UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Concentrate to annihilate : a strategic analysis of the Rwandan genocide Stonehouse, Jeff 2016

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


24-ubc_2016_februrary_Stonehouse_Jeff.pdf [ 1022.52kB ]
JSON: 24-1.0223351.json
JSON-LD: 24-1.0223351-ld.json
RDF/XML (Pretty): 24-1.0223351-rdf.xml
RDF/JSON: 24-1.0223351-rdf.json
Turtle: 24-1.0223351-turtle.txt
N-Triples: 24-1.0223351-rdf-ntriples.txt
Original Record: 24-1.0223351-source.json
Full Text

Full Text

  Concentrate to Annihilate: A Strategic Analysis of the Rwandan Genocide  By  Jeff Stonehouse  B.A., University of British Columbia, 2009      A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE   REQUIREMENTS FOR THE DEGREE OF     MASTER OF ARTS     In     The College of Graduate Studies   (Interdisciplinary Studies)   THE UNIVERSITY OF BRITISH COLUMBIA   (Okanagan)      January 2016      © Jeff Stonehouse, 2016  ii  Abstract   This thesis has two goals. First, drawing upon centuries of strategic theory, I develop an analytical framework called the ‘strategy hierarchy’. This heuristic device simplifies the diffuse field of strategic theory by focusing analysis on three interconnected themes: policy, grand strategy and military strategy. I argue that by using this framework it is possible to transfer the contributions of conventional strategic theory to episodes of genocidal violence.   Second, I use the ‘strategy hierarchy’ as a lens through which to base an analysis of the Rwandan genocide (1994). The ‘policy’ that preceded the genocide was developed by an entrenched political, economic and social elite called the akazu. In the late 1980s, this group of elites encountered three threats to their continued dominance: calls for multipartyism, an economic downturn and an invasion from the Rwandan Patriotic Front (RPF).   The diverse methods of meeting this threat were the domain of ‘grand strategy’. Prior to the outbreak of mass violence, the akazu and their network of supporters tackled these challenges politically, economically, socially, culturally and diplomatically. One of the key processes paralleling this multifaceted assault was the development of a genocidal ideology. According to this belief system, non-violent threats from the political opposition and from the Tutsi ethnicity were conflated with the armed and aggressive RPF.   On 6 April, 1994, President Habyarimana was assassinated. This ‘strategic shock’ resulted in a dramatic shift in grand strategic priorities. Pursuing elite policy, ‘military strategy’ took precedence over non-martial means. Two types of military strategy emerged. The primary strategy targeted Hutu politicians and rival elites. This strategy was largely successful within the first week of the genocide. Then, a new strategy emerged. The organizers of the genocide took  iii  advantage of and encouraged the traditional Tutsi practice of converging upon the churches for sanctuary during times of political upheaval. With thousands gathered at public locations, the perpetrators commissioned a series of extensive and high fatality massacres.                                      iv  Table of Contents  Abstract....................................................................................................................................... ii Table of Contents ....................................................................................................................... iv List of Tables ............................................................................................................................ vii List of Figures .......................................................................................................................... viii Acknowledgements .................................................................................................................... ix Chapter 1: Introduction ............................................................................................................... 1 1.1 A Place Called Nyarubuye ................................................................................................ 1 1.2 Overview of the Thesis ..................................................................................................... 3 1.3 Purpose and Research Question ...................................................................................... 5 1.4 Analytical Framework ....................................................................................................... 6 1.5 Argument .......................................................................................................................... 7 1.6 Significance of the Study .................................................................................................. 9 1.7 Research Design and Sources ......................................................................................... 9 1.8 Organization of Thesis .....................................................................................................11 Chapter 2: Literature Review .....................................................................................................14 2.1.1 Lemkin and the United Nations Genocide Convention ..............................................14 2.1.2 First Generation Comparative Genocide Studies .......................................................16 2.1.3 Definitional Conundrums ...........................................................................................17 2.2 War and Genocide ...........................................................................................................19 2.2.1 War as a Catalyst for Genocide .................................................................................20 2.2.2 War as Inspiration for Genocide ................................................................................21 2.2.3 Mark Levene’s Typology of War ................................................................................22 2.2.4 Martin Shaw and ‘Degenerate War’ ...........................................................................24 2.3 Massacre .........................................................................................................................25 2.3.1 Jacques Sémelin and Massacre ................................................................................28 2.3.2 ‘Genocidal Massacres’ as a Problematic Concept .....................................................29 2.4 Uses of ‘Strategy’ in Genocide Literature .........................................................................30 2.4.1 Benjamin Valentino and the ‘Strategic Perspective’ ...................................................31 Chapter 3: The ‘Strategy Hierarchy’ Analytical Framework .......................................................33 3.1 Why Link Strategy and Genocide?...................................................................................34 3.2 Origins of Strategy ...........................................................................................................36  v  3.3 Definitions of Strategy......................................................................................................37 3.4 The ‘Strategy Hierarchy’ Analytical Framework................................................................40 3.5 Strategic Behaviors and the Intentionalist/Functionalist Debate .......................................42 3.6 Strategy: Science or Ideal Type? .....................................................................................44 3.7 A Caveat: Adversarial and Non-Adversarial Enemies ......................................................45 3.8 Non-Adversaries and Strategy .........................................................................................48 3.9 Summary .........................................................................................................................50 Chapter 4: The Strategic Context:  Political Violence in Rwanda (1959 to 1990) .......................52 4.1 “Race” in Pre-Colonial and Colonial Rwanda ...................................................................53 4.2 Revolutionary Violence (1959) .........................................................................................55 4.3 Inyenzi Raids and Retaliation against the Tutsi................................................................57 4.4 Genocide in Burundi and the Habyarimana Coup ............................................................61 4.5 The Role of Sanctuaries ..................................................................................................63 4.6 Summary .........................................................................................................................65 Chapter 5: Policy, Ideology and the Rwandan Elite ...................................................................67 5.1 Policy and Policymakers ..................................................................................................68 5.2 Policy’s Normative Control over the Strategy Hierarchy ...................................................70 5.3 Genocidal Ideology and Policy .........................................................................................72 5.4 Rwanda: The Emergence of the Akazu ...........................................................................75 5.5 Rwanda: Strategic Shocks ...............................................................................................76 5.6 Creating the Adversary ....................................................................................................80 5.6.1 Print: Kangura and the ‘Hutu Ten Commandments’ ..................................................81 5.6.2 Radio: Radio-Télévision Libre des Mille Collines .......................................................84 5.6.3 The ENI Commission ................................................................................................85 5.6.4 Léon Mugesera's Party Speech.................................................................................86 5.7 Summary .........................................................................................................................87 Chapter 6: Grand Strategy and the Preparations for Genocide .................................................90 6.1 Defining Grand Strategy ..................................................................................................91 6.2 Comparing ‘Grand Strategy’ and Lemkin’s ‘Genocide’ .....................................................94 6.3 Grand Strategy and Rwanda (October 1st 1990 to April 6 1994) .....................................97 6.3.1 Military Strategies ......................................................................................................99 6.3.2 Diplomatic Strategies .............................................................................................. 101 6.3.3 Political Strategies ................................................................................................... 103  vi  6.3.4 Legal and Moral Strategies ..................................................................................... 104 6.3.5 Economic Strategies ............................................................................................... 105 6.3.6 Cultural Strategies ................................................................................................... 106 6.4 Summary ....................................................................................................................... 107 Chapter 7: Military Strategy and the Implementation of Genocide ........................................... 110 7.2 Engagements: Battles and Massacres ........................................................................... 111 7.3 Pre-Genocide Massacres and the Last Strategic Shock ................................................ 114 7.5 Military Strategy and the Rwandan Genocide ................................................................ 117 7.5.1 Original Strategy: Targeting the Opposition ............................................................. 118 7.5.2 Emergent Strategy: Targeting the Tutsi ................................................................... 121 7.5.2a Phase One: Concentration .................................................................................... 122 7.5.2b Phase Two: Degradation and Containment ........................................................... 126 7.5.2c Phase Three: Eradication ...................................................................................... 129 7.5.2d Phase Four: Annihilation ....................................................................................... 132 7.6 Summary ....................................................................................................................... 133 Chapter 8: Conclusion ............................................................................................................. 135 8.1. Question 1: ‘How does strategic theory assist in our understanding of the organization and implementation of genocidal violence and massacres?’ ................................................ 135 8.2 Question 2: ‘How does strategic theory help to explain the synchronized pattern of violence that gripped Rwanda for 100 days in 1994?’ .......................................................... 137 8.3 Limitations ..................................................................................................................... 139 8.4 Future Directions for Research ...................................................................................... 140 Bibliography ............................................................................................................................ 142     vii    List of Tables  Table 3-1 The Strategy Hierarchy .............................................................................................40  Table 3-2 Adversarial vs. Non-Adversarial Enemies .................................................................49  Table 7-1 Types of War and Engagements ............................................................................. 112         viii  List of Figures  Figure 3-1 The Strategy Hierarchy ............................................................................................42  Figure 5-1 The Strategy Hierarchy (Policy) ...............................................................................67  Figure 6-1 The Strategy Hierarchy (Grand Strategy) .................................................................90  Figure 7-1 The Strategy Hierarchy (Military Strategy).............................................................. 110      ix  Acknowledgements  I would like to express my sincerest gratitude to my supervisor, Dr. Adam Jones. Without his encouragement, his kindness, and his expertise, I doubt I would have made it this far. His willingness to proofread my occasionally disastrous grammar came as a great relief. I would also like to thank my thesis committee, Dr. Carl Hodge and Dr. Janet MacArthur.   I would also like to thank the International Association of Genocide Scholars (IAGS). Since joining this organization in 2013 I have had the pleasure of attending and presenting at conferences in Winnipeg and Yerevan. At both conferences I received positive feedback from my fellow IAGS researchers. As a result, I am emboldened by their infectious enthusiasm.   I owe my initial interest in Rwanda to non-profit organization SHOUT Canada. Under their aegis I was able to travel to Rwanda and conduct research at the various massacre sites covered in this thesis. Specifically, I owe a debt of gratitude to Kurt Goddard, Rachel Grace, Audrey Madsen and Faustin Bismarck Murangwa.   I would like to acknowledge the financial support UBC’s Office of Research Services. With their funding I was able to travel to Armenia to present at the IAGS conference. I would also like to thank the College of Graduate Studies and their dedication to Progress Reports.   Finally, without the support and patience of my wife, daughter and parents I would not have managed to write a single page. I appreciate their willingness to acceptance my unusual choice of research topic.         Chapter 1: Introduction  1.1 A Place Called Nyarubuye   For one hundred blood-soaked days in 1994, much of Rwanda was beset by massacres of unfathomable scale and barbarity. One such massacre transpired in and around Nyarubuye parish. This instance of mass slaughter, and many others like it, proceeded according to a common and consistent pattern. This first section provides an anecdotal account of the actions of Sylvestre Gacumbitsi, bourgmestre of Rusomo commune.  Deep within the communal office of Gacumbitsi were thousands of ID cards, each bearing photographs, home addresses and ethnic identity. For weeks prior to the outbreak of genocidal violence, the names and locations of resident Tutsi and moderate Hutu were circulated amongst the interahamwe militia and police. ‘I have no doubt,’ wrote the journalist Fergal Keane shortly after massacres, ‘that this is an index list for murder, prepared years in advance and held in readiness for the day when the Tutsis might need to be sorted out.’1   Like other bourgmestres, Gacumbitsi was a very influential figure and the highest ranked local administrative official.2 Like most of the civilian leadership, he was a member of the Mouvement républicain national pour la démocratie et le dévelopement (MRND), the single party that had dominated Rwandan politics since 1975. And, like other bourgmestres, he had significant sway over the police and army. His influence and connections would ultimately prove decisive in determining the organization and implementation of violence in his home commune.   In the months leading up to the genocide, a climate of fear was firmly established across Rwanda. Newspapers and radio, especially the extremist Radio-Télévision Libre des Mille                                                1Fergal Keane, Season of Blood: A Rwandan Journey (London: Viking, 1995), 86.  2 Judgment, Sylvestre Gacumbitsi (ICTR-2001-64-A), Appeals Chamber, 7 July 2006, § 2.  2  Collines (RTLM), had been spewing out hateful propaganda, and militias had been training and amassing arms for some time. The distribution of ‘kill lists’ was widely known, and Tutsi were frequently harassed solely on the basis of their ethnic identity.   On 6 April 1994, President Juvénal Habyarimana was assassinated when his plane was shot down by a surface-to-air missile over the Rwandan capital, Kigali. Tutsi and Hutu alike were rightly aware that the killing of the President did not bode well. In the immediate aftermath, militias spread throughout Gacumbitsi’s commune, blowing intimidating whistles, setting homes ablaze while stealing Tutsi cattle and possessions.  Killing had begun immediately in the capital, and it seemed only a matter of time before the violence would spread. To this end, on April 9th, Gacumbitsi urged his conseillers de secteur ‘to incite the Hutu to kill the Tutsi,’ and on three occasions on April 13 and 14 he appealed directly to groups of dozens of Hutu to engage in violence.3 The final judgment by the International Criminal Tribunal for Rwanda (ICTR) found that Gacumbitsi drove around his commune and announced by megaphone that Tutsi women should be raped and sexually degraded.4  It quickly became evident that the Tutsi ethnic group were no longer safe in their homes. Because only a fortunate few could find refuge with their Hutu neighbors, thousands took flight with whatever possessions they could carry. In the midst of chaos, the refugees looked to their civil leaders for succor. A delegation of Tutsi asked Gacumbitsi for help, but he declined to protect them; instead, he told them to assemble at the church of Nyarubuye. Unbeknownst to them, this was a ruse meant to lead and concentrate the fleeing Tutsi into an elaborate trap.   Taking Gacumbitsi at his word, thousands of Tutsi and Hutu converged upon Nyarubuye parish. The deception was so successful that word of the refuge spread across the commune and beyond. As more and more refugees gathered from further afield, interahamwe militia and                                                3 Judgment, Sylvestre Gacumbitsi, § 42.  4 Judgment, Sylvestre Gacumbitsi, § 107.  3  local villagers surrounded the church. Priests negotiated with the militia, hoping to arrange a safe passage for the refugees. At first the interahamwe seemed willing to negotiate; however, this was a deception. They were biding their time until they were reinforced by the Presidential Guard and the armed militia from the prefecture’s capital, Kibungo.5  Ultimately, many thousands were massacred in their ‘sanctuary,’ first by militia, then by members of the FAR (Rwandan army), neighbors, and other armed entities beholden to a genocidal ethos. At the forefront was the bourgmestre, Gacumbitsi, leading a troop of soldiers, police, militia, and villagers. He distributed weapons and was said to have personally engaged in some of the killing.6 Gacumbitsi killed a Tutsi named Murefu on 15 April 1994, and this signaled ‘the beginning of the attack at Nyarubuye Parish’.7 The massacre began at 5 in the morning and the killings lasted two days.  In June 2001, Gacumbitsi was arrested while hiding in the Mukugwa refugee camp in Tanzania.8 Brought before the ICTR, he was given life imprisonment for his role in the massacres. Estimates of the number killed vary widely. The generally agreed upon figure is an almost incomprehensible 20,000 slain. Even if these estimates are exaggerated, it is undeniable that thousands were murdered according to a systematic process in and around the parish.   1.2 Overview of the Thesis  The violence at Nyarubuye parish represents a single episode in a coordinated campaign of violence against civilian targets. Rwanda experienced many more episodes of large-scale killing at analogous public sites. Spanning the width of the country, dozens of                                                5 Shaharyar M. Khan,The Shallow Graves of Rwanda (London: I.B. Tauris & Co. Ltd, 2000), 62. 6 Philip Gourevitch, We Wish to Inform You That Tomorrow We Will Be Killed With Our Families (New York: Farrar, Straus and Giroux, 1998) 18. 7 Judgment, Sylvestre Gacumbitsi, § 42. 8 “Two Former Rwandan Mayors Arrested for Genocide.” hirondellenews.com. Hirondelle News Agency, 21 June 2001.   4  landmarks and public buildings became synonymous with massacre. The sheer numbers killed at these sites speak volumes: Ntarama, 5,000; Nyamata, 2,500; Murambi, 27,000; and Nyarubuye, 20,000 to name but a few.9 While this wanton slaughter of innocents appears incoherent to an outside observer, the Rwandan genocide evinced an internal rationality and an overriding purpose. Violence was not a frenzied and directionless outburst, but an instrument used purposefully to achieve certain goals. This thesis contends that an analysis of perpetrator strategy provides a key to unlocking the dynamics of structured genocidal violence.  However, emphasizing strategy as a frame of reference is problematic. When referring to ‘strategy’ the term is conventionally reserved to describe the conduct of war.10 The great theorists of strategy, from Carl von Clausewitz to Colin S. Gray, focus on the conduct of states and armies in conflict. They habitually discount instances of states eliminating non-adversarial social groups. Strategic discourse, as represented by its foremost promoters, possesses a normative bias. Strategy’s purview is limited to contested battles, eschewing one-sided massacres as beneath contempt. There is a ubiquitous tendency to remember the great generals over the great butchers; not a single massacre adorns Sir Edward Shepherd Creasy’s classic 1851 work, The Fifteen Decisive Battles of the World.   Because of this normative bias, strategic thought possess an inherently awkward relationship with genocide and massacre. Awkward, but not incompatible: if we reorient our understanding of genocide and conceptualize the phenomena as a type of war, it is possible to interpret episodes of largely one-sided mass killing in terms of strategy. By equating war with genocide, it is possible to conceive of both as organized and instrumentalized forms of large scale violence. War and genocide alike are waged with a view to a preconceived end that is                                                9 Statistics gleaned from the Kigali Memorial Centre, available from: http://www.kigalimemorialcentre.org/old/centre/other.html 10 While ‘strategy’ has spread to subjects as diverse as business, video games and sport, none of these explicitly involve the use of violence.   5  realized via the strategic preparation, distribution and implementation of means. Before and during the outbreaks of both conventional war and most11 genocidal violence, the state (or power centre) exercises its influence in a variety of public spheres - including the political, social, economic - and manipulate them to create an optimal context for the actual engagement.  While a number of genocide scholars have appropriated the term strategy, they use a vulgarized form of the concept; rarely is the approach defined more than superficially, and even more rarely is it rooted in the seminal contributions of strategic thought. By failing to draw upon strategic theory, contemporary genocide scholarship eschews theoretical insights from a relevant and closely related discipline. Conversely, by failing to consider genocides and massacres, strategic theory is likewise blinded to a fuller account of instrumental violence.    1.3 Purpose and Research Question    The onset of genocide in Rwanda was sudden, but not spontaneous. The 100 days of mass murder were preceded by months of physical and psychological preparation. Even if strategic theory finds genocide a troubling subject, it is more than appropriate to posit the existence of a strategy guiding the course of genocidal violence. The overarching purpose of this thesis is to reconcile strategic thought with genocide studies. In order to address a perceived gap in existing knowledge, this thesis develops an analytical framework based upon the fundamentals of strategic theory. While this framework is equally applicable to both war and genocide, the focus here is on the latter.  The following chapters attempt to answer the following research questions, one broadly conceptual, the other rooted in a specific case. The first question is, ‘How does strategic theory assist in our understanding of the organization and implementation of genocidal violence and                                                11 The emphasis upon ‘most’ here is important because ‘strategy’ of settler violence on an indigenous frontier, for example, would be quite different from the centrally organized discussed herein. Personal correspondence with Adam Jones.    6  massacres?’ The abstractness of this question is countered by a focus on a specific case: ‘How does strategic theory help to explain the synchronized pattern of violence that gripped Rwanda for 100 days in 1994?’   1.4 Analytical Framework   In order to answer the two research questions posed above, it is necessary to define ‘strategy’ according to the prevailing themes of strategic theory. Because of the concept’s complexity, ‘strategy’ cannot be effectively defined in a single sentence. Accordingly, this thesis conceptualizes ‘strategy’ as an analytical framework built around a hierarchy of ends, ways, and means. In times of peace, war and genocide, strategy is instrumental in marshalling available means - including the use of force - in the pursuit of policy goals. The goal of the ‘strategy hierarchy’ analytical framework is threefold:   - To show how policy (the ‘ends’ to be pursued) and policymakers are normatively in control of strategy.  - To show how during both peace and war, ‘Grand Strategy’ encompasses all of the diverse means at a state’s disposal.  - To show how ‘Military Strategy’ is unique among types of strategy because it is responsible for the instrumental use of violence.   This ‘strategy hierarchy’ framework is properly understood as a Weberian ideal type. This designation enables the comparison of a specific case to a paradigmatic model. The value in this methodology is that it ‘acknowledges the historical individuality of social phenomena, but at the same time deploys abstract conceptual constructs to highlight certain significant aspects of empirical reality and enable comparison across cases; without, however, claiming to offer a  7  direct representation of that reality.’12 Accordingly, this thesis aims to juxtapose the Rwandan genocide (including the years leading up to it) with the ‘strategy hierarchy’ ideal type.  A final and significant methodological note must be added. The analytical framework developed herein attempts to present the essence of strategic thought without straying too far from established conceptualizations. However, the task of reconciling strategy and genocide comes with a serious caveat: strategy, as it is conventionally defined, assumes an adversarial opponent. In other words, strategy takes for granted a foe that is armed, organized and capable of mounting their own defensive and offensive behavior. These traits do not apply to the poorly armed and unorganized targets of genocidal violence; and yet, during episodes of mass violence, these groups are treated as threats necessitating a forceful solution. Instead of presenting an insurmountable barrier, this thesis incorporates the contributions of genocide scholars, such as Martin Shaw, who have conceptualized the change in perpetrator perceptions whereby social groups become construed as a military threat.  The clearest manifestation of this caveat is to be found in an expanded conception of the engagement. An engagement is morally neutral encounter between two incompatible entities. Its moral designation changes according to the actors involved: it is called a battle in the context of a conventional war between two or more adversaries; or conversely, it is called a massacre when an armed and organized entity annihilates a mostly defenseless civilian category. Except for this caveat, strategic theory finds a straightforward application to genocidal campaigns.   1.5 Argument   By applying the strategy hierarchy analytical framework to the specific case of Rwanda, it is possible to come to the following conclusions. While the Rwandan genocide is                                                12Asaf Kedar, "Ideal Type" Encyclopedia of Political Theory (Thousand Oaks: Sage Publications Inc., 2010), 678-679.   8  conventionally framed in terms of three-and-a-half months of unremitting slaughter, the most lethal phase occurred relatively early on, during April 1994.  The extreme violence during this period was shaped by two military strategies, one original, and the other emergent. The first strategy targeted specific individuals, usually political elites and opposition figures. This original strategy was prepared well in advance and was largely successful by the end of the first week of the genocide.   The second strategy materialized by the second week of April. This emergent strategy focused upon exploiting the physical concentration of fleeing Tutsi at traditional public sites of refuge. The perpetrators took advantage and eventually encouraged these gatherings in order to facilitate large-scale massacres. Both of these military strategies were rooted in the policy goals and the long term grand strategy of the Hutu power clique. This dominant clique, known as the akazu, aspired to maintain and protect its power and privileges against a multifaceted assault on its interests. These ‘strategic shocks’ consisted of invasion by the RPF, dramatic economic downturn, and calls for democratization. In response, the ‘grand strategy’ of the akazu aimed at confronting this multilayered threat via political, social, economic, cultural, moral, and legal mechanisms.  Until the final ‘strategic shock’ - the assassination President Habyarimana - the military component of ‘grand strategy’ was focused upon low levels of violence while developing armed organizations loyal to the akazu. After the assassination, the original politicidal strategy was put into motion. The violence may have ended here were it not for the genocidal ideological discourse propagated by the akazu via their influence over the media, politics and the military. The Hutu Power ideology conflated the Tutsi ethnic group, political opposition and the RPF into a single threat necessitating a violent solution. Thus, after the original strategy targeting political opposition was complete, strategy turned to the next ‘enemy’ in the form of Tutsi.   9  1.6 Significance of the Study  The intended contribution of this thesis is threefold. First, it aims to bring ‘strategy’ out of the shadows of the more entrenched disciplines. Comparative genocide scholarship prides itself on its inherently multi- and interdisciplinary focus. The last three decades have seen contributions from, inter alia, anthropology, sociology, psychology, and political science. The field of strategic theory, however, has traditionally been relegated to a marginal position, or else subsumed by these established and more encompassing academic fields. And even as scholars attempt to overcome the moral opposition to using ‘strategy’ to discuss genocidal violence, they still manage to conceptualize ‘strategy’ in a theoretical vacuum largely uninformed by millennia of strategic thought.  Second, beyond disciplinary matters, the framework developed in this thesis assists in our understanding of the particular strategies used before and during the Rwandan genocide. In relation to this specific case, the strategy hierarchy provides an alternative lens through which to view and interpret perpetrator actions well before the commencement of generalized violence in April 1994. Third, the value of this analytical framework lies in its applicability to a diverse array of cases that are traditionally ignored by strategic theory. Episodes of genocidal violence are seldom studied from an explicitly strategic perspective. The framework developed in this thesis has a bearing on cases of genocidal violence and massacre far beyond Rwanda.   1.7 Research Design and Sources    The research design that this thesis follows has two parts. First, many of the most salient aspects of strategic theory are synthesized in the analytical framework. Second, the framework is applied to data and accounts of the Rwanda’s history of violence. By these means, we may  10  gain a greater understanding of how and when perpetrator strategy adheres to this ideal type framework. This research design is indebted to three categories of sources. First, this thesis is thoroughly rooted in comparative genocide and massacre theory. Chapter 2 is devoted to highlighting the diverse contributions from this field.  Second, the sections on strategy draw from the canonical documents and other authoritative works that have historically dominated the field of strategic theory and military history.13 These include the writings of Sun Tzu, Thucydides, Niccolò Machiavelli, Antoine-Henri Jomini and Carl von Clausewitz. Additionally, the thesis deploys recent works of strategy by influential authors from the past two centuries, most prominently B. H. Liddell Hart and Colin S. Gray. Joining them are Hans Delbrück, Helmuth von Moltke the Elder, Edward Luttwak, Bernard Brodie, Peter Paret, Michael Howard, Hew Strachan, and J.C. Wylie. As strategy and military history invariably overlap, contributions from strategy-oriented military historians are included, including John Keegan, Martin Van Creveld, Edward Mead Earle, Lawrence Freedman, Barry Posen, Michael I. Handel, and Paul Kennedy. Finally, although he comes from a business and management perspective, Henry Mintzberg contributed to a number of relevant strategic topics.   Third, the case study chapter on the Rwandan genocide draws on varied sources. Three documents issued in the years following the genocide require special mention. They are Leave None to Tell the Story by Human Rights Watch and Alison Des Forges, Rwanda: Death, Despair and Defiance by African Rights, and The Commission for the Memorial of the Genocide and Massacre in Rwanda, by the Rwandan Ministry of Higher Education, Scientific Research and Culture. These documents draw upon interviews with numerous survivors, and each focuses upon the preparation and commission of the massacres, among other themes.                                                 13 By “canonical” I do not mean that these select authors are held as absolute authorities in a religious sense, merely that these documents have historically (and to the present) dominated our discourse on the practice and theory of strategy.   11   Joining these three important documents are works by journalists, including Fergal Keane and Philip Gourevitch; academics, such as Scott Straus, Linda Melvern, and Philip Verwimp; and area specialists, including Gérard Prunier and Rene Lemarchand. Additionally important is the wealth of court documents produced by the International Criminal Tribunal for Rwanda (ICTR), specifically the factual findings in the numerous indictments and judgments. Finally, it should be noted that the author traveled personally to Rwanda in 2012, in order to visit and study numerous massacre sites, including Ntarama, Nyamata, Murambi, Bisesero, and Nyarubuye.   1.8 Organization of Thesis  The literature review in Chapter 2 addresses four interrelated subjects that are integral to subsequent theory building. First, the term ‘genocide’ needs to be unpacked and defined. The concept is shown to derive from the legal scholar Raphael Lemkin and the United Nations Genocide Convention that he helped to engineer. This is followed by an overview of first-generation genocide scholarship. The second part of the literature review focuses upon the interconnectedness of war and genocide. War has been viewed variously as a catalyst for genocide, and also as an inspiration for mass violence. This thesis follows a third interconnecting thread developed by Mark Levene and Martin Shaw that understands genocide as a kind of war. The third section provides an overview of the relatively small literature on the phenomenon of massacre, before turning to a critique of the concept of ‘genocidal massacre.’ The fourth and final section of the literature review dissects a number of ill-defined uses of strategy in the literature on genocide.  Chapter 3 operationalizes the ‘strategy hierarchy’ as an analytical framework. It first asks why we should link strategy and genocide. It then turns to a survey of the historic origins and the diverse definitions of ‘strategy’. Incorporating common definitional themes, the ‘strategy  12  hierarchy’ is presented as a vertical chain of ends, ways and means. It is presented as an ideal type rather than a scientific framework. The chapter then turns to a focus upon a major caveat. This section moves beyond existing strategic theory and attempts to reconcile strategy to the notion of non-adversarial threats.   