In Burundi do the tall ones kill the short ones or the short ones kill the tall ones? I know Burundi does one thing and Ruanda the other.

Shiva Naipaul, North of South: An African Journey


Wherever you go in Rwanda and Burundi—to a private house, a bar, a government office, a refugee camp—drinks are served with the bottle caps on, and opened only before the eyes of the drinker. It is a ritual that honors the fear of poison. An open bottle is unacceptable. When, as with the potent banana beer which is drunk in quantity in rural areas, a drink comes unbottled from a common pot, or when a drink is to be shared, the provider must take the first sip—like a food taster in a medieval court—to prove that it is safe.

Tales of poisonings punctuate the lore of pre-colonial and colonial Rwanda and Burundi. Indeed, Marc Vincent, a pediatrician from Brussels who practiced in the Belgian colony of Ruanda-Urundi during the early 1950s, observed that the locals tended to regard poisoning or sorcery as the main cause of fatal illness. In his monograph L’Enfant au Ruanda-Urundi, Vincent recalls overhearing a very sick ten-year-old Burundian boy tell his father, “When I die, you must see who poisoned me.” And an eight-year-old boy told Vincent, “Yes, death exists, but all those who die here, it’s not ordinary death, it’s sorcery. When you spit on the ground, one takes your saliva, one takes the dust on which you walked. My parents told me to watch out.” Such attitudes, Vincent reports, pervaded all levels of society.

Family hatreds and vendettas were extremely tenacious and members of enemy families would not risk casually to drink beer at each other’s homes. One chief of Urundi, young and raised in European ways, genuinely disengaged from the Bantu beliefs, told me one day: “We can’t ignore these family enmities…just as you see me, I wouldn’t dare to go drink or eat with old Tutsis, or the elders of an enemy family, because I know that they are capable of slipping me some poison. Yet, if you saw me meeting them in public, you could take us for the best friends in the world, so completely would they and I speak civilities and politeness.” This declaration is not unique, and must be taken to reflect the exact situation. The natives see poisoners everywhere.1

Even today, after the most extreme political violence in all of post-colonial Africa, Rwandans and Burundians still frequently blame invisible poisoners when the deaths of public figures cannot be otherwise explained. Poison, then, takes on the quality of a metaphor for societies in which there must be little expectation of a natural end. When death is the work of enemies, distrust and subterfuge are the means of survival, and politics itself can become a poison.

We saw the effects of this poison in Rwanda in the spring and summer of 1994, when Hutus were provoked by their leaders to slaughter at least 800,000 and perhaps as many as one million people from the Tutsi minority, as well as tens of thousands of Hutus who were perceived as enemies of the extermination program. Now a host of voices—in the UN, the State Department, humanitarian aid groups, and the press—warn that Burundi, which has the same mix as Rwanda (roughly 85 percent Hutu, 14 percent Tutsi, and 1 percent Twa pygmies), is also at risk of erupting into a frenzy of genocidal mass killing. As René Lemarchand’s lucid political history of Burundi makes clear, it would not be the first time; the poison of “ethnic” violence has already claimed the lives of as many as half a million Burundians since the country gained independence from Belgium in 1962. Fighting between the Tutsi-dominated military and an increasingly well-armed and tightly organized Hutu insurgency claimed around one thousand lives a month in 1995, and the situation is rapidly deteriorating. Reports that twenty Hutu civilians were killed on May 5 by Tutsi soldiers or militias in northern Burundi followed accounts of a massacre on April 26 of 234 Hutu civilians in the Gitega district of central Burundi, including 137 women and 87 children. At the end of April, one of the main roads into the capital city of Bujumbura—now an “ethnically cleansed” Tutsi enclave ringed by Hutu rebels—was closed for three days because of land mines laid by rebel insurgents. The murder of political officials has become an almost daily occurrence. Staff members of three international humanitarian relief agencies which have not already left Burundi are frequently attacked. As Iwrite, Bujumbura is without electricity and has been for the past week, and there is a severe water shortage. In response to Hutu rebel activities, Antoine Nduwayo, the country’s Tutsi prime minister, recently urged his young followers “not to hesitate to destroy people they would qualify as suspects.”


