Military history


Political Reform after Genocide

I BEGAN this book with a discussion on the need to distinguish between three different kinds of identities: cultural, market based, and political. By highlighting Hutu and Tutsi as political identities, I distanced myself from alternate views that see Hutu and Tutsi as either cultural or market-based identities. Both points of view tend to gloss over the existence of specifically political identities: one does it by subordinating the political to the cultural, the other by presuming that political identities are nothing but a version of market-based identities.

I pointed out that prominent among those who see culture as the basis of politics are the core theorists of the nation-state. For nation-state theorists, the “self” in the notion of self-determination is a cultural self. This core premise came to be shared by most theorists across the ideological spectrum of nineteenth-century Europe, from Max Weber to V. I. Lenin. The doctrine of self-determination proclaims the right of every “nation”—defined variously as a group with a common culture—to its own state. It translates cultural into political identities, innocently and unproblematically. Where the identities are explicitly political, as with Hutu and Tutsi, it assumes them to be cultural.

Those who still think consistently along “nation-state” lines call for a separation of present-day Rwanda into two political entities, one a Hutuland, the other a Tutsiland—or some version of this proposal.1 The proposal will not solve the problem of a minority in the context of a nation-state. It will only ensure a balance by reproducing the problem. Since it will produce two states out of one—a Hutuland with a Tutsi minority and a Tutsiland with a Hutu minority—it will allow each identity to be a permanent majority in one state while being held hostage as a permanent minority in the other. The idea is that the majority which wishes fair treatment for its cultural brethren who are a minority in another state will have no choice but to treat fairly cultural others who are a minority within its own borders. By institutionalizing Hutu and Tutsi as political identities in the state, the solution makes permanent the civil war between them.

Those who would give political primacy to market-based identities also think of the state as a state of the majority or minority, except that in this case the majority or the minority in question is identified with classes gelled through market-based relationships or, occasionally, with occupational groups such as pastoralists or agriculturalists. As with nation-state theorists, the market-state theorists too do not see political majorities and minorities as outcomes of distinct political processes, but of prior processes. For both, then, the political process is simply a litmus test that confirms vital prepolitical identities, either cultural or market based. From this point of view, politics appears as fundamentally a noncreative activity. Instead of being seen as outcomes of political processes, political majorities and minorities are seen as permanent features of political life introduced into it from the outside. Neither point of view has methodological room for a serious reflection on the question of political reform.

Anyone interested in the question of political reform after the genocide will need to keep in mind three salient features of the Rwandan situation: first, its starting point, the genocide; second, the consequence of the genocide, a tension-ridden polity and society; and third, that these consequences have overflowed the boundaries of Rwanda, making it the epicenter of the crisis of the African Great Lakes. The starting point of reform in Rwanda, the genocide, is radically different from that in any other country in the region. Like molten lava, the genocide has imparted to the Rwandan polity, and particularly to Hutu/Tutsi relations, tensions that are volcanic in nature. To contain these tensions will not only require a drawn-out cooling-off period and an approach that puts reform in Rwanda in the context of a regional reform agenda; it will also require a commitment and a responsibility that is international, not just regional. Whatever the nature and the length of this transition period, this custodianship, its direction will be shaped by its end goal. Will Rwanda follow the example of Israel and create a separate political community of Tutsi, alongside another of Hutu? Will it follow the example of Zanzibar and merge in a larger union with the tendency to dissolve bipolar political identities—Arab and African in Zanzibar, Tutsi and Hutu in Rwanda—in a wider arena with multiple political identities? Or will it charter a third course—without a wider merger and without creating a Tutsiland, a course we have come to associate with postapartheid South Africa—by trying to forge a political identity that transcends Hutu and Tutsi? Whatever course of action Rwanda chooses, its people will have to begin by making one basic choice: between political union and political divorce.


Postgenocide Rwanda presents a sharp contrast to postapartheid South Africa. In the white population in Apartheid South Africa, there were few perpetrators but many beneficiaries. Among the Hutu in Rwanda of the genocide, there were fewer beneficiaries and many more perpetrators. If it is true that hundreds of thousands of Hutu participated directly in the killings, then reconciliation presents a dilemma, morally and politically. Even a cursory visit to postgenocide Rwanda brings one face-to-face with this dilemma. Every time I visited postgenocide Rwanda, I would ask responsible state officials—sometimes a minister—as to how many ordinary civilians they thought had participated in the genocide. Every time, the answer was in the millions. Even more troubling, the estimate grew with each visit. The first time I went, a minister suggested a practical way to apportion blame and mete out justice: “Categorize according to responsibility. Let those with responsibility be shot in the national stadium. Then go ahead and say that for all those who participated, the three to four million, let them say we did the wrong thing.”2 From “three to four million” in 1995, the figure had grown to “four to five million” in 1997, when another minister told me that “80 percent of those [Hutu] alive had participated in the killing.”3 What was the point of these growing estimates? Was it an attempt by those in power to underscore that the majority of Hutu in Rwanda are guilty of genocide? Or, was it also a claim that this guilty majority be deprived of political rights as punishment for its crimes? I am concerned less with the truth of the claim than with its political significance. Rwanda’s key dilemma is how to build a democracy that can incorporate a guilty majority alongside an aggrieved and fearful minority in a single political community.