Chapter 4 describes the historical context within Rwanda prior to the 1990 invasion by the Rwandan Patriotic Front (RPF). It surveys pre-colonial Rwanda, revolutionary violence of the late 1950’s, inyenzi raids of the 1960’s, genocide in Burundi and Habyarimana’s coup in the early 1970’s. It concludes with a discussion on the role played by sanctuaries in protecting non-adversarial groups during times of political turmoil.    Chapters 5, 6 and 7 discuss in greater detail the three core aspects of the strategy hierarchy framework. Chapter 5 focuses upon the role of ‘policy’ and ‘policy makers’ in providing direction for strategy. Policy is held to be normatively in control of strategy. It is argued that when a genocidal ideology motivates policymakers, then their policies and strategies reflect this radical orientation. The chapter then turns to the Rwandan case by defining the elites and the situational milieu of ‘strategic shocks’ which challenged their continued dominance. The chapter concludes by highlighting the numerous processes by which Tutsi and opposition Hutu were transformed into adversarial threats.   Chapter 6 turns to ‘grand strategy,’ understood as the diverse means at the disposal of the state. This multifaceted and encompassing approach to strategy is shown to parallel Raphael Lemkin’s original understanding of ‘genocide.’ Between October 1990 and April 1994, the grand strategy of the Rwandan elite aimed to maintain its power. This is evident in the domains of the military, diplomacy, politics, law, morality, the economy, and culture.   Chapter 7 narrows the focus to ‘military strategy’ and covers the 100 days of genocide beginning in April 1994. It begins with a theoretical discussion on the ‘engagement,’ one of the primary components of military strategy. It then turns to massacres prior to April 1994. Then, the  13  two military strategies are discussed: first, the original strategy targeting the political opposition, then second, the emergent strategy focused upon concentrating and massacring the Tutsi.     14  Chapter 2: Literature Review  The following literature review is divided into four parts. The first section begins with an historical and theoretical overview of the concept of genocide; it then turns to the contentious task of defining the phenomenon in question. The second section narrows the focus to a particular aspect of genocide, specifically its connections to war. Three themes dominate this discussion: war as a catalyst for genocide, war as an inspiration for genocide, and war as a type of genocide. The third sections turns to scholarly approaches to specific manifestation of violence generally referred to as massacres. The fourth and final section attempts to illustrate the diverse uses (and abuses) of the term ‘strategy’ in recent genocide studies.   2.1.1 Lemkin and the United Nations Genocide Convention   Every conceptual discussion of genocide must begin with an overview of the twin pillars of the discipline: Raphael Lemkin’s original definition, and its codification into international law, significantly modified, with the United Nations Convention on the Prevention and Punishment of the Crime of Genocide.14 Both definitions, one sociological, the other legal, structure much of the subsequent debate.  ‘In the beginning,’ writes Dirk Moses, ‘was Raphael Lemkin.’15 Lemkin coined the term ‘genocide’ in his 1944 Axis Rule in Occupied Europe: Laws of Occupation, Analysis of Government, Proposals for Redress. Here he presents a neologism that pairs the Greek genos (tribe, race) with the Latin cide (to kill). To Lemkin, genocide meant ‘the destruction of a nation or of an ethnic group’ via a ‘coordinated plan of different actions aiming at the destruction of                                                14 Ann Curthoys and John Docker, ‘Defining Genocide,’ in The Historiography of Genocide, ed. Dan Stone (London: Palgrave Macmillan, 2008), 13. 15 Dirk Moses, “Raphael Lemkin, Culture and the Concept of Genocide,’ in The Oxford Handbook of Genocide Studies, eds. Donald Bloxham and Dirk Moses (Oxford: Oxford University Press, 2010), 19.  15  essential foundations of the life of national groups, with the aim of annihilating the groups themselves.’16 Such actions, he emphasized, do ‘not necessarily mean the immediate destruction of a nation, except when accomplished by mass killings of all members of a nation.’17 Against a subject population, perpetrator governments implement ‘techniques’ of genocide to ensure the 'disintegration of the political and social institutions, of culture, language, national feelings, religion, and the economic existence of national groups.’18 Genocide, in this sense, refers to ‘the destruction of the personal security, liberty, health, dignity, and even the lives of the individuals belonging to such groups.’19  The 1948 UN Genocide Convention (UNGC) emerged in light of the International Military Tribunal’s failure to denounce genocidal acts outside the context of the Second World War.20 The Convention that ultimately emerged overcame this obstacle by enshrining in Article 1 of the UNGC the clause that states genocide is a crime ‘whether committed in time of peace or in time of war.’21 The UNGC is largely based upon Lemkin’s formulation, albeit with the modification of numerous foundational elements due to political compromises and legal technicalities. Leo Kuper notes that there were major controversies regarding the groups to be protected, the question of intent, the inclusion of cultural genocide, the problem of enforcement and punishment, the extent of destruction which would constitute genocide, and the essential nature of the crime.22 While the UNGC includes national, ethnical, racial or religious groups, notably                                                16 Raphael Lemkin, Axis Rule in Occupied Europe: Laws of Occupation - Analysis of Government - Proposals for Redress (Washington, D.C.:  Carnegie Endowment for International Peace, 1944), 79 - 95. 17 Lemkin, Axis Rule in Occupied Europe, 79 - 95.  18 Lemkin, Axis Rule in Occupied Europe, 79 - 95 19 Lemkin, Axis Rule in Occupied Europe, 79 - 95 20William A. Schabas, "Convention for the Prevention and Punishment of the Crime of Genocide," in United Nations Audiovisual Library of International Law, http://www.un.org/law/avl, (2008): 1.   21 UN General Assembly, Convention on the Prevention and Punishment of the Crime of Genocide, 9 December 1948, United Nations, Treaty Series, vol. 78, p. 277. 22 Leo Kuper, Genocide: Its Political Use in the Twentieth Century (New Haven: Yale University press, 1982), 19 - 40  16  absent were social and political categories.23  Additionally, the UNGC emphasized the role of violence and killing over Lemkin’s more comprehensive notion of a ‘synchronized attack’ on the foundation of a group’s existence. For Lemkin, genocide did not necessarily entail mass killing. In contrast, Article II of the convention includes actions ‘committed with intent to destroy,’ and foremost among them is ‘(a) Killing members of the group.’24 The next two subsequent acts - causing serious bodily or mental harm and deliberately inflicting conditions to bring about a groups destruction - both emphasize the physical element, rather than a gradual undermining of the cultural basis of group cohesion.   2.1.2 First Generation Comparative Genocide Studies   Until the late 1970’s, genocide studies tended towards neglect.25 A new generation of scholars emerged at that time, including Leo Kuper, Irving Louis Horowitz, Ervin Staub, Helen Fein, Vahakn Dadrian, Robert Melson, Israel Charny, R.J. Rummel, Frank Chalk and Kurt Jonassohn.26 While the Holocaust was their frame of reference, comparative research (as opposed to single case monographs) did begin to take root.   Scott Straus has classified these authors as part of the first generation of genocide scholars. While each approaches the subject in diverse ways, Straus argues that there are some commonalities among them. In their interpretations of genocide, these may be understood as a combination of hatred, totalitarianism, and scapegoating.27 Many of these authors                                                23 Helen Fein, “Defining Genocide as a Sociological Concept” in Current Sociology 38: 8 (1990): 10 - 12.   24 UN General Assembly, Convention on the Prevention and Punishment of the Crime of Genocide, 277.  25 This isn’t to say there was no writing at all on the subject. See Curthoys and Docker on Jessie Bernard, We Charge genocide, Hannah Arendt, Pieter Drost and Jean-Paul Sartre in Curthoys and Docker, “Defining Genocide,” 14-26. 26 Helen Fein reviews much of this literature in her Current Sociology piece. Fein, “Defining Genocide as a Sociological Concept.”  27 Scott Straus, "Destroy Them to Save Us": Theories of Genocide and the Logics of Political Violence" in Terrorism and Political Violence 24:4 (2012): 546.  17  (especially Fein and Kuper) share the opinion that deep social divisions, including cultures of prejudice and entrenched practices of discrimination, are structural causes of genocide.28 In addition, a number of these scholars (including Horowitz and Rummel) ascertain a link between authoritarian regime type and genocide.29 A third feature of the first generation of genocide scholars (especially Staub) is the connection it isolated between deprivation and social stress, and the assigning to certain groups of collective blame for society’s ills.30  2.1.3 Definitional Conundrums   Definitions of genocide are inherently fraught with controversy. Cognizant of the rigidity of the UNGC definition, scholars have variously expressed devotion to the legal concept, adopted a version of Lemkin’s thesis, or invented their own definition of genocide. The introduction of new terms, however, has resulted in a ‘conceptual proliferation.’31 Among the ‘many ‘-cides’ of genocide,’ Martin Shaw identifies murderous cleansing, democide, ethnocide, cultural genocide, gendercide, politicide, urbicide, classicide, and auto-genocide.32 By determining which instances may qualify as constituting a ‘genocide,’ the particular definition utilized has implications for the type of cases that may be analyzed. This debate is reflected in the twin poles of exclusivity and inclusiveness that Henry Huttenbach calls the ‘Katz–Charny Conundrum.’33 For Steven T. Katz, the Holocaust is the sole genuine case of genocide in history. Other instances of mass killing may approach qualifying as genocide and share features with it, but these are lesser instances incomparable to the supreme instance of                                                28Scott Straus, “Second-Generation Comparative Research on Genocide" in World Politics 59:3 (2007): 480.  29 Straus, “Second Generation Comparative Research on Genocide,” 481. 30 Straus, “Second Generation Comparative Research on Genocide,” 482. 31 Martin Shaw, What is Genocide? (Cambridge: Polity Press, 2007).  32 Shaw, What is Genocide? 63-78.  33 Henry R. Huttenbach, "Towards a Conceptual Definition of Genocide" in Journal of Genocide Research 4:2 (2002): 168.  18  the Holocaust. In contrast, Israel W. Charny’s approach encompasses a multitude of diverse cases whereby almost every act of organized violence may be construed as an act of genocide.  The definition of genocide used throughout this thesis will follow that of Adam Jones. He contends that genocide is ‘the actualization of the intent, however successfully carried out, to murder in whole or in substantial part any national, ethnic, racial, religious, political, social, gender or economic group, as these groups have been defined by the perpetrator, by whatever means.’34 This is a modification of Katz’s original (1994) definition.35 It breaks with other definitions (especially those adhering to the UNGC) by including not only political groups, but social, gender and economic categories as well. The inclusiveness of groups covered is modified by an important caveat: target groups may have no real internal coherence and may be the fictional creations of the perpetrators themselves. This reflects Frank Chalk and Kurt Jonassohn’s earlier sociological contribution to the definition.36  Additionally, Jones has altered Katz’s murder ‘in its totality’ to ‘in whole or in substantial part.’ Since no genocide has been absolute, Katz’s original definition places unnecessary limits on the cases being analyzed. By Katz’s own definition, even the Holocaust would not qualify as a genocide, for the mass killing was not realized ‘in its totality.’  Jones’s approach is ‘soft’ and ‘inclusive,’ handling issues of intent, requisite numbers of deaths, and perpetrator agency more flexibly than narrower definitions. Jones, however, is strict on a ‘narrower conception of genocidal strategy’ that understands ‘mass killing to be definitional to genocide.’37 Thus, notions of cultural genocide - or ethnocide - are to be excluded from analysis when not accompanied by mass killing. In sum, Jones’s definition overcomes the Katz–                                               34 Adam Jones, Genocide: A Comprehensive Introduction, 2nd ed. (New York: Routledge/Taylor and Francis Publishers, 2010), 22. Italics in original.   35 Steven Katz, The Holocaust in Historical Context, Vol. 1: The Holocaust and Mass Death before the Modern Age (New York: Oxford University Press, 1994), 131. 36 Frank Chalk and Kurt Jonassohn, The History and Sociology of Genocide: Analyses and Case Studies (New Haven: Yale University press, 1990), 23 37 Jones, Genocide: A Comprehensive Introduction, 22. Italics in original.   19  Charny divide via both inclusiveness and limitations.38   2.2 War and Genocide   The repeated concurrence of genocide and war is historically verifiable.  Many of the most substantial genocides of the past century have occurred in the context of war: Armenians were targeted by the Turkish state during the First World War, the Jewish Holocaust and extermination of Roma occurred during the Second World War, and the Rwandan genocide transpired in the context of a war between the Hutu dominated government and the RPF rebel army.  Regrettably, the exact nature of the connection between war and genocide is somewhat nebulous. Helen Fein, aware of this confusion, writes that ‘genocides both lead to war and war leads to genocide.’39 This is not to say that war necessarily accompanies genocide. The Ukrainian famine, the Indonesian purge of communists, and the destruction of the Ache in Paraguay attest to that.40  Nonetheless, war and genocide are intrinsically linked. The connections have been rigorously analyzed in the scholarly literature. Scott Straus contends that one of the most substantial contributions of the second generation genocide scholarship is the ‘strategic or                                                38 There is one shortcoming to this definition of ‘genocide’, albeit one that is easily remedied. The author of this thesis would add former soldiers to the groups singled out for genocide. As they are no longer armed or capable of organized resistance, they are the target of mass violence in a similar fashion to the groups listed above. This category includes soldiers held captive in enemy camps, as well as soldiers from a collapsed army fleeing retribution. Even though this group is ostensibly considered under the category of ‘war crimes,’ the strategies used to destroy formers soldiers frequently coincide with the methods used on the other targeted categories. Examples of this include the mass starvation of Soviet prisoners of war in Nazi camps, and the massacre of thousands of defeated Chinese soldiers by the Japanese invaders in and around Nanking. 39 Helen Fein, "Accounting for Genocide After 1945: Theories and Some Findings" in International Journal on Group Rights 1 (1993): 79.  40 Even though these incidents occurred outside the a ‘direct context of war,’ Shaw reminds us that in these instances, among others, the ‘perpetrator states were militarized states with militaristic ideologies, and have recently experienced wars.’ Martin Shaw, War and Genocide: Organised Killing in Modern Society (Wiley, 2003), 44.  20  rationalist approach.’41 ‘[T]he empirical connection between genocide and war,’ writes Straus, ‘is arguably the most robust empirical finding in the most recent literature.42 But, as we will show, the connection between war and genocide was by no means ignored by earlier scholarship.   For now, it is necessary to elaborate first generation views on the connection between war and genocide as elaborated by Markusen, Kopf and Alvarez. They have described war as a catalyst for genocide. Second, Bartrop understands war (World War I in particular) as an inspiration for genocide. Finally, following Raphael Lemkin, Levene and Shaw are representative of current second generation scholarship, in which genocide is understood as a form or type of war.   2.2.1 War as a Catalyst for Genocide   Markusen, Kopf and Alvarez understand the potential for war to act as an enabler and impetus for an environment in which genocide may ultimately emerge. Fist, war ‘produces widespread psychological and social disequilibrium.’43 Markusen and Kopf contend that previously dormant cleavages may flare up as a section of society assumes the role of scapegoat. This minority may be variously blamed for military setbacks and aiding an external foe. Alvarez adds that war creates ‘psychological and spatial dislocation as entire populations are uprooted and set adrift.’44 War breaks traditional social structures and scapegoat minorities in this context may be ‘attacked as a consequence of the weakened sense of community and diminished social control.’45 This view draws from the earlier ‘first generation’ work of Fein and Kuper.                                                  41 Straus, “Destroy them to Save Us”, 546. 42 Straus, “Destroy them to Save Us”, 546. 43 Eric Markusen, "Genocide and Warfare" in Genocide, War and Human Survival, eds. Charles B. Strozier and Michael Flynn (Lanham: Rowman and Littlefield Publishers, Inc., 1996) 44 Alex Alvarez, Governments, Citizens, and Genocide: A Comparative and Interdisciplinary Approach (Bloomington:  Indiana University Press, 2001): 69. 45Álvarez, Governments, Citizens, and Genocide, 69.  21  Second, governments engaged in total war become more ‘centralized, secret and powerful.’46 States engage in censorship in order to limit public knowledge of atrocities, while simultaneously producing propaganda to vilify the enemy. R.J. Rummel has expressed this tendency as the ‘Power Principle.’ Rummel contends that ‘[t]he more power a government has, the more it can act arbitrarily according to the whims and desires of the elite, the more it will make war on others and murder its foreign and domestic subjects.’47 Third, the soldiers called up to fight a war may be used as perpetrators of genocide. As they have already been inculcated with notions of killing for one’s nation, it is a logical step to expand the purview of military operations to include civilian targets. The armed forces, Alvarez emphasizes, already have the ‘necessary resources, organization, logistics, and personnel to carry out genocide.’48 Fourth, minorities become ‘isolated and marginalized’ and increasingly vulnerable vis-à-vis the expanding influence of the state apparatus.49 These groups are excluded from sources of social and political power and are, to use Fein’s term, stranded outside the perpetrator’s ‘universe of responsibility.’ Finally, war ‘brutalizes populations and desensitizes them to violence.’50 War leads people to be acclimatized to previously inconceivable horrors.    2.2.2 War as Inspiration for Genocide   Another connection between war and genocide has been cited.  Bartrop has linked war - in particular, the First World War - with generating among human beings ‘the belief that they                                                46 Markusen "Genocide and Warfare," 78.  47 R.J. Rummel, Death By Government (New Brunswick, N.J: Transaction Publishers, 1994).  48 Álvarez, Governments, Citizens, and Genocide, 70.  49 Álvarez, Governments, Citizens, and Genocide, 70.  50 Álvarez, Governments, Citizens, and Genocide, 71.   22  could, with impunity, slaughter massive numbers of their own kind.’51 With the ‘Great War’ of 1914-18, people ‘became conscious not only of the idea of massive man-made death but also of the ease with which it could be achieved.’52 Bartrop’s contention is that the idea of ‘massive military death’ was transferred to possibility of ‘massive political or ideological death.’53 In his view, while there is a recurring connection between genocide and war, the two are not synonymous: genocide may occur outside of war (he cites the Ukrainian Famine, the Indonesian massacres, and the Chinese takeover of Tibet), and war may occur without the occurrence of genocide (for instance, the war in the Falklands).54 Nonetheless, it was the ‘concept of mass death achievable by human actions’, exemplified by the First World War that gave inspiration to later genocidal episodes.55   2.2.3 Mark Levene’s Typology of War   Beyond serving as a catalyst or an inspiration, a third thread links war and genocide: the view that genocide is a kind of war. This connection, argues Mark Levene, is reflective of Lemkin’s original understanding of genocide. Lemkin framed his crusade in terms of extending international law to include a new and previously excluded category of war crimes.56 The chief difference between war and genocide, as Lemkin saw it, lay in whom the state was waging conflict against. Conventional war occurred between states and professional armies, while genocide pitted a state against a ‘national, ethnic or some comparable population.’57  Levene extends Lemkin’s framing by distinguishing among three types of war waged by                                                51 Paul Bartrop, "The Relationship Between War and Genocide in the Twentieth Century: A Consideration" in Journal of Genocide Research 4:4 (2002): 522 52 Bartrop, “The Relationship Between War and Genocide in the Twentieth Century,” 522.  53 Bartrop, “The Relationship Between War and Genocide in the Twentieth Century,” 524.  54 Bartrop, “The Relationship Between War and Genocide in the Twentieth Century,” 527 and 530.  55 Bartrop, “The Relationship Between War and Genocide in the Twentieth Century,” 531.  56 Mark Levene, Genocide in the Age of the Nation State Vol. 1: The Meaning of Genocide (London: I.B. Tauris, 2005), 51.  57 Levene, Genocide in the Age of the Nation State Vol.1, 55.   23  states. The first genre is between mutually recognized sovereign states. Using examples from the Second World War, Levene includes the Pacific war between the United States and Imperial Japan. This first type represents the standard ‘normative’ form of war that, according to Levene, is ‘acceptable... in a Clausewitzian sense.’58 The second type of war pits a sovereign state against an external state or nation that the former perceives as being ‘illegitimate.’ Examples of this include Germany vis-a-vis Poland, and Japan vis-a-vis China. In these conflicts, the armed forces and, frequently, the civilian population of the perceived ‘inferior’ enemy are targeted for destruction.  The third type of war Levene describes involves a sovereign entity that targets an internal national group depicted as alien and/or illegitimate. Here, Levene argues, ‘the disparity between the state’s military capacity to kill the targeted population and the latter’s ability to resist is usually so marked that the state’s killing operatives can undertake their task not only for the most part with a high degree of impunity but often with no serious threat to their physical existence whatsoever.’59  While all manner of horrible tactics and techniques may be present, no single type of war necessarily devolves into genocide. Ostensibly, even in this third type of war, organized resistance is possible. Levene cites the Irish national liberation as an example of a type 3 war that had all the ingredients for genocide. Rather than militarily unfeasible operations and civilian massacre, the British opted for a negotiated settlement.60 Levene concludes that while ‘genocide could have emanated from this War Type Three’, its absence in practice ‘would suggest that the two are not automatically identical or congruent.’61                                                  58 Levene, Genocide in the Age of the Nation State Vol.1, 56.  59 Levene, Genocide in the Age of the Nation State Vol.1, 58.  60 Levene, Genocide in the Age of the Nation State Vol.1, 63. 61 Levene, Genocide in the Age of the Nation State Vol.1, 63.  24  2.2.4 Martin Shaw and ‘Degenerate War’  A useful addition to Levene’s typology is Martin Shaw’s notion of ‘degenerate war.’ In each of Levene’s types of war, combat invariably results in civilian fatalities. Sometimes this takes the form of incidental ‘collateral damage’ where the deaths were not intended, but the result of legitimate military action against another armed actor. Alternatively, states and their armed forces may intentionally target civilians. These conflicts, once civilian targeting comes to be an integral part of strategy, may be deemed ‘degenerate wars.’  Shaw defines these wars as the ‘deliberate and systematic extension of war against an organized enemy to war against a largely unarmed civilian population.’62 Combat operations target both civilian groups and military forces. While strategies and the modes of violence employed against armed and unarmed groups are fundamentally different, they operate alongside each other and are often complementary.  A degenerate war, like the three types outlined by Levene, does not necessarily lead to genocide. Violence against civilian targets may be a form of collective punishment or a means of encouraging cooperation. Extermination need not be the pursued rationale. Once ‘victory’ is achieved over the external or internal enemy, some sort of ‘normalcy’ may prevail. As an example, Levene cites the Nigerian conflict with Biafra. This conflict qualifies as a degenerate war because it targeted both soldiers and civilians; moreover, it qualifies as a Type 3 war because it was waged against an illegitimate internal foe. While the Nigerian state utilized systematic starvation against Biafra, once the conflict was over, the rebellious region was readmitted into the federal fold.63 While ‘normalcy’ may mean decades of oppression and subordinate status, it does not entail the ‘extermination’ that is central to a genocidal programme.                                                62 Shaw, War and Genocide, 5.  63 Levene, Genocide in the Age of the Nation State Vol.1, 61.  25  Conversely, degenerate wars may be transformed ultimately into full-fledged genocides when violence against civilians, instead of being incidental or instrumental in defeating an armed threat, becomes the objective in itself. In practice, all three Levene’s types of war overlap with Shaw’s definition of degenerate war. Most wars have, to widely varying degrees, their degenerate and conventional phases.    2.3 Massacre   As a discrete unit of analysis, the subject of ‘massacre’ has generated few standout scholarly works. Only infrequently is the phenomenon defined, examined comparatively, and/or distinguished from other forms of killing. Over the past two decades the field has seen scholarly contributions from Eric Carlton, Brenda Uekert, Mark Levene, Penny Roberts, and Jacques Sémelin. Eric Carlton was one of the first to explore the varieties of massacre throughout history. He defines massacre as ‘the indiscriminate killing of unresisting or defenseless people, which can be subsumed under the more general term atrocity which normally refers to some horrible or cruel act.’64 His stated intention is to ‘examine the situational contexts in which these incidents took place in order to appreciate why they happened as they did.’65 This approach is useful for approaching instances of mass killing without an explicitly political agenda, including such cases as the Roman gladiatorial contests.  He presents a typology of massacres because this ‘exemplifies the phenomenon in the context of a number of different cultures and situations.’66                                                64 Eric Carlton, Massacres: An Historical Perspective (Hants: Scolar Press, 1994): 1. 65 Carlton, Massacres, 2. 66 Carlton, Massacres, 1.  26  Regrettably, Carlton’s short introduction only alludes to a massacre’s function in relation to individual case studies. No effort is made to fit these episodes into a comparative framework, and the cases seem to be chosen arbitrarily. In addition, the concept of ‘massacre’ is underdeveloped and defined in the broadest of strokes. A ‘massacre’, for Carlton, is essentially any atrocity involving intentional killing, thus expanding the number of cases to seemingly Charny-esque proportions. Brenda Uekert’s Rivers of Blood overcomes some of the shortcomings of Carlton’s book by explicitly framing massacres as a subject for comparative study. Working within a framework of state terrorism, Uekert discerns two main targets for government massacre: first, genocidal massacres67 refer to ‘massacres of persons based on their communal characteristics, such as their racial, ethnic, national or religious status’; and second, politicidal massacres, which are conducted when ‘persons are targeted for murder because of their real or perceived opposition to regime and dominant group.’68  Uekert’s definition of ‘massacre’ is stronger than Carlton’s. For Uekert, massacres are intended not only to eliminate ‘supposed or real opponents, but to terrorize the population into refraining from organizing themselves socially or politically.’69 Massacres are characterized by ‘violence and cruel and degrading treatment,’ and are never random, but carried out in a ‘systematic fashion’ against victims who are ‘invariably defenseless.’70 One of the highlights of Uekert’s text is her typology of genocidal and politicidal massacres classed according to underlying motivation. A state will target political groups in order to ‘consolidate the power of the central authority’, or will target communal groups to ‘maintain or enhance the role of the dominant ethno-religious group in the stratification                                                67 This use of the term ‘genocidal massacres’ is not the same the one discussed below.  68 Brenda K. Uekert, Rivers of Blood: A Comparative Study of Government Massacres (London: Praeger, 1995): 7.  69 Uekart, Rivers of Blood, 6 70 Uekart, Rivers of Blood, 6  27  system.’71 Alternatively, in order to preserve or re-establish social control, states will target political groups to ‘maintain oligarchic control of the political and economic systems,’ or they will target communal groups to ‘establish government control over a separatist nationality or indigenous population.’72  As a basis for comparing various cases, Uekert’s typology is useful for contrasting the desired political ends. However, Uekert does not describe the massacres in detail; instead, like Carlton, she focuses upon the context of the events. While context is certainly crucial, an overriding focus upon background factors leaves the reader ignorant as to the manner of preparation and the conduct of the massacre itself. In other words, while Uekert does focus upon the nature of the perpetrator and victim, the dynamics of massacre are subsumed by the larger set of circumstances. Additionally, her selection of massacres range from eight killed in San Salvador to over 5,000 killed in Halabja. While both indeed constitute ‘massacres,’ the difference in the scale of the killing is by no means negligible. Finally, her definition lacks any notion of time and space: for example, did the killing occur over a few days or months, within tight confines or spread widely?  While Mark Levene and Penny Roberts’ edited book The Massacre in History is slightly uneven (as is inevitable when multiple writers take on a thousand years of mass violence), some of the essays stand out, including Levene’s introductory chapter. Levene defines massacre as when ‘a group of animals or people lacking in self-defense, at least at that given moment, are killed - usually by another group... who have the physical means, the power, with which to undertake the killing without physical danger to themselves.’73 While this definition is useful it lacks any mention of massacre as a functional activity, part of a plan to realize specific ends. On the positive side, Levene recognizes that massacres occur in time and space; they                                                71 Uekart, Rivers of Blood, 168. 72 Uekart, Rivers of Blood, 168. 73 Mark Levene, “Introduction” in The Massacre in History, eds. Mark Levene and Penny Roberts (Oxford: Berghahn Books, 1999): 5.  28  occur in ‘quite specific spatial and temporal dimensions. Massacre implies an event which takes place in a limited, though not defined geographic arena, as well as in a limited, though again not clearly defined time period.’74  2.3.1 Jacques Sémelin and Massacre  It is in the writings of Jacques Sémelin that massacre as a ‘lexical unit of reference’ is most thoroughly developed.75 He attempts to narrow the definition of massacre to the specificity required that social-scientific analysis requires, posing questions about proximity (is the murder face-to-face or at a distance?), qualitative versus quantitative definitions (does the murder of a single individual constitute a massacre?), and the nature of the perpetrator (is it the state or militant organization?).76 For Sémelin, the linkages between massacres and strategy are particularly noteworthy. He writes: ‘[t]he primacy of the political in the understanding of massacres leads to the prioritization of the analysis of their strategic function as a means to the seizure and acquisition of power.’77 He later expands his theory to include a typology of objectives, including actions aimed at destroying to subjugate (the partial annihilation of a group to force the remainder into acquiescence), destroying to eradicate (the elimination of a community from a territory), and destroying to revolt (acts by non-state actors in order to influence government policy).78 Sémelin understands massacre as part of a process. He writes: ‘The study of massacres must necessarily fit within the broader framework of a destructive process, of which it is the most spectacular expression.’ He continues:  ‘As a form of extreme violence, a massacre needs                                                74 Levene, “Introduction,” 6. 75 Jacques Sémelin, Purify and Destroy: The Political Uses of Massacre and Genocide (New York: Columbia University Press, 2007): 323 76 Jacques Sémelin, “In Consideration of Massacres” in Journal of Genocide Research 3:3 (2001): 378 - 379. 77 Sémelin,“In Consideration of Massacres,” 380. 78 Sémelin, Purify and Destroy, 327-361.  29  to be understood in the context of the dynamic of violence that precedes and accompanies it.’79 Sémelin argues that emphasizing the physical outcome – ‘the act of killing’ - obscures our understanding; as a result, different ‘forms of violence occurring prior to a massacre might simply be ignored or played down.’80 Examining massacre as a process allows for an appreciation of the preparations it requires. This moves the focus beyond (while still encompassing) what Lee Ann Fujii has described as ‘extra lethal violence,’ i.e. the ‘cruelties, punishments and brutalities’ that precede (and proceed from) the actual massacre.’81  2.3.2 ‘Genocidal Massacres’ as a Problematic Concept  Some comments are merited, however, about the term ‘genocidal massacre.’ Leo Kuper coined this phrase as a means of including cases that fall outside the terms of the UN Genocide Convention.82 It is frequently utilized when questions of the scale of casualties, or the nature and intent of the perpetrator are difficult to ascertain, but with an understanding that a horrendous act of organized violence has occurred.83  Additionally, the concept is used to describe circumstances where large-scale mass killing is undertaken by states that do not explicitly pursue an exterminatory agenda. Stark, for instance, contends that the deaths of 200,000 Hungarian prisoners in Soviet prison camps ought to be considered a genocidal massacre.84 While many died in conditions of forced labour, Stark contends the UNGC does not cover this case because there was no established policy                                                79 Sémelin, Purify and Destroy, 324. 80 Jacques Sémelin, “From Massacre to the genocidal process” in International Social Science Journal, No.174 (2002): 435. 81 Lee Ann Fujii, “The Puzzle of Extra Lethal Violence” Seminar Series on Civil War and the Challenge of Political Order (2011).  82 Kuper, Genocide: Its Political Use in the Twentieth Century, 10.  83 Eric Markusen and David Kopf, The Holocaust and Strategic Bombing: Genocide and Total War in the Twentieth Century (Westview Press, 1995), 63. 84 Tamás Stark, "Genocide or Genocidal Massacre?: The Case of Hungarian Prisoners in Soviet Custody" in Human Rights Review, April-June (2000): 115.  30  objective to annihilate them (i.e. to destroy Hungarians in whole or in part).85 The problem with the term ‘genocidal massacres’ is that, depending on the methods of killing utilized, actual massacres may be absent. This is a problem with Stark’s usage of the term. In other words, there is a major difference between prisoner deaths resulting from a massacre and ‘tacit consent to the inhuman living conditions.’86 Even though mass death was the end result, massacres were not overtly present. While the treatment of Hungarian prisoners ought to be recognized as genocidal including the term ‘massacre’ obscures the nature and methods of violence. For these reasons, this thesis will not deploy the term ‘genocidal massacres.’ Rather, it will utilize a more expansive definition of ‘genocide’ that makes it possible to incorporate the kind of cases excluded by the UNGC.      2.4 Uses of ‘Strategy’ in Genocide Literature  The word ‘strategy’ is frequently invoked in the literature on comparative genocide, either in passing or as a central element of the argument. Unfortunately, when it is used, it is used clumsily. Seldom is the term defined or operationalized, and even less often is it rooted in existing strategic thought. The concept of ‘strategy’ has been used primarily in two forms: it is conflated with either the desired ends pursued or the means utilized to achieve this goal. These uses of ‘strategy’ obscure the role that strategy plays in linking ends and means.  First, strategy is understood as instrumental in realizing certain political goals.  According to Helen Fein, genocide is ‘a strategy that ruling elites use to resolve real solidarity                                                85 Stark, "Genocide or Genocidal Massacre?”, 115. 86 Stark, "Genocide or Genocidal Massacre?”, 116.  