Lemarchand, a French-born political scientist, has been tracking the conflicts between Hutus and Tutsis since 1960, two years before Rwanda and Burundi gained independence from Belgian colonial rule. The central question he has been forced to confront is how these two countries, which are unique among modern sub-Saharan African states in having existed as cohesive political societies, more or less within their current borders, for centuries before colonialism, have recently produced the continent’s most unambiguously genocidal politics.

The precise origins of Hutus and Tutsis have never been reliably established. According to the prevailing legend, Hutus are a Bantu people who settled the Great Lakes Region of central Africa well before the Tutsis arrived, migrating, it is supposed, from Ethiopia or elsewhere in the horn of Africa four or five hundred years ago. Hutus were cultivators and Tutsis were cattle-herders, and this appears to be the main indisputable distinction between the two groups, apart from their different, but highly variable, physical characteristics. By the time anything like a reliable historical record can begin to be cited, around the early to mid-nineteenth century, Hutus and Tutsis had developed a common culture. They shared the same languages (Kirundi in Burundi, Kinyarwanda in Rwanda), the same spirit faiths, folk customs, political institutions, and cuisine, and they lived intermingled, with few territorial disputes and without taboos against intermarriage. For these reasons, they cannot properly be said to fit the anthropological definition of separate tribes or ethnic groups.

Rwanda and Burundi, moreover, had been organized as separate political entities before the colonial period. In Rwanda, Tutsis emerged as the dominant monarchical clans and the economic elite. In Burundi, the picture was more complex: the monarchs came from a dynastic lineage known as the ganwa, who were distinct from Hutu and Tutsi and regarded them both as collectively inferior. Frequent feuding among rival ganwa clans kept any single king from consolidating his power. In addition, there were two classes of Tutsis in Burundi—the Banyaruguru, known colloquially as the “high Tutsis,” and the Hima, or “low Tutsis”—and poor Tutsis often found themselves subservient to wealthy Hutus. Under the circumstances, Lemarchand writes, “Even the most superficial reading of Burundi’s precolonial history reveals unmistakable evidence of widespread social discontent; yet seldom does it express itself in the form of a sharp and persistent Hutu-Tutsi opposition.”

Then the Europeans came—Germans in the last years of the nineteenth century, and Belgians after World War I—bringing with them the fiction of “race science” in a particularly ludicrous form, the so-called “Hamitic hypothesis,” which posited that Tutsis are the lost tribe of Ham, a tall, beautiful master race, and that Hutus are a slavish Negroid sub-species. To be sure, despite intermarriage and social mobility, Hutus and Tutsis had, to varying degrees, retained their distinctive physical characteristics, and although these attributes are a notoriously unreliable means of identification, the Europeans imagined a logical correspondence between the shapes of their bodies and their status. Marc Vincent expressed the conventional wisdom of his day when he described Tutsis as “nobles”—slow-moving, lazy, “slender,” “fine featured,” “diplomats and sophisticates”—and Hutus as their “feudal vassals”—stocky, coarse, “poor, loutish, dirty and ill-clad.”

This was the myth with which the Belgians helped to create tribal divisions in the territory they administered as Ruanda-Urundi. They pursued a policy of divide-and-rule and imposed an apartheid-style system that dismantled the traditional monarchical structures which had maintained considerable social cohesion in the country. Under the Belgians, “ethnic” identity cards were issued, and Tutsis were favored for educational, professional, and political opportunities. However much Tutsis may have exploited their privileged status, as Lemarchand makes clear, it was “the European colonizer who introduced this vector of disunity,” thereby writing “the script” for the political polarization and violence that filled the vacuum of power that was created when the Belgians left the country to self-rule.

On October 13, 1961, on the eve of independence from Belgium, the man who had emerged from legislative elections as Burundi’s prime minister designate, Prince Louis Rwagasore—a son of Burundi’s last ganwa king, Mwambutsa, and a nationalist much loved by both Hutus and Tutsis—was assassinated at a restaurant overlooking Lake Tanganyika. The gunman was a Greek named Jean Kageorgis, who was later discovered to have been hired by two sons of Mwambutsa’s leading rival to the throne, Pierre Baranyanka. In January 1963, the two brothers and three co-conspirators were hanged before a crowd of about ten thousand spectators in Bujumbura by order of the high court of the Republic.

