Military history

Chapter Seven

The Civil War and the Genocide

IF RWANDA was the genocide that happened, then South Africa was the genocide that didn’t. The contrast was marked by two defining events in the first half of 1994: just as a tidal wave of genocidal violence engulfed Rwanda, South Africa held elections marking the transition to a postapartheid era. More than any other, these twin developments marked the end of innocence for the African intelligentsia. For if some seer had told us in the late 1980s that there would be a genocide in one of these two places, I wonder how many among us would have managed to identify correctly its location. Yet, this failure would also be testimony to the creative—and not just the destructive—side of politics.

To historicize the Rwandan genocide from this vantage point is to begin by identifying key differences between South Africa and Rwanda. From the standpoint of post-1994 Africa, I find one difference telling: if South Africa has millions of beneficiaries and few perpetrators, Rwanda has perpetrators at least in the hundreds of thousands and few beneficiaries. The difference highlights a salient political fact: that the genocide was carried out by subaltern masses, even if organized by state functionaries. I will reflect on this morally troublesome fact in this chapter, and return to think through its significance for postgenocide Rwanda in the conclusion.


The civil war profoundly changed all those who took part in it. The Rwanda Patriotic Front (RPF) went into it as an army of liberation and came out of it as an army of occupation. The Habyarimana regime entered the war pledged to a policy of ethnic reconciliation and came out of it pledged to uphold Hutu Power. From a marginal tendency in the constellation of forces supporting the regime in 1990, the war turned Hutu Power into a central tendency in Hutu politics. With defeat looming on the horizon, the Hutu Power tendency differentiated even further: the genocidal tendency was born of the crisis of Hutu Power.

RPF: From Liberation to Occupation and Displacement

The trained military cadre of the RPF numbered some 4,000 Banyarwanda who left the National Resistance Army (NRA) barracks in absolute secrecy in the dark of the night of 30 September 1990, a week before Uganda’s day of independence, 8 October. The initial RPF attack occurred on 1 October 1990. By any reckoning, it was a failure. From day one, disagreement over methods and tactics led to infighting among guerrilla leaders. First, the legendary hero of the NRA guerrilla struggle, Fred Rwigyema, and then Baingana, died in the space of but a few weeks. Once it recovered from the shock of the initial attack, the Rwandan army was able to repulse the invasion—with support from French, Belgian, and Zairean troops. While guerrilla ranks were in great disorder, the RPF attack allowed the Habyarimana regime to put on the mantle of the defender of the nation in the face of a Tutsi threat. Its legitimacy rose overnight in ordinary Hutu eyes. This fact, however, had yet to register with the combatants of the RPF.

By the end of November 1990, many RPF soldiers had been killed and thousands were scattered by the counterassault of the Rwandan army. About this time, Major Paul Kagame interrupted his military training course in the United States and took charge of the RPF. Pulling together some 2,000 men, he withdrew into the cold but heavily forested Virunga mountains in northwest Rwanda, along the Uganda border. By the middle of 1991, he had reportedly rebuilt the RPF to a 15,000-strong force. By the end of the year, the RPF had taken control of a strip of territory along the Uganda border stretching some 32 kilometers into Rwanda.1

It is often said that political movements are shaped more by adversity—such as Mao’s Long March—from which they draw their vital lessons, than by the dulling effect of success. The political education of the RPF, however, took place in the context of military victories, not losses. From the end of 1991, the RPF entered a period in which every military victory brought home the same bitter lesson about the political realities of Rwanda. The RPF consistently failed to translate military victory on the field into political gains within the population. The reason was simple. Every time the RPF captured a new area and established military control, the population fled. “Contrary to the expectations of the RPF,” wrote Gérard Prunier in an account otherwise highly sympathetic to the RPF, “local Hutu peasants showed no enthusiasm for being ‘liberated’ by them—they had run away from the area of guerrilla operations.”2 With every RPF advance, the numbers of the displaced multiplied. From an estimated 80,000 in late 1990 to 350,000 following the Byumba offensive in 1992, the numbers of the displaced swelled to roughly 950,000 after the February 1993 offensive, when the RPF doubled the size of territory under its control.3 At the peak of the war, when the rebels entered Gitarama in June 1994, the town emptied, as if on cue.

A number of journalists visited RPF-controlled areas in 1992 and 1993. All agree on one thing: that desolate calm prevailed in areas held by the RPF. Interestingly, all of them try to picture this lifeless calm with the same adjective: eerie. Writing towards the end of 1992 of the “immense suffering” unleashed by the war in government-controlled Rwanda, Kampala-based journalist Catherine Watson concluded with an observation on the guerrilla-held part of the country: “In contrast, the area under the RPF is eerily calm. One of the most densely populated regions of Africa in peacetime, it now holds a mere 2,600 civilians grouped by the RPF into two ‘safe’ villages.”4 Visiting after the February 1993 battles, Gérard Prunier also found these places “eerily empty of life.” “RPF soldiers had not looted anything and houses could be seen with chairs still set around a table and mouldy food on the plates where people had fled so hurriedly as not to eat their last meal. The RPF admitted that only 1,800 Hutu peasants were left in an area which had had a population of about 800,000 before the war.”5 Roughly at that time, the veteran Ugandan journalist Charles Onyango-Obbo used almost the same words to picture the overall ambiance in guerrilla-held areas: “In RPF-controlled areas of Rwanda, there is an eerie calm.” He then gave an additional reason for the exodus of Hutu peasants: “The rebels have asked all civilians to leave, because they don’t want the responsibility of caring for them and fear infiltrators. Privately, some officers say they hope that as the number of displaced people swells, pressure will grow on Habyarimana to reach a settlement in the war.”6

There is no contradiction between some journalist reports that peasants fled as the RPF approached, and other reports, like that of Onyango-Obbo, that the RPF asked peasants to leave. Both are true: peasants left RPF areas both because of their own volition and in response to administrative encouragement from above. Peasant attitudes shifted dramatically with the civil war, for the civil war seemed to have brought to life memories long since buried under the weight of day-to-day concerns. We can get an idea of this from an account by Catharine Newbury and David Newbury. Recalling their search in the late 1980s, they write of “the extraordinary degree to which the Revolution of a generation before seemed almost to have been removed from the collective historical consciousness.” They then follow with an account of conversations with refugees in camps in Tanzania in July and August of 1994: “But in 1994, time and again conversation in the refugee camps returned to focus intensely on the monarchical regime before Independence and the 1959 Revolution. The RPF was seen by many as the reincarnation of the pre-Revolutionary power structure.”7 Memories that would have seemed esoteric in the heyday of the second postrevolutionary republic, fitting material for intellectual reflection but no guide for day-to-day endeavors, came alive as the civil war progressed.

The RPF, too, changed. From recognizing that peasants distrusted them to a distrust of peasants, a sort of mutual distrust, was but a short step. For those in the RPF leadership convinced that peasants were anyway “backward” and “ignorant”—as reported by many a journalist who interviewed leaders in the RPF—this was an easy step to take.8 From the initial expectation of a relationship of political tutelage that was meant to translate shared interests into shared perspectives, RPF cadres had to come to grips with a relationship in which the role of coercion seemed to increase in direct proportion to military success. Looking back at the record, Human Rights Watch reported that the RPF had “forcibly moved hundreds, perhaps thousands, of people from Rwanda into Uganda in order to create free-fire zones.” It had also “pillaged and destroyed their property.” And finally, it had “recruited boys and men against their will to serve the RPF as porters and cattle herders.” “This abuse,” the report added, “began with the RPF invasion in October 1990. Although it continued on a lesser scale throughout the war, the abuse increased sharply again with the RPF offensive in February 1993.”9

If RPF’s expectations were based on what the NRA accomplished in the Luwero Triangle in 1981–85, the reality of the guerrilla struggle in Rwanda turned out to be dramatically different from that in Uganda. In sharp contrast to the NRA experience, there were in Rwanda no liberated zones where alternate modes of governance were introduced under the benevolent eye of a new administration. There were no Resistance Councils and Committees as in the Luwero Triangle, no effort to reach out to mobilize peasants politically, so as to transform them into a human resource for the struggle. There was not even an effort to establish administrative structures in the areas over which the RPF had military control.10 If anything, there was—unlike in most previous cases—a distrust of the peasantry, for the peasants were predominantly Hutu and they showed no enthusiasm at being “liberated.”11 “The RPF’s unconventional guerrilla strategy,” writes William Cyrus Reed in a euphemistic vein, “was accompanied by the large-scale flight of the peasantry, rather than their politicisation.”12 The object of this kind of liberation was no longer the population, but the territory. Thus, liberation turned out to be a combination of occupation and displacement: occupation of the land and displacement of the people.

Birth of “Hutu Power”

A fringe preoccupation among Rwandan Hutu by the late eighties, Hutu power became a mainstream ideology in the early nineties. The birth of “Hutu Power”—really, Hutu Powa, rather than Hutu Power, which would be an English-language slogan in a francophone milieu—as a formal organized tendency signified a sea change. Its context was the civil war. “Hutu Nation” had been the rallying cry of the 1959 Revolution and the foundation claim of the First Republic of Kayibanda. In contrast, the Second Republic promised “reconciliation” between Hutu and Tutsi. Branded an alien minority under the First Republic, the Tutsi were redefined, even rehabilitated, as a Rwandese minority under the Second Republic. The more the possibility of Tutsi power receded into a dim history, the less Hutu power had to offer as an organizing ideology, and the bleaker seemed its future.

The 1990 RPF invasion changed this context dramatically. For the first time since the inyenzi raids of the early 1960s, the 1990 invasion raised the specter of Tutsi Power inside Rwanda. This, unsurprisingly, is how the Rwandan government portrayed the invasion to the population inside and the world outside. In an address to the foreign diplomatic corps in Kigali, Foreign Minister Casimir Bizimungu accused the invaders of seeking “a reversal of history” which could only mean a return to “forced labour and feudal servitude.”13 And the fact was that many inside the country agreed that RPF rule would mean nothing but the return of Tutsi domination. The irony was that the more successful the RPF was on the battlefield, the more this view came to define the political center stage, bringing Hutu Power back from a fringe preoccupation to the mainstream of respectable politics. Hitherto, the demand that power must remain Hutu had been the rallying cry of those opposed to President Juvénal Habyarimana’s line of “ethnic reconciliation” between Hutu and Tutsi. Its last major public assertion had been associated with the failed Lizinde coup of almost a decade ago. Now, its proponents finally had an object worthy of public attention: to prevent the return of Tutsi Power because, surely, no worse calamity could befall Rwanda.

At the core of the ideology of Hutu Power was the conviction that the Tutsi were a race alien to Rwanda, and not an indigenous ethnic group. The shift in political vocabulary was a return to the vision of the colonial period. That the Tutsi were a race not indigenous to Rwanda was both central to colonial ideology and a key idea that had propelled forward the 1959 Revolution. The same notion had been part of the ideological baggage of the First Republic. This is where the Second Republic made a difference: Habyarimana spoke of the Tutsi as an ethnic group, not a race; as aRwandan, and not an alien, minority. The claim that the Hutu constituted a democratic majority because they were the ethnic majority would have made no sense from the point of view of Hutu Power. Because for Hutu Power, the Hutu were not just the majority, they were the nation. This is why the birth of Hutu Power as an organized political tendency went alongside a comprehensive propaganda effort discrediting Habyarimana’s effort at reconciliation. Hutu Power had to undo Habyarimana’s attempt to rehabilitate the Tutsi as an ethnic minority in Rwandan society.