31  and legitimacy conflicts or challenges to their interests.’87 More recently, Straus writes that ‘strategies of mass violence are developed in response to real and perceived threats to the maintenance of political power.’88 Second, genocidal strategy and means are equated. Jurkovich, for instance, describes ‘the use of starvation as a primary killing strategy versus more active methods such as death by firearm or bayonet.’89 The Office of the UN Special Adviser on the Prevention of Genocide (OSAPG) states that genocide is ‘a “tool” or “strategy” to achieve military goals in an operation whose primary objective may be unrelated to the targeted group.’90 Sartre invoked genocide as ‘the foundation of anti-guerrilla strategy.’91  R.J. Rummel argues that ‘democide’ (his concept that includes both genocide and politicide) may ‘become part of the strategy for achieving victory.’92 In Rwanda, Gerald Caplan writes that ‘[b]oth physical and rhetorical violence [were used] against the Tutsi,’ adding that ‘that at some point it turned into a strategy for genocide.93  2.4.1 Benjamin Valentino and the ‘Strategic Perspective’    Benjamin Valentino has gone further than most in applying strategy to genocidal violence. He extolls the virtues of a ‘strategic’ perspective, whereby ‘mass killing’ [his altered definition of genocide] occurs when powerful groups come to believe it is the best available                                                87 Helen Fein, "Revolutionary and Antirevolutionary Genocides: A Comparison of State Murders in Democratic Kampuchea, 1975 to 1979, and in Indonesia, 1965 to 1966" in Comparative Studies in Society and History 35:4 (1993): 813.  88 Straus, “Destroy them to save us,” 546.  89 Michelle Jurkovich, "When Food is the Weapon: Examining the Strategy of Starvation in Cases of Genocide" APSA 2010 Conference, 2. 90 Office of the UN Special Advisor on the Prevention of Genocide, Analysis Framework availible at: http://www.un.org/en/preventgenocide/adviser/pdf/osapg_analysis_framework.pdf  91Jean-Paul Sartre, “Statement 'On Genocide'” Second Session of the Bertrand Russell International War Crimes Tribunal on Vietnam (1967). 92 R.J. Rummel, "Democracy, power, genocide, and mass murder," in The Journal of Conflict Resolution 39 (March 1995): 3-26. 93 African Union, Rwanda: The Preventable Genocide (July 2000): 39.  32  means to accomplish certain radical goals, counter specific types of threats, or solve difficult military problems.’94 Mass killing is understood as both a means to an end and as an ‘instrumental policy.’95  While Valentino’s scholarship is top notch, his definition of ‘strategy’ exemplifies the pitfalls already noted. He examines strategy in terms of ‘ends’ and ‘means,’ rather than situating it as a link between the two. His empirical analysis, meanwhile, present a different problem. While his section on Rwanda is excellent in describing the perpetrator’s grand strategy, Valentino devotes only a few brief sentences to the actual conduct of violence. Even then, he pays little heed to the military strategies utilized.96 In sum, while Valentino should be applauded for popularizing ‘strategy in the discussions of mass killing, he fails to sufficiently flesh out the concept. This is the task of the next chapter.                                                            94 Benjamin A. Valentino, Final Solutions: Mass Killing and Genocide in the 20th Century (Ithaca: Cornell University Press, 2004), 66.  95 Valentino, Final Solutions, 67. 96 Valentino, Final Solutions, 187.  33   Chapter 3: The ‘Strategy Hierarchy’ Analytical Framework   The primary aim of this chapter is to construct an analytical framework that reconciles mainstream strategic thought with the unconventional violence that characterizes genocide. The ‘strategy hierarchy’ analytical framework developed herein understands ‘strategy’ as functionally linking ends with means. This framework guides the subsequent discussion on the Rwandan genocide.   The first section attempts to answer the question of why we should link strategic theory with the phenomenon of genocide. The key to answering this question resides in Hannah Arendt’s understanding of instrumental violence. Next, the historical origins and the diverse definitions of strategy are presented. Three themes are found to be key to defining strategy: its subordination to policy, its division into ‘grand’ and ‘military’ spheres, and its utilization of the military engagement. This discussion is incorporated into a conceptualization of strategy as a hierarchy of ends, ways and means. Strategy is understood here as coordinating an array of actions that are deployed with particular end goals in mind. This ‘strategy hierarchy’ serves as the basis of the analytical framework of this thesis.   I contend that it is wrong to view strategy as a science, and instead prefer to use the label ‘ideal type’. In this manner we may juxtapose unique strategies with the timeless elements of strategy. Additionally, this section breaks with one problematic strategic truism: that strategy must be adversarial. It argues that perpetrators, mediated by ideology, come to perceive targeted groups as threats in an explicitly military sense. Resistance, even if ineffectual, reinforces this dynamic.    34  3.1 Why Link Strategy and Genocide?    It is not particularly controversial to claim that war has long been studied and practiced in terms that would eventually be labeled as strategic. ‘Even if the word was not always in use,’ writes Freedman, ‘we can now look back and observe how personalities engaged in activities that would later be called strategy.’97 For centuries, scholars and soldiers have reflected upon the act of war making by various entities. Whether describing the conduct of city-states, empires, nations or tribes, their collective efforts have led to an understanding (imperfect though it may be) of how to prepare for war, and how to wage it once hostilities have commenced.      Likewise, the practices of genocide possess an ancient lineage that has long accompanied human history. It is in this sense that Leo Kuper’s makes the apt claim that ‘[t]he word is new, the concept is ancient.’98 Roger Smith concurs in his reckoning that ‘[w]hat we today call “genocide” is as old as history.’99 Lemkin’s earliest cases reach back as far as Biblical times and the ancient Assyrians, while Ben Kiernan begins his magnum opus with Sparta.100 Chalk and Jonassohn include in their edited volume The History and Sociology of Genocide numerous cases rooted in antiquity and the medieval era, such as the destruction of Melos and Carthage, and later the Albigensian Crusade.101  That the links between war and genocide are more than incidental has been borne out by the history of political violence. As detailed in section 2.2, Mark Levene and Martin Shaw both contend that genocide ought to be understood as a type of war. This conceptualization                                                97 Lawrence Freedman, Strategy: A History (Oxford: Oxford University press, 2013), p. 72.   98 Leo Kuper, Genocide: Its Political Use in the Twentieth Century (Harmondsworth: Penguin, 1981), p. 9 99 Roger W. Smith, ‘As Old as History’, in Will Genocide Ever End? eds. Carol Rittner, John K. Roth and James M. Smith (St. Paul: Paragon House, 2002): p. 31. 100 Raphael Lemkin, Lemkin on Genocide (Lanham: Lexington Books, 2012)) and Ben Kiernan, Blood and Soil: A World History of Genocide and Extermination from Sparta to Darfur (New Haven: Yale University Press, 2007) 101 Frank Chalk and Kurt Jonassohn, The History and Sociology of Genocide: Analyses and Case Studies (New Haven: Yale University press, 1990).  35  holds both war and genocide to be two sides of the same coin: they are comparably lethal practices, even as they differ fundamentally regarding the adversary to be fought. This leads to the question: what common factor(s) underpin a typology of war? One answer is the ubiquitous practice of instrumental violence. Organized acts of force for a political purpose dominates the course of both conventional and unconventional conflicts. ‘Violence,’ writes Hannah Arendt, ‘is by nature instrumental; like all means, it always stands in need of guidance and justification through the end it pursues.’102 While the scope and intensity of conventional war, degenerate war and genocide all vary to wide degrees, what these types of political violence all share is a calculation by one or more parties to use bloodshed. If the academic study of mass killing views genocide as a type of war, then analysis should inherit and adapt the conceptual tools accumulated and developed by centuries of strategic thought. To this end, Arendt’s quotation may be understood in strategic terms as describing a hierarchy of ends, ways and means. The ‘justification’ she cites is the domain of policy and ideology; her use of the term ‘guidance’ is interchangeable with the definition of ‘strategy’ developed below; and whether it be against soldiers or civilians, violence is necessarily implicit in any means that prioritise the purposeful use of force. This ‘strategy hierarchy’ is expanded below and provides the foundation for the analytical framework that follows.  Arendt’s notion of ‘guidance’ is essential to genocide. Barbara Harff argues that ‘[i]n genocides and politicides killings are never accidental, nor are they acts of individuals. The key is that they are carried out at the explicit or tacit direction of state authorities, or those who claim state authority.’103 Along similar lines, Valentino contends that ‘most scholars studying political violence now understand it to be primarily, if not exclusively, instrumental and coordinated by                                                102 Hannah Arendt, On Violence (Orlando: Harcourt, Inc., 1970): 51. Italics added.  103 Barbara Harff, 'No Lessons Learned from the Holocaust? Assessing Risks of Genocide and Political Mass Murder Since 1965" in The American Political Science Review 97:1, p.58. Italics added.   36  powerful actors seeking to achieve tangible political or military objectives.’104 Regardless of whether one uses the term ‘guidance,’ ‘direction, or ‘coordination,’ both war and genocide assume that at least one party to the violence is behaving strategically.  In sum, the linkages between strategy, war and genocide are ancient but perennially relevant. The argument pursued above adheres to the following syllogism: ‘War is shaped by strategy. Genocide is a type of war. Therefore, genocide is shaped by strategy.’ This simple logic allows genocide scholarship to take on a strategic perspective and adopt the conceptual tools traditionally reserved for the analysis of war. The common factor that unites a typology of war (conventional war, degenerate war, and genocide) is the ubiquity of instrumental violence.   3.2 Origins of Strategy  As argued above, the recurring themes of strategic thought may be traced back to an ancient lineage. Lawrence Freedman concurs with this view, contending that strategy has three historical origins. First he cites the ‘evolutionary’ genesis, showing that chimpanzees display strategic behavior in their instrumental use of violence. Then there is the ‘Biblical’ origin, whereby God and His laws determine the nature of organized violence and its limits. Third, Freedman locates the ‘Hellenic’ ancestry of strategy in the ancient Greeks and the Homeric tradition.105  Arguably, Freedman’s interpretation leans towards eurocentricity. A less western-biased search for the origin of strategic thought focuses upon ancient Chinese military writings, including the Seven Military Classics, of which Sun Tzu’s Art of War is the oldest and the best-known today.106 From the Indian subcontinent, the Arthashastra touches upon military                                                104 Benjamin Valentino, 'Why We Kill: The Political Science of Political Violence Against Civilians' in Annual Review of Political Science 17:2014, p. 91. Italics added.  105 Freedman, p. 3-41.  106 Ralph D. Sawyer, The Seven Military Classics of Ancient China (Boulder: Westview Press, Inc., 1993).   37  strategy.107 S. K. Malik has written on the Quranic concept of war, and rooted strategic thought in the seventh-century Islamic text.108 In the end, an appreciation of multiple, independent origins of strategy lends credence to the timelessness and durability of its elemental features.  Regardless of geographic origin, these authors all approach the same subject: the instrumental use of violence. ‘There is no such thing as an exclusively ‘Western’ or ‘Eastern’ approach to strategy and politics,’ contends Michael Handel.109 In his extensive comparison of Sun Tzu and Clausewitz, Handel concludes that the ‘basic logic of strategy, like that of political behavior, is universal.’110 He claims that Machiavelli’s works acted as a bridge between the classical thinkers like Sun Tzu and Thucydides, and later thinkers like Clausewitz and Jomini. This link, he claims, ‘illustrates the similarities, conceptual unity, and universality of strategic thought.’111  3.3 Definitions of Strategy   Hew Strachan is correct to lament that ‘[t]he word ‘strategy’ has  acquired  a  universality which  has  robbed  it  of  meaning,  and  left  it  only with banalities.’112 Originally used solely in a martial sense, ‘strategy’ has invaded the discourse of non-military activities; as a result, the term is used widely, from the business realm to video games.  As argued in chapter 2, a vague conception of ‘strategy’ has been endemic within genocide scholarship. Seldom, though, is the term rooted in anything more than its colloquial use. The prevalence of the term ‘strategy’ and                                                107 Kautilya, Arthasastra, trans. R. Shamasastry. Available online: https://archive.org/details /Arthasastra_English_Translation 108 S.K. Malil, The Quranic Concept of War (Delhi: Adam Publishers and Distributors, 1992). 109 Michael I. Handel, Masters of War: Classical Strategic Thought (Portland: Frank Cass Publishers, 2001), p.2. 110 Handel, p. xiv.  111 Handel, p. 3.  112 Hew Strachan, "The Lost Meaning of Strategy" in Survival: Global Politics and Strategy 47:3, p. 34.   38  its indistinct position within genocide scholarship acts to complicate the task of discerning a concise and broadly applicable definition of ‘strategy.’  Therefore, to the question, ‘What is strategy?’ we have a diversity of answers. Management and business theorist Henry Mintzberg argues that we ‘cannot afford to rely upon a single definition of strategy.’ Instead, he provides five definitions: strategy as plan, ploy, pattern, position and perspective.113  Various landmark documents of strategy have each contributed their own definitions. Starting in the early nineteenth century, Carl von Clausewitz offered a terse definition, identifying strategy as ‘the use of engagements for the object of war.’114 For Clausewitz, strategy consisted of directing a sequence of battles. Strachan writes, ‘strategy was consequentialist for Clausewitz: one thing followed another, and outcomes and events shaped the next step.’115  Clausewitz’s contemporary (and rival) Antoine-Henri Jomini provided another definition, considering strategy ‘the art of making war upon the map.’116 Later commentators would criticize Clausewitz for his exclusive focus upon the military engagement and upon geometric abstractions. Hart especially thought that battle was not ‘the only means to the strategical end.’117 Yet, despite this admonition, focus upon the engagement cannot be ignored.    Although Jomini was initially the more influential theorist by far, Clausewitz would ultimately surpass him. In addition to focusing upon the military, Clausewitz also argued for strategy’s subordination to policy. War, according to Clausewitz’s well-known aphorism, is ‘merely the continuation of policy by other means.’118  By the mid-nineteenth century, this                                                113 Henry Mintzberg, 'The Strategy Concept I: Five Ps For Strategy' in California Management Review, Fall 1987, p. 11.   114 Carl Von Clausewitz, On War. trans. Michael Howard and peter Paret (Oxford: Oxford University Press, 1976): p. 74. 115 Hew Strachan, 'Strategy and Contingency' in International Affairs 87:6, p. 1287.  116 Antoine Henri de Jomini, The Art of War. trans. G.H. Mendell and W.P. Craighill (Rockville: Arc Manor, 2007), p. 51. 117 B.H. Liddell Hart, Strategy (New York: Meridian, 1991), p.319.  118 Clausewitz (1976), p. 28.  39  influence was discernible in the writings of Helmuth von Moltke the Elder, who defined strategy as ‘[t]he practical adaptation of the means placed at a general’s disposal to the attainment of the object in view.’119 The impact of Clausewitz is evident: the military leader controls the instruments of force, but not the war’s purpose. Strategy serves policy, not vice-versa.   In the twentieth century, Edward Mead Earle conceived of strategy as a more general and encompassing phenomenon, viewing it as ‘ the art of controlling and utilizing the resources of a nation  - or a coalition of nations - including its armed forces, to the end that its vital interests shall be effectively promoted and secured against enemies, actual, potential or merely presumed.’120 This definition, however, tends to conflate ‘grand’ and ‘military’ strategy. In other words, this definition does not make the separation between conduct at the highest levels of power (including the articulation of policy) and martial actions on the ground.  Hart and Gray both define ‘strategy’ in a similar Clausewitzian mould. In essence, Hart views strategy as ‘the art of distributing and applying military means to fulfill the ends of policy.’121 Gray’s definition is almost identical: strategy is understood as ‘the use that is made of force and the threat of force for the ends of policy.’122 Here both Hart and Gray are explicitly concerned with military strategy. They hold ‘grand strategy’ to be conceptually distinct from the actual implementation of organized violence. Likewise, both Hart and Gray draw upon Clausewitz, and emphasise that strategy is subordinated to political ends, while tactics provides the means.   Clearly, ‘strategy’ cannot be succinctly encapsulated by a single definition. Nonetheless, we may identify a number of key themes that each of the cited definitions touch upon. First, certain types of strategy emphasize preparing for and fighting the engagement. Violence is                                                119 Hart, p.320. 120Edward Mead Earle, ‘Introduction,’ in Makers of Modern Strategy (Princeton: Princeton University press, 1943), viii.    121 Hart, 321.  122 Colin S. Gray, The Strategy Bridge: Theory for Practice (Oxford: Oxford University Press, 2010), p. 18.   40  primary in some strategies, and plays a subordinate role in others. Second, strategy should be recognized as subordinate to policy, but still able to organize and manage the diversity of means placed at its disposal. Practically and normatively speaking, why a war is fought ought to be determined by a political elite, not by those whose task it is to implement force. Third, strategy itself is divided into ‘grand’ and ‘military’ strategies. Military strategy is but a component of grand strategy, but it is the most powerful one, because it controls the instrumental use of violence.  3.4 The ‘Strategy Hierarchy’ Analytical Framework   The analytical framework outlined in this section incorporates the three themes highlighted above: strategy’s subordination to policy, the division of grand strategy into multiple overlapping domains of action, and the role of the engagement in military strategy. These elements are understood as foundational to strategy, and each is elaborated in greater detail in chapters 5, 6 and 7. For now, I focus on modelling ‘strategy’ as the central tier of a hierarchy. This ‘strategy hierarchy’ positions strategy between policy (the ends) and the instruments of coercion (the means). Table 3.1 presents this division: Table 0-1 The Strategy Hierarchy Policy (and Ideology)  Ends Grand Strategy  ● Social ● Political ● Economic ● Cultural ● Military  Ways Strategic Behaviors (Tactics and Techniques) Means  Edward Luttwak has identified this hierarchy as the ‘vertical dimension’ of strategy. His framework places the technical and tactical levels at the foundation, followed by operational and  41  theatre strategy, all of which is guided by grand strategy.123 Above grand strategy is the ultimate level, that of policy. Policy provides purpose, without which conflict is nothing more than aimless violence.  Paraphrasing General Maxwell D. Taylor, Lykke writes that ‘[s]trategy equals ends (objectives toward which one strives) plus ways (courses of action) plus means (instruments by which some end can be achieved).’124 These three levels - policy, strategy and tactics - must be integrated in order for each component to be successful. The tripartite division into ends, ways and means may be understood in terms of the three basic questions posed by J. Boone Bartholomees: ‘what is it I want to do, what do I have or what can I reasonably get that might help me do what I want to do, and what is the best way to use what I have to do what I want to do?’125  Another way of viewing this hierarchal division is via the metaphor of a sandwich (Strachan) or a chain (Posen) - where strategy is bounded by policy and tactics, with each separate but kept in delicate balance.126 Gray uses the term ‘strategy bridge’ to signify the bidirectionality of strategy. In his interpretation, strategy is ‘the only bridge built and held to connect policy purposefully with the military and other instruments of power and influence.’127 If a state promotes a particular policy, but lacks the means to implement it, then strategy cannot realize the direction set for it. Conversely, if a state has ample force at its disposal, but lacks a clear purpose, then violence becomes mindless, and strategy has no meaning. Strategy thus bridges ends and means, policy and tactics, each of which exerts influence over the other.                                                 123Edward N. Luttwak, Strategy: The Logic of War and Peace (Cambridge: Belknap Press, 1987), p. 70. 124 Arthur F. Lykke, “Defining Military Strategy,” Military Review 69:5, p.3.  125 J. Boone Bartholomees, Jr., 'A Survey of Strategic Thought' in U.S. Army War College Guide to National Security Policy and Strategy (Strategic Studies Institute, 2004), p. 81.  126 Strachan, The Lost Meaning of Strategy, p. 36. 127 Gray, p. 29. This is his Dictum #3  42   Figure 0-1 The Strategy Hierarchy  3.5 Strategic Behaviors and the Intentionalist/Functionalist Debate  In both war and genocide, it is easy to misrepresent the behavior of the state and its agents as being wholly intentional, and to assert that every action is explicitly animated by preordained ends. Sometimes actors follow the goals set for them; at other times, their actions reflect their own needs and desires, or context exerts an inexorable power over their decision-making. Within Holocaust studies, Bessel notes that the 'intentionalist’ and ‘functionalist’ debate poses the question of ‘whether one should regard the actions of the "National Socialist Regime" as the unfolding of the ideology and expressed intentions of its leadership... or whether one instead should focus on the dynamics of decision-making processes and the institutional pressures inherent in the Nazi system of government.’ 128 While some scholars (such as Ian                                                128 Richard Bessel, “Functionalists vs. Intentionalists: The Debate Twenty Years on or Whatever Happened to Functionalism and Intentionalism?” in German Studies Review, 26:1; p. 15.  43  Kershaw) have argued for the need to ‘synthesize’ these two elements rather than place them in opposition, the distinction is still analytically significant. 129 The contrast between purposeful action and to structural constraints is by no means unique to the Holocaust. The behaviors of those responsible for other genocides - including that in Rwanda - reflect a combination of a) an aggressively pursued, prearranged plan of action, and b) a series of improvisations and expediencies shaped by the context and structures within which actors are immersed. The strategic understanding dimension of the intentionalist-functionalist debate may be summarized in terms of the difference between an original strategy, and an emergent one. Original strategies are understood in terms of plans, or articulated blueprints of coordinated action. In contrast, according to Mintzberg, an emergent strategy refers to ‘patterns developed in the absence of intentions, or despite them.’130 When complex plans confront reality, they face unanticipated ‘frictions,’ the perfidy of self-willed actors, and a host of unforeseen opportunities and obstacles. An original strategy thus derives from intent, while an emergent strategy is the product of function and structure. The intentionalist-functionalist debate finds ready application to the Rwandan genocide, especially when viewed from the standpoint of ‘original’ and ‘emergent’ strategies. On the one hand, the architects of the genocide established ‘kill lists’, made preparations to ensure the survival of the akazu, and struck against the Hutu opposition when the time was right. On the other hand, the concentration of the Tutsi began without central planning, and this ‘emergent’ structure gave shape to the subsequent acts of mass killing.                                                 129 Ian Kershaw, The Nazi Dictatorship. Problems and Perspectives of Interpretation (London: Edward Arnold, 1985), 81. 130 Mintzberg, p.13.  44  3.6 Strategy: Science or Ideal Type?    Strategy invariably pervades the conduct of violent conflict throughout history. Jomini claims that ‘[m]ethods change, but principles are unchanging.’131 Lawrence Freedman traces the ‘elemental features of human strategy that are common across time and space.’132 Luttwak likewise has ‘demonstrat[ed] the existence of strategy as body of recurring objective phenomena that arise from human conflict.’133 Proceeding from this understanding, if ‘strategy’ has transhistorical characteristics, can it be studied scientifically? Clausewitz himself cautioned that the ‘term ‘‘science’’ should be kept for disciplines such as mathematics or astronomy,’ preferring the terms ‘art of war’ over ‘science of war.’134 In practice, strategic behavior is always in flux. Control over the conduct of war is inevitably limited, and unavoidable ‘frictions’ tend to collapse the artificial boundaries that theory imposes. Instead of conceptualizing strategy as a science, we may follow Shaw’s suggestion that we ‘use the conceptual framework offered by the understanding of war, derived initially from strategic theory but developed in a fundamentally sociological framework.’135 Rather than assuming that the strategy hierarchy is amenable to scientific investigation, it is better understood as a Weberian ideal type.136 In this manner, the framework acknowledges both the perennially relevant traits of strategy and the historical contingency of its expression.                                                131 Bernard Brodie, ‘The Continuing Relevance of On War’ in On War, trans. Michael Howard and Peter Paret (Oxford: Oxford University Press, 1989): p. 57. 132 Freedman, p.3,  133 Luttwak, p.267.  134Clausewitz (1976), p.99.  135 Martin Shaw, 'Conceptual and Theoretical Frameworks for Organised Violence', International Journal of Conflict and Violence 3:1, p. 104.  136 Max Weber, “Objectivity in Social Science and Social Policy” in The Methodology of the Social Sciences, E. A. Shils and H. A. Finch (ed. and trans.) (New York: Free Press, 1904/1949), p 90.  45  Gray has captured this duality between ideal and practice in his contrasting of ‘strategy’ with ‘strategies.’ While arguing that ‘[s]trategy has a permanent nature,’ he defines ‘strategies’ as ‘plans, formal or informal,’ that ‘have a variable character driven, but not mandated, by their unique and changing contexts.’137 In other words, he suggests that while strategy possesses a permanent and unbending body of eternal truths; conversely, strategies are specific, contextualized episodes of organized violence By examining strategy as an ideal type, we may highlight particular strategies deployed at certain moments, and show how they were affected by influences such as culture and personality. For instance, we can speak of General Vo Nguyen Giap’s role in Vietnamese strategies to confront the United States, or the cultural context of France’s pre-WWII strategies of permanent fortifications. Set within distinct historical circumstances, these context-dependent manifestations retain a timeless essence. We can look beyond the particular context, and perceive how strategies follow certain principles, especially those laid down in Gray’s twenty-one dicta.  In short, ‘strategy’ is an ideal type, an unattainable state that reflects general features (such as those encapsulated in Gray’s 21 dicta). By contrast, ‘strategies’ represent real-world practices that adhere - in widely varying degrees - to this universal standard. Strategy, in short, is not a science but an aspiration.   3.7 A Caveat: Adversarial and Non-Adversarial Enemies   Now that strategy has been defined and operationalized, it is necessary to highlight an important caveat. War and strategy assume the presence of an adversary, i.e. a foe capable of opposition and resistance. For Clausewitz, war is ‘nothing but a duel on a larger scale.’138 The                                                137 Gray, p. 41.  138 Clausewitz (1976), p.13.   46  enemy could represent an existential risk to a polity, or it could comprise a perennial, albeit mostly harmless, nuisance. Regardless, the enemy is in some way armed, organized, and capable of offensive and defensive action. This conventional conception of war assumes, as Walzer puts it, ‘combat between combatants.’139 The same cannot be said for targeted populations that are unarmed or poorly armed, and disorganized. These groups are largely incapable of collective action, and their ‘resistance’ to genocide is generally localized, ineffectual, and one-sided. ‘War,’ writes Clausewitz, ‘... is not the action of a living force upon a lifeless mass (total nonresistance would be no war at all) but always the collision of two living forces.’140 This general symmetry of force capabilities fundamentally distinguishes conventional war from genocide. Because of what Dadrian calls a ‘critical disparity of power relations,’ genocidal violence appears as non-adversarial.141 Or, to rephrase Clausewitz’s metaphor, genocide is not a duel, but an execution on a gargantuan scale.  This leaves us with a fundamental conundrum: if strategy assumes an adversary, and genocide is conducted against non-adversarial targets, then is it possible for genocide to be framed as strategic? Fortunately, Martin Shaw has already accomplished the intellectual heavy lifting on this theme. It will be recalled that Shaw understands genocide as a type of war. Echoing Clausewitz, he argues that both war and genocide ‘aim to destroy the power of the enemy through means that are pre-eminently violent.’142 The ‘key difference’ between war and genocide ‘is the nature of the enemy: in war, another state or armed force; in genocide, a civilian social group.’143 According to Shaw, genocide and degenerate wars occur when non-adversarial groups are targeted and identified ‘as an enemy in an essentially military (rather                                                139 Michael Walzer, Just and Unjust Wars (New York: Basic Books, Inc., 1977), p. 42. Italics in original.  140 Clausewitz (1976), 15-16.  141 Vahakn N. Dadrian, 'The Determinants of the Armenian Genocide' Journal of Genocide Research 1:1, p. 65. 142 Martin Shaw, What is Genocide? (Cambridge: Polity Press, 2007), p. 35.  143 Shaw (2007), p.   47  than political, economic, or cultural) sense.’144 It is this process of singling out, Shaw writes, that specifies ‘against whom it is justified to use violence in a comprehensive and systematic way.’145  ‘Since genocide is also a kind of war,’ writes Shaw, ‘the notion of combatants does not disappear.’146 He continues:   The killing of civilians generally accompanies or follows the destruction of armed elements among the enemy population. And since terror and slaughter always invoke resistance, there are nearly always some real combatants. But genocidal power usually imagines that these are far more numerous than they really are: that every young man is an actual or potential resister, that every woman and older man is aiding and abetting resistance, that every community and settlement provides support.147   The key to understanding the genocide victim’s transition from member of a harmless social category to military threat resides in the perpetrator’s ideology and genocidal logic. This logic ‘expands to embrace whole populations, from apparently conventional military starting-points. Quasi-military discrimination of “combatants” and potential combatants among the enemy population leads, through circuits of ideology as well as the crudities of practice, to wholesale terror and slaughter.’148 How this ‘logic’ comes to determine policy is elaborated in chapter 5.   Sémelin echoes Shaw’s argument, further stressing the role of radical beliefs. ‘There is no doubt,’ he writes, ‘ that in situations of serious crisis, the ideological discourse propagated by                                                144 Shaw (2007), p. 35. 145 Shaw (2007), p. 35. 146 Shaw (2007), p. 35. Italics added.  147 Martin Shaw, War and Genocide: Organised Killing in Modern Society (Wiley, 2003), p. 175.  148Shaw (2003), p. 175,   48  the power in place (or by groups seeking this power) proposes a “reading” of the situation, designates the “threats” and calls for a collective mobilization towards their destruction.’149  Ideology is understood to be crucial in designating who constitutes a threat. International law ostensibly protects those groups defined in the Convention on the Prevention and Punishment of the Crime of Genocide, namely, national, ethnical, racial or religious groups. However, in practice, perpetrator regimes extend this definition to include groups not covered by the Convention, or bearing multiple and overlapping identities.150 As mentioned in the definition of genocide in the introductory chapter, we may add political, social, gender and economic groups to this list. Ideology informed the targeting of Armenian and Greek Christians in Ottoman Turkey, of Buddhists, Cham, Vietnamese and ‘new people’ in Democratic Kampuchea, and of Jews, Gypsies and Slavs in the Third Reich. In the end, regardless of who constitutes the enemy, it is the perpetrator that defines who, as Weitz puts it, ‘could be categorized as dangers to the well-being of the dominant groups.’151   3.8 Non-Adversaries and Strategy    Ideology not only outlines a vision of a better world or superior order; it also specifies the social categories that stand in the way of this vision. Ideology that identifies non-adversarial enemies perverts policy when it penetrates the corridors of power and is internalized by a cadre influential elites. As strategy is beholden to policy and the ends it sets, strategy comes to assume that the ideological ‘other’ is a threat that requires a violent counter. In other words,                                                149 Jacques Semelin, 'Toward a Vocabulary of Massacre and Genocide,' Journal of Genocide Research 5:2.  p. 196.  150 Adam Jones, Genocide: A Comprehensive Introduction (New York: Routledge/Taylor and Francis Publishers, 2010), p. 34.  151 Eric D. Weitz, A Century of Genocide: Utopias of Race and Nation (Princeton: Princeton University Press, 2003), p. 237.    49  instead of targeting soldiers and military infrastructure, strategy extends its violence to manifestly non-adversarial groups.   What all groups targeted for genocidal violence share (other than a common identity, real or imputed) is that they lack the means of effective self-defence or counteroffensive. Table 3.3 presents traits that distinguish adversarial from non-adversarial enemies.  This thesis extends Shaw and Sémelin’s threat-creation argument by emphasizing the role that strategy plays. The concept of strategy, as we have seen, includes both ‘grand strategy’ and ‘military strategy.’ When grand strategy dominates, these targeted groups are recognized as a threat, and structural or other indirect means are implemented to diminish their strength. This is the phase in which groups, in Shaw’s words, are defined as enemies in a ‘political, economic, or cultural’ sense.152 Military means may still be present, but as background elements. However, once groups are understood as a military threat, that is, as combatants or adversaries, then military strategy comes to feature more prominently.   Table 0-2 Adversarial vs. Non-Adversarial Enemies Adversarial enemies Non-Adversarial enemies Members of an established organization (State actor, violent non-state actor, state proxy, guerrilla force, etc)  Members of a perpetrator-defined category (national, ethnic, racial, religious, political, socio-economic, former combatant, etc.)  Organized (hierarchical): command and control structure, strategic planning and implementation)  Dis- or un-organized (while they may possess some coherence at a community level, or at minimum, they are effectively defined and categorized by the perpetrator)  Armed (they are in physical possession of the means of violence, are trained to use them properly, and are supplied by a logistic support apparatus)  Unarmed (Not the same as defenseless. Non-adversarial targets are capable of self, but they generally lack the means, the training, and ability to sustain defense or go on the counteroffensive).  Found on the frontlines or in friendly territory as infiltrators Found in situ, in flight, in hiding or in concentrations                                                  152 Shaw (2007), p. 35.   50  Although non-adversarial groups lack the means and wherewithal to resist effectively, they do not necessarily respond with passivity and inaction. Even if the situation appears hopeless, targeted groups will attempt to resist destruction, often with extraordinary vigour and determination. Examples include the Warsaw ghetto uprising, the siege of Musa Dagh during the Armenian genocide, and (as will be shown below) defense among the hills of Bisesero in Rwanda. Alongside these famous episodes are countless acts of small-scale resistance that may never be documented. Regrettably, these moments of resistance, both large and small, bolster the perpetrator's ideological mindset. Artificial threats become reified in acts of resistance, as ideology amplifies perceptions of risk.    With this recognition, we may transcend the ‘caveat’ discussed earlier, and appreciate the ideological process by which non-adversarial enemies are transformed into military threats. By defining a group as such a threat, even if it remains an imaginary one, a military strategy is increasingly likely to be pursued. To return to the question posed earlier: can genocide be considered strategic even if its primary focus is upon non-adversarial targets? The answer is affirmative. From the perpetrator’s perspective, these enemies, despite their objective innocence, represent a collective threat requiring a violent solution.  3.9 Summary  This chapter has sought to reconcile strategic theory with an understanding of the genocidal process. The task is rendered more complex both by the sheer volume of literature on strategy, and by strategic studies’ inherent aversion to examining violence against civilians. Following Arendt, the link that binds war, genocide and strategy is the use of instrumental violence. Such a purposeful use of force functionally requires strategy to reconcile ends and means.  