The Rwandan state generally avoids the use of Hutu and Tutsi as political identities. But it has adopted a “genocide framework” from which to categorize the population politically, meaning that “the 1994 genocide is singled out as an event producing the only politically correct categories for identification and guidelines” for state policy.4 The state language in Rwanda, the language one hears from all officials, and also from many who are not, divides the population into five categories: returnees, refugees, victims, survivors, and perpetrators. The returnees are, first and foremost, the mainly Tutsi (and some Hutu) exiles who returned with the Rwanda Patriotic Front (RPF). The refugees are divided into two: the “old case load” refers to mainly Tutsi pregenocide refugees, whereas the “new case load” refers to the wholly Hutu postgenocide refugees. The terminology is also used by UN and NGO circles. The victims are said to be both Tutsi and Hutu—the latter victims of the massacres of the internal political opposition. But when it comes to identifying living victims, this identification is limited to the “Tutsi genocide survivors” and “old caseload refugees”; “new caseload are not considered victims and as such are often not entitled to assistance for the construction of homes.”5 Finally, survivoris a term applied only to Tutsi. This is because the genocide was aimed at only the Tutsi, I was told. From this point of view, the “survivor” is a Tutsi who had been in the country at the time of the genocide and who is alive today. The word is not used for any Hutu then in the country. The assumption is that every Hutu who opposed the genocide was killed. The flip side of this assumption is that every living Hutu was either an active participant or a passive onlooker in the genocide. Morally, if not legally, both are culpable. The dilemma is that to be a Hutu in contemporary Rwanda is to be presumed a perpetrator.

Associated with this is another obvious fact: that political violence in the Rwandan genocide had an open, mass, and perversely popular character, as opposed to the secret, cloak-and-dagger nature of political violence in South Africa. Killings in Rwanda were not done by shadowy death squads, but by mobs of machete-wielding citizens. Killings did not happen under cover of darkness, with hardly a witness in sight, and with every effort to destroy the evidence. Instead, they happened in broad daylight, for all to see, and with no effort to destroy the evidence. In a nutshell, while the identity of the perpetrator was not always known in South Africa, it is known in Rwanda.

True, there are many more perpetrators than there are beneficiaries in Rwanda, unlike in South Africa, and their identity also tends to be more public. And yet, neither the identity of the perpetrator nor that of the survivor is as transparent in Rwanda as these differences would lead one to think. This is because the identification of both perpetrator and survivor is contingent on one’s historical perspective. This is why it is not possible to think of reconciliation between Hutu and Tutsi in Rwanda without a prior reconciliation with history. In a 1996 visit to Kigali, I requested to be taken to a school so I could talk to a history teacher. My host, an aide to the vice-president, said this would be difficult since history teaching in schools had stopped. I asked why. Because there is no agreement on what should be taught as history, was the reply. History in Rwanda comes in two versions: Hutu and Tutsi. Ever since the colonial period, the cycle of violence has been fed by a victim psychology on both sides. Every round of perpetrators has justified the use of violence as the only effective guarantee against being victimized yet again. For the unreconciled victim of yesterday’s violence, the struggle continues. The continuing tragedy of Rwanda is that each round of violence gives us yet another set of victims-turned-perpetrators.

To break the stranglehold of Hutu Power and Tutsi Power on Rwanda’s politics, one also needs to break their stranglehold on Rwanda’s history writing, and thus history making. This exercise requires putting the truth of the genocide, the truth of mass killings, in a historical context. To find a way out of this cycle, it is necessary to link political outcomes more to political institutions and less to political agency. The tendency has been the opposite: indeed, to so individualize and decontextualize the truth of the genocide—South Africa-style—that it escapes comprehension. What would it mean to contextualize the truth? It would be, first of all, to connect it to the civil war. This means to avoid two pitfalls: neither to merge and dissolve the genocide in the civil war, in which case it would cease to exist analytically, nor to sever it so completely from the civil war that the act of killing would become devoid of motivation. To see the genocide as one outcome of defeat in the civil war would be to see it as political violence, an outcome of a power struggle between Hutu and Tutsi elites. That would mean both to recognize Hutu and Tutsi as political identities and to recognize that the problem of Rwanda is first and foremost one of political power. There can be no reconciliation without a reorganization of power.