Hutu Power propagandists claimed to be radical nationalist and populist. Yet, in defining the Tutsi as a foreign race, even if without knowing it, they were reaffirming the colonial legacy and construing themselves the same way that Belgian colonialism had construed them prior to independence. At the same time, the emergence of Hutu Power as a radical nationalist tendency in postcolonial Rwanda was evidence enough that the anticolonial struggle did not succeed in reconfiguring Hutu and Tutsi as political identities. The objective of their propaganda effort was toreracialize the Tutsi, as they had been in the colonial period, and under the first postrevolutionary republic of Kayibanda from 1961 to 1973. To recast the Tutsi as a race was to confirm that they were aliens in Rwanda. Two propaganda organs were central to this effort: the radio RTLM (Radio et Télévision Libres des Mille Collines) and the newspaper Kangura. Funded by members and friends who gathered around the person of the president’s wife and constituted a key power group referred to as the akazu (little house), RTLM began broadcasting from Kigali only four days after the signing of the Arusha Agreement. Shortly after the RPF invasion, Kangura published the widely circulated “Hutu Ten Commandments.” The commandments forbade Hutu from entering into a wide range of relations with Tutsi, whether in sex, business, or state affairs. “The Hutu should stop having mercy on the Tutsi,” went the eighth commandment.14

For Hutu Power propagandists, the Tutsi question was not one of rights, but of power. The growing appeal of Hutu Power propaganda among the Hutu masses was in direct proportion to the spreading conviction that the real aim of the RPF was not rights for all Rwandans, but power for the Tutsi. This is why one needs to recognize that it was not greed—not even hatred—but fear which was the reason why the multitude responded to the call of Hutu Power the closer the war came to home. Hutu Power extremists prevailed not because they promised farmers more land if they killed their Tutsi neighbors—which they did—but because they told farmers that the alternative would be to let RPF take their land and return it to the Tutsi who had been expropriated after 1959.15 Increasingly, the war shaped the context of daily lives. The war, said the government, was about keeping the threat of Tutsi Power at bay. “Defend your rights and rise up against those who want to oppress you,” the singer on Radio Libre des Mille Collines repeated as drums beat and guitars strummed a traditional melody. At the receiving end of this message were men and women like Kiruhara, an illiterate twenty-seven-year-old peasant who had spent most of his life cultivating sorghum and sweet potatoes on the steep mountain slopes of Kibunga Prefecture in eastern Rwanda. He had joined the Interahamwe when it was set up in 1992 as a youth militia of the ruling party. The stations “were always telling people that if the RPF, the rebel Rwandan Patriotic Front, comes, it will return Rwanda to feudalism, that it would bring oppression,” Kiruhara told his captors in 1994.16 Interviewed in the Benaco refugee camp in Tanzania, Bénédicte Ndagijimana, a college freshman majoring in English, explained the impact of such a truth on ordinary lives: “They hear over and over that the Tutsis are out to kill them, and that is reality. So they act not out of hate as fear. They think they have only the choice to kill or be killed.”17 We shall later see how others have sought to explain these developments: the economists arguing for a direct link between an increasing resource crunch and the genocide, and the culturalists claiming that the link is really with an ingrained culture of obedience or fear. Both, however, forget the central role that fear—not as a relatively timeless cultural reflex but as a much more time-bound response to a rapidly shifting political and social context—played in providing the ideologues of Hutu Power a mass following for the first time since 1963.

The more credibility Hutu Power ideologues gained among the Hutu multitudes, the more they were able to turn the Tutsi minority inside Rwanda into a hostage population. In the context of the war, there were actually two hostage populations, not one: not only the Tutsi in government-held areas but also the Hutu in RPF-held areas. The difference lies in how each was treated by its captors. While there is evidence of RPF slaughter of Hutu civilians, it suggests select killings that were more in the nature of reprisals or revenge. After a brief summary of these killings, Human Rights Watch concluded in May 1994 that “there is at present no credible evidence that the RPF has engaged in any widespread slaughter of civilian populations, although there are reports of less systematic abuses.”18 But, as we have seen, the RPF did resort to widespread displacement of Hutu civilians, to pillage, and even to conscription for coerced labor. Whereas the RPF resorted to displacement of Hutu peasants to pressure the regime into concessions and compromise, proponents of Hutu Power sought to achieve a similar objective through periodic massacres directed at ordinary Tutsi citizens.

So the massacres, which had ceased in 1964—once the question of power had been firmly settled—came back to life as the RPF invasion once again brought self-appointed custodians of the Hutu revolution face-to-face with the specter of Tutsi power. Each massacre was carefully timed and deliberately organized to follow a turning point, either in the civil war or in the negotiations that accompanied it. Four massacres occurred in the three and a half years that separated the RPF invasion and the onset of the genocide. The first took place in the weeks immediately following the October 1990 invasion, when an estimated three hundred Tutsi were massacred in cold blood in Kibilira. The second massacre started in Bugogwe and was a direct response to the January 1991 RPF raid on the town of Ruhengeri. This time at least a thousand Bugogwe cattle herders and their families were slaughtered. The third massacre at Bugesera in March 1992 was of a different type. It was less a retaliation than an offensive. The numbers killed were in the hundreds. But this time the killings had been prepared for in advance: the civilian Hutu population was urged and organized to defend itself against an expected massacre by the RPF and its civilian collaborators.19 The international commission of inquiry that visited Rwanda in January 1993 found evidence that these deaths were carried out by death squads directed by the security services in the office of the president.20 That same month, on 9 January, a key protocol relating to powersharing was signed in Arusha. The visit of the international commission notwithstanding, the fourth massacre followed: some three hundred Tutsi were killed in Gisenyi Prefecture the very next month. It was believed to be a response by Hutu Power to those who championed the call to share power. In all, an estimated 3,000 Tutsi were killed in massacres between 1990 and 1993.

It makes sense to see these pogroms not as projects of the government itself, but of an extremist tendency, Hutu Power, that linked some in the central government with others in local officialdom in different parts of the country. When the first massacre followed the October invasion, the government tried to check it. According to the U.S. State Department’s 1990 Human Rights Report, “The Government, with the help of the Catholic Church, sent troops and teams of clerics to trouble spots in Gisenyi, Gikongoro and Gitarama prefectures and insisted that local officials prevent further violence or face dismissal and punishment.”21 Instigated by a combination of central and local officials, these massacres further cemented an organized link between all those wedded to violence in the pursuit of a political agenda.

Much has been written about a centrally organized apparatus of genocide being in place as early as 1992. Professor Filip Reyntjens organized a press conference at the Belgian Senate in October 1992 giving evidence that a civil-military organization, code-named “Zero Network,” indeed functioned as a death squad. It had taken part in the Bugesera massacres of March 1992 and planned various political killings.22 We have seen that the International Commission of January 1993 also made reference to the existence of a death squad. But a death squad is not quite an apparatus of genocide. Death squads have existed in a variety of contexts, from Latin American rightist dictatorships to apartheid South Africa, but they have not perpetrated anything approaching a genocide. Death squads carry out individual assassinations, even group massacres, not the elimination of an entire people.

Later in this chapter, I will return to this question of advance preparation for the genocide. For the moment, however, we need to look at some key features of how the 1990–93 pogroms were organized. While initiated from the center, every massacre was executed locally. Just as it was the apparatus of the local state that had organized the flight of the Tutsi during the pogroms of 1963–64 and 1973, and then redistributed their property, so local authorities also organized the massacres that followed the RPF invasion of October 1990.23 Over time, a pattern could be discerned. Prefects and burgomasters organized Hutu militants who identified and targeted Tutsi “collaborators,” took over the land of those who were killed or fled, and redistributed it to militants. The use of local authorities was not simply a matter of using whatever instrument was available and at hand. There was, rather, a deliberate effort to use the “customary” as opposed to the “civic” apparatus of the state. In my view, the relevant distinction between the two was that while “customary” power highlighted the obligations of those indigenous to the land, civic power recognized the rights of all those resident on the land. We need to recall that customary power was employed through the colonial and postcolonial periods to enforce obligations on entire communities—such as coerced labor (umuganda) and compulsory villagization (paysannate)—in the name of observing custom. When it came to pressing ordinary people into a violent political campaign, it was not at all surprising that thegénocidaire tendency decided on “customary” power as the agency most suited to cleanse the community of threatening alien influences.

The use of the “customary” apparatus went alongside that of the language of “customary” obligation. Right from the first massacre at Kibilira that followed the October 1990 invasion, local officials were instructed to kill Tutsi as part of their communal work obligation. Killings came to be referred to as umuganda (communal work), chopping up men as “bush clearing,” and slaughtering women and children as “pulling out the roots of the bad weeds.”24 “In one commune,” writes Timothy Longman, “a massacre occurred when the burgomaster called the Hutu peasants to gather with machetes for umuganda, ostensibly to clear bush, then, with gendarmes present, sent them to kill their Tutsi neighbours.”25 After the slaughter of thousands of Tutsi at the church in Cyahinda in Nyakizu, the burgomaster told local people that burying the dead was required as umuganda.26The use of the language of “custom” was highly significant. After all, was not customary obligation supposed to distinguish the indigenous from the nonindigenous? With clearing the land of those branded alien considered a “customary” obligation, the genocide would ultimately be presented as a community project.

Two trends gathered over time. One accelerated the element of spontaneity, the other reinforced organization. The spread of massacres gave free reign to forces of banditry and pillage. As banditry and pillage grew, so did random killings. Yet, while there were reports of the poor attacking the well-off, the killings remained directed in the main at those identified as the political enemy, not the class enemy.27 As it developed, the genocide was undoubtedly the outcome of an interaction between dynamics that were both central and local. As in 1973, there were developments in several localities that tended to turn the focus of the conflict away from Hutu against Tutsi to poor against rich. At the same time, we know from the outcome that the predominant tendency in most localities was to target the Tutsi as the enemy.

The enemy was, first, the Tutsi, all of them, with the RPF considered the spearhead of Tutsi power, and then it was those Hutu branded as their accomplices. That the enemy was defined in the context of a war situation gave the massacres a degree of coherence they would otherwise have lacked. The more the war grew in scale, and the closer it got to Kigali, the more the country was put on a war footing. Everyone had to contribute, at first to the war effort, then to the war itself. The leaders of Hutu Power decreed that the war was everywhere, since the Tutsi—the enemy—were everywhere. “Everyone was asked to keep guard—to go to the barricades,” explained a Hutu resident of Kigali to a Christian Science Monitor reporter. “If you stayed at home, you risked being labeled an accomplice.”28 As in any war, but particularly in this one, there could be no neutrality. No wonder perpetrators who defend the genocide usually explain the massacres as inevitable excesses in a war situation.

If we are to understand the context of the mass killings that together constitute the hundred-days genocide, we need to move away from an assumption of the genocide as simply a conspiracy from above to an understanding of how perceptions could radically shift in response to an equally radical change in forces and circumstances—by making the genocide thinkable. I will try and make this point with reference to two key participants in the leadership of the genocide. The first is Léon Mugesera, a leading ideologue of the genocidal tendency in Hutu Power. The second is Stanislas Mbonampeka, a leading member of the Parti Libéral (PL) and an outspoken opponent of Léon Mugesera in 1992.

A Canadian-educated linguist, Mugesera was reportedly the first to air publicly the notion of eliminating the Tutsi physically as a final solution to the question of Tutsi Power. “We the people are obliged to take responsibility ourselves and wipe out this scum. No matter what you do, do not let them get away,” Mugesera invoked in a notorious 1992 speech in northwestern Rwanda, one that has been taken as the clarion call for the genocide that followed two years later. He went on to advise that the Tutsi be returned to Ethiopia, from where they had come anyway, but this time by “the river route,” specifically by way of the Nyabarongo River, which feeds into the Nile.29

By then, Rwanda had a coalition government. Mbonampeka, the minister of justice, came from the Parti Libéral, a critic of the government. He issued a warrant for the arrest of Mugesera, but was unable to carry it out because Mugesera fled to a military camp where he remained hidden until he escaped from the country. Fed up with what he understood as political vacillation on the part of President Habyarimana, Mbonampeka resigned as minister of justice. From then on, his position shifted radically, as radically as did the constellation of circumstances and forces in Rwanda. From one who was ready to arrest Mugesera in December 1992, Mbonampeka moved to a partisan anti-RPF position as the war advanced, and then to a defense of the genocide, finally emerging in the rump post-1994 government in exile as its minister of justice.30 Philip Gourevitch interviewed him in Goma, eastern Zaire, in June 1995. Mbonampeka explained the genocide as violence that inevitably accompanies war. “In a war, you can’t be neutral. If you are not for your country, are you not for its attackers?” He passed off the genocide as crimes of war committed in the course of civil defense. “This was not a conventional war. The enemies were everywhere. It wasn’t genocide. Personally, I don’t believe in the genocide. There were massacres within which there were crimes against humanity or crimes of war. But the Tutsis were not killed as Tutsis, only as sympathizers of the RPF. 90% of Tutsis were pro-RPF.” To drive the point home, he added: “Think about it. When the Germans attacked France, France defended itself against Germany. They understood that all Germans were the enemy. The Germans killed women and children, so you do, too.” In a space of but two years, Mbonampeka had moved from issuing a warrant of arrest for the prime ideologue of Hutu Power to appearing on a government of Kigali list of 414 “suspected commanders, organizers and authors of genocide.”31 It is this shift, and not the premeditated conspiracy of a Léon Mugesera which can provide a clue to the question: Why did hundreds of thousands, and perhaps more, of Hutu respond to the call of Hutu Power?


To explain the mass involvement in the genocide, writers have accentuated one of two factors: the economic and the cultural. Without necessarily denying the significance of either, I shall shift accent away from both to the political aspect of the genocide. Hegel once said that humans are distinguished from animals by the fact that they are willing to give life for a reason higher than life. He should have added that humans, unlike animals, are also willing to take life for a reason they consider higher than life. In both cases, there is a demarcation between life that is considered worth taking (or giving) and life considered worth preserving (or enriching). When the life in question is that of groups, involving large numbers, the decision is inevitably political. Though it may be taken under the pressure of necessity (economy) or the force of habit (culture), we need to highlight the decision as conscious, as the result of a deliberation. If not, we risk losing sight of any difference between humans and animals. My critique of those who tend to accent the economic or the cultural in the understanding of the genocide is that their explanation obscures the moment of decision, of choice, as if human action, even—or, shall I say, particularly—at its most dastardly or heroic, can be explained by necessity alone. Though we need to take into account circumstances that constrain or facilitate—that is, necessity—we must resist the temptation to present necessity as choice and thereby strip human action of both the dimension of possibility and that of responsibility.