Throughout the colonial period, the would-be king, Baranyanka, had allied himself with the Belgian administrators against Mwambutsa, Prince Rwagasore’s father, and he had become an advocate of Tutsi supremacy. Conversely, many Hutus gravitated toward the view that Mwambutsa was their true champion—a “just king” thwarted by the colonial system. With independence, as organized party politics developed, the old princely rivalries took a different form and Rwagasore’s murder quickly acquired “ethnic” overtones.2 Moreover, the shadow of Rwanda’s bloody Hutu revolution of 1959-1962, in which tens of thousands of Tutsis were massacred and hundreds of thousands were driven into exile, fell heavily on Burundi, making Tutsis particularly distrustful of Hutu advances.


Rwagasore’s assassination, then, set the pattern for an all-engulfing Jacobean drama of infighting among Burundi’s new political elites, who increasingly pursued their quest for power by exploiting ethnic divisions. By 1963, with “the National Assembly split down the middle between Hutu and Tutsi, and the government almost paralyzed,” King Mwambutsa attempted to become a broker between the parties. But his efforts were short-lived.

In January 1965, Pierre Ngendandumwe, a Hutu, was appointed prime minister by King Mwambutsa, only to be gunned down three days later by Tutsi extremists. Ngendandumwe’s killing set off a new round of unrest, which led to a coup attempt late in 1965 by Hutu military officers. Tutsi officers immediately launched a successful counter-coup, seizing control of the previously integrated military, purging Hutu officers, arresting virtually every Hutu leader in Bujumbura, and executing a great many of them. The Tutsi repression, according to Lemarchand, resulted in “the physical elimination of the entire first generation of Hutu leaders,” and as many as ten thousand Hutu deaths nationwide. In 1966, a Tutsi army officer, Captain Michel Micombero, abolished the monarchy and established himself as president. Micombero’s consolidation of power in a Tutsi-dominated military dictatorship has continued to define Burundian political life to this day. As Lemarchand writes, the specter of a “Hutu peril” was also useful in keeping a lid on intra-Tutsi struggle at the higher levels of state power.

Burundi’s definitive catastrophe, however, did not come until 1972, when, once again, a brief Hutu uprising provoked a drastic response from the military. In the name of restoring “peace and order,” at least one hundred thousand Hutus were slaughtered and as many as two hundred thousand were driven into exile. The property of the victims was transferred to Tutsis. While the killing was often indiscriminate, Lemarchand argues that the Tutsi leaders who planned the massacres were bent on the “systematic elimination of Hutu elites.” And, indeed, the pattern of what Lemarchand calls “selective genocide” was consistent with the 1965 killings; after three months “the army, the government, and the economy [were] virtually purged of Hutu elements.” As radio broadcasts urged on the killers, Tutsi schoolchildren were asked to draw up lists of their Hutu classmates; trucks in Bujumbura rounded up every identifiable Hutu white-collar or skilled worker and took them to jail, where they were “either shot or beaten to death with rifle butts and clubs”; Tutsi youth militias called thousands of Hutu students out of college classes, and they were never seen again. The definition of a Hutu, which once required a Hutu father, was expanded to include anyone with one Hutu grandparent. A program of mass executions reduced the already slim ranks of Hutus in the army from five hundred soldiers to no more than one hundred.

As a comment on his account of these atrocities, Lemarchand bitterly quotes from a diplomatic cable, sent on May 29, 1972, from the US embassy in Bujumbura to the State Department:

Trucks ply the road to the airport every night with a fresh contribution to the mass grave…. But the real concern is coffee…. If the coffee harvest fails, the modern sector of the economy will be sorely buffeted, if not destroyed.


Lemarchand sees the carnage of 1972 as a watershed in Burundian history. Mass death had made the colonial fiction of tribalism a reality. “For the next fifteen years,” Lemarchand writes, “only Tutsi were qualified to gain access to power, influence, wealth,” and “what was left of Hutu society was systematically excluded from the army, the civil service, and the university.”

