Boston Review
table of contents
new democracy forum
new fiction forum
rave reviews
writers’ guidelines
bookstore locator
literary links


Search this site or the web Powered by FreeFind

Site Web


The Legacies of Collective Violence

The Rwandan genocide and the limits of law.

Helena Cobban

8 The 1998 Rome Treaty, which set the rules for a new, permanent International Criminal Court (ICC) will soon receive the last of the sixty ratifications required to trigger the ICC's establishment. This development will crown a decade-long campaign by human-rights groups, aimed at creating a transnational tribunal capable of enforcing human rights declarations to which governments around the globe have given rhetorical support since the 1940s.

Now that the establishment of the ICC is imminent, it is crucial that we understand just what kind of contribution it might make to the well-being of the citizens in its global jurisdiction. Fortunately, we can draw some preliminary lessons from over eight years of experience with the ICC's two ad-hoc precursors: the International Criminal Tribunal for former Yugoslavia (ICTY), based in The Hague, and the International Criminal Tribunal for Rwanda (ICTR), based in Arusha, Tanzania.

Much evidence from the former Yugoslavia and Rwanda—two areas devastated by ethnic cleansing and genocidal mass murder—should make us nervous about the role to be played by the future ICC. When it established ICTY and ICTR, the UN Security Council expressed its hope that these bodies would help to bring inter-group reconciliation to the societies in question. That goal has not been achieved, if it was even pursued. Indeed, given the adversarial nature of proceedings in a Western-style criminal court—a system designed to produce clear "winners" and "losers"—it is hard to see how the courts could have achieved it, absent any broader political strategy for fostering peace and reconciliation.

The problem—for the existing ad-hoc courts and for the ICC—is more fundamental than just the polarizing effect of criminal-court proceedings. A criminal justice system by its very design crowds out, or at least asserts "primacy of jurisdiction" over, any other possible approach that societies or governments might use address the legacies of communal violence. This tends to rule out, for example, the creation of South-African-style 'truth commissions.' In late 1998, when leaders of the three ethnic groups in Bosnia were discussing creating a joint truth commission to establish a common record of the past decade, they were told bluntly by ICTY's Chief Prosecutor, Louise Arbour, that such an effort would contaminate her evidence. Arbour also told aid donors not to support the Bosnian initiative, which set back its plans considerably.1 Once the permanent ICC is established, it will become far more difficult, perhaps impossible, for members of war-torn societies to opt for any reconciliation process that might, like the South African process, offer amnesty to former rights abusers in order to pursue broad communal goals.

My colleagues in the human-rights movement are all, I am confident, dedicated to the broad goal of fostering human well-being. (Wasn't that what brought us to the movement in the first place?) I submit that the well-being of our fellow humans in conflict-torn societies like the former Yugoslavia and Rwanda requires first and foremost that they escape the cycles of retributive violence in which they have become so deeply entangled. Left unbroken, these cycles can continue indefinitely, mocking any chance these people may have of winning meaningful assurance of any of their civil, political, economic, or social rights.

However, reducing violence, achieving the stable coexistence of different social groups, and moving toward long-term reconciliation are all political, not juridical, tasks. Dealing effectively and compassionately with the legacies of past violence is an important part of this broader political process—and perhaps criminal courts can help. But the record from Rwanda, the former Yugoslavia, and other countries suggests that their role is far more limited than many of us had supposed.

The Rwandan experience in particular offers a number of rich lessons about the efficacy (or lack thereof) of the juridical approach to the legacies of violence. First, there is the record of the ICTR court in Arusha: a record that a recent authoritative study summed up as "lamentable." Second, there is the experience of Rwanda's post-genocide government, in its attempt to prosecute alleged genocide participants in its own national courts—an effort which ran parallel to the ICTR prosecutions. By 1999, the limited effectiveness of the prosecutions-focused approach had become evident to Rwanda's leaders. By then, roughly 125,000 alleged génocidaires were crammed into overcrowded detention camps. Thousands had been languishing there, untried, for several years and few could expect a trial in the country's hard-stretched courts any time in the next twenty to thirty years. In the face of this crisis, the government decided to move most of these cases to a new hearing system, based on a traditional mechanism for resolving disputes called gacaca (gä chä´ chä), or "the lawn." The first gacaca courts will be launched this May or June.

The Rwandans' experience is of interest, too, because of what it suggests about the phenomenology of genocide itself. In the 1940s, Hannah Arendt argued that the European Holocaust was not a "normal" event, and that although the instruments of a criminal justice system designed for settled times might be necessary to confront it, they could never be sufficient. In Rwanda, the mass-participatory aspect of the 1994 genocide made it an even less "normal" episode in human history than the Shoah.

The Rwandan genocide throws into profound relief many of the cosmological and ethical assumptions—about the nature of individual responsibility, the purpose of punishment, and the normal conditions of human life—upon which our contemporary criminal-court system is based. We in the West seldom examine these assumptions. But the Rwandan case challenges them deeply and calls on us to tread lightly and carefully before we spread the mantle (or strait-jacket) of our criminal justice system over populations or situations to which it may be fundamentally unsuited.

                                                                *      *      *      

When we arrived at Nyanza, they told us to sit down. They divided us into two groups and ordered us to take up sleeping positions on the road. The soldiers told the interahamwe [militia members] to wait. Clearly they wanted to begin the massacre themselves. Then the soldiers began by throwing grenades. They would sometimes throw fifteen grenades at a time. There was blood everywhere and people moaning in agony….

   I was lying between a lot of people. I was not touched either by the grenades or the guns….I heard one of the soldiers say to a group of interamhamwe , 'Go and see if there is still anyone alive.' They had machetes, knives, spears and bows and arrows. They used these weapons to finish off the wounded. They did not even spare children and old people who were wounded. Then they came to me. A huge machete landed on my neck and my right hand. As if that was not enough, they cut me up with a knife all over. Finally they left in the evening.

   They came back very early the next morning. This time it was to loot.2

The legacies of genocide

Rwanda's genocide started on April 6, 994, and continued for ten weeks. It was precipitated by a still-mysterious rocket attack that brought down a small plane carrying the country's president, Juvénal Habyarimana and Cyprien Ntaryamira, president of neighboring Burundi. Both presidents perished in the crash. Both were members of the region's Hutu caste, and the killings that followed—which were carried out by Hutus mainly against members of the Tutsi caste—may at first have looked to some like spontaneous acts of revenge.3

But there is now (and was at the time) considerable evidence that the genocide had been meticulously planned and prepared during the weeks or even months prior to April 6. Members of the country's Hutu-dominated armed forces and the extremist 'Hutu power' militias had been acquiring and distributing weapons. They had drawn up and circulated lists that prioritized those who were to be killed. Members of the Tutsi intellectual elite and Hutus thought to be pro-Tutsi topped these lists. Also listed by name, down to the level of individual villages and neighborhoods, were members of Tutsi families—men, women, elders, and children. For months, the extremists had used the country's printing-presses and airwaves to disseminate hate-filled propaganda that routinely referred to Tutsis as "cockroaches" (inyenzi) who should be eliminated in the interest of national hygiene—and to any Hutus who might disagree with this program as "traitors" (ibyitso).

In the ten weeks that followed April 6, some 800,000 people—around 10 percent of the national population—were killed. More than 90 percent of the victims were Tutsis. These killings were carried out intentionally and systematically—in stadiums to which the pre-designated victims had been herded to await their deaths; in churches, hospitals, and local government buildings where terrified Tutsis had fled for sanctuary; and in homes and at roadblocks throughout the country where drunken groups of interahamwe militiamen used machetes and nail-studded maces to bludgeon their targeted victims to death.

It was so routine, so sustained, that the perpetrators simply called it "work." It was so well-organized and so efficient that the killing rate was three to five times the rate at which, at the height of the Holocaust in Europe, Adolph Hitler's industrialized killing system was able to dispatch its victims.

Unlike the European Holocaust, however, this genocide was—according to its very design—a highly public affair, whose organizers actively sought to enlist in its deadly "work" the maximum possible number of Hutu participants. Human-rights researchers have documented much evidence of this broad, public mobilization. For example, one researcher interviewed Jean Bosco Bugingo, a married father of five who lived in the Gitaziga sector of Muhazi commune. "I personally macheted five people to death," Bugingo acknowledged. He underscored the active role that local officials called 'councillors' played in persuading people in Gitaziga to take part:

Our councillor, Joseph Munyaneza, told the Hutus to rise up and defend ourselves….At the meeting the councillors said Hutus must kill Tutsis and everybody who was against the government, so that when the [Rwandese Patriotic Front rebels] came they would not have anyone to rule. Many Hutus pointed out that they had lived with and intermarried with these people they are being told to kill. Munyaneza said: "Either you kill them or you will be killed." He told us that we would be moving around with soldiers and "They would see to it that either you killed or you die." In the end, even those who had hesitations had to kill too. They killed with less zeal, but they killed.4

The compilers of this account suggest that descriptions such as Bugingo's contain "no special pleading: many survivors readily admitted that some of the participants were reluctant murderers."5 (It is also important to note the many cases of Hutus who withstood the enormous pressure to participate or who ran considerable risks to save the lives of those targeted. Numerous Hutus lost their lives as they attempted to withstand the pressures of the génocidaires in these ways.)6

What was the intent of this mass mobilization for genocide? Was it intended simply to complete the killing of the country's Tutsi population as rapidly as possible? Or were the organizers of the genocide also hoping to give the largest possible number of Hutus the bonding experience of participating in a pan-Hutu baptism in the blood of their foes? Was it designed to implicate as many individuals as possible in the killings, and thus to make any future assignation of responsibility for specific acts of genocide just about impossible?

