaging local ‘self-defence’ militias in rural communities. These efforts can be considered a form of remilitarization and starkly juxtapose themselves against the intentions to build up the ANA and ANP as the monopoly providers of security. One of the most interesting aspects of these attempts at remilitarization in support of ongoing counter-insurgency efforts is the lack of engagement with the Taliban insurgency itself. Rather than the co-option, with ‘reconciliation’ being largely euphemistic, of insurgent groups, the focus has been on supporting the self-defence efforts of local communities against insurgents. Even the Soviets put a great deal of emphasis on co-opting insurgents to stop fighting them, as did the USA in Iraq. In contrast the USA and Kabul have made no broad, long-term effort at co-option, and indeed, as one commentator noted, ‘the message from Washington and its Afghan allies could hardly have been clearer: hold out an olive branch and you will go straight to jail’.28 For example, several ex-Taliban commanders who have sought rapprochement with or even reintegration into the Afghan government have been imprisoned, such 148

149

7 | Afghanistan’s experience with militias

as a former Taliban foreign minister and a deputy minister of intelligence. This lack of effort to co-opt insurgents is somewhat surprising because by some accounts 90 per cent of the Taliban are ‘potentially cooptable’ (Kilcullen 2009: 49). There have been sporadic announcements of reconciliation efforts, such as when Karzai announced in March 2005 that there would be a ‘Peace and Reconciliation Commission’, but these intermittent and limited efforts overall produced very little. With the January 2010 London conference on Afghanistan declaring reconcili­ ation a strategic priority, furthered at the Kabul conference in May 2010, there may be more systematic efforts to reach out to insurgent groups at a local level. If that were to happen, some efforts would inherently need to be made to effect the transition of these armed groups, most likely, as is common, into semi-official militias in their home areas, then subsuming them into the Afghan security forces or ‘DDR-ing them’ into civilian life.

Conclusion: militias and the search for local security

This book has sought to critically examine militias, their origins and methods, and the approaches adopted by scholars to conceptualize them, as well as the policies employed by the international community aimed at either managing them or addressing the ‘problem’ of militias. By grounding our work in extensive field research and, concurrently, applying a basic framework that allows for rough comparison across a diverse set of case studies, we believe that this study has been able to illuminate key features of the dynamics that play out between cycles of conflict, the emergence of actors and the forms of international intervention. As argued throughout the book, we contend that security strategies that ignore or underplay the local political, social and economic context within which militias are fostered and operate are inevitably flawed. Recognizing the possible points of convergence between local interests and broader aims sets the stage for more convincingly rooted collaboration. This convergence is, as the empirical work has shown, founded on the search for local security, and can be successfully mobilized only insofar as international actors are able to frame their aims in terms that can translate into goals that are recognized and meaningful to local actors. The case studies covered in this volume highlight the many contextual factors that shape these different undertakings: the alignment of regional interests and influences bearing on a DDR process supposedly taking place within a country; the political economy of war and peace distinctive to each conflict zone and cutting across formal borders; the unique characteristics of militias and their leadership; and the evolving global and normative environment in which DDR is conceived and carried out (Berdal and Ucko 2009). Our original focus on post-conflict settings has, necessarily, been conditioned by these same changing dynamics of the near-conflict situations which we investigated, highlighting the contingent nature of peace, be it formal or social, and the changing response of local actors to these conditions. This dynamism is something we hope to have captured in our analysis, informing our understanding of what is too often seen 150

Militias and the search for local security  The most significant thread of continuity found throughout all of the case studies in the book is the emphasis on self-defence. Framed as it is by local concerns, conditions and actors – in short, the local context – the importance of this notion should not be underestimated. A primary challenge of demilitarization efforts everywhere is to draw a balance between the needs of state-building and the primacy of strengthening national government institutions with the demands of local communities for immediate security provisions. These seemingly congruent imperatives can, as demonstrated in situations as varied as eastern DRC and Afghanistan, be disturbingly at odds with one another as national governments seek to impose their authority on security-seeking local communities. Unfortunately, the international community suffers from a chronic inability to achieve this balance, often giving more emphasis to local disarmament and at times de­cidedly myopic attempts at government capacity-building, especially of the security sector at the local level. This often leaves local populations effectively detached from – and even threatened by – a process that is ostensibly all about them. While arguments by militias for self-defence can be dubious and self-serving in many cases, the notion itself should nonetheless not be dismissed wholesale, especially when considering the motives of rank-and-file militiamen versus senior leaders. For example, the SSDF presented itself as an organization focused on protecting assorted communities in South Sudan, both from the Khartoum government and the SPLA rebel movement. While it is easy to question the sincerity of SSDF senior commanders who were safely ensconced in Khartoum villas pontificating on their ostensibly altruistic motives from a safe distance 151

Conclusion

by external observers to be a rigid, timeless set of conditions rooted in ideas about the essentialist nature of ‘deep-set ethnic conflict’ in a given region or the barbarity of traditional practices of politics and warfare. This is not to say that these features aren’t significant accelerants of conflict, fuelled or appropriated by militias as the case may be, but rather that they play a part in a world that is framed by long-standing local or regional factors that, at the same time, are firmly rooted in modernity. It is a very contemporary world where cellular telephones, the World Wide Web, missile technology, humanitarian ideals and diaspora communities become both instruments and resources of conflict. And it is one in which the international policy-making community – naive, self-serving and cynical in equal measure – actively intervenes under the banner of promoter of peace.

from the war, the authenticity of normal militiamen within South Sudan bears more consideration. For your average SSDF militiaman on the ground, the earnestness of their argument was much easier to believe in the context of chronic anarchy spread over decades. More ­believable still were the arguments of the White Army militiamen of South Sudan, who felt their local village militias had an essential role to play in their own lives, even after the CPA had come into effect. This was because of a fear of a rebel movement, the SPLA, which had become the government but was still regarded by many communities as ethnic­ally biased and predatory, while the broader situation remained that of a UN peacekeeping presence and a nascent security sector that was widely incapable of providing protection in rural areas. Similarly, in the DRC, the imperatives for security provisions are particularly pressing, considering that so much of the ongoing violence in the east and the proliferation of militia groups so often stems from the extreme violence of the Rwandan genocide. The emergence of the CNDP, for example, was not only to secure resource-rich areas in the Kivus, but also to provide protection to the local Kinya-rwandan Tutsi population against the FDLR (which contains ex-FAR and interahamwe elements) and Mayi-Mayi. Similarly, the Mayi-Mayi militias have historic­ ally, and in the post-2003 period, been considered a significant part of the local defence force – against rebel movements and foreign armies. Having fought alongside the government during the two Congolese wars, Mayi-Mayi have largely been motivated by the need to protect local communities, packaging this in a very staunch ‘Congolese’ nationalist trajectory – as such, in many circumstances, the Mayi-Mayi have fought alongside the FARDC against the FDLR and CNDP. In Afghanistan the US military’s efforts to ‘remilitarize’ rural areas through the formation of militias is building upon compelling concerns within rural communities to be directly involved in their own security provisions. The sincerity of their claims of self-defence is all the more believable considering the inability of the government to meet the ‘promises’ by the state to immediately step into the security void created in the aftermath of DDR. In this sense, for many Afghan locals, the DDR processes in mid-decade were simply conducted too early, considering the weakness of the state’s security sector at the local level. Furthermore, right or wrong, the US military’s ongoing militia effort is building on local perceptions that there is a substantial difference between a small, localized militia acting in its own community and the large militias that ravaged the country in the late 1980s and 1990s, such as Abdul Rashid Dostum’s Jowzyani militia, which was composed 152

Research and policy implications  The result of this comparative research study is that two broad agendas have emerged, one with an academic orientation and set of considerations and a second aimed more directly at the international policy-making community. A scholarly agenda

As can be seen in the review of literature, modelling conflict for predictive purposes has an important place in the established approaches to militias. However, following Karen Ballentine and Jake Sherman, who argue for deeper analysis of causality in assessing the relationship between motivation and pursuit of conflict, this project has identified a number of factors which play a crucial (and sometimes unrecognized) part in delineating actors and their actions (Ballentine and Sherman 2003). Given that, this book suggests that recognition of the following 153

Conclusion

of upwards of ten thousand men and acted violently in broad swathes of the country. In this sense, one man’s local ‘militia’ is understood to be totally different from how the international community tends to understand all militias, i.e. as morally repugnant and universally bad. This is not to say that the actions of militias should be forgiven or that the US military is right to encourage militia formation in Afghan­istan. Rather, the argument is merely that there is consistently a strong d ­ ynamic of self-defence to their motives, at least from their perspective, even if that is often overwhelmed by other motives, such as financial gain. Nonetheless, rather than treating the local impulse towards security as insincere or purely secondary to other imperatives, it is something that should be given greater consideration by inter­ national policy-makers. Instead of arguing that militias should be armed, the local impulse towards security could be built upon in other, more benign ways, such as involving communities more directly in decision-making processes regarding DDR programming, including sequencing and time frames. Furthermore, other evolved types of community actors could be involved to replace militias, such as ‘community watch’ organizations. At a most base level, membership in a militia is about empowerment through participation. That is something that is clearly craved by local communities, which so often find themselves utterly marginalized between national governments, international actors and security programming that do not give them wider opportunities for participation. Communities should surely be given broader involvement in their own security, which would be one of the best ways to mitigate interest in militias in the first place.

themes will do much to capture the key features that distinguish militias and their relationship to conflict and its potential resolution.

Context matters  One neglected aspect of understanding militias is the role of context in shaping actors and agendas in conflict situations. To generalize, militias arise and operate in an environment of weakened to non-existent governing authority, both within and across the terri­torial boundaries of a recognized sovereign state. Concurrently, the spatial dimensions of conflict are deeply affected by matters such as state infrastructure, resource distribution and population disbursement. A predatory core and an anarchic periphery mark the ‘classic’ depiction of a conflict-ridden, weak state, though particulars and dynamics vary with each situation. The context in which militias operate clearly conditions their subjective sense of motive and action as well as the ‘objective’ perception of them on the part of international actors. Alliance formulation de­ velops in this setting on the basis of expediency mediated through local networks and customary practices. It is our belief that the operational expediency found in alliance formation is replicated (or is likely to be replicated) with regard to the instrumentality of violence. These groups use it because it has purpose; the particular mode employed may have special local and/or cultural resonances (severing of limbs, rape, etc.) but also because this is all they can do in a particular context. This reflects the relative transitory nature of their power (hard, not soft, to use international relations terminology), which needs to be asserted frequently through visual demonstrations and actions. This may also point to the way of resolving their integration back into society in a variety of different ways.  The other aspect of context is that it conditions where these groups are on a spectrum of violence. The shifting of approaches to achieving aims (be they narrow short-term survival, resource extraction or otherwise) is linked, on the one hand, to the changing context within which they operate and, on the other hand, to redefinitions of their underlying aims. What becomes fixed is the perception of them, probably in the capital and held (and reified) by NGOs, international organizations and foreign governments – not without more than a little spin by a ‘trusted’ local interlocutor – so that these groups become locked into a zero-sum game that does not enhance the possibility of reintegration (Khmer Rouge is always Khmer Rouge because Pol Pot is always Pol Pot). The change of approach adopted by the US-led coalition in Iraq towards ex-Baathists after 2004 made possible the coalition’s reconciliation 154

Rationality and motivation redux  A further observation on commercial rationales beloved of the behaviourists: it seems self-evident – despite the existing literature – that not enough attention is paid to the meaning of the commercial/greed rationale and its impact on the conduct and formation (or reformation) of militias. Doing business requires some very basic forms of reciprocal trust to develop between trading partners. 155

Conclusion

with the local Sunni community at large and, with that, established the basis for growing stability in the country. This strongly suggests that militias are not always irredeemably committed to ‘spoiling’ the peace process and perpetrating heinous acts, but that in some cases it is the international community which does not allow them to reinvent themselves in any other way.  In addition, aims themselves are not fixed and ‘roving bandits’ (in Tilly’s words) can recognize profitability in becoming ‘stationary bandits’, hence moving beyond the narrow claims of security (see Tilly 1990). They become enmeshed in commercial pursuits, beholden to outside interests, subject to parochial factional disputes and, in the course of this, alter their aims and founding rationale. Unlike the state, which is grounded in a bureaucratic-administrative logic that privileges pro­ cedure, routine and precedent over innovation and contingency, militias operate in a setting that rewards tactical change and other short-term behaviour strategies. It is – and this goes against the theoretical position of the systems approach – only in the context of enhancing the relationship to fixed territoriality (see below) that one begins to see the state-like behaviour take root. Finally, the impulse towards sovereignty remains the most important measure of militias as it conditions their conduct, aspirations and the possibility of international influence. Reflecting this, territoriality in its spatial aspects as well as in terms of their relationship to a given population is the most significant division between various militias. Those groups which are divorced from the long-term aspirations of sovereignty hold little possibility of acclimatizing to international standards of conduct. However, those which cast their aspirations towards sovereignty are, perforce, oriented to external norms and thus open to change. This need not be in the sense of aiming towards a separatist agenda but rather of a willingness to identify oneself with or participate (even if this is instrumentalist in origin) in state structures. In this sense, the normative project for the international community of states clearly has to be to transform ‘roving bandits’ into ‘stationary bandits’ (Chabal and Daloz 1999).

In this lie the procedural seeds of an administrative apparatus, one which can be bargained with and is able to conduct itself on terms which roughly accord with those of the wider society. This transition point, if reached, needs to be examined closely as a proverbial ‘window of opportunity’ rather than seen exclusively as a source of conflict perpetuation – poachers can become gamekeepers. Collier and Hoeffler’s assertion that: ‘[G]rievance is to a rebel organization what image is to a business [… and while] popular perceptions see rebellion as a protest motivated by genuine and extreme grievance […] economic analysis sees rebellion as […] the ultimate manifestation of organised crime’ (Collier and Hoeffler 1998: 152, 144). Patrick Chabal and Jean-Pascal Daloz’s insightful assessment of motivation in African politics, that is to say that much of what is seen externally as chaos has an inner systemic logic bound to the political instrumentality of disorder (despite a tendency towards reductionism), is worth bearing in mind when divining motive. As articulated earlier, militias must be seen as lucid actors, whose rational use of violence exemplifies their various strategic objectives. This highlights that overlapping motivations, ranging from local defence and personal enrichment to ethnic/religious protection, underlie militia behaviour. 