The purge of Hutu elites meant that for the first time Tutsi power stood unchallenged, except for the nagging “crisis of legitimacy” that invariably plagues ethnocratic minority regimes. Before 1972, Hutu polemicists had taken to comparing their plight as an oppressed majority with that of black South Africans struggling against apartheid, and the Tutsi dictatorship, recognizing its rhetorical disadvantage, retaliated by branding Hutu activists as “tribalists.” As the government’s slaughter of Hutus continued, the Burundian Mission at the UN circulated an Orwellian document called The White Paper on the Real Causes and Consequences of the Attempted Genocide against the Tutsi Ethnicity in Burundi, which argued baldly that Hutus had been attempting genocide against Tutsis.3

Of course, instead of using the word “Hutu,” the White Paper preferred to describe the “enemy” as “criminals,” “evil forces,” “rebels,” and “gangs,” and it used the non-ethnic discourse that it could expect to play well in the West:

Tribalism was unknown before the arrival of the whites…. Before colonialisation, our society had reached a degree of cohesion and national unity that many European countries lacked…. In a few years you [i.e., Europeans] destroyed the secular product of our ancestors. You distinguished between the Burundese citizens libelling [sic] them as Hutu and Tutsi. You did not stop there. You convinced the Hutu of the necessity of massacring Tutsi.

Before the year was out, the spirit of the White Paper was encoded in laws which prohibited all mention of identity labels—Hutu and Tutsi, majority and minority, etc. Thus, Lemarchand writes, “Tutsi regimes removed the critical issue of ethnic hegemony and discrimination from legitimate debate.” Burundi was by then a one-party state; every citizen belonged to the party, everyone in government belonged to the party, ergo a Burundian was a Burundian, no more, no less.

Lemarchand cites the case of a Burundian graduate student at an American university in the 1980s who refused an assignment to write about the historical “relationship between pastoralists and agriculturalists” in the Great Lakes Region because, he told his teacher, “such a paper would be illegal in Burundi.” Yet if the big lie that the Burundian government was ethnically neutral served the Tutsi supremacists’ myth that they alone were fit to rule, it simultaneously inspired Hutu ideologues in exile to assert their identity as a “martyred community” by revising the rhetoric of the Hamitic myth to subvert the logic of the White Paper.

On first glance it may seem bizarre that Hutus who had just run for their lives from Tutsi violence would wind up articulating the very colonial race theories that had caused them so much suffering. But the tactic fits the classic revolutionary logic explained so succinctly by the character Gisors in André Malraux’s La Condition Humaine:

A civilization becomes transformed, you see, when its most oppressed element—the humiliation of the slave, the work of the modern worker—suddenly becomes a value, when the oppressed ceases to attempt to escape his humiliation, and seeks his salvation in it…seeks in it his reason for being.

Lemarchand’s account of the Hutu counterreaction in Burundian politics relies heavily on the work of Liisa Malkki, an American anthropologist, now a professor at the University of California at Irvine, who came to the Hutu-Tutsi conflict long after violence had become its defining feature. As a Ph.D. candidate at Harvard, Malkki spent the academic year 1985-1986 among Hutu refugees who had fled Burundi during and after the massacres of 1972 and settled in the Kigoma region of western Tanzania. Malkki lived among some 35,000 Burundian Hutus at the Mishamo refugee camp administered by the Tanzanian government and several international relief agencies. The huge camp was made up of sixteen villages spread over an area the size of Zanzibar (about 2,050 square kilometers).

Mishamo was widely regarded as a model camp; the refugees had established themselves as subsistence farmers, and were soon producing a lucrative surplus of such staple crops as rice and sorghum. But Mishamo’s isolated physical setting was singularly forbidding: carved out of a dense, virtually uninhabited forest infested with tsetse flies, it was hours by car to the nearest town, and during much of the rainy season it was entirely cut off. By way of comparison, Malkki also spent some time in Kigoma Town, an ethnically mixed Tanzanian mercantile center south of Bujumbura on the eastern shore of Lake Tanganyika, where a sizable population of Burundian Hutus had settled, quite independently. In both places, Malkki spent her time asking Hutus to account for themselves. Her basic question was: Who do you think you are?

Not surprisingly, the camp refugees and the town refugees had completely different responses, and the difference underscores what must be obvious to anyone who has ever visited any of the refugee camps that ring the troubled regions of the contemporary world: life in camps tends to breed the extremist politics that individual refugees would often be eager to escape. Malkki’s town refugees themselves regarded the camp as a kind of political prison, an extension of the conflict they had fled from in Burundi. Their story was much closer to that of classic migrant groups, who enter a new country in the hope of never again having to encounter their old-country woes. They were so eager to assimilate into Tanzanian society that they were often reluctant to identify themselves as Burundians, let alone as Hutus. The Mishamo refugees, on the other hand, were defined by their isolation in the camp and thought of themselves almost entirely in terms of their past. They had no opportunity to develop any other identity.