The second consequence of contextualizing the truth would be to understand the civil war in Rwanda as the development of a regional dynamic. I have argued that the critical impetus behind the RPF crossing the border in October 1990 was the confluence of a citizenship crisis in Uganda with that in Rwanda. The invasion was at the same time an armed repatriation from Uganda and an armed return home. Having embraced the Banyarwanda refugees as “comrades-in-arms” during their hour of need in the Luwero Triangle, Ugandan guerrillas-turned-government did not hesitate to “solve” their first major crisis in power by dispensing with the same comrades. The combination of loss of state positions for the elite and refusal to give grazing land to ordinary pastoralists highlighted the continuing homelessness of the generation born of 1959 Tutsi exiles. This same development also dramatically underscored the limits of Habyarimana’s internal reform. Thus the Museveni government exported Uganda’s internal crisis to Rwanda.

The third consequence of contextualizing the truth would be to put the question of power in a historical context. It would be to trace the genesis of Tutsi Power to two moments in Rwanda’s recent history. The first is Rwabugiri’s reforms at the turn of the century, which mark the starting point of a process with two related outcomes, the degradation of the Hutu and the genesis of Tutsi privilege. The second is the colonial reforms of 1926–36 that racialized the Tutsi identity and hardened Tutsi privilege into a crust, giving it an apartheid-like quality. It would undercut both the Tutsi version of Rwanda’s history that Tutsi privilege was exclusively a colonial creation, and the Hutu version that Tutsi privilege is as old as the presence of Tutsi on Rwandan soil.

The fourth consequence of contextualizing the truth would be to reflect on the complicity between the imperial project in twentieth-century Rwanda and history writing in and about Rwanda. In racializing Rwandan society and polity, the imperial project also racialized the parameters within which most historians pursued knowledge most of the time. If the colonial state underscored racial origins as a key attribute of citizenship and rights, historians became preoccupied with the search for origins. If official racism presumed that migration was central to the spread of civilization, particularly statecraft, historians seemed content to center their scholarly pursuits on the question of migration. And finally, if the colonial state defined the subject population as Hutu and Tutsi (and Twa)—regardless of the extent of intermarriage—historians presumed an equallyunproblematized link between ancestral Hutu and Tutsi and those contemporarily so identified. Historians preoccupied with the search for origins read cultural differences from facts of migration and translated cultural into political difference. To differentiate cultural from political identity, and thereby to depoliticize historical facts of migration, it seems to me, is a prerequisite to rethinking the question of citizenship in postgenocide Rwanda.

The fifth consequence of contextualizing the truth would be to problematize both the 1959 Revolution and the ideology born of it, Hutu Power. To do so would be to recognize both the historical legitimacy of the revolution and its historical limitation. The 1959 Revolution was the antidote to Tutsi privilege that had crystallized between the time of Rwabugiri and the end of the colonial period, just as Hutu Power was the antidote to the Hamitic hypothesis that provided the civilizational rationale for Tutsi Power. Hutu Power had undertones of a subaltern ideology, similar to Black Power in the United States, Black Consciousness in South Africa, or Dalit Power in India. The obvious difference was that by the time Hutu Power was formulated as ideology, its proponents were already in power. Theirs was a call to defend power, not to take it. Yet, like its counterparts elsewhere, Hutu Power needs to be understood as a contradictory possibility. An outcome of struggle in the world of the rat and the cat, it had not only the potential of liberating the rat from the terror of the cat, but also of locking the rat forever in a world driven by fear of the cat. That world is one in which there are no other—and bigger—dangers or possibilities. To be locked into the claustrophobia of intimate differences, and to be blind to larger possibilities, is the historical limitation of Hutu Power. The historical failure of Hutu Power was that it failed to transform the political legacy of colonialism. Instead, it built on the very racialized political identities generated by colonial rule: of Hutu as Bantu and Tutsi as Hamites.

The sixth consequence of contextualizing the truth would be to distinguish between Hutu Power and génocidaire, as ideology and as political tendency. While Hutu Power was a broad and contradictory tendency born of the hope of the 1959 Revolution,génocidaire is a narrow tendency coalesced by the desperation of defeat in the civil war. True, the latter is born of the former, of its negative side. Yet, this child of adversity cannot be confused with the parent. While Hutu Power reconciled itself to living in the bipolar world of Hutu and Tutsi by a political struggle with Tutsi Power so as to acquire political supremacy over it, the génocidaires looked for a final solution to this bipolarity in the physical elimination of the Tutsi.

The final advantage of contextualizing the truth would be to recognize that Rwanda is once again at a historical crossroads where its political leadership is faced by two clear options. The first is a continuation of the civil war, as those defeated in the last round prepare for battle in the next; the second is its termination through a political reconciliation that rejects both victory and defeat and looks for a third and more viable possibility. Each of these possibilities is linked to a different form of justice and a different form of the state. The first is victor’s justice, the second survivor’s justice.