A Resource Crunch

Two kinds of explanations highlight the world of necessity. For economists, necessity is a resource crunch. For culturalists, it is a closure of the mind. When it comes to writings on the genocide of 1994, both have figured, though the economic has tended to predominate. The economic standpoint highlights internally and externally generated constraints closing in on ordinary people in the decade before the genocide, like a growing sense of claustrophobia in a crowded commuter train. The resource crunch was said to be a result of rapidly increasing numbers of people having to look for a piece of cultivable land from a relatively stagnant pool. A postgenocide study on population growth and agricultural change commissioned by USAID—titled “More People, More Trouble”—pointed out that average farm holdings in Rwanda had shrunk by 12 percent from 1984 to 1989.32 Another evaluation, this one following the genocide, pointed out that 57 percent of rural households were already having to farm less than one hectare of land in 1984, while 25 percent of these had less than half a hectare. At the same time, these shrinking land parcels had to feed an average family of five people. Prevailing inheritance practices required that a family divide land among all sons. The result was not only diminishing but fragmented parcels: “Thus, in the beginning of the 1990s, the average Rwandese household farmed at least five plots of land.”33 These studies point at land conflict as the inevitable result of increasing human pressure on a fixed pool of land. The USAID-commissioned study summarizes existing research thus: “Land scarcity in Rwanda has resulted in intense rivalry and conflict among neighbors, with frequent fights and lawsuits over disputed land and frequent thefts from fields.” The conclusion: “Disputes over land are reported to have been a major motivation for Rwandans to denounce neighbors during the ethnic conflicts of 1994.”34

If land conflict was one outcome of land scarcity, diminishing food production was its other consequence. Food production declined after 1985: postgenocide research pointed out that kilocalories produced by Rwandan farmers had dropped from 2,055 per person per day to 1,509 over the period 1984–91. Even the “severe and moderate malnutrition (which) remained stable up to 1993” increased dramatically thereafter.35 The link between rising numbers, worsening poverty, and political extremism was made by Tipper Gore, the wife of the U.S. vice-president, Al Gore, at the Cairo World Population Conference in 1994: “Rwanda is a tragedy and a warning. It is a warning about the way in which extremists can manipulate the fears of a population threatened by its own numbers and by its massive poverty.”36

No matter how depressing these facts may seem, we need to keep in mind that there is no necessary connection between a drastic reduction in resources and deadly human conflict. One only needs to read the social history of natural disasters—be it drought, flood, or hurricanes—to recognize that countries have suffered a worse crunch than did Rwanda between 1989 and 1993 without the population turning in on itself, with one part devouring another. The connection between the constraints under which we live and the choices we do make is mediated through how we understand and explain these constraints and the resources we can muster to change them. As always, humans shape their world based on human consciousness and human capacities.

A Closure of the Mind

Few have dared to argue that the Holocaust was linked to a resource crunch in Germany. In contrast, there have been many explorations into German culture and psychology. In a well-known study, Shulamit Volkov deemed “anti-Semitism” a “cultural code,” created in the first decade of the German Reich as a convenient abbreviation for a broad “cluster of ideas, values and norms” that opposed “liberal, capitalist, democratic, and internationalist currents associated with the nineteenth century emancipation of Jews” in the name of “militant nationalism, imperial expansion, racism, anti-socialism, militarism, and support for a strong authoritarian government.” Arguing that the existence of an anti-Semitic tradition did not “require” the murder of Jews, Michael Marrus observed: “In the end it was Hitler, and his own determination to realize his anti-Semetic fantasies, that made the difference; in brief, ‘No Hitler, No Holocaust.’ ” This is precisely why, he argued, “for historians of the Holocaust, the greatest challenge has not been making sense of Hitler, but rather understanding why so many followed him down his murderous path.”37

If an understanding of what motivated functionaries to participate in the Holocaust has puzzled its historians, the riddle is perhaps even greater for those trying to make sense of mass participation in the Rwandan genocide. The final report of the Kigali conference observed: “The massive participation of the population in the Rwandan genocide is virtually without historical precedent.”38 What, indeed, explains the participation of vast masses, neighbors, coworkers, friends, even family members in the slaughter of those they knew well only yesterday? The response of those who write of the genocide divides into two: one focuses on society, the other on the state. Society-based explanations stress the historical legacy of racism in the Hutu population of the country, but without grounding racial perspectives in institutions, or distinguishing race from ethnicity as political identities, or even appreciating the historical dynamic making for the shift from one to the other. In an article that tries to understand the link between “demographic entrapment” and the genocide, Peter Uvin writes: “The most profound factor fueling the transmission of genocidal ideology from the regime to the masses, however, was the longstanding and deeply ingrained racism of Rwandan society.” He then continues: “For decades, Rwandan society had been profoundly racist. The image of the Tutsi as inherently evil and exploitative was, and still is, deeply rooted in the psyche of most Rwandans; this image was a founding pillar of the genocide to come. Although ethnic peace had prevailed during most of the regime, the racist nature of Rwandan society had not changed.”39

The second explanation shifts attention from society to the state as the lead actor and the active agent in the genocide. It focuses on the dead weight of cultural traditions that demand conformity to power. The proverb cited most often to sum up the cultural compulsion ordinary people felt to obey authority goes as follows: an order is as heavy as a stone. Gérard Prunier speaks of a “Rwandese political tradition” through the ages, before, during and after colonialism, as “one of systematic, centralised and unconditional obedience to authority.” Then he adds: “Most people were illiterate. Given their authoritarian tradition, they tended to believe what the authorities told them.”40 The idea that tradition and illiteracy make for a powerful mix is not uncommon: the more illiterate the population, the more it is said to be held within the grip of a mindless tradition. The implicit identification is of literacy with reason and tradition with unreason.41

In articles for the New Yorker, Philip Gourevitch let survivors of the genocide, both Tutsi and Hutu, speak for themselves. Laurent Nkongoli, a Tutsi survivor who became the vice-president of the National Assembly in the post-1994 RPF government, reflected on why the Tutsi were resigned to death. “There were four thousand Tutsis killed here at Kacyiru. The soldiers brought them here and told them to sit down because they were going to throw grenades. And they sat…. Rwandan culture is a culture of fear.” For François-Xavier Nkurunziza, a Hutu lawyer, the dilemma was how so many Hutu had allowed themselves to kill. “Conformity is deep, very developed here. In Rwandan culture, everyone obeys authority. People revere power, and there isn’t enough education. You take a poor, ignorant population, and give them arms, and say, ‘It’s yours. Kill.’ They’ll obey.”42Fear and obedience are like flip sides of a single coin: common to them is the claim that the person involved has ceased to think. When he got to talk to the leaders of the RPF, Gourevitch described their take on the genocide “as a crime committed by masterminds and slave bodies.”43

To believe that ordinary Rwandans killed, in their hundreds and thousands, and perhaps more, because of a congenital transhistorical condition—“a culture of fear” or of “deep conformity”—would require stretching one’s sense of credibility. For the period under discussion, the early years of the 1990s, was precisely when these very multitudes responded to a democratic opening with a growing defiance of authority, by uprooting coffee trees and refusing to perform compulsory communal labor. Yet, we need to remember that the culture of conformity is not an original construction of Gourevitch; it is, rather, Gourevitch the reporter quoting Rwandan respondents, both Hutu and Tutsi. Could it be that these Rwandan respondents were regurgitating as truisms “the mythical imagery of racism”: in the words of Peter Uvin, “the old myths of the Hutu as obedient and docile and the Tutsi as commandeering and cunning”?44

The notion of an unthinking participant, whether killer or victim, whether caught in the grip of fear or tradition, does not rule out the notion of a calculating individual, but one who calculates without reasoning, whose response to stimuli is predictable. From this point of view, calculation is short term, understandable under the circumstances because it is an adaptation to circumstances, but hardly reasonable since it also suggests a capitulation to the force of circumstance. This, then, is how the two kinds of explanations of the mass nature of genocidal violence are brought together without contradiction: the Rwandese peasant is portrayed as capable of both obedience to authority and calculated self-interest. This is more or less how Prunier sums up his exploration into the “mechanics of the genocide”: “unquestioning obedience to authority, fear of the Tutsi devils and the hope of grabbing something for oneself in the general confusion.”45

Without caricaturing this point of view or dismissing it outright, I want to point out its limits. The observation that ordinary participants in the genocide hoped to gain from it has been made by many a source, usually citing perpetrators of violence. Reporting in January 1994 on sporadic massacres that predated the systematic killing of Tutsi, Human Rights Watch noted that combat operations by the Rwanda army were often accompanied by civilian atrocities: “Civilian groups, composed of majority Hutu, committed widespread acts of ethnic violence against Tutsi. These rampaging crowds were incited and led by local administrators and by militia attached to Rwanda’s longtime ruling political party, the National Republican Movement for Democracy and Development (MRNDD). They destroyed crops, stole food, slaughtered cattle, burned homes and attacked their neighbours using machetes, spears and clubs.”46 Alison Des Forges of Human Rights Watch/Africa explained how those in authority could convince “those reluctant to kill” to join the killing spree anyway: “They also offered attractive incentives to people who are very poor, giving license to loot and promising them the land and businesses of the victims. In some cases, local officials even decided ahead of time the disposition of the most attractive items of movable property. Everyone knew who had a refrigerator, a plush sofa, a radio, and assailants were guaranteed their rewards before attacking.”47 Rakia Omaar, the director of Africa Rights, which provided detailed accounts by witnesses and survivors pointing accusing fingers at members of the educated Hutu elite, agreed: “The motive was often to secure a coveted job or property.”48 A Tutsi who survived because he had been away at a conference in Uganda, but who returned to find his wife and two children murdered, had this to say: “It was politics. Politicians told the people: kill, and you will get your neighbours’ goods and land.”49

After all, had there not been precedents to vindicate the expectation that a pogrom would enrich its participants? None would deny that the mass exodus triggered by the killings of Tutsi in 1959–63 had opened up vast tracts of land for landless Hutu. The link between political violence and social redistribution has been key to revolutionary politics everywhere. What distinguishes social redistribution from individual greed and theft is precisely the difference between two kinds of action: one social, the other individual; one political and extralegal, the other apolitical and illegal. Precisely this difference has been at the root of another distinction, one between two kinds of violence: revolutionary and criminal, the former an instrument of justice but the latter a source of injustice. The connection between political violence and social redistribution has been central to radical discourse and practice of justice in postcolonial African politics. It is the connection to justice that explains a range of reactions to revolutionary violence, from ambivalence to enthusiasm and romance. Since it involved large masses of people in common action, and not small groups or individuals in a conspiracy, the violence needs to be understood as political, and not criminal. At the same time, this violence did not pit the poor against the rich. If anything, it divided the poor—as it did the rich—into antagonistic groups: Hutu against Tutsi. It had little to do with either revolutionary violence or class struggle.

Neither does the reverse argument, that the violence was some kind of a desperate bid for survival in a resource-impoverished environment, hold. The crudest formulation I have seen of this argument was in a submission to the United Nations tribunal in Rwanda. It argued that those who killed “were engaged in a desperate scramble for survival at each other’s expense, of a kind that is all too familiar in the economic wastelands of Africa.”50 But the genocide was not a collapse of power and authority, a free-for-all in which everyone turned against their neighbor, with all thrown into some sort of a Hobbesian state of nature where life had turned “nasty, brutish and short.” The target of the genocide was clearly defined: not anyone, only the Tutsi. The truly disturbing aspect of the genocide is that the definition of the enemy appeared credible to many ordinary Hutu. To explain why this was so, we need to understand the violence of the genocide as a political violence born of the civil war that was a struggle for power within an elite once again fractured between Hutu and Tutsi. In many ways, the civil war of 1990–94 was a repeat of 1959–63, with one difference: this time, it is the Hutu political elite that was internally fractured, while its Tutsi counterpart showed greater political and ideological cohesion. To understand the circumstances that shaped the political violence and its genocidal magnitude, we need to join our historical discussion of political identities in Rwanda to an understanding of how the war reshaped these political identities. For the 1990–94 civil war changed not only those who directly participated in it but also those who suffered its consequences. It changed not only the political elite but also ordinary working people, both Hutu and Tutsi.


The genocide was born of civil war but it also marked a rupture in the civil war. Its perpetrators understood the multiple massacres that ultimately added up to genocide as a continuation of the civil war. Without keeping this fact in mind, it will be difficult either to understand the dynamic that propelled Rwanda into its darkest hour ever, or to explore ways out of it.