51  Strategy, war and genocide are conceived of as ancient and interrelated practices. As such, ‘strategy’ is a diffuse concept, difficult to pin down analytically. Drawing upon the contributions of multiple scholars, I argue (1) that ‘strategy’ is subordinated to policy, or the ends set by elites in power; (2) that ‘grand strategy’ utilizes all of the instruments placed at a state’s disposal, while ‘military strategy’ is primarily concerned with coordinating the implementation of force; and (3) that the engagement, or actual moment of fighting, is vital to military strategy, even as it is of variable significance in the grand strategic and policy framework. The heuristic of the ‘strategy hierarchy’ seeks to clarify the interrelationship among ends, ways and means.  This chapter departs from one rule prevalent in strategic thought. Strategy assumes an adversary, a foe capable of credible opposition and intrigues of their own. During episodes of genocide, genuine adversaries are absent; instead, via a process of adversary creation, genocidal ideology comes to interpret social categories as an existential threat requiring a violent solution. In short, non-adversarial groups are transformed into ‘legitimate’ targets. With this framework in place, we may begin to analyze specific genocidal episodes, such as the 1994 holocaust in Rwanda, in strategic terms.   52  Chapter 4: The Strategic Context:  Political Violence in Rwanda (1959 to 1990)  As argued in section 3.8, in order to understand the dynamics of particular strategies, it is necessary to explore their unique contexts. To reiterate, Gray argues that the ‘general theory of strategy applies to all projects in strategy-making and execution, at all times and places.’ In contrast, the ‘content of those strategies and their fates must be individual.’153 Strategies, therefore, are always unique. According to Gray, they are the product of seven contexts: political, social-cultural, economic, technological, military, geographical and historical.154 Each of these contexts adds to the singularity of both deliberate and emergent strategies.   The purpose of this section is to establish the context within which an extremist elite, beholden to Hutu Power ideology, came to comprehend certain non-adversarial social categories as strategic threats. This task entails highlighting key episodes of politically motivated violence against Tutsi and opposition figures.  The story of Rwanda in the twentieth century may be summarized in terms of social revolution and revenge. Initially favored by Belgian colonial authorities, the Tutsi ruling minority lost their command of the state’s institutions to the majority Hutus. After independence in 1962, the hegemony of the numerically superior Hutu over the Tutsi was to be a pernicious factor in Rwandan politics. While the 1973 coup by Juvénal Habyarimana brought some respite to the beleaguered Tutsi, the situation was irrevocably altered with the RPF invasion from Uganda in October 1990.  In order to understand the violence that enveloped Rwanda in 1994 it is necessary to reconstruct the milieu that spawned it. This is true for three reasons. First, this history of erstwhile Tutsi superiority and subsequent refugee raids provided the foundation for extremist ideological discourse. A selective interpretation of the past justified both structural discrimination                                                153 Colin S. Gray, The Strategy Bridge: Theory for Practice (Oxford: Oxford University Press, 2010), p. 39.  154 Gray, p. 38. This is Gray’s 8th Dictum.   53  and massacre against the Tutsi. It would later (see chapter 5) facilitate a transformation in perpetrator perceptions whereby Tutsi and opposition figures were converted from a social irritant to a military threat. Second, analysis of the historical context reveals a continuity of particular strategic behaviors against both ethnic Tutsi and opposition figures. It also reveals a discontinuity: up until the genocide, churches provided sanctuary from politically motivated violence. These violent activities anticipated the genocide of 1994. Third, the Habyarimana coup shifted power to a coterie of Hutu elites in the north-east, largely centered on Gisenyi. This regional faction of Hutu extremists controlled the civil service and armed forces, and would ultimately come to constitute the hard core of the génocidaire leadership.    4.1 “Race” in Pre-Colonial and Colonial Rwanda   Long before the arrival of Western powers, the principal factor which determined an individual’s social classification was occupation. The Hutu were considered cultivators who worked the land, while the Tutsi were cattle-herders. As cattle ownership came to be seen as a sign of wealth, occupational differences gradually evolved into social cleavages. The Tutsi, as the cattle owning class, came to dominate the Hutu across the spectrum of social relations. Economically the Tutsi held the wealth in the form of livestock; politically, the Tutsi directed the centralization of a nascent state; and culturally the Hutu were portrayed as inferiors. Defining the Hutu-Tutsi distinction along racial lines is highly problematic. They are both Kinyarwanda-speaking Banyarwanda people that possess the same culture and occupy the same physical territory.155 They were not separate “tribes”. Rather, identification was malleable and could change according to an individual’s socio-economic status. Both groups coexisted, intermarried and went to the same schools and churches. The physiological differences that                                                155 Mahmood Mamdani, When Victims Become Killers: Colonialism, Nativism, and the Genocide in Rwanda (Princeton: Princeton University Press, 2001), p. 52.  54  evolved between the two groups were the result of different diets: the Hutu ate grains, while the Tutsi had a steady diet of meat.  By the late nineteenth century, ‘racially-obsessed’ European anthropologists attempted to explain the distinction.156  The Hutu were portrayed as simple, jovial and thick-set; in contrast, the Tutsi were described as possessing ‘a refinement of feelings which is rare among primitive people.’157 Adding pseudo-scientific credibility to this view was John Hanning Speke. In his 1863 Journal of the Discovery of the Source of the Nile, Speke describes his ‘theory of conquest of inferior by superior races.’158 According to this view, the Tutsi were a ‘conquering race’ with a ‘superior civilization’ from southern Ethiopia.159 According to Prunier, the problem with Speke’s ‘Hamitic origins’ hypothesis was that it was theorized ‘without a shred of evidence.’160 In the closing decades of the nineteenth century, two waves of colonial administration - first Imperial Germany, and then Belgium - reinforced the prevailing view that the Tutsi were intrinsically superior to their Hutu kin. The Belgians entrenched Tutsi hegemony by replacing the vast majority of Hutu chiefs with Tutsi. By 1959, 43 out of 45 chiefs and 549 out of 559 sub-chiefs were Tutsi.161 Adding to the subordinate status of the Hutu was the introduction in 1933 of mandatory passports that specified an individual’s ethnicity. Significantly, even after decolonization, these passports were retained by the new government and were still in place by 1994.  The reinforcement of this artificial “racial” distinction had serious repercussions. Prunier writes that ‘[t]he result of this heavy bombardment with highly value-laden stereotypes for some sixty years ended by inflating the Tutsi cultural ego inordinately and crushing Hutu feelings until                                                156 Gerard Prunier, The Rwanda Crisis: A History of Genocide (Columbia University Press, 1995), p. 6.  157 Prunier, p. 6.  158 Prunier, p. 6.  159 Prunier, p. 7.  160 Prunier, p. 7.  161 Prunier, p. 27.   55  they coalesced into an aggressively resentful inferiority complex.’162 Ultimately, during the 1990s, racial categories were utilized and exploited as prevalent themes within the extremists’ ideological discourse. By April 1994, writes Scott Straus, ‘acute insecurity and orders from above ignited a categorical logic of race and ethnicity… neighbors became enemies in war and under the authorities’ direction.’163  4.2 Revolutionary Violence (1959)  o the end of the 1950s, the traditional Tutsi hold upon state power was overthrown by a Belgian-backed social revolution. In July 1959, the reigning mwami (king) died under mysterious circumstances, and his successor was appointed by an elite clique of Tutsi aristocracy.164 Hutu activists were incensed by what they perceived as a coup. It is in this context that the political party MDR-PARMEHUTU - the ‘Party of the Hutu Emancipation Movement’ - was established.  On November 1st, 1959, members of the Tutsi party UNAR and anti-Belgian monarchists were rumored to have assaulted a Hutu sub-chief and important PARMEHUTU member Dominique Mbonyumutwa in Gitarama.165 PARMEHUTU turned the Mbonyumutwa incident into a rallying call for anti-Tutsi, anti-elite violence. Almost immediately, riots and mob violence against the Tutsi spread throughout Rwanda. A contemporary witness wrote that ‘PARMEHUTU mobilised its followers and for the first time the Hutus rose en masse against their feudal lords.’166 Hutu retaliated by attacking a Tutsi chief who belonged to UNAR, and the                                                162 Prunier, p. 9.  163 Scott Straus, The Order of Genocide : Race, Power, and War in Rwanda (Ithaca: Cornell University Press, 2006): p. 173.  164 Aaron Segal, 'Massacre in Rwanda' Fabian Research Series 240 (April, 1964), p. 9.  165 Charles K. Mironko, 'Ibitero : Means and Motive in the Rwandan Genocide' in Genocide in Cambodia and Rwanda: New Perspectives ed. Susan E. Cook (New Brunswick: Transaction Publishers, 2009), p. 174. 166 Segal, p. 9.   56  situation degenerated into a state of civil war.167 By 1960, a visiting UN mission estimated that 200 people had been slain.168 Fearing reprisals from Tutsi authorities against the Hutu majority, Belgium stepped in to suppress the violence. Their imposed solution ultimately took the form of a social revolution: half of Tutsi chiefs were replaced by Hutus, helping to ensure a Hutu victory in Rwanda’s first elections. By 1961, Grégoire Kayibanda, founder of PARMEHUTU, was installed as the first President of a Rwandan republic.169  While the primary targets of this revolt were Tutsi chiefs and local officials, it would be wrong to construe these events as spurred merely by ethnic tensions. Rather, Caplan argues that ‘this series of events is most accurately regarded as a class uprising rather than as a first step toward genocide.’170  Nonetheless, perceptions of class and ethnicity were inseparable, as the ethnic Tutsi were habitually equated with the abuse of power and privilege.  Violence in 1959 and after was marked by the burning of houses, theft of property, the killing of cattle and the forcing into flight of thousands of Tutsi. This process would be repeated, with varying intensity, up until April 1994. Violence often occurred in broad daylight and in full view of the police.171 For instance, in mid-October 1960, after a fracas between a police officer and a Tutsi drifter in Kibingo commune, thirteen Tutsi were massacred.172 For the Tutsi, fear of Hutu impunity increasingly came to cast a shadow over interethnic relations.   It would be wrong to contend that this violence was solely the result of spontaneous and disorganized outbursts. Mironko argues that it was not a knee-jerk response at a popular level,                                                167 Dixon Kamukama, Rwanda Conflict: Its Roots and Regional Implications (Kampala: Fountain Publishers Ltd., 1997), p. 31. 168 Mamdani, p. 123.  169 Mamdani, p. 124.  170 International Panel of Eminent Personalities, Rwanda: The Preventable Genocide (African Union, 2000), p. 17.  171 Linda Melvern, A People Betrayed: The Role of the West in Rwanda's Genocide (London: Zed Books, 2009), p. 17.  172 Prunier, p. 52.   57  but the enactment of the Rwandan institution of ibitero, Kinyarwandan for a ‘group response to a situation of danger.’ He writes that ‘[a]s a chronic form of violence against Tutsi, [ibitero] has been misidentified as spontaneous rioting... when, in fact, it has a recognizable history as a mechanism of organizing group attacks dating back to 1959.’173 As a method for inciting violence at the level of the cellule (the lowest administrative level), the ibitero was an effective means for mobilizing and directing communities to serve the particular needs of those in power - or, in 1959, those aspiring to power. PARMEHUTU developed the technique in which they would ‘contact the most influential Hutu on each isolated hillside and make him responsible for his location.’174  When harsher measures were required, organized and directed violence could essentially be ‘turned on and off’: the word used to describe the nature of these massacres was muyaga, or wind, as violence would ‘come suddenly and forcefully and then, just as suddenly as it came, it would stop.’175   4.3 Inyenzi Raids and Retaliation against the Tutsi  The ‘social revolution’ and the establishment of the Rwandan republic led to the flight of tens of thousands of Tutsi into neighboring countries. Forced abroad, these refugees harbored revanchist sentiments, and some formed military bands that launched raids into Rwanda from foreign bases along the border. The Tutsi exodus planted the seeds for the recurrence of conflict so long as these refugees remained armed, angry, and dispossessed of their homeland. The truism that ‘[c]onflicts that generate refugees can easily lead to conflicts generated by refugees’ returned repeatedly to hamper reconciliation between the Hutu and the Tutsi.176                                                 173 Mironko, p. 174.  174 Segal, p. 10.  175 Joseph Sebarenzi, God Sleeps in Rwanda: A Journey of Transformation (Simon and Schuster, 2009), p. 14.  176 ‘Preventable Genocide,’ p. 18.    58  Over the years 1959 to 1973, three hundred thousand Tutsi fled to neighboring Uganda, Tanganyika (later Tanzania), Burundi and Zaire.177 They left Rwanda in three waves: 1959-61 1963-64, and 1973.178 The numbers of refugees increased as the newly installed Hutu government of Grégoire Kayibanda retaliated against Tutsi who remained in Rwanda. Domestically, the Tutsi UNAR party was severely repressed by the PARMEHUTU regime. UNAR members told a U.N. commission in 1962 of ‘mass arrests of their members and acts of brutality against its officials.’179  Killings were deemed to be ‘systematic and seemed designed to eliminate the UNAR.’180   After the establishment of the Republic in 1961, members of UNAR and other aggrieved exiles organized themselves as a guerrilla force and conducted cross-border raids against the newly installed Hutu officials.181 They called themselves inyenzi (cockroaches) in tribute to their alleged resilience. For about eighteen months, the inyenzi limited their activities to agitation and cattle rustling, and violence between the Rwandan government and the exiles was limited.182 Against the Tutsi who stayed in Rwanda, however, violence continued: 150 Tutsi were killed in September-October 1961 in Butare.183 Between March 1961 and July 1966, the inyenzi carried out ten major armed assaults into Rwanda proper.184 These raids had no hope of succeeding so long as Belgium continued to support the Rwandan army. Regardless, their impact was felt on the countryside, inculcating a climate of fear among the rural population.185 In November 1963, 3,000 inyenzi, including women and children, invaded Rwanda from                                                177 Sebarenzi, p. 13.  178 Mamdani, p. 160.  179 Melvern (2009), p. 19.  180 Melvern (2009), p. 19. 181 Kamukama, p. 32.  182 Segal, p. 11.  183 Prunier (1995), p. 53. 184 Kamukama, p. 32.  185 African Rights, Rwanda: Death, Despair and Defiance (London: African Rights, 1995), p. 12.  59  Burundi. At the time, it was uncertain what their motivations were and whether ‘it was the intention of the inyenzi leaders actually to invade Rwanda or to provoke a slaughter of civilians by Rwandan troops in the hope that this would incite world opinion against the Rwanda[n] regime.’186 Ultimately, the invaders were routed by the Belgian-led National Guard.187  Another inyenzi raid occurred in December of the same year. Armed with bows and arrows and homemade rifles, they again crossed from Burundi to the vicinity Nemba, in Bugesera prefecture. They gathered Tutsi still living in Rwanda to their cause, and with around 600 men they attacked the Agako military base, then under construction. The invasion was stopped fifteen miles shy of Kigali, whereupon the refugee force was largely eliminated by Rwandan soldiers and Belgian military advisors.188 Again, the exiles suffered defeat with the loss of hundreds of lives.  Des Forges contends that Hutu leaders turned these raids into mechanisms ‘to bolster the sense of Hutu solidarity, to solidify their own control and to eradicate the last vestiges of respect for Tutsi authority.’189 The reaction of the First Republic to these inyenzi raids was to target Tutsi still in the country for further repression. Tutsi were harassed because they were considered ‘active or potential support for the inyenzi,’ as they would again be after the invasion of 1990.190 Such vilification of all Tutsi became standard practice. The government responded to these raids with various forms of repression. Prominent Tutsis were rounded up and imprisoned. A list of names was found on the body of a Congolese mercenary who had accompanied the Tutsi incursions. According to the Fabian Society, ‘[t]he list presumably contained the names of the government the invaders hoped to install and the Rwandan authorities promptly executed the leading members of the opposition monarchist                                                186 Segal, p. 13.  187 Melvern (2009), p. 20.  188 Segal, p. 13.  189 Alison Des Forges, ‘Leave None to tell the Story’: Genocide in Rwanda (New York: Human Rights Watch, 1999), p. 36. 190 Mamdani, p. 129.  60  UNAR ministers, UNAR deputies in the National Assembly and several leading Tutsi civil servants mentioned in the list.’191 At this time, roadblocks emerged as a form of controlling movement, and travel permits (which were conspicuously denied to members of opposition parties) were required to travel between prefectures.192   A severe instance of repression took place between February and May 1962 in the region of Byumba. Violence erupted after two inyenzi raids caused the deaths of half a dozen Hutu, including police and civil servants. The response by local Hutus was astonishingly disproportionate to the initial offense. Lemarchand writes that ‘between 1,000 and 2,000 Tutsi men, women and children were massacred and buried on the spot, their huts burned and pillaged and their property divided among the Hutu population.’193 Additionally, after the aborted December 1963 invasion, an estimated 10,000 Tutsi were killed over the next two months.194  After 1963, Mamdani suggests, ‘the repression of Biumba [sic] provided a norm: a raid turned into a signal for the massacre of the local Tutsi population, and for the distribution of their property among those organized as the local self-defence group.’195 Such use of organized massacre was found to be very effective in halting inyenzi attacks, which ceased altogether in 1967.196 However, the decline of the military threat was not matched by a relaxation of Tutsi oppression. There were sporadic massacres in 1967.197  By that same year, Des Forges writes, ‘when both the incursions and the attacks on Tutsi within Rwanda ended, Tutsi were at risk of attack for the simple fact of being Tutsi.’198                                                 191 Segal, p. 14.  192 Melvern (2009), p. 19.  193 René Lemarchand,  Rwanda and Burundi (London: Pall Mall Press, 1970), p. 219. 194 Prunier (1995), p. 57. 195 Mamdani, p. 129. 196Alan J. Kuperman, The Limits of Humanitarian Intervention: Genocide in Rwanda (Washington D.C.: Brookings Institution Press, 2001), p. 7.   197 African Rights, p. 12.  198 Des Forges (1999), p. 37.   61  4.4 Genocide in Burundi and the Habyarimana Coup  Events to the south in Burundi just prior to the Habyarimana coup had tremendous implications for Rwanda. Both Burundi and Rwanda shared the same Tutsi-Hutu divide, and both gained independence from Belgium in the same year. However, the difference between the two states was fundamental. In Burundi the Tutsi retained and extended their power, and frequently used violent methods to repress the Hutu majority. In many ways, each neighbor embodied the other’s greatest fear: the domination of a rival ethnicity at the expense of the other.  Additionally, in both countries, this fear was manipulated to serve specific political purposes. Caplan writes that, ‘[v]ictimization of the Tutsi in one country was first aggravated by, and then used to justify, persecution of the Hutu in the other country and vice versa. Each act of repression in the one state became the pretext for a renewed round of killing in the other.’199  In response to a rural Hutu uprising in late April 1972, the government and army of Burundi began a systematic attempt to exterminate every Hutu accused of being implicated. What followed, according to Lemarchand, ‘was not so much a repression as a hideous slaughter of Hutu populations.’200 Hutus considered to be ‘elite’ (i.e. ‘local administrators, chauffeurs, clerks and skilled workers’) were targeted, rounded up, and concentrated in jails, where they faced beatings and murder.201 Singled out for particular brutality were those involved in education: thousands of primary and secondary school teachers were slaughtered along with their Hutu students. By August 1972, an estimated 200,000 Hutu were estimated to have been slain during these months of murder.202                                                 199 ‘Preventable Genocide, p. 20. 200 Lemarchand (1975), p. 11. 201 Lemarchand (1975), p. 11. 202 René Lemarchand, ‘The Burundi Killings of 1972’, Online Encyclopedia of Mass Violence (27 June 2008), p. 6.  62  Politically active Hutus in Rwanda closely followed events in Burundi. Both the incumbent Kayibanda regime and those aspiring to power manipulated news of the violence in Burundi to strengthen their positions.203 In particular, the PARMEHUTU government used the violence to bolster its authority. The regime was increasingly perceived as incompetent and generally favoring their base in southern Rwanda – much to the chagrin of Hutu in other regions, particularly north-westerners from Gisenyi.204 Conversely, Juvénal Habyarimana - a northerner - used the climate of violence as a justification to restore order and establish his rule.    In Rwanda, violence began as an attempt the purge the seminaries and university of Tutsi. Since opportunities for employment in the state and military were largely reserved for Hutus, Tutsi sought advancement in other institutions, such as the church and in education. In 1973, Hutu students attending the National University in Butare turned against Tutsi students and teachers. This pattern extended into persecution of Tutsis working at banks, for parastatal companies and in private businesses.205 Significantly, at this time Radio Rwanda was first used to incite Hutus against the Tutsi.206 According to African Rights, ‘[t]he government disclaimed responsibility, but its hand was clearly to be seen behind the violence.’207 Conversely, it was rumored at the time that Habyarimana had instigated much of the violence in order to justify his coup d'etat.208    Habyarimana’s ascension to the presidency shifted power from southern Rwanda to the northeast. Two years after the coup, in 1975, he formed the National Republican Movement for Democracy and Development (MRND). Every Rwandan, regardless of age, was automatically a member of the party.209 His government established ten administrative districts, or prefectures,                                                203 Melvern (2009), p. 25. 204 Kuperman, p. 7.  205 Mamdani, p. 137. 206 Mamdani, p. 137.  207 African Rights, p. 13.  208 African Rights, p. 13. 209 Des Forges (1999), p. 42.   63  each headed by a préfet; each of these units was further divided into separate communes, led by bourgmestres. Although bourgmestres ranked below the préfets, Des Forges claims in practice they ‘exercised more immediate and pervasive power over the ordinary people than did [their] superiors.’210 They were charged with resolving property disputes and family grievances, and hiring local employees and communal policemen. Each bourgmestres was appointed directly by the President Habyarimana. ‘At the local level,’ Des Forge writes, ‘he was clearly and directly the president’s man out on the hills.211 While it may sound disingenuous to say conditions for the Tutsi improved with the Habyarimana regime, especially given the horrors of 1994, violence against Tutsis subsided until the 1990 RFP invasion. Prunier calls these ‘The Good Years,’212 and not without reason: while systematic discrimination was pervasive and entrenched, there were no pogroms after the 1973 coup.213  While there were reports of massacres of Tutsi refugees around the Luwero Triangle in Uganda in 1982, Rwanda’s Tutsis did not suffer direct state-organized violence.214  4.5 The Role of Sanctuaries  From a post-genocide perspective, possibly the most significant aspect of the violence prior to 1994 was the role of sanctuaries in limiting the number of fatalities. During times of conflict and upheaval, both Tutsi and Hutu would take shelter in the churches scattered across the country. Philip Gourevitch quotes one Rwandan on this tradition: ‘When there were problems, people always went to the church… the pastors were Christians. One trusted that                                                210 Des Forges (1999), p. 43.  211 Des Forges (1999), p. 43.  212 Prunier (1995), p.74. 213 African Rights, p. 13. 214 Ogenga Otunnu, ‘Rwandese Refugees and Immigrants in Uganda’, in The Path of a Genocide: The Rwanda Crisis from Uganda to Zaire, eds. Howard Adelman and Astri Suhrke (New Brunswick, New Jersey: Transaction Publishers, 2000), p.17.   64  nothing would happen at their place.’215 During the tumult of the social revolution, inyenzi raids, genocide to the south and the Habyarimana coup, Rwandans were able to find reliable protection at these locations. Roméo Dallaire writes that ‘[c]hurches had always been a place of sanctuary in Rwanda.’216 During the recurring cycles of violence, refuge at these sites became a protective tradition, and it was understood to be taboo to harm those assembled in these sanctuaries. While hundreds were killed from 1959 to the early 1970’s, there were no extensive massacres at these locations.  After the first inyenzi invasion in 1963, for example, harassed Tutsi fled to the churches. Segal remarks that ‘[w]ithin Rwanda thousands of Tutsi took refuge at Catholic Mission stations where they believed they would be safe... There were no reprisals against these internal refugees.’217 While the Kayibanda government desired that the refugees return home (in one instance unsuccessfully nailing church doors shut to keep people out), the Tutsi remained unmolested.  A decade later in 1973, while Tutsi were attacked in educational and other public institutions, they found refuge in the church. Timothy Longman notes that ‘the fact that Tutsi who fled to churches in 1973 received sanctuary and that the sanctity of churches was generally accepted assured many Tutsi when the genocide broke out two decades later that they would again be safe in churches. Leaders of the genocide actually exploited this belief in order to gather their intended victims in a central location to facilitate their slaughter.’218 The high number of violent deaths in 1994 was the result of the explicit violation of this norm. Assuming that tradition still held, hundreds of thousands of Rwandans gathered en                                                215 Philip Gourevitch, We Wish to Inform You That Tomorrow We Will Be Killed With Our Families (New York: Farrar, Straus and Giroux, 1998), p.26.  216 Romeo Dallaire, Shake Hands With the Devil: The Failure of Humanity in Rwanda (Random House, 2008): p. 190.  217 Segal, p. 16.  218 Timothy Longman, Christianity and Genocide in Rwanda (Cambridge: Cambridge University Press, 2010): p.87.   65  masse at these sites, where many were invariably slaughtered. These massacres marked a major departure in the history of violence in Rwanda, and represent a significant strategic innovation.   4.6 Summary   This brief overview of Rwandan history highlights episodes of politically motivated violence from independence in 1959 until the October 1990 invasion by the Rwandan Patriotic Front. The primary aim of this background survey is to develop the strategic context that preceded the development of genocide in 1994. Each of the historical episodes covered - the social revolution, the inyenzi raids, genocide in Burundi, and the establishment of the Habyarimana regime - had a significant impact upon the strategies that shaped the 1990-94 war and the subsequent full-scale genocide. From this background analysis we may identify recurring patterns of behavior that foreshadowed the genocide of 1994. These include the use of violence against both Tutsi and oppositional elites for instrumental ends. The violence was not ‘spontaneous’ and frequently evinced central control; it often involved the complicity of local elites who used their authority to bestow impunity to the perpetrators; and similar tactics (looting, arson, abuse) were used to inculcate a climate of fear. Later, we will show how other historically rooted institutions were utilized against the ethnic Tutsi and opposition figures, including the use of the ethnic identity cards, roadblocks and the use of the radio to incite and organize massacre. While it is possible to discern a historical continuity in the types of violence deployed, one salient feature did not carry over to the genocide: the protective role of churches. The tradition of sanctuary limited the extent of violence until the genocide’s outbreak, and after the assassination of the President Habyarimana, it came to facilitate a skyrocketing mortality.  Two categories of threat took root in the extremist imagination. First, the Tutsi and their  66  camps on foreign soil posed an ever-present menace. It was an easy task for the propagandist to link the inyenzi raids to a conspiracy of irredentist Tutsi seeking to overturn the social revolution of 1959. Second, opposition from rival Hutu elites challenged the regime. After the Habyarimana coup, former members of MDR-PARMEHUTU, especially from its southern powerbase, were viewed as suspect. It is in this context that non-adversarial social categories were ultimately misconstrued as military threats.        ’              67  Chapter 5: Policy, Ideology and the Rwandan Elite   Figure 0-1 The Strategy Hierarchy (Policy)  This chapter first aims to define policy and the elites who are ultimately responsible for its development. Policy is understood as an ensemble of objectives formulated by a class of self-interested political elites. In non-democratic regimes, one of the primary aspirations of these elites is to protect and preserve their status and privileges vis-à-vis those perceived to challenge their continued dominance.  The chapter then turns to the debate between Clausewitz and Jomini regarding who should be in control of strategy and the conduct of war: military leaders or political elites. Clausewitz takes the normative position that political elites should control war. By extending Clausewitz’s logic, this section contends that, like war, genocide is a thoroughly political affair. When political elites control the conduct of war and genocide, their ideological viewpoints may have an impact on the course of the conflict. For instance, when elites are exposed to a genocidal ideology, non-adversarial social categories gradually become perceived as an existential threat. Even though the targeted groups are poorly armed, unaggressive, and disorganized, genocidal ideology informs strategy (via policy) regarding who should be targeted.   68  The second half of the chapter utilizes this framework to identify the elites who dominated Rwanda prior to the genocide, as well as their self-serving policies, and the ideological foundations of those policies. The ‘akazu’ was a clique of Hutu from the north-west of Rwanda that possessed inordinate influence over politics, the economy, the army and the administration. However, from the mid-1980s, the elite was beset by three connected ‘strategic shocks’: economic decline, invasion by the RPF, and demands for democratic concessions. This elite group viewed the challenge to their rule through an ideological prism, ultimately conflating the legitimate adversary (the RPF) with non-adversarial social groups (ethnic Tutsi and opposition Hutu).   Enemy creation requires a steady stream of propaganda and demagoguery if it is to be successfully internalized. The chapter concludes by documenting the campaign of misinformation and hatemongering waged against both Tutsi and moderate Hutu. Four sources of enemy creation are identified: the newspaper Kangura, broadcasts over RTLM, circulation of a hate discourse by military elites through the ENI commission, and Mugesera’s incendiary rhetoric, which spread these messages to the party faithful.   5.1 Policy and Policymakers  This section defines ‘policy’ as a series of goals formulated by a class of self-interested elite actors. These goals structure both what is desirable and what is attainable via grand strategy. Policy shares with strategy and genocide a veritable glut of definitions and lack of conceptual clarity. One conventional definition holds it to be ‘[t]he general principles which guide the making of laws, administration, and executive acts of government in domestic and international affairs.’219 In a specifically martial sense, Koliopoulos and Platias write that policy is                                                219 Roger Scruton, The Palgrave Macmillan Dictionary of Political Thought (London: Palgrave Macmillan, 2007), p. 544.  69  the ‘governing mind behind strategy [that] sets the aims that strategy will subsequently be called upon to achieve.’220 Likewise, Gray defines policy as the ‘objectives that provide the purposes of particular historical strategies.’221 Policy is rooted in a vision of what a state hopes to achieve, and is held to be a ‘desired condition that serves to inspire.’222 Policy, in short, provides the justification for why a course of action (i.e. strategy) is undertaken.  Policies and the goals they prescribe are inseparable from the elite actors who construct them. It is this group of elites that possesses the de facto authority to determine the desired ends to be achieved. In non-democratic regimes, this group constitutes a hegemonic stratum composed of the architects and chief beneficiaries of policy objectives. Gaetano Mosca defines this preeminent segment of society as the ‘political class.’ Mosca testifies to the influence and Power of this group: ‘[w]e all know that, in our own country, which ever it may be, the management of public affairs is in the hands of a minority of influential persons, to which management, willingly or unwillingly, the majority defer.’223 Members of this class dominate the state apparatus through both official and unofficial positions. This ‘political class,’ in sum, is chiefly responsible for the formulation of policy. One of the primary aspirations of the preeminent class is to establish policies which protect and preserve their status and privileges. In other words, the maintenance and preservation of power is integral to policy. The US Marine Corps doctrinal text on strategy states that the ‘making of policy is a conscious effort by a distinct political body to use whatever power it possesses to accomplish some purpose—if only the mere continuation or increase of its own power.’224 The policies developed by this political body are intended to realize the shared vital interests of the political class.                                                 220 Athanassios G. Platias & Constantinos Koliopoulos, Thucydides on Strategy (London: Hurst, 2010), p. 4.   221 Colin S. Gray, The Strategy Bridge: Theory for Practice (Oxford: Oxford University Press, 2010), p. 18. 222 Gray, p. 18.  223Gaetano Mosca, The Ruling Class (New York: McGraw-Hill Book Company, Inc. 1939), p. 50.  224 U.S. Marine Corp, MCDP 1-1: Strategy (Washington D.C: Department of the Navy, 1997): p. 11-12.  70  Additionally, policy goals establish the framework for permissible action by both restraining and enabling grand strategy. What matters to the political class is whether or not policy promotes strategies that best protect their interests. If these strategies successfully bring about the realization of policy aspirations, then the actions are to be encouraged; alternatively, coordinated activities that run contrary to policy goals are to be eschewed. Along these lines, Kenneth Waltz writes that ‘success is the ultimate test of policy, and success is defined as preserving and strengthening the state.’225    5.2 Policy’s Normative Control over the Strategy Hierarchy  As outlined in chapter 3, policy stands at the apex of the strategy hierarchy. This is a normative position, and is the product of a longstanding debate over who should control strategy: political elites or military leaders.  Statesmen and theorists have long been familiar with the interplay between the recognized sovereign and his or her use of violence in wartime. For instance, Louis XIV was fond of the expression Ultima ratio regum (force is the last argument of kings), which reflects both the instrumental nature of organized force and its relation to a higher political authority.226 However, after the Napoleonic wars, the previously mentioned rivalry (see 3.3) between Carl von Clausewitz and Antoine-Henri Jomini came to dominate the discourse on strategic theory. Both of these seminal writers analyzed the relationship between policy and strategy, but disagreed about which party should be ascendant in the actual conduct of war.   For Jomini, while it was the responsibility of the political leadership to broadly set the overall purpose of a conflict, it should refrain from meddling with the minutiae of waging war.                                                225 Kenneth N Waltz, Theory of international politics (Reading, Mass.: Addison-Wesley Pub. Co., 1979): p. 117.  226 Samuel Arthur Bent, Familiar Short Sayings of Great Men (Boston: Ticknor and Co., 1887).   71  Jomini writes in the Précis de l'art de la guerre (Art of War) that ‘[a] general whose genius and hands are tied by an Aulic Council five hundred miles distant cannot be a match for one who has liberty of action, other things being equal.’227 Jomini’s principles held, according to Strachan, that ‘once war was declared the statesmen should fall silent until the general delivered the victory.’228 Handel, likewise, depicts Jomini as affirming that ‘operational and military considerations’ should have primacy ‘over political control once war has begun.’229 While the general staff (or more generically, those who control the use of force) do not determine the initial purpose of war, Jomini suggests that they ought to be free to fight and win on their own terms. In short, Jomini argues for war free from extraneous political interference.   