Whatever the reasoning of the organizers, the mass-participatory aspect of this genocide gave it a psychosocial content significantly different from that of the European Holocaust with its more 'sanitized,' mechanized, and secret methods of killing. Unlike the Shoah, too, once this genocide had been brought to a halt there was no clear place for survivors to flee to or regroup. In early July, forces of the Rwandan Patriotic Front (RPF), which had grown up among Tutsi exiles in nearby Uganda, were able to capture the Rwandan capital, Kigali. As the RPF gained control of the country, traumatized survivors of the genocide came out of the swamps or attics where they'd been hiding and received some succor for their wounds. The genocidal 'Hutu power' leaders who had been running Rwanda since April 6 fled to neighboring countries, taking an estimated 2 million terrified Hutu civilians with them. But most of the country's Hutus either remained in their home villages, or returned to them in the months that followed. Meanwhile, the survivors of the genocide—who were pitiably few in number—haunted the country like ghosts. Most found that their homes had been burned, their family members killed, and their communities destroyed or changed forever. The new administration was dominated by Tutsis who, like the RPF leaders, had grown up in exile, and who streamed back into Rwanda after the RPF seized power.7 The new government claimed to speak for the survivors, but was often operating out of an entirely different set of experiences and motives.

Members of the post-1994 government insist that they are dedicated to the interests of all of the country's 7 million citizens, regardless of caste or other divisions. But caste allegiance remains strong, perhaps inevitably after the horrors of 1994. Meanwhile, all the government's efforts to de-emphasize caste have been unable to mask the fact that even after the return of hundreds of thousands of previously exiled Tutsis, Tutsis still constitute only about one seventh of the national population. Nearly all the rest are Hutus. (About one percent of Rwandans belong to the tiny 'Twa' group.)

Rwanda's Tutsis desperately need to find a way to coexist with the Hutus: there is no place either inside or outside the country where they can hope to regroup in a compact and self-supportive way. But given the events of 1994 and the sharp demographic imbalance between the two groups, finding a mechanism to bring this about presents a profound political challenge—for Rwanda's leaders and for their friends in the international community.

Western guilt

In the summer of 1994, the Cold War had only recently sputtered to a halt and policymakers worldwide were still trying to work out the global organizing principles that might replace it. Building an 'international rule of law' seemed like a real prospect. Just one year before Rwanda's genocide, the UN Security Council had established ICTY, giving it a mandate to prosecute "persons responsible for serious violations of international humanitarian law committed in the territory of the former Yugoslavia since 1991."8

The establishment of ICTY had two consequences that would be of key significance for Rwanda. First, it took the decisions on whether and how to prosecute alleged rights abusers out of the hands of the domestic governments concerned, and into those of a new international Prosecutor appointed by the Security Council. Second, it sent a clear message that henceforth, neither the interests of diplomacy nor any other 'reasons of state' would be allowed to stand in the way of such prosecutions.

In the summer of 1994, it was still too early to assess the effectiveness of the ICTY experiment. Indeed, the new court would take considerably longer than that to assemble its infrastructure and issue its first indictment. In 1994, human-rights activists and their allies in Western governments were still optimistic that ICTY could help 'end the culture of impunity' for atrocities, by holding former perpetrators strictly accountable for their deeds.

In Rwanda, the genocide persisted through April, May, and much of June 1994. As the killing continued, two major aspects of the response from the American-led 'international community' became increasingly clear. First, none of the rich, Western governments would take any of the steps—entirely within their capability—that could have either halted the genocide in its tracks, or reduced its murderous consequences considerably.9 Second, given the tension between this failure to act (or decision not to act) and the self-image of many of these governments (including the first Clinton administration) as being committed to liberal internationalism, the policies that they did adopt toward Rwanda would almost inevitably be a mishmash of guilt-driven and poorly-thought-through gestures.

One such gesture was the decision to send massive humanitarian aid to Rwanda, and to Rwandan refugees, in the immediate aftermath of the genocide. That decision was evidently well-intentioned. But much of the aid was sent to help the new waves of refugees who streamed out of Rwanda in advance of the progressing RPF forces, and aid workers in the refugee camps rapidly came to realize that the primary effect of the aid distribution was to enable the génocidaires to regroup. This posed searing dilemmas. Conditions in the camps were often terrible, with the weakest refugees—women, children, the elderly—ravaged by disease. Nevertheless, the aid workers, their bosses in humanitarian organizations, and Western governments struggled to find ways to distribute aid that, instead of benefiting the génocidaires, would support the new Rwandan government's efforts to disarm the refugees and attract them back to their homes.10

The miasma of guilt that hovered over Western governments in the summer of 1994 resulted in another major policy initiative that was, perhaps, just as poorly planned. This was a strong—and again, well-intentioned—drive to extend the new norms of 'accountability' and 'ending impunity' to the perpetrators of atrocities in Rwanda, as in the Balkans. In Rwanda, this drive took two forms. At the strong urging of Western governments, the Security Council established a new ad-hoc tribunal for Rwanda, parallel to ICTY, and sharing some facilities with it. ICTR was created in November, 1994, with the aim of prosecuting the top-level organizers of the genocide.

Throughout the negotiations over ICTR's establishment, however, Rwanda's post-genocide government expressed reservations about many aspects of the future court's work. By chance, Rwanda had a seat on the Security Council in 1994; in November, it cast the only vote there opposing the creation of ICTR. Broadly speaking, the Rwandans objected to the idea that officials of the new international court, and not the Rwandan government, should make the very important decisions on who should be tried in cases related to the genocide carried out in their country, and how this should be done.11 Accordingly, even after the ICTR resolution passed, the Rwandan government continued with its plan to prosecute as many participants and planners of the genocide as possible in its own national courts. As for Western governments, their attitudes to the government in Kigali were still driven largely by guilt. So they enthusiastically supported the government's plans to pursue national-level prosecutions alongside those carried out by ICTR.12

Thus, by the end of 1994, it was clear that, in attempting to deal with the legacies of the genocide, both the new Rwandan government and the international community would be focusing their efforts on prosecutions. There would be no room in Rwanda for other approaches to dealing with the aftermath of grave political violence—like the community-healing approach that Mozambique's people were using at that time to deal with the repercussions of their lengthy civil war, or the truth-seeking approach that South Africa's new democratic government adopted in 1994–95 to deal with the legacies of apartheid. Rwanda would instead be a test-case for a new world order based on strict application of the norms of criminal justice.

The limits of prosecution

Judge Robert Jackson, the Chief Prosecutor at the Nuremberg Tribunal, once memorably remarked that, "Courts try cases—but cases also try courts." The genocide-related cases taken up by ICTR and the Rwandan national courts have certainly tried both those institutions, and both have been found wanting.

Regarding ICTR, a study released by the prestigious International Crisis Group in June 2001 judged its performance "lamentable." The study noted that seven years after its establishment, ICTR had handed down verdicts on only nine defendants—despite the fact that it boasted "more than 800 employees, three trial chambers presided over by nine judges, and a budget of around $90 million."13 (This figure was ICTR's annual budget around the turn of the century. That amounts to 22 percent of the foreign aid for Rwanda in 1999, and is considerably more than the total foreign aid received by the 6.3 million citizens of Rwanda's neighbor, Burundi, or that received by numerous other impoverished African nations.)

The Crisis Group study did acknowledge that ICTR had achieved "some symbolic results." These included the first conviction of a defendant on a charge of genocide following a substantial trial before an international court and some contribution to neutralizing the ability of the 'Hutu power' apparatus to regroup internationally after 1994. However, the study also noted a number of failures, including serious questions about the case-management abilities of ICTR's judges, the competence of the prosecutorial staff, and the management of the defense lawyers and their budgets. Beyond these procedural problems, the Crisis Group reported the tribunal's failure to fulfill the hope the Security Council had expressed, that it might "contribute to the process of national reconciliation and to the restoration and maintenance of peace." The study noted that "[t]he tribunal's contribution to national reconciliation was non-existent as long as it was still perceived to be delivering victors' justice."14

Most Rwandans have expressed comparably negative assessments of ICTR's performance. In 1996, André Sibomana, a seasoned Rwandan human-rights activist, journalist, and Catholic priest, concluded that ICTR's main effect had been

to enable the international community, or rather the countries within it, to save face and give the impression to the public that the crime they watched without intervening would not go unpunished….

   I have met some of the ICTR officials; I am amazed by their incompetence. They are very intelligent people, but completely incapable of carrying out research. They don't speak Kinyarwanda—which is understandable—but nor do they know how to employ competent interpreters….They are incapable of approaching those who lived through the genocide. They don't ask the right questions. People are offended by their attitudes and their discourse. Rwandans had invested great hope in the ICTR. They are very disappointed.15

As for the Rwandan government, it has continued to be sharply critical of ICTR's work. An official assessment of ICTR posted on the web site of the Rwandan embassy in Washington, D.C. states that the international court "has since its inception operated without any discernible strategy for investigation and prosecution of the crimes covered by its mandate. The prosecutors have never determined [a] policy as to whom the tribunal should pursue. They have never indicated the kind of cases they wish to prosecute before the tribunal and those they expect to be tried by national courts."16

Given these criticisms, it is scarcely surprising that the Rwandan government continued its own parallel effort to bring as many accused perpetrators as possible to "justice." According to the conception of justice that dominates Western societies, and that long dominated the thinking of the Rwandan government, this required four sequential steps: detaining the accused génocidaires, bringing concrete indictments against the detainees (and other suspects who might still be at large), trying the indicted individuals in courts operating according to acceptable standards, and finally, administering punishment to those found guilty.