The war is crucial for several reasons. Its consequences for the civilian population were drastic. Coming at the end of a decade of economic decline, the war disrupted agricultural production, the infrastructure of communication, and thus the distribution and availability of food. It created widespread hunger and starvation. In addition, the war also displaced a substantial minority, so large that it came to include one out of every seven Rwandans. To this group were added Hutu refugees from Burundi. Living in camps scattered around the country, the internally displaced and the refugees were like so many bundles of dry tinderwood, awaiting but a spark to light a conflagration.

The civil war not only generated the raw material for the coming conflagration, it also provided the spark that would light it. To begin with, the government lost the civil war. The war discredited the army, fragmented the political class, and divided it into two hostile sections, each blaming the other for losing the war. Every major political party divided into two: those supporting a power-sharing deal with the RPF, and those opposed to it. Major opposition parties—the Mouvement Démocratique Républicain (MDR), the Parti-Libéral (PL), and the Parti Social-Démocrate (PSD)—were not immune from this trend. Each split into two factions, one called “power” and the other called “moderate”; one identified with the defense of Hutu Power, the other with power sharing. Each held the other responsible for defeat. Defeat discredited the army. Its leadership, in turn, held the political opposition responsible for dressing up defeat as power sharing and disguising national betrayal as democratic opposition. But the army too split in the face of defeat and disgrace at the battlefront. The erosion of the army’s esprit de corps, and the cohesion of its leadership, set the context in which the president’s plane was brought down. The hour had struck for the most ardent champions of Hutu Power—for those whose patriotic zeal knew no limits—to call the nation to arms against those they considered to have betrayed it. The enemy within were the Tutsi and their objective accomplices, the Hutu political opposition; in a word, the inyenzi and their abeyitso. We shall see that the killing of the abeyitso, the Hutu opposition that was branded as having betrayed the nation, was the curtain-raiser to the genocide.

The Displaced and the Refugees

For the civilian population, the war translated, first and foremost, into day-to-day hardships. It became that much more difficult to find bare means of daily survival. When the RPF attacked the most fertile part of the country in January and February 1993, the supply of agricultural produce reaching markets dropped by a drastic 15 percent.51 The second consequence for the civilian population was displacement. Every time the RPF scored a military victory and gained territory, another group of Hutu peasants and civilians was flushed out of the “liberated” areas. They flooded refugee camps that in turn mushroomed like so many sore and poisonous outgrowths around the capital city. The growth in their numbers was dramatic: from 80,000 in late 1990 to 350,000 in May 1992 to 950,000 after the February 1993 offensive. In its 1993 Human Rights Report, the U.S. State Department estimated that 650,000 of them were displaced for the first time but 350,000 were “re-displaced, some for the fourth time.”52 Other sources gave higher estimates of the numbers displaced, at 1.1 million, reaching beyond 15 percent of the total population of the country.53 The United Nations Information Center reported in August 1993 that most of the displaced “were living in and around 30 camps where serious malnutrition and disease were prevalent.”54

The liberalization of political activity drew two volatile constituencies into the political arena: the unemployed youth and those displaced by the civil war. Starting with the ruling party, one by one, political parties began reaching out to this constituency, incorporating it into its youth wing. The ruling party created its own youth wing in early 1992. It was called the Interahamwe, variously translated as “those who work together” or “those who attack together.” Once it had formed its own political party, the Coalition pour la Défense de la République (CDR), the extreme Hutu Power tendency followed by organizing their own youth wing. This was the Impuzamugambi (“those who have the same goal”). Its leader, Robert Kajuka, was a Tutsi.55 In the context of the civil war, unarmed youth wings rapidly transformed into so many armed militia. Africa Rights notes that the arming of the militia intensified following the February 1993 RPF offensive.56 The U.S. State Department noted in its 1993 Human Rights Report that armed militias were involved in the massacre that followed the RPF offensive.57

To the growing pool of internally displaced and unemployed youth was added a third volatile influence. This was the stream of Hutu refugees from Burundi, also victims of political violence. The flow of refugees from Burundi into Rwanda, as from Rwanda into Burundi, had a long history, in the case of Rwanda stretching to the 1959 Revolution, in the case of Burundi beginning with the massacres of 1972. Refugees from Burundi were mainly Hutu fleeing army-perpetrated massacres after each major political crisis. Of these, three were key: 1972, 1988, and 1993. As each wave of refugees subsided, and as the political situation at home returned to normal and realities of a normal refugee existence began to sink in, many returned home. But each time a minority would remain, holding on to bitter memories of home while nurturing sharp night-and-day type of distinctions between good and evil. Nineteen thousand such refugees from 1972 remained in Rwanda in 1990. They were joined by between 47,000 (U.S. Department of State) and 55,000 (U.N.) others in mid-1988. But unlike the 1972 group, they were refused refugee status by the Rwanda government, which granted them temporary asylum, insisting that they return to Burundi by the end of the year. By 1990, approximately one thousand had refused to return and remained within Rwanda.58

The last wave of Barundi refugees entering Rwanda was, however, also the biggest ever. It was a consequence of the violence that spread in the wake of the assassination of President Ndadaye of Burundi in October 1993. When conditions in Rwanda began to turn nasty in April 1994, about 200,000 people were in refugee camps in southern Rwanda. As the slaughter grew in magnitude, the violence taking on the force of a tornado, many fled to Tanzania or even returned home. Human Rights Watch estimated in May 1994 that “as many as 80,000 may still be left in Rwanda.” Other estimates were much higher, even as high as 400,000.59 Soon after their arrival into Rwanda, many were recruited into the Interahamwe. The UNHCR complained, but with little effect.60

Many an account speaks of the catalytic role of the Barundi in starting massacres in south-central Rwanda, a part of the country considered a stronghold of the opposition and difficult terrain for Hutu Power ideologues. Here, the long history of Hutu/Tutsi relations predated by centuries the bipolar racial mold in which colonialism came to cast it. Hutu/Tutsi intermarriage had not only predated the colonial period, it had also flourished in the postcolonial period. This kind of history tended to give rise to a plurality of notions about how Hutu and Tutsi may live together, in turn generating ambivalence toward ideas associated with proponents of Hutu Power. While this ambivalence tended to dissolve in the crucible of civil war, it never really vanished completely. In contrast, the historical memory of the refugees was sharp and simple. For them, the Tutsi were responsible for their continuing misfortune. Even those who believed in coexistence with the Tutsi would not tolerate any thought of sharing power. Not only did many of the Barundi refugees take an active part in the political violence of 1994, those who did were also responsible for some of the most gruesome tortures that marked the genocide. When I talked to survivors at the Church in Ntarama in 1995, I was told of the “Barundi torture” as the most cruel: beginning with the heel, a part of the body would be cut daily, a process that usually took a week as the victim bled to death. It is as if they were settling old scores, even if across the border.

A Defeated Army

The Rwandan army grew in size dramatically, from 5,000 to more than 30,000 in the course of a few years of civil war. New recruits were poorly trained and badly disciplined. At the same time, small arms were easily and cheaply available in local open-air markets.61 Following the model of many radical regimes in the region, the Habyarimana government in 1991 began a program to arm civilians and create “self-defense” forces. The program began as a pilot project, confined to the four border communes of Muvumba, Ngarama, Muhura, and Bwisige. Its aim was to provide a gun for every administrative unit of ten households, and to train the civilians who would be expected to handle the gun as part of “self-defense.” After February 1993, when the RPF doubled the size of territories under its control, the program was extended from border communes to interior communes. It is these “self-defense units,” found in every commune and village, that formed the civilian core of the machinery that came to carry out the genocide. They killed in response to orders from above because most believed in the moral rightness of obeying one’s government, particularly in a war situation, that is, when confronting an “enemy.” While an integral part of the machinery that carried out the genocide, it is most unlikely that they were created in the first place as machinery to execute a genocide. Rather, like much of the administrative innovation under the Second Republic, they were initially borrowed from the experience of regimes in the region, and only later adapted in response to changing circumstances.

The government responded to battlefield losses with a strategy to expand the army and to train and arm the civilian population following the ten-cell strategy made popular in Tanzania. The state initiative to create armed civil units for self-defense was different from the initiative of the ruling party and its allies in CDR. We have seen how political parties created youth militias, which began to take on paramilitary functions as the civil war expanded. An easily available anchor for frustrated and unemployed youth, the militia began to proliferate throughout the country. By early 1994, some 30,000 to 50,000 youth were estimated to belong to militias.62 From becoming active participants in the political process that expanded as the democratic opening broadened, they soon turned into perpetrators of the violence that began to consume that same political process.

Crucial to this souring of reform, to its turning inside-out from a democratic opening to open political violence, was the specter of defeat. Here, it may be useful to note a comparison between Rwanda and Somalia. Neither Rwanda nor Somalia had a military in the colonial period.63 In both countries, the army was a child of revolution: in Rwanda, of the 1959 Revolution, and in Somalia, of the 1969 revolutionary coup. No wonder the army came to consider itself not only the child but also the privileged custodian of thenational revolution. Military defeat not only demoralized the army; it disgraced the army and fragmented it. After defeat in the Ogaden, the Somali army ceased to have a national project. From a project of pan-Somali nationalism, the fragments of the Somali political leadership—including Siad Barre—embraced the fragment of the nation, the clan, as the vehicle for their political ambitions, in the process turning in on the nation. Not surprisingly, the leadership for each clan militia that contended for supremacy a decade later came, not from outside the army, but from one or another commander of the army defeated in the Ogaden.64

In Rwanda, too, as defeat disgraced it, the army exploded, as if into so many fragments of a cluster bomb. Rather than simply deflate the esprit de corps and the sense of mission of the army, defeat seemed to energize its parts, such as the Presidential Guard, and its attachments, such as the Interahamwe and the Impuzamugambi. From confronting the enemy that seemed to advance relentlessly on the battlefield or on the diplomatic frontier, they turned around to face the enemy within. Rather than forsaking the nation for an object worthy of its diminished capacity, as did the army in Somalia, the Rwandan army and its paramilitary attachments went on to purify the nation and rid it of all impurities that detracted from its strength. Even then, we shall see that this was not the project of the entire army, but of a fragment. This fragment was the génocidaire tendency. Born of defeat in the civil war, this fragment re-created a sense of national unity—unity of the Hutu Nation—and lived its moment of national glory through a shadowy “struggle” in which it locked defenseless civilians into a deadly embrace. For the perpetrators of the genocide, the enemy within were the Tutsi and the Hutu political opposition, accomplices who had betrayed the national cause in the name of democracy. It is in this sense that the genocide was both a continuation of the civil war and marked a rupture with it.

Political Democracy as National Betrayal

The great paradox of Rwanda of the 1990s is that democratic reforms blossomed at the same time as the civil war raged. The former fed aspirations for individual and group freedom, the latter gave rise to demands for loyalty to the nation. The two processes could not continue side by side, except through generating great tension. As war intensified and defeat loomed on the horizon, more and more of those in power, and even those in the population, came to see dissent not only as a luxury but, at a time of national crisis, as betrayal. Defeat in civil war spelled an end to both the democratic opening and to the democratic movement and its torchbearers. After the fires of war had consumed democracy, its burning ashes extinguished life itself.

Though political party activity predated the beginning of the civil war, political parties were only legalized in July 1991. In the opening phase, opposition parties demanded a recasting of political life along plural lines: some wanted a national conference, as in Benin or Congo, others wanted elections. To dramatize the extent of public support behind the demand for continuing reform, the three major opposition parties formed a coalition and organized a series of public demonstrations. The first was held in November 1991, to highlight the demand for a national conference when Habyarimana named a man from his own party as prime minister. It drew 13,000 persons. The second followed a month later, when the new prime minister named a cabinet in which all positions but one were held by members of the ruling party. This time, the attendance was estimated at 50,000. The government agreed to a Church-mediated negotiation. A few months later, in March 1992, President Habyarimana appointed a prime minister from the largest opposition party, the MDR, being the successor to Kayibanda’s PARMEHUTU.65

It is unlikely that the introduction of multiparty reform was wholly or even mainly a government response to opposition activity on the streets. Though the level of opposition organization had reached noticeable proportions, oppositional activity was still mainly an urban affair in a predominantly rural country. It is far more likely that official reform was a response to a combination of factors, internal and external. At a time when the Rwandan economy was fragile and deeply indebted, and things were getting even worse, the most significant of these factors must have been direct pressure from key creditors, including France. It is this which provides a clue to the great paradox of 1990: that a government drawn into civil war would respond, not as most governments in similar situations would, by curtailing civil liberties and declaring a national emergency, but by ushering in the dawn of multiparty democracy.

“Multiparty competition at a time of civil war put Rwanda in a state of permanent tension,” the head of the Parti Libéral (PL), one of the three main opposition parties, told me in Kigali in 1995. Its immediate effect was to strain unity in the ruling party. Those who believed that the nation deserved loyalty and not dissent in the hour of its need began to organize publicly and separately. To organize against reform, they left the ruling Mouvement Révolutionnaire National pour le Développement (MRND) to form a new party on the right, the Coalition pour le Défense de la République (CDR). From an ideological tendency inside the ruling party, Hutu Power was rapidly beginning to create its own institutions—at first a radio/television (Radio-Télévision Libres des Mille Collines) and a newspaper (Kangura), then a political party (CDR) and its youth militia (Impuzamugambi). At the same time, it battled to take over existing institutions of power: the media of the ruling party and its youth militia (Interahamwe).