Recognition and legitimacy  The importance of ‘recognition’ for actors in international politics remains a relatively unexamined area. This is des­ pite the fact that, as is the case with other political actors, the drive for recognition permeates the character and conduct of militias. A ­ cquiring recognition is not only an acknowledgement of material power attained – gained through violent means – but is reified through the appropriation of familiar symbols (ethnic badges, historical grievances, etc.) that provide (often local) meaning and context to the activities of militias and armed groups. Christopher Cramer’s point is worth repeating: ‘violence also makes sense as a form of communication, often where other means of communication have broken down’ (Cramer 2006). Certainly classical tracts on revolutionary warfare understood the vital communicative function of public acts of violence as declarations of being and purpose aimed at the general public. Moreover, the authority accrued by militias through such violence allows for more efficient allocation of power: no longer are acts of gratuitous violence necessary to secure compliance from targeted communities and/or international actors. In their stead, the use of physical sources of influence over a population can be supplanted by ritualized expressions of authority. These ritualized forms of recognition-cum156

Organizational structure and strategies  The structure of decisionmaking within militias, something that has preoccupied Policzer (2002), deserves greater attention as well. In this, the basic schema used in the systems approach is a helpful starting point but needs to be modified to account for social networks. For example, a recognition of the place of networks and, in those instances where ethnicity has served as a key organizing principle for armed groups, social structures is important to understanding how particular choices are made and implemented (and here the southern Sudanese provide a case in point). Furthermore, there has been little analysis of preference formation, beyond the simple dichotomy of ‘greed and grievance’, and in particular hierarchies of choice for militias. If one asserts that militias are essentially constructed according to the charismatic leadership model (following from classic Weberian organizational types: charismatic, feudal, bureaucratic), then this holds implications for understanding choice in militias, implying a hierarchical form of decision-making. As noted above, the application of some of the insights of principalagent theory are strongly suggestive of new interpretations of organ­ izational structure and relations, as well as introducing new possible avenues of engagement for the international community. It could be argued that just as institutional theorists see socialization of states ­occurring through increasingly assertive and autonomous intergovernmental organizations, so too some similar process takes place between 157

Conclusion

authority are important markers in the evolution of militias as they, much like the aforementioned drive for territory, and however laden with the symbols of violence, lay down the preconditions for social stability. Following from this, the instrumentalization of ‘legitimacy’ by the international community is a precious commodity that deserves more consideration. This was demonstrated most clearly by the UN Special Representative Aldo Ajello in his management of the feared guerrilla force Renamo during the 1992–94 UN peacekeeping operation in Mozam­bique. He carefully calibrated the use of recognition and symbols of international authority, including Renamo’s access to authorities such as the Vatican, to demonstrable changes in conduct by the Renamo leadership and its followers. Judicious application of the acknowledgement of leaders and movements has often not been a feature of the international community and, as a result, has elevated groups that provided little support for the peace process. Perverse incentives such as providing finance to parties or movements that ‘spoil’ peace negotiations only encourages other actors to behave in the same way.

the state (‘principal’) and the militia (‘agents’). Socialization could be said to occur through norms of interaction, with the state taking the role of socializing agent for the militia, effecting adaptation to the norms of the international community, and the militia seeking to extend the boundaries of action implied by the purported dependency exercised by the principal. Like states and intergovernmental organizations, these two entities engage in consultation over the rules of conduct and ­­barriers to direct action, seeking a tacit form of consensus on behaviour. The effort by sponsor states to transform militias into recognizable political organizations represents one expression of this; concurrently, the impulse to gain recognition from the international community on the part of militias, despite their sometimes egregious conduct breaking key international norms, is another. This suggests that an important arena of action is the negotiation over terms of legitimacy. Moreover, it needs to be recognized that militias are exceptionally dynamic organizations, and appreciating their evolution and the causes driving a respective militia’s change is critical. For example, the SSDF was merely an umbrella organization of assorted militias that essentially started as small, local, ethnically based militias. It achieved its prom­inence only through its coalescence as a large proxy force for the ­Khartoum government. That evolution from small ethnic militias f­ ocused primarily on local self-defence to broader motives more amenable to partnership with dictatorial Khartoum governments was profoundly important in understanding the militia in its last form. Similarly, the Mayi-Mayi militias, which have focused mainly on protecting land and defending local populations, have not only worked in coordination with Kinshasa to combat domestic and foreign threats, but have become very much entrenched in the autochthony/­allochthony discourse. As such, the Mayi-Mayi assert themselves as ‘authentic’ Congolese (i.e. autochthons) and reconstruct their identity according to the broader Great Lakes rwandaphone ethnic configuration – which reinforces the divisive and polarized nature of local politics in the ­Kivus. Perhaps even more profound was the example of the mutineers in Timor-Leste. Their formation was triggered by the political crisis surrounding the so-called petitioners, deserting members of the Timor-Leste army complaining of discrimination within the military. While the mutineers may initially have been relatively open to dialogue and compromise with the RamosHorta and Gusmão governments, they ultimately attacked both leaders. In sum, they underwent profound changes to their driving rationales from limited issues revolving around ethnic tensions within the security sector to direct attempts to change the very leadership of the state. 158

A policy-making agenda

The policy implications of the research conducted by the authors, though clearly linked to some of the analytical insights raised above, nonetheless need to be laid out. Three spheres in particular are worth highlighting.

Towards greater flexibility in DDR  There is a basic need to appreciate that militias are going to be very diverse in form, motives, histories, etc., and that this requires an equivalent flexibility in determining policy responses. Excessively linear DDR and SSR policies that do not have this flexibility and are not tailored to local circumstances are bound to fail. One solution is to do more research in advance of formulating these programmes to allow for ‘scenario mapping’ and from that building ‘best practices’. At the same time, there needs to be fairly constant on-site review of programmes which can take into account the shifting rationales and conditions which drive militia conduct and adjust programming accordingly. Moreover, as the case of southern Sudan suggests, sometimes the best approach to militias may be to avoid formal, rigid demilitarization processes. The SSDF was successfully controlled through a combination of internal political negotiations and willingness to effectively abandon them to the hinterland with no discernible purpose. They were never propped up by the recognition provided by an internationally backed 159

Conclusion

In this regard, a key need for the conceptual framework articulated in Chapter 2 and applied in the case studies is to include a specific level of analysis focused on the teleogical development of militias. A better understanding of the changes militias evolve through would open up more policy options for responding to them. As it is, militias are often treated as static organizations needing to be summarily disbanded. Appreciating the subtleties of the changes could assist in devising more effective, context-specific policy responses. For instance, at least in the specific context of eastern Jonglei state in mid-2006, the resident White Army militias were disarmed relatively effectively because they were engaged by indigenous NGOs with the heavy involvement of local communities. This success was achieved largely because the policy response appreciated the origins of the youth militias from within the broader communities and used that as leverage. This contrasted sharply with SPLA efforts to forcibly disarm the militias in western Jonglei. Overall, the inclusion of understanding the teleology of militias in future fieldwork would strengthen the quality of future research produced.

process, and hence had less ability to generate the disruptive ‘spoiler’type behaviour seen elsewhere. Finally, the transition from the tenuous stability induced by a relatively successful DDR programme to a long-term form of post-conflict stability requires a commitment to security sector reform (SSR). International resources need to be fixed on this dimension of the process. As our case studies suggest, SSR processes that are not emphasized and given national priority can have catastrophic consequences – to the point where they can even prolong the life of militias, as in Timor-Leste.  In the case of the DRC, the fusing of the DDR and SSR programmes proved to be ineffective for a number of reasons. The overemphasis on disarmament and demobilization not only drained the DDR budget but also reinforced the notion that reintegration is something of an afterthought. Besides poor management at the local level (via CONADER), and poor planning and funding at the international level (via the World Bank and MONUC), the SSR process was implemented in a hasty and superficial manner – leading to a weak and fractured FARDC, especially in the Kivus. Finally, divorcing the DDR and SSR processes from the socio-economic and political dynamics of the eastern DRC, including abject poverty (especially high levels of unemployment among young men), impunity and weak judicial frameworks, politicization of identity at the local and provincial levels (especially in government), and transborder economic networks and the role of regional actors, eventually led to the failure of demilitarization and SSR efforts.

Provisions for local security  Militias are important because they revolve around the notion of ‘self-defence’ even if this is heavily manipulated by the militia itself or its state sponsors. Militias continue to have promin­ ence as a phenomenon where the state cannot impose a monopoly of violence. This is especially true in post-conflict situations in the aftermath of DDR processes, which, though often deemed by those in the capital as formally complete, have not necessarily filled the security void for local communities with state-sponsored protection. Indeed, given that the state itself may in recent memory have been the perpetrator of violence and insecurity for these local communities, it stands to reason that assertions by international actors that the reintroduction of state authority to a region will quell the need for security are deeply flawed. In this regard, as our case studies demonstrate, the most consistent failing of the international community has been to undertake DDR programmes but not provide the required security to make them worthwhile. This argues for, in select situations, the application of a relatively 160

Reasserting the ethical yardstick  The Darwinian quality of scholarly work on militias is striking – a kind of tautology is used which suggests that the only groups worth recognizing are those which demonstrate survivability. In this sense, though it is fashionable to debunk the ideological dimension in ascribing international legitimacy to militias by the superpowers as mere instrumentalism in the service of national interest, the fact of the matter is that ideology did serve a number of important normative functions. For example, broadly construed public platforms and organizational structures (Leninist vanguard, democratic centralism, pluralist party system, etc.) served to prescribe a positive relationship between militias, their goals and the communities in which they found themselves. Naked stripping of resources and brutal exercising of power, while obviously not unknown, were generally discouraged by sponsor states in the build-up to winning control over territory and in some instances (where overarching ideological goals were jeopardized) inspired shifts in support away from violators of these norms. Again, this only underscores the importance of assessing the relationship of militias to aspirations that conform to or contain the requisite impulse to sovereignty, as well as giving consideration to their established ties to local communities and external actors as key parameters for affording them some form of recognition. Identifying those militias which exhibit stabilizing tendencies in their interactions with local communities, which seek out patterned relationships which increasingly trade on symbolic forms of authority, may be a crude device for distinguishing between the militias or armed groups that could play a constructive role in peace and those whose tactics suggest no significant recognition by local communities. As has been the case with British military intervention in Sierra Leone, this could serve as the basis for adopting concerted action against certain militias. Conclusion

Militias are ultimately configured around local actors and communities of interest. Understanding the context within which militias arise, 161

Conclusion

large blanketing of the conflict zones by UN peacekeeping forces with a robust mandate as a necessary step to develop the conditions for achieving a sense of security. However, as evidence has demonstrated, the role of the UN as ‘neutral’ peacekeeper has been compromised time and again by the actions of its designated forces on the ground – from Srebrenica to the Kivus – so, in advocating this, we are quite conscious of the patent shortcomings of such an approach.

the patterns of behaviour that militias follow, and the differing impacts emerging from the implementation of international policy prescriptions is critical to addressing their role in post-conflict settings. Such better understanding can pave the way for more constructive policy responses based not on unfettered optimism or undue negativity, but on the realistic possibilities offered by a given situation. Studied insights into particular contexts and actors can reveal points of entry for a ‘quick impact’ that may induce significant changes in the behaviour of militias. The flexibility in policy formulation and application, however, is as dependent upon dynamic local and international institutions as implementing agents in post-conflict settings as it is upon the muta­ bility of local leaders and organizations engaged in forms of political violence. Without both, the possibilities of achieving the long-term peace remain remote.

162

Notes

1 Introducing militias and demilitarization

1  Certain countries have to address demilitarization not only of ‘national’ and domestic militias, but also foreign-based militias and armed groups. In the Democratic ­Republic of Congo (DRC), the ­national DDR process operated parallel to the disarmament, demobilization, repatriation, reintegration or resettlement (DDRRR) process for foreign armed groups and militias. The DDRRR process was guided by the Multi-Country Demobilization and Reintegration Programme (MDRP), which was spearheaded by the World Bank and UNDP. 2  According to Hill, state and transnational actor relationships can be categorized as: normal, bargaining relationship; competitive, power relationship; or transcending relationship. See Hill (2003: ch. 8, esp. p. 205). 3  We recognize that using a particular label may be value-laden, but we felt it was important to adopt a specific label as it will allow us to help define the context, highlight what factors are important, and code the abundant information on armed groups so as to provide policy directions. This does not, however, take away from the fact that labels and terminology are value-laden, contextual, impose patterns and foster an element of exclusion; and that ‘naming’ a specific actor in the international

system not only removes them from the unknown but also assigns them a set of characteristics, motives, values and behaviours. However, for conceptual clarity we felt it was important to use a term we believed best encompassed the diversity of the militias in our research, while also understanding that the identity of militias is fluid and shaped by the context and how the world/discourse sees them. See Bhatia (2005). 4  For a lucid analysis of the LRA, see Borzello (2009). 5  For an updated analysis, see Restrepo and Muggah (2009). 6  For a cogent analysis, see Jett (2001). 2  Conceptualizing militias

1  Owing to the complexity of certain conflicts, ‘foreign’ armed groups are also present in the conflict and post-conflict environments. In the DRC, the DDR programme was undertaken simultaneously with the DDRRR (disarmament, demobilization, repatriation, reintegration and resettlement) programme. 2  This is in line with the growing literature on demilitarization. See especially Colletta and Muggah (2009) and Torjesen (2009b). 3  See Collier et al. (2009); Collier and Hoeffler (2004); Keen (1998); Berdal and Malone (2000). 4  Herbert A. Simon defines rationality as denoting a ‘style of behaviour that is appropriate to the achievements of given goals,