The Rwandan government's wholehearted adoption of the prosecutorial approach had troubling consequences. In the early post-genocide period, this was undoubtedly due to the sheer numbers of those accused, and the logistical inability of the Rwandan legal system to move the vast majority of these cases anywhere beyond the first of the four steps listed above.17 Detentions of alleged génocidaires were widespread in the eighteen months following the RPF's takeover of the country—and again, during subsequent periods of unrest between 1996 and 1998. In the immediate aftermath of the genocide, it is true, there had been numerous cases of accused génocidaires being summarily executed, either by members of the RPF military or by civilians. In comparison with that outcome, perhaps detaining the suspects and giving them some hope of due process was more humane. But as the numbers of detainees mounted, it became increasingly clear that conditions in the detention camps were extremely inhumane, and in many cases life-threatening. Meanwhile, the court system itself had also been a major victim of the genocide: André Sibomana reported that after 1994, fewer than 250 of Rwanda's previous roster of 719 magistrates were left in the country. Because of the continuing frailty of the Rwandan justice system (and perhaps for other reasons, too), only a tiny number of the detainees ever had any concrete charges brought against them—and fewer still ever saw the inside of a courtroom.

The number of detainees rose rapidly above 100,000. In early 1995, Sibomana visited Gitarama Prison for the first time. "What I saw defied imagination," he told two interviewers in 1996:

There were three layers of prisoners: at the bottom, lying on the ground, there were the dead, rotting on the muddy floor of the prison. Just above them, crouched down, there were the sick, the wounded, those whose strength had drained away. They were waiting to die. Their bodies had begun to rot and their hope of survival was reduced to a matter of days or even hours. Finally, at the top, standing up, there were those who were still healthy. They were standing straight and moving from one foot to the other, half asleep. Why? Simply because that's where they happened to be living. Whenever a man fell over, it was a gift to the survivors: a few extra centimeters of space. I remember a man who was standing on his shins: his feet had rotted away.18

Sibomana reported that within nine months, almost 1,000 of Gitarama's 7,000 detainees had died as a result of this ill-treatment, which he described as "deliberate." He noted that as soon as his diocese gained permission to intervene in the prison in spring 1995, the mortality rate started dropping: the church distributed makeshift sandals to the prisoners, which allowed them to avoid gangrene simply by keeping their feet dry.

Sibomana cited the Rwandan Constitution's explicit guarantee that "[e]very person is presumed innocent of the offences of which he is accused until a final sentence has been issued" and noted that this presumption of innocence is also spelled out in numerous international conventions and treaties to which the government was party. "As a human rights defender and as administrator of the diocese of Kabgayi, it was my duty to think of actions which would ensure that prisons did not turn into death-traps," he said. "The prison guards accused me of wanting to help the interahamwe. In their minds, even before they had been tried or simply heard, these people had ceased to be human beings."19

Despite occasional releases of prisoners, the number of detained genocide suspects continued to rise. By late 1998 it reportedly topped 140,000. Further programs of controlled releases, and a slight pick-up in the speed at which the courts processed cases, brought the number down to around 125,000 by mid-2001.

By then, many of the detainees had been in the camps for seven years. Perhaps most of these people were indeed participants—enthusiastic participants even—in the genocide. But certainly, many had been wrongly accused—on the whim of a jealous neighbor, perhaps, or simply by mistake. And until trials could be held, there was no way to separate the falsely accused from the guilty. By 2001, and even after international aid donors had poured large amounts of money into the Rwandan justice system, the courts were still only able to process cases at a rate of 1,500–2,000 per year (and Human Rights Watch was still reporting that "[a]uthorities recognized corruption in the judiciary as widespread and serious").20

Nearly all of the 100,000-plus detainees are able-bodied people of breadwinning age who have, we can assume, a number of dependent family members. It is hard not to conclude that the continued separation of these people from productive life places a huge burden on a country whose total population is just over 7 million. In addition, nearly every single detainee is Hutu. The government that has incarcerated them is widely perceived as being dominated by members of the Tutsi minority. The prolonged detentions have thus exacerbated the very inter-group tensions that the government claims it wants to erase.

In the late 1990s, to the government's considerable credit, its members started to realize that their single-minded pursuit of a Western-style, highly individualized, prosecutorial approach to the maelstrom of 1994 was proving unworkable or even counter-productive. By 1998, the country's behind-the-scenes strongman (later president), Paul Kagame, was starting to consider an alternative based on the country's tradition of gacaca. As Kagame told Philip Gourevitch around that time,

"With those who masterminded the genocide, it's clear-cut…. They must face justice directly. I'm not as worried about these ordinary peasants who took machetes and cut people in pieces like animals." [Kagame] explained that "long ago" Rwandan justice was conducted in village hearings, where fines were the preferred penalties. "The guy who did the crime can give some salt or something, and that can bring the people back together."21

At that stage, Kagame seems to have been fairly vague about the exact nature of the "village hearings" he was talking about, although gacaca hearings were not merely a feature of Rwanda's distant past, but had continued throughout the 1980s and—presumably with some interruption during the genocide—the 1990s. The fact that Kagame had spent nearly all his life until 1994 in exile in Uganda may help explain his vagueness.

Kagame and his inner circle continued to explore the idea of incorporating some facets of gacaca into their response to the genocide. By 1999, they were canvassing the idea quite openly. It came under intense criticism from Western rights organizations almost immediately. In April 2000, Amnesty International published a report that stated: "Fundamental aspects of the gacaca proposals do not conform to basic international standards for fair trials guaranteed in international treaties which Rwanda has ratified." It added (somewhat imperiously), "[t]he draft law on gacaca must be amended to ensure that standards of justice in Rwanda are not further undermined."22 But the Rwandan government stuck to the essentials of its plan. Later in 2000, Justice Minister Gérald Gahima told an interviewer that, with the upcoming enactment of the Gacaca Law, the government was now putting, "political considerations ahead of the message of [ending] impunity," and even that "[w]e have to put the genocide behind us."23 The final passage of the Gacaca Law, in early 2001, signaled not only a government consensus that the previous stress on prosecutions was no longer desirable for Rwanda, but also a willingness to try to incorporate elements of a very different, 'restorative' approach to issues of justice and wrongdoing into its policy. The new system is expected to go into full operation in May or June of this year.

Gacaca then and now

Filip Reyntjens, a Belgian expert in Rwandan and international law, has studied gacaca since the 1970s. He has written a lively description of one gacaca hearing that he observed, in a rural area, in 1986. A local woman had accused a neighbor of having failed to pay a debt of twenty Rwandan francs. The 'defendant' recognized that there was a debt, but claimed it was only ten francs. She also claimed that the 'plaintiff' had insulted her in public. The plaintiff responded that the defendant had insulted her, too. It was generally agreed that the two womens' insults had caused a disturbance of the peace, and that was why the case had been brought to the local 'lawn.'

According to Reyntjens, about sixty men and women of varied ages gathered around the hearing, joining in with shouts to express support for or opposition to the arguments made. Nine local officials sitting on a bench directed the proceedings—which had to be moved into a small building when it started raining. Finally, after about an hour of discussion, the defendant admitted that the debt had indeed come to twenty francs, and said she would repay the whole amount. The people gathered round then debated what 'damages' she would have to pay to the community for having caused a disturbance of the peace. After some discussion, the total was set at ten bottles of banana beer, which would be contributed to the community. (Reyntjens noted that this penalty would cost her 300 francs.) "The reconciliation would then be celebrated later in the course of a little feast," he wrote.24

He noted that in the Rwandan canton that he studied most closely in 1986, around 1,200 cases were heard in gacaca forums during an eight-month period, compared with 83 handled by the canton's formal court system. There was, he wrote, some distinction between the severity of the cases heard in gacaca and those heard in the courts, but the two systems seemed to be well-integrated. Sometimes, parties dissatisfied with a judgment made at gacaca would take an 'appeal' against it to the formal court, and all judgments rendered at gacaca hearings were registered with the local authorities.

Reyntjens argued in a June 2001 interview that using a gacaca-style system to hear cases of alleged genocide would be a significant departure from how gacaca was traditionally used. Rwanda's Ambassador to the United States, Richard Sezibera, had made the same point to me a couple of weeks earlier. Sitting in the sparsely furnished town-house that is his embassy in Washington, Sezibera said that

"Traditionally, gacaca would be a method of arbitration between families, a situation in which wise men would sit together. It would only deal with minor infractions, or minor disputes….When we revived gacaca, we wanted to tap into the tradition of arbitration that it embodied—but to use it to deal with the genocide. But note that we do call what we're setting up 'gacaca courts,' not just plain 'gacaca'….We've looked at other arbitration systems, in the U.S. and elsewhere. But no one has ever tried to use such a system to deal with a genocide."

His voice trailed off.

"Mainly, we're trying to merge our traditions with current international law."

Ambassador Sezibera is a quietly intense and scholarly person who sets great store by being well-organized. He marshals facts and arguments in numbered order. I asked him to explain the reasoning behind the move to gacaca. "If we want to have justice and reconciliation in our country, we need to build our policy on four pillars," he said.

"The first of these is that we need the deconstruction of the myths that have been built up, in order to help strengthen national unity and national reconciliation. Secondly, we need the punishment of the crimes committed. Third, we need the rehabilitation of those who were victims. And fourth, we need the construction of a new Rwandan national identity…."