To Malkki, the Mishamo population was an ideal laboratory in which to explore her notions of “refugeeness” and “Hutuness,” and she quickly found that her many informants had a tendency, when asked for their stories, to explain themselves in sweeping cosmological narratives, which reached from a time before recorded history to their present moment of displacement. In these narratives Hutus figured as a noble and violated people and Tutsis as the embodiment of evil.

What have they [Tutsis] stolen from us? First of all, our country. The Tutsi are of Nilotic provenance. They came from Somalia. And then [they stole] that which exists in a country—the livestock, cows, chickens, domestic animals,…let us say, the living things—even the birds, the fish, the trees, the banana fields, whatnot…. All of the wealth of the country, you understand, was ours. Because we were the natives of the country.

Collecting such stories from diverse informants, Malkki began to recognize so many similarities of theme and detail that she concluded that she was hearing one larger story—the narrative of Hutu “collective identity” as it was understood in Mishamo. She calls this larger narrative “the mythico-history,” a sort of morality play “constructed in opposition to other versions of what was ostensibly the same world, or the same past.” (By “other versions,” Malkki appears here to mean partcularly the official line of the Tutsi-dominated Burundian government at the time, as presented in the White Paper.) Regrettably, she makes no effort to place her narrative within Hutu or Tutsi folkloric tradition, or to trace the diverse narrative influences—other than the Hamitic myth—that converge in the refugees’ stories. Malkki is concerned simply with the way that, in the fourteen years between their harrowing flight from Burundi in 1972 and her visit, the various refugees she spoke with at Mishamo had come up with a remarkably consistent story—one that tried to explain their devastation and assert their righteousness as victims of ancient ethnic antagonisms.

The Mishamo “mythico-history” begins “one hundred years before Jesus Christ,” with an ancestral figure named Burundi, a Bantu from the Central African Republic, who got lost while following his hunting dogs and wound up in the country that would come to bear his name. Burundi himself was not a Hutu; according to the story, that name, like all things destructive of the true Burundi nation, was introduced by the Tutsis, who came in the sixteenth century—and to Tutsis, the word “Hutu” meant “slave.”4 Tutsis are referred to as Hamites, a group of conquering invaders who ensnared the Hutus into bondage by giving them cattle to tend. “Once the Hutu receives a cow from the Tutsi, all his descendants have to work for this Tutsi,” according to the mythico-history. And so it went, until the arrival of the Belgians, who appear in the narratives as saviors of a sort:

The Belgians were just; they were educated. They were just without regarding race [sic]! If a Tutsi had done something bad, he shall be punished. If a Hutu has done something bad, he shall be punished equally. The Belgians, they kept the Tutsi in the truth.

That Hutus should celebrate the Belgians, who considered them an inferior and subordinate race, is entirely consistent with their embrace of the Hamitic fantasy. Thus Tutsis, according to the Mishamo refugees, are “beautiful,” much more beautiful than Hutus. Tutsis are tall, thin, frail, and astoundingly lazy boozers who “wish only to engage in the secret art of statecraft, administration, and chiefly duties.” Hutus by contrast describe themselves here as energetic workaholics with an aversion to drink. “They [Tutsis] do not do anything,” one of Malkki’s informants reported. “They want that we, the Hutu, be the workers…. We are their granaries. We call them insects; they do not work; they just live, like insects. They eat our sweat.” Worse yet, Tutsi set their beautiful and lethargic women out as lures to trap Hutus into servitude. And woe to the Hutu who takes the bait. “The wife will be lazy while her husband will work, work, work, provide, provide, provide!” Furthermore, a Hutu married to a Tutsi can expect to be among the first killed whenever there is a massacre.