In contrast, Clausewitz and his book vom Kriege (On War) diverged from Jomini, contending that policy is inseparable from the conduct of war. For Clausewitz, war is a rational and instrumental act only insofar as it serves a political purpose. His famous aphorism that ‘[w]ar is not a mere act of policy but a true political instrument, a continuation of political activity by other means,’ dramatically departs from Jomini’s thesis.230 Clausewitz contends that ‘When whole communities go to war… the reason always lies in some political situation, and the occasion is always due to some political object.’231 Clausewitz’s originality arguably resides in his understanding that ‘[policy] will permeate all military operations, and, in so far as their violent nature will admit, it will have a continuous influence on them.’232 Politics does not take a backseat to those conducting strategy, as it might in Jomini’s writings, but continually exerts an                                                227 Antoine Henri de Jomini, The Art of War trans. Capt. G.H. Mendell and Lieut. W.P Craighill (Rockville, MD: Arc Manor, 2007): p. 31. 228 Hew Strachan, "The Lost Meaning of Strategy" in Survival: Global Politics and Strategy 47:3, p. 36.  229 Michael I. Handel, Masters of War: Classical Strategic Thought (Portland: Frank Cass Publishers, 2001), p.54. 230 Carl Von Clausewitz, On War. trans. Michael Howard and peter Paret (Oxford: Oxford University Press, 1976): p. 86-87. 231 Clausewitz, 86–87.  232 Clausewitz, 86–87.  See also Daniel Moran Strategic, ‘Theory and the History of War’ in Strategy in the Contemporary World (New York: 2003):p. 18.   72  influence. ‘Subordinating the political point of view to the military would be absurd,’ he writes, ‘for it is policy that has created war.’233 Clausewitz is careful to clarify that policy does not meddle with ‘operational details’ (such as posting of guards or the employment of patrols). It is, however, ‘influential in the planning of war, of the campaign, and often even of the battle.’234 ‘If policy reads the course of military events correctly,’ he writes, ‘it is wholly and exclusively entitled to decide which events and trends are best for the objectives of the war.’235  Both Jomini and Clausewitz’s positions on the proper role of policy can be grafted onto the strategy hierarchy. Clausewitz represents the hierarchy in its ideal form, with policy dominant over war strategy. In contrast, Jomini represents a broken hierarchy in which policy has much less of an impact, and elites are much less influential.   5.3 Genocidal Ideology and Policy  Following the argument made earlier (see 2.2) that genocide is a type of war, we may extend Clausewitz's line of reasoning to say that policy impacts upon every aspect of both: just as policy shapes the conduct of war, policy is intrinsic to the genocidal process. The impact of political elites can typically be felt in every aspect of genocide, from its preparation to its implementation and subsequent cover-up. Daniel Goldhagen echoes this view, arguing that ‘[g]enocide is a political act; it is a policy of the state.’236  The setting of policy goals is seldom a purely objective or rational process. Rather, the choice of objectives is shaped by the perceptions and the idiosyncratic worldview held by elite actors. These groups subjectively define what is important. Bernard Brodie is right to contend                                                233 Clausewitz, p. 254.  234 Clausewitz, p. 253. Italics added.  235 Clausewitz  p. 255.  236 Daniel Jonah Goldhagen, Worse Than War: Genocide, Eliminationism and the Ongoing Assault on Humanity (New York: PublicAffairs, 2009): p. 68-69.   73  that ‘the importance of vital interests comes not necessarily from some intrinsic quality, but rather from what we are ready to do about some infringement of them, real or imagined.’237 For elites determining policy, the dividing line between ‘real or imagined’ is often obscured by the dominant ideology.  While all power is guided by ideology and ideological thought is not inherently harmful238, it assumes negative traits when harmless civilian groups are identified as obstructing policy goals. An ideology becomes explicitly genocidal when it identifies non-adversarial groups as threats necessitating a generalized coercive solution. Helen Fein coined the term ‘ideological genocide’, which she defines as motivated by ‘an ideology, myth, or an articulated social goal which enjoins or justifies the destruction of the victim.’239  A genocidal ideology impacts upon policy in three ways. First, it furnishes policy with goals. Second, it determines who the enemy is. Third, it distorts how crises are interpreted and handled. Genocidal ideologies stand in stark contrast to their milder cousins (such as liberalism) because they provide a comprehensive set of beliefs while prescribing an idealistic (and sometimes lethally grandiose) course of action. When extremist beliefs are held as articles of faith by a political elite, ideology provides the fundamental basis for policy. Ideology and policy become indistinguishable as these beliefs are incrementally internalized by a radicalized elite, spread by sympathetic ideologues, and implemented by a committed bureaucracy. Valentino notes that ‘what the ideologies that lead to mass killing share is not their specific content but the magnitude, scope and speed of changes they force upon large groups of people.’240  In addition to influencing the ends of policy, ideology identifies adversaries, such as vulnerable minorities and opposition groups that are seen to challenge these imaginative                                                237 Bernard Brodie, War and Politics (London : Cassell, 1973): p. 342. Italics added. 238 Adam Jones, Genocide: A Comprehensive Introduction (New York: Routledge/Taylor and Francis Publishers, 2010): p. 596-598.  239 Helen Fein, “Defining Genocide as a Sociological Concept” in Current Sociology 38: 8 (1990): p. 27.  240 Benjamin A. Valentino, Final Solutions: Mass Killing and Genocide in the 20th Century (Ithaca: Cornell University Press, 2004), p. 72.  74  schemes or to threaten the power base of those in control. Parties to large-scale mass killing typically exhibit exclusionary beliefs and values alongside any positive aspects of their programme. National Socialism, Communism, Pan-Turanism, and Hutu Power, each in their own perverted way, supplied an ethos that informed and motivated actions while identifying threats. This aspect of the enemy-creation role of ideology has already been elaborated (sections 3.9 and 3.10).  Ideology also shapes how elites react to unexpected events or crises. According to Freier, strategic shocks ‘jolt convention to such an extent that they force sudden, unanticipated change in the... perceptions about threat, vulnerability, and strategic response.241 Other scholars have defined these shocks, or ‘surprises,’ as ‘game-changing events,’ wherein ‘the rules that were previously in place no longer apply.’242 What these events share is their dramatic impact on policy and grand strategy. They may elicit a change in relative priorities towards - or away from - the use of instrumental violence.  Examples of strategic shocks include the 1941 Japanese attack on Pearl Harbor, or the attacks by al-Qaeda on September 11, 2001. In both cases (and many more besides), policy was reoriented from a nominal focus on peaceful strategies to a state of war. Some of the most dramatic strategic shocks may thus act as a casus belli, provoking and justifying war. Alternatively, strategic shocks may de-escalate conflicts. One recent example is the collapse of the Soviet Union; the policies and grand strategy of both superpowers and their dependents changed priority, and war or the threat of war came to play less of a role in their relations.  Elites interpret strategic shocks through an ideological prism. With the admixture of a genocidal ideology, events that can be explained by mundane causes are exaggerated and                                                241 Nathan Freier, Known Unknowns: Unconventional ‘Strategic Shocks’ in Defense (Carlisle, Pa: U.S. Army War College, 2008): p. vii.  242Peter Schwartz and Doug Randall, "Ahead of the Curve: Anticipating Strategic Shocks" in Blindside: How to Anticipate Forcing Events and Wild Cards in Global Politics, ed. Francis Fukuyama (Washington, D.C.: Brookings Institution, 2007): p. 94.   75  willfully misinterpreted to fit the elite’s weltanschauung. Moreover, ideology tends to exacerbate the effects of strategic shocks by inventing conspiracies and designating responsible enemies. For example, the Gallipoli Campaign arguably sparked the roundup of Armenian intellectuals, ushering in the genocide of Ottoman Christians. Another example: the Holocaust gained intensity after unanticipated German defeats in the Soviet Union. Finally, the assassination of Habyarimana in Rwanda was a strategic shock that ushered in 100 days of mass killing.      5.4 Rwanda: The Emergence of the Akazu   Following the Clausewitzian ideal outlined above, the Rwandan genocide was a thoroughly political enterprise. Policy, shaped by the Hutu Power ideology, was dominant in shaping the subsequent course of events: it prioritized the goal of regime survival, it identified non-adversarial enemies in the form of Tutsi and opposition Hutu, and it distorted how the elites responded to unanticipated events. The Rwandan genocide, following the model established in chapter 3, represents the strategy hierarchy in its ideal form. Policy, informed by a genocidal ideology, provided direction for grand strategy and the diverse means at the state’s disposal. This section first identifies the Rwandan political elites who orchestrated the genocide, before turning to the series of crises that threatened their dominance.  Juvénal Habyarimana’s coup in 1973 shifted political power from Gitarama and Butare in southern Rwanda to an alternative power base in the northeast. This move unseated Grégoire Kayibanda and his PARMEHUTU party (see chapter 4), while distancing southern Hutu from positions of influence and power. Habyarimana solidified his position by formally establishing the Mouvement Révolutionaire National pour le Développement (MRND) in 1975. Surrounding the MRND was a cadre of regional elites who came to dominate the regime behind the scenes. Based in Gisenyi and Ruhengeri, this clique was centered on the family of the president’s wife,  76  Madame Agathe Habyarimana.243 Although she possessed no formal authority, she has been widely regarded as the true source of power behind Juvénal’s throne.244 This group has been variously named ‘Clan de Madame’ and the ‘akazu’ (‘little house’). This elite clique claimed historical legitimacy from its alleged connection to a previous Hutu dynasty centered in the north-east.245  This class of political elites actively cultivated - and jealously protected - its privileged existence. As a direct consequence of its political dominance, Prunier contends that northern Hutu were assured a steady stream of ‘cabinet posts, economic opportunity, and foreign scholarships.’246 He identifies two sources by which this elite grew rich: exporting coffee, tea, and tin, and skimming off foreign aid.247 Longman notes that Habyarimana’s home prefectures of Gisenyi and Ruhengeri received public monies vastly disproportionate to their populations.248The akazu also included an extensive network of influential retainers, including Théoneste Bagosora and other figures instrumental in orchestrating the genocide.249 This group benefited from appointments to military leadership and offices at the prefectural and communal level.   5.5 Rwanda: Strategic Shocks   Beginning in the late 1980s, a series of interconnected challenges threatened the entrenched northern elite.  Initially, this shock was felt as, to use Straus’s phrase, a ‘deep,                                                243African Rights, Rwanda: Death, Despair and Defiance (London: African Rights, 1995), p. 22. 244 Linda Melvern, A People Betrayed: The Role of the West in Rwanda's Genocide (London: Zed Books, 2009), p. 47. 245 Helen M. Hintjens, "Explaining the 1994 Genocide in Rwanda" in The Journal of Modern African Studies 37:2, p. 259.  246 Gerard Prunier, The Rwanda Crisis: A History of Genocide (Columbia University Press, 1995), p. 86.  247 Prunier, p. 85.  248 Timothy Longman, Christianity and Genocide in Rwanda (Cambridge: Cambridge University Press, 2010): p. 120.  249 Prunier, p. 85.   77  multifaceted, and escalating crisis.’250 This crisis derived from three sources: economic decline, military invasion from the north, and demands for political liberalization.  First, in addition to an AIDS crisis, famine in 1988-89, and rampant overpopulation, the dominant clique confronted a dramatic economic downturn.251 When both coffee and tin prices collapsed in the mid-1980s, these elites were drained of a major source of income.252 This crisis was further exacerbated by foreign donors pressuring for ‘structural adjustment’ packages.253 The dominant clique was warned that, unless it amended its penchant for corruption and nepotism, it would lose access to this windfall of hard currency.  The second aspect of this crisis began on 1 October, 1990, when a largely Tutsi force of refugees calling itself the Rwandan Patriotic Front (RPF) invaded from Uganda. The presence of Banyarwanda (see 4.1.) in the Ugandan National Resistance Army (NRA) had been a source of concern to the MRND since the mid-1980s. These fears were confirmed when a large number of soldiers splintered from the NRA to form the RPF.254 Initially, the RPF invasion was a failure. The leader of the military invasion, Maj. Gen. Fred Rwigyema, was killed only a day after setting foot in Rwanda.255 Additionally, with French support, the Forces Armées Rwandaises (FAR) was able to repel RPF advances, forcing the refugee army to flee to Akagera National Park to regroup.256 But under the leadership of Major Paul Kagame, the RPF renewed its offensive in early 1991. The RPF transformed itself into a credible force with a renewed focus on guerrilla engagements.  In January 1993, RPF forces attacked the northern town of Ruhengeri and vanquished FAR forces there with relative ease.257 With each RPF victory came                                                250 Scott Straus, The Order of Genocide : Race, Power, and War in Rwanda (Ithaca: Cornell University Press, 2006): p. 31.  251 Longman, p. 125.  252 Prunier, p. 85.  253 African Rights, p. 23.  254 Dixon Kamukama, Rwanda Conflict: Its Roots and Regional Implications (Kampala: Fountain Publishers Ltd., 1997): p. 44. 255 Kamukama, p. 46. 256 Prunier, p. 96. 257 Kamukama, p. 54.   78  greater political demands from the rebel army, and every inch of Rwandan territory taken represented a mounting existential threat to the continued rule of the Habyarimana government, the akazu and its supporters.  The third component of the crisis came in the form of demands for multipartyism. The end of the Cold War brought with it a wave of democratization across Africa, and individuals within and outside of Rwanda began calling for an end to the MRND’s unopposed political monopoly. Four months prior to the RPF invasion, Habyarimana agreed to form a commission to examine the question of democratization.258 After the RPF attack, the nascent political opposition was emboldened to push for more and faster change. Additionally, while French military aid was critical in keeping the RPF at bay, it came with strings attached: the French warned that such assistance would not last forever, and that if the government were to survive, Rwanda would have to democratize.259 In June 1991, Habyarimana begrudgingly accepted a constitutional amendment that allowed the formation of independent political parties.260 Among these parties was the Mouvement Démocratique Républicain (MDR), which assumed the historic mantle of Kayibanda’s PARMEHUTU party. These parties immediately began agitating for a share of the power and pushed for inclusion in a coalition government, a demand grudgingly conceded in April 1992.261  The domestic political situation was further exacerbated by the demands of the RPF made during negotiations at Arusha, Tanzania. The Arusha Accords, signed in August 1993, further threatened the powerbase of the akazu. According to the terms of the agreement, the MRND had to share power in a Broad Based Transitional Government (BBTG) before                                                258 Alison Des Forges, ‘Leave None to tell the Story’: Genocide in Rwanda (New York: Human Rights Watch, 1999): p. 41. 259 Alan J. Kuperman, The Limits of Humanitarian Intervention: Genocide in Rwanda (Washington D.C.: Brookings Institution Press, 2001), p. 9.  260 Des Forges, Leave None, p. 44. 261 Des Forges, Leave None, p. 44. and Straus, Order of Genocide, p. 24.  79  nationwide multiparty elections were held.262 The Arusha Accords also challenged the MRND’s control over the military; the RPF was to assume half of the officerships and supply 40% of the regular soldiers.263 The power-sharing agreement would strip Habyarimana of most of his executive powers and transfer them to a cabinet of opposition parties, including the MDR.264  Each of these crises threatened to undermine the hegemony of the political elite. Not only did they fear losing their privileges, there was the real possibility that under a new regime they would be held accountable for their criminal practices. Kinzer identifies these criminal acts as ‘drug rackets, prostitution, extortion, bribery and murder’.265 In 1993, an international team issued a report on human rights violations since 1990 that specifically named elite Hutus responsible for killings and massacres.266 Worse yet, elites feared that a Tutsi takeover of government could cost them their lives. In 1993, the first Hutu President of Burundi, Melchior Ndadaye, was assassinated by Tutsi military officers.267 The next month, in Rwanda, Frouald Karamira, the head of MDR-Power faction, was also assassinated.268  The Tutsi, from the elite’s perspective, would never accept Hutu rule. MRND stalwarts and akazu loyalists did not respond passively. Instead, Caplan contends that they turned the assault against their powerbase into an ‘an opportunity to consolidate their eroding support and to mobilize international backing for the war the invaders had begun.’269 In the aftermath of the October invasion, Habyarimana pursued what Catharine Newbury calls a ‘two-track policy.’ ‘On the one hand, responding to pressure from western donors, he permitted a gradual process of political liberalization and made concessions to an active internal pro-democracy movement.                                                262 Straus, Order of genocide, p. 24. 263 Straus, Order of genocide, p. 24. 264 Kamukama, p. 61.  265 Stephen Kinzer, A Thousand Hills : Rwanda's Rebirth and the Man who Dreamed it (Hoboken, N.J.: John Wiley & Sons, 2008): p. 92. 266 African Rights, p. 33.  267 Straus, Order of Genocide, p. 29-30.  268 Straus, Order of Genocide, p. 30. 269 International Panel of Eminent Personalities, Rwanda: The Preventable Genocide (African Union, 2000), sec. 7.6.   80  Simultaneously Habyarimana permitted (or pursued) a policy of internal repression as part of the war strategy, and he allowed (or encouraged) a proliferation of human rights abuses.’270 How this ‘two-track policy’ took shape is elaborated in the following chapter.   5.6 Creating the Adversary    The interrelated crises highlighted above encouraged the growth of a genocidal ideology that would provide the ultimate foundation for policy and the mass killing that followed. The three components of this strategic shock - the loss of economic privileges, RPF invasion, and push for democracy- all contributed to the enemy-creation process.  From the perspective of the MRND and the akazu, three identifiable formations threatened the regime. The first was the nascent political opposition, especially those from the former MDR-PARMEHUTU stronghold in southern and central Rwanda. The second was the armed incursions of the Rwandan Patriotic Front. As the RPF grew in strength, it came to threaten the regime in Kigali directly. The third threat derived from a pervasive suspicion of Tutsi supporters of the RPF within Rwanda. During the four years from 1990 to 1994, this category ultimately expanded to include every Tutsi as a potential collaborator. In the minds of the extremists, the RPF invasion initiated the Tutsi transition from ethnic minority to what Prunier describes as an ‘ontological ibyitso’ (accomplice).271  As emphasized above, ideology is not inherently harmful, or even avoidable. However, once ideology identifies non-adversarial groups as enemies requiring destruction, then an otherwise benign system of thought may turn malevolent. In Rwanda, the growth of a genocidal ideology facilitated the generalized transition of ethnic Tutsi and political opposition into the                                                270 Catharine Newbury, "Background to Genocide: Rwanda" in Issue: A Journal of Opinion 23:2 (1995): p. 14.  271 Prunier, p. 231.  81  same adversarial threat that the RPF represented. All three categories of enemies were amalgamated in extremist ideological discourse and propaganda. Over these four years, the entire Tutsi ethnicity and the political opposition came to be associated with the RPF, and the more they challenged elite hegemony, the more they came to be perceived as a military threat requiring a military solution.   By examining the propaganda spread by Hutu Power ideologues, we may identify the ideological process of creating non-adversarial enemies. Propaganda was insidious and pervasive. Caplan writes that it came variously from ‘political rallies, government speeches, newspapers, and a flashy, new radio station, [which] poured vicious, pornographic, inflammatory rhetoric designed to demonize and dehumanize all Tutsi.’272 By analysing this extremist discourse, it is possible to document the interrelated and evolving demonization of ethnic Tutsi and the political opposition. Significantly, all of these sources of enemy creation can be linked to the MRND and akazu loyalists. While by no means comprehensive, four sources of enemy creation stand out: articles in Kangura, RTLM broadcasts, the ENI commission, and the Mugesera speech.   5.6.1 Print: Kangura and the ‘Hutu Ten Commandments’    One critical aspect of democratization during the early 1990s was the relaxation of state control over the press.273 This liberalization resulted in a proliferation of publications, including the extremist paper Kangura (‘Wake It Up’), first published in 1990. The paper was described by Hassan Ngeze, the owner, founder and editor, as 'a voice of the Hutu.'274 Ngeze was commissioned by Madame Agathe Habyarimana and the akazu to counter the recently                                                272 International Panel of Eminent Personalities, sec. 7.20.  273 International Panel of Eminent Personalities, sec. 7.19.  274 Judgement, Ferdinand Nahimana, Jean-Bosco Barayagwiza and Hassan Ngeze (ICTR-99-52-T), Trial Chamber 1, 3 December 2003 § 136.  82  established liberal press.275 The akazu used the publication as an ideological mouthpiece to push the agenda of the rubanda nyamwinshi (Kinyarwanda for ‘majority people,’ i.e. the Hutu).276 Although Ngeze’s editorial line occasionally ran counter to akazu interests, the periodical espoused a persistent and virulently anti-Tutsi discourse.277 In order to reach the widest possible audience, the paper was distributed to the nation’s burgermeisters at no charge.278 Two months after the October 1990 invasion, Kangura printed the infamous ‘Hutu Ten Commandments.’ It contained harrowing claims that ‘[t]he Batutsi are bloodthirsty and power-hungry and want to impose their hegemony on the people of Rwanda using armed force.’279 According to this document, Hutu men should be forbidden to associate with Tutsi women, and every Tutsi was dishonest and only looking out for his or her ethnic kin.280 Ominously, the Hutu were encouraged to ‘cease having pity for the Tutsi.’281 The ‘fifth commandment’ possessed a grand strategic element: it states that ‘[s]trategic positions in the political, administrative, economic, military and security domain should, to a large extent, be entrusted to Hutus.’282 It added that education and the military should be exclusively Hutu domains. The final commandment contained a call to action: it insisted that the ‘1959 social revolution, the 1961 referendum [that abolished the monarchy] and the Hutu ideology must be taught to Hutus at all levels. Every Hutu must propagate the present ideology widely.’283 To this end, the ‘ten commandments’ were repeated ad nauseam on RTLM, and read                                                275 Philip Gourevitch, We Wish to Inform You That Tomorrow We Will Be Killed With Our Families (New York: Farrar, Straus and Giroux, 1998): p. 85. 276 Judgement, Ferdinand Nahimana, Jean-Bosco Barayagwiza and Hassan Ngeze § 136. 277 Gourevitch, p. 86.  278 Gourevitch, p. 87.  279 Kangura, Appeal to the Bahutu Conscience (December 6 1990). Online: http://www.rwandafile.com/Kangura/k06a.html 280 Kangura, Appeal to the Bahutu Conscience.  281 Kangura, Appeal to the Bahutu Conscience.   282 Kangura, Appeal to the Bahutu Conscience.  283 Kangura, Appeal to the Bahutu Conscience.   83  by Hutu and Tutsi alike.284 One ICTR prosecution witness claims that because of the Ten Commandments, ‘Hutu started perceiving the Tutsi as enemies instead of seeing them as citizens.’285 In the pages of Kangura, regular themes included explicit calls for massacre, vitriolic attacks against the Tutsi, lists of perceived enemies that required elimination, and threats against Hutu who continued to cooperate with Tutsi.286 Kangura published other equally incendiary articles and diatribes against domestic opposition parties. In one piece, a writer for Kangura made the link between RPF intrigues and the political opposition. The author claimed to have intercepted a document that ‘lays out a strategy that shows them [the RPF] conspiring with other Parties from the so-called opposition’ and aiming to ‘destabilize the government and the socio-economic sector of the nation.’287 There was some truth to this claim: the MDR Party, successor to PARMEHUTU, had a mutually beneficial relationship with the RPF.288 However, the extent of the collusion was exaggerated. Nonetheless, the article promoted a discourse that explicitly linked the political opposition to a hostile armed adversary.  The cover of Kangura No.26 was singled out for special attention during the ‘Media Case’ at the post-genocide Arusha tribunal.289 The first headline stated, ‘If We Re-Launched the 1959 Hutu Revolution to Triumph Over the Tutsi Cockroaches,’ and was followed by a second headline asking, ‘What Weapons Shall We Use to Conquer the Cockroaches Once and for All?’290 These words were evocatively positioned alongside a drawing of a machete, a photo                                                284 Kinzer, p. 93.  285 Judgement, Ferdinand Nahimana, Jean-Bosco Barayagwiza and Hassan Ngeze § 141. 286 International Panel of Eminent Personalities, sec. 7.62.  287 Raphael Zayana, "The Inkotanyi’s New Strategy of Widespread Destabilization" in Kangura 8 (1992). Online: http://www.rwandafile.com/Kangura/k08a.html 288 Alan J. Kuperman, The Limits of Humanitarian Intervention: Genocide in Rwanda (Washington D.C.: Brookings Institution Press, 2001), p. 10.   289 Judgement, Ferdinand Nahimana, Jean-Bosco Barayagwiza and Hassan Ngeze § 160. 290 Judgement, Ferdinand Nahimana, Jean-Bosco Barayagwiza and Hassan Ngeze § 160.  84  President Kayibanda, and heavily armed soldiers.291 While the question raised was not explicitly answered, it clearly amounted to a call for a second revolution.292  5.6.2 Radio: Radio-Télévision Libre des Mille Collines    A significant portion of the Rwandan population was illiterate, making radio an essential means of reaching the masses.293 In mid-1993, Radio-Télévision Libre des Mille Collines (RTLM) took to the airwaves. Like Kangura, RTLM received funding and was owned by influential members of the akazu. These included close relatives of the president, two Cabinet ministers, and top militia leaders.294 RTLM rapidly gained a large following by combining popular music with entertaining programmes entirely in Kinyarwanda, the language spoken by every Rwandan.295  Ferdinand Nahimana was a prominent public figure who used RTLM to inculcate a genocidal ideology. He was a virulently anti-Tutsi historian who was responsible for much of the work of setting up the radio station.296 In 1992, while speaking on Radio Rwanda, he spread the rumor of an impending attack on the Hutu in the Bugesera region. Spurred on by these ‘warnings,’ three hundred Tutsi were killed.297 On RTLM, Nahimana stepped up his sectarian rhetoric. According to Caplan, in the months prior to the genocide, Nahimana and his co-presenters ‘raised the bar of permissible hatemongering’, and their broadcasts ‘played a prominent role in keeping passions at a fever                                                291 Kangura No. 26. Online: http://www.rwandafile.com/Kangura/k26.html 292Judgement, Ferdinand Nahimana, Jean-Bosco Barayagwiza and Hassan Ngeze § 161.  293 Alison Des Forges, "Call to Genocide: Radio in Rwanda, 1994" in The Media and the Rwanda Genocide ed. Allan Thompson (Kampala: Fountain Publishers, 2007): p.42.  294International Panel of Eminent Personalities, sec. 7.21. 295Alain Destexhe, Rwanda and genocide in the twentieth century (New York: New York University Press, 1995): p. 30. 296 Des Forges "Call to Genocide: Radio in Rwanda, 1994", p. 42.   297 Linda Melvern, Conspiracy to Murder: The Rwandan Genocide (Verso, 2006): p. 28.  85  pitch.’298 Current events were willfully misrepresented in order to spread the message of fear and divisiveness. For example, after the assassination of the Burundian Hutu president by Tutsi military officers, RTLM claimed he had been tortured and castrated. This erroneous description of his death (he was killed by a bayonet to the chest) was intended to elicit repugnant memories of Tutsi superiority and to link them to the present crisis.299  Des Forges lists the recurring themes in RTLM’s broadcasts, including ‘the inherent differences between Hutu and Tutsi, the foreign origin of Tutsi and, hence, their lack of rights to claim to be Rwandan, the disproportionate share of wealth and power held by Tutsi and the horrors of past Tutsi rule.’300 Moreover, she writes, RTLM repeated the message that Hutu ought to be vigilant against Tutsi subversion, and urged Hutu to prepare to protect their interests.301 In addition to demonizing the Tutsi, Nahimana bolstered the legitimacy of his political patrons. Hintjens writes that Nahimana’s ‘studies of pre-colonial Bahutu kingships in the north-west of Rwanda provided the intellectual respectability and historical legitimacy that the akazu craved, by supporting their claims of Bantu purity.’302  5.6.3 The ENI Commission   In December 1991, President Habyarimana established a think-tank of ten military officers who were given the task of studying and responding to the question: ‘What must be done in order to defeat the enemy militarily, in the media, and politically[?]’303 The most ominous aspect of the commission is how ‘the enemy’ came to be defined. According to a document produced by this ‘Bagosora Commission,’ the principal enemy was ‘the Tutsi inside or outside                                                298 International Panel of Eminent Personalities, sec. 7.23. 299 Des Forges "Call to Genocide: Radio in Rwanda, 1994", p. 45. 300 Des Forges "Call to Genocide: Radio in Rwanda, 1994", p. 45. 301 Des Forges "Call to Genocide: Radio in Rwanda, 1994", p. 45. 302 Hintjens, p.259. 303 Judgement, Theoneste Bagosora, Gratien Kabiligi, Aloys Ntabakuze, Anatole Nsengiyumva (ICTR-98-41-T 18), Trial Chamber 1, 18 December 2008 § 198.    86  the country, extremist and nostalgic for power, who have NEVER recognized and will NEVER recognize the realities of the 1959 social revolution and who wish to reconquer power by all means necessary, including arms.’ The commission established categories of people from whom enemy partisans were said to be recruited, including Tutsi inside the country, Hutu political opponents, and foreigners married to Tutsi wives.304 While the results of the commission were initially distributed to the military elites, the document was intentionally spread to the rank-and-file in September 1992, shortly after the signing of the first Arusha Accord.305 According to the prosecution at Théoneste Bagosora’s trial, ‘this document is evidence of conspiracy because the final document took a legitimate purpose (defining the enemy) and shifted it to an illegitimate, criminal purpose (targeting the Tutsis).’306 Des Forges contends that ‘[n]owhere did it caution against confusing the RPF as a political group with Tutsi as an ethnic group.’307 Moreover, she notes that ‘the word “Tutsi” is used 14 times in the document and interchangeably in some places with “enemy”, and there are generalisations which may indicate that the Tutsis were unified behind the single ideology of Tutsi hegemony.’308  5.6.4 Léon Mugesera's Party Speech   During a party meeting on 21 November 1992, MRND vice president Léon Mugesera delivered his infamous speech.309 Mugesera was a high ranking elite, and his utterances were                                                304 Alison Des Forges, ‘Leave None to tell the Story’: Genocide in Rwanda (New York: Human Rights Watch, 1999), p. 62-63.  305 Judgement, Theoneste Bagosora, Gratien Kabiligi, Aloys Ntabakuze, Anatole Nsengiyumva § 198.  306 Judgement, Theoneste Bagosora, Gratien Kabiligi, Aloys Ntabakuze, Anatole Nsengiyumva § 199.  307  Des Forges, ‘Leave None’ p.  308 Judgement, Theoneste Bagosora, Gratien Kabiligi, Aloys Ntabakuze, Anatole Nsengiyumva § 204. 309 Melvern, Conspiracy, p. 39.    87  tantamount to an official policy statement. This speech was widely circulated and, according to Gourevitch, ‘most Rwandans can still quote from his famous speech quite accurately.’310  Mugesera first turned his attention to the opposition parties, who, he claimed, were ‘busy trying to injure the President of the Republic’ and ‘working against us.’311 Parties such as the MDR were ‘accomplices of the inyenzis.’312 Mugesera turned his wrath upon the Tutsi within Rwanda and declared that they should be returned to Ethiopia via the Nyabarongo River.313 He cautioned that ‘there are “Inyenzis” in the country who have taken the opportunity of sending their children to the front, to go and help the Inkotanyi.’314 During his diatribe, Mugesera persistently stressed that threats abounded, and that the Hutu should rise up to defend their lives and interests. ‘Do not be afraid,’ he declared; ‘know that anyone whose neck you do not cut is the one who will cut your neck.’315   5.7 Summary  The first part of this chapter focused upon policy, policymakers, and their relationship to strategy. Policy, it was argued, provides goals to be pursued strategically. Behind these policies lie a class of political elites who establish the broad policy objectives, typically aimed at preserving their status and privileges. The choice of objectives may be shaped by elite ideology. Such an ideology may be benign; however, if it identifies non-adversarial enemies as requiring destruction, then it becomes genocidal.  The contrast between Jomini and Clausewitz provides two perspectives on the proper role of policy. Following Jomini, while the political elites may set the overall purpose for a war,                                                310 Gourevitch, p. 96.  311 http://www.law.utoronto.ca/documents/Mackin/mugesera.pdf 17 (Need to figure out how to cite this properly)  312 http://www.law.utoronto.ca/documents/Mackin/mugesera.pdf 18 313 http://www.law.utoronto.ca/documents/Mackin/mugesera.pdf 23 314 http://www.law.utoronto.ca/documents/Mackin/mugesera.pdf 20 315 http://www.law.utoronto.ca/documents/Mackin/mugesera.pdf 24  88  they should refrain from interfering once war has commenced. In contrast, Clausewitz assumes that war is a political instrument, and that policy should not be detached from the war itself. Comparing these two perspectives is important, because genocide is an intrinsically political activity. Genocide is a kind of war, and like war, it is the product of elite policy. Unlike conventional war, however, policy comes to be distorted by a genocidal ideology. Instead of an exclusive focus on a legitimate threat, policy directs strategy to concentrate on non-adversarial categories of foe.    In the years prior to the 1994 genocide, a group of Hutu elites drawn from northwestern Rwanda sought to cling to power. Called the ‘akazu’, this coterie of influential individuals dominated politics, the military and the economy. However, beginning in the late 1980s, a series of interconnected ‘shocks’ threatened the basis of their power. This shock came from three sources: a dramatic economic downturn, armed invasion by the RPF, and pressure for political liberalization.  What set the akazu apart from other imperiled elites was its members’ adherence to a genocidal ideology. This ideology encouraged chauvinistic dominance by the Hutu over the Tutsi minority, as well as disdain for the nascent political opposition. It became genocidal once it began identifying Tutsi and opposition Hutu as belonging to the same category of adversaries as the RPF. As the three were conflated, unarmed social groups became identified as credible threats.  Charting the enemy creation process is made possible by examining the ideological discourse propagated by the extremists tied to the akazu. While the Kangura articles, RTLM broadcasts, the ENI memorandum, and the Mugesera speech did not cause the transition from non-adversarial civilian group to an adversarial one, they created the milieu in which it become hazardous to be a Tutsi or an opposition figure. All these examples of scurrilous invective reflected the positions held by Rwandan Hutu elites, both political and military. The identification  89  of the Tutsi as a military threat was a necessary step toward imposing a military solution. What united all of these diatribes and hatemongering was their common ideological underpinning.      90  Chapter 6: Grand Strategy and the Preparations for Genocide   Figure 0-1 The Strategy Hierarchy (Grand Strategy)  Mass killing is inseparable from genocide. But to claim that violence is the only instrument used by perpetrator states is to overlook a host of other destructive mechanisms. This chapter aims to link genocide to the notion of ‘grand strategy’ by presenting wholesale group extermination as a multifaceted process implicating the domains of politics, diplomacy, culture, religion, morality, and economics. The most violent stage of genocide occurs when military strategy is given priority; however, prior to this point, the state will undertake various strategies to weaken the ‘enemy’ and make preparations in anticipation of their destruction.  