Even in this list of basic principles, the classic goals of a criminal justice system—punishment and rehabilitation—were sandwiched between the much more political goals of deconstructing myths and constructing a new identity. In Rwanda, indeed—as in many or most other Third World countries—the whole edifice of a national criminal justice system sits uneasily atop a morass of unresolved but unavoidable issues of power, control, representation, legitimacy, identity, and identification: in short, of politics.

The Ambassador continued his explanation. He recalled that the Genocide Law that his government passed in 1996 divided the cases of the accused génocidaires into four categories. Category 1 is reserved for the highest level of genocide architects, and for those accused of rape and other sexual tortures; Category 2 includes those accused of committing one or more killings during the genocide, but who were not ringleaders; Category 3 comprises those who committed assaults without the intention to kill; and Category 4 includes people accused only of looting or destruction of property. Under the Gacaca Law of 2001, Category 1 detainees—roughly 2,000 to 3,000 people—would still be handled by the country's criminal courts.25 But suspects from Categories 2 through 4 would now, he explained, be dealt with "in ways that enhance national unity and national reconciliation."

Sezibera noted that even before the introduction of gacaca, the courts had been trying to foster these broader objectives by offering incentives—mainly in the form of reduced sentences—for defendants who confessed fully to their deeds.26 "Now, we'll just be strengthening that philosophy with gacaca," he said.

Regarding the second pillar, punishment, Sezibera referred to the need to end the 'culture of impunity' in Rwanda. "If you committed a crime, you will be punished proportionately," he stressed. "Not commensurately, of course. But proportionately. We need to send the message: if you committed a crime, there will be consequences."

On the third pillar, the rehabilitation of survivors, he said it is important that survivors "feel they are a part of the process of healing." Up to that point, he admitted, survivors had caused "problems" when the government tried to reduce the sentences of accused génocidaires. "So we have to make sure that the survivors themselves are a part of the process." He noted that many survivors had lost both people they loved and property during the genocide. "We can't replace the people who were lost. But we must replace the property that was lost, and the whole community must take part in that. Rehabilitation has to be both a moral and a physical undertaking."

Finally, Sezibera addressed the fourth pillar: construction of a determinedly national identity for all Rwandans. "Our system should not be polarizing," he said. "We should not end up entrenching the rigid identities that had been promoted in the past." (Here, as elsewhere in our discussion, he abstained from naming the suspect identities in question, which both of us understood to be 'Hutu' and 'Tutsi.')

"Taken together," he continued, "the Genocide Law and the Gacaca Law are designed to pick out the small numbers of perpetrators that society can't deal with, and then to reintegrate the rest. Indeed, the communities will actually help the culprits do the community service they need to do, in terms of rebuilding homes and farms destroyed during the genocide."

Creating the new gacaca system has been a truly massive undertaking—one in which the Rwandan government has sought to involve the country's entire adult population. The first step, in October 2001, was to elect the new gacaca 'judges.' Rwanda has for many generations been a highly hierarchical society.27 At the base of its pyramid of tight social organization are 11,000 units called 'cells' (localities or hamlets). In early October 2001, all Rwandan adults were given a day's holiday to take part in open-air meetings of their cells, at which the new gacaca judges were elected. Each cell elected nineteen gacaca judges to work at its local level, and a further six judges to work at the higher district level. The elections, which were spread over four days, thus resulted in the election of more than 250,000 judges.28

In accordance with the Gacaca Law, judges were chosen by their communities on the basis of their "high moral integrity," and had to be free from any suspicion of participation in the genocide. Their qualifications notably did not include any formal legal or administrative training; indeed, many were illiterate. The next step, which was scheduled to start in February 2002, was to provide these judges with some training. In December 2001, Aloysie Cyanzayire, the head of the Supreme Court's Gacaca Department, announced a plan to provide six days of training to each of the new judges. They should then, she said, be ready to start hearing cases in May 2002.29

The gacaca courts are designed to operate at four administrative levels: cell, sector, district, and province. Courts at the cell level will try only the lowest level suspects (Category 4), but they will also review the categorization of all the suspects who lived in that cell during the genocide. Category 3 suspects will go to sector-level gacaca courts and Category 2 suspects will go to district-level hearings. Province-level gacaca courts will deal only with appeals arising out of judgments at the district level. Category 1 suspects will remain in the regular court system. Only defendants in Categories 1 and 2 will have access to trained legal advisors.30

Questions of justice

Establishing the gacaca court system is an ambitious undertaking—not only because of the logistics involved, but also because of the conceptual and philosophical agility required to design this path-breaking venture and justify it to skeptical Western aid donors.

As Ambassador Sezibera admitted, the new system aims at a synthesis of the traditional gacaca gathering and a formal, Western-style court system. But it is still not clear whether the two systems can be merged in the way the Kigali government hopes. In a Western, prosecutorial system, the main focus of judicial enquiry is on individual perpetrators and on determining their responsibility for intentional acts they may have committed that contravene existing law. The concerns of traditional gacaca are very different. One researcher has written that gacaca's main objective is "not to determine guilt nor to apply state law…but to restore harmony and social order in a given society, and to re-include the person who was the source of the disorder."31 From this point of view, traditional gacaca looks similar to other 'restorative justice' systems used in numerous societies around the world. Building on an essentially communitarian view of the relationship between the individual and society, such systems pursue the restoration of harmony within the community as the main goal, rather than the examination and punishment of individual wrongdoers.32 In a restorative-justice process, the perpetrator of a harmful act will often be expected to offer some form of restitution or reparation for those harmed by her acts. But this does not constitute a punishment dictated by a governing authority. Instead, it consists of a series of relationship-repairing actions collectively agreed upon by the perpetrator, her victims, and the community.

The new gacaca courts will preserve some of this emphasis on restitution and reparation. The London-based group Penal Reform International (PRI), which has been advising Rwandan ministries on the introduction of the gacaca courts, reports that half of any term of imprisonment for Category 3 defendants (or for Category 2 defendants who confess their crimes) can served in the form of community service work. Reactions to this proposal have been mixed. The PRI researchers reported that "[t]he Tutsi population (survivors and returnees) were in general shocked when they heard that even persons who had killed could be released conditionally in this way and would therefore return as neighbors to their old communities." The report noted that the government could do a lot more to describe and explain the new system to its own people.33

These two approaches to justice, the restorative and punitive, reflect widely divergent views about the nature of justice and different visions of public good. Which is why, even if the Rwandans are able to master the logistical challenges of implementing the gacaca courts, they will still face huge political and conceptual challenges in doing so.

They face continuing problems, too, trying to explain their change of approach to Western rights activists and the Western governments on whose largesse they are still heavily dependent. They are not the first to face such problems. In 1995, when post-apartheid South Africa established the Truth and Reconciliation Commission (TRC)—the mechanism with which South Africans had decided to address the sequelae of apartheid-era violence—it too came in for criticism from Western rights activists, who were outraged that the TRC would be offering amnesties to the perpetrators of earlier atrocities. But as the South Africans designed the TRC's revolutionary procedures and defended it in the 'court' of world opinion, they had considerable cultural, social, and political resources to draw on. They rolled out such impeccable advocates of the restorative approach as TRC Chairman (and former Anglican Archbishop) Desmond Tutu, who eloquently explained that the South Africans considered the establishment of the truth about the abuses of the apartheid era as an important way-station along the path to reaching the broader goal of inter-racial reconciliation—and that in order to get at this truth, they were prepared to offer amnesties to former perpetrators who told everything they knew. Tutu pointed out that this approach was based not only on traditional Christian ideas of forgiveness, but also on indigenous African concepts like ubuntu, a communitarian view of the self that holds that "a human is human inasmuch as he recognizes the humanity of other humans." He also argued that for former perpetrators, the shame they experienced when standing in a public forum and recounting all the heinous acts they had undertaken itself constituted something like a punishment.

Tutu also suggested that there were solid, realpolitik calculations behind the willingness of African National Congress leaders to allow their former tormentors to get off virtually scot-free. He once told a group of foreign diplomats that it was "certain" that the leaders of the apartheid government "would not have agreed to a peaceful transition from repression to democracy if they had known that we were going to have Nuremberg-type trials."34 (It was generally left to people other than Tutu to point out that one of the main reasons the TRC worked as well as it did—gathering and publishing thousands of confessions that added considerably to the public's knowledge of the apartheid regime's crimes—was that the perpetrators continued to face a threat of prosecution in South Africa's formal criminal courts if they failed to tell all to the TRC.)

In 1995, soon after the Rwandan genocide, Tutu and a number of other advocates of restorative justice visited Kigali, where they urged the country's post-genocide rulers to emphasize restorative policies over the punitive ones they were intent on pursuing at the time. But that attempt at persuasion failed. It took a few more years—and the deaths of thousands in Rwanda's bloated detention camps—before the Rwandan authorities recognized that some measured use of the restorative-justice approach might indeed better serve the country's needs.

But when Rwandans have tried to defend their introduction of gacaca courts against the criticisms of Western rights advocates, they have suffered from a number of handicaps that tend to undermine their credibility. First, because they are relatively recent converts to the idea that something other than a strictly prosecutorial approach might be best for their country, their commitment to alternative strategies sometimes appears half-hearted. Second, the cultural and philosophical foundation of Rwanda's alternative solution does not seem to be as broad, as deeply embedded, and as thoroughly-considered as the foundation Tutu and his colleagues were able to adduce for the TRC. And finally, for all the Rwandan government's claims about seeking national reconciliation and ensuring respect for all citizens' human rights, their credibility on both these counts has been questioned by many of those who know the country best.

