On Exile and Postcolonial Nationhood in Rwanda and Burundi

Burundi 2015/1972

The crisis began after the president of Burundi, the late Pierre Nkurunziza, decided to run for a third term. Segments of the population contested this choice, particularly in the capital city of Bujumbura, where Claude lived. Nkurunziza had come to power in 2005 by parliamentary nomination, with the Burundian parliament acting as an electoral college. He was the nominee of the Conseil National pour la Défense de la Démocratie-Forces pour la Défense de la Démocratie, the political party that gained a majority in the national elections of the same year. According to protesters like Claude, by running a third time in 2015, the president violated the two-term presidency clause in the civil war’s 2010 peace agreement that had brought him to power. This agreement stipulated presidential terms be limited to two five-year terms. Nkurunziza justified his 2015 run by exploiting a loophole, stating that his first term had in fact not resulted from a national election, but from parliamentary nomination.¹ This argument, compounded by documented targeted attacks on journalists and radio channels, the manipulation of election results, and general discord about the management of the country, did not sit well with the civilian population, particularly in Bujumbura, and people took to the streets. For a little over a month, the streets of the capital filled with demonstrators against the president’s campaign, using the slogans “Sindumuja” (I am not a slave) and “NonAu3emeMandat” (No to the 3rd term).² The government responded violently to the protests by jailing and kidnapping the perceived dissenters.

Claude, however, insisted that the situation in Burundi was in fact “revenge from the past . . . 1972, that is what they like to say often,” he remarked in one of our encounters. “Hutus were killed; all [Tutsi] governments always discriminate against Hutus” was the constant refrain hurled at people like him. I asked him who said this—all Hutus? “Not all,” he answered, “only those in power; the others we don’t know, they don’t have a way of showing it yet.” I first met Claude in Rwanda in July 2019 on my way to the eastern town of Nyamata. We were seated next to each other on the bus, and I asked Claude if he was going to Nyamata. He didn’t hear my question and responded, “sa?” I repeated the question but also asked if he was Burundian. He was indeed going to Nyamata and yes, he was Burundian and asked how I knew. “Your use of sa gave it away,” I told him jokingly. This word, a shortened version of sabwe and generally used by Kirundi speakers, can be used to signal presence or to signal incomprehension, as in the expression pardon me? Claude was following a lead on a potential job. Intrigued, I told him who I was and my reason for being in Rwanda (doing research on post-1994 Rwanda), and asked about a potential collaboration. He agreed, but with caveats. We opted for semi-structured interviews at Claude’s request. I delineated possible ways of collaborating: formal interviews with predetermined questions, semi-structured interviews, or regular ethnography, given that I already lived close to his neighborhood of Gikondo in Kigali city. Claude, unlike Muzehe, refused the classic format of the ethnographer following him around. We would meet regularly over a period of three months after our initial encounter. Our meetings, one-on-one in quiet restaurants, lasted several hours. In these meetings, we agreed that Claude would update me on his job hunt and news of political meetings about the situation in Burundi. He would share what he deemed sensitive information only under these circumstances. His life felt precarious enough as it was. I respected this request and took what he told me, his own narration of his life in Rwanda and his nostalgia for home in Burundi, at face value, but with the recognition that his retelling was exactly that, a retelling, and thus in part mediated through his own interpretation of the events in his life.

“I didn’t say goodbye to anyone or pack anything,” Claude explained. In March 2016, he took a bus from Bujumbura for the southern border with Rwanda and never looked back. Claude would later reveal that he had in fact also fled to Rwanda in 1998 and again in 2003. His first stop in 2015 was Kigali, where he stayed with a Rwandan childhood friend. The friend in question was born in Burundi to Rwandan parents. This friend had returned to Rwanda in the mid 1990s. Claude’s network in Kigali was constituted, in other words, by individuals also familiar with having to flee. Rwandan families who had fled to Burundi at some point starting with the first wave of political violence in Rwanda in 1959. “Why did you go to Rwanda?” I asked him. There was a network of “old friends and political allies in Rwanda,” he said, and in Kigali in particular, that he could plug into precisely because of this postcolonial history of displacement. Claude’s other network comprised Burundian refugees like him, who arrived in Rwanda beginning in 2015.