163

within the limits imposed by given conditions and constraints’ (Simon 1984: 408). 5  The strategies include: attrition; intimidation; provocation; outbidding; and spoiling. See Kydd and Walter (2006). 6  See Wendt (1992: 391–425); Wendt (1994: 384–96); Wendt (1996); Wendt (1999). 7  Burchill et al. (2001: 194–6). For example, according to Campbell, the violence of ethnic cleansing in former Yugoslavia in pursuit of a ‘pure, homogeneous political ­identity’ is simply a ‘continuation, albeit extreme, of the same political project inherent in any modern nation-state’ (Campbell 1998: 23). 3  The South Sudan Defence Force

1  For a further review of the SSDF’s evolution circa 2005/06, see Young (2006). 2  Interview with SSDF soldier, Ket­bek, Upper Nile state, 24 July 2006. 3  UNMIS briefing on security/ political situation by UNMIS military officers at Sector III Headquarters, Malakal, 9 August 2006. 4  Interview with Anthony Nyawelo, SPLM Security Adviser for Upper Nile State, Malakal, Upper Nile state, 11 August 2006. 5  Interview with Stephen Pal Kun, Longochuk County Commissioner, Malakal, Upper Nile state, 11 August 2006. 6  Interview with SPLA general officer, Malakal, Upper Nile state, 3 August 2006. 7  Interview with Charles Lam, SPLA brigadier general, Nasir, Upper Nile state, 24 July 2006. 8  Interview with Anthony Nyawelo, SPLM Security Adviser for

Upper Nile State, Malakal, Upper Nile state, 11 August 2006. 9  Interviews with SSDF officers and soldiers, Ketbek, Upper Nile state, 30 July 2006. 10  Interview with Anthony Nyawelo, SPLM Security Adviser for Upper Nile State, Malakal, Upper Nile state, 14 July 2006. 11  Interview with SSDF officer, Nasir, Upper Nile state, 23 July 2006. 12  Interview with SSDF officers, Malakal, Upper Nile state, 29 July 2006. 13  Interview with SPLM official, Panyikang county government, ­Mala­kal, Upper Nile state, 3 August 2006. 14  Interview with SPLM official (ex-SSDF), Nasir, Upper Nile state, 23 July 2006. 15  Interview with SSDF officer, Nairobi, 4 July 2006. 16  Interview with SPLM official (ex-SSDF), Nasir, Upper Nile state, 23 July 2006. 17 Ibid. 18 Ibid. 19  Juba Post, Juba, 20–27 July 2006. 20  Interview with SSDF soldier, Ketbek, Upper Nile state, 27 July 2006. 21  Interview with SSDF officer, Nairobi, 6 July 2006. 22 Ibid. 23  Interview with SSDF officer, Mala­kal, Upper Nile state, 18 July 2006. 24 Ibid. 25  Interview with SSDF officer, Nairobi, 4 July 2006. 26  Interview with SSDF officer, Nasir, Upper Nile state, 23 July 2006. 27  Interview with SSDF officer, Malakal, Upper Nile state, 18 July 2006. 28 Ibid.

164

SPLA officer, Malakal, Upper Nile state, 3 August 2006. 44  Interview with SSDF officer, Nairobi, 3 July 2006. 45  SSDF press release, ‘SSDF Press Release: SPLM/A declares genocidal war against civilians in greater Upper Nile region’, 16 April 2006, www.sudantribune. com/spip.php?article15091&var_ recherche=SSDF%2014%20april%20 2006, accessed 16 April 2006. 46  Interview with SSDF soldier, Ketbek, Upper Nile state, 31 July 2006. 47  Interview with Murle SSDF officers, Malakal, Upper Nile state, 19 July 2006. 48  Sudan Mirror, Juba, 14–27 August 2006. 49  Interview with SSDF officer, Nasir, Upper Nile state, 30 July 2006. 50  Interview with SSDF officer, Nairobi, 4 July 2006. 51  Interview with UNMIS official, Malakal, Upper Nile state, 6 August 2006. 52  Interview with SSDF officers and soldiers, Ketbek, Upper Nile state, 30 June 2006. 53  Interview with leaders of northern Murle communities, Lok­on­ gole, Jonglei state, 23 August 2006. 54  UNMIS Mediation Briefs, official notes of UNMIS Sector 3 Area Joint Military Committee from 16 May 2006, 30 May 2006 and 27 June 2006, Malakal, Upper Nile state. 55  Interview with SSDF officer, Nasir, Upper Nile state, 23 July 2006. 56  Interview with Peter Aduok, Member of Sudan Parliament, Mala­ kal, Upper Nile state, 16 July 2006. 57  Interview with UNMIS official, Malakal, Upper Nile state, 9 August 2006. 58  Interview with SPLM official, Malakal, Upper Nile state, 11 August 2006.

165

Notes

29  Interview with SSDF officer, Nairobi, 4 July 2006. 30  Southern Eye, Juba, 16–21 July 2006. 31  Interview with SSDF soldier, Ketbek, Upper Nile state, 31 July 2006. 32  Interview with SSDF soldier, Malakal, Upper Nile state, 19 July 2006. 33  Interview with UNMIS official, Malakal, Upper Nile state, 17 July 2006. 34  Interview with Murle SSDF officer, Malakal, Upper Nile state, 19 July 2006. 35  Interview with SSDF officer, Ketbek, Upper Nile state, 23 July 2006. 36  SSDF press release, ‘SSUDA Press Release: Jarch Group seeks pipeline routes from South Sudan oil fields’, 5 March 2006, www. sudantribune.com/spip.php? article14370&var_recherche=SSDF %20oil%20, accessed 5 March 2006. 37  Interview with SSDF officer, Malakal, Upper Nile state, 19 July 2006. 38  Interview with SPLM official (ex-SSDF), Nasir, Upper Nile state, 23 July 2006. 39  Interview with SPLM official, Upper Nile State Assembly, Malakal, Upper Nile state, 2 August 2006. 40  Interview with UNMIS official, Malakal, Upper Nile state, 9 August 2006. 41  Interview with Peter Aduok, Member of Sudan Parliament, Mala­ kal, Upper Nile state, 16 August 2006. 42  Interview with UNMIS official, Malakal, Upper Nile state, 17 June 2006; focus group with Murle civilians, Lokongole, Jonglei state, 23 August 2006. 43  Interviews with SSDF officers, Ketbek and Nasir, Upper Nile state, 23–24 July 2006; and interview with

59  SPLA communiqué addressed to the SPLA chief of staff, Lieutenant General Oyay Deng Ajak, from SPLA General Headquarters, Juba, 4 February 2006. 60  Interview with Charles Lam, SPLA brigadier general, Nasir, Upper Nile state, 24 July 2006. 61  Southern Eye, Juba, 16–21 July 2006. 62  Sudan Mirror, Juba, 31 July–13 August 2006. 63 Ibid. 64  Interview with SPLM official, Jonglei state government, Akobo, Jonglei state, 22 August 2006. 4  The White Army militias 1  Interview with Lou Nuer elder, Malakal, Upper Nile state, 5 August 2006. 2  For another review of the White Army militias and a more detailed description of their origins, see Young (2007). 3  Interview with group of Lou Nuer elders who resided in Malakal, Upper Nile state, 5 August 2006. 4  UNMIS UMAC Brief, ‘Security related events in Jonglei during May and June 2006’, 5 July 2006. 5  ‘Fragile disarmament in the south’, IRIN News, 18 August 2006. Most of these deaths were not combat deaths but from thirst and starvation after participants were scattered following the fighting. 6  ‘White Army, case for disarm­ ament’, Sudan Mirror, 14–27 August 2006, p. 19. 7  Interviews with UNMIS sources, Malakal, Upper Nile state, 6 August 2006. 8  Interview with UN source, UNDP DDR Unit, Malakal, Upper Nile state, 4 August 2006. 9  ‘Briefing on security and arms control in Jonglei’, Pact Sudan

pres­entation to UNMIS, Malakal, late April 2006. 10 Ibid. 11  Interview with SPLA brigadier general Charles Lam, Nassir, Upper Nile state, 24 July 2006. 12  Interview with youth, Akobo, Jonglei state, 22 August 2006. 13  Focus group with ex-White Army members, Akobo, Jonglei state, 22 August 2006. 14  Focus group with ex-White Army members, Walgak, Jonglei state, 20 August 2006. 15  SPLM security adviser and Shilluk tribal elder, Malakal, Upper Nile state, 16 July 2006. 16 Ibid. 17  This insight was articulated by a UNDP consultant who had undertaken extensive interviewing of White Army youth. Interview with UNDP consultant, Juba, South Sudan, 17 August 2006. 18  Interview with White Army youth, Walgak, Jonglei state, 20 ­August 2006. 19 Ibid. 20  Interview with UNMIS source, Malakal, Upper Nile state, 9 August 2006; and interview with UNDP source, Juba, South Sudan, 17 August 2006. 21  Focus group with Lou Nuer elders, Malakal, Upper Nile State, 16 July 2006. 22  Interview with the Speaker of the State Assembly, Malakal, Upper Nile state, 2 August 2006. 23  Interview with recently disarmed youth, Akobo, Jonglei state, 22 August 2006. 24  Interview with youth, Akobo, Jonglei state, 22 August 2006. 25  Focus group with group of youth in the market, Akobo, Jonglei state, 22 August 2006. 26 Ibid.

166

‘Security related events in Jonglei during May and June 2006’, 5 July 2006, stated: ‘SPLA assesses that SAF continues support to former SSDF of Yuai under the command of Simon Weijang Reth, deputy of Cdr. Simon Gatwic and is able to reorganize the scattered White Armies.’ 33  Interview with Lou Nuer member of SSDF, Malakal, Upper Nile state, 29 July 2006. 34  Interview with UNMIS source, Malakal, Upper Nile state, 9 August 2006. 35  Interview with Akobo Women’s Association, Akobo, Jonglei state, 23 August 2006. 36  Interview with UNDP DDR officer, Juba, South Sudan, 17 August 2006. 37  Interview with youth in market, Akobo, Jonglei state, 22 August 2006. 38  ‘White Army, case for disarmament’, Sudan Mirror, 14–27 August 2006, p. 19. 39  Interview with Lou elder, Malakal, Upper Nile state, 5 August 2006. 40  This idea was raised in an interview with a UNDP DDR officer, Juba, South Sudan, 17 August 2006. 41  Interview with UNDP DDR officer, Malakal, Upper Nile State, 4 August 2006. 42  Interview with UNDP DDR officer, Juba, South Sudan, 17 August 2006. 43  Interview with local SPLM security official, Malakal, Upper Nile State, 16 July 2006. 44  Interview with Akobo county commissioner, Akobo, Jonglei state, 22 August 2006. 45  Interview with Jonglei state deputy governor, Akobo, Jonglei state, 22 August 2006. 46  Focus group with ex-White

167

Notes

27  This insight was articulated by a UNDP consultant who had undertaken extensive interviewing of White Army youth. Interview with UNDP consultant, Juba, South Sudan, 17 August 2006. 28  This tendency of the militias to temporarily unify against larger enemies mirrors the organizational make-up of the Nuer nation more broadly. As anthropologist E. E. ­Evans-Pritchard noted: ‘The Nuer […] are divided into a number of tribes which have no common organization or central administration and these peoples may be said to be, politically, a congeries of tribes, which sometimes form loose federations’ (Evans-Pritchard 1940: 5). 29  For instance, there was an allegation that an SAF helicopter was used to directly resupply White Army militiamen during their conflicts with the SPLA. Interview with Nasir county commissioner, Nasir, Upper Nile state, 31 July 2006. This was also mentioned in an UNMIS brief entitled ‘Security related events in Jonglei during May and June 2006’, 5 July 2006, p. 4. 30  Interview with Jonglei state deputy governor, Akobo, Jonglei state, 22 August 2006. Furthermore, it is notable that after the White Army remnants left Yuai in May they went to Dolieb Hill area near Malakal, allegedly with the original intention of recuperating and replenishing their freshly depleted stocks of armaments from Thomas Mabior’s SSDF unit barracked there. Interview with UNMIS source, Malakal, Upper Nile state, 19 July 2006. 31  Interview with Lou Nuer member of SSDF, Malakal, Upper Nile state, 29 July 2006. 32  An UNMIS brief entitled

Army youth, Walgak, Jonglei state, 2 August 2006. 47 Ibid. 48  Interview with Akobo county commissioner, Akobo, Jonglei state, 22 August 2006. 49  Interview with UNDP DDR officer, Juba, South Sudan, 7 August 2006. 50 Ibid. 51  Interview with UNDP DDR officer, Malakal, Upper Nile State, 4 August 2006. 52  Interview with traders, Akobo, Jonglei state, 22 August 2006. 53  Interview with NHDF programme officer, Akobo, Jonglei state, 22 August 2006. 54  Interview with Pact Sudan programme officer, Akobo, Jonglei state, 20 August 2006. 55  Interview with Pact Sudan community development officer, Akobo, Jonglei state, 22 August 2006. 56 Ibid. 57  Interview with NHDF programme officer, Akobo, Jonglei state, 22 August 2006. 58  Interview with Pact Sudan community development officer, Akobo, Jonglei state, 22 August 2006. 59  Interview with NHDF programme officer, Akobo, Jonglei state, 22 August 2006. 60  Interview with Pact Sudan training consultant, Akobo, Jonglei state, 22 August 2006. 61 Ibid. 62  Interview with Akobo Youth Association staff, Akobo, Jonglei state, 22 August 2006. 63  Interview with Jonglei state deputy governor, Akobo, Jonglei state, 22 August 2006. 64  Interview with Nuer elder, Akobo, Jonglei state, 22 August 2006. 65  Interview with local resident, Akobo, Jonglei state, 22 August 2006.