This chapter first aims to define ‘grand strategy’ in terms of the strategy-hierarchy heuristic developed in chapter 3. ‘Grand strategy’ is understood as a policy-driven coordination of all means available to a state, including (but not solely) the military apparatus. The next section attempts to trace the conceptual parallels between ‘grand strategy’ and Raphael Lemkin’s original understanding of ‘genocide’.  The second half of the chapter applies this framework to the diverse actions undertaken by the Hutu elite in Rwanda prior to April 6, 1994. While some of these preparations may not  91  initially have evinced a genocidal intent, they were ultimately manipulated to realize the goal of group destruction.     6.1 Defining Grand Strategy  Grand strategy represents the highest level of planning. It precedes and accompanies both war and genocide, encompassing all defensive and offensive arrangements made before, during, and after the outbreak of instrumental violence. While there are numerous definitions of ‘grand strategy’, they share four recurring qualities. First, grand strategy is hierarchically subordinate to the policy it serves. Second, military strategy is a distinct component of grand strategy. Third, grand strategy has shifting priorities that can change suddenly or gradually over time. Finally, the scope of grand strategy is all-encompassing in its reach. First, grand strategy is subordinate to the policy which gives it direction. This theme has already been explored in the previous chapter.  According to Hart, the role of grand strategy is to ‘coordinate and direct all of the resources of a nation, or a band of nations, towards the attainment of the political object of the war - the goal defined by fundamental policy.’316 Grand strategy can be understood as ‘policy in execution’: it subsumes a diversity of interrelated actions, in both peace and war that comprehensively aim towards the ends set by policy.317 Along these lines, Kennedy argues that ‘[t]he crux of grand strategy lies therefore in policy, that is, in the capacity of the nation’s leaders to bring together all of the elements of the nation’s long-term (that is, in wartime and peacetime) best interests.’318 Second, military strategy is viewed as a distinct and subordinate component of grand strategy. Military strategy must share the stage with implicitly non-martial strategies. It is the role                                                316 B.H. Liddell Hart, Strategy (New York: Meridian, 1991), p. 322.   317 Hart, p. 322.  318 Paul  Kennedy, “Grand Strategy in War and Peace:Toward a Broader Definition” in Grand Strategies in War and Peace ed. Paul Kennedy (New Haven: Yale University Press, 1991): p. 5.   92  of grand strategy, writes Platias and Koliopoulos, to ‘align the military strategy of the war with the political, diplomatic and economic strategies that form part of the war effort, making sure that they interact harmoniously and that one of  these strategies does not have a detrimental impact on another.’319 While it is often held to be the strategy with the most direct impact (because it controls the instrumental use of violence), military strategy is nonetheless a constituent part of overarching grand strategy.  Brodie’s alternative (but complementary) definition assumes that grand strategy is ‘the basic but all-embracing features of a plan of war, as distinct from either the details of a war plan or the strategy of a particular campaign.’320 The difference between grand and military strategies is further emphasized by Strachan. He argues that ‘operational plans of military strategy look to the near term, and work with specific situations. Grand strategy, on the other hand, can entertain ambitions and goals which are more visionary and aspirational than pragmatic and immediate.’321  Third, grand strategy is also shaped by changing priorities. As a consequence of its subordination to the policy goals set by an elite authority, grand strategy changes priority according to the variable needs of policy. During periods of peace, non-martial strategies prevail; when a state is engaged in armed conflict, the organized use of force is dominant, even while non-combat oriented strategies continue in the background.  ‘Military strategy,’ writes Gray, ‘can dominate grand strategy, while alternatively it may be assigned relatively too limited a role.’322  Additionally, by operating in an environment ‘full of imponderables and “unforeseen frictions,”’ grand strategies may change relative priorities very suddenly.323 Frequently, a                                                319  Athanassios G. Platias & Constantinos Koliopoulos, Thucydides on Strategy (London: Hurst, 2010), p. 5.  320 Bernard Brodie, "Strategy as a Science" in World Politics 1:4 (July 1949): p. 477 321 Hew Strachan, 'Strategy and Contingency' in International Affairs 87:6, p. 1281.  322 Colin S. Gray, The Strategy Bridge: Theory for Practice (Oxford: Oxford University Press, 2010), p. 29.  323 Kennedy, p. 5.   93  dramatic stimulus, such as revolution or a foreign invasion, needs to transform elites’ outlooks before a change in grand strategy’s relative priorities can be effected. In order words, the emphasis placed on martial methods over less violent ones may change dramatically in the context of strategic shocks (see 5.3).  Finally, grand strategy encompasses a wide array of means. The diverse actions of grand strategy, according to Paul Kennedy, consist of ‘much more than the supervision of battles.’324 Gray defines grand strategy as ‘the direction and use made of any or all the assets of a security community, including its military instrument, for the purposes of policy as decided by politics.’325 Gray’s emphasis upon the utilization of ‘all the assets of a security’ stresses the numerous and varied components of grand strategy. Such strategy is said to embrace a plethora of interrelated means across multiple, overlapping subfields of state influence. Under the rubric of ‘grand strategy’, it is possible to discuss specific political, economic, diplomatic, cultural and military strategies. It is, writes Freedman, ‘where all instruments of policy were weighed, one against each other.’326  What does this mean in practice? Before resorting to the use of arms, governments will employ diplomacy to isolate the enemy, form alliances, and keep neutral actors noncommittal; they will exert financial pressures on the adversary, rally economic resources, and develop logistical capacity; and they will engage in moral and religious struggle, glorifying their particular cause, while demonizing that of the foe. For Earle, the sum of these practices of grand strategy reflects a state’s theory of how it can ‘cause’ security for itself.327                                                  324 Kennedy, p. 2.  325 Gray, The Strategy Bridge, p. 28.  This is Dictum #1  326 Lawrence Freedman, Strategy: A History (Oxford: Oxford University press, 2013), p. 136.  327 Edward Mead Earle, 'Introduction' in The Makers of Modern Strategy, ed. Edward Mead Earle (Princeton: Princeton University Press, 1943): p.3-23.   94  6.2 Comparing ‘Grand Strategy’ and Lemkin’s ‘Genocide’    This section aims to draw out similarities between ‘grand strategy’ and ‘genocide’. First, both can be understood hierarchically, as they involve a chain of ends and means. Second, both attempt to shape the various domains of politics, diplomacy, economics, law, morality, and culture. Finally, while neither immediately prioritizes the use of violence, both have a special regard for military actions. First, for Lemkin, genocide was essentially a strategic endeavour: he defined it as a ‘coordinated plan of different actions aiming at the destruction of essential foundations of the life of national groups, with the aim of annihilating the groups themselves.’328 This gradual process encompasses a diverse array of intentional acts – techniques, to use Lemkin’s term - within the domains of politics, society, culture, religion, morality and the economy. ‘The objectives of such a plan,’ writes Lemkin, ‘would be [the] disintegration of the political and social institutions of culture, language, national feelings, religion, and the economic existence of national groups, and the destruction of the personal security, liberty, health, dignity, and even the lives of the individuals belonging to such groups.’329  Lemkin’s conception of genocide is readily converted to the strategy hierarchy framework.  The ‘objective’ (policy) gives guidance to the ‘coordinated plan’ (strategy), which in turns directs the use of ‘techniques’ (tactics or means) to realize this overriding purpose. When analyzed from the perspective of strategic theory, we may interpret Lemkin’s definition of genocide as a theory of grand strategy that assumes the targeting of non-adversarial groups. In other words, following Lemkin’s elaboration, genocide is a type of grand strategy that channels an ideological policy.                                                328 Raphael Lemkin, Axis Rule in Occupied Europe: Laws of Occupation - Analysis of Government - Proposals for Redress (Washington, D.C.:  Carnegie Endowment for International Peace, 1944), 79 - 95. 329 Lemkin, Axis Rule.   95  Second, both grand strategy and genocide attempt to realize policy goals by coordinating actions in the domains of politics, diplomacy, economics, law, morality, culture. The diverse techniques highlighted by Lemkin parallel the domains of grand strategy at a state’s disposal. He writes that genocide may ‘begin with political disenfranchisement, economic displacement, cultural undermining and control, the destruction of leadership, the breakup of families, and the prevention of propagation.’330   The ‘techniques’ of genocide and the various actions of grand strategy are both intended to serve policy. Like grand strategy, genocide encompasses myriad actions to confront perceived security threats. Each strategic sphere embraces what Roth calls a ‘multitude of sins.’331 These genocidal techniques complement each other and represent, in Lemkin’s words, ‘a synchronized attack on different aspects of life.’332 Daniel Feierstein has defined these techniques as ‘policies of systematic weakening.’ He describes them as   strategies of both physical destruction through overcrowding, malnutrition, epidemics, lack of healthcare, torture, and sporadic killings; and psychological destruction, manifested in humiliation, abuse, the harassment or killing of family members, attempts to undermine solidarity through collective punishment, encouraging collaboration in the categorization and classification of prisoners, and peer denunciation and peer abuse.333  Finally, paralleling these definitions of grand strategy, the use of force is not necessarily primary at the outset. ‘Cumulative discrimination,’ writes Shaw, precedes ‘indiscriminate                                                330 Raphael Lemkin, Lemkin on Genocide (Lanham: Lexington Books, 2012): p.35.  331 John K. Roth, “The Politics of Definition” in Will Genocide Ever End? eds. Carol Rittner, John K. Roth and James M. Smith (St. Paul: Paragon House, 2002): p. 25. 332 Lemkin, Axis Rule, p. xi. 333Daniel Feierstein, 'The Concept of 'Genocidal Social Practices' in New Directions in Genocide Research ed. Adam Jones (Routledge, 2011): p. 26.   96  slaughter.’334 Before an event is classified as genocidal (i.e. prior to widespread and systematic massacres), a state will implement numerous discriminatory measures. Alex Alvarez writes that ‘[i]n order to fully understand genocide... it is necessary to appreciate that the killing sites where the murders are carried out only represent the endpoint of a long sequence of steps involving many institutions, agencies and actors.’335 Alvarez’s notion of an ‘endpoint’ concluding a long process resonates with comments made by Arendt. She writes that ‘[t]he insane mass manufacture of corpses is preceded by the historically and politically intelligible preparation of living corpses.’336 Understanding the role of various techniques of oppression prior to systematic genocidal violence is crucial. According to Levene, this perspective demonstrates Lemkin’s view that ‘annihilation as such can never be viewed in isolation but has a developmental, cumulative trajectory in which a variety of coercive strategies might be applied to the targeted group by way of ‘preparatory’ measures.337 Moreover, the benefits of this ‘contextual approach,’ according to Verdeja, is that it ‘draw[s] attention to the broader host of perpetrator repressive and destructive policies (beyond those leading to genocide), show the interactive effects and general processes of perpetrator policy radicalization, and shed light on the different steps in the continuum of repression.’ 338  Both ‘genocide’ and ‘grand strategy’ have a special regard for military strategies, because they direct the instrumental use of violence. Lemkin understood the ‘physical’ aspect of genocide as the techniques that bring about group destruction.339 This emphasis may be                                                334 Martin Shaw, War and Genocide: Organised Killing in Modern Society (Wiley, 2003), 40.  335 Alex Alvarez, Governments, Citizens, and Genocide: A Comparative and Interdisciplinary Approach (Bloomington: Indiana University Press, 2001): p. 88. Italics added.  336 Hannah Arendt, The Origins of Totalitarianism (Orlando: Harcourt Brace & Company, 1979): p. 447.  337 Mark Levene, Genocide in the Age of the Nation State Vol. 1: The Meaning of Genocide (London: I.B. Tauris, 2005): p. 46.  338Ernesto Verdeja, "On Situating the Study of Genocide within Political Violence" in Genocide Studies and Prevention, 7:1 (Spring 2012): p. 83.  339 Lemkin, Axis Rule.   97  considered analogous to the distinction between the varieties of grand strategy and the special place of military strategy. While Lemkin emphasized the role of other techniques of genocide, he acknowledged that ‘[a]ctual physical destruction is the last and most effective phase of genocide.’340 As discriminatory regimes gain intensity and evolve to incorporate a more radical eliminationist341 agenda, the organization and implementation of the physical means of violence takes precedence. The shift from largely structural oppression to naked violence may be gradual, or, if motivated by an abrupt or unforeseen exigency, tragically sudden. The most violent stage of genocide does not necessarily commence because of an identifiable strategic shock; nonetheless, many instances of mass killing originate in some dramatic change. Some genocides, such as the Holocaust, begin gradually and build up a lethal momentum; others, such as in Rwanda, begin (and end) with disturbing abruptness.    6.3 Grand Strategy and Rwanda (October 1st 1990 to April 6 1994)  As discussed in the previous chapter, the policy of the Rwandan elites was to defend their privileges against a multifaceted assault. Hintjens notes that ‘[b]y the early 1990s, akazu members had come to dominate the most strategic positions both in central ministries and in regional government.’342 They used their dominant position to deploy every imaginable means to ensure their survival. The coordination of these ‘means’ or ‘techniques’ may be viewed under the rubric of grand strategy.                                                   340 Lemkin, Lemkin on Genocide, p. 35.  341 Daniel Jonah Goldhagen, Worse Than War: Genocide, Eliminationism and the Ongoing Assault on Humanity (New York: PublicAffairs, 2009) 342 Helen M. Hintjens, "Explaining the 1994 Genocide in Rwanda" in The Journal of Modern African Studies 37:2, p. 259.  98   Following the overlapping grand-strategic spheres described above, we may discern the implementation of numerous grand strategic behaviors and genocidal techniques across the various domains of military, politics, diplomacy, economics, law, morality, culture. (1) Military strategy was focused upon the preparation for war, and included the expansion of the army, and the creation of the Presidential Guard, civil defence programs, and party militias. (2) Diplomatic strategy attempted to isolate the RPF and undermine the Arusha Accords and UNAMIR, while maintaining solid relations with foreign allies such as France. (3) Political strategy aimed to sabotage the nascent multiparty system by creating openly sectarian pro-Hutu Power parties, while also dividing the new opposition parties into Hutu Power factions. (4) Legal strategy aimed to create a climate of impunity in which Tutsi and opposition Hutu were denied justice from an impartial judiciary. (5) Moral strategy was conducted via the churches, who were historically close to the government, and lent their moral authority to its actions. (6) Economic strategy attempted to manage the declining sources of elite income while marginalizing Tutsi and southern Hutu. (7) Finally, as highlighted in the previous chapter, Cultural strategy aimed at spreading a virulently anti-Tutsi, anti-opposition ideology.  The proceeding analysis incorporates the period between the RPF invasion in October 1990 and the assassination of President Habyarimana in April 1994. It must be stressed that many of these behaviors were deliberately motivated by particular exterminatory ends, while others emerged in an ad hoc manner that was ultimately co-opted by the genocidaires. The difference between the two is not always clear. Caplan writes that ‘from October 1, 1990, Rwanda endured three and a half years of violent anti-Tutsi incidents, each of which in retrospect can easily be interpreted as a deliberate step in a vast conspiracy… But all such interpretations remain speculative… nor can it be demonstrated that the countless manifestations of anti-Tutsi sentiment in these years were part of a diabolical master plan.’343                                                343 International Panel of Eminent Personalities, Rwanda: The Preventable Genocide (African Union, 2000): sec. 7.4.   99  Despite this need for caution, it is still possible to detail the diverse actions and techniques that comprise the grand strategy of the elite clique centred on the akazu, the MRND, and exponents of Hutu Power ideology.  6.3.1 Military Strategies  Military strategy in this lead-up period to the genocide concentrated upon preparing the instruments of coercion and maintaining low levels of violence. This task involved the expansion of the army, and the creation of the Presidential Guard, civil defence programs, and party militias.  Although the initial October 1990 RPF invasion was repulsed, it was a stark reminder of the inherent weakness of the Forces armées rwandaises (FAR). Within six months of the assault, the FAR grew from five to six thousand inexperienced soldiers to between eighteen and thirty-five thousand new recruits.344 Reyntjens estimates that the army grew up to 40,000 recruits.345 The rapid expansion of the army was accomplished by reducing its recruiting and training regimen.346 Additionally, new specialized armed groups were developed, including the elite Presidential Guard (GP). The GP was tasked with protecting President Habyarimana, and according the ICTR, ‘had a separate chain of command linked directly to him.’347 The GP expanded to over 1,500 men drawn largely from regional base of the akazu in the north and east.348                                                 344 Stephen Kinzer, A Thousand Hills : Rwanda's Rebirth and the Man who Dreamed it (Hoboken, N.J.: John Wiley & Sons, 2008): p. 94.  See also: African Rights, Rwanda: Death, Despair and Defiance (London: African Rights, 1995).  Two sources diverge on numbers. 345 Judgement, Theoneste Bagosora, Gratien Kabiligi, Aloys Ntabakuze, Anatole Nsengiyumva (ICTR-98-41-T 18), Trial Chamber 1, 18 December 2008 § 150.  346 Judgement, Theoneste Bagosora, Gratien Kabiligi, Aloys Ntabakuze, Anatole Nsengiyumva § 150. 347 Judgement, Theoneste Bagosora, Gratien Kabiligi, Aloys Ntabakuze, Anatole Nsengiyumva § 162. 348 African Rights, p. 49.   100  Paired with the expansion of the FAR and the creation of the GP was the establishment of nationwide civil defence programs. According to Straus, this program had two main components.349 First, it introduced neighborhood patrols of about ten men each who, carrying traditional weapons, would make rounds within the commune and man roadblocks. This was done ostensibly in order to intercept infiltrators. Second, arms were distributed to civilian administrators, communal police and army reservists. Taken together, Straus contends that ‘[t]he central idea behind the civilian defense program was that to combat the rebel threat, which purportedly had clandestine internal support, the government would arm its civilian administration.’350 While organized locally in every prefecture, the organization of the civil defence would take orders directly from President Habyarimana and his National Security Council.351 Then there was the creation of youth militias, such as the MRND’s interahamwe and the CDR’s impuzamugambi. The economic downturn of the late 1980’s had resulted in a substantial pool of unemployed and aggrieved youths from which to draw upon.352 The rationale behind the creation of the militias was the need for a force at the rural communal level exclusively loyal to hardline politicians.353 While every party developed its own youth wings, the MRND and CDR militias received arms, uniforms, and military training.354  In January 1994, Faustin Twagiramungu, first Prime Minister after the genocide, accused the Defense Ministry of providing military training to over 1,000 interahamwe.355 Arms distribution increased after                                                349 Scott Straus, The Order of Genocide : Race, Power, and War in Rwanda (Ithaca: Cornell University Press, 2006): p. 26. 350 Straus, Order of Genocide, p. 26.  351 Linda Melvern, Conspiracy to Murder: The Rwandan Genocide (Verso, 2006): p. 21. 352 Philip Gourevitch, We Wish to Inform You That Tomorrow We Will Be Killed With Our Families (New York: Farrar, Straus and Giroux, 1998): p. 93.  353 African Rights, p. 54.  354 African Rights, p. 55.  355 Orth, p. 221.   101  October 1993 as the militias received firearms, grenades and machetes.356 When they weren’t training, the militias disrupted rallies by opposition parties, blocked traffic, picked fights, engaged in rape and robberies, and were implicated in a series of bomb attacks.357 By the time of the genocide, membership in these militias ranged from 10,000 to 30,000.358   6.3.2 Diplomatic Strategies  Diplomatic strategy aimed at maintaining strong relations with foreign allies, especially the French, while isolating the RPF and undermining both the Arusha Accords and the deployment of UNAMIR. The Habyarimana regime sought to maintain positive ties within the Francophone community and with other regional allies. A week after the October 1990 RPF invasion, a fake attack on Kigali was staged and was widely viewed as credible by foreign observers.359 This phony attack on the capital convinced Rwanda’s allies there was an imminent security threat. As a result, military support from France, Belgium and Zaire helped to repulse the initial RPF assault.  Habyarimana sought support from regional allies, such as Kenya’s Daniel arap Moi and Zaire’s Mobutu Sese Seko.360 However, it was France who remained Rwanda’s most important backer. After the RPF invasion, France remained in Rwanda and maintained a military deployment there until being replaced by UNAMIR at the end of 1993.361 Kinzer claims that during the early 1990s, ‘France sold the Rwandan regime more than $20 million worth of                                                356Judgement, Ferdinand Nahimana, Jean-Bosco Barayagwiza and Hassan Ngeze (ICTR-99-52-T), Trial Chamber 1, 3 December 2003 § 113. 357 International Panel of Eminent Personalities, sec. 7.29.  358  Orth, p. 220.  359 International Panel of Eminent Personalities, sec. 7.9  360Dixon Kamukama, Rwanda Conflict: Its Roots and Regional Implications (Kampala: Fountain Publishers Ltd., 1997), p. 54. 361 Alan J. Kuperman, The Limits of Humanitarian Intervention: Genocide in Rwanda (Washington D.C.: Brookings Institution Press, 2001), p. 9.  See also: African Rights, p. 67.   102  weaponry and help it buy five times that amount from arms dealers in Egypt and South Africa.’362 The French provided mortars, light artillery, armoured cars and helicopters.363 France was also key in training the new GP from 1990 to late 1993.364  Additionally, efforts were made to derail the Arusha peace talks. The agenda set during Arusha talks challenged elite control; it focused on ceasefires, legal principles, power sharing, the integration of the army and the repatriation of refugees.365 According to Orth, the government used the ‘tactic of political squabbling’ in order to prolong negotiations and prevent the implementation of the transitional government.366  Ultimately, the MRND’s negotiators could prevaricate no longer, and conceded to the agreement. One of the stipulations was the formation of the United Nations Assistance Mission for Rwanda (UNAMIR). Established by the UN Security Council in October 1993, it was given the task of establishing a safe environment for the establishment of the Broad Based Transitional Government (BBTG).367 True to form, the efforts of UNAMIR were sabotaged from the start by Rwanda’s elite vested interests. Propaganda, especially through RTLM, criticised UNAMIR as a pro-RPF colonial oppressor.368 The government for its part did its best to feign cooperation while doing - sometimes quite openly - the very opposite.                                                    362 Kinzer, p. 94.  363 African Rights, p. 67.  364 African Rights, p. 49.  365 Bruce D. Jones, ‘The Arusha Peace Process’, in The Path of a Genocide: The Rwanda Crisis from Uganda to Zaire, eds. Howard Adelman and Astri Suhrke (New Brunswick, New Jersey: Transaction Publishers, 2000),  p.137.  366 Richard Orth, ‘Rwanda’s Hutu Extremist Insurgency: An Eyewitness Perspective' in Genocide in Cambodia and Rwanda: New Perspectives ed. Susan E. Cook (New Brunswick: Transaction Publishers, 2009), p. 221.  367 Judgement, Theoneste Bagosora, Gratien Kabiligi, Aloys Ntabakuze, Anatole Nsengiyumva § 177. 368 Judgement, Theoneste Bagosora, Gratien Kabiligi, Aloys Ntabakuze, Anatole Nsengiyumva § 192.  103  6.3.3 Political Strategies  In 1991, the government gave in to pressure and begrudgingly drafted a new constitution. It allowed for a number of concessions, including the creation of independent political parties, the new office of prime minister, and freedom of the press.369 The Rwandan elites had other ideas. ‘From the very beginning,’ notes African Rights, ‘the government described opposition politicians as “accomplices” of the RPF. Opposition activity was seen as war by other means.’370 Political strategy sought to undermine multipartyism by establishing openly sectarian pro-Hutu parties while also infiltrating the new opposition parties. New parties were factionalized and split in half into camps of moderates and those devoted to the Hutu Power ideology. In order to sabotage multipartyism, President Habyarimana and the akazu established a number of political parties aimed, according to Gourevitch, at ‘sow[ing] confusion and mak[ing] a mockery of the pluralist enterprise.’371 The worst of these puppets was the Coalition for the Defense of the Republic (CDR). The CDR was to the far right of the MRND, and was openly sectarian. African Rights describes the formation of the CDR as part of ‘a strategy for obstructing the peace and democracy process.’372 Additionally, Longman notes that Hutu Power activists used ethnicity as a wedge to ‘divide... the opposition parties, co-opting large factions of each party into a broader reconfigured coalition united in opposition to Tutsi.’373 Parties were infiltrated by MRND loyalists who acted to divide Hutu Power and non-Hutu Power factions.374 By means of this divide-and-                                               369 Melvern, Conspiracy to Murder, p. 33-34.  370 African Rights, p. 31.  371 Gourevitch, p. 92.  372 African Rights, p. 37.  373 Longman, p. 162.  374 Joan Kakwenzire and Dixon Kamukama, ‘The Development and Consolidation of Extremist Forces in Rwanda 1990 - 1994’, in The Path of a Genocide: The Rwanda Crisis from Uganda to Zaire, eds. Howard Adelman and Astri Suhrke (New Brunswick, New Jersey: Transaction Publishers, 2000), p.69.  104  conquer method, Habyarimana caused the MDR party to implode.375 Factions beholden to ‘Hutu Power’ ideology were encouraged, while those taking a more moderate position were sidelined, intimidated into compliance, or assassinated.376  6.3.4 Legal and Moral Strategies   Legal strategies created a climate of impunity in which Tutsi and opposition Hutu were denied justice through an impartial judiciary. During the four years prior to the genocide, African Rights contends that impunity became institutionalized.377 Immediately after the October 1990 RPF invasion and the faked ‘attack’ on Kigali, between 8,000 and 13,000 Tutsi and Hutus critical of the regime were imprisoned, held in deplorable conditions, tortured, and held for months without charges.378 Killings of Tutsi and opposition figures were seldom investigated, and those responsible for the early massacres (before 1994) were expeditiously released from jail.379 The gendarmerie (communal police) did little to prevent the violence.380 Impunity created a climate of fear in which, Caplan writes, ‘[m]urder, rape, harassment or imprisonment could befall any Tutsi at any time.’381 Opposition figures launched a National Commission of Inquiry into civil servants tasked with identifying public officials culpable for these killings.382 The commission was, regrettably, stillborn from the outset as MRND members blocked their recommendations. By 1993, African Rights claims, ‘the judiciary was effectively paralyzed by continuing harassment and interference from the government.383 It was clear that this system                                                375 African Rights, p. 86.  376 African Rights, p. 86.  377 African Rights, p. 50.  378 International Panel of Eminent Personalities, sec.7.8. See also: African Rights, p. 29.  379 African Rights, p. 50.  380 African Rights, p. 52.  381 International Panel of Eminent Personalities, sec. 7.17.  382 African Rights, p. 50-51.  383 African Rights, 51  105  protected the extremists. The Minister of Justice, unable to bring to justice Leon Mugesera (see 5.6.4), resigned.384 In addition to this legal climate, the church lent its moral authority to whitewash the actions of the regime. The church, according to Longman, ‘played an essential role in facilitate [the] descent into violence.’ They ‘not only failed to speak out forcefully against the increasing exclusion of Tutsi and the growing violence against Hutu activists and Tutsi in the years leading up to the genocide but also lent strong support to the regime that was encouraging the exclusion and violence.’385 This silence in the face of government tyranny would prove fateful.  Longman contends that  the failure of the Catholic bishops to address attacks on church personnel and church buildings proved particularly important in setting a precedent…. In Kibilira and Bugesera, Tutsi sought refuge in church buildings, but the bishops never openly supported the principle of sanctuary. When a church school at Muramba in Gisenyi was attacked in January 1993, the bishops’ failure to defend the sanctity of the church sent a strong message to many people.386   6.3.5 Economic Strategies    Economic strategies attempted to manage the dwindling sources of elite income while benefiting from the marginalization of the Tutsi and southern Hutu. By the time of the October 1990 RPF invasion, the northwestern Hutu elites were faced with a generalised economic downturn, decreased demand for exports, currency devaluation, and international pressures to                                                384 Melvern, Conspiracy to Murder, p. 39.  385 Timothy Longman, Christianity and Genocide in Rwanda (Cambridge: Cambridge University Press, 2010): p.162.  386 Longman, p. 174.   106  reform a corrupt system. Despite these obstacles, the elites found alternative means of filling their coffers, such as illicit currency transactions, generous commissions on government contracts, drug deals, and money laundering.387 They perpetuated their dominance through the senior positions they held in parastatal companies.388 The elite also continued the practice of skimming off foreign aid. Uvin contends that rural development projects funded by foreign aid ‘act[ed] as mechanisms for exclusion and for reproduction of the privileges of a small elite.’389 Money was habitually redirected to serve the interests of the elites. In 1992, the government was allotted emergency financial assistance in order to pay for imports of food and medicine, but these monies were allegedly redirected to finance arms purchases.390 Additionally, the Habyarimana regime perpetuated a system of ‘ethnic equilibrium’ whereby quotas limited the number of Tutsi admitted to education and particular occupations. Only 9% of appointments to secondary school were reserved for Tutsi.391 Very few Tutsi, Longman notes, were able to serve in the military or in the civil service.392  Of the 143 bourgmestres across Rwanda, there was not a single Tutsi, and of the 11 prefects, only one was Tutsi.393 Unofficially, this system of exclusion extended to southern Hutu as well.394   6.3.6 Cultural Strategies   The cultural component of the Rwandan elite’s grand strategy was explored in the previous chapter (5.6). By insinuating their message through the media the, extremists linked to                                                387 International Panel of Eminent Personalities, sec. 5.18. See also: Hintjens, p. 257. 388 African Rights, p. 23.  389 Peter Uvin, Aiding Violence: The Development Enterprise in Rwanda (West Hartford, Connecticut: Kumarian Press, Inc., 1998): p. 118. 390 Hintjens, p.257.  391 African Rights, p. 22.  392 Longman, p. 122.  393 African Rights, p. 22. This single Tutsi prefect would later play a significant role in delaying, albeit temporarily, the subsequent massacres.  394  Hintjens, p. 259.   107  the akazu were able to spread their divisive ideology. The cultural aspect of the elite’s grand strategy was evident in their multifarious attempts at propaganda. Through both radio (Radio Rwanda and RTLM) and print media (Kangura), the strategy aimed to shape a discourse of hatred and division. This was clearly evident even at the level of institutional (Mugesera and MRND) and military culture (the ENI commission)   6.4 Summary   This chapter has attempted to conceptualize ‘grand strategy’ and to reconcile it with Lemkin’s original understanding of ‘genocide.’ In this manner, we may utilize grand strategy and the strategy hierarchy to understand the preparatory period beginning on 1 October 1990 with the RPF invasion and concluding with the 6 April 1994 assassination of President Habyarimana.  While policy provides the goals to be pursued, grand strategy encompasses the wide array of instruments at a state’s disposal. Grand strategy operates in the overlapping domains of the military, diplomacy, politics, law, morality, the economy and culture. This interpretation of grand strategy parallels Lemkin’s original understanding of genocide. This chapter has noted three similarities between ‘grand strategy’ and ‘genocide. First, both are directed by the ends set by policy. Second, they both assume a wide array of means across similar domains of action. Finally, the military component of grand strategy, while significant, is not always immediately primary.   The second part of this chapter utilizes the concept of grand strategy to frame the various actions undertaken by the dominant Rwandan elites in their pursuit of a policy of self-preservation. Militarily, the army was expanded, specialized armed groups were established, party militias were created, and civic defense units were instituted across the country. Diplomatically, grand strategy sought to keep allies (especially the French) on the side of the government, while undermining both the negotiations at Arusha and later the UNAMIR  108  deployment. Politically, the elites sought to disrupt multipartyism by dividing the opposition parties into factions loyal to Hutu Power, while also creating their own extremist party in the form of the CDR. Legally and morally, grand strategy used the justice system and the church to create an atmosphere of legal impunity backed by the moral authority of the church. Economically, the elites sought to perpetuate their unequal access to state resources while weakening that of their enemies. Finally, cultural grand strategy was manifested in extremist propaganda on the radio and in print, but also institutionally within the MRND and the armed forces. Taken individually, many these actions do not provide evidence of an exterminatory programme. Each of these ‘techniques’ can be justified in terms of elites taking measures to protect their interests and status. However, Butcher notes that ‘none of the individual techniques that Lemkin described are genocidal in and of themselves — even mass murder. Rather, they become genocidal when they are part of an integrated policy of genocide—one that also incorporates multiple other techniques (which themselves also become genocidal through participation in the policy).’395 Thus, measures that are not initially exterminatory in intent may become exterminatory in consequence as a genocidal ideology comes to dominate policy.  In Rwanda, the elites’ primary objective was to preserve their status and privileges in the face of economic troubles, external aggression and political liberalization. These interrelated threats were exemplified by three groups: the RPF, the political opposition, and the Tutsi. Of these groups, only the RPF represented a true armed and organized adversary. However, as the genocidal Hutu Power ideology came to penetrate the making of policy, these three groups became conflated. The innocent, poorly armed, and disorganized Tutsi and Hutu opposition                                                395Thomas M. Butcher, "A ‘Synchronized Attack’: On Raphael Lemkin's Holistic Conception of Genocide" in Journal of Genocide Research 15:3 (2013): p. 258.   109  figures became, in the minds of the ideologically indoctrinated policy makers, a menace equivalent to the RPF.   Following Butcher, the grand-strategic measures undertaken by elite policymakers in Rwanda do not obviously resemble the ‘techniques’ highlighted by Lemkin. For instance, the expansion of the army can be explained as a necessary response to a foreign aggressor, and the disruption of democratic politics may be justified in terms of self-interest. What unifies all of these actions is the preeminence of a genocidal ideology. In hindsight, the various actions of grand strategy highlighted above contributed to the ultimate campaign of genocide. They constitute ‘techniques’ of genocide because they were bound together by policy; even if their original intention was obscured, the various means of grand strategy were directed towards policy goals.     