Well-founded suspicions

The sense of guilt that motivated many Western governments' policies in the post-genocide period continued, well after 1994, to color the views of Western policymakers about the Rwandan government. This government was thought not only to represent the interests of the survivors of a genocide, but also (with even more foundation), to have its origins in the only fighting force anywhere on the planet that in 1994 took concrete and successful action to suppress the Rwandan genocide—a task that, under the terms of the 1948 Genocide Convention, should have been mandatory for all the governments that had signed the Convention. (That included the United States.)

Many members of Western elites continued, long after 1994, to give the government in Kigali a fairly easy ride, even when it committed rights abuses that elsewhere would have been a cause for considerable alarm. "Until now," Belgian researcher Peter Uvin wrote in early 2001, "the international community has by and large closed its eyes to [Rwanda's] human rights abuses, or attempted to justify them."35 The temptation to do this has certainly been increased by the existence of a still-vocal cadre of propagandists in many parts of the (Hutu) Rwandan diaspora in Africa and Europe, who either deny the facts of the genocide altogether or try to lessen Hutu culpability by claiming that what happened in 1994 was only one half of a 'double genocide,'(the other half being a program of genocide directed against the Hutus by Tutsis and others in the present Rwandan government).

The heated rhetoric of this debate makes it necessary to double-check facts and to be cautious when endorsing accusations against any party. Nevertheless, a careful reading of reports on the Rwandan government by reputable human-rights organizations, by aid groups, and by UN-sponsored teams of investigators, reveals a chronic pattern of serious rights abuse. These abuses have been committed both inside Rwanda and in Rwanda's vast neighbor to the West, the Democratic Republic of Congo (DRC)—formerly Zaire—where the Rwandan army has carried out multiple operations of varying intensity since 1996.

Most Western researchers who write about these matters maintain a careful distinction between the actions carried out since 1994 by the RPF-led government and its armed forces—which on occasion they may refer to as "massacres," "war crimes," or even "crimes against humanity"—and the anti-Tutsi killings of 1994, which, all now agree, clearly constituted genocide. The distinction is important, because (as President Clinton and his staff recognized in 1994) a finding of genocide places an immediate obligation on signatories to the Genocide Convention to intervene to suppress it. An accusation or finding of genocide also has great symbolic importance. In the opening statement that Carla del Ponte, Chief Prosecutor of both ICTR and ICTY, made in the trial of Slobodan Milosevic, she referred to genocide as, "the crime of crimes." Certainly, genocide has always justly occupied a central place in the crimography of the post-Shoah West.

But massacres, war crimes, and crimes against humanity are bad enough.36

Gérard Prunier is a French researcher who has studied Rwanda for many years and has written one of the best accounts of the origins and events of the genocide.37 In June 2001, I visited his attic apartment in Paris and listened to his extended and passionate analysis of Rwanda's tangled internal politics. Like most other specialists on the country, Prunier drew a distinction between the anti-Tutsi genocide of 1994 and the series of anti-Hutu massacres that he attributed to the present government in Kigali. He estimated that between 120,000 and 150,000 people had been killed in anti-Hutu massacres committed by government forces inside Rwanda since 1994; that more than 200,000 Hutus disappeared in eastern DRC (then Zaire) during massacres carried out by the Rwandan army there in 1996; and that the Rwandan government was also implicated in the disappearance of an additional 300,000 Hutus in the DRC in the months after its Congolese ally, Laurent Kabila, came to power there in 1997.

These events in Congo have formed an important component of the policy of the post-genocide Rwandan government (and indeed, they have caused some Western governments to start criticizing Kigali more openly). A UN report published in June 1998 described the killings of 1996 and 1997 as having constituted "crimes against humanity," and possibly also "genocide," depending on the intentions of those who carried them out.38 (Intentionality is a doubly key component of any finding of genocide, which requires not only that the perpetrator intended to commit crimes such as killing, maiming, or starving his victims, but also that his intention in committing them was, "to destroy, in whole or in part, a national, ethnical, racial or religious group, as such.")39

The consequences of the state failure that the DRC experienced in the late 1990s have been truly horrifying. An estimated 3 million people have died there since 1996 because of what Prunier describes as the "continental war" being waged inside the country. The Rwandan government, like many other African governments, has been deeply involved. Since 1998, the Rwandan army has maintained a brutal military occupation over an area in eastern DRC that is sixteen times larger than Rwanda itself.40

Prunier used blunt language to explain the Kigali government's actions in the DRC : "Especially after the genocide, the Tutsis became crazy as a group." This craziness, he suggested, stemmed from the terrifying existential crisis faced by survivors of genocide:

If a single person dies, or has a close brush with death, that's one thing. But if you're a survivor of genocide, then you know that you came so close to a situation where not only your self, but also everything that constitutes your world, everything that makes your life worth living—your work, your family, your children; everything that might provide a vehicle for your legacy in the world—all that was on the point of being wiped out, too. Of course it makes you crazy.

   Yes, the Tutsis are crazy as a group. Massacres will get them nowhere. They are bound to lose….There are 7 million Hutus in Rwanda, and at most a million Tutsis. It is a system of minority rule, like apartheid.

   Can you imagine how crazy the white South Africans would have become if the blacks had killed 800,000 of them?

This was the context in which Prunier viewed the Rwandan government's long-term detention of 125,000 Hutus accused of involvement in the genocide. "The government doesn't care if they are guilty or not. They are jailed in order to keep pressure on the Hutus. That's all."41

He scoffed at the idea that any criminal-court proceeding, either domestically or in the form of ICTR, could ever have been expected to deal adequately with the repercussions of the genocide. "To be successful, the action should have been swift and political—and it would have had to violate all our understandings of international law," he said. "They should have captured the top leaders of the genocide, and hanged them quickly and visibly. That might have worked."

A few days later, I asked Filip Reyntjens what he thought of Prunier's theory that perhaps the detainees inside Rwanda had been held merely as political hostages. "Just those 125,000?" he exclaimed. "The government is holding seven and a half million hostages in Rwanda. The entire population."42

In the United States, the general level of public interest in Rwandan affairs is considerably lower than it is in Belgium or France. Perhaps that is understandable. Americans were shocked by the images that came out of Rwanda in 1994. Many of us felt uncomfortable that our government had not done more to prevent or suppress the genocide; we gained some comfort from the impression that the country seemed to provide a suitable test-case for the UN's second ad-hoc war-crimes court; and then we moved on. Certainly, very few Americans in or out of government have seemed to pay much heed to the tumultuous social and political conflicts that have continued to roil the country since 1994. This extremely unsettled political situation has, however, had a far greater impact on the well-being of Rwanda's people and their neighbors than any number of "historic" rulings emanating from the Arusha court, or from Rwanda's own judicial system. Indeed, the country's political instability has made it virtually impossible for either of these court systems to operate with any reasonable level of integrity or efficiency.

Too many Americans, I fear, have remained fixated on assessing the minutiae of the workings of the judicial institutions. And meanwhile, in the collines of Rwanda and the forests of the DRC, hundreds of thousands of people have continued to die.

Alternatives to criminal justice?

Any possible solution to the problems faced by Rwanda will raise serious ethical questions. The current policies of the Rwandan government underscore the depth of this dilemma. But it seems likely that the country's problems have been aggravated and its options unnecessarily limited by the persistence and single-mindedness with which most of the international community focused its approach to the legacies of the genocide on a criminal justice strategy. It seems probable, indeed, that this strategy misidentified central aspects of what was happening in Rwanda during those ten bloody weeks in 1994.

The criminal justice system, after all, deals with intentional acts provably undertaken by individuals considered as rational agents. With regard to the frenzy of killing in Rwanda, however, it is virtually impossible to assign individual responsibility for the innumerable acts that caused the deaths of roughly 800,000 Tutsis. In the period leading up to April 1994, some people started publicly advancing the view that the Tutsis should "be taught a lesson," or even "be killed." Some people made plans to bring this about: they imported the guns and the machetes, and drew up lengthy lists of targets. Then, once the killing started, some people directed it from the center, with continued incitement and administrative action. Some people staffed the roadblocks or went to find the intended victims at their listed addresses. Some people wielded the machetes—with some of these taking pleasure in prolonging the torture of the doomed Tutsis before killing them while others reportedly acted only as "reluctant killers." Some raped Tutsi women before killing them or took them as sex-slaves. Some rounded up reluctant participants for the "work." Some brought food and beer for the killers and shouted their encouragement. Some people organized prison inmates into "clean-up crews" and forced them to move, bury, or otherwise dispose of the bodies. Some people burned down the houses of the Tutsis to erase any trace of them. Some people simply moved into the homes of dead Tutsis or stole their cows or other belongings….

How can anyone start to assign a specific amount of responsibility to each one of the hundreds of thousands—probably millions—of people who took part in one or more of these ways? Moreover, how does one identify the role of rationality and intention in the behavior of most of the grassroots participants in the blood orgy of those terrifyingly irrational weeks? How, in short, can one apply conventional methods of legal judgment to the great mass of the Hutu people who were caught up in a profoundly abnormal reality?

What happened in Rwanda was a human tragedy of unimaginable dimensions. It was a collective frenzy of hatred and killing in which the bonds of community that normally are directed toward the sustenance of human life were directed instead toward the extermination of an entire social caste. Millions of men and women lost, for some period of time, all recognition of the humanity of their victims, and even of their own self-respect and humanity. It was a tragedy that made reconnecting the bonds of community—not only across the caste-line, but also within each caste—nearly impossible. The Rwandan genocide clearly left the survivors deeply scarred.43 But it left participants scarred as well.