66 Ibid. 67  Interview with Akobo county commissioner, Akobo, Jonglei state, 22 August 2006. 68 Ibid. 69  Interview with UNDP DDR officer, Malakal, Upper Nile state, 4 August 2006. 70  ‘South Sudan, Ethiopia discuss Lou-Nuer peace process’, Sudan Tribune, 6 October 2006. 71  Interview with Akobo county commissioner, Akobo, Jonglei state, 22 August 2006. This development was also noted in numerous UNMIS briefings and reports. 72  Ibid.; and interview with UNDP DDR officer, Malakal, Upper Nile State, 4 August 2006. 73  Interview with Akobo county commissioner, Akobo, Jonglei state, 22 August 2006. 74  Interview with Lou Nuer elder, Malakal, Upper Nile state, 6 July 2006. 75  UNMIS UNMO briefing in Malakal, Upper Nile state, 11 August 2006. 76  Interview with UNDP DDR officer, Malakal, Upper Nile state, 4 August 2006. 77  UNMIS brief, ‘Security related events in Jonglei during May and June 2006’, 5 July 2006. 78  Interview with Murle youth, Lokongole, Jonglei state, 23 August 2006. 79  Interview with Murle chiefs, Akobo, Jonglei state, 23 August 2006. 80  Interview with Murle chiefs, Lokongole, Jonglei state, 20 August 2006. 81 Ibid. 82  UN Office for the Coordination of Humanitarian Affairs, ‘Humanitarian action in Southern Sudan report’, 20 November 2009.

168

9  Interview with Joaquim Fonseca, special adviser to Prime Minister Gusmão, Dili, 15 April 2008. 10 Ibid. 11  Prime Minister’s Office, Government of Timor-Leste, Strategy to Resolve the Petitioners’ Case (Dili, n.d., but a statement released to the diplomatic community in late March 2008), p. 2. 12  Interview with Ed Rees, Country Director of the Peace Dividend Trust NGO, Dili, 7 April 2008. 13  Prime Minister’s Office, Strategy to Resolve the Petitioners’ Case, p. 3. 14  Interview with Joaquim Fonseca, special adviser to Prime Minister Gusmão, Dili, 15 April 2008. 15  Prime Minister’s Office, Strategy to Resolve the Petitioners’ Case, pp. 4–5. 16  S. Montlake, ‘East Timor rebel leader surrenders’, Christian Science Monitor, 29 April 2008. 17  Interview with international observer, Dili, 17 April 2008. 18  Interview with Joaquim Fonseca, special adviser to Prime Minister Gusmão, Dili, 15 April 2008. 19  Telephone interview with international observer, Dili, 5 ­November 2008. 20  Interview with international observer, Dili, 22 April 2008. 21  ‘Ramos-Horta attack “was coup attempt”’, The Australian, 21 May 2008. 22  See the YouTube video of an interview with Reinado in January 2008, www.youtube.com/watch?v=lgg JoQtAxCw&feature=related, accessed 15 May 2008. 23  To view a copy of the badge and supporting letter, see wikileaks. org/leak/reinado.pdf, accessed 5 November 2008. 24  Interview with Alison Cooper,

169

Notes

5  The mutineers of Timor-Leste 1  For a comparative review of the wide assortment of armed groups, including youth gangs, in Timor-Leste, see Timor-Leste Armed Violence Assessment (2009). 2  For a comprehensive review of DDR programming in Timor-Leste since independence, see Peake (2009). 3  Australian Defence Force, Operation Astute, www.defence.gov. au/opEx/global/opastute/index.htm, accessed 10 November 2008. 4  The most detailed description of the attack can be found in a leaked UN report investigating its response to it. See wikileaks.org.uk/leak/unmit -ramos-horta-shooting-2008.pdf. 5  There are many conflicting interpretations of the attack on RamosHorta, ranging from assassination attempt and foiled kidnapping attempt to it being a periodic meeting which, for unknown reasons, exploded into violence. There have been no public, formal conclusions made regarding the incident and the Timor-Leste government has resisted calls for an independent investigation while a criminal one is ongoing. There was a coroner’s report on Reinado’s death which raised even more controversy as it highlighted the possibility that Reinado may actually have been executed or ambushed. The best source of documents surrounding the attack are to be found at en.wikipedia. org/wiki/2008_East_Timorese_ ­assassination_attempts#cite_note37, accessed 15 May 2008. 6  Interview with international observer, Dili, 17 April 2008. 7 Ibid. 8  This was a common theme throughout the interviews conducted in Dili.

UNMIT spokeswoman, Dili, 15 April 2008. 25  Interview with Major Phil Pyke, ISF spokesman, Australian Defence Force, Dili, 21 April 2008. 26  See the ISF’s Memorandum of Understanding, signed between the Australian and Timor-Leste governments and the UN in 2007, www.lao hamutuk.org/reports/UN/07MOU-TLUN-Aus.html, accessed 22 May 2008. 27  Interview with international observer, Dili, 22 April 2008. 28  YouTube video of Reinado, www.youtube.com/watch?v=lggJoQt AxCw&feature=related, accessed 15 May 2008. 29  Interview with international observer, Dili, 22 April 2008. 30  Interview with Joaquim Fonseca, special adviser to Prime Minister Gusmão, Dili, 15 April 2008. 31  Interview with Antonio Cardosso, FRETILIN Member of Parliament, Dili, 14 April 2008. 32  J. Head, ‘East Timor’s struggle to build a nation’, BBC News, 15 ­February 2008, news.bbc. co.uk/2/hi/asia-pacific/7246109.stm, accessed 18 February 2008. 33  Interview with Ed Rees, Country Director of the Peace Dividend Trust NGO, Dili, 7 April 2008. 34  Interview with Joaquim Fonseca, special adviser to Prime Minister Gusmão, Dili, 15 April 2008. 35  Interview with Ed Rees, Country Director of the Peace Dividend Trust NGO, Dili, 7 April 2008. 36  Interview with international military observer, Dili, 10 April 2008. 37  Interview with international observer, Dili, 17 April 2008. 38 Ibid. 39 Ibid. 40  Interview with Juan Carlos Arevulo, Deputy Commissioner of UN Police, Dili, 17 April 2008.

41  Interview with international observer, Dili, 17 April 2008. 42  H. Beech, ‘East Timor’s broken promises’, Time (online), 8 March 2007, www.time.com/time/ magazine/article/0,9171,1597237,00. html, accessed 9 March 2007. 43  L. Williamson, ‘Ramos-Horta wants UN to stay in East Timor’, BBC News, 12 May 2008, news.bbc. co.uk/2/hi/asia-pacific/7397472.stm, accessed 13 May 2008. 6  Militias in the eastern DRC 1  See International Rescue Committee (2008) and UNHCR, ‘DR Congo: UNHCR fears military build-up could create massive displacement’, UNHCR News, 23 January 2009. 2  Fieldwork was conducted in North Kivu, South Kivu, Ituri and Kinshasa from January to February 2007. 3  For a discussion of the DDR process of groups in the Lusaka agreement, refer to Marriage (2009). 4  For a more elaborate discussion of the connection between DDR and SSR, refer to Bryden (2007). 5  Of the $200 million provided, $100 million was provided by the International Development ­Association/World Bank; and the remaining $100 million by the Multi-Country Demobilization and Re­integ­ration Programme (MDRP), which has been financed by a number of donors, including Belgium, Canada, France, Germany, the European Union, the United Kingdom, etc.). The MDRP provided an additional $18 million for child combatants, and $5 million for community projects (CONADER 2007). 6  The targeted groups included: the Front des Nationalistes Intégra­ tionnistes (FNI); the Union des

170

17  Interviews with FARDC official, Goma, 19 January 2007, and CONADER official, Goma, 18 January 2007. 18  Interview with European Communications Security and Evaluation Agency (EUSEC) official, Goma, 18 January 2007; focus group with former RCD-Goma combatants, Goma, 17 January 2007. 19  John Kanyunyu and Joe Bavier, ‘Congo rebel leader Nkunda arrested in Rwanda’, Reuters, 23 January 2009. 20  Ntaganda was formerly leading the Union des patriotes congolais-Lubanga (UPC-L) in the Ituri region, after Thomas Lubanga was arrested and taken to the ICC to face his criminal trial. Ntaganda is said to have very close relations with officials from the Ugandan government and army. Interviews with FARDC official, Bunia, 31 January 2007, and MONUC official, Bunia, 1 February 2007. Ntaganda, who is also known as the ‘Terminator’, has an International Criminal Court warrant issued against him for war crime charges (see Joe Bavier, ‘EU, rights groups call for Congo warlord’s arrest’, Reuters, 30 January 2009). 21  IRIN, ‘Kivus move closer to peace – but risks remain’, IRIN/ OCHA, 26 March 2009. 22  Historically, the Mayi-Mayi conception of warfare was defensive, and they were held together by a ­local ‘diviner’ (mganga), who provided fetishes so that when guns were fired, the bullet and/or the person possessing the fetish would turn into water. This use of the o ­ ccult in warfare was first used in the MayiMayi revolt against the German army in German East Africa, 1905–07 (see Gwassa 1972).

171

Notes

Patriotes Congolais (UPC); the Parti de l’Unité et la Sauvegarde de l’Intégrité du Congo (PUSIC); the Forces Armées du Peuple Congolais (FAPC); the Front Populaire pour la Démocratie au Congo (FPDC); and the Front de Résistance Populaire de l’Ituri (FRPI). 7  Currently, Thomas Lubanga (UPC), Germain Katanga (FRPI) and Mathieu Ngudojolo Chui (MRC) are in the custody of the ICC facing charges of war crimes and crimes against humanity; while Bosco Ntaganda (CNDP) remains a fugitive. 8  This amount had been decided upon by the World Bank and UNDP. 9  CONADER (2007); interview with Daniel Kwata, Coordinator General, CONADER, Kinshasa, 9 January 2007. 10  CONADER (2007: 15); interview with CONADER official, Kinshasa, 9 January 2007. 11  Interview with MONUC ­official, Goma, 15 January 2007. 12  CONADER (2007: 14); interview with CONADER official, Kinshasa, 10 January 2007. 13  Interview with MONUC DDRRR official, Kinshasa, 11 January 2007. 14  For an examination of other armed groups which began opera­ ting in the post-2003 period, and are currently undergoing demilitarization, refer to Thakur (2008a, 2008b). 15  The Mayi-Mayi have posed a serious security threat in the province of Katanga, especially northern Kat­ anga. It is reported that more than nineteen warlords command armed bands with a collective strength of 5,000–8,000 combatants. See Inter­ national Crisis Group (2006c). 16  IRIN, ‘Dissident general seeks troops’ integration into national army’, IRIN/OCHA, 2007.

23  Interview with FARDC official, Goma, 19 January 2007. 24  Interviews with FARDC official, Goma, 19 January 2007; MONUC official, Goma, 15 January 2007; FARDC, Bukavu, 25 January 2007; and MONUC official, Bukavu, 24 January 2007. 25  IRIN, ‘DRC: aid delivery under threat in Kivus’, 23 March 2010. 26  An example of how quickly alliances change: the FARDC and FDLR collaborated in 2007 to combat Kinya-rwandan armed collectives, including the CNDP. See United Nations (2008: 25). 27  For example, in some villages FDLR combatants are embedded within the civilian population, usually through marrying local women, but they intermittently take up arms for the FDLR. In other areas, it is reported that Rwandan refugees were held ‘hostage’ by the rebels, who compel them to remain outside of Rwanda as returning would result in punishment; although in other refugee areas, the local community claims the FDLR provides protection. Also, the FDLR reside in military camps in isolation, and it is in these areas that they are most violent, as killings, rapes and abductions are widespread. See MONUC, DDRRR Office, G2, FDLR in South Kivu, B ­ ukavu, 2007. Interviews with MONUC officials, Goma, 18 January 2007, and Bukavu, 27 January 2007; focus group with former FDLR combatants, DDRRR MONUC transit site, Goma, 19 January 2007. 28  Laure Pichegru, ‘Pursuing rebels at what price’, Inter Press Service (IPS), 17 June 2010. 29  Interview with former RCD-Goma combatant, Bukavu, 24 January 2007. 30  For example, Pierre Mulele

and Gaston Soumialot, with support from Laurent Kabila, led the first Mayi-Mayi rebellions against the government after Lumumba’s assassination in 1961. 31  International Crisis Group (2005b: 2–3); also, focus group with former FDLR combatants, DDRRR MONUC transit site, Goma, 19 January 2007. 32  Interviews with MONUC officials, Goma, 18 January 2007, and Bukavu, 27 January 2007. 33  Focus group with Mayi-Mayi ex-combatants, Goma, 17 January 2007, and Bukavu, 27 January 2007. 34  Focus group with former FDLR combatants, DDRRR MONUC transit site, Goma, 19 January 2007. The group was waiting to be repatriated (back to Rwanda) and reintegrated as part of the DDRRR process. 35  See International Alert (2010) and Global Witness (2009). 36  As one FDLR youth commented, ‘We were in ninja-mode, operating in the dark forests.’ Focus group with former FDLR com­batants, Masisi, North Kivu, 20 January 2007. 37  Focus group with former FDLR combatants, DDRRR MONUC transit site, Goma, 19 January 2007; interview with MONUC official, Goma, 16 January 2007. 38  Bemba is currently facing war crimes charges at the International Criminal Court. Agence FrancePress, 3 July 2008, ‘Former DRC rebel arrives in the Hague to face war crimes charges’. 39  Interview with MONUC ­official, Bukavu, 25 January 2007. 40  In addition, there are 681 military observers, 1,074 police personnel, 969 international civilian personnel, 2,154 local civilian staff

172

7 Afghanistan’s experience with militias 1  For a comprehensive review of DDR programmes in Afghanistan, see Bhatia and Muggah (2009). 2  Gossman (2009: 2). For further analysis of this dilemma, see also Torjesen (2009a). 3  For a comprehensive analysis of the re-emergence of the Taliban, see Giustozzi (2007). 4  M. Dearing and M. Dupee, ‘Mobilizing Afghan militias: civil defense forces vs. “tribal militias”’, Middle East Times, 16 February 2009. 5  Cited in G. Bruno, ‘Afghan defense chief unhappy with Obama

plan’, Council on Foreign Relations interview, 16 April 2009, www.cfr.org/ publications/19116/afghan_defense_ chief_unhappy_with_obama_plan. html, accessed 8 February 2010. 6  A. Mulrine, ‘US military to launch pilot program to recruit new local Afghan militias’, US News and World Report, 16 December 2008, www.usnews.com/articles/news/ iraq/2008/12/16/us-military-tolaunch-pilot-program-to-recruit-newlocal-afghan-militias_print.html, accessed 12 February 2010. 7  C. J. Radin, ‘The Afghan Public Protection Force is underway’, Long War Journal, 25 March 2009, www. longwarjournal.org/archves/2009/03/ afghan_public_protec-print.php, accessed 11 February 2010. 8  A. Gopal and Y. Dreazen, ‘Afghanistan enlists tribal militia forces’, Wall Street Journal Onlne, 12 August 2009, online.wsj.com/ article/SB124994313594220571. html#articleTabs=article. 9  J. Michaels, ‘Security plan looks to Afghan villages in fight against Taliban’, USA Today, 11 November 2009, www.usatoday.com/news/ world/2009-11-11-afghan-tribes_N. html, accessed 10 February 2010. 10  D. Filkins, ‘Afghan militias battle Taliban with aid of US’, New York Times, 22 November 2009, nytimes.com/2009/11/22/world/ asia/22militias.html, accessed 11 February 2010. 11  S. Ackerman, ‘Afghan human rights official criticizes McChrystal “tribes” initiative’, Washington Independent, 1 December 2009, washingtonindependent.com/69131/ afghan-human-rights-officialcriticizes-mcchrystal-tribes-initiative, accessed 12 February 2010. 12  D. Filkins, ‘Afghan militias battle Taliban with aid of US’.