Victor’s Justice

To pursue victor’s justice would be to follow the example of Israel. It would be to build a Zionist-type state on the ashes of the genocide. This is indeed what is happening in contemporary Rwanda. Three convictions underline the character of postgenocide power in Rwanda. The first is an overwhelming sense of moral responsibility for the very survival of all remaining Tutsi, globally. This gives postgenocide power its first distinguishing characteristic: it is defined by a diasporic, rather than a territorial, notion of political obligation and political community. The second conviction—also a direct outcome of the experience of the genocide—is that Tutsi Power is the minimum condition for Tutsi survival. Tutsi will only be protected if they have a state of their own. I found this conviction shared by both the Congolese Tutsi legal adviser to the secretary-general of the Alliance of Democratic Forces in Kabila’s Congo, and the newly appointed Rwandese commander of the Congolese national army in 1997. This point of view marks postgenocide power with yet a third conviction: that the only peace possible between Tutsi and Hutu is an armed peace. It also lends credibility to those in the opposition who argue that the Hutu must be armed if they are not to return to the servile condition of pre-1959 Rwanda.

Thus, even the moderate opposition to the RPF complains that not only are structures of power in Rwanda being Tutsified, civic organizations—from the media to nongovernmental organizations—are being cleansed of any but a nominal Hutu presence.6 On its part, postgenocide power is determined to remove from the soil of Rwanda any trace of conditions that could possibly lead to a repeat of the genocide. Its unswerving motto recalls the claim that made post-Holocaust power in Israel immune to any moral doubts when it came to atrocities against Palestinians: NEVER AGAIN. Ironically, the conviction that Tutsi Power is the precondition for Tutsi survival means that life itself can be subordinated to this supreme goal, the survival of Tutsi Power.

The founding ideology of Tutsi Power in postgenocide Rwanda is the memory of the genocide and the moral compulsion never to let it happen again. The pursuit of the génocidaires is the raison d’être of the postgenocide state, the one permanent part of its agenda. In the real world of state politics, however, the word génocidaire may be used to label any Hutu seen as an opponent, or even a critic, of Tutsi Power. Arrests can be made on the basis of denunciation, not investigation. Even if the crowded jails of Rwanda take a daily toll on the lives of those incarcerated within, this does not disturb moral sensibilities.7 The moral certainty about preventing another genocide imparts a moral justification to the pursuit of power with impunity.

Most recognize that the precondition for victor’s justice is, clearly, victory. Few, however, recognize its price. The victor must remain on constant guard, lest the spoils of victory be snatched yet again. Just as a jailer comes to be tied to the jail as much as is the prisoner, so a victor must live in anticipation and fear of the next round of battle, why adversaries often tend to get locked into a single cycle more securely than do friends. The price of victor’s justice is either a continuing civil war or a permanent divorce. It is worth remembering that it is not simply German defeat in the Second World War that made Nuremburg possible, but also the effective divorce between Gentiles and Jews in Germany, since most surviving German Jews departed for either America or Israel. In the absence of this effective divorce, anything resembling Zionist power in Germany would have been a recipe for triggering a civil war. In this sense, we need to bear in mind that while the RPF won the war, there has been no divorce between Tutsi and Hutu in Rwanda. The price of victor’s justice, in Rwanda, must thus be yet another round of a continuing civil war.

It is also worth remembering a second difference between the Nazi Holocaust and the Rwandan genocide. Though both were designed from above, from within the state, the genocide alone unfolded as wave upon wave of mass killings, where not only victims but perpetrators too were drawn from civil society. As a state project that was carried out by many in society, the Rwandan genocide resembles apartheid more than it does the Holocaust. This is why victor’s justice—the Tutsification of state institutions—cannot be an effective guarantee against a repeat of genocidal violence in Rwandan society. If anything, it will keep alive the specter of yet another round of genocidal violence.

Survivor’s Justice

The form of justice flows from the form of power. If victor’s justice requires victor’s power, then is not victor’s justice simply revenge masquerading as justice? To get away from this dilemma, we need to explore answers to two questions. Is a form of justice possible that is not at the same time victor’s justice? Is a form of reconciliation possible that is not at the same time an absence of justice, and thus an embrace of evil? These questions provide a clue to finding a way out of the dialectic of civil war. That way has to be anchored in an alternative form of justice that I will callsurvivor’s justice.

The prerequisite for survivor’s justice, as for victor’s justice, may also be victory. For victory presents alternatives to the victor, which it does not to the vanquished. Only the victor has the choice of reaching out to the vanquished on terms that have the potential of transcending an earlier opposition between the two, by defining both as survivors of the civil war. To transcend the terms of the earlier opposition is to forge a new community of survivors of the civil war. From this point of view, the term “survivor” does not refer to surviving victims—which, as I have pointed out, is how it is used in contemporary Rwanda—but to all those who continue to be blessed with life in the aftermath of the civil war.8 The notion of survivor seeks to transcend the bipolar notions of victim and perpetrator.