How could normal life be rebuilt after a tragedy that perverted the very concept of normality?

Humankind has faced such issues before. In the aftermath of the Holocaust, the questions of how best to characterize those events and how to deal with their legacy were the focus of intense intellectual debate. These questions formed a major topic in the correspondence between Hannah Arendt and her former professor at Heidelberg University, Karl Jaspers, who had remained in Heidelberg throughout the war.

In August 1946, at the close of the Nuremberg trial of two dozen ringleaders of the Nazi atrocities, Arendt wrote to Jaspers:

Your definition of Nazi policy as a crime…strikes me as questionable. The Nazi crimes, it seems to me, explode the limits of the law; and that is precisely what constitutes their monstrousness. For these crimes, no punishment is severe enough….[T]his guilt, in contrast to all criminal guilt, oversteps and shatters any and all legal systems….We are simply not equipped to deal, on a human, political level, with a guilt that is beyond crime and an innocence that is beyond goodness or virtue….I don't know how we will ever get out of it, for the Germans are burdened now with thousands or tens of thousands or hundreds of thousands of people who cannot be adequately punished within the legal system; and we Jews are burdened with millions of innocents, by reason of which every Jew alive today can see himself as innocence personified.44

That October—just three days after the executioners at Nuremberg hanged the eleven men condemned by the Allied judges—Jaspers wrote back:

I'm not altogether comfortable with your view, because a guilt that goes beyond all criminal guilt inevitably takes on a streak of 'greatness'—of satanic greatness— which is, for me, as inappropriate for the Nazis as all the talk about the 'demonic' element in Hitler and so forth. It seems to me that we have to see these things in their total banality, in their prosaic triviality, because that's what really characterizes them….I regard any hint of myth and legend with horror, and everything unspecific is just such a hint.45

Karl Jaspers was a fair-minded, conscientious man who had stuck by his humanistic principles (and his German-Jewish wife) throughout the twelve years of Nazi rule, and had suffered considerably for this exercise of love and conscience. It is understandable that in 1946, he should still be particularly sensitive to any kind of reasoning that might revive any notion of the Third Reich's "satanic greatness"—an idea that had often been promulgated by pro-Nazi intellectuals during the war. In December 1946, Arendt wrote that she found Jasper's criticism on this point "half convincing; that is, I realize completely that in the way I've expressed this up to now I come dangerously close to that 'satanic greatness' that I, like you, totally reject." But she suggested that, even without subscribing to such a myth, there was still, "a difference between a man who sets out to murder his old aunt and people who without considering the economic usefulness of their actions at all…built factories to produce corpses." She suggested another possible way of characterizing this difference: "Perhaps what is behind it all is only that individual human beings did not kill other individual human beings for human reasons, but that an organized attempt was made to eradicate the concept of the human being."46

In 1960, Israel captured Adolph Eichmann and announced that it would put him on trial for "crimes against the Jewish people" and "crimes against humanity." Arendt and Jaspers re-opened their discussion of whether a criminal proceeding was the best way to deal with the phenomenon of the Holocaust. "I have this foolishly simplistic idea," Jaspers wrote to her. "It would be wonderful to do without the trial altogether and make it instead into a process of examination and clarification. The goal would be the best possible objectification of the historical facts. The end result would not be the judges' sentence, but certainty about the facts, to the extent that such certainty can be attained."

This might seem like an early proposal for something like a truth and reconciliation commission, or at least a truth commission. (Jaspers did, however, go on to suggest that after having established "the facts," the Israeli body should pass Eichmann on to an international court to undergo a formal trial for "crimes against humanity.") At this point, Jaspers seems to have been generally convinced by Arendt's arguments that what had happened during the Holocaust could not be dealt with solely through criminal prosecutions. He also stressed that political factors would inevitably underlie the decision to hold trials:

Judgments passed by the victors on the vanquished have, in the past, been regarded as political actions and distinct from legal ones….The political realm is of an importance that cannot be captured in legal terms (the attempt to do so is Anglo-Saxon and a self-deception that masks a basic fact in the functionings of political existence).47

In her response, Arendt admitted that

as far as the role of the law is concerned, I have been infected by the Anglo-Saxon influence. But quite apart from that, it seems to me to be in the nature of this case that we have no tools to hand except legal ones with which we have to judge and pass sentence on something that cannot even be adequately represented either in legal terms or in political terms.48

Soon enough, she was on her way to Jerusalem, where she reported on the Eichmann trial for The New Yorker in a series of pieces that subsequently became her book, Eichmann in Jerusalem: A Report on the Banality of Evil. Despite this reference to Jaspers' notion of the 'banality of evil,' Arendt seems to argue in the book that what happened during the Holocaust was not merely the commission of 'banal' (that is, justiciable) evil, but went beyond that. In the Postscript she wrote, "I held and hold the opinion that this trial had to take place in the interests of justice and nothing else." But despite that, she concluded that many aspects of the trial itself demonstrated "the inadequacy of the prevailing legal system and of current juridical concepts to deal with the facts of administrative massacres organized by the state apparatus."49

In Jerusalem, as in Nuremberg, Arendt (and Jaspers) judged that criminal prosecutions might be a necessary element in a response to state-sponsored atrocities—but they could not be sufficient. In Rwanda, in our current era, the distinctive features of the Rwandan genocide made an exclusive reliance on legal methods even less appropriate than it was in the case of the Holocaust.

Crimes—or sickness—or both?

In both the Rwandan genocide and the Shoah, a state-supported set of organizations committed a centrally coordinated series of actions that were intended to destroy an entire group of people.

One of the chief aspects of the 1994 genocide that differentiated it from the Shoah, however, was the collective frenzy of the killings and the sheer number of the people directly involved in them. Arendt and Jaspers both recognized the limited ability of standard criminal justice procedures to satisfactorily address the legacy of the Shoah's highly professionalized and industrialized mass murder. This assessment suggests that in Rwanda (but also, perhaps, in the DRC? Sierra Leone? Afghanistan? the Balkans?), where the mass-participatory nature of the killing makes individual accountability even harder to disentangle, the prosecutorial method may need even more help from other disciplines and traditions.

The realm of politics and diplomacy is one evident place to look. Though some rights activists view the peace-brokering efforts of diplomats with barely disguised scorn, bringing about a solid, workable peace can often halt atrocities and lay the groundwork for future reconciliation.

Consider Mozambique, where the mix of policies adopted in 1992 to deal with the aftermath of a punishing civil war provides a complete contrast with the prosecutions-focused approach pursued by post-1994 Rwanda. At one level, the peace agreement that brought an end to Mozambique's seventeen years of atrocity-laden conflict was a classic, diplomat-driven deal in which all participants in the war were given complete amnesty for acts committed during the war. It is important not to gloss over the severity of those acts. During the war, between 600,000 and 1 million of the country's 16 million people lost their lives. A Human Rights Watch report published in July 1992 concluded that, "The war has involved widespread violence against civilians, including both the systematic use of mutilations and killings and indiscriminate violence during sweeps through contested areas… Such incidents appear to be a central part of Renamo's strategy of…terrorizing the local population and the opposing soldiers. Government soldiers are also responsible for acts of mutilation."50 (Renamo was the rebel faction that had been backed by the apartheid regime in neighboring South Africa and at some points, by the Reagan administration.)

The main policy recommendation in the HRW report was, "To bring to trial those members of the armed forces, on both sides, who are primarily responsible for gross abuses of human rights during the war." But the diplomats who were already working to negotiate the country's peace deal paid this prescription little heed. Soon after the report was published, they announced the conclusion of a final peace agreement. And days after that, the country's parliament passed legislation permitting the blanket amnesty.

Mozambique's peace agreement stipulated that most members of the fighting forces, from both sides, would be demobilized, and those who remained in service would be melded into a new national army. All Mozambicans, including demobilized combatants and the many displaced noncombatants, were encouraged to return to their home communities. Renamo would become a political party and would compete peacefully in post-demobilization elections. The United Nations and Western aid donors underwrote much of the process. During the next few years, aid donors invested $80 million to help demobilize and reintegrate some 93,000 former combatants.51

Those demobilized combatants doubtless included many who had committed the atrocities documented by Human Rights Watch. But in 1992, before the Milosevic prosecution and the ICTY, the international community had not yet acquired its present enthusiasm for war-crimes courts. Thus, in Mozambique, the UN had few qualms about supporting a peace agreement that allowed for a blanket amnesty and even for the smooth reintegration into civilian society of individuals who were clearly responsible (according to Western judicial norms) for atrocities. In 1994, and again in 1999, longstanding Renamo leader Afonso Dhlakama led his party in nationwide elections—and ended up winning nearly 50 percent of the vote.

In addition to the UN's permissive attitude toward perpetrators of atrocities, what was also notable about the peace agreement in Mozambique was what happened after it in the communities to which the former 'war criminals' and their victims all returned. What came to the aid of these communities was not criminal courts and not Western aid donors, but a set of cultural traditions: to be precise, a whole range of indigenous cosmologies that allowed for the swift healing of deep personal and communal wounds. These cosmologies, roughly speaking, present the essence of the human condition as constituted by the set of relationships that the individual sustains with his family, the broader community, and a spirit world that includes his ancestors.52 According to these cosmologies, violence is a clear aberration in the condition of being human. It is not something that humans can be considered personally responsible for. Rather, it is an extra-human force—generally considered as a malevolent, spiritual force—that possesses humans, and from whose clutches they must be saved.