The complex of physical and behavioral traits becomes most meaningful in the Mishamo refugees’ narrative when Tutsis are identifying Hutus for slaughter. Malkki, who has published her dissertation as a book without translating much of the unfortunate graduate-student jargon into recognizable English, calls this “The Concrete Enactment of the Body Maps and the Life/Death Axis,” an absurdly pretentious title for the chilling narrative she presents as evidence:

There are symbols for recognizing a Hutu…. In the hand of the Hutu there was an “M.” Like this, you see? [Informant shows the lines forming an “M” on his palm.] Between the hands of the Tutsi and the hands of the Hutu there is a difference…. Like this they were able to recognize him who was the Hutu. But there were also other symbols…. The Tutsi do not have these bones…. [Informant bends to pull up the leg of his trousers, and points to his ankle bones]…. The Tutsi have a straight line here…. One observed above all the gums…. Especially the Tutsi have black gums. Exactly the majority of the Tutsi! Sometimes a few little red parts, but otherwise black…. For the Hutu, the gums and the tongue are always red or pink. Also, there are Hutu who have had the good fortune of having black gums and all the Tutsi symbols. Like this they have not been killed.

The list of Hutu and Tutsi attributes goes on to cover manners of speech and of walking (Hutus fast, Tutsis slow—“like he who goes where he does not want to”). Once a Hutu has been identified, the narratives gruesomely record the varieties of death dealt out by Tutsis:

With the sisters [nuns]—you are familiar with bamboo?[…]They were split into two parts—of the length of 1.80 meters, 1.90 meters, or 2 meters, if you will. They were prepared with machetes, until the bamboo was pointed like a nail. So, a Hutu is placed on the ground. The bamboo is pushed from the anus to the head. It was like that they did—to sisters, or padres—or to pastors. There are two other fashions. There are large nails, six centimeters, long, fat. It is planted with a hammer. The nail is placed on the head of a Hutu. Once hit, it begins to penetrate the head.

This list, too, goes on. For pregnant women, there is disembowelment, and sometimes mothers are made to cannibalize their fetuses. Fathers and daughters are roped together in sexual positions, then dragged behind a car. People are roasted by a fire until they burst. Bullets, of course, must be used only very sparingly because in this story everything a Tutsi does is secretive, and every effort is made to hide his evil from the prying eyes of the outside world. The Hutus, meanwhile, are presented as the most docile people imaginable:

We are the descendants of Israelites. History says so…. We are the first Bantus who came to Africa. We are Bantus, thus Israelites…. This demonstrates that we had a democracy [before the Tutsi came]. It was good. There were no distinctions of race.

The narrative of these refugees is clearly the product of people who have endured extreme horror, and whose minds have been poisoned by the experience. Yet one wonders who, exactly, were Malkki’s anonymous informants? Once in a while she mentions that somebody at Mishamo will say “our party,” and she tells us toward the end of her book that one of her contacts happened to be Gahutu Rémi, the founder and president of the Parti pour la Libération du Peuple Hutu (Palipehutu). But she does not explain that Palipehutu was, in fact, an extreme expression of Hutu nationalism, which originated in Mishamo in 1980, and which the Burundian regime regards as an illegal rebel organization. This is a surprising omission. According to Lemarchand, Palipehutu held up “the vision of a Rwanda-type Hutu-dominated republic,” and was considered by many Hutus as well as most Tutsis as promoting an unwise and dangerously exclusivist ideology. Malkki’s failure to examine the role that such a potent political force might have had in shaping the narratives she collected at the Mishamo camp calls into question the authority of her reporting.

I have visited close to a dozen camps, in Tanzania, Zaire, and Burundi, inhabited by Hutu refugees from Rwanda, and in all of them virtually everybody I met told me an almost identical version of the history of Hutu-Tutsi relations. While the history put forward by the Rwandan Hutu refugees bore some similarities to that of the Burundian refugees in Mishamo—the heavy Hamitic overtones, and the blame on Tutsis for their country’s woes—the situations of the two groups could not have been more different. Rwandan Hutus, after all, had just committed the most extensive genocide since the Nazi Holocaust. The refugees in the camps I visited last year were under tremendous political pressure, and often direct physical threat, to subscribe to the propaganda of their camp leaders, who were determined to cover up their own crimes. The Burundian Hutus whom Malkki found in Mishamo were, of course, in quite the opposite position, having fled the extermination of 1972. They did not need to mythologize their victimization; the simple facts would have done quite well in most cases.

Yet while much of what Malkki was told at Mishamo was probably propaganda fed to her informants and to herself by Palipehutu activists, it may still be an accurate portrait of what many Hutus believe. Certainly some of their descriptions of recent massacres correspond to the reports of more dispassionate witnesses. But while Malkki is careful not to present her narrative as the truth, only as the truth according to the Hutu refugees, the historical distortions call many of its claims into question.