Ironically, the first phase of the democratic opposition gave Hutu Power the legal space to organize its own institutions and to develop the capacity for independent political initiative. Not surprisingly, the first phase also came to an end once Hutu Power got organized as an independent force. Henceforth, every time the reform tendency registered its presence publicly—such as at the time of the March 1992 negotiations for a multiparty government or the December 1992/January 1993 negotiations between the multiparty government and the RPF—the institutions of Hutu Power responded with organized massacres of Tutsi. From the time a genuine coalition government was organized in March 1992 and Hutu Power responded by organizing its own institutions, there began a struggle between these two tendencies for political leadership. In July 1992, a former minister of information from the president’s home region and one of his closest associates, Christophe Mfizi, resigned from the MRND. In a public letter explaining his resignation, he accused the government of coming “under the control of a narrow group of extremists who [now] dominated all aspects of Rwanda’s public life for their personal gain and were fighting to protect their hold on power.”66 As the civil war progressed, the middle ground—that defined by Habyarimana’s project of “ethnic reconciliation”—eroded. As Hutu and Tutsi once again polarized as political identities, the battle for popular support was lost by the democratic opposition and won by the proponents of Hutu Power.

The Arusha Talks

A month after the inauguration of the first real coalition government in Kigali, talks were held between three opposition parties (MDR, PSD, and PL) and the Rwanda Patriotic Front (RPF) in Paris. The May–June 1992 discussions led to an agreement to start peace negotiations at a formal and governmental level. The Arusha talks began on 10 August 1992.67

The first thing to note about the Arusha negotiations was the process itself. The government delegation comprised at least three different tendencies, each responding to a separate center of power: the opposition parties in the power-sharing arrangement in Kigali took their orders from the prime minister, a member of the opposition MDR; others “were more clearly Habyarimana’s men,” with “still other delegates representing Hutu extremist groups in Kigali.”68 The most notorious of the last group was Colonel Théoneste Bagosora, a senior member of the extremist CDR who would later emerge as the key coordinator of the genocide.

It was not difficult for the Hutu Power tendency to portray the Arusha negotiations as talks between the internal opposition and the RPF, that is, between the RPF and its internal Hutu accomplices. Based on personal interviews in Rwanda in 1992–93, Timothy Longman observed: “Portraying the opposition parties as sympathetic to the RPF effectively served to discredit them with a large portion of the population.”69 For this reason, the Arusha talks were doomed from the outset. They unfolded in two phases: the first was led by the minister of foreign affairs and cooperation, Boniface Ngulinzaire of the main opposition party, MDR; and the second by the minister of national defense, James Gasana of Habyarimana’s party, the MRNDD.70 Shortly after the signing of the accord in 1993, James Gasana had to run for his life to Switzerland, but the unfortunate Boniface Ngulinzaire stayed behind and was among the first to be slaughtered in April 1994.71

Hutu Power claims were vindicated by an outcome suggesting that the RPF had won at the conference table what it had yet to win on the battlefield. Three parts to the final agreement fed this conclusion.72 The first was the provision on merging the two armies: these stipulated that the RPF would provide 40 percent of the soldiers in the new national army, but 50 percent of the officer corps. When the army had ballooned into six times its original size in four years of civil war, and when the country was plagued with massive unemployment, this was literally like serving an unemployment notice to young recruits in the army. The second was that the RPF was given charge of the important Ministry of the Interior. Together, these provisions gave the RPF decisive control over forces of coercion in the new state, especially at the leadership level. Third, while the RPF was to have eleven of seventy seats in parliament and five of twenty-one ministries, the power-sharing agreement excluded the organized Hutu Power tendency, CDR, from taking any seats in Parliament. A fourth relevant part of the peace agreement was the provision that recognized the right of return of all refugees. This was the fodder that Hutu Power media used to convince the population that the opposition had in fact sold the nation, and that all the gains of the 1959 Revolution—particularly land to the tiller and power to the Hutu—were now in imminent danger.73 The irony is that while Arusha was central to the opposition claim that it held the key to end the civil war and usher in the dawn of national reconciliation and peace, Arusha in reality confirmed the Hutu Power claim that the opposition had betrayed the nation. In doing so, Arusha sealed the political fate of the opposition.

Arusha was the reason why the first coalition government fell, and the reason why the second coalition government never really got established. The first coalition government fell in July 1993 when Prime Minister Dismas Nsengiyarmye strongly and publicly criticized President Habyarimana for his resistance to a negotiated settlement. In response, the president dismissed him and appointed a new prime minister, also a member of the MDR, but one considered less sympathetic to the RPF and closer to the political center. It is when the main opposition parties refused to support Agathe Uwilingiyimana’s appointment as prime minister that each split into two factions over the next two months, one called “power,” the other “moderate.” One called for a defense of Hutu Power, the other for a negotiated settlement with the RPF.

The Arusha Agreement was signed stillborn, mainly because it failed to take account of the extremist CDR, either by including it or by containing it. Instead, the peace agreement wholly excluded the CDR, even from the transitional government. Strong in both the government and the army, the extremists faced a double loss: of the government to the opposition and of the army to the RPF. Not surprisingly, when the opportunity presented itself, the extremists struck out viciously—at both. Nor was this outcome quite unforeseen by others at the Arusha talks. According to close observers, “All of the major third parties involved in the Arusha process”—the Tanzanians, the Americans, and even its principal regional support, the Ugandans—warned the RPF against “winning pyrrhic victories at the negotiating table,” and particularly against excluding the CDR from power and splitting the army disproportionately in its favor. The Tanzanians and the Americans were even said to share a metaphor: “If the hardliners weren’t brought into the tent, they would burn the tent down.” Some put it in a more colorful version: that it was “better to have the hardliners inside the tent, pissing out, than outside of the tent, pissing in.”74 One recalls the more sober advice offered by a wise old Tutsi man to a young RPF fighter who had come to “liberate” him in Ruhengeri in January 1991: “You want power? You will get it. But here we will all die. Is it worth it to you?”75

The Arusha Agreement was signed on 3 August 1993. Because the Uwilingiyimana government was unable to form a coalition government, the peace accord could never be implemented. While the “moderate” factions of the opposition refused to support a prime minister and a government not wholly behind the power-sharing agreement, the “power” factions of the same parties—now rapidly getting absorbed in a fast expanding movement to defend and assert Hutu Power—went on the offensive. The newspaperKangura, set up by Hassan Ngeze in 1990, had already broadcast genocide as a political solution in its January 1994 issue. Three months before the mass killings began, Kangura reported: “We will begin by getting rid of the enemies inside the country. The Tutsi ‘cockroaches’ should know what will happen, they will disappear.”76 Three months later, when precisely that agenda unfolded, Radio Mille Collines began calling for the assassination of those who had betrayed the nation. Its chilling broadcasts invited listeners to join in the killing, as if this were an appeal to patriotism: “The grave is only half-full. Who will help us fill it?”77 It would then caution its listeners to do a thorough cleanup: this time—as opposed to the last time, 1961–63—even the children should not be spared.

The war provided the context in which the Interahamwe transformed, from a youth organization at its founding in 1990 to a vigilante group in 1991–92, and ultimately into a death squad whose members led the house-to-house search for identifying and killing Tutsi in 1994. Robert Kajuga, a founding member and by 1994 the national president of Interahamwe, was born of a multiethnic family. His father was a Hutu, an Anglican priest, and his mother a Tutsi. Interviewed by an English journalist at the height of the genocide, Kajuga’s defense of violence was simple: “It’s a war against the Tutsis because they want to take power, and we Hutus are more numerous. Most Tutsis support the RPF, so they fight and they kill. We have to defend our country. The government authorises us. We go in behind the army. We watch them and learn.”78 The coalescence of civil war and a democratic opening put Rwanda “in a state of permanent tension.” The surest sign of it was the fast-eroding middle ground.

Having subjected Rwanda to a series of conditions throughout the course of the four-year civil war, the West sensed that the end of the regime was fast approaching and prepared to leave. The arrival of a 1,000-strong UN peacekeeping force in November was followed by the departure of French forces the next month. It signaled the complete isolation of the government in Kigali. That same month, the UN contingent escorted an RPF battalion into UN premises in Kigali. Meanwhile, Prime Minister Uwilingiyimana was still failing to put together the transitional government: it was less a sign of lack of will, then of mounting tension on the verge of defeat in the civil war.

The more the paralysis of government continued, the more tension increased, incredibly and palpably. Killing, burning, and looting spread through Kigali in the last weeks of February 1994. Many were killed right in front of UN troops, who just stood by and let it happen. The slaughter of the prime minister and ten Belgian soldiers sent to protect her, right inside the UN compound,79 presented the UN with a clear choice: either increase the size of the United Nations Assistance Mission to Rwanda (UNAMIR) force and change its mandate, or pull out. The UN chose to all but pull out. The Security Council met on 21 April. While the secretary-general requested more than a doubling of the size of the contingent, from the original 2,500 to 5,500, the major powers hesitated: led by the United States, the Security Council decided to leave behind a derisory force of only 270 soldiers. The message to the government was clear: implement the Arusha Agreement or else the UN will pull out and the RPF take power. By putting in place the final squeeze, the UN had succeeded in fully polarizing the situation.80

Could the UN have done otherwise and thus prevented the genocide? Opinions varied. Mr. J. Brian Atwood of USAID, in a press comment after the genocide, maintained “it would have been virtually impossible” to do anything under the circumstances, especially given that planned attacks were augmented by “irrational forces at work.” General Roméo Dallaire, the Canadian commander in charge of UN forces in Rwanda, had thought otherwise. On 11 January 1994, he sent a cable to the office of the secretary-general, concluding on the basis of the testimony of an informant in the Interahamwe: “Principle aim of the Interahamwe in the past was to protect Kigali from RPF. Since UNAMIR mandate he has been ordered to register all Tutsi in Kigali. He suspects it is for their extermination. Example he gave was that in 20 minutes his personnel could kill up to 1000 Tutsis.”81 A high-ranking UN official in New York also disagreed with USAID.82 We may never know the answer to the question: Could the UN have prevented the genocide? What we do know is that the UN took no action indicating that it took a serious view of the shedding of African blood. If anything, it gave the opposite message. When the slaughter began in April, a story familiar to African ears played itself out again. Belgian and French paratroopers swooped in to repatriate those who were least in danger: their own nationals and those from other rich countries who wanted to leave. They categorically refused to consider evacuating any Rwandans unless, of course, they were married to Europeans.83 When the French did return to Rwanda, it was toward the end of the genocide, as benefactors of a UN-sanctioned “humanitarian” intervention called Opération Turquoise. Ostensibly undertaken to save the remnants of the Tutsi population running for their lives, it also turned out to be a protective umbrella for those in the leadership of the genocide running for cover.

Let us return to consider the fate of the democratic opposition. If we are to go by the threefold increase in those joining opposition public demonstrations in a single month in late 1991, the democratic opposition seemed to be enjoying a groundswell of urban support at the time. But in the aftermath of the signing of the Arusha Agreement, precisely when it may have hoped to reap the harvest of popular support, the opposition was instead thrown to the margins. In my view, this outcome was the result of a confluence of three forces: a reckless internal opposition, an irresponsible donor community, and a naive RPF. The internal opposition was reckless because it acted without a sense of the balance of forces within the country. Had it taken the balance of forces into account, it would have moderated both the nature of its demands and the pace of its momentum. It did not because of a false confidence that it derived from the backing of donors whom it knew the government could not ignore because of dire financial need. The donor community force-fed Rwanda a reform agenda out of a textbook, without regard to the situation on the ground and secure in the knowledge that they would not have to suffer the consequences of their actions. Joined to immunity, power bred a reckless irresponsibility. The RPF had even less contact with the situation on the ground than did the internal opposition or the donors. The naivité of the RPF was fed by leaders who were mostly born outside the country and whose sense of possibilities was shaped by their experiences in Uganda, and not in Rwanda.

Prelude to Genocide

Neither political assassinations nor massacres make for genocide. It will also not help to equate Hutu Power as a whole with a genocidal tendency. The genocidal tendency arose out of a double crisis, of both the democratic opposition and of Hutu Power. Faced with a military defeat that seemed to sound the very death knell of Hutu Power, the génocidaires chose to embrace death itself as an alternative to life without power. Turning away from the enemy on the battlefield—the enemy it could not defeat—it looked for an enemy within. There were two turning points in defining the shift of focus from an armed target on the battlefield to unarmed and defenseless civilians within. The first was the assassination of the first Hutu president in neighboring Burundi, Melchior Ndadaye. Thesecond was the assassination of Rwanda’s own Hutu president, Juvénal Habyarimana, and the murder of the prime minister, Agathe Uwilingiyimana. If Ndadaye’s death was taken as a prophetic lesson that the only alternative for the Hutu was between power and servitude, that there could be no power sharing between Hutu and Tutsi, Habyarimana’s death was a signal that the hour to choose between power and servitude had indeed struck. Finally, the death of the president and the killing of the prime minister removed precisely those leaders who had publicly championed an agenda for an “ethnic reconciliation” between Hutu and Tutsi.