Only a few outside researchers have ever described the kinds of rituals that Mozambican communities and their curandeiros (traditional healers) used to heal the many wounds of the civil war. But it seems clear that as individual sons or daughters, or whole families, came back to their home communities from the battle-front, from exile, or from internal displacement, their communities relied on rituals whose central aims were to "take the violence out of the person" and to facilitate reintegration into the community.53 These rituals were held for all those who had been exposed to atrocious violence, perpetrators and survivors alike. In general, the curandeiros seem not to have made too much of a distinction between members of these two groups. Former perpetrators were seen, perhaps with some justification, as also being survivors of the violence, along with those other survivors who had never been perpetrators. It was agreed that all of the affetados (those affected by violence) needed to be purified from its effects and ritually welcomed back into the state of being fully human—in productive community with other humans.

These Mozambican cosmologies are, of course, very different from the highly individualized and punitive cosmologies that motivate Western-style justice systems. In Mozambique, however, the combination of amnesties and community-based healing seems to have worked. Certainly, it has been remarkably effective in ending the cycle of atrocious violence in which, until 1992, the country's people seemed so inextricably trapped.

In the future, it may be useful to compare the outcome in Mozambique with outcomes in the countries that have been subject to UN-mandated international criminal courts: Rwanda and former Yugoslavia. Such a comparison might include, too, the costs of different efforts to the international community: a $300 million investment in ICTR through the end of 2000 versus an $80 million investment in the demobilization and reintegration effort in Mozambique.54

Whatever the final outcome of that comparison, the Mozambican example suggests that in the aftermath of atrocious violence all those affected by it are in need of some form of psycho-social healing. Drawing this conclusion does not depend on a romantic glorification of "traditional" cultures or a grandiose psychological theory. It is based, rather, on a lesson already well known to those who have worked with Vietnam veterans or their predecessors from earlier wars. Were such veterans perpetrators of violence or victims of it? Or rather, did every one of them lie somewhere along a spectrum in which all were, in some sense, victims, but many were also, to varying degrees, perpetrators? If we admit the latter possibility with respect to those members of our own society whom our government has trained to kill, how can we try to impose on other agents of violence elsewhere in the world a prosecutorial system that simplistically divides humankind into victims and perpetrators?

With respect to Rwanda, I believe we can identify social psychoses as having made a crucial contribution at two different stages. In the first stage, the planning and commission of the 1994 killings can be seen as symptoms of a particular kind of social psychosis that had a truly genocidal intent. The collective frenzy of the street-level perpetrators has already been remarked. But how much 'sicker' than they were those people who coolly planned and prepared for all those events in advance? Then, at the second stage, after the broad suppression of the genocide, at least three different kinds of psychosis came into play and helped frame the continuing nightmare of post-1994 Rwanda. There was, first, the mental trauma of Tutsis who directly survived the killings. Second, there was the trauma of the Tutsis who were 'vicarious' survivors. For many of these individuals, returning to their homes must have involved witnessing and attempting to handle—practically and psychologically—the thousands of Tutsi corpses that still lay unburied or that clogged wells and latrines, the devastated homes and community infrastructure, and the challenge of encountering Hutus wherever they went.

But we should not forget the traumas that the perpetrators of the genocide carried with them after it ended. Some dedicated psychologists who have done trauma-recovery work in Rwanda in recent years are frank about this component of post-genocide trauma. Laurie Pearlman works with the Healing through Connection and Understanding Project, which has operated in Rwanda since 1999. "Perpetrators too may experience traumatic stress," she writes. "Killing other people or participating in violence in other ways is unnatural and dehumanizing behavior. Those who did so in Rwanda may have symptoms of trauma."55

Most people in the human-rights community do not like to think about former perpetrators as being, in their own way, subject to post-atrocity trauma: to do so would make it harder to advocate a policy of widespread prosecutions. Where rights activists admit this possibility at all, it is usually only with regard to former killers who belong to the group that Westerners like to endow with a unique capacity for innocence: children. But why killing or mutilating another person at, say, age fourteen should be considered a potentially traumatizing experience, while committing the same actions at age twenty-one is not, has never been fully explained. Far better, surely, to suppose that for just about anyone, at any age, having committed such acts leaves a legacy of dehumanization and trauma.

But where does this leave us? Am I attempting to undermine norms of individual responsibility and individual accountability to such an extent that the entire edifice of Western criminal justice collapses? No, my intention is simply to assert that we need a distinction between policy responses to events in more settled times and those adopted after times of atrocity.

In this latter context there may be something to be said for the policy response that Prunier imagined—a limited number of more or less summary executions—over the tortuous and deeply divisive path that the UN and the Rwandan government adopted instead. But we should also, I think, conclude that punishment should be only a small (or even non-existent) part of the response in such circumstances. What is far more important is a generous and thoughtful set of policy responses that facilitate healing at the many levels where healing is urgently needed: for traumatized individuals, between families, between social groups, between a country and its neighbors. Or, between Western countries that bear guilt for egregious sins of omission and commission and the rest of the international community.

So what role is left for punishment? A small but significant role, perhaps. For even if we view the vast body of rank-and-file 'perpetrators' as people who, after atrocities, also need healing themselves, that still leaves a possible role for punishment at the highest levels of leadership. In policy terms, we can say that after an event like the Rwandan genocide, as after the Shoah, there is an urgent need to dismantle any capabilities the genocidal organizations might have to re-group and shed more blood. Such incapacitation may be best achieved through a broad re-education effort aimed at the mass of street-level perpetrators and condoners—combined, perhaps, with a swift 'decapitation' effort aimed at the ringleaders and preferably carried out through imprisonment. (Note, though, that no such decapitation was undertaken with regard to the Renamo leadership; yet even without it the process seemed to work.)

In the cosmological terms which are no less important in this process, we could perhaps agree with Hannah Arendt, Karl Jaspers, and the Mozambican curandeiros that a time of mass killings is not a normal time in human history. The curandeiros (like Archbishop Tutu and many other people from around the world) would argue that the normal human condition is to be in healthy relationships with other humans. So perhaps we could conclude that the worst thing the ringleaders of atrocities have done—the thing that may justify us treating them differently from the great mass of perpetrators—is that they have subverted the bonds of human society by twisting them to distinctly inhuman ends.

Or, indeed, as Hannah Arendt put it, that they have sought "to eradicate the concept of the human being." <

Helena Cobban
is global affairs columnist for The Christian Science Monitor and Al-Hayat (London), and a member of the Middle East advisory committee of Human Rights Watch.


1 See Priscilla Hayner, Unspeakable Truths: Confronting State Terror and Atrocity (New York and London: Routledge, 2001), 207–8. In May 2001, ICTY president Claude Jorda retreated somewhat from this position, though he still strongly asserted the primacy of ICTY's jurisdiction. See ICTY press release #591 of 17 May 2001, [20 March 2002].

2 Testimony of Pierre Simon Hitiyisa, 23, in African Rights, Rwanda: Death, Despair and Defiance, revised ed., (London: African Rights, August 1995), 567–8. Another excellent collection of testimonies is Alison Des Forges, "Leave None to Tell the Story": Genocide in Rwanda (New York, London, and Brussels: Human Rights Watch, 1999).

3 I have used the term "caste" to describe the Tutsi-Hutu distinction after much consideration. Members of both groups speak the same language (Kinyarwanda), practise the same set of religions (mainly Catholic Christianity), and participate in nearly all the same social practices. The main distinction between them was, historically, one of occupation.

4 Testimony of Jean Bosco Bugingo; in African Rights, Death, Despair and Defiance, 998.

5Ibid., 999.

6 For numerous accounts of such acts of bravery, see Ibid., 1024–61; also see André Sibomana, Hope for Rwanda, trans. Carina Tertsakian (London and Sterling, Va.: Pluto Press, and Dar es Salaam: Mkuki na Nyota Publishers, 1997), 68–9, 104–5.

7 According to some accounts, the number of Tutsi 'returnees' nearly equaled the number of those killed in the genocide. It was significant for future political relations inside the country that many of the returnees preferred English as their 'main' foreign language, rather than the French that previously dominated inside Rwanda, and that many of the returnees spoke the languages of neighboring African countries (where they had spent decades of exile) better than they spoke the national language, Kinyarwanda.

8 The statute that established ICTY is Security Council Resolution 827, 25 May 1993, [12 March 2002].

9 Samantha Power has written a detailed account of the Clinton administration's actions (and decisions to refrain from action) throughout the genocide. She records how, in addition to refusing to send American forces to beef up the small UN presence already in the country, the administration refused to try to persuade other governments to send peacekeepers—and indeed, worked hard at the Security Council to persuade Council members to dismantle the peacekeeping force completely. See Samantha Power, "Bystanders to genocide," Atlantic Monthly, September 2001.

10 For an excellent description of the issues facing aid workers and the rest of the international community after the genocide see Ian Martin, "Hard choices after genocide: Human rights and political failures in Rwanda," in Hard Choices: Moral Dilemmas in Humanitarian Intervention, ed. Jonathan Moore (Lanham, Md., and Oxford: Rowman & Littlefield, 1998). Martin, previously the secretary-general of Amnesty International, worked as chief of the UN's Human Rights Field Operation in Rwanda after the genocide.

11 They also objected to many specifics of the new court's procedure, including the fact that—in line with emerging international norms regarding criminal justice—it did not allow for any death penalty sentences for those convicted of even the most heinous crimes.