173

Notes

and 606 UN volunteers. See MONUC (2009). 41  Interview with CONADER official, Kinshasa, 9 January 2007. 42  Focus group with former RCD-Goma, MLC and Mayi-Mayi combatants, Goma, 18 January 2007. 43  Interviews with CONADER official, Goma, 18 January 2007; EUSEC official, Goma, 18 January 2007; and MONUC official, Bukavu, 23 January 2007. 44  Interview with senior government official, Kinshasa, 15 July 2004. 45  Interview with senior government official, Kinshasa, 14 July 2004. 46  Interview with senior government official, Kinshasa, 5 July 2004. 47  Interview with Heile Menkeros, Deputy Special Representative of the Secretary-General (DSRSG), MONUC, Kinshasa, 6 February 2007. 48 Ibid. 49  Interview with MONUC officials, Goma, 15 January 2007, and Bukavu, 23 January 2007. 50  Interview with EUSEC official, Kinshasa, 12 January 2007. 51  Interview with EUSEC official, Goma, 18 January 2007.

13  J. Boone, ‘US pours millions into anti-Taliban militias in Afghanistan’, Guardian, 22 November 2009, www.guardian. co.uk/world/2009/nov/22/us-antitaliban-militias-afghanistan, accessed 12 February 2010. 14  D. Filkins, ‘Afghan militias battle Taliban with aid of US’. 15  H. Vogt, ‘Afghan tribe takes first step in anti-Taliban pact’, AP, 11 February 2010, www.msnbc.msn. com/id/35343170/ns/world_news_ south_and_central_asia/, accessed 11 February 2010. 16  A. Rubin, ‘Afghans to form local forces to fight Taliban’, New York Times, 14 July 2010, www. nytimes.com/2010/07/15/world/ asia/15afghan.html, accessed 3 August 2010. 17  R. Chandrasekaran and J. Partlow, ‘Karzai approves village defence forces’, Washington Post, 14 July 2010, voices.washingtonpost.com/ checkpoint-washington/2010/07/ karzai_approves_village_defens. html, accessed 3 August 2010. 18  Gopal and Dreazen, ‘Afghan­ istan enlists tribal militia forces’. 19  Boone, ‘US pours millions into anti-Taliban militias in Afghanistan’. 20  Gopal and Dreazen, ‘Afghan­ istan enlists tribal militia forces’. 21  Ackerman, ‘Afghan human rights official criticizes McChrystal “tribes” initiative’.

22  J. Partlow and G. Greg Jaffe, ‘US’s good intentions go awry in Afghan tribal area’, Washington Post, 15 May 2010, seattletimes.nwsource. com/html/nationworld/2011869228_ afghantribe16.html, accessed 5 August 2010. 23  G. Jaffe and R. Chan­ drasekaran, ‘US ambassador puts brakes on plan to utilize Afghan militias against Taliban’, Washington Post, 22 January 2010, www.washingtonpost.com/ wp-dyn/content/article/2010/01/21/ AR2010012101926_pf.html, accessed 11 February 2010. 24  D. Blatt, E. Long, B. Mul­ hern and M. Ploskunak, ‘Tribal engagement in Afghanistan’, Special Warfare, January/February 2009, p. 20. 25  D. Filkins, ‘In recruiting an Afghan militia, US faces a test’, New York Times, 15 April 2009, www. nytimes.com/2009/04/15/world/ asia/15afghan.html, accessed 15 February 2010. 26  Blatt et al., ‘Tribal engagement in Afghanistan’, p. 24. 27  G. Niazmand, ‘Afghan village armies fight Taliban’, AFP, 17 November 2009, http://www. google.com/hosted news/afp/article/ ALeqM5j4bcr4jJmoaSSo6XrH5Bdt Sf0c2w, accessed 11 February 2010. 28  F. Christia and M. Semple, ‘Flipping the Taliban: how to win in Afghanistan’, Foreign Affairs, July/ August 2009, p. 38.

174

Bibliography

Ahady, A. (1995) ‘The decline of Pashtuns in Afghanistan’, Asian Survey, 35(7), July. Alden, C. (2002) ‘Making old soldiers fade away: lessons from the reintegration of demobilised soldiers in Mozambique’, Security Dialogue, 33(3). Amnesty International (2007) Democratic Republic of Congo: Disarmament, Demobilization and Reintegration (DDR) and Reform of the Army, AFR 62/001/2007. Angstrom, J. (2008) ‘Inviting the levia­than: external forces, war and state-building in Afghan­ istan’, Small Wars and Insurgencies, 19(3). Arnold, M. B. (2008) ‘“Who is my friend, who is my enemy?” Youth and statebuilding in TimorLeste’, International Peacekeeping, 16(3), June. Arnold, M. B. and C. Alden (2007) ‘“This gun is our food”: disarming the White Army militias of South Sudan’, Conflict, Security and Development, 7(3). Assad, D. (2008) ‘The Peshmerga: militia or regular army?’, Niqash: Briefings from Inside and Across Iraq, 21 May, www.niqash.org/ content.php?contentTypeID= 75&id=2205&lang=0, accessed 12 August 2010. Ballentine, K. and H. Nitzschke (2003) ‘Beyond greed and grievances: policy lessons from studies in the political economy of armed conflict’, Working Paper,

New York: International Peace Institute. Ballentine, K. and J. Sherman (eds) (2003) The Political Economy of Armed Conflict: Beyond Greed and Grievance, Boulder, CO: Lynne Rienner. Bandura, A. (1990) ‘Mechanisms of moral disengagement’, in W. Reich (ed.), Origins of Terror­ ism: Psychologies, Ideologies, Theologies, States of Mind, Washington, DC: Woodrow Wilson Center Press. Bayart, J.-F. (1993) The State in Africa: Politics of the Belly, New York: Longman. — (2000) ‘Africa in the world: a history of extraversion’, African Affairs, 99(395). Beehner, L. (2006) ‘Iraq’s militia groups: backgrounder’, Council on Foreign Relations, 26 October, www.cfr.org/publication/11824/ iraqs_militia_groups.html, ­accessed 12 August 2010. Berdal, M. (1996) Disarmament and Demobilization after Civil War, Adelphi Paper 303, London: IISS. Berdal, M. and D. Malone (eds) (2000) Greed and Grievance: Economic Agendas in Civil Wars, Boulder, CO: Lynne Rienner. Berdal, M. and D. H. Ucko (2009) ‘Introduction: the political re­ integration of armed groups after war’, in M. Berdal and D. H. Ucko (eds), Reintegrating Armed Groups after Conflict: Politics, Violence and Transition, Abingdon: Routledge.

175

Bhatia, M. (2005) ‘Fighting words: naming terrorists, bandits, rebels and other violent actors’, Third World Quarterly, 26(1). Bhatia, M. and R. Muggah (2009) ‘The politics of demobilization in Afghanistan’, in R. Muggah (ed.), Security and Post-Conflict Reconstruction: Dealing with Fighters in the Aftermath of War, Abingdon: Routledge. Bhatia, M. and M. Sedra (2008) Afghanistan, Arms and Conflict: Armed Groups, Disarmament and Security in a Post-War Society, London: Routledge. Biddle, S. (2005/06) ‘Allies, airpower, and modern warfare: the Afghan model in Afghanistan and Iraq’, International Security, 30(3), Winter. Blatt, D., E. Long, B. Mulhern and M. Ploskunak (2009) ‘Tribal engagement in Afghanistan’, Special Warfare, January/February. Borzello, A. (2009) ‘The challenges of DDR in northern Uganda: the Lord’s Resistance Army’, in M. Berdal and D. H. Ucko (eds), Reintegrating Armed Groups after Conflict: Politics, Violence and Transition, Abindgon: Routledge. Bouta, T. (2005) Assessment of the Ituri Disarmament and Community Reinsertion Programme, Conflict Research Unit, Netherlands Institute of International Relations/ Clingendael Institute. Bruno, G. (2008) A Tribal Strategy for Afghanistan, Council on Foreign Relations Background Paper, 7 November. Bryden, A. (2007) ‘Understanding the DDR–SSR nexus: building sustainable peace in Africa’, Issue Paper presented at the Second International Conference on DDR and Stability in Africa, Kinshasa,

June, www.un.org/africa/osaa/ speeches/ddr-ssr.pdf, accessed 15 July 2009. Burchill, S. et al. (2001) Theories of International Relations, New York: Palgrave. Busumtwi-Sam, J. (2004) ‘Development and peacebuilding: conceptual and operational deficits in international assistance’, in T. M. Ali and R. O. Matthews (eds), Durable Peace: Challenges for Peacebuilding in Africa, Toronto: University of Toronto Press. Campbell, D. (1998) National Deconstruction: Violence, Identity and Justice in Bosnia, Minneapolis: University of Minnesota Press. Chabal, P. and J.-P. Daloz (1999) Africa Works: Disorder as Political Instrument, Oxford: James Currey. Chesterman, S. (2002) ‘East Timor in transition’, International Peacekeeping, 9(2), Spring. Chopra, J. (2002) ‘Building state failure in East Timor’, Development and Change, 33(5), November. Christia, F. and M. Semple (2009) ‘Flipping the Taliban: how to win in Afghanistan’, Foreign Affairs, July/August. Colletta, N. J. and R. Muggah (2009) ‘Context matters: interim stabilisation and second generation a ­ pproaches to security promotion’, Conflict, Security and Development, 9(4). Collier, P. and A. Hoeffler (1998) ‘On economic causes of civil war’, Oxford Economic Papers, 50(4), Oxford: Oxford University Press. — (2002) ‘Greed and grievance in civil war’, Working Paper, Centre for the Study of African Economies, University of Oxford. — (2004) ‘Greed and grievance in civil war’, Oxford Economic Papers,

176

ship in Afghanistan, 1978–97’, American Ethnologist, 25(4), November. Ellis, K. C. (1999) ‘Lebanon: the struggle of a small country in a regional context’, Arab Studies Quarterly, 21(1). Evans-Pritchard, E. E. (1940) The Nuer: A Description of the Modes of Livelihood and Political Institutions of a Nilotic People, Oxford: Oxford University Press. Ferguson, Y. H. and R. W. Mansbach (1996) Polities: Authority, Identities and Change, Columbia: University of South Carolina Press. Flint, J. (2009) ‘Beyond “Janjaweed”: understanding the militias of Darfur’, Small Arms Survey, June. Giustozzi, A. (2000) War, Politics and Society in Afghanistan: 1978–1992, Washington, DC: University Press. — (2007) Koran, Kalishnikov and Laptop: The Neo-Taliban Insurgency in Afghanistan, London: Hurst. — (2009) ‘Bureaucratic façade and political realities of disarmament and demobilisation in Afghan­ istan’, in M. Berdal and D. Ucko (eds), Reintegrating Armed Groups after Conflict: Politics, Violence and Transition, Abingdon: Routledge, p. 85. Global Witness (2009) War and the Militarisation of Mining in Eastern Congo, London. Gorjão, P. (2002) ‘The legacy and lessons of the United Nations transitional administration in East Timor’, Contemporary Southeast Asia, 24(2), August. Gossman, P. (2009) Transitional Justice and DDR: The Case of ­Afghanistan, Research Brief, International Centre for Transitional Justice, June. Gould, E. R. (2006) ‘Delegating IMF conditionality: understanding

177

Bibliography

56(4), Oxford: Oxford University Press. Collier, P., A. Hoeffler and D. Roher (2009) ‘Beyond greed and grievance: feasibility and civil war’, 61(1), Oxford Economic Papers, Oxford: Oxford University Press. CONADER (Commission Nationale pour la Démobilisation et la Ré­ insertion) (2007) Rapport Mensuel 2007, Kinshasa: CONADER. Cotton, J. (2007) ‘Timor-Leste and the discourse of state failure’, Australian Journal of International Affairs, 61(4), December. Council on Foreign Relations (2009) ‘Afghan defense chief unhappy with Obama plan’, Council on Foreign Relations interview, 16 April. Cramer, C. (2006) Civil War is Not a Stupid Thing, London: Hurst. Davis, D. E. (2009) ‘Non-state armed actors, new imagined communities, and shifting patterns of sovereignty and insecurity in the modern world’, Contemporary Security Policy, 30(2), August. De Soysa, I. (2002) ‘Paradise is a bazaar? Greed, creed and governance in civil war, 1989–99’, Journal of Peace Research, 39(4), July. De Waal, A. (2004) ‘Counterinsurgency on the cheap’, London Review of Books, 5 August. — (2007) ‘Deaths in Darfur: keeping ourselves honest’, SSRC Blog, 16 August, blogs.ssrc.org/ sudan/2007/08/16/deaths-indarfur-keeping-ourselves-honest/, accessed 12 August 2010. Duffield, M. (2007) Development, Security and Unending War: Governing the World of Peoples, Cambridge: Polity. Edwards, D. (1998) ‘Learning from the Swat Pathans: political leader-

variations in control and conforminty’, in D. G. Hawkins, D. A. Lake, D. L. Nielson and M. J. Tierney (eds), Delegation and Agency in International Organizations, Cambridge: Cambridge University Press. Guha-Sapirm, D. and O. Degomme (2005) ‘Darfur: counting the death, what are the trends’, WHO Collaborating Centre for Research on the Epidemiology of Disasters (CRED), Université catholique de Louvain, December. Gurr, T. (1970) Why Men Rebel, Princeton, NJ: Princeton University Press. — (1993) Minorities at Risk, Washington, DC: Institute of Peace. Gwassa, G. C. K. (1972) ‘African methods of warfare during the Maji Maji War’, in B. A. Ogot (ed.), War and Society in Africa: Ten Studies, London: Frank Cass. Herbst, J. (2000) ‘Economic incentives, natural resources, and conflict in Africa’, Journal of African Economies, 9(3). Hill, C. (2003) The Changing Politics of Foreign Policy, Houndsmills: Palgrave Macmillan. Hoebeke, H. et al. (2008) ‘Assessing security sector reform and its impact on the Kivu provinces’, Institute for Security Studies Situation Report, Institute for Security Studies, South Africa. Hoffman, B. and G. H. McCormick (2004) ‘Terrorism, signaling and suicide attack’, Studies in Conflict and Terrorism, 27. Homer-Dixon, T. (1999) Environment, Scarcity and Violence, Princeton, NJ: Princeton University Press. Hood, L. (2006) ‘Security sector reform in East Timor, 1999–2004’, International Peacekeeping, 13(1), March.