In spite of a history that presented a milder version of Tutsi power and Hutu servitude, Burundi had come to be seen by Rwandan Hutu and Tutsi alike as some sort of an accursed Siamese twin.84 For Hutu in post-1959 Rwanda, Burundi presented a real life portrayal of what it would be like for Hutu to continue to live under Tutsi power. The mass killing of Hutu schoolchildren and intellectuals in 1972—estimated at around two hundred thousand—gave the Great Lakes region the first sight of a genocidal wave of killing. That was until the political reform under President Pierre Buyoya, which brought universal franchise with one-person–one-vote in a system of multi-party competition. In the ensuing election of 1993, Melchior Ndadaye was elected president. For the first time in its postindependence history, Burundi had a Hutu as president and, even more incredibly, a Hutu-dominated government with a nearly all-Tutsi army. But only two months after the signing of the Arusha Agreement—on 21 October 1993—the promise of the Burundi reform turned into a nightmare when elements from the all-Tutsi army murdered the Hutu president. Political violence swept the country and some 200,000 panicstruck Hutu crossed the border into Rwanda. The core message of Hutu Power began to sound credible to ordinary Hutu ears in Rwanda: power sharing was just another name for political suicide. History had ruled out political coexistence between Hutu and Tutsi.85

As tension mounted, arms proliferated and training in the use of arms became open and prevalent. Every organized political tendency, and not just the advocates of Hutu Power, felt it necessary to have an armed militia attached to it. If the ruling party had the Interahamwe and the CDR its own Impuzamugambi (those with a single purpose), then the MDR had the Inkuba (thunder) and the PSD its Abakombozi (the liberators).86 It came to be accepted that only armed politics was credible politics. A pastoral letter issued in December 1993 by the Catholic bishop of Nyundo, the diocese where Habyarimana’s home was located, also the diocese where a number of massacres had been carried out between 1900 and 1993, criticized the distribution of arms to youth militia. But the level of political violence continued to escalate. In March 1994, the head of one of the three main opposition parties, the PSD, was assassinated. The next day, a mob of PSD supporters lynched the head of CDR.87

The last major signpost of “ethnic reconciliation” in Rwanda—even in the context of Hutu Power—was in the person of President Habyarimana himself, and of his prime minister, Agathe Uwilingiyimana.88 Representing the “power” wing of the largest opposition party, the MDR, Uwilingiyimana had been appointed prime minister a year before by Habyarimana in place of a moderate-opposition prime minister who appeared to be too closely identified to the RPF. Though defenders of Hutu Power, Habyarimana and Uwilingiyimana defined the middle ground that thegénocidaire tendency needed to clear if it was to assume power. With the shooting down of Habyarimana’s plane on 6 April and the brutal killing of his prime minister that same day, the agenda of “ethnic reconciliation” ceased to exist. The genocidal tendency was now in a position to take over the reigns of political power. The génocidaires were not synonymous with the army, but with a faction that cut across the army, the political, and the civil elite. They were different from other defenders of Hutu Power in the means they advocated: for the génocidaires were the faction who advocated genocide as the only effective—and remaining—way of defending power.

One needs to exercise the same caution when it comes to an understanding of what has been called “the institutional machinery of genocide”: the Presidential Guard in the army, the youth militia Interahamwe and the Impuzamugambi, the local-level ten-cell civil defense units, and the machinery of local government and “customary” rule. “By 1992, the institutional apparatus of genocide was already in place,” wrote René Lemarchand.89 Just as the génocidaires were a political tendency born of civil war—and not simply one that had marked time awaiting a suitable opportunity—so were none of the instruments used to perpetrate genocide not created for that purpose from the outset, but were turned to that purpose in the face of defeat in the civil war. The Presidential Guard had been created in 1992 to wage war, not genocide. The administrative ten-cell groups were created as units for civil defense. The youth wing of each political party was created to expand political participation in the course of multiparty competition. And the machinery of local administration, which was built in the colonial period and had preceded all others, was created to enable colonial government, to give it teeth, while masking this project as customary. Though they perpetrated the worst political violence in the history of Rwanda, none of these had been created for that purpose. To understand how their original purpose came to be subverted, one needs to understand—as I have argued—the consequences of civil war and defeat.

By portraying opponents as potential perpetrators and ourselves as potential victims, war tends to demonize opponents and sanctify aggression as protective and defensive. The Rwandan genocide was carried out by two different groups. The first were actual victims of the RPF war and of the massacres that followed Ndadaye’s assassination in Burundi, the former displaced from territories RPF captured in the northeast of Rwanda and the latter refugees from Burundi. The second were those who were convinced that they would surely, even if potentially, be victims if Tutsi Power won. They included two classes of beneficiaries of the 1959 Revolution, one prosperous but the other poor, the former members of the post-1959 Hutu middle class, and the latter ordinary peasants who would not have had access to land but for the land reform that followed 1959.

The president’s death and the prime minister’s murder presented the génocidaires with an opportunity for a coup d’état. It is said that the head of the Presidential Guard, Colonel Théoneste Bagosora, and his circle, tried to take power in their own right, but failed to secure the support of “a number of officers as well as ranking representatives of the UN in Rwanda.” Their next move was “to install a regime of extremists” other than themselves; the move was successful.90 Without their hands on the levers of power, they could have carried on with political massacres in different areas, but they could not have unleashed a countrywide terror whose object was to annihilate the entire community of Tutsi so as to remove any trace of a Tutsi presence from the soil of Rwanda. Only from a position of state power could they put the stamp of law and authority on a genocidal design. To consolidate that grip, they began with the physical elimination of all Hutu in public life who were identified with the middle ground. The elimination of moderate Hutu, estimated to number 50,000, cleared the ground of any significant group that could mount a challenge to the tidal wave of violence that was about to follow.


A Central Design

The génocidaires’ tendency had to do more than simply capture power to implement its design. To embark on an agenda calling for the total physical elimination of the Tutsi and of the “moderate” Hutu political and civil elite—those branded as “accomplices” because they were likely to stand in the way of such a gruesome “solution” to civil war—they needed a reliable machinery. To do that, they had to ensure the loyalty of both the political machinery of “customary” power that would direct the violence, and the coercive machinery, mainly the Presidential Guard and the youth militias, that would enforce it. The first step was to remove those with suspect loyalties from positions of power. The presidential plane was shot down on 6 April. Public appeals were made to a meeting of prefects on 11 April and on the radio the next day, to the effect that partisan interests must be set aside to fight the common enemy, the Tutsi. All this while, the Presidential Guard was busy identifying and eliminating members of the political opposition. On 16 and 17 April, they replaced “the military Chief of Staff and the prefects best known for opposing the killings.” This included “one who was murdered with his family” and another who “was later imprisoned and executed.” Also slain were “three burgomasters and a number of other officials who sought to stop the killings.”91

The pivotal role of a centrally-coordinated plan for the unfolding of ground-level violence—as well as its limits—comes out most clearly where popular participation was absent, or was too weak to provide the necessary deadly force. Legendary in accounts of the genocide is the case of Butare, whose local authority actually refused to carry out the orders to kill the internal “enemy”. This extraordinary courage came from its prefect, Jean-Baptiste Habyarimana, the only Tutsi prefect in the country. For two weeks, Butare remained a calm oasis while a storm raged in the rest of the country. At that point, the president of the interim government, himself a Butare man, came to rebuke the population for “sleeping” through the war, and appointed a new prefect whose enthusiasm would equal the demands of the situation. He came to town with soldiers of the Presidential Guard and with militia members from Kiaka Cooperative in Gisenyi in the far north of the country. As the town population gathered to witness the investiture of the new prefect, the former prefect and his family were murdered by the Presidential Guard. For those harboring moral ambivalence about the new direction of the “war,” the signal was clear: any wavering in the face of orders from above would mean certain death. Fear could silence opposition, but it could not generate enthusiasm. To sustain the killing in Butare, the new officials had to bring in armed groups from the outside and unemployed youth from the hills.92

The role of outside forces in triggering the violence appears to have been the greatest in the prefectures of Butare and Gitarama, both at the center of the old Rwanda kingdom. This is where there had been the highest incidence of intermarriage, especially among the elites, and where the colonial history of more or less unaccountable Tutsi power in the local authority was but an interlude in a centuries-old history.93 It is in this region that the presence of a substantial number of Hutu refugees who had fled violence in Burundi the previous year made the job easier: “Angry about being forced to flee their own homes by Tutsi in Burundi and concentrated in a camp where they had nothing to do with their time and energy, the people from Burundi (known as Barundi) offered the ideal recruits for launching an attack on the Tutsi of Nyakizu.”94

No matter how prominent and predominant the role of external forces in particular locations, their presence should not detract from the decisive role of locals in the genocide. To understand the variety of local responses, we need to make a number of distinctions. The first is regional. The genocide, we must remember, took place mainly in central and southern Rwanda, and hardly in the north. After the first RPF attack in 1990, Tutsi in the north were killed. Not only were the Tutsi in post-1990 Rwanda resident mainly in the center and south of the country, so were the parties that comprised the internal political opposition and that had allied with the RPF at Arusha. The genocide thus divided the southern Hutu more than it did the northern Hutu into two sharply opposed groups: those who joined the massacres and those who got massacred. The two extremes were joined by middling reactions that tried to combine moral extremes. The predicament of ordinary Hutu is clear from a single fact: it is not only the political opposition that got massacred in the days that followed the president’s assassination. As they grew in scope, the massacres targeted anyone, peasant or professional, who refused to join in the melée. Kodjo Ankrah of Church World Action recounted to me what happened when soldiers entered a church in Ruhengeri and asked that Hutu step on one side, and Tutsi on another: “People refused; when they said, Tutsis this side, all moved. When they said Hutus that side, all moved. Eventually, soldiers killed them all, 200 to 300 people in all.”95 Professionals who refused to join in the killing also met the same fate. Take, for example, the parents of François Nsansuwera, deputy attorney general under Habyarimana, later appointed to the same post under the RPF. Nsansuwera’s father was a retired army officer. When thirty Hutu and Tutsi gathered to seek shelter at their house, the militia called in the army; all thirty, including his parents and father-in-law, were killed. Of his family of nine, Nsansuwera said only two survived, himself and a younger brother who had gone through Burundi to join the RPF.96

The Hutu who were slaughtered in the opening days of the genocide—either because they were identified with the political opposition or because they simply refused to join in the killing—were the first of three groups among the southern Hutu. The second were Hutu who not only saved their Tutsi neighbors, but also got away with it. Some were indeed text-booklike heroes, such as the retired soldier-turned-policeman I met at the church in Ntarama. He hid eleven people in different locations: from ceilings up above to pits down below. When I asked him whether he knew of anyone else who had helped people by hiding them, he said “no.” When I asked why he thought there had not been others like him in the area, he simply said: “People don’t have the same mind.”

Later on, I heard of a different kind of response from Faustin, my guide during my 1995 visit. This ambivalent response illustrates the third type. He narrated an account he had heard from a survivor from Kibuye, on the eastern shores of Lake Kivu. As in several other places, there too the attack against the Tutsi began with the arrival of a group of Interahamwe from outside. This time though, the entire community—Tutsi, Hutu, and Twa—got together and fought them with stones. The Interahamwe retreated, tried again on a second day, and failed yet again. On the third day, the Interahamwe sent political cadres, one by one, to approach local Hutu and promised not only that their lives would be spared if they didn’t join Tutsi in the fight, but also that they would benefit from the distribution of Tutsi property. The next time the Interahamwe attacked, the Tutsi found themselves isolated. Some 20,000, they retreated into the ridges, up in the hills. There they fought for two months, with the fighting taking on a regular daily pattern: from 8:00 in the morning till 12:30, when the Interahamwe would descend the valley for lunch, and from 2:00 in the afternoon until 5:00 or 6:00. The hours of sleeping, shooting, and eating were known to all. So were the means of defense: the Interahamwe shot, whereas the Tutsi simply ducked bullets by running around the hill in concentric circles, at most throwing stones. By the time French paratroopers arrived and stopped the massacre, two months had passed and nearly 19,000 were dead and two hundred injured, with some eight hundred still alive. If in this instance the Hutu deserted their Tutsi neighbors, leaving them to face Interahamwe rifles alone, at least they did not join the Interahamwe in massacring the Tutsi.97

The most ambivalent stories of the genocide I heard from survivors were about the Hutu who saved a friend or a colleague in one place, only to go and join the killings in another. A lecturer at the University of Butare rescued the child of a Tutsi colleague, and then ran for shelter in the Zone Turquoise, only to find herself accused upon return of having participated in the genocide. Whereas she was released after four months in prison, her husband was not. When I spoke to her in July 1995, she told me her husband was not alone but languished in jail with ten other lecturers, all accused of participation in the genocide.98 I never did find out the truth of the allegations against them. But I did often wonder: How could the same person risk his or her own life to save another at one time in one place, and yet take life at another time in another place? As I heard more and similar stories, I posed the same question to those I met, and mulled over it myself. As so often happens in postgenocide Rwanda, I usually got two answers to the question, one Hutu and the other Tutsi, depending on the respondent. The Hutu answer was that these were stories made up by Tutsi survivors reluctant to recognize any Hutu savior, and the Tutsi answer was that the act of mercy lacked sincerity and was more an alibi for the future. Over the next year, as I regularly perused discussions on Rwandanet, an email listserve, I realized that these cases were not solitary. Many had combined saving in one place with killing in another. Could they have killed under duress—knowing that if they refused or even appeared reluctant, they would surely be killed—and saved a life when the opportunity presented itself? Was this not more representative of humanity in the ordinary? They were less than heroic under stress, yet humane in ordinary circumstances—perhaps one reason their experiences are not be celebrated in the open and without reservation.