The difference between victor’s justice and survivor’s justice is clear if we look at the two major postwar paradigms of justice: de-Nazification and de-Sovietization. The former came into being at the onset of the Cold War. The latter marked the end of the Cold War. Simply put, the logic of de-Nazification is to blame the agent, that of de-Sovietization is to blame the system; de-nazification requires identifying both victims and perpetrators. De-Sovietization is anchored first and foremost in the identity of survivors; it acknowledges victims, but not perpetrators. From this point of view, to identify individuals as perpetrators would be to demonize them. To pursue the logic of de-Nazification in contemporary Rwanda would be to identify the leadership of the genocide so as to hold it accountable. Such, indeed, is the purpose of the international court in Arusha and the local courts inside Rwanda. To pursue the logic of de-Sovietization would be to put emphasis, first and foremost, on the institutions of rule in Rwanda. Where survivors—victims and perpetrators from an earlier round of struggle—must learn to live together, ways must be found to reconcile the logic of reconciliation with that of justice.

Survivor’s justice is different from revolutionary justice. It makes sense only in contexts where there have been few beneficiaries in the preceding civil war. I have already commented on the difference between South Africa and Rwanda on this score: one is struck by how few were the perpetrators of apartheid, and how many its beneficiaries, and conversely, how many were the perpetrators in Rwanda’s genocide and how few its beneficiaries. Where beneficiaries are many, reconciliation has to be social to be durable, which is the same thing as saying there can be no durable reconciliation without some form of social justice.9 But where beneficiaries are few, the key to reconciliation is political reconciliation. The prime requirement of political reconciliation is neither criminal justice nor social justice, but political justice. It requires not only shifting the primary focus of reform from individuals to institutions, but also recognizing that the key to institutional reform is the reform of institutions of rule. Thus the question: What would it mean to reform institutions of rule so as to give survivors of the genocide another chance?


The genocide retrenched Hutu and Tutsi as salient political identities. The dilemma of postgenocide Rwanda lies in the chasm that divides Hutu as a political majority from Tutsi as a political minority. While the minority demands justice, the majority calls for democracy. The two demands appear as irreconcilable, for the minority sees democracy as an agenda for completing the genocide, and the majority sees justice as a self-serving mask for fortifying minority power. To break out of this logjam, I suggest we link both political justice and political democracy to a reform of institutions of rule.


The question of political justice goes beyond holding the perpetrators of the genocide accountable. Ultimately, it is about the definition of political identities. I have argued that European colonialism in twentieth-century Africa turned indigeneity into the litmus test of rights. Every postindependence regime vowed to change the political world of the settler and the native. Every one of them pledged to deracialize civic rights by making them available to all citizens regardless of color. That is where similarities ended.

While everyone agreed that the settler’s prerogative had to go, not everyone was agreed that the native too was a colonial construct that needed to be reformed just as urgently. Could the political identity “settler” be done away with when its bipolar twin “native” was embraced? Anticolonial nationalism was divided on this question. Radical nationalism—as championed by Julius Nyerere, for example—was determined to reform citizenship consistently, both to deracialize and to deethnicize it. From this point of view, it was not enough to do away with just the settler’s prerogative; all prerogatives, racial as well as ethnic, would need to be abolished. The predominant trend in African postcolonialism was otherwise: for conservative nationalism, the point of independence was precisely to replace the settler’s prerogative by the native’s prerogative.

Even though the political prerogative was transferred to the native, the continued legal representation of the indigenous population as natives showed that the colonial political legacy had yet to be fully transcended. Where colonial rule had been indirect, as in Uganda and Congo, the native prerogative was defined as ethnic. But where colonialism had imposed a version of direct rule—a halfway house, as I have said, in the case of Rwanda—the prerogative was racial. The 1959 Revolution in Rwanda against the Tutsi, like the 1964 Revolution in Zanzibar against Arabs and the 1972 expulsion of Asians in Uganda, belongs to this second category. Targeted in 1959 as an alien race, the Tutsi were recognized as an indigenous ethnicity by the Second Republic after 1973, but reconstructed as an alien race by the génocidaires after the coup of April 1994. As in the 1959 Revolution, so in the 1994 genocide too, the Tutsi were targeted as an alien race. Political justice for the Tutsi cannot mean simply identifying and holding the perpetrators of massacres accountable. By itself, that would return them to the world of the rat and the cat. It also requires a juridical and institutional reform that ceases to make a distinction between two kinds of citizens: one indigenous, the other not.