Horta, L. (2006) ‘East Timor riots expose a political divide’, Asia Times, 18 May. Human Rights Watch (2001) ‘Uganda in eastern DRC: fuelling political and ethnic strife’, 13(2). Indermaur, D. (1996) ‘Perceptions of violence’, Psychiatry, Psychology and the Law, 3(2). Institute for Security Studies (2004) The South Sudan Defence Force (SSDF): A Challenge to the Sudan Peace Process, Pretoria, April. International Alert (2010) The Role of the Exploitation of Natural ­Resources in Fuelling and Pro­ longing Crises in the Eastern DRC, London. International Crisis Group (2005a) The Congo’s Transition is Failing: Crisis in the Kivus, Africa Report no. 91, 30 March. — (2005b) Solving the FDLR Problem Once and for All, Africa Briefing no. 25, 12 May. — (2006a) Resolving Timor-Leste’s Crisis, Asia Report no. 120, 10 October. — (2006b) Security Sector Reform in the Congo, Africa Report no. 104, 13 February. — (2006c) Katanga: The Congo’s Forgotten Crisis, Africa Report no. 103, 9 January. — (2007a) Hizbollah and the Lebanese Crisis, Middle East Report no. 69, 10 October. — (2007b) Timor-Leste’s Parliamentary Elections, Asia Briefing no. 65, 13 June. — (2007c) Congo: Staying Engaged after the Elections, Africa Briefing no. 44, 9 January. — (2008a) Iraq after the Surge II: The Need for a New Political Strategy, Middle East Report no. 75, 30 April. — (2008b) Timor-Leste’s Displace-

178

and W. Wallace (eds), Non-State Actors in World Politics, Basingstoke: Palgrave. Kakeyi, S. (2010) ‘The progress of the Peshmerga forces and their role in post-2003 Iraq’, Kurdish Aspect, 24 June, www.kurdishaspect.com/ doc062410SK.html, accessed 12 August 2010. Kaldor, M. (2006) New and Old Wars: Organized Violence in a Global Era, Palo Alto, CA: Stanford University Press. Keen, D. (1997) ‘A rational kind of madness’, Oxford Development Studies, 25(1). — (1998) The Economic Functions of Violence in Civil Wars, Adelphi Paper 320, London: IISS. — (2002) ‘“Since I am a dog, beware my fangs”: beyond a “rational violence” framework in the Sierra Leonean war’, Working Paper no. 14, Crisis States Programme, DESTIN, London School of Economics. Kilcullen, D. (2009) The Accidental Guerilla, Oxford: Oxford University Press. Kingma, K. (2000) Demobilisation in Sub-Saharan Africa: The development and security impact, New York: St Martin’s Press. Kydd, A. H. and B. F. Walter (2006) ‘The strategies of terrorism’, International Security, 31(1). Leach, M. (2006) ‘History on the line: East Timorese history after independence’, History Workshop Journal, 61(1). Lemarchand, R. (1992) ‘Uncivil states and civil societies: how illusion became reality’, Journal of Modern African Studies, 30(2). — (2001) ‘Foreign policy making in the Great Lakes region’, in G. M. Khadiagala and T. Lyons (eds), African Foreign Policies: Power

179

Bibliography

ment Crisis, Asia Report no. 148, 31 March. — (2008c) Timor-Leste: Security Sector Reform, Asia Report no. 143, 17 January. — (2009a) Jonglei’s Tribal Conflicts: Countering Insecurity in South Sudan, Africa Report no. 154, 23 December. — (2009b) Handing Back Respon­ sibility to Timor-Leste’s Police, Asia Report no. 180, 3 December. — (2009c) Timor-Leste: No Time for Complacency, Asia Update Briefing no. 87, 9 February. International Rescue Committee (2008) Mortality in the Democratic Republic of Congo: An Ongoing Crisis. Iqbal, Z. and H. Starr (2008) ‘Bad neighbors: failed states and their consequences’, Conflict Management and Peace Science, 25(4), September. Jackson, S. (2006) ‘Sons of which soil? The language and politics of auto­ chthony in eastern D.R. Congo’, African Studies Review, 49(2). Jepperson, R. L., A. Wendt and P. J. Katzenstein (1996) ‘Norms, identity and culture in national security’, in P. J. Katzenstein (ed.), The Culture of National Security: Norms and Identity in World Politics, New York: Columbia University Press. Jett, D. C. (2001) Why Peacekeeping Fails, New York: Palgrave ­Macmillan. Johnson, D. (2004) The Root Causes of Sudan’s Civil Wars, Oxford: James Currey. Jones, S. (2006) ‘Averting failure in Afghanistan’, Survival, 48(1), Spring. Josselin, D. and W. Wallace (2001) ‘Non-state actors in world politics: a framework’, in D. Josselin

and Process, Boulder, CO: Lynne Rienner. — (2003) ‘The Democratic Republic of the Congo: from failure to potential reconstruction’, in R. I. Rotberg (ed.), State Failure and State Weakness in a Time of Terror, Cambridge, MA: World Peace Foundation. Lusaka Ceasefire Agreement (1999) www.iss.co.za/af/profiles/ drcongo/cdreader/bin/2lusaka. pdf, accessed 5 July 2010. Lyon, D. (2009) In Afghanistan: Two Hundred Years of British, Russian and American Occupation, New York: Palgrave Macmillan. Lyons, T. (2002) ‘The role of post­settlement elections’, in S. J. Stedman, D. Rothchild and E. M. Cousens (eds), Ending Civil Wars: The Implementation of Peace Agreements, Boulder, CO: Lynne Rienner. Mamdani, M. (1996) Citizen and Subject: Contemporary Africa and the Legacy of Late Colonialism, Princeton, NJ: Princeton University Press. Marriage, Z. (2009) ‘Flip-flop rebel, dollar soldier: demobilisation in the Democratic Republic of Congo’, in M. Berdal and D. H. Ucko (eds), Reintegrating Armed Groups after Conflict: Politics, Violence and Transition, Abingdon: Routledge. Martin, I. and A. Mayer-Rieckh (2005) ‘The United Nations and East Timor: from self-determination to state-building’, International Peacekeeping, 12(1), Spring. Mishal, S. and M. Rosenthal (2005) ‘Al-Qaeda as a dune organization: toward a typology of Islamic terrorist organizations’, Studies in Conflict and Terrorism, 28(4), July. MONUC (2009) ‘Facts and figures’, www.un.org/Depts/dpko/­

missions/monuc/facts.html, accessed 20 July 2009. Muggah, R. (2009) ‘Introduction: the emperor’s clothes’, in R. Muggah (ed.), Security and Post-Conflict Reconstruction: Dealing with Fighters in the Aftermath of War, Abingdon: Routledge. Nevins, J. (2007) ‘Timor-Leste in 2006: the end of the postindependence honeymoon’, Asian Survey, 47(1), January/February. O’Ballance, E. (1993) Afghan Wars 1839–1992: What Britain Gave Up and the Soviet Union Lost, London: Brassey’s. — (2000) Sudan, Civil War and Terrorism, 1956–99, New York: St Martins Press. Onana, R. and H. Taylor (2008) ‘­MONUC and SSR in the Demo­cratic Republic of Congo’, International Peacekeeping, 15(3). Ortiz, R. D. (2002) ‘Insurgent strategies in the post-Cold War: the case of Revolutionary Armed Forces of Colombia’, Studies in Conflict and Terrorism, 25(2). Peake, G. (2009) ‘What the Timorese veterans say: unpacking DDR in Timor-Leste’, in R. Muggah (ed.), Security and Post-Conflict Reconstruction: Dealing with Fighters in the Aftermath of War, Abingdon: Routledge. Philpott, S. (2006) ‘East Timor’s double life: smells like Westphalian spirit’, Third World Quarterly, 27(1), February. Policzer, P. (2002) ‘Human rights and armed groups: toward a new policy architecture’, Working ­Paper no. 1, Armed Groups Project, University of Calgary, Alberta. Porto, J. G. (2002), ‘Contemporary conflict analysis in perspective’, in J. Lind and K. Sturman (eds), Scarcity and Surfeit: The Ecology of

180

‘Liberal peacebuilding in TimorLeste: the emperor’s new clothes’, International Peacekeeping, 15(2), April. Risse-Kappan, T. (1995) ‘Bring transnational relations back in: introduction’, in T. Risse-Kappan (ed.), Bringing Transnational Relations Back In: Non-State Actors, Domestic Structures and International Institutions, Cambridge: Cambridge University Press. Rogier, E. (2005) No More Hills Ahead? The Sudan’s Tortuous Ascent to Heights of Peace, Clingendael Security Paper no. 1, The Hague: Netherlands Institute of International Relations, August. Rubin, B. (2002) The Fragmentation of Afghanistan: State Formation and Collapse in the International System, 2nd edn, New Haven, CT: Yale University Press. Saikal, A. (2006) ‘Afghanistan’s transition: ISAF’s stabilization role?’, Third World Quarterly, 27(3). Schmidt, V. A. (2005) ‘Institutionalism and the state’, in C. Hay, D. Marsh and M. Lister (eds), The State: Theories and Issues, Basingstoke: Palgrave. Shaw, T. (2003) ‘Regional dimensions of conflict and peace-building in contemporary Africa’, Journal of International Development, 15(4), May. Shultz, R. H. and A. J. Dew (2006) Insurgents, Terrorists and Militias: The Warriors of Contemporary Combat, New York: Columbia University Press. Simon, H. A. (1984) Models of Bounded Rationality, vol. 2, Cambridge, MA: MIT Press. Sisk, T. D. (2003) ‘Democratization and peacebuilding: perils and promises’, in C. A. Crocker, F. O. Hampson and P. Aall (eds),

181

Bibliography

African Conflicts, Pretoria: Institute of Security Studies. Prendergast, J. and E. Plumb (2002) ‘Building local capacity: from implementation to peacebuilding’, in S. J. Stedman, D. Rothchild and E. M. Cousens (eds), Ending Civil Wars: The Implementation of Peace Agreements, Boulder, CO: Lynne Rienner. Prunier, G. (1995) The Rwanda Crisis, 1959–1994: History of a Genocide, Kamala: Fountain Publishers. — (2009) Africa’s World War: Congo, the Rwandan Genocide and the Making of a Continental Catas­ trophe, Oxford: Oxford University Press. Pugh, M. and N. Cooper with J. Goodhand (2004) War Economies in a Regional Context: Challenges of Transformation, Boulder, CO: Lynne Rienner. Rashid, A. (2008) Descent into Chaos: The United States and the Failure of Nation Building in Pakistan, Afghanistan and Central Asia, London: Viking. République Démocratique du Congo and Programme des Nations Unies pour le Développement (2008) Rapport National sur le Développement Humain 2008: Restauration de la Paix et Reconstruction, www.undp. org.cd/downloads/PNUD%20 Rapport%20National%20sur%20 le%20d%C3%A9veloppement%20 Humain.pdf, accessed 27 December 2009. Restrepo, J. A. and R. Muggah (2009) ‘Colombia’s quiet demobilization: a security dividend?’, in R. Muggah (ed.), Security and Post-Conflict Reconstruction: Dealing with Fighters in the Aftermath of War, Abingdon: Routledge. Richmond, O. and J. Franks (2008)

Turbulent Peace: The Challenges of Managing International Conflict, Washington, DC: US Institute of Peace Press. Small Arms Survey (2008) Allies and Defectors: An Update on Armed Group Integration and Proxy Force Activity, Sudan Issues Brief no. 11, Geneva, May. Spear, J. (2002) ‘Disarmament and demobilization’, in S. J. Stedman, D. Rothchild and E. M. Cousens (eds), Ending Civil Wars: The Implementation of Peace Agreements, Boulder, CO: Lynne Rienner. Stedman, S. J. (1997) ‘Spoiler problems in peace processes’, International Security, 22(2), Autumn. Steele, J. (2002) ‘Nation building in East Timor’, World Policy Journal, 19(2), Summer. Stewart, F. (2002) ‘Horizontal inequalities as a source of conflict’, in F. O. Hampson and D. M. Malone (eds), From Reaction to Conflict Prevention: Opportunities for the UN System, Boulder, CO: Lynne Rienner. Tariq, M. (2008) ‘Tribal security system (arbakai) in southeast Afghanistan’, Occasional Paper no. 7, LSE Crisis States Research Centre, December. Thakur, M. (2005) ‘Ugandan foreign policy in the DRC conflict and Great Lakes region of Africa’, Unpublished PhD thesis, Department of International Relations, London School of Economics and Political Science. — (2008a) ‘Demilitarising militias in the Kivus’, African Security Review, 17(1). — (2008b) ‘Elusive peace in the DR Congo: armed groups and ­militias in the Kivus and Ituri’, Paper presented at the