Mectilde’s Story

The ambiguity of responses in the real world of the genocide comes together in the experience of a single Tutsi survivor, Mectilde.99 I met Mectilde during my second visit to postgenocide Rwanda, in December 1995. She was working with a UN agency and was engaged to be married. At the time of the genocide, Mectilde was a typist with an international company in Kigali. She lived in Kiyovu Joc, a youth hostel for Catholic women workers which housed sixty women. “On 6 April, I went to study computer at Ministry of Planning during the morning, returned to the office in the afternoon, and went home at five with a colleague. About 8:30 P.M., we heard bangs. My neighbor said it was the plane of President Habyarimana exploding. I closed my door. I had a one-room apartment and was alone. I turned my radio on at 9 P.M. Radio RTLM gave news of the president’s death.” The first group she turned to for support were neighbors. “I invited them to come to my room. We waited for the news, heard it together, and decided to search where to go that night.” The next morning, the Presidential Guard entered the hostel. “They killed four with guns. My other neighbors asked me not to stay there, but come with us to a priest. Our neighbors were mixed, Hutu and Tutsi. But at that time, no one was thinking of ethnie. We were thinking only of the death of the president.”

“On the second day, they began to kill Tutsi.” She then turned to a Canadian priest, spending two nights in his compound, along with three hundred others, until the priest suggested they leave. “He said we should go out because he had heard that the Interahamwe had killed all priests and refugees at another center. He said, pray, and you die, God will receive you.”

Her next refuge was with two girlfriends, “one mixed with a Hutu father and a Tutsi mother, and the second a Hutu.” They spent four days in the hostel, living on tea and rice. On the morning of the fifth day, the president of the Catholic Workers Youth, a Hutu, came to her room and broke the news that she was number three on the list of those to be killed that day. She and her girlfriends decided to go to the town of Gitarama. The problem was that whereas her girl friends had Hutu ID cards, she didn’t. The Youth Association president said he would help pass her through the roadblocks if she went with his younger brother, who was in the military. She agreed. But when the brother arrived, he refused. “He said no, I was very very Tutsi and he could not help; they will kill us both. I said OK, go. The president said no, you come. They said, take nothing, try and be as a peasant. I went with a trouser and slippers. I put out my glasses. I wrote my name on a paper saying I am a wife of a military staying in Gisenyi. They said I should wait in Gitarama for the other girls while they continued to Gisenyi.”

She took a taxi jammed with passengers, negotiating roadblocks, first at Kiyovu, then at Muhima. “When I arrived at Muhima, near the workshop, I met a woman with a gun. She asked me to leave the car and said she saw me once at CMD [the military camp of 600 RPF soldiers near Kigali]. I said it was not me. She insisted, said the car could go without me. I pleaded. She said no, but then allowed me to go.”

The next roadblock was manned by Interahamwe. “They said I must be a Tutsi because I had no card. They took me out. The military (in the taxi) said no, it is a mistake, for this is the wife of our colleague who is in Gisenyi and came without an ID. They took out their machete. Someone with a gun said he will do it. The military said you can’t kill someone with us; if you kill her, we will kill you. They said, OK, you are Ikotanyi. In our army, there are some, we know. But we know you can’t pass the roadblock of Nyabarongo, manned by the Presidential Guard. The driver said, now, we will pass the next roadblock and we won’t discuss with them, because if we do, they will kill us all. So just pray. When we arrived, the Presidential Guard asked where we were going. We said Gisenyi. They said you can pass, without any discussion. I had passed the last step of death.”

The taxi dropped her at Gitarama, where she had to pass the night before meeting up with her friends. “It was 4 P.M. when they left me on the road. I had 2,000 francs only in my pocket. I had a Fanta and returned to the road. I saw someone who was working near my office. He said, why are you here? What have you done to be here? I said I had no idea where to go. He invited me to his apartment. There were forty-two people, three families, inside the apartment. They were both Hutu and Tutsi.” She spent the night in the apartment and returned to the road the next day. “I waited until 2 P.M., when I saw a taxi with my two friends. Their families were in Gitarama in a neighboring commune. We were the first to arrive from Kigali since the war began. I went with my friend who had a Tutsi mother and a Hutu father. I thought I wouldn’t have a big problem since the mother was a Tutsi, better than going into a Hutu family. I arrived there. They took us as survivors. We received many visitors who asked news of Kigali. We spent a week without any problem.”

The following week, a neighbor reported that they had begun to kill Tutsi in the commune. The family also had a daughter married to a Tutsi. “They went to see if he [her husband] was alive. Now we were two who were a problem to the family, the son-in-law and me. They decided to dig a hole in the kraal. We would spend the whole night in that hole, with the cows there also. During the night, about one or two, they would put us out, and give us something to drink after a bath. We would go to bed for one or two hours and return to our hole.”

The hole-in-the-ground routine went on for three weeks. “Then the son-in-law said, it’s enough for me. I don’t want to cause trouble for this family. He told his father-in-law that he preferred to be killed as a man who respects himself. They said, where can you pass with your ID card? They took the son’s ID card, put the photo of the man, and tried to draw a stamp. Then they gave him the ID. He took it. He passed three communes and arrived in Butare without any trouble. He spent a month there. Once he went out and met his former student who was a military. He asked for his ID, and when he saw it he accused him of lying and killed him with a gun.”

Mectilde continued to stay in her hole, but not for too long. On 27 April, the Ikotanyi came and took them to the refugee camp at Musamo. There, they spent a month and then returned to Gitarama. Another month there, and the préfet said she could go home. “When I arrived in Kigali, I saw so many people. We passed the night at my girlfriend’s place. The second day, I went to town to see if my brother was still alive. He had gone to Bujumbura. Our house was empty. My brother said my father, mother, younger brother, all were killed. Only the three sisters are alive. He gave me 20,000 francs.” Mectilde returned to work. “Before I used to be a senior personnel officer, but I was asked to be a receptionist because it was very difficult to think, to make a report.”

I asked Mectilde if many Hutu had helped Tutsi. She said, “ten percent helped. A Hutu can help you in Kigali, but in Butare he can begin to kill Tutsi. Take the president of Catholic Workers Youth we went with to Gitarama. When he arrived in Butare, he killed people. I don’t know why.” I asked her, what about the remaining 90% Hutu? She gave me a rough count: “Ten percent helped; 30 percent were forced to kill; 20 percent killed reluctantly; 40 percent killed enthusiastically.” It is the 40 percent, those who “killed enthusiastically,” who represent the real moral and political dilemma of the Rwandan genocide. Mectilde is, of course, one person. My purpose in citing her estimates of how many were killed—whether enthusiastically, reluctantly, or under duress—is not to give these a stamp of approval. Later, we shall see that some of the most responsible leaders in the RPF also estimate the killers in the millions. I will later discuss the political uses to which such estimates can be put and which in turn can lead to inflated estimates. Here, I want to make two different points. One, from the point of view of the minority in postgenocide Rwanda, the majority is guilty, either of killing, or condoning, or just looking elsewhere while the killing happened. Second, even if we can never know the numbers of those who killed, there is no escaping the disturbing fact that many did enthusiastically join in the killing. The genocide was not simply a state project. Had the killing been the work of state functionaries and those bribed by them, it would have translated into no more than a string of massacres perpetrated by death squads. Without massacres by machete-wielding civilian mobs, in the hundreds and thousands, there would have been no genocide. We now turn to the social underbelly of the genocide: the participation of those who killed with a purpose, for whom the violence of the genocide and its target held meaning.

Women Killers and Child Accomplices

By September 1995, several hundred of the 10,000 inmates in Kigali’s sweltering central prison were women. Rakiya Omaar of African Rights told an Associated Press journalist that some “were actively involved, killing with machetes and guns” while others “acted in support roles—allowing murder squads access to hospitals and homes, cheering on male killers, stripping the dead and looting their houses.”100 Calixte, the survivor I talked with at the church in Ntarama, gave a ground-level view of female agency in the massacres. “Women killed a few, but mainly waited for Tutsi women crossing the river with a kid on the back, so they would take the kid and throw it in water. Or they did espionage, reporting on who was hiding where.”101 Aloysius Inyumba, the RPF minister in charge of women’s affairs in 1995, gave me a more global picture. “I have 838 women in prison. One woman said to me, I have only killed eight; there are people who have killed many and are free.”102 Alongside women there were children, eight hundred according to one report. Their ages ranged from seven to seventeen years. The vast majority of the older children “were charged with genocide.”103

Killings Sanctified

Just as the killing in Rwanda was not done by shadowy death squads but by mobs of ordinary people guided by armed militia and trained infantrymen, the killing also did not happen in secluded but in public places. Most often, the killings happened in places of worship. Contrary to Gérard Prunier’s contention that “the bystanders were mostly the churches,”104 the church was a direct participant in the genocide. Rather than a passive mirror reflecting tensions, the Church was more of an epicenter radiating tensions.

Like the middle class of which they were a prominent part, priests were also divided between those who were targeted in the killings and those who led or facilitated the killings. Here, too, there was hardly any middle ground. A Lutheran minister recalled what the gangs told him: “You can have religion afterwards.” Explaining why he walked around with a club, the minister told a reporter: “Everyone had to participate. To prove that you weren’t RPF, you had to walk around with a club. Being a pastor was not an excuse.”105 Priests who had condemned the government’s use of ethnic quotas in education and the civil service were among the first victims of the massacres. In all, 105 priests and 120 nuns, “at least a quarter of the clergy,” are believed to have been killed. But priests were not only among those killed, they were also among the killers. Investigators with the UN Center for Human Rights claimed “strong evidence” that “about a dozen priests actually killed.” Others were accused of “supervising gangs of young killers.”106Hugh McCullum, a former editor of the United Church Observer and author of The Angels Have Left Us, told The Toronto Starthat two-thirds of Rwanda’s Catholic priests either died during the genocide or ran into exile after the genocide. He said the figures were similar for the Protestant churches.107 Twenty-seven of the priests in Goma sent a letter to the pope on 2 August 1994, claiming that the massacres in Rwanda were “the result of the provocation and of the harassment of the Rwandese people by the R.P.F.” They continued in the same vein: “We dare even to confirm that the number of Hutu civilians killed by the army of the R.P.F. exceeds by far the Tutsi victims of the ethnic troubles.”108 At a press conference in Nairobi in early June, more than two months into the genocide, the Anglican archbishop refused unequivocally to denounce the interim government. The Catholic archbishop even moved with the interim government from Kigali to Gitarama.109

How low the moral terpitude of the clergy had sunk is illustrated by a story Jean Carbonnarre, honorary president of the Paris-based NGO Survie, narrated to correspondents of the Inter-Press Service: “André Karamaga, president of the Anglican Church in Rwanda, told me that he went to the Taba commune near Kigali to settle a dispute between two priests quarreling over who should run the parish. The first priest told Karamaga that he was more deserving because the second priest had killed 15 people. When Karamaga challenged the second priest, he admitted the killings, but still maintained that he deserved to run the parish as, he said, the other priest had killed even more.”110 Father Wenceslas Munyashyaka, the curate of Sainte-Famille church, sheltered eight thousand refugees but provided the militia members with lists of those he alleged had expressed sympathy for the RPF and agreed to let them come and pick off those they wanted. Wearing a flak jacket and toting a pistol during the massacres, he fled to Goma with the interim government and was one of the twenty-seven who wrote to the pope defending the Rwanda army and blaming the RPF for massacres.111 In an open letter to His Holiness, Pope John Paul II, Rakiya Omaar, director of African Rights, listed the most shocking instances of clergy organizing massacres, and summed up participation of the Church in the genocide: “Christians who slay other Christians before the alter, bishops who remain silent in the fact of genocide and fail to protect their own clergy, priests who participate in the murder of their parishioners and nuns who hand people over to be killed cannot leave the Church indifferent.”112