It is saddening to have to doubt a people’s account of its sufferings, but when, for example, the town refugees Malkki interviewed describe commuting back and forth to Burundi for repeated visits, one wonders why informants at the camp say that every Hutu returnee has been killed. Malkki leaves such matters unexamined. Don’t wonder, she says.

The more challenging approach to such narratives, in my view, is not to sort out “true facts” from “distortions” but to examine what is taken to be the truth by different social groups, and why.

Proceeding from this premise, however, Malkki is unable to reckon with the profoundly political nature of her material. By resisting the challenge of sorting out those portions of her informants’ stories which must reflect actual events from the larger web of historical falsehoods, Malkki seems to reduce history to the status of legend, and to elevate legend to the status of history.


Lemarchand regards the kind of extremist ethnic identity narratives that Malkki presents as “little more than caricatures of a more complex reality,” and maintains that the notion that “ancestral enmities and atavistic hatreds” have caused the present conflict “runs counter to every shred of historical evidence.” What matters is simply that with each new bout of killing more Burundians come to believe in such forces. Indeed, while there is nothing in the story of pre-independence Burundi that leads one to expect the present violence, in the past thirty-five years the enmities have become almost literally ancestral, as cyclical killings carry an ever-heavier burden of personal and collective revenge.

Around the time when Malkki left Mishamo in 1987, a Tutsi general named Pierre Buyoya ascended to the presidency of Burundi in the traditional fashion—by ousting the sitting president, Jean-Baptiste Bagaza, who was Buyoya’s cousin and who had himself taken office in a coup. Buyoya, however, quickly made it clear that he would not stand for business as usual. He struck the laws prohibiting reference to ethnicity from the books, freed hundreds of Hutu political prisoners, brought several corrupt Tutsi officials to trial, and—helped by the Tanzanian government—he called for the return of Hutu refugees. Buyoya’s reforms proceeded haltingly at first. Tutsi extremists and chary Hutus alike suspected treachery. In August 1988, a confrontation between Tutsi communal administrators and Hutu civilians in northern Burundi led to violence, and Hutu mobs began slaughtering Tutsis indiscriminately. As ever, the army’s retaliation was overwhelming. Amnesty International describes “reprisals aimed at the Hutu civilian population as a whole and carried out to punish and eliminate them rather than just to restore public order,” and Lemarchand reports an estimated death toll of fifteen thousand Hutus, while an additional fifty thousand Hutus fled to Rwanda.

Lemarchand attributes the 1988 killings to the mutually provocative positions of extremist Tutsis and Hutus, whose shared interest in creating instability in order to justify their exclusionary politics now seemed to converge as much as to collide. “Anticipated violence,” he writes, “inexorably led to preemptive violence.” Still, Buyoya continued to pursue his reformist program, and for a while in the early 1990s the news from Burundi was promising. The country’s first popular presidential election was held in June 1993. Melchior Ndadaye, a Hutu, won the election, and Buyoya packed up to leave office. At Ndadaye’s inauguration in July, the outgoing president embraced his successor. Then, one hundred days after his inauguration, President Ndadaye was assassinated by Tutsi military men.5 Hutus throughout the country rose up against their Tutsi neighbors, and the army launched a counter-offensive. Lemarchand, who was in the middle of writing the preface to the first edition of Burundi, announced, in italics:

As this book goes to press, Burundi is again lurching into a grim spiral of ethnic violence on a scale reminiscent of the 1972 carnage…, an ethnic butchery in which Hutu and Tutsi seem to have reached equal status….

How an all-Tutsi army can be made to coexist with a predominantly Hutu government will remain the central issue of Burundi politics for the foreseeable future.

Certainly that remains the central issue in Burundi to this day. At least fifty thousand Burundians—Hutus and Tutsis—were killed, and hundreds of thousands fled into exile, or were internally displaced, in the four months after Ndadaye’s assassination, and although a semblance of order was restored, the fighting between the army and Hutu insurgents—now allied with Rwandan Hutu extremists from the refugee camps in Zaire—has never stopped. At least fifty thousand more Burundians have been killed in the past two years, and the number who have fled into exile or been internally displaced hovers around half a million. The great majority of the dead are non-combatants. The killing is often hand to hand and characterized by mutilation; homes, and sometimes whole villages, are sacked and burned. While Bujumbura is now a besieged Tutsi enclave with no route of escape except across Lake Tanganyika, much of the rest of the country is under the control of Hutu insurgents. Most rural Tutsis are living in displaced-persons camps around army bases. In a country where 90 percent of the population lives by subsistence farming, and foreign exchange is earned by the export of coffee and tea, the economy has collapsed.