International Studies Association Annual Conference, March, San Francisco, CA. Thakur, M. and R. Singh (2006) ‘God’s warriors and Lucifer’s army: a comparative analysis of “terrorist” organizations and African militias’, Paper presented at the International Studies Association Annual Conference, March, San Diego, CA. Thompson, E. P. (1963) The Making of the English Working Class, New York: Vintage. Tilly, C. (1990) Coercion, Capital and the European State, AD 990–1990, Cambridge: Basil Blackwell. Timor-Leste Armed Violence Assessment (2009) Groups, Gangs, and Armed Violence in Timor-Leste, Issue Brief no. 2, April. Torjesen, S. (2009a) ‘The political economy of disarmament, demobilisation and reintegration (DDR)’, Working Paper no. 709, Nor­wegian Institute of Inter­ national Affairs, Oslo. — (2009b) ‘New avenues for research in the study of DDR’, Conflict, Security and Development, 9(4). Torjesen, S. and N. MacFarlane (2009) ‘Reintegration before ­disarmament: the case of post conflict reintegration in Tajikistan’, in M. Berdal and D. H. Ucko (eds), Reintegrating Armed Groups after Conflict: Politics, Violence and Transition, Abingdon: Routledge. Ucko, D. H. (2009) ‘Militias, tribes and insurgents: the challenge of political reintegration in Iraq’, in M. Berdal and D. H. Ucko (eds), Reintegrating Armed Groups after Conflict: Politics, Violence and Transition, Abingdon: Routledge. UK Prime Minister’s Office (2007) ‘Statement on Afghanistan’,

182

Wendt, A. (1987) ‘The agent-structure debate in international relations’, International Organization, 21(3), Summer. — (1992) ‘Anarchy is what states make of it: the social construction of power politics’, International Organization, 46(2), Spring. — (1994) ‘Collective identity formation and the international state’, American Political Science Review, 88(2). — (1996) ‘Identity and structural change in international politics’, in Y. Lapid and F. Kratochwil (eds), Return of Culture and Identity in IR Theory, Boulder, CO: Lynne Rienner. — (1999) Social Theory of International Politics, Cambridge: Cambridge University Press. Young, J. (2006) The South Sudan Defence Forces in the Wake of the Juba Declaration, Sudan Human Security Baseline Assessment Project Working Paper no. 1, Geneva: Small Arms Survey. — (2007) The White Army: An Introduction and Overview, Sudan Human Security Baseline Assessment Working Paper no. 5, Geneva: Small Arms Survey. Zahara, M.-J. (2002) ‘Peace by unconventional means: Lebanon’s Ta’if Agreement’, in S. J. Stedman, D. Rothchild and E. M. Cousens (eds), Ending Civil Wars: The Implementation of Peace Agreements, Boulder, CO: Lynne Rienner. Zartman, I. W. (1989) Ripe for Resolution: Conflict and Intervention in Africa, Oxford: Oxford University Press.

183

Bibliography

7 December, www.number10.gov. uk/Page14050, accessed 8 February 2010. United Nations (2006a) Report of the Independent Special Commission of Inquiry for Timor-Leste, Geneva, 2 October. — (2006b) Report of the United Nations Expert Panel on the Arms Embargo in the Democratic Republic of Congo, S/2006/53. — (2006c) Integrated Disarmament, Demobilisation and Reintegration Standards (IDDRS), New York: United Nations. — (2008) Final Report of Group of Experts on the Democratic Republic of Congo, S/2008/773. — (2009) Twenty-seventh Report of the Secretary-General on the United Nations Organization Mission in the Democratic Republic of Congo, S/2009/160. US Department of State (2010), Background Note: Timor-Leste, 12 February, www.state.gov/r/ pa/ei/bgn/35878.htm, accessed 12 August 2010. Uvin, P. (2001) ‘Difficult choices in the post-conflict agenda: the international community in Rwanda after the genocide’, Third World Quarterly, 22(2), April. Vlassenroot, K. and T. Raey­maekers (2004) ‘The politics of rebellion and intervention in Ituri: the emergence of a new p ­ olitical complex?’, African ­Affairs, 103(412). Walzer, M. (1996) Thick and Thin: Moral Argument at Home and Abroad, Notre Dame, IN: University of Notre Dame Press.

Index

106, Colonel, 115 Abdul, Colonel, 115 ‘accidental guerrillas’, 147 accountability of militias, 32–3 Adar, oil production in, 59–60 Afghan Military Forces (AMF), 15, 131, 133, 148 Afghan National Army (ANA), 132, 138, 148 Afghan National Auxiliary Police (ANAP), 140, 143 Afghan National Police (ANP), 132, 138, 148 Afghan New Beginnings Programme (ANBP), 132–3, 134 Afghan Public Protection Programme (AP3), 140–2, 146 Afghan Social Outreach Programme (ASOP), 139 Afghanistan, x, xi, 28, 36, 39, 41, 109, 152; militias in, 128–49 African Development Bank, 109 African Union (AU), 39 Ajello, Aldo, 157 Akatomba, Captain, 113–14 Akobo county (Sudan), 75–7, 83; as disarmament point, 81 Akol, Lam, 43, 55 Alkatiri, Mari, 89, 90, 92–3, 95, 97, 98, 105 Alliance des forces démocratiques pour la libération du Congo (ADLF), 121 Alliance of Majority in Parliament (AMP) (Timor-Leste), 91, 99 Allied Democratic Forces–National Army for the Liberation of Uganda (ADF-NALU), 115 Amani Leo operation (DRC), 116

Amani Pact (DRC), 113, 122 Amin, Hafizullah, 128 amnesty, 13, 108 Ampatuan clan, 12 Amum, Pagan, 61 analytical framework, 25–34 Angola, 18, 21 Anuak people, 81 Anyanya II (Sudan), 43, 45, 53 arbakai institution (Afghanistan), 137, 142–3, 144, 145 Arkan’s Tigers (former Yugoslavia), 14 arms trafficking, 29 Atmar, Hanif, 141 Atta, Mohammad, 131 Australia, 105 Australian Defence Force, 90 Bagosora, Théoneste, 7 Ballentine, Karen, 153 bandits, roving and stationary, 155 Banyamulenge ethnic group, 118 Banyarwanda ethnic group, 118 Baraka, Colonel, 114 Bashir, Omar Hassan al-, 7 Bemba, Jean-Pierre, 119 blood feuds, 79, 80 Bolivarian militia (Venezuela), 13 brassage, 111; centres de brassage, 110, 125 Brown, Gordon, 137 Cambodia, 21 cantonment, 94 capacity-building, 151 cattle raiding, 68, 69, 70–1, 75, 79–80, 82 Centre for Humanitarian Dialogue, 96 184

context, importance of, 154–5, 162 co-opting of insurgents, 148–9 corruption, 141 criminal activities of militias, 2, 115 Croatian Defence Forces, 14 Darfur, 7, 8, 45, 84 decision-making in militias, structure of, 157 demilitarization, ix, 13, 16–18, 40, 65–86; contextualization of, 1–19 (economic context, 36–7; historical context, 34–5; models of, 34–41; political context, 35–6; regional dynamics of, 38–9); in Afghanistan, since 2001, 131–6; in DRC, 107–27 passim; problems facing, 127; rigid, avoidance of, 159 demobilization, emphasis placed on, 14 democracy, creation of, 35 Democratic Party (Timor-Leste), 91, 99 Democratic Republic of Congo (DRC), ix, 28, 35, 36, 38, 39, 41, 107–27; death toll in, 107 Deng, Gatlauk, 44 diasporas, influence on resurgent conflict, 25 Dinka people, 56–7, 66, 71, 73, 84 disappearance and retreat, 31 disarmament: challenges to, 79–83; incentives for, 74; meaning of, 73; rolling wave of, 81–3, 85; universal, 75, 81; voluntary, 77, 111 disarmament, demobilization and reintegration (DDR), x–xi, 2, 17, 20, 37, 40, 42, 109, 111, 112, 126, 137, 148, 150; community involvement in, 111, 153; complexity of, xi; greater flexibility in, 159–60; in Afghanistan, 132–3; operations of, 15; resistance to, 1 disarmament, demobilization, repatriation, reintegration or

185

Index

Chávez, Hugo, 13 Chetniks, 13–14 children, recruitment of, 118 Citizen Armed Forces Geographical Units (Philippines), 11 civil society, role of, 40–1 Civilian Volunteers Organizations (Philippines), 11 civilians, attacks on, 30 cocaine industry, in Colombia, 12–13 Cold War, 16, 22, 23, 32, 130 Collier, Paul, 23 Colombia, militias in, 12–13 Comité de Gestion des Fonds de Désarmement, Démobilisation et Réinsertion (CGFDR) (DRC), 110 Comité Inter-Ministériel chargé de la Conception et de l’Orientation en matière de Désarmement, Démobilisation et Réinsertion (DRC), 109–10 commercial activities of militias, 1, 53–5, 155 Commission Nationale pour la Démobilisation et la Réinsertion (CONADER) (DRC), 37, 109, 111, 112, 123, 124, 160 community involvement in DDR, 73–5, 111, 153 community militias, 128, 138–9, 143, 146, 148 community security, 137, 140–1 community watch systems, 140, 153 COMPASS Security company, 143 compensation: for cattle loss, 80; for killings, 79; for reinsertion, 112; for surrendered weapons, 68 Comprehensive Peace Agreement (CPA) (Sudan), 42, 44, 48–9, 52, 62, 63–4, 82, 152 Congrès national pour la défense du peuple (CNDP) (DRC), xii, 8, 108, 110, 113, 114, 116, 117, 118, 121, 126, 152 Conseil national pour la défense de la démocratie – Forces pour la défense de la démocratie (CNDDFDD) (DRC), 115

resettlement (DDRRR), 112, 115, 123 Disarmament and Community Reinsertion Programme (DCF) (DRC), 111 Disbandment of Illegal Armed Groups (DIAG) (Afghanistan), 132, 133, 134 disempowerment, 147 Dostum, Abdul Rashid, 129, 130, 131, 133, 139, 148, 152 drug trafficking, 132 Dunia, Colonel, 115 East Timor see Timor-Leste elections, 142; importance of, 35 empirical mapping of militias, 20, 150; in DRC, 113–19; in TimorLeste, 92–7; of White Army, 68–7 empowerment, 153 enfants associés aux forces et aux groupes armés (EAFGA) (DRC), 112 Equatorian Defence Force (EDF), 57 ethical yardstick, reassertion of, 161 Ethiopia, relations with Sudan, 80–1 ethnicity, 23, 116–17, 158; invention of, 38; role of, 56 extraversion, 22 Fahim, Qasim, 131, 133, 148 FALANTIL see National Liberation Forces of an Independent East Timor Fashoda Agreement, 43, 54 Force combattant abachunguzi (FOCA) (DRC), 115 Forces armées congolaises (FAC), 111 Forces armées de la République Démocratique du Congo (FARDC), 108–9, 110, 113, 114, 116, 124, 125, 152; criminal activities of, 126, 160 Forces armées rwandaises (FAR), 115 Forces démocratiques pour la libération du Rwanda (FDLR), xii, 8, 108, 112, 113, 115–16, 117, 118, 121, 122, 126, 152 Fountaine, General La, 115

France, militias in, 6 FRETILIN party (Timor-Leste), 90, 91, 97, 98, 99, 100, 102 Front de libération nationale (FLN) (Burundi), 115 Front de résistance populaire de l’Ituri (FRPI) (DRC), 111 Gambellan Peoples Liberation Movement (GPLM) (Sudan), 81 Garang, John, 43, 49, 52–3 Gates, Robert M., 137, 139 genocide, 8 see also Rwanda, genocide in Ghund-i Qawmi (Afghanistan), 129 Global and Inclusive Agreement (DRC), 121 government, role of, in disarmament, 75–6 Government of Southern Sudan (GoSS), 47, 48, 49, 50–3, 75, 83; overall strategy of, 61 Government of Sudan (GoS), 43, 57–9 ‘greed and grievance’ school of thought, 1, 23, 25, 27–8, 156, 157 Green Berets (Bosnian), 14 group dynamics of militias, 24 Group of 47 (DRC), 113 Guevara, Ernesto ‘Che’, 33 guns: AK-47, 71; ownership of, 71 (rationale for, 69–70) Gurr, Ted, 23 Gusmão, Xanana, 88, 90, 91, 92–3, 94, 95–6, 97, 98, 99, 100, 102, 104, 158 Guterres, Francisco ‘Lu-Olo’, 91 Habyarimana, Juvenal, 7 Hezbollah (Lebanon), 9 Hizbi Islami (Afghanistan), 130 Hizbi Wahdat (Afghanistan), 130 Ho Chi Minh, 22 Hoeffler, Anke, 23 Hutu people, 7, 8, 38, 115, 118, 121 identity: not essentialist, 37–8; politicization of, 20, 28, 29, 30, 32, 37–8, 72–3, 109, 160

186

Jackson, Yemeni, 114 Jamiat-e Islami (Afghanistan), 131 Janjaweed militias (Sudan), 7, 45 jirgas (Afghanistan), 145 Joint Integrated Units (JIU) (Sudan), 52 Jonglei state, 42, 44, 46, 49, 53, 62, 65–86 passim, 159 Jowzyani militia (Afghanistan), 129, 152 Juba Declaration (Sudan), 44, 45, 48, 49, 50–3 passim, 55, 57, 60, 61, 62, 64 Justice and Equality Movement (Darfur), 7