In contrast to colonial Rwanda, where race was the salient political identity, Congo and Uganda were indirect-rule colonies where both race and ethnicity defined political identity. If the settler identity was racialized, the native identity was ethnicized. Did it not follow that, in indirect-rule colonies such as Uganda and Congo, decolonization would require a combination of deracialization and deethnicization, as indeed Nyerere had championed in Tanzania? On this question, too, nationalism was differentiated. The mainstream—conservative—view was that the world of the “customary” as defined by colonialism was indeed the world of African tradition, and so the conviction that it must be preserved.10 A reform executed from this point of view did two things. While civic law and civic authority were deracialized in the name of a universal rights culture, an ethnically defined “customary” law and an authority to enforce it were retained as particular to the tradition of those indigenous to Africa. Independent governments also vowed to end the perversion of colonialism by restoring the political prerogative of those indigenous over strangers. The result was to reproduce the bifurcated world created by colonialism: the distinction between indigenous and non-indigenous, abolished in the civic sphere, remained in the ethnic sphere. Even if turned upside down, the political world remained as designed by the settler.

The antidote to the embrace of colonially constructed custom as authentic African tradition came from among the postindependence oppositional political movements that had to contend with the rights of ethnic strangers. As one would expect, the most promising initiatives came from those that stood to lose the most from an uncritical reproduction of the colonial legacy. Not surprisingly, the most creative departures have come from those movements strongly influenced by Rwandan Tutsi: the Banyamulenge in Congo, and the National Resistance Army in Uganda. Of the two, we have seen that the most radical solution to this dilemma came from the latter, born of the guerrilla struggle in the Luwero Triangle. Luwero had an extremely heterogeneous population: anywhere from a third to a half of its residents had immigrated from outside the area. To continue to define rights on the basis of indigeneity in such a socially heterogeneous context was bound to be politically explosive and disruptive—regardless of whether one leaned in favor of those indigenous or those not. Welding an alliance between locals and migrants required a political identity that could encompass both. The National Resistance Army found this identity in the criterion of resident. When it came to deciding who would be a member of a village council and who could run for office on the ten-person village committee, what mattered was residence, not the circumstances of one’s birth or ethnic belonging.

To leave the test of indigeneity for one of residence as the basis for political identity and political rights is to take leave of the world of the rat and the cat, of ethnicity and race, of the native and the settler, as political identities. This, in turn, would require making a clear distinction between cultural and political identities so as to redress the dialectic between the past and the future. To ground political rights in cultural identities is to accent the past—of which a shared culture is one outcome—as a guide to limiting future possibilities. To differentiate political from cultural identities, however, is to accent the commitment to live under a common roof over the recognition of a common history—no matter what the overlap between them—as the real basis for a shared future.


Just as the question of political justice goes beyond holding the perpetrators of the genocide accountable, so the question of democracy goes beyond that of who should govern to deciding how they should govern—through what kinds of institutions. To address the institutional basis of rule is to address a dual and combined legacy. To the colonial legacy of administering local communities through despotic forms of power, nationalism has added the legacy of equating democratic rule with unqualified majority rule, in a setup in which the winner takes all and power is the prized and unchecked possession of the majority.

I have restated here with reference to Uganda and Congo—and, with qualifications, to Rwanda—the argument that I elaborated in an earlier book with reference to indirect-rule colonialism in twentieth-century Africa: the true seat of colonial despotism was not the central state and civic law which limited the regime of rights to races, but the local state wherein an unaccountable authority called “customary” enforced an authoritarian version of custom as “customary” law.11 The conquest state removed any trace of democratic accountability to those below, and reinforced every sign of bureaucratic accountability to those above. In Rwanda of 1959, this “customary” authority was totally discredited since it was identified with Tutsi Power and Tutsi privilege. The revolution not only replaced Tutsi with Hutu chiefs, it also made the Hutu functionaries accountable to a popular mandate through regular elections, but without disaggregating the despotic power—legislative, executive, judicial, administrative—that these agents exercised. The Second Republic eliminated local elections and re-created the despotic local state constructed under colonialism—this time, though, as not only “customary” but also “revolutionary.” Power and authority defined as “customary” in the colonial period have not only been at the center of coercive day-to-day practices in much of postindependence Africa, they have also—in the context of a racialized bipolar difference as in Rwanda—orchestrated and organized the mass slaughter that led to the genocide.

To this, some may say: Why not just go ahead and junk custom? My argument is that when a particular version of history (custom) is found wanting, in this case because it builds on the authoritarian strand as if it were the entire past, this surely cannot be reason to junk the very notion of history. From a reified language fortifying a despotic authority, custom needs to be rethought as a thread of life, not only one that makes us but also one that we make. To smash one version of the past as a prison dressed in the language of custom, one needs to turn that very past—the entire treasure house called custom, and not simply the authoritarian strands in it that colonial power welded into a “customary” law—into a plural resource for more open futures. There has been little effort at a comprehensive and critical rethinking of custom and its relationship to law, and it is not the purpose of this book to do so.