How could it be that most major massacres of the genocide took place in churches? 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? Médecins sans Frontières, a medical charity, pulled out of the University Hospital in Kigali after its patients kept disappearing. The British Medical Journal quoted the testimony of Dr. Claude-Emile Rwagasonza: “The extremist doctors were also asking patients for their identity cards before treating them. They refused to treat sick Tutsis. Also, many people were coming to the hospital to hide. The extremist doctors prevented many of these people from hiding in the hospital.” A medical doctor, a member of the hospital staff, directed the militia into the hospital in Kibeho and shut off the power supply so that the massacre may proceed in darkness. Some of “the most horrific massacres occurred in maternity clinics, where people gathered in the belief that no one would kill mothers and new-born babies.”113 “The percentage of doctors who became ‘killers par excellence’ was very high,” concluded African Rights on the basis of extensive investigations. They included persons as highly qualified as Dr. Sosthène Munyemana, a gynecologist at the University Hospital of Butare, Rwanda’s principal teaching hospital. “A huge number of the most qualified and experienced doctors in the country, men as well as women—including surgeons, physicians, paediatricians, gynaecologists, anaesthetists, public health specialists and hospital administrators—participated in the murder of their own Tutsi colleagues, patients, the wounded and terrified refugees who had sought shelter in their hospitals, as well as their neighbours and strangers.” In a sector as small as Tumba, three doctors played a central part. Of these, one was a doctor at Groupe Scolaire Hospital, and the other, her husband, was the health director for Butare. “Two of the most active assassins in Tumba” were a medical assistant and his wife, a nurse.114

Close on the heels of priests and doctors as prime enthusiasts of the genocide were teachers, and even some human rights activists. When I visited the National University at Butare in 1995, I was told of Hutu staff and students who betrayed their Tutsi colleagues and joined in their physical elimination. Teachers commonly denounced students to the militia or killed students themselves. A Hutu teacher told a French journalist without any seeming compunction: “A lot of people got killed here. I myself killed some of the children…. We had eighty kids in the first year. There are twenty-five left. All the others, we killed them or they have run away.”115 African Rights compiled a fifty-nine-page dossier charging Innocent Mazimpaka, who was in April 1994 the chairman of the League for the Promotion and Defence of Human Rights in Rwanda (LIPRODHOR) and simultaneously an employee of a Dutch aid organization, SNV, with responsibility for the genocide. Along with his younger brother, the burgomaster of Gatare commune, he was charged with the slaughter of all but twenty-one of Gatare’s Tutsi population of 12,263.116 Rakiya Omaar pointed out that “several members of human rights groups are now known to have participated” in the killings, refuting “the notion that an independent civil society—of which the educated and the political opposition were the backbone—resisted the project of genocide.”117

That victims looking for a sanctuary should seek out churches, schools and hospitals as places for shelter is totally understandable. But that they should be killed without any let or hindrance—even lured to these places for that purpose—is not at all understandable. As places of shelter turned into slaughterhouses, those pledged to heal or nurture life set about extinguishing it methodically and deliberately. That the professions most closely associated with valuing life—doctors and nurses, priests and teachers, human rights activists—got embroiled in taking it is probably the most troubling question of the Rwandan genocide.

One could go on narrating atrocity stories ad infinitum, and indeed some have.118 The point of such an exercise may be to show how base human nature can be, or it may be, I fear, more self-serving: to show how base is the nature of some humans, usually some others, not us. This is not my purpose, nor do I wish to shut my eyes to atrocity when atrocities have indeed been perpetrated. My point, though, is that atrocity cannot be its own explanation. Violence cannot be allowed to speak for itself, for violence is not its own meaning. To be made thinkable, it needs to be historicized. My preoccupation is not with the universal character of evil, with describing acts of cruelty to underline the fact that people—or some people—are capable of unspeakable cruelty. It is, rather, with trying to understand the political nature of violence—that its targets are those defined as public enemy by perpetrators who see themselves as the people—and thus with the process that leads to it and the specific conditions that make this possible.

This study has located the genesis of Hutu/Tutsi violence in the colonial period, specifically around the 1959 Revolution and its recurrence at times when Hutu and Tutsi emerged as identities of groups contending for power, as in 1959–63 and 1990–94. True, elements of Tutsi Power began to gel over a long period, from the sixteenth century onwards, but the identity “Tutsi” became associated with privilege only in the reign of Rwabugiri toward the end of nineteenth century. Yet, it was not until Belgian colonialism that the local state structures were fully Tutsified, and that Tutsi hardened into a category signifying local privilege. From being a transethnic distinction of local significance, Belgian colonialism inserted Tutsi and Hutu into the world of races and indigeneity. Key to the political impact of Belgian colonialism is the opposite ways in which it constructed Tutsi and Hutu—Hutu as indigenous and Tutsi as alien—thereby racializing the difference between them. Unlike at any time in Rwanda’s history, the Tutsi were presented as both a nonindigenous and a civilizing influence, as Caucasians of a lesser breed. The identities “Tutsi” and “Hutu” were politically enforced through state-issued identity cards. The educational system separated Tutsi from Hutu—and not just chiefs from commoners, as in other colonies—and nurtured its beneficiaries in notions of the Hamitic hypothesis. Hutu were effectively excluded from recruitment in local government and in the priesthood, both of which were completely Tutsified. More than ever in the history of Rwanda, the colonial world effectively sealed the Hutu into a servile status: while the mass of Hutu were compelled to do forced labor (ubureetwa) for Tutsi chiefs, the few who would have been ritually ennobled in a previous era were also branded into a servile condition. Their only salvation lay in politics.

The Hutu middle class had a different history from that of the Tutsi middle class. This different history was also adversarial: it made of them a counterelite. While many Tutsi could date their climb to a middle-class status to educational opportunity and civil service appointment in the colonial period, few Hutu could put forth such a claim. Their history was more likely one of individuals whose aspirations had been frustrated by the racialized structure of colonial Rwanda. Unlike their Tutsi counterparts, most members of the Hutu middle class dated their genesis from the 1959 Revolution, and not from the colonial period. They not only provided the leadership of the 1959 Revolution, they were also its main beneficiaries. The Hutu middle class was both the child of the 1959 Revolution and its proud inheritor. In time, many came to criticize postrevolutionary power, but hardly any voiced criticism of the revolution. When youthful members of this class began calling for democratic rights toward the end of the 1980s, they presumed a context where the gains of the revolution were secure. The civil war that began with the RPF invasion of 1990 removed this certainty. It put the Hutu middle class in a state of crisis, and it hurled them yet again into a world where Hutu and Tutsi were names of corporate groups contending for political power. It was a world in which democratic opposition came to be synonymous with treason.

POWER struggles in the bipolar world of Hutu and Tutsi are marked by a truism: not only are members of the middle class the main beneficiaries of every victory, they are also the core victims in every defeat. And since victory for one is defeat for the other, every struggle bears the hallmark of a life-and-death tussle. Three events in recent history bear testimony to this truism: 1959 in Rwanda, and 1972 and 1993 in Burundi. Few Rwandans could have been unaware that when the Tutsi army in neighboring Burundi unleashed terror on Hutu in 1972 and killed nearly 200,000, it did not go for the lives of ordinary Hutu but for those of school-going youth. The objective was to crush the flower from which would come tomorrow’s intelligentsia. Anyone who had come out of the colonial period understood that the existence of an intelligentsia was the prerequisite for initiative, independence, and leadership.

The 1972 massacre of Hutu school youth in Burundi stirred demands in Rwanda. Critics said that while the Hutu revolution of 1959 had managed to transform the state, its impact on society had been superficial. While Hutu had gained political supremacy, they had yet to win social supremacy, whether in education or in the marketplace. This unrest was the context of the emergence of the Second Republic, a fact all the more remarkable since it was the Second Republic under President Habyarimana that began the process of deracializing the Tutsi and of reintegrating them into the Rwandan polity as an ethnicity. From one based on the distinction between two races, one indigenous and the other alien, the political discourse in the Second Republic turned around the distinctionbetween majority and minority. From a nonindigenous race, the Tutsi became an indigenous minority. As an ethnic group, they could aspire to rights like other Rwandans; but as a minority, they would have to give up any claim to power.

The life of the Second Republic was cut short by the RPF invasion of 1990. The invasion literally reversed the dynamic of the Second Republic. By highlighting the distinction between the struggle for rights and the pursuit of power, it once again polarized Hutu and Tutsi as political identities. Key to this shift, this repolarization of Hutu and Tutsi as political identities, was the growing realization that the real objective of the RPF invasion was not rights but power—specifically Tutsi Power. The assassination of the newly elected Hutu president, Melchior Ndadaye, in Burundi in 1993 by a Tutsi army merely confirmed this realization as a truism. It is, after all, defeat in the civil war, and the specter of Tutsi Power, that provided the context in which a tendency born of Hutu Power—the génocidaire—chose to embrace death in preference to life. For the Hutu from Rwanda’s postrevolutionary middle class, and those from other sectors, who followed the lead of the génocidaires, the true stake in the civil war was the key gain of the 1959 Revolution. That gain was a future for the Hutu, synonymous with the existence and expansion of a Hutu middle class. While the existence of the middle class could be said to be of direct interest to its members only, its expansion surely interested all those of humble origin, peasants or artisans, who looked with hope to their children and to the future. This point of view begins to make some sense of why so many hundreds and thousands who had never before killed participated in the genocide; why they included not only the victims of Hutu/Tutsi civil wars (the displaced and the refugees), but many more yet to be touched by the expanding civil war; why they thought of massacres of neighbors not only as a continuation and culmination of the civil war but also as a defense of the gains of the revolution; and why so many of these came disproportionately from Rwanda’s Hutu middle class.

One needs to remember that the Tutsi were killed as Hamites, not as Tutsi. Whereas the Tutsi of precolonial vintage never claimed political privilege because they came from elsewhere, the Hamites of colonial vintage were said to be “a civilizing influence” for no other reason than that they were said to have come from elsewhere. Whereas ethnicized Tutsi existed before colonialism, the racialized Hamites were creatures of colonialism. As a political identity, ethnicity marked an internal difference, whereas race signified an external difference. Ethnic conflict does not breed genocide; at most, it can give rise to massacres. The difference is that only with genocide is there an attempt to obliterate the other physically and totally. When it comes to ethnicities, there is no question about the legitimacy of their presence in the political arena: as indigenous groups they belong, all of them. The points of conflict concern borderline issues, matters of excess. It is only with racialized groups, those constructed as nonindigenous, that their very presence in the political arena is considered illegitimate. And the racialization of the Tutsi, we must not forget, was the joint work of the state and the Church.

Herein lies the clue as to why the violence was marked by a greater fury in the Church than in any other institution in Rwandan society. The Church was the original ethnographer of Rwanda. It was the original author of the Hamitic hypothesis. The Church provided the lay personnel that permeated every local community and helped distinguish Hutu from Tutsi in every neighborhood: without the Church, there would have been no “racial” census in Rwanda. At the same time, the Church was the womb that nurtured the leadership of the insurgent Hutu movement. It provided the intellectual and organizational backup for this movement, from talent as ghostwriters to funding for the cooperative movement which oiled the tentacles that ran through Rwandan society like so many arteries through a body politic. It is from the ranks of the Church-connected movement that the leadership of the 1959 Revolution was drawn. On the morrow of the revolution, that same leadership used the power of the state to establish control over church education, both in terms of its content and its personnel. The fusion of the Church and the state, both in personnel and in vision, was symbolized by the fact that the archbishop of Rwanda sat as a formal member of the Central Committee of the ruling party until he was forced to resign in the early 1990s, on the eve of the pope’s visit to Rwanda.

But the Church also differed from every other state-connected institution in Rwanda. While Hutu came to occupy the top echelons of the Church hierarchy, its middle level continued to include a substantial number of Tutsi. The civil war brought the power struggle in the Church to a climax. In the Church, there could be no middle ground, no sanctuary. Rather than a place of healing, the Church turned into a battleground for settling scores.

“A religious community which wages wars,” wrote Carl Schmitt, “is already more than a religious community; it is a political entity.”119 So indeed was the Church in Rwanda. Let us recall that there was no single institutional home, no mortuary, bigger than the Church for the multiple massacres that marked the Rwandan horror. After all, but for the army and the Church, the two prime movers, the two organizing and leading forces, one located in the state and the other in society, there would have been no genocide.

If it is the struggle for power that explains the motivation of those who crafted the genocide, then it is the combined fear of a return to servitude and of reprisals thereafter that energized the foot soldiers of the genocide. The irony is that—whether in the Church, in hospitals, or in human rights groups, as in fields and homes—the perpetrators of the genocide saw themselves as the true victims of an ongoing political drama, victims of yesterday who may yet be victims again. That moral certainty explains the easy transition from yesterday’s victims to killers the morning after.

If you find an error please notify us in the comments. Thank you!