110  Chapter 7: Military Strategy and the Implementation of Genocide  Figure 0-1 The Strategy Hierarchy (Military Strategy)  The previous two chapters have applied the strategy-hierarchy heuristic to the period prior to April 1994. The focus thus far has been upon the policies and grand strategy of Rwanda’s entrenched elite. This chapter turns to the period after 6 April, and analyzes the two military strategies that animated the genocidal violence.  War and conventional military strategy emphasize the importance of the engagement. While genocide likewise consists of engagements, they are qualitatively different. Depending upon who is participating in the violence, it is preferable to describe them as either battles or massacres. The next section highlights some of the massacres that preceded the genocide, before turning to the crucial ‘strategic shock’ that triggered the beginning of the mass killing: the assassination of President Habyarimana. The Rwandan genocide is framed in terms of a series of massacres directed by two military strategies. The first and primary strategy focused on the selective elimination of Rwanda’s moderate opposition. The second military strategy began a week later. This emergent strategy targeted the Tutsi for wholescale annihilation; it relied upon and encouraged the mass concentration of Tutsi upon supposed sites of ‘refugee’. Once assembled, the perpetrators  111  sought to restrain and degrade the ability of the Tutsi to defend themselves, before ultimately eradicating them. Once the crime was complete, the killers moved on to besiege the next concentration until, ultimately, hundreds of thousands of Tutsi were murdered.     7.2 Engagements: Battles and Massacres  The aim of this section is to define the military ‘engagement’ and its two chief forms, battles and massacres. Military strategy is shown to be responsible for managing the conduct of armed groups during the engagement. Such strategy, it is argued, is concerned with both the implementation of instrumental violence and the creation of the optimal operational context.    A military ‘engagement’ represents the coming together of two (or more) groups whereby one (or more) party exercises force against the other. The engagement is understood here as the actualized moment of coercion when lethal violence is realized. Following Clausewitz, military strategy uses engagements in order to secure the object of the war. ‘Fighting,’ he writes, ‘is the central military act... Engagements mean fighting. The object of fighting is the destruction or defeat of the enemy.’396   Engagements respect boundaries of time and space. An engagement usually lasts hours, or at most days. While sieges stretch into months or years, the actual moments of direct violence are more limited. Additionally, engagements take place within certain spatial limits. While conflict may occur within an expansive theater of operations, fighting takes place in a specific – and often limited – physical context. Engagements do not last forever, or occur everywhere at once. It is the task of military strategy to ensure that policy objectives are realized within these constraints.                                                     396 Carl Von Clausewitz, On War. trans. Michael Howard and peter Paret (Oxford: Oxford University Press, 1976): p. 226-229.   112  The overall form that violence takes is a direct product of who is participating in an engagement. For this reason, the term ‘engagement’ is morally neutral; it merely connotes the violent collision of actors, without restricting the agents involved. This morally neutral form of ‘engagement’ is dominant in strategic studies.  Once the notion of the ‘engagement’ is reconciled with the practice of targeting non-adversarial groups, two forms of engagement may be distinguished. ‘Battles’ assume a contest between armed and organized adversaries, and consist of fighting between antagonistic forces. ‘Massacres’, by contrast, are largely one-sided episodes of violence against a nonthreatening and superficial threat.   Table 0-1 Types of War and Engagements   This distinction between types of engagements parallels the kinds of war highlighted in chapter 2 (see Table 7.1). Conventional war assumes battles between armed and organized combatants. Conversely, genocide implies one-sided massacres against an unwilling participant.  Both battles and massacres are variations of engagements, and share much in common. They consist of two (or more) essentially irreconcilable groups in which at least one party desires the destruction or submission of the other. Both involve lethal violence, whether small-War Engagements Adversary versus Adversary  Adversary versus Non-Adversary  Conventional War Battles Massacres Genocide  113  scale or extensive; and both are generally fought for instrumental purposes. Moreover, just as battles shape our historical understanding of particular wars, so massacres define certain genocides. While D-Day and Srebrenica were both engagements, and both involved mass killing, the latter is justifiably considered an atrocity, while the former is not. Likewise, there is a substantial moral and historiographical gap between the battles of Marathon and Gettysburg on one hand, and the massacres of Babi Yar and Nanking on the other.   Both battles and massacres reflect different orders of magnitude. Small engagements are significant in their own right. Clausewitz emphasized this point, and wrote that ‘our wars today consist of a large number of engagements, great and small, simultaneous or consecutive.’397 Even though Clausewitz’s ‘today’ referred to conflict in the early nineteenth century, his observation has transhistorical validity. From the perspective of the adversarial actor, military strategy determines how armed forces are utilized during engagements. Distinct from other types of strategy, ‘military strategy’ is concerned with the deliberate implementation of instrumental violence. It harnesses violence in order to achieve the ends of policy, as described in chapter 5. It will be recalled that ‘military strategy’ is part of an all-encompassing ‘grand strategy.’  As highlighted in chapter 6, the military aspect must share the stage with other strategies from the domains of politics, economics, culture, and society. However, during times of war – and genocide – military strategy becomes dominant, the primary strategy for realizing the aspirations of policy.  Gray’s second dictum defines ‘military strategy’ as the ‘direction and use made of force and the threat of force for the purposes of policy as decided by politics.’398 However, military strategy is not focused exclusively on fighting. Rather, according to Hart, ‘the aim of [military] strategy must be to bring about this battle under the most advantageous circumstances.’399                                                397 Clausewitz, p. 227.  398 Colin S. Gray, The Strategy Bridge: Theory for Practice (Oxford: Oxford University Press, 2010).  399 B.H. Liddell Hart, Strategy (New York: Meridian, 1991), p. 324.   114  Thus, before fighting begins, this aim is achieved by ‘diminish[ing] the possibility of resistance.’400 Prior to the engagement, it is the role of military strategy to ensure that the actual moment of violence occurs in a favourable context. This favorable context helps to ensure both the destruction of the enemy and the safety of the attacking force. Resistance precedes both battles and massacres, and if successful, it may threaten the designs of strategic planners. Ideally, Hart writes, ‘[t]he perfection of strategy would be, therefore, to produce a decision without any serious fighting.’401 In summary, ‘engagements’ take two primary forms. First, ‘battles’ refer to conventional violence between two (or more) adversarial groups. Second, ‘massacres’ are engagements that consist of an armed and organized force against a poorly armed and unorganized civilian group. ‘Military strategy’ is concerned with the utilization of armed forces aimed at realizing the goals of policy. One of the key means of realizing policy goals is winning engagements by either destroying or subduing the enemy. Military strategy focuses not only upon the fighting of engagements, but on the creation of an optimal context for this fighting.   7.3 Pre-Genocide Massacres and the Last Strategic Shock  Immediately following the October 1990 RPF invasion, the Habyarimana regime instigated a series of massacres. Describing these atrocities, a United Nations Commission on Human Rights claims that ‘[i]t has been shown time and time again that government officials were involved, either directly by encouraging, planning, directing or participating in the violence, or indirectly through incompetence, negligence or deliberate inaction.’402                                                 400 Hart, p. 323.  401 Hart, p. 324.  402 B.W. Ndiaye, "Question of the Violation of Human Rights and Fundamental Freedoms in Any Part of the World, With Particular Reference to Colonial and Other Dependent Countries and Territories: Extrajudicial, Summary or Arbitrary Executions” United Nations Economic and Social Council (April 1993).   115  The first massacre occurred forty-eight hours after the invasion, when hundreds of homes were burned and at least 348 Bagogwe Tutsi were killed in Kibilira.403 They were targeted again in early 1991 when the Bagogwe were blamed for assisting the RPF in their attack on the town of Ruhengeri.404 The Bugesera massacre in March 1992 was significant because of the variety of actors involved: militia, local authorities (including the bourgmestres), gendarmes, elements of the armed forces, and the Presidential Guard.405 It was preceded by the intentional spreading of rumors that the Tutsi planned to kill prominent Hutu, and that many Tutsi had crossed the border to Burundi in order to join the RPF.406 Hassan Ngeze, a chief propagandist of the Hutu Power ideology, was instrumental in spreading this divisive message. Des Forges writes that he was responsible for distributing a polemical tract that accused the Liberal Party leader of being a rebel and an assassin.407 Over the course of five days, over 300 people were killed. The Bugesera massacres marked an important change in the overall pattern of killing. Instead of focusing solely on Tutsi, oppositionist Hutu were also identified as accomplices and targeted for violence.408 These killings have been described as ‘practice massacres’ in anticipation of the genocide. For instance, Longman notes that ‘these periodic massacres provided an important precedent for the 1994 genocide, serving as a dress rehearsal to test out methods later duplicated throughout the country.’409  In contrast, the actual military strategy utilized during                                                403 Ndiaye, “Question”, p. 3; Timothy Longman, Christianity and Genocide in Rwanda (Cambridge: Cambridge University Press, 2010): p. 136.  404 Alison Des Forges, ‘Leave None to tell the Story’: Genocide in Rwanda (New York: Human Rights Watch, 1999), p. 71.  405 Linda Melvern, Conspiracy to Murder: The Rwandan Genocide (Verso, 2006): p. 26.  406 Melvern, “Conspiracy”, p.27; Des Forges, “Leave None”, p.  71.  407 Des Forges, “Leave None”, p.  71. 408 Longman, “Christianity and Genocide”, p. 170.  409 Longman, “Christianity and Genocide”, p. 169.   116  these pre-genocide massacres differs from that which marked the bloodshed in April and early May 1994. Des Forges notes that  In the northwest [against the Bagogwe] and in Bugesera, civilian and military authorities occasionally rounded up groups of several dozen people to be massacred all at one time at a site such as a communal office. But for the most part, they did not attack large groups who gathered spontaneously at such sites—particularly at churches. Instead they cut their access to food and water to force them to return home. They were not yet ready to launch the large-scale attacks that became usual during the 1994 genocide.410  In both Kibilira and Bugesera, the Tutsi sought refuge in the churches and public buildings.411 As Des Forges notes, these concentrations were not targeted for open violence. Prior to the genocide, military strategy eschewed these large gatherings, focusing instead upon easier targets.  Everything changed on the evening of 6 April 1994. President Habyarimana’s plane was destroyed by a surface-to-air missile as it approached Kigali airport. Despite numerous international inquiries, there is little agreement on who perpetrated the crime. For present purposes, however, the question of responsibility is moot. Since the extremists controlled the airwaves, the media, the army and militias, and much of the civil administration, they could communicate a singular message: the RPF, their Tutsi accomplices, and the backstabbing Hutu opposition had all conspired to kill the President.  The killing of the President was the fourth strategic shock. The previous three — economic decline, RPF invasion, and democratization — alerted the Rwandan elite to                                                410 Des Forges, “Leave none”, p. 72.  411 Longman, “Christianity and Genocide”, p.174.   117  challenges to their continued rule. As this group propagated the Hutu Power worldview, the real threat — the Rwandan Patriotic Front — was amalgamated with two distinctly non-adversarial groups: the nascent and largely Hutu political opposition, and the entire Tutsi ethnic group. According to the genocidal ideological discourse of Hutu Power, opposition politicians and Tutsi were nothing more than inkotanyi (accomplices) of the RPF. When President Habyarimana was assassinated, these accomplices were seen as a fifth column requiring immediate destruction. In short, after the assassination, these two non-adversarial groups were identified as an enemy in an essentially military sense, i.e. a group against which it was justified to employ systematic physical violence.412  7.5 Military Strategy and the Rwandan Genocide  Gérard Prunier contends that ‘the Rwandese [sic] genocide is of a mixed type - partly a classical genocide with the systematic massacre of an allegedly alien population, and partly political with the systematic killing of political opponents.’413 Utilizing Prunier’s distinction, the Rwandan genocide may be understood as conducted according to two different strategies. The original strategy selectively targeted the political opposition. Its members were singled out on the basis of their perceived threat to the regime and the continued dominance of the akazu. This initial strategy was largely implemented during the first week of the genocide. The second military strategy took shape by mid-April. It collectively targeted the entire Tutsi ethnicity, and took the form of a series of extraordinarily lethal massacres within and around public buildings. During times of political turmoil, Tutsi (and Hutu) would gather at these traditional sites of refuge. For the first few days after the assassination of Habyarimana, these locations were not explicitly targeted.                                                 412 Martin Shaw, War and Genocide: Organised Killing in Modern Society (Wiley, 2003).  413 Gerard Prunier, The Rwanda Crisis: A History of Genocide (Columbia University Press, 1995), p. 239.  118  Once the primary political targets were disposed of, however, the génocidaires redirected their attention to the next ‘enemy’ on their list. Instead of repulsing the RPF advance, the regime turned to the destruction of a much easier target: civilian Tutsi. From the perspective of the perpetrators, this task was made easier by the self-concentrating movements of the Tutsi in locations that could be transformed into elaborate traps.  These two military strategies by no means exhaust the diverse modalities of killing. Straus notes that thousands were killed at road-blocks, during house-to-house searches, and many more were hunted down in the swamps and forests.414 Nonetheless, these forms of killing were ancillary to the dominant modes of assassination and extermination.  In short, while both of these military strategies — the original against the political opposition and the emergent against the concentrated Tutsi — ultimately resulted in massacres, the methods animating these strategies were fundamentally different. The next two sections elaborate these differences.   7.5.1 Original Strategy: Targeting the Opposition  Habyarimana’s assassination triggered the initiation of the military strategy against Rwanda’s opposition moderates. Within an hour of the killing, the Presidential Guards (PG), the elite unit loyal to the akazu, set up roadblocks across Kigali.415 Leading opposition politicians were ordered to stay at home.416 By the next morning, the PG were going door-to-door hunting                                                414Scott Straus, 'How many perpetrators were there in the Rwandan genocide? An estimate' in Journal of Genocide Research 6:1, p. 89.  415 Fred Grunfeld and Anke Huijboom, The Failure to Prevent Genocide in Rwanda: The Role of Bystanders (Leiden: Transnational Publishers, 2007): p. 156.  416 International Panel of Eminent Personalities, Rwanda: The Preventable Genocide (African Union, 2000), sec. 14.9.    119  selected individuals according to pre-established lists.417 At 6:00 a.m., Radio Rwanda announced both the death of the president and the imposition of a curfew.418  Radio RLTM was used to broadcast the names of people to be killed, as well as indicate where they might be found.419 According to Des Forges, personal information was communicated, including ‘addresses, places of work, and where [targets] spent their leisure time.’420 Uncertain and afraid, many of those targeted remained in their homes in Kigali, awaiting the killers. According to René Lemarchand, these killings were focused upon two categories of individuals: (a)  moderate  (as distinct  from  "Hutu power") Hutu politicians from  the  south/central regions, most  of them  affiliated  to  the  Mouvement  Democratique  Republicain (MDR),  and  (b)  opposition leaders  (Hutu  and  Tutsi) identified  with  the  Parti  Liberal  (PL)  or  Parti  Social Democrate  (PSD).421 These individuals were not ‘selected  on  the  basis  of ethnic criteria’; rather, they were seen ‘as animated by a sense  of compromise and  conciliation  towards  the  RPF, in  short as potential traitors.’422 Adding to this list, African Rights includes ‘senior civil servants, wealthy businessmen with ties to the political opposition, critical journalists and human rights activists.’423 Most of these initial targets were Hutu, and much of the killing was performed by the PG.  Among the first opposition politicians targeted were those who potentially stood in way of the desired transition of power. Chief among them was Prime Minister Agathe Uwilingiyimana. Upon Habyarimana’s death, Uwilingiyimana became de jure the new head of the government. In the MDR party, her ascension to power was roundly rejected by Théoneste Bagosora and the                                                417 Grunfeld and Huijboom, p. 159.  418African Rights, Rwanda: Death, Despair and Defiance (London: African Rights, 1995), p. 177.  419 Des Forges, “Leave none”, p. 158. 420 Des Forges, “Leave none”, p. 158.  421 René Lemarchand, "Rwanda: The Rationality of Genocide" in Issue: A Journal of Opinion 23:2 (1995): p. 10-11.  422 Lemarchand, p. 10-11.  423 African Rights, p. 178.   120  ‘Crisis Committee.’424 In the morning of 7 April, ten Belgian paratroopers and five Ghanaian peacekeepers were sent to her home, tasked with escorting her to the radio station. When they arrived, she sensed the threat, and refused to leave without further reinforcements. By 7 a.m., her house was surrounded by twenty armed soldiers. The U.N. forces were subdued, and the Prime Minister and her husband fled. But they were set upon and killed.425  The ‘Crisis Committee’ murdered others who stood in their way, including Joseph Kavaruganda, president of the constitutional court and the individual responsible for swearing in new politicians.426 Additionally, Faustin Rucogoza, Minister of Information, was slain because of his control of Radio Rwanda and his threats to shut down RTLM.427 Much of the killing was along party lines. More than three quarters of the PSD’s ‘political bureau’ were killed over three days, including Frédéric Nzamurambaho, party president and the Minister of Agriculture and Livestock, and his first and second vice presidents.428 Both the PL and the MDR were divided into moderate and Hutu Power factions. Among those killed were the First Vice-President of the PL and Minister of Labour and Social Affairs, Landoald Ndasingwa, and Faustin Rucogoza, the MDR’s Minister of Information.429 According to the rules established by the Arusha Accords, Ndasingwa was to replace Habyarimana as president.430  The murder of 10 Belgian peacekeepers merits special mention. The peacekeepers were sent to protect Prime Minister Uwilingiyimana, but after coming under fire they were apprehended by FAR soldiers. They were taken to Camp Kigali; the five Ghanaian troopers were separated while the Belgians were murdered. This assassination was all pre-planned. As foreshadowed by General Roméo Dallaire’s infamous ‘genocide fax’ from three months prior,                                                424 Grunfeld and Huijboom, p.57.  425 Grunfeld and Huijboom, p.158-9 426 African Rights, p. 177.  427 African Rights, p. 177.  428 African Rights, p. 178.  429 African Rights, p. 183, 189.  430 Des Forges, “Leave none”, p. 150.   121  the targeting of Belgian peacekeepers was meant to encourage the withdrawal of the country’s soldiers.431 While not Rwandan themselves, the Belgian peacekeepers represented potential international opposition to the schemes of the akazu. The plan was largely successful: according to Caplan, after the killings ‘the world witnessed the unprecedented phenomenon of a UN peacekeeping mission actually sharply reducing its forces in the midst of a genocide.’432 In sum, the first few days after the assassination were not genocidal but eliticidal or, to use Barbara Harff’s term, politicidal.433 Caplan maintains that it is ‘arguable that a coup by the radicals against the coalition government, not genocide, was the original aim in the immediate wake of the crash [of Habyarimana’s plane].’434 However, after less than a week, the second military strategy clearly began to emerge.   7.5.2 Emergent Strategy: Targeting the Tutsi   As argued above, much of the initial violence during early April was marked by the targeted killings of opposition elites, especially in Kigali. The emphasis on targeting specific individuals, however, would soon change. ‘By the middle of the first week of the genocide,’ writes Alison Des Forges, ‘organizers began implementing a different strategy: driving Tutsi out of their homes to government offices, churches, schools or other public sites, where they would subsequently be massacred in large-scale operations.’435  According to this new military strategy, the organizers of the genocide exploited and encouraged the civilian tendency to gather at public sites during times of crisis. With large numbers of Tutsi concentrated in limited confines, the perpetrators expeditiously implemented a                                                431 For a full transcript of the fax, see: http://www.pbs.org/wgbh/pages/frontline/shows/evil/warning/cable.html 432 International Panel of Eminent Personalities, sec. 14.41. 433 Barbara Harff, 'No Lessons Learned from the Holocaust? Assessing Risks of Genocide and Political Mass Murder Since 1965" in The American Political Science Review 97:1.  434 International Panel of Eminent Personalities, sec.14.3.  435 Des Forges, “Leave none”, p.   122  four-phased military strategy that ended in the annihilation of the Tutsi. This emergent strategy resulted in numerous high-fatality, large-scale and short-duration massacres.  This particular strategy was most evident throughout April 1994. By the end of the month, Linda Melvern writes, ‘most of the large scale massacres were over.’436 Mamdani asks, ‘how could all those institutions that we associate with nurturing life -- not only churches, but schools and even hospitals -- be turned into places where life was taken with impunity and facility?’437 Following the conceptual understanding of the ‘engagement’ and military strategy developed above, the answer to Mamdani’s question is twofold. First, these sites were traditionally understood as sanctuaries where people could gather and find refuge from political turbulence. Second, once the génocidaires completed their original strategy against opposition targets, they came to perceive these concentrations of the hated Tutsi as an opportunity. They began encouraging Tutsi to gather at such locations in anticipation of their ultimate destruction. The overall rationale behind these efforts was, according to a dreadfully apt Rwandan expression, to ‘sweep... dry banana leaves into a pile to burn them more easily.’438 Unlike the military strategy against the opposition, which deliberately targeted specific individuals, the aim of the emergent military strategy was to facilitate the massacre of concentrations of non-adversarial Tutsi in specific public locations. Many of these large-scale massacres evince a pattern repeated throughout the country, with devastating consequences. This pattern is suggestive of a four-phased military strategy: (1) concentration, (2) degradation/containment, (3) eradication, and (4) annihilation.   7.5.2a Phase One: Concentration                                                   436Melvern, “Conspiracy”, p. 241. 437 Mahmood Mamdani, When Victims Become Killers: Colonialism, Nativism, and the Genocide in Rwanda (Princeton: Princeton University Press, 2001), p. 227.  438 Des Forges, “Leave none”, p. 301.   123  Immediately after the assassination of Juvénal Habyarimana on April 6, 1994, both Tutsi and Hutu converged upon public buildings, such as churches, schools, stadiums, sports fields, hospitals, and government offices. At first, they did so of their own volition: the country was plunged into chaos, and these locations ostensibly promised some form of sanctuary.  As emphasised in chapter 4, churches had customarily provided sanctuary during times of political turmoil. Initially, this assumption prompted the mass flight to these locations. One survivor interviewed by Mamdani said, ‘[w]e thought this was God’s house, no one would attack us here.’439 Another survivor related that ‘[w]e considered the church… a place… to be safe, a place where people can’t kill you.’440 This sentiment was widespread across the country. However, Janzen and Janzen claim that historical notions of security were no longer warranted: ‘[t]he church buildings that had afforded sanctuary in earlier attacks on Tutsi became sites of mass murder, a strategy that the officials of the genocide used willingly.’441  Regrettably, then, the large-scale flight to sites of refuge proved a great boon to the organizers of the genocide. According to Alan Kuperman, ‘[c]oncentrating a dispersed populace is an essential step in modern genocides that is usually orchestrated by the killers. The fact that Tutsi in Rwanda self-congregated rapidly saved the extremists a step and may explain the unprecedented speed of the genocide.’442 Within a few short days, Kuperman continues, ‘most of Rwanda’s more than half-million Tutsi were congregated at sites throughout the country, in groups typically ranging from a few hundred to tens of thousands.”443 For example, according to                                                439 Mamdani, p. 4.  440Samuel Totten and Rafiki Ubaldo. We Cannot Forget: Interviews with Survivors of the 1994 Genocide in Rwanda (New Brunswick: Rutgers University Press, 2011): p. 117.  441 John M. Janzen and Reinhild Kauenhoven Janzen, Do I Still Have a Life? Voices from the Aftermath of War in Rwanda and Burundi (Lawrence, Kansas; University of Kansas, 2000): p.  Italics added.   442 Alan J. Kuperman, The Limits of Humanitarian Intervention: Genocide in Rwanda (Washington D.C.: Brookings Institution Press, 2001), p. 131.  443 Kuperman, p. 15.   124  one survivor of the Gikongoro church massacre, by April 8th there were 40,000 people assembled at this single location.444 By the second week of the killing, the génocidaire strategy began, according to Des Forges, to ‘exploit… the Tutsi impulse to seek refuge and promised them protection if they would assemble in designated sites.’445 Until this point, movement towards these locations was largely the reflexive choice of the fleeing Tutsi. Local elites frequently told people to return to their home. At Ndera, for instance, a survivor claimed that the vice-president of the Hutu Power wing of the Liberal Party asked the priests to ‘chase us out so that we could go back to our homes.’446  Then, a week into the genocide, various principals actively encouraged people to assemble. This encouragement motivated people who were still in their homes, and who may have initially held back, to gather at these concentration sites. One survivor summarized this process, claiming that ‘[t]he refugees who had gathered at the commune office of Musubiya managed to arrive at Kaduha without casualties on the way. We kept wondering why the local authorities kept letting refugees, even from other communes, come to the parish. It was, without a doubt, to make it easier to kill everyone [in one go].’447 Local elites beholden to Hutu Power ideology facilitated the process of concentration by using their authority and influence to deceive the fleeing Tutsi. The prefect of Kibuye, Dr. Clement Kayishema, encouraged Tutsi to gather in parishes and at Gatwaro stadium.448 He positioned gendarmes throughout the prefecture with ‘explicit instructions’ to inform the fleeing Tutsi that they should seek refuge at these locations.449  Some authorities were particularly eager. In Kivumu commune, Des Forges writes, ‘the burgomaster reportedly drove a white pick-                                               444 Totten and Ubaldo, p. 81. 445Des Forges, “Leave none”, p. 301.  446 African Rights, p. 284.  447 African Rights, p. 320.  448 African Rights, p. 132.  449 African Rights, p. 132.   125  up truck around to gather Tutsi who were straggling along the road.’450 Sylvestre Gacumbitsi, the bourgmestre described in the introduction, sent Tutsi to die by the thousands at Nyarubuye. Additionally, influential church leaders lent their authority to concentrate Tutsi; for instance, Pastor Elizaphan Ntakirutimana was observed to be ‘personally instructing Tutsis to gather at the Adventist complex’ in Mugonero.451 According to Timothy Longman’s research in the prefectures of Gikongoro, Butare, and Kibuye, ‘a number of clergy and other church employees lured people into churches knowing that they would be killed.’452 Even soldiers were incorporated into this scheme. For instance, one survivor recalls how members of the FAR told refugees to gather at the Parish of Gahanga because they ‘were going to start shelling the area in order to hold back the RPF.’453  Concentration meant that thousands of refugees converged upon these sites. They were mostly confined within buildings and courtyards. At Kayumba, one survivor commented that ‘there were so many people crammed into the church that you could not find anywhere to put your feet.’454 As the month went on, more and more people arrived from further afield, many of them escaping other concentrations and massacres.455  At some such concentrations, the génocidaires bided their time until more Tutsi had gathered. One of the more infamous killers, John Yusufu Munyakiza, while describing the Parish of Nyabitimbo, said it was not worth his time to travel there to kill ‘such a small number of Tutsi.’456 The local interahamwe encouraged more Tutsi to gather at this site in order to make it ‘worthwhile’ for Yusufu.457  At Kaduha, gendarmes actually managed to hold back an                                                450 Des Forges, “Leave none”, p. 301.  451 Philip Gourevitch, We Wish to Inform You That Tomorrow We Will Be Killed With Our Families (New York: Farrar, Straus and Giroux, 1998), p. 27.  452Timothy Longman, “Church Politics and the Genocide in Rwanda” in Journal of Religion in Africa 31:2 (2001), p. 181.  453 African Rights, p. 277.  454 African Rights, p. 271.  455 African Rights, p. 327.  456 African Rights, p. 488. 457 African Rights, p. 489. .   126  interahamwe raid. Their motive, however, was not to protect the refugees, but to encourage the attackers to return with reinforcements.458  Finally, it must be noted that the drive to concentrate Tutsi was not immediately heeded or universally implemented. At the Parish of Nkanka, attackers initially insisted that the refugees return home, telling them that peace had been restored.459 On 20 April, a survivor of Hanika Parish recalled that more than a week into the concentration strategy, ‘[t]he préfet did not want to hear about refugees being piled up in parishes. They were to be finished off in cold blood in their sectors of origin.’460 Despite some exceptions, the push to gather as many Tutsi at these locations dominated the first phase of the perpetrators’ emergent military strategy. This pattern of concentration was, in Des Forges’s words, ‘so general throughout the country that it must have reflected orders from above.’   7.5.2b Phase Two: Degradation and Containment   Once assembled, the Tutsi were gradually and collectively weakened by the besieging génocidaires. Control over movement, previously focused on bringing Tutsi to such sites, turned to preventing escape from these elaborate traps. Deceptive and coercive techniques aimed at degrading the Tutsi ability to resist, while containing them in anticipation of their slaughter. This aspect of the emergent military strategy aimed to make the Tutsi less capable of resisting once massacre commenced. As argued above, the aim of military strategy during this phase was to establish the most favourable possible context for the engagement.  The génocidaires achieved the goals of degradation and containment through a variety of concurrent means. First, gendarmes were positioned around the sites, supposedly to ‘guard’                                                458 African Rights, p. 321. 459 African Rights, p. 492.  460 African Rights, p. 517.   127  those within. This was a ruse. The police were there not to protect Tutsi, but rather to keep them corralled. Initially, there seemed to be no reason to distrust the gendarmes. One survivor recalled how they were lulled into thinking that the killing was ‘between political parties’: ‘since we were simple people and not members of political parties we believed [the police] would protect us.’461 Later, many of these gendarmes ‘joined the ranks of the assassins and turned their arms upon the people they were ostensibly there to defend.’462  Second, officials and people in authority visited the sites and assured Tutsi of their safety. In almost every instance, this was a deception to lull the Tutsi into passivity. According to a survivor at Hanika Parish:  At about 4:00 pm, we saw the bourgmestre of Gatare, Fabien Rugwizangoga. He told us to calm ourselves down, saying that he was ready to understand our grievances. We stood around him. He told us that we should ask each other’s pardon. It was a way of disarming us because the militia had taken advantage of the occasion to surround us. The meeting did not lead to anything. He left and the militia invaded all the buildings of the parish.463    Third, the assembled Tutsi were disarmed. Many brought with them farming implements, such as hoes and machetes. They gathered rocks and sticks as a means of defense. At Nyamata, pews were torn apart in a desperate attempt to make wooden staves.464 A refugee describes the disarming process:                                                 461 Totten and Ubaldo, p. 26.   462 African Rights, p. 307.  463 African Rights, p. 517.  464 African Rights, p. 270.  128  The bourgmestre ordered a detailed search. The policemen looked in our pockets and bags. No guns were found. They took a few machetes and spears. Since many of the people had been sleeping in the bushes before they came to the parish, they kept these [instruments] as defense. The bourgmestre and his allies left us there without any object of defense.465  Fourth, the refugees were deprived of both food and water. After repelling an interahamwe attack on the Parish of Mibilizi, Tutsi supplies of water were cut off in an attempt to force them out.466 Likewise, the water to Shangi Parish was cut. One survivor of Shangi commented that ‘[f]or all the refugees, thirst was something intolerable.’467 Even when food was available, cooking outside was made difficult by harassment from soldiers and the interahamwe.468 Many of the refugees brought with them livestock such as cows, goats and sheep.469 These animals were stolen during militia raids, or used by refugees to purchase a quick death.  One survivor recalled that ‘[t]hey [the interehamwe] were standing outside the fence and they said, “Those of you who do not wish to be hacked to death with a hoe or massue, throw a goat or chicken over the fence, and you will be shot.”’470 When food supplies dwindled, the Tutsi were unable to leave to find sustenance. The gendarmes, supported by a cordon of villagers and roadblocks, prevented the refugees from leaving to seek food, water, and wood.471 Fifth, the refugees were infiltrated by spies and agitators. One survivor described a spy named ‘Ruteruzi.’ ‘B]etween the 9th and the 12th,’ she writes, Ruteruzi ‘was seen a lot at the                                                465 African Rights, p. 494.   466 African Rights, p. 525.  467 African Rights, p. 532.  468 African Rights, p. 272.  469 African Rights, p. 326.  470 Totten and Ubaldo, p. 26.  471 African Rights, p. 316  129  Parish of Hanika. He had come to spy and find out the number of refugees so that he could ensure sufficient distribution of weapons to the assassins.’472 Additionally, infiltrators systematically separated Hutu from Tutsi. Many Hutu gathered at these sites for sanctuary, often because they had a Tutsi spouse or relative. Infiltrators attempted to convince the Hutu that the target was the Tutsi, and that they were encouraged to leave prior to the impending massacre.473 These Hutu who sought sanctuary with the Tutsi often turned against them, sharing their weaknesses and defensive strategies. Members of the FAR, interahamwe, and gendarmerie reportedly checked everyone’s ethnic ID card to ensure that only Tutsi remained.474  7.5.2c Phase Three: Eradication  Once the assembled Tutsi were degraded and contained and the killers were sufficiently reinforced, the final annihilatory push could commence. In smaller concentrations, the defenders had invariably been overrun earlier by militia raids. At the larger concentrations, the sheer volume of people allowed Tutsi to fight off the attackers, albeit with major loss of life.   Perpetrator reinforcements profoundly changed the dynamic of violence. Previously, the defending Tutsi were armed with comparable agricultural implements and improvised weaponry.  