The region of the African Great Lakes, provides one important clue, by way of experience, to dismantling the local machinery of despotism that went by the name of Native Authority. We have seen that the guerrilla struggle in the Luwero Triangle recognized local power by disaggregating the moments of power that had been fused into one. The chief thus ceased to have the right to pass a by-law; instead, this legislative power became a prerogative of popular local councils, from the village at the lowest level to the district at the highest. Similarly, judicial and executive power was also transferred from the chief to other organs. Where the chief remained, he retained only administrative power as a paid agent of the state. Whether for the Hutu majority or for the entire people, democracy has to mean first of all an institutional reform that unravels this armed fist and separates each moment in this fused power (executive, legislative, judicial, administrative) and makes it accountable.

Both in forging a majority and in waging a struggle for unqualified majority power, the region offers two lessons to postgenocide Rwanda. The first comes from indirect-rule colonies; the second from colonies with more of a legacy of direct rule. The distinctive feature of the indirect-rule state, such as in Uganda and Congo, was that it fractured the identity of the colonized majority into so many ethnicities, each a minority. It was thus said that, unlike with nations in Europe, there were no natural majorities in African colonies; everyone was said naturally to belong to a minority. The statement contained an element of truth: everyone did belong to a minority, except that the minorities were not natural; ethnic identity was a political artifact of state power. In such a context, democracy was likely to be a recipe for instability, and there was a strong temptation to see a benevolent dictator as the only realistic source of stability, even for rule of law. To forge a political majority in this context would require dismantling the ethnically organized apparatus of indirect rule. Before that could be done—indeed, so that it may be done—a way had to be found to put together a transitional majority.

In the political vocabulary current in the region of the African Great Lakes, the search for an agency of rule that can bring stability to a post–civil war context has come to be known as the search for “a broad base.” Where no political movement could marshal a consensus and where there was a history of bitter fragmentation, as in Uganda, the practice of coalition government came to be seen as necessary to ensuring a sufficiently broad base for rule in a period of transition. The practice of the “broad base” made a clear distinction between means and ends. All political tendencies—whether monarchist or “tribalist,” even when identified with as brutal a dictatorship as that of Amin—were welcomed into the broad base provided they gave up violence as a means for attaining their objectives.

The lesson for postgenocide Rwanda is indeed radical. It would mean making a distinction between proponents of Hutu Power and perpetrators of the genocide. This would mean making a distinction between ends and means, politics and ideology, and thus between those proponents of Hutu Power willing to give up violence as a means and those not willing to do so: the former would be invited into the “broad base,” the latter would be left out of it. This, indeed, is also the lesson of Rwanda’s political history. The last attempt to put an end to the cycle of civil wars in Rwanda was the Arusha Agreement. Its key lesson is that one cannot put an end to the civil war by excluding one party to it, especially the party most entrenched in its partisan ideology. The lesson is to be inclusive, to recognize the right of all ideological currents—without exception—to compete in the marketplace of ideas, leaving out in the cold only those unwilling to disarm as a precondition to gaining entry to the reform process.

In colonies with more of a tradition of direct rule, as in Rwanda, the majority classified as indigenous was literally panel-beated by the apparatus of rule into a single, racialized mass. This is why in instances like Rwanda, Zanzibar, and preapartheid South Africa, the “Hutu,” the “African,” or the “Bantu” appeared as natural, prepolitical, majorities. It is also in these very countries that privileged but vulnerable minorities—such as Arabs in Zanzibar or whites in South Africa—concluded that power was indeed their only guarantee for life, liberty, and, indeed, property. Does this not echo the central conclusion that the RPF seems to have drawn from the history of postindependence Rwanda: that the Tutsi cannot survive without power? But, here too, the weight of experience seems to point in a different direction. Rather than think that power is the precondition for survival, the Tutsi will sooner or later have to consider the opposite possibility: that the prerequisite to cohabitation, to reconciliation, and a common political future may indeed be to give up the monopoly of power. Like the Arabs of Zanzibar, and even the whites of South Africa, the Tutsi of Rwanda may also have to learn that, so long as Hutu and Tutsi remain alive as political identities, giving up political power may be a surer guarantee of survival than holding on to it.

Yet, we cannot ignore the one fact that must weigh like a nightmare on the minds of Tutsi survivors: neither the Arabs of Zanzibar nor the whites of South Africa have gone through the experience of genocidal violence as have the Tutsi of Rwanda. To find historical parallels to this situation, where an imperiled minority fears to come under the thumb of a guilty majority yet again—even if the thumb print reads “democracy”—we have to take leave of the boundaries of Africa for the New World. Only in the erstwhile settler colonies of the New World do we have a comparable history of violence that has rendered the majority guilty in the eyes of victimized minorities. Such, indeed, has been the aftermath of genocide and slavery: the genocide of indigenous populations in the Americas, Australia, and New Zealand, and the slavery of Africans in the Americas.