Kabila, Joseph, 119, 121 Kabila, Laurent, 120, 121 Kagame, Paul, 8, 117 Kandak Auxiliary Uruzgan (Afghanistan), 143 Karmal, Babrak, 128 Karzai, Hamid, 131, 133, 136, 142, 143, 149 Katanga, 113 Khan, Ismael, 139 Khan, Jan Mohammad, 143 Khartoum Agreement, 43–4, 49, 54, 60 Khmer Rouge, 17, 18, 154 Kiir, Salva, 44, 51, 61 Kivus (DRC), 107–27 Kong, Gordon, 44, 46, 63 Kony, Joseph, 8 Konyi, Ismail, 46, 55, 57, 58, 62, 83 Kurdish Democratic Party, 10 lashkars, 12 leadership models, 157 Lebanon, 9 legitimacy of militias, 32–3, 156–7 Liquiçá Church Massacre (East Timor), 11 local communities, role of, in disarmament, 73–5 Local Defence Initiative (LDI) (Afghanistan), 140–1, 145 Local Police Force (Afghanistan), 144, 146 local/stranger duality, 119 London School of Economics and Political Science (LSE), ix longe durée perspective, 35 looting, 29, 30, 115 Lord’s Resistance Army (LRA), 8, 115 Lubanga, Thomas, 122 Mabior, Thomas, 46, 58 Macapagal-Arroyo, Gloria, 12 Machar, Riek, 43–4 Mahdi Army militia, 9, 10 Malaita Eagle Force (Solomon Islands), 11 Malakal (Sudan), 46–7, 60

187

Index

identity politics, and SSDF, 56–7 impunity, 29, 125, 126, 127, 160 impuzamugambi (Rwanda), 7 Independent Directorate for the Protection of Public Properties and Highways by Tribal Support (IDPPPH) (Afghanistan), 143 Indonesia, occupation of East Timor, 10–11 institutionalization levels of militias, 31–2 Inter-Congolese Dialogue, 114 Inter-Governmental Authority on Development (IGAD), 39 interahamwe (Rwanda), 7, 8, 115, 152 international community, role of, 39–40 International Criminal Court (ICC), 7, 111, 114, 122 International Crisis Group, 98, 102, 104 International Security Assistance Force (ISAF), 135, 136 International Stabilization Force (ISF), 90, 96–7, 105 Iraq: armed groups transform into political parties, 36; militias in, 9–10 Isatabu Freedom Movement (Solomon Islands), 11 Israel, attacks on, 9 Ituri (DRC), 107–27

malek traditional leaders (Afghanistan), 147 Mamdani, M., 38 Mao Zedong, 33 Marxism-Leninism, 32 Massoud, Ahmad, 131 Matip, Paulino, 44, 46, 51, 61, 63 Mayi-Mayi (DRC), xii, 8, 110, 113, 114–15, 117, 118–19, 121, 152, 158 McChrystal, Stanley, 141, 145 McKiernan, David, 138 McNeil, Dan, 137 milice française, 6 Milishia-i Sahard (Afghanistan), 129 military capabilities analysis, 25 militias: as rational players, 25; assymmetric weakness of, 118; attractiveness of, 136–8; conceptualization of, 20–41, 150 (model for, 25–34); contemporary examples of, 7–14; contextualization of, 1–19; definition of, 3–4 (behaviourist approach, 23; ideological approach, 22; systems approach, 23–4); dynamic organizations, 158; historical examples of, 4–7; in Afghanistan, 128–49; legitimacy and accountability of, 26; motivations of see motivations of militias; nocturnal strategy of, 29; organization of, 26; rationality of, 26–9; relations with other actors, 33–4; seasonal, 139; strategies of, 118; structures of, 31–2; tactics and strategies of, 26, 29–30; typology of, xii, 3; understanding of, x, 2, 4 see also community militias mining activities, illegal, 2 Mission de l’Organisation des Nations Unies pour la stabilisation en République Démocratique du Congo (MONUSCO), 123 Mission des Nations Unies en République Démocratique du

Congo (MONUC), 108, 37, 110, 112, 113, 116, 117, 122–6, 160 mixage process, 110 Mobutu, Sese Seko, 107, 120 Moramvia Group, 113 motivations of militias, 25, 26–9, 48–50, 68–9, 155–6 (conceptualization of, 28; of personal enrichment, 116) Mouvement pour la libération du Congo (MLC), 111, 121 Mouvement révolutionnaire du Congo (MRC), 111, 122 Mozambican National Resistance (Renamo), 17, 157 Mozambique, 21 Muggah, R., 16 mujahedin, 128–49 Multi-Country Demobilization and Reintegration Programme (MDRP), 112, 115 Murle people, 55, 71, 81, 82, 84; disarmament of, 83 Mutebusi, Jules, 113 mutineers of Timor-Leste, xii, 87–106, 158 Najibullah, Mohammad, 130, 131 narcotics trading, 2 National Congress for Timorese Re­ construction (CNRT), 91, 98, 99 National Congress Party (NCP) (Sudan), 48, 60–1 National Liberation Forces of an Independent East Timor (FALANTIL), 88, 92, 93, 101 National Police of Timor-Leste (PNTL), 89, 96, 100, 101, 103, 105; disintegration of, 104 National Union for the Total Independence of Angola (UNITA), 17, 18 natural resources, 59; management of, 37 networks: chain networks, 31; multichannel, 31; of militias, 30, 31; role of, 157 New Zealand Defence Force, 90

188

Pibor Defence Force (PDF) (Sudan), 57, 58, 83 police forces, 85; development of, 82–3 Policzer, Pablo, 23–4 post-conflict contexts, xi, 2, 36, 108, 125, 150, 160; definition of, 127; reconstruction in, 39 poverty, 160 principal-agent theory, 24 Programme Implementation Unit (PIU) (DRC), 109 Programme National pour le Désarmement, la Démobilisation et la Réinsertion (PNDDR) (DRC), 109 Pronk, Jan, 57–8 prostitution, 2 Qaeda, al-, 9, 39, 131

Obama, Barack, 138 oil companies, relations with SSDF, 59–60 oil production, 59; in South Sudan, 53 Organization of American States (OAS), 39 Oromo Liberation Front (OLF) (Sudan), 81 Pakistan, militias in, 12 Palestinian militias, 9 Pashtuns, 129–30, 145 patrimonialism, 120 Patriotic League (Bosnian), 14 Patriotic Union of Kurdistan (PUK), 10 patronage, 41 peace, definition of, 127 Peace and Reconciliation Commission (Afghanistan), 149 Peshmarga militia, 10 petitioners (Timor-Leste), 88, 92–7; categorization of, 94; subsidence of movement, 106 Petraeus, David, 141–2 Petrodar oil company, 59 Philippines, militias encouraged by government, 11–12

Rabbani, Burhanuddin, 135 Ramos-Horta, José, 90, 91–2, 93, 95, 96, 99, 106, 158; evacuation of, 105 rape, 115 Rassemblement congolais pour la démocratie (RCD) (DRC), 121 Rassemblement congolais pour la démocratie-Goma (RCD-Goma) (DRC), 111, 113, 118 rationality of militias, 41, 155–6; two models of, 27 recognition, of actors, importance of, 156–7 reconciliation programmes, 149 Reinado, Alfredo, 89, 90–1, 93, 95–7, 99; escape from jail, 105 reintegration, 35, 37, 123–4, 154; support programmes, 15 remilitarization strategy, in Afghanistan, 128, 136–47, 148, 152; seen as counterproductive, 144 resistance to disarmament, rationale for, 70–2 Revolutionary Army of Colombia (FARC), 12

189

Index

Ngojolo, Mathieu, 122 Nile Hope and Development Forum (NHDF), 77 Nkunda, Laurent, 110, 113, 117 non-governmental organizations (NGOs), role of, 84 (in disarmament, 77–8) Northern Alliance (Afghanistan), 15, 131, 133, 135, 143, 148 Norwegian Institute of International Affairs (NUPI), ix Ntaganda, Bosco, 114, 117 Ntaryamira, Cyprien, 7 Nuer people, 55, 56–7, 69, 73, 79; Jikany Nuer, 71, 85; Lou Nuer, 66, 70–1, 74, 75, 81, 82, 84, 85 (compensation method for homicide, 76) Nurzai, Aref Khan, 143–4

Revolutionary United Front (RUF) (Sierra Leone), 24 Ruhara, Kalala, 115 Rwanda, 7–8, 114, 117; genocide in, 115, 120–1, 152; militias in, 18 Rwandan Patriot Front (RPF), 8 Sadr, Muqtada al-, 10 Sadrist Movement, 10 Salsinha, Gastão, 88, 93, 95, 97; prison sentence of, 106 Savimbi, Jonas, 22 Scorpions (former Yugoslavia), 14 security, local, xi, 3, 160–1, 162; search for, 150–62 security sector reform (SSR), 15–16, 108, 109, 110, 122–6, 127, 159, 160–1 self-defence, emphasis on, 4, 151, 152–3 Self-Defence Groups (Afghanistan), 129 selfhood, construction of, 37 Sherman, Jake, 153 Shinwari tribe (Afghanistan), 145 shuras, formation of (Afghanistan), 139 Sierra Leone, 24, 161 smuggling, by militias, 30 Social Democratic Party (TimorLeste), 91 Solomon Islands, militias in, 11 Somalia, 21 Sons of Iraq programme, 136, 146 South African Development Community (SADC), 39 South Sudan, 15, 40, 41, 65–86, 151–2, 157, 159 South Sudan Defence Force (SSDF), xi, 42–64, 79, 151–2, 158; and identity politics, 56–7; commercial and financial interests of, 53–5; financial interests of, 54–5; hold-out activities of, 50–3, 55; motivations of, 48–50; operational mode of, 45–8; origins of, 43–5; relations with external actors, 57–60;

relations with oil companies, 59–60 South Sudan Independence Movement (SSIM), 43 South Sudan Police Service, 85 South West Africa People’s Organization (Swapo), 18 spoilers of peace deals, 2, 21–5, 28, 119, 155, 157; definition of, 21 Srebrenica massacre, 14 state: capacity of, 36; endorsement of militias, 142; negotiation with armed elements, 28; rogue states, 25; role of, 3, 24 (in socializing militias, 158); weak, 32 (Congolese, 108, 119–22, 127) state-building, 136, 138, 146 Stedman, Steve, 21 Structure Militaire d’Intégration (SMI) (DRC), 110, 125 Sudan, ix, 8, 28, 35, 39; relations with Ethiopia, 80–1; South see South Sudan Sudan People’s Liberation Army (SPLA), 8, 44, 46–7, 48–9, 52, 56, 57, 60–2, 63–4, 65, 70, 73, 74, 79, 81, 83–6, 152, 159; as Dinka militia, 82; disarmament activity of, 67–8, 72, 78, 82, 83, 84 (rolling wave of, 81–3) Sudan People’s Liberation Movement (SPLM), 46, 44, 48, 51 Sudanese Armed Forces (SAF), 42, 45, 47, 50, 59, 63, 86; support for militias, 72 (for SSDF, 55, 58) Sunni Awakening Councils, 9 Switzerland, militia structure of armed forces, 4, 6 Taliban, xii, 138, 147; counterinsurgency against, 128, 137, 144; emergence of, 130–1, 148; in Afghanistan, 39; in Pakistan, 12; neo-Taliban, 136 Tang, Gabriel, 46, 58, 61, 63 tanzim groups (Afghanistan), 139 Taur Matan Ruak, 88, 100, 104 tax collection, by militias, 29

190

Uganda, 109 Ulufa’alu, Bartholomew, 11 Union des patriotes congolais (UPC), 111, 122 Union of Soviet Socialist Republics (USSR), 22; demise of, 135; intervention in Afghanistan, 128–31, 147 ­United Nations (UN), 17, 65, 68, 74, 88, 89, 90, 95, 96–7, 101, 102, 103, 135; mission in Mozambique, 18; mission to Somalia, 18; peacekeeping operations of, 17; role of, 161 (in disarmament, 76–7) UN Angola Verification Missions (UNAVEM), 17, 18 UN Children’s Fund (UNICEF), 112 UN Civil Police (UNPOL), 96–7, 103, 104 UN Development Programme (UNDP), 76, 84, 109, 111, 112; DDR Unit, 68 UN Integrated Disarmament, Demobilization and Reintegra­ tion Standards (IDDRS), 16, 122 UN Mission in the Sudan (UNMIS), 77, 83, 84, 86; Military Observers, 68 UN Transition Assistance Group (UNTAG), 17, 18 UN Transitional Authority in Cambodia (UNTAC), 17, 18 United Self-Defence Forces of Colombia (AUC), 13

United States of America (USA), 22, 136; Constitution of, Second Amendment, 5; military, 152 (lack of understanding of Afghanistan, 145); militia efforts in Afghanistan, 138–42; militias in, 5–6 (limited to state level, 5); strategies of, 134–5 US Special Forces, 131, 141, 143, 145 Upper Nile state, 42, 44, 46, 48, 49, 53, 60, 61, 62, 65–86 passim Venezuela, 4; militias in, 13 village militias, 65, 130 violence: as a form of communication, 156; justification of, 32–3; legitimate, state’s ­monopoly of, 4, 86, 139; levels of, in South Sudan, 84; sexual, against women, 107; spectrum of, 154 wages of militias, 112 war economy, decreasing impact of, 37 war on terror, 135 Wardak, Abdul Rahim, 140 warlords, 144, 148; backed by USA, 134–5; class of, 128, 130, 134; demilitarization of, 133–4 White Army militias (Sudan), xii, 40, 50, 62, 65–86, 152, 159; history of, 66–7; operational modes of, 72–3 White Eagles (former Yugoslavia), 14 White Nile Oil company, 54 Wood, William, 137 World Bank, 108, 109, 110, 112, 160; payments for demilitarization, 124 Yellow Wasps (former Yugoslavia), 14 youth: as members of militias, 66, 69–70; encouraged into new activities, 78; sensitization of, 78, 84 Yugoslavia, militias in, 13–14 Zartman, William, 21

191

Index

timber activities, illegal, 2 Timor-Leste, Democratic Republic of, x, xi, 10–11, 28, 41, 109, 158, 160; establishment, 87; militias in, 87–106; state-building in, 97–102 Timor-Leste Defence Force (FDTL), 87–106 passim Timorese Social Democratic Association, 91 Total company, 54 transition from war to peace, 20 Tutsi people, 7–8, 38, 113–14, 117