Albeit at a terrible cost, the Tutsi defenders initially held their own. However, the arrival of guns and explosives and trained soldiers, though few in number, shifted the tide in favor the génocidaires, ultimately crushing Tutsi resistance.  In many instances, the killing commenced along gender lines, with men and boys targeted first. At the commune office in Kamembe, a survivor reported that ‘[the interahamwe]                                                472 African Rights, p. 514-15 473 Des Forges, “Leave none”, p. 161.  474 Des Forges, “Leave none”, p. 161.   130  said that they first wanted male refugees. A communal policemen, Gatera, made all the men walk out. Some had tried to disguise themselves. But it was no use. Gatera knew them and their features. All the men and boys were killed that day.’475  The interahamwe ‘did not hesitate to lift skirts and dresses to check the sex of their victims.’476  Another survivor describes a similar situation at the Parish of Mibilizi:  They took all the men and boys, everyone masculine from about the age of two. Any boy who could walk was taken. They put them to one side. They were particularly interested in men who looked like students, civil servants, in short anyone who looked as if he had education or money. They left only very poor men, those who were already wounded, and tiny babies. Not even the very old were spared. They were all killed with machetes, spears and swords.477   Afterwards, one of the interahamwe was heard to say: ‘[w]e were ordered to kill only Tutsi men and boys for today.'478 In other instances, the targeted killings of the original military strategy were carried over. While the primary opposition targets were eliminated in Kigali over the course of a few days, there were still many locally influential individuals left at the communal level. A survivor of the Mugonero hospital massacre reported that names were read over a loudspeaker prior to the commencement of the killings.479 After the massacres, the génocidaires would comb through the bodies in an effort to identify particular individuals on their kill lists.480                                                 475 African Rights, p. 508  476 African Rights, p. 517.  477 African Rights, p. 522.  478 African Rights, p. 522.   479 Des Forges, “Leave none”, p. 159.  480 Des Forges, “Leave none”, p. 159.   131  For the duration of the genocide, firearms and bullets were a finite commodity. Verwimp notes that ‘[t]he objective of the regime was to kill as many Tutsi as possible under the constraint that firearms and bullets were in short supply.’481 Thus, guns were used sparingly, but to great effect. Often it was a member of the local elite who used firearms to signal the beginning of the massacre. For instance, the préfet of Kibuye, Clement Kayishema, reportedly fired the first shot that ‘triggered the carnage at Gatwaro stadium.’482 Guns were used to target the strongest fighters, usually men on the frontline. In addition to their effectiveness as killing instruments, firearms were used to shock the captive Tutsi into confused submission. For instance, at Gahanga, soldiers fired into the air to get the defenders to drop the rocks they were using to protect themselves.483  In addition to firearms, grenades were responsible for many of the deaths. They were thrown into the crowded buildings through windows and roofs; in this way, much of the killing was committed from outside.484 With hundreds or thousands of people tightly gathered in buildings, grenades could devastate those assembled within. Not only explosive grenades were used, but also pepper bombs that ‘made [victims] cough and choke.’485 After saturating buildings with bullets and grenades, interahamwe and local Hutu peasants stormed inside with machetes, swords, spears, and massues (nailed clubs). The génocidaires could kill without the further expenditure of bullets. This combination of firearms, explosives, and melee weapons created chaos among the refugees, hindering both defense and flight. An anonymous survivor interviewed by Samuel Totten and Rafiki Ubaldo describes the mortal effect of crowds:                                                 481Philip Verwimp, "Machetes and Firearms: The Organization of Massacres in Rwanda" in Journal of Peace Research 43:1 (2006): p. 7-8. 482 African Rights, p. 132. 483 African Rights, p. 277.  484 African Rights, p. 270.   485 African Rights, p. 272.   132   [T]here was total confusion. No control of anything; swarms of people [moving everywhere]. The militia were striking people with machetes and killing them, others were yanking people into the larger crowd; if a person next to you got pulled into the crowd, there was nothing you could do, it was over. It was chaos with people moving everywhere, being beaten, screaming, and crying… If you stepped in a single direction you could not reverse. You were part of the movement and you are not walking like you do every day, you are being stepped on, moved forward [against your will], and there is nothing you can do to choose which way you will go.486  Tutsi were inexorably drawn into this writhing morass, unable to escape and left at the mercy of the killers.   This phase was marked by extremes of cruelty and sadism. Caplan identifies some of the most horrendous practices: pregnant woman had their wombs ripped open, family members were forced to kill one another, people were thrown alive into latrine pits, and children were forced to watch their parents killed.487 The killers spent hours at these grisly tasks, often working from 9 in the morning to 5 at night.   7.5.2d Phase Four: Annihilation  The génocidaires sought to ensure that not a single Tutsi survived the massacres. Among the corpses that littered the interior and exterior of these public sites, a few individuals were able to hide amongst the dead. At Gahanga, the killers deployed cows to walk amongst                                                486 Totten and Ubaldo, p. 70-71.   487 International Panel of Eminent Personalities, sec. 14.27.   133  the dead. If the animal trod on a survivor who showed signs of life, the victim was quickly dispatched by handheld implements.488 Another method of finding survivors was the daubing of hot piri piri chili into open wounds. Again, if this evoked a sign of life, the Tutsi was immediately killed. At Mugonero, tear-gas canisters were lobbed among the bodies to make the survivors cough.489 Elsewhere, such as at the Parish of Kibeho, the killers simply set the building on fire, immolating the dead and living within.490 Nearby at the College of Kibeho, Tutsi were reportedly buried alive.491 With almost all Tutsi dead, it remained to remove the evidence of both the crime and the victim. At a number of locations, bulldozers were employed to load bodies into the backs of large trucks.492 Many took up Mugesera’s call to send the Tutsi back to Ethiopia via the Nyabarongo river. Thousands of bodies were thrown into the waterway.  At many locations, bodies were unceremoniously buried in mass graves. In the aftermath of the genocide, the new Ministry of Higher Education, Scientific Research and Culture established a commission to identify the main massacre sites. Its final reports lists dozens of mass graves, many hidden within septic tanks. The number of bodies in these tanks ranged from a dozen to a few hundred.493   7.6 Summary   This chapter’s focus upon ‘military strategy’ supplies the final component of the strategy-hierarchy framework. Policy, grand strategy, and military strategy are all constitutive parts of the                                                488 African Rights, p. 277.  489 Des Forges, “Leave None”, p. 162.  490 African Rights, p. 307.  491 African Rights, p. 310.   492 Totten and Ubaldo, p. 27.  493 Ministry of Higher Education, Scientific Research and Culture, Preliminary Report on Identification of Sites of the Genocide and Massacres That Took Place in Rwanda From April to July 1994 (Kigali: 1996).  134  ‘strategy’ implemented by Rwanda’s political elite. This chapter has argued that massacres generally focus on non-adversarial targets. While massacres had occurred prior to April 1994, they lacked the element of concentration, which kept the overall death toll relatively low. The assassination of President Habyarimana, however, initiated a dramatic shift in military strategy. First, moderate Hutu and members of the political opposition were assassinated. These were people who could potentially interfere with the machinations of the akazu.  Once these targets were largely eliminated, the génocidaires perceived an opportunity. Hundreds of thousands of Tutsi had converged upon sites of refuge. According to the genocidal ideology of Hutu Power, the political opposition and the Tutsi ethnic group were accomplices to the RPF. With the political opposition devastated, attention could turn to the next enemy: the Tutsi. Thus, a week after the assassination of Habyarimana, the perpetrators began implementing their second military strategy, which consisted of four phases. First, Tutsi were gathered at specific spots. Second, they were weakened and contained. Third, they were murdered in massacres, often on a huge scale. Finally, survivors were eliminated, along with evidence of the crime.    135  Chapter 8: Conclusion   A ‘strategic’ understanding of the Rwandan genocide allows us to trace its development, beginning with its policy origins, continuing to the diverse grand strategic means used to realize this objective, and concluding with the ultimate instrumental use of violence.  The original strategy against the opposition and the emergent strategy against the Tutsi cannot be understood in isolation from the policy goals of the entrenched Hutu elite. Reeling from three ‘strategic shocks’, they developed and promoted the Hutu Power ideology which identified the ‘enemy’ (i.e. the cause of their problems) as the Tutsi, the political opposition, and the RPF. Before turning to mass killing, the Rwandan elite pursued its policy goal of regime survival via diverse grand-strategic means.  After the assassination of President Habyarimana, the architects of the grand strategy shifted their focus to military strategy. Two such strategies took shape: the first targeted the opposition, while the second emerged a week into the killing and focused upon the Tutsi ethnic group. Unlike the RPF, both groups were non-adversarial, and thus profoundly vulnerable.  In this conclusion, I aim to answer the two research questions posed in the introduction. I then explore the limitations of the study, and conclude with suggestions for future research.   8.1. Question 1: ‘How does strategic theory assist in our understanding of the organization and implementation of genocidal violence and massacres?’  In answering this question, I assert that strategic theory provides a foundation to develop an analytical framework that applies to both war and genocide. While strategic theory is traditionally reserved for conventional war and battles, this thesis argues that strategic theory is equally relevant to genocide and massacre. At its core, strategy concerns the management of  136  instrumental violence. It is relevant to both war and genocide because participants in both utilize and direct force for a specific purpose. Thus, genocide - like war - is a strategic activity. ‘Strategy’ is not a concept conducive to a simple definition. Instead, this thesis draws on key themes in strategic theory to develop a ‘strategy hierarchy’ framework. This framework is a heuristic device that breaks down ‘strategy’ into its component parts. ‘Strategy’ is divided into three dimensions: policy, grand strategy, and military strategy. Each aspect is nested within the overarching strategy hierarchy. Policy is dominant, and determines the overall objectives to be pursued. Grand strategy utilizes the diverse means at its disposal and directs them to realize these goals; finally, while military strategy is a single subordinate component within an overarching grand strategy, its role in managing the instrumental use of violence makes it especially relevant.  This framework – subsuming policy, grand strategy and military strategy – may be applied to both conventional war and genocide. However, one caveat must be highlighted. Strategic thought assumes that conflict will take the form of adversary versus adversary. In other words, each side in a conflict is assumed to be armed, organized, and capable of offensive and defensive behavior. Genocide, in contrast, pits an adversary against a non-adversarial target; the ‘enemy’ is a poorly armed and unorganized group unable to offer credible resistance.  This caveat, however, does not fundamentally obstruct the application of strategic theory to genocide. Perpetrators of genocide identify certain categories of people as a threat; this assessment is justified by an ideological worldview held and propagated by elite policymakers. Ideology takes on genocidal connotations when a non-adversarial group is identified, and efforts are made to impose a violent solution. Thus, from the perpetrator’s perspective, the targets of genocide represent such a threat that they are seen as analogous to legitimate adversarial targets.   137  8.2 Question 2: ‘How does strategic theory help to explain the synchronized pattern of violence that gripped Rwanda for 100 days in 1994?’  The second question can be addressed by applying this three-part analytical framework to the Rwandan case. Chapter 5 discussed the policy dimension. From 1973 to July 1994, Rwanda was dominated by a clique of influential Hutu known as the akazu. After his ascension to power in a coup, President Juvénal Habyarimana ruled the country while favoring those from his base in north-western Rwanda. From this region was drawn an elite group that dominated politics, the economy, and the civil service. They grew comfortable with these perks, and lashed out against all challenges to their dominance.  Beginning in the late 1980’s, three ‘strategic shocks’ shook the foundations of akazu hegemony. First came the economic difficulties that accompanied the crash of international coffee prices, and the elites were prevented from skimming off foreign aid. Second, the end of the Cold War set off a wave of democratization across the continent. There were international and domestic calls to reform the MRND’s one-party state and to allow multiparty elections. Third, the Rwandan Patriotic Front, a rebel army of refugee Tutsi, invaded from Uganda. They threatened to overthrow the akazu and return the Hutu to a subordinate role.  Taken together, these three ‘strategic shocks’ represented an existential threat to Rwandans who benefited from the status quo. In this milieu, the Hutu Power ideology emerged. This worldview gave expression to the fears of the akazu and their supporters. Significantly, it identified the chief culprits allegedly behind the multifaceted challenge to the regime. Three groups emerged as threats in this ideological discourse: the RPF army, the Hutu opposition, and their alleged Tutsi accomplices. One of these groups is adversarial: the RPF was armed and organized, and was certainly willing to engage in combat. In contrast, the two other groups - the Hutu opposition and civilian Tutsi - were explicitly non-adversarial. When the final ‘strategic  138  shock’ came in the form of President Habyarimana’s assassination, all three groups were amalgamated in the regime’s mind, and targeted for destruction. After policy, the next component of the ‘strategy hierarchy’ analyzed in this thesis was grand strategy. Chapter 6 focused upon akuzu grand strategy from the 1 October 1990 RPF invasion to the 6 April assassination of President Habyarimana. During this time, the regime used every means at its disposal - short of extensive killings - in order to prevent a further erosion of their powers. Before turning to massively violent means, military strategy was concerned with developing capabilities, including expanding the army, establishing specialized fighting units such as the Presidential Guard, and creating loyal party-based militias such as the interahamwe. Diplomatic strategy attempted to shore up support from foreign allies, isolate the RPF, and derail international attempts to solve Rwanda’s internal challenges, including those posed by the Arusha Accords and the introduction of UNAMIR peacekeepers. Political strategy involved creating far-right parties, such at as the CDR, aimed at disrupting the democratic process, as well as dividing opposition parties from within. Legal and other strategies attempted to establish a climate of impunity that was also buttressed by the Rwandan Catholic Church. Economic strategy focused on buttressing and expanding the resource base of the akazu at the expense of the rest of the population. Finally, cultural strategies saw the development of hate radio and publications propagating the Hutu Power ideology. Influential party members and high-ranking officials further diffused a culture of anti-Tutsi and anti-opposition discourse. Taken together, these various strategies were complementary. Each sought to protect the interests of the akazu while preparing the ground for the ultimate use of instrumental violence.  The final aspect of the strategy hierarchy is military strategy. Such strategy is unique in that it harnesses the violence resources at the disposal of the state. Two types of military strategy emerged after the assassination of President Habyarimana. The first and original one focused on individuals perceived as opposing the continued dominance of the akazu. These  139  targets were largely Hutu, the killing was performed by the Presidential Guards, and the strategy was largely implemented within the first week.  At that point, a second military strategy emerged. As outlined in chapter 4, during times of political turmoil Tutsi (and Hutu) traditionally fled to the churches and other sites of refuge. Until April 1994, those who took refuge in such locations were largely unmolested. After the assassination, hundreds of thousands of Tutsi converged upon such sites, assuming they would receive sanctuary. After eliminating the political opposition, the killers were confronted with an opportunity to target their remaining ‘adversary’.  The massacres of the Tutsi typically followed four phases. First, Tutsi were concentrated at specific locations. This was facilitated by exploiting the traditional sanctuaries and mobilizing the Hutu civilian population. Once Tutsi had assembled, the perpetrators sought to weaken them and prevent their escape. This facilitated the next phase, the eradication of the Tutsi through mass violence, featuring the use of firearms, explosives and handheld implements. Finally, the survivors were eliminated and the bodies disposed of.   8.3 Limitations   A limitation inherent in almost all comparative genocide studies is a paucity of honest perpetrator accounts. In order to avoid justice or retribution, those responsible for genocide obfuscate their role, and seldom describe in detail how they organized and implemented a campaign of mass violence against a civilian population. There are rare exceptions to this rule. For instance, Jean Hatzfeld interviews a number of perpetrators in his book Machete Season: The Killers in Rwanda Speak. Scott Straus in The Order of Genocide: Race, Power and War in Rwanda goes a step further, drawing upon perpetrator testimony as the basis for theory development. These books, however, focuses exclusively upon low-level perpetrators. Regrettably, we do not have access to verified accounts  140  from high-level génocidaires, i.e. those individuals responsible for formulating policy, setting grand-strategic priorities, and implementing military strategy.   Analyzes of perpetrator strategy must thus be derived from other sources. For instance, it is possible to discern patterns of perpetrator actions that were consistently reported in survivor accounts. This thesis attempts to ‘recreate’ perpetrator strategy by drawing upon such testimony. Describing the massacre process, these accounts highlight broadly analogous experiences. Descriptions of the killing process at various killing sites inform the analysis of the four-phased emergent military strategy in the previous chapter.  In contrast, traditional strategic studies seldom utilizes the accounts of victims and survivors to inform their theories. Generals frequently become autobiographers and historians, and are seldom reticent about emphasizing their own military prowess. There is no parallel for this trend and scholarly resource in genocide studies. In short, while ‘genocide’ is amenable to strategic analysis (as this thesis has sought to show), it must rely upon alternative sources of data and evidence.   8.4 Future Directions for Research   Strongly motivating the development of the ‘strategy hierarchy’ heuristic is its potential universality. Even with its inherent limitations, the framework is readily applicable to other instances of genocide. By focusing upon the interconnections of policy, grand strategy and military strategy, it is possible to analyze historical and current genocides - along with similar eruptions of mass violence - along three crucial dimensions.  The Rwandan genocide was animated by a clear policy, pursued by a rigorous grand strategy, and ultimately realized via two deliberate military strategies. In other words, Rwanda typifies the ideal strategy hierarchy. However, not every episode labeled as ‘genocide’ clearly accords with each of these dimensions. For instance, an explicit policy that requires the  141  extermination of a category of people may be absent. Rather, deaths may accrue as a result of willful negligence, such as during the famines in China and North Korea. Alternatively, there may be no articulated grand strategy directing destruction. For instance, colonial genocides may be defined by localized settler violence, rather than by a coordinated plan for comprehensive destruction. Finally, genocide may lack a military strategy. Instances of this include the Ukrainian Holodomor, during which millions of lives were destroyed, but not as a result of military campaigning. Additionally, the ‘cultural genocide’ of First Nations in Canada was not carried out via extensive massacres, but by other insidious means.  By analyzing mass killing along these three dimensions - policy, grand strategy and military strategy - past, present, and future instances of genocide may be analyzed and better understood. This thesis has shown how a key instance of genocide is amenable to a strategic analysis. It is to be hoped that others will join the author in applying this framework to a broader set of cases.     142  Bibliography   African Rights. Rwanda: Death, Despair and Defiance. London: African Rights, 1995.   African Union, Rwanda: The Preventable Genocide, July 2000.   Alvarez, Alex. Governments, Citizens, and Genocide: A Comparative and Interdisciplinary Approach. Bloomington: Indiana University Press, 2001.   Arendt, Hannah. On Violence. Orlando: Harcourt, Inc., 1970.   - - - The Origins of Totalitarianism. Orlando: Harcourt Brace & Company, 1979.    Bartholomees, Jr., J. Boone.  'A Survey of Strategic Thought.' In U.S. Army War College Guide to National Security Policy and Strategy. Strategic Studies Institute, 2004.   Bartrop, Paul. "The Relationship Between War and Genocide in the Twentieth Century: A Consideration." In Journal of Genocide Research 4:4 (2002).   Bent, Samuel Arthur. Familiar Short Sayings of Great Men. Boston: Ticknor and Co., 1887.    Bessel, Richard. “Functionalists vs. Intentionalists: The Debate Twenty Years on or Whatever Happened to Functionalism and Intentionalism?” In German Studies Review, 26:1.   Brodie, Bernard. "Strategy as a Science." In World Politics 1:4 (July 1949).    - - -  ‘The Continuing Relevance of On War.’ In On War, trans. Michael Howard and Peter Paret. Oxford: Oxford University Press, 1989.  - - - War and Politics. London: Cassell, 1973.  Butcher, Thomas M. "A ‘Synchronized Attack’: On Raphael Lemkin's Holistic Conception of Genocide." In Journal of Genocide Research 15:3 (2013).    Carlton, Eric. Massacres: An Historical Perspective. Hants: Scolar Press, 1994.   Chalk, Frank and Jonassohn, Kurt. The History and Sociology of Genocide: Analyses and Case Studies. New Haven: Yale University press, 1990.   Clausewitz, Carl Von.  On War. trans. Michael Howard and peter Paret. Oxford: Oxford University Press, 1976.   Curthoys, Ann  and Docker, John. “Defining Genocide,” in The Historiography of Genocide, ed. Dan Stone. London: Palgrave Macmillan, 2008.   Dadrian, Vahakn N. 'The Determinants of the Armenian Genocide.' In Journal of Genocide Research 1:1.   143   Dallaire, Romeo. Shake Hands With the Devil: The Failure of Humanity in Rwanda. Random House, 2008.   Des Forges, Alison. "Call to Genocide: Radio in Rwanda, 1994." In The Media and the Rwanda Genocide ed. Allan Thompson. Kampala: Fountain Publishers, 2007.  - - - ‘Leave None to tell the Story’: Genocide in Rwanda. New York: Human Rights Watch, 1999.   Destexhe, Alain. Rwanda and Genocide in the Twentieth Century. New York : New York University Press, 1995.   Earle, Edward Mead. ‘Introduction.’ In Makers of Modern Strategy. Princeton: Princeton University press, 1943.    Eccles, Henry E.  Military Concepts and Philosophy. New Brunswick, N.J.: Rutgers Univ. Press, 1965.   Feierstein, Daniel. 'The Concept of 'Genocidal Social Practices.' In New Directions in Genocide Research ed. Adam Jones. Routledge, 2011.   Fein, Helen. "Accounting for Genocide After 1945: Theories and Some Findings." In International Journal on Group Rights 1 (1993).   - - - “Defining Genocide as a Sociological Concept.” In Current Sociology 38: 8 (1990).  - - - "Revolutionary and Antirevolutionary Genocides: A Comparison of State Murders in Democratic Kampuchea, 1975 to 1979, and in Indonesia, 1965 to 1966." In Comparative Studies in Society and History 35:4 (1993).    Freedman, Lawrence. Strategy: A History. Oxford: Oxford University press, 2013.   Freier, Nathan. Known Unknowns: Unconventional ‘Strategic Shocks’ in Defense. Carlisle, Pa: U.S. Army War College, 2008.   Fujii, Lee Ann. “The Puzzle of Extra Lethal Violence” in Seminar Series on Civil War and the Challenge of Political Order (2011).   Goldhagen, Daniel Jonah. Worse Than War: Genocide, Eliminationism and the Ongoing Assault on Humanity. New York: PublicAffairs, 2009.    Gourevitch, Philip. We Wish to Inform You That Tomorrow We Will Be Killed With Our Families. New York: Farrar, Straus and Giroux, 1998.   Gray, Colin S. The Strategy Bridge: Theory for Practice. Oxford: Oxford University Press, 2010.   144  Grunfeld, Fred and Huijboom, Anke. The Failure to Prevent Genocide in Rwanda: The Role of Bystanders. Leiden: Transnational Publishers, 2007.    Handel, Michael I. Masters of War: Classical Strategic Thought. Portland: Frank Cass Publishers, 2001.   Harff, Barbara , 'No Lessons Learned from the Holocaust? Assessing Risks of Genocide and Political Mass Murder Since 1965." In The American Political Science Review 97:1.    Harff, Barbara and Gurr, Ted Robert. "Toward Empirical Theory of Genocides and Politicides: Identification and Measurement of Cases Since 1945" in International Studies Quarterly 32:3 (1988).   Hart, Liddell. Strategy. New York: Meridian, 1991.   Hintjens, Helen M. "Explaining the 1994 Genocide in Rwanda." In The Journal of Modern African Studies 37:2.   Hirondelle News Agency, “Two Former Rwandan Mayors Arrested for Genocide.”  hirondellenews.com. 21 June 2001.  Huttenbach, Henry R. "In Favor of the Genocidal: A Caveat for the Academic." In Journal of Genocide Research 1:3 (1999).   - - - "Towards a Conceptual Definition of Genocide." In Journal of Genocide Research 4:2 (2002).   International Panel of Eminent Personalities, Rwanda: The Preventable Genocide. African Union, 2000.    Janzen, John M. and Janzen, Reinhild Kauenhoven. Do I Still Have a Life? Voices from the Aftermath of War in Rwanda and Burundi. Lawrence, Kansas; University of Kansas, 2000.  Jones, Adam. Genocide: A Comprehensive Introduction. New York: Routledge/Taylor and Francis Publishers, 2010.   Jones, Bruce D. ‘The Arusha Peace Process.’ In The Path of a Genocide: The Rwanda Crisis from Uganda to Zaire, eds. Howard Adelman and Astri Suhrke. New Brunswick, New Jersey: Transaction Publishers, 2000.   Jomini, Antoine Henri de. The Art of War. trans. G.H. Mendell and W.P. Craighill. Rockville: Arc Manor, 2007.   Judgement, Ferdinand Nahimana, Jean-Bosco Barayagwiza and Hassan Ngeze (ICTR-99-52-T), Trial Chamber 1, 3 December 2003.   Judgement, Sylvestre Gacumbitsi (ICTR-2001-64-A), Appeals Chamber, 7 July 2006.  145   Judgement, Theoneste Bagosora, Gratien Kabiligi, Aloys Ntabakuze, Anatole Nsengiyumva (ICTR-98-41-T 18), Trial Chamber 1, 18 December 2008.   Jurkovich, Michelle. "When Food is the Weapon: Examining the Strategy of Starvation in Cases of Genocide." APSA Conference (2010).   Kakwenzire, Joan and Kamukama, Dixon. ‘The Development and Consolidation of Extremist Forces in Rwanda 1990 - 1994.’ In The Path of a Genocide: The Rwanda Crisis from Uganda to Zaire, eds. Howard Adelman and Astri Suhrke. New Brunswick, New Jersey: Transaction Publishers, 2000.   Kamukama, Dixon. Rwanda Conflict: Its Roots and Regional Implications. Kampala: Fountain Publishers Ltd., 1997.   Kangura, Appeal to the Bahutu Conscience (December 6 1990).   Kangura, Issue No. 26.   Katz, Steven. The Holocaust in Historical Context, Vol. 1: The Holocaust and Mass Death before the Modern Age. New York: Oxford University Press, 1994.  Kautilya, Arthasastra, trans. R. Shamasastry. Available online: https://archive.org/details /Arthasastra_English_Translation   Keane, Fergal. Season of Blood: A Rwandan Journey. London: Viking, 1995.    Kedar, Asaf. "Ideal Type." In Encyclopedia of Political Theory. Thousand Oaks: Sage Publications Inc., 2010.   Kennedy, Paul. “Grand Strategy in War and Peace:Toward a Broader Definition.” In Grand Strategies in War and Peace ed. Paul Kennedy. New Haven: Yale University Press, 1991.  Kershaw, Ian. The Nazi Dictatorship. Problems and Perspectives of Interpretation. London: Edward Arnold, 1985.   Khan, Shaharyar M. The Shallow Graves of Rwanda. London: I.B. Tauris & Co. Ltd, 2000.   Kiernan, Ben. Blood and Soil: A World History of Genocide and Extermination from Sparta to Darfur. New Haven: Yale University Press, 2007.   Kinzer, Stephen. A Thousand Hills: Rwanda's Rebirth and the Man who Dreamed it. Hoboken, N.J.: John Wiley & Sons, 2008.   Kuper, Leo. Genocide: Its Political Use in the Twentieth Century. New Haven: Yale University press, 1982.   146  Kuperman, Alan J. The Limits of Humanitarian Intervention: Genocide in Rwanda. Washington D.C.: Brookings Institution Press, 2001.    Lemarchand, René. Rwanda and Burundi. London: Pall Mall Press, 1970.    - - - "Rwanda: The Rationality of Genocide" in Issue: A Journal of Opinion 23:2 (1995).   - - - ‘The Burundi Killings of 1972.’ In Online Encyclopedia of Mass Violence (27 June 2008).   Lemkin, Raphael. Axis Rule in Occupied Europe: Laws of Occupation - Analysis of Government - Proposals for Redress. Washington, D.C.:  Carnegie Endowment for International Peace, 1944.   - - - Lemkin on Genocide. Lanham: Lexington Books, 2012.   Levene, Mark. Genocide in the Age of the Nation State Vol. 1: The Meaning of Genocide. London: I.B. Tauris, 2005.   - - - “Introduction.” In The Massacre in History. Eds. Mark Levene and Penny Roberts. Oxford: Berghahn Books, 1999.   Longman, Timothy. Christianity and Genocide in Rwanda. Cambridge: Cambridge University Press, 2010.   - - - “Church Politics and the Genocide in Rwanda.” In Journal of Religion in Africa 31:2 (2001).  Luttwak, Edward N. Strategy: The Logic of War and Peace. Cambridge: Belknap Press, 1987.   Lykke, Arthur F. “Defining Military Strategy.” In Military Review 69:5.   Malik, S.K. The Quranic Concept of War. Delhi: Adam Publishers and Distributors, 1992.  Mamdani, Mahmood. When Victims Become Killers: Colonialism, Nativism, and the Genocide in Rwanda. Princeton: Princeton University Press, 2001.    Markusen, Eric. "Genocide and Warfare." In Genocide, War and Human Survival, eds. Charles B. Strozier and Michael Flynn. Lanham: Rowman and Littlefield Publishers, Inc., 1996.   Markusen, Eric and Kopf, David. The Holocaust and Strategic Bombing: Genocide and Total War in the Twentieth Century. Westview Press, 1995.   Melvern, Linda. A People Betrayed: The Role of the West in Rwanda's Genocide. London: Zed Books, 2009.   - - - Conspiracy to Murder: The Rwandan Genocide. Verso, 2006.    147  Ministry of Higher Education, Scientific Research and Culture, Preliminary Report on Identification of Sites of the Genocide and Massacres That Took Place in Rwanda From April to July 1994. Kigali: 1996.    Mintzberg, Henry. 'The Strategy Concept I: Five Ps For Strategy.' In California Management Review (Fall 1987).   Mironko, Charles K. 'Ibitero : Means and Motive in the Rwandan Genocide.' In Genocide in Cambodia and Rwanda: New Perspectives ed. Susan E. Cook. New Brunswick: Transaction Publishers, 2009.   Moran, Daniel Strategic. ‘Theory and the History of War.’ In Strategy in the Contemporary World. New York: 2003.   Mosca, Gaetano. The Ruling Class. New York: McGraw-Hill Book Company, Inc. 1939.  Moses, Dirk. “Raphael Lemkin, Culture and the Concept of Genocide,’ in The Oxford Handbook of Genocide Studies, eds. Donald Bloxham and Dirk Moses. Oxford: Oxford University Press, 2010.   Ndiaye, B.W. "Question of the Violation of Human Rights and Fundamental Freedoms in Any Part of the World, With Particular Reference to Colonial and Other Dependent Countries and Territories: Extrajudicial, Summary or Arbitrary Executions” United Nations Economic and Social Council (April 1993).    Newbury, Catharine. "Background to Genocide: Rwanda." In Issue: A Journal of Opinion 23:2 (1995).    Office of the UN Special Advisor on the Prevention of Genocide, Analysis Framework. Online:  http://www.un.org/en/preventgenocide/adviser/pdf/osapg_analysis_framework.pdf.   Orth, Richard. ‘Rwanda’s Hutu Extremist Insurgency: An Eyewitness Perspective.' In Genocide in Cambodia and Rwanda: New Perspectives ed. Susan E. Cook. New Brunswick: Transaction Publishers, 2009.   Otunnu, Ogenga . ‘Rwandese Refugees and Immigrants in Uganda.’ In The Path of a Genocide: The Rwanda Crisis from Uganda to Zaire, eds. Howard Adelman and Astri Suhrke. New Brunswick, New Jersey: Transaction Publishers, 2000.   Platias, Athanassios G.  and Koliopoulos, Constantinos. Thucydides on Strategy. London: Hurst, 2010.   Prunier, Gerard. The Rwanda Crisis: A History of Genocide. Columbia University Press, 1995.    Roth, John K. “The Politics of Definition.” In Will Genocide Ever End? eds. Carol Rittner, John K. Roth and James M. Smith. St. Paul: Paragon House, 2002.    148  Rummel, R.J. "Democracy, power, genocide, and mass murder." In The Journal of Conflict Resolution 39 (1995).   Sartre, Jean-Paul. “Statement 'On Genocide'” in Second Session of the Bertrand Russell International War Crimes Tribunal on Vietnam (1967).   Sawyer, Ralph D. The Seven Military Classics of Ancient China. Boulder: Westview Press, Inc., 1993.    Schabas, William A. "Convention for the Prevention and Punishment of the Crime of Genocide." In United Nations Audiovisual Library of International Law, http://www.un.org/law/avl. 2008.    Schwartz, Peter and Randall, Doug . "Ahead of the Curve: Anticipating Strategic Shocks." In Blindside: How to Anticipate Forcing Events and Wild Cards in Global Politics, ed. Francis Fukuyama. Washington, D.C.: Brookings Institution, 2007.   Scruton, Roger. The Palgrave Macmillan Dictionary of Political Thought. London: Palgrave Macmillan, 2007.   Sebarenzi, Joseph. God Sleeps in Rwanda: A Journey of Transformation. Simon and Schuster, 2009.   Segal, Aaron. 'Massacre in Rwanda.' In Fabian Research Series 240 (April, 1964).    Semelin, Jacques. “From Massacre to the genocidal process.” In International Social Science Journal, No.174 (2002).   - - - “In Consideration of Massacres.” In Journal of Genocide Research Vol. 3, No. 3 (2001).   - - - Purify and Destroy: The Political Uses of Massacre and Genocide. New York: Columbia University Press, 2007.   - - - 'Toward a Vocabulary of Massacre and Genocide.' In Journal of Genocide Research 5:2.   Shaw, Martin. 'Conceptual and Theoretical Frameworks for Organised Violence.' In International Journal of Conflict and Violence 3:1.   - - - War and Genocide: Organised Killing in Modern Society. Wiley, 2003.   - - - What is Genocide? Cambridge: Polity Press, 2007.   Smith, Roger W. ‘As Old as History.’ In Will Genocide Ever End? eds. Carol Rittner, John K. Roth and James M. Smith. St. Paul: Paragon House, 2002.   Stark, Tamás. "Genocide or Genocidal Massacre?: The Case of Hungarian Prisoners in Soviet Custody." In Human Rights Review, April-June (2000).    149  Strachan, Hew. "The Lost Meaning of Strategy." In Survival: Global Politics and Strategy 47:3.  - - -  ‘Strategy and Contingency.' In International Affairs 87:6.   Straus, Scott. "Destroy Them to Save Us": Theories of Genocide and the Logics of Political Violence." In Terrorism and Political Violence 24:4 (2012).   - - - 'How many perpetrators were there in the Rwandan genocide? An estimate.' In Journal of Genocide Research 6:1.   - - -  “Second-Generation Comparative Research on Genocide" in World Politics 59:3 (2007).   - - - The order of genocide : race, power, and war in Rwanda. Ithaca: Cornell University Press, 2006.   Totten, Samuel . and Ubaldo, Rafiki. We Cannot Forget: Interviews with Survivors of the 1994 Genocide in Rwanda. New Brunswick: Rutgers University Press, 2011.   Uekert, Brenda K. Rivers of Blood: A Comparative Study of Government Massacres. London: Praeger, 1995.   United Nations General Assembly, Convention on the Prevention and Punishment of the Crime of Genocide, United Nations Treaty Series, vol. 78. 9 December 1948. .  U.S. Marine Corp, MCDP 1-1: Strategy. Washington D.C: Department of the Navy, 1997.  Uvin, Peter. Aiding Violence: The Development Enterprise in Rwanda. West Hartford, Connecticut: Kumarian Press, Inc., 1998.   Valentino, Benjamin A. Final Solutions: Mass Killing and Genocide in the 20th Century. Ithaca: Cornell University Press, 2004.   - - - 'Why We Kill: The Political Science of Political Violence Against Civilians.' In Annual Review of Political Science 17 (2014).  Verdeja, Ernesto. "On Situating the Study of Genocide within Political Violence." In Genocide Studies and Prevention, 7:1 (Spring 2012).  Verwimp, Philip. "Machetes and Firearms: The Organization of Massacres in Rwanda." In Journal of Peace Research 43:1 (2006).   Waltz, Kenneth N. Theory of international politics. Reading, Mass.: Addison-Wesley Pub. Co., 1979.   Walzer, Michael. Just and Unjust Wars. New York: Basic Books, Inc., 1977.     150  Weber, Max. “Objectivity in Social Science and Social Policy.” In The Methodology of the Social Sciences, E. A. Shils and H. A. Finch (ed. and trans.) New York: Free Press, 1904/1949.   Weitz, Eric D. A Century of Genocide: Utopias of Race and Nation. Princeton: Princeton University Press, 2003.   Zayana, Raphael. "The Inkotanyi’s New Strategy of Widespread Destabilization" in Kangura 8 (1992).                      


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items