In a New Year’s address to the nation, Burundi’s Tutsi prime minister, Antoine Nduwayo, noted how sabotage and neglect have devastated “some basic functions: health, education, employment, security, salaries, etc.” And he warned of worse to come: noting that “the ideology of exclusion and genocide is gaining ground,” he urged that everyone “must be ready to make sacrifices in the common war effort.” The figurehead president, Sylvestre Ntibantunganya, a Hutu, offers no greater hope, and with good reason. According to United Nations and Western diplomatic sources, he has spent most of the past half year under imminent threat of being ousted by Tutsi oppositionists, allegedly backed by the military. Indeed, observers tend to regard the beleaguered civilian government as barely functional; state power remains with the military, and there is an atmosphere of violent intimidation for which Tutsi youth militias, who have army support, are as responsible as Hutu rebels. Indeed, those who speak of an imminent genocide in Burundi seem at a loss to explain which group is expected to carry it out.

Perhaps when people speak of Burundi as a Rwanda in the making they also mean that nobody in the outside world is doing anything to stop the killing. The UNSecretary General, recalling the UN peacekeepers’ desertion of Rwanda before the machete-wielding mobs, warned in February that “preventive diplomacy alone may no longer be sufficient” to save Burundi. He is now recommending that the Security Council prepare a contingency force to intervene there, a proposal supported by the Clinton administration. What is missing from the discussion is not only an idea of how bad it has to get in Burundi before this hypothetical force would be deployed but, far more importantly, what its objectives would be. Not surprisingly, no country has yet volunteered troops for a Burundi mission.

Indeed it is doubtful that a UN-style military intervention would resolve Burundi’s woes. The country’s Hutu and Tutsi leaders alike have protested the idea, promising that foreign troops would meet violent resistance, and a number of foreign diplomats in Bujumbura believe that intervention could provoke the extremist belligerents and trigger a disaster. One African diplomat who has followed the situation closely says, “If you do send troops, they”—the Burundians—“will just kill each other before you get there.” As it is, they are killing each other at a rate of at least one hundred a week.

We all inhabit ‘constructs’ of the world,” V.S. Naipaul wrote. Lemarchand and Malkki would agree. The suggestion implicit in their assessments of Burundi is that anyone who hopes to make sense of the place—let alone tries to make peace there—must begin by acknowledging the authority of the myths according to which Hutus and Tutsis have constructed a mutually uninhabitable world. But it is hard to imagine how a recipe for political sanity can be sought in such madness.

Twenty-five years ago Lemarchand described Burundi as “a cultural environment in which concealing or distorting the truth [is] traditionally regarded as both a virtue and an art.” And so long as Hutus and Tutsis continue to live in their mutually exclusionary narratives, the tragedy there will continue.

“Burundi,” Lemarchand writes, “has become one of the African continent’s most intractable basket cases.” There is no disputing him. Colonial race policies were the original divisionist sin, but most of the blame since lies with the extremist politicians who have used violence to make the myth of ancestral conflict into a contemporary reality. Burundi needs a new account of itself, one that both Hutus and Tutsis can accept, without the lies that inspire bloodshed. When Naipaul writes about our “constructs” of the world, he says, “Every culture has its own: men are infinitely malleable.” If there is hope in this remark, as well as despair, that hope requires the sort of assertive sanity that is nearly impossible to locate on the Burundian political scene these days.

In March, when Tutsi soldiers from the Burundian army attacked a gathering of displaced Hutu civilians in retaliation for a raid by Hutu militias in Zaire, another group of army soldiers turned against their colleagues to defend the Hutus. It was the first instance, at least in recent memory, of internal fighting in the Tutsi military. But what were those men who protected the Hutus fighting for? What were their politics? What kind of Burundi did they imagine that made them defend as countrymen the people they were called upon to attack? We may never know; all of the mutineers were killed, along with dozens of the men, women, and children they had sought to protect.

—Kigali, May 9, 1996

This Issue

June 6, 1996