If we are to go by these experiences, we have to admit both that the attainment of self-enlightenment by guilty majorities has been a painfully gradual process, and that even the little progress made along that road has been as much a result of international pressure as it has been of internal struggle. Rwanda is different from the New World in one important respect: the cycle of violence has alternated positions between victims and perpetrators. The internal pressure in Rwanda is now joined to a regional dynamic as two diasporas—one Hutu, the other Tutsi—confront each other in a life-and-death encounter. Both diasporas are animated, not simply by the cycle of revenge in Rwanda but also by the common regional inheritance that has been translated into a mode of citizenship that denies full citizenship to residents it brands as ethnic strangers. Not surprisingly, Rwanda has become the epicenter of the wider crisis in the African Great Lakes. Tied together by the thread of a common colonial legacy—one that politicized indigeneity as a basis for rights—the region has little choice but to address the Rwandan dilemma, if only to address its own dilemma. To do so will mean, first of all, to reform the state and citizenship within their own borders so that power recognizes equal citizenship rights for all based on a single criterion: residence.

To reform Rwanda, the epicenter of the crisis, will require a regional approach through a regional agenda that approaches the center as firefighters would approach the heart of a raging fire, from the outside in. If a regional reform of citizenship needs to be its first step, its second step may have to focus on Rwanda’s splitting political image, Burundi. Precisely because Rwanda and Burundi read developments in each other’s backyard as prophetic signs of their common fate, reform in Burundi can serve as a compelling example for Rwanda. For that reason if for no other, it is in Burundi that the regional and international community would be wise to invest physical resources alongside political guarantees to bring a political reform. For without a reform in power, one that recognizes both the importance of a majority in politics and the need for fearful minorities to participate in the exercise of power, there can be no sustained reconciliation between Hutu and Tutsi. To do so will be to recognize that neither the tragedy of Rwanda nor its possible salvation can be exclusively, or even mainly, a Rwandan responsibility.

ONE needs to close with a sense of real political obstacles that will face any attempt to democratize public life in postgenocide Rwanda. Where there is an uneasy coexistence between guilty majorities and fearful minorities, the possibility of a democratic transition is likely to appear more as a threat than a promise to the minorities concerned—why vulnerable minorities tend to fear rather than welcome democracy. The experience of the Tutsi, too, is likely to reinforce an ambivalent attitude to democracy. Were not the Tutsi liberators inside Uganda’s NRA sidelined on the morrow of the guerilla victory precisely because they came from a vulnerable minority? Did not the dawn of democracy in Zaire, signified by the coming together of a National Conference of civil and political society in the early 1990s, complete the process leading to the disenfranchising of the Banyarwanda minority? Was not the Rwandan genocide driven forward by the energy of popular mobs mobilized to defend Hutu Power? By itself, majority rule provides no guarantee for minorities that fear majority domination. My point is that if we go by the experience of Banyarwanda—and more specifically Tutsi Banyarwanda—in the African Great Lakes, majority rule can be turned into a bedrock for domination over fragile minorities.

How to foreclose the possibility of a democratic despotism remains our toughest challenge yet. While this question is not directly the subject of this book, I believe its subject does bring us a step closer to addressing this question. I began the book with the claim that, even when they mimic preexisting identities—whether cultural or market based—political identities need to be understood as a product of the political process. From this point of view, Hutu and Tutsi need to be understood both as historical identities and aspolitical identities. As majority and minority, Hutu and Tutsi are not natural identities brought into the political realm; they are political artifacts of a particular form of the state.

If the immediate challenge in Rwanda is to undercut Hutu and Tutsi as political identities, I have argued that this will not happen so long as the minority monopolizes power. If anything, it will be the surest way of locking the Banyarwanda into the world of the rat and the cat, and giving these identities a longer lease on life. The region provides us two examples of how a minority may give up power. The first is Zanzibar, the second South Africa. For a minority gripped by the fear of extinction, the Zanzibari example is likely to have greater resonance, for at least one reason: it involved longer-term political concessions by both the minority and the majority. Not only did the “Arab” minority cede power, the “African” majority in Zanzibar also ceded full claim on power as the country merged with mainland Tanganyika to form a wider union, Tanzania. The union set in motion a new dynamic tending to dissolve the identities “Arab” and “African” in a wider crucible, over time generating a “Zanzibar” identity. Is a dynamic possible that may undercut the legacy of Hutu and Tutsi as binary political identities, dissolving them in the crucible of a larger Banyarwanda identity in the short run and, other identities we may not imagine today, in the medium run? If yes, it will require us to question the hitherto presumed equation of the democratic project with the national project. Indeed, if it is to be, it will need to draw on energies that go beyond any national assertion. Such a dynamic will need to be the result of a regional initiative, backed up by international support, which in turn needs to be driven by the urgent need to defuse a simmering volcano before it blows up yet again, this time engulfing the wider region.

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