12 An excellent account of the role of post-genocide guilt in determining the policies Western governments adopted toward Rwanda in the late 1990s is Peter Uvin, "Ethics and the new post-conflict agenda: The international community in Rwanda after the genocide," Third World Quarterly 22 (2001).

13 International Crisis Group, International Criminal Tribunal for Rwanda: Justice Delayed (Nairobi, Arusha, and Brussels: International Crisis Group, 2001), ii.

14 International Crisis Group, Justice Delayed,18. The statute that established ICTR was UN Security Council resolution 955, 8 November 1994, [12 March 2002].

15 Sibomana, Hope for Rwanda, 110–11.

16 See position.htm [18 March 2002].

17 The huge number of suspects was, of course, a consequence of the genocide leaders' pursuit of the widest possible Hutu participation in the killing of Tutsis.

18 Sibomana, Hope for Rwanda, 108–9. Scott Peterson has given a similar description of a visit he made to Gitarama in mid-1995. See Scott Peterson, Me Against my Brother: At War in Somalia, Sudan, and Rwanda (New York and London: Routledge, 2000), 318. By the time Philip Gourevitch visited in mid-1996, conditions had gotten a little better, but apparently not much. See Philip Gourevitch, We wish to inform you that tomorrow we will be killed with our families: Stories from Rwanda (New York: Farrar, Straus and Giroux, 1998), 246–8.

19 Sibomana, Hope for Rwanda, 109–10.

20 Human Rights Watch, World Report 2002 (New York: Human Rights Watch, 2002), 79. The report covers events in 2001.

21 Gourevitch, We wish to inform you, 249. One of the unfortunate weaknesses of Gourevitch's book is its a-historicity: he doesn't clearly indicate when many statements by Kagame and others were made.

22 Amnesty International, "Rwanda: The troubled course of justice," 26 April 2000, <\rwanda> [8 March 2002]. The report fails to specify how, with the resources available, Rwanda could be expected to provide the elements of 'fair trials' that the report specified, such as legal representation for all defendants.

23 Gahima, quoted in International Crisis Group, Justice Delayed, 36.

24 Filip Reyntjens, "Le gacaca ou la justice du gazon au Rwanda," in Politique Africaine 40 (1990): 31–41.

25 It is in this category of accused persons that there has been some competition between the Rwandan courts and ICTR for jurisdiction. See, for example, International Crisis Group, Justice Delayed, 19.

26 A body called a "prison gacaca commission" was reportedly established inside Kigali Prison (and ostensibly other prisons), as early as late 1999. By October 2001, the commission in Kigali had reportedly gathered confessions from 1,127 of the prison's 6,000 genocide suspects. The confessions were then handed over to the Prosecutor's office. See Foundation Hirondelle, "Rwandan genocide suspects welcome new system," 7 October 2001. Hirondelle, a Swiss-financed venture reporting on Central Africa and other conflict zones, has a section on its web site, , devoted to "Rwanda/justice" where this and other articles can be found.

27 That same fact made the centralized organization and coordination of the genocide much easier than it would have been in less well-organized societies.

28 Participation in the elections was estimated by organizers at more than 90 percent. For a good eye-witness description of an election meeting, see Julia Crawford, "Women take centre stage in election of 'people's judges,'" 5 October 2001, Foundation Hirondelle, [8 March 2002].

29 See "First 'gacaca' genocide trials to start in May," 10 December 2001, Foundation Hirondelle, [8 March 2002].

30 Information from ibid., and from interview with Amb. Sezibera, June 2001.

31 Stef Vandeginste, cited in Klaas de Jonge, Interim report on research on Gacaca jurisdictions and its preparations (July–December 2001) (Kigali: Penal Reform International, January 2002), 8–9.

32 For a basic exploration of these issues, see Wesley Cragg, The Practice of Punishment: Towards a Theory of Restorative Justice (London and New York: Routledge, 1992). For an account of an illuminating debate between advocates of the two different approaches, see chapter 7 of Helena Cobban, The Moral Architecture of World Peace: Nobel Laureates Discuss our Global Future (Charlottesville, Va.: University Press of Virginia, 2000).

33 Klaas de Jonge, Interim report, 26–7. The report also suggested that those convicted would be able to count time served before the hearings against their terms of imprisonment.

34 "Tutu says amnesty process exposed previous regime," South African Press Association, 23 June 1998, 141. Reprinted in Foreign Broadcasts Information Service, FBIS-AFR-98-174.

35 Uvin, "Ethics and the new post-conflict agenda." Uvin also noted that the issue of how to view Rwanda's actions had, "become a source of deep divisions between and within the human rights, humanitarian, and foreign policy circles of the West."

36 Just as the term "genocide" was coined by a Jewish American (Raphael Lemkin) to describe the atrocities committed by the Nazis in Europe, the term 'crimes against humanity' seems to have been coined—regarding atrocities committed under Belgian auspices against the native peoples of eastern Congo in the 1880s and 1890s—by an African-American preacher and reformer called George Washington Williams. See Adam Hochschild, King Leopold's Ghost: A Story of Greed, Terror, and Heroism in Colonial Africa (Boston and New York: Houghton Mifflin, 1998), 112.

37 Gérard Prunier, The Rwanda Crisis: History of a Genocide, 2nd ed., (New York: Columbia University Press, 1997).

38 See UN Security Council, Report of the Secretary-General's Investigative Team charged with investigating serious violations of human rights and international humanitarian law in the Democratic Republic of Congo, 29 June 1998, S/1998/581, p. 2. See also Filip Reyntjens, "Briefing: The Democratic Republic of Congo," African Affairs 100 (2001): 311–7.

39 Article 2 of the 1948 Genocide Convention.

40 For some details of this brutality, see Amnesty International, "Rwandese-controlled eastern DRC: Devastating human toll," AFR 62/015/2001, 19 June 2001, Amnesty International On-line, [1 April 2002].

41 André Sibomana expressed a similar view in Hope for Rwanda, 105–7.

42 Despite the evident skepticism with which he viewed the political record and ongoing actions of the Kigali government, Reyntjens still expressed qualified support for the move to gacaca courts. "There is no possibility of a 'perfect' policy choice here," he said in the interview. "And on balance, between the imperfect choices available, I think we should support gacaca." Peter Uvin, who also largely supports the move to gacaca courts, has produced a systematic description of the risks involved. See Peter Uvin, "The introduction of a modernized gacaca for judging suspects of participation in the genocide and the massacres of 1994 in Rwanda," (discussion paper produced for the Belgian Secretary of State for development Cooperation, 2001).

43 Here, it is probably important to include not only the 'direct' survivors of the killings, but also the 'vicarious' survivors: those Tutsis who were still in exile throughout most of the genocide but lost family members. It was returned exiles, rather than 'direct' survivors of the genocide, who dominated decision-making after July 1994.

44 Hannah Arendt, letter to Karl Jaspers of 17 August 1946, in Hannah Arendt and Karl Jaspers, Hannah Arendt Karl Jaspers: Correspondence, 1926–1969 (New York: Harcourt Brace Jovanovich, 1992), 54.

45 Jaspers, letter to Arendt of 19 October 1946, in Correspondence, 62.

46 Arendt, letter to Jaspers of 17 December 1946, in Correspondence, 69, my emphasis.

47 Jaspers, letter to Arendt of 16 December 1960, in Correspondence, 413.

48 Arendt, letter to Jaspers of 23 December 1960, in Correspondence, 417.

49 Hannah Arendt, Eichmann in Jerusalem: A Report on the Banality of Evil (Harmondworth, England: Penguin Books, 1963), 186, 294.

50 Africa Watch, Conspicuous Destruction: War, Famine and the Reform Process in Mozambique (New York: Human Rights Watch, 1992), 3.

51 Kees Kingma, Demobilization in Sub-Saharan Africa: The Development and Security Impacts (New York: St. Martin's Press, 2000), 218, 219.

52 Mozambique has many cultural continuities with parts of neighboring South Africa, so it is not surprising that these cosmologies have much in common with ubuntu tradition described by Archbishop Tutu.

53 Some good descriptions of such rituals can be found in Carolyn Nordstrom, A Different Kind of War Story (Philadelphia: University of Pennsylvania Press, 1997); and in Kees Kingma, Demobilization, 197–200. Bóia Efraime Júnior and Alcinda Honwana are two excellent Mozambican researchers on this topic, but their work is harder to find in the United States.

54 It is difficult to ascertain the actual levels spent on either ICTR or ICTY from public records. I have tried to make a conservative estimate. Of course, the parameters of a comprehensive cost-benefit analysis should probably be broader than those used here. We might add in the 'cost' of the numerous lives lost or blighted in Rwanda (and at Rwandan hands in the DRC), and in former Yugoslavia, due to failures at the political-diplomatic level and the misapplication of prosecutorial methods. Or we might add in the costs of sustaining the international military and humanitarian efforts in former Yugoslavia from 1992 to the present, or the humanitarian aid that has been generously pumped into Rwanda since 1994. (The spending continues in both countries, at a rate far, far higher than that in generally peaceful Mozambique.)

55 Laurie Anne Pearlman, "Psychological trauma," %7egubin/rwanda/lec4.htm[12 March 2002], lecture notes.

Originally published in the April/May 2002 issue of Boston Review

Copyright Boston Review, 1993–2005. All rights reserved. Please do not reproduce without permission.

 | home | new democracy forum | fiction, film, poetry | archives | masthead | subscribe |