Haunted Presents

Ikiza (a grand calamity) is the 1972 event in Burundi that Claude referenced as the historical antecedent to his escape. Historians are divided on what exactly transpired in 1972. To some, this event constituted a “selective genocide” that targeted an educated Hutu class and was motivated by decades of repression perpetrated by a predominantly Tutsi elite. Others, perhaps mindful of the legalities of a morally and politically loaded term, speak of “proximity to genocide” or simply retain the Kirundi word, ikiza (Chrétien and Dupaquier 2007). What is certain is that on April 29, 1972, an insurgency against the regime of then president Michel Micombero, who had been in power since independence in 1962, started in Nyanza-Lac and Rumonge near Bujumbura and rapidly escalated to other parts of the country. “Roving bands of Hutu attacked civilians,” Lemarchand writes, and “countless atrocities were reported by eyewitnesses” (Lemarchand 1996, 4). The decisive factor behind the insurgency was President Micombero’s “rigid discriminatory policies” against Hutus (Lemarchand 1996, 4). The government met these attacks with a powerful suppression of the civilian population. Within three months, between 100,000 and 200,000 Burundian Hutus had been murdered. Thousands fled the country for refugee camps and cities in neighboring countries—“no sector of society was left untouched” (Lemarchand 1996, 6). President Nkurunziza, the key protagonist of the 2015 crisis, also lost his father to armed repression in 1972. He later, in the early 1990s, fled during yet another civil war, and returned in the 2000s.³

The 1972 massacres and their consequences have haunted Burundian politics ever since. Like Claude in 2015, decades before, thousands of families found themselves in the same situation. Liisa Malkki’s Purity and Exile tells the tale of some of these families in the refugee camp of Mishamo and the town of Kigoma in neighboring Tanzania in the 1990s. In Malkki’s account, Burundian refugees apprehend the reality of exile differently depending on whether they live in the camp or town. In the camp, the history of Hutu exclusion in Burundi was extended beyond politics to create what Malkki (1995a) calls a “mythico-history” and a cosmological order in which Tutsi signify moral corruption and Hutu-ness is synonymous with “purity.” The camp served as a “place of purification,” transforming its residents from servants of Tutsi in Burundi to refugees in a foreign land, to final ascension into a collective free Hutu people. For those in town, the past, present, or future were not “firmly localized in Burundi or anywhere.” This “challenging cosmopolitanism” suited the interests of the refuges in Kigoma town. They juggled multiple identities, preferring to be Burundians rather than simply Hutu (Malkki 1995a, 161). For this reason, Malkki insists, the town refugees did not find themselves in a “state of exile” per se, because their narratives did not insist on collective return to a “homeland, an origin or a proper place.” The “mythico-history” in the camp was “replaced by a partial, potential immigrant mythology” (Malkki 1995a, 193). Though both groups share “the fact of exile,” the cosmopolitan town refugees denied this history and past. The rooting in Burundi signaled a kind of “parochialism” (Malkki 1995a, 233).

While Malkki emphasizes the differences between those in the camp and those in the town, I would like to underscore their shared expulsion from nationalist citizenry. Though resulting in different “mythologies,” both groups share a fundamental and significant factor in their life as political subjects: because of what they are perceived to be (Hutu), they have been targeted and expelled from their political community and in exile must cope repeatedly with what this means as political subjects. Political violence erupted again in Burundi in 1988, 1993, 2001, and finally in 2015, when Claude fled. Moreover, as Malkki’s (1995a, 277) postscript to the book illustrates, in 1987, the Tanzanian government re-signified the cosmopolitan town refugees as “illegal immigrants,” targeting them for “repatriation” to Burundi. A few years later, in 1992, the Tanzanian government gave the Burundian exiled “three months” to choose between two options: repatriate to Burundi or naturalize in Tanzania. That is, it demanded them to anchor into one place, naturalizing here or there, but in one place. This choice resulted from political negotiations between the governments of Burundi and Tanzania (Malkki 1995a, 277).

This political landscape that insists on the fact of exclusion and its corollary nationalist citizenry makes it difficult if not impossible to maintain an “immigrant mythology.” The nation (its meaning and content) remains the hovering overarching narrative with each episode of violence and exclusion. The camp site and the town share a history of repeated marking of belonging and exclusion from it. Both groups of exiled may appear to live “the consequences differently” (Malkki 1995a, 234), but they can just as simultaneously be interpellated back into the “national order of things” despite their individual desires. The cosmopolitan path is contingent on larger political structures and the state’s attempts at capturing and claiming national stability by reinscribing a rootedness into one place. Both town and camp eventually became locations from which both groups were marked as political subjects “out of place,” with an “ultimatum” to repatriate or naturalize (Malkki 1995a, 277). Once more, the people in exile were reminded of the nationalist logic precisely because the state logic in Burundi and Tanzania insisted on rooting them in one place, whatever immigrant mythology they desired.

Originary Fragmentation

These movements to and from the postcolonial nation merely ten years after political independence from colonial Belgium make exile a constitutive aspect of postcolonial political life. Exile for the refugees in the camp marks one step toward their ascension into a collective Hutu people and the regaining of their “homeland” as free political subjects. For the people in Kigoma town, the experience of exile makes them itinerant Burundian migrants, rather than simply Hutu subjects, perhaps for a fleeting moment until they become re-signified again as Burundians—in exile, out of place.

Nationhood here is shaped through and by the experience or memory of repeated flight. The idea of a homeland is somewhat paradoxical. There is at once an attachment to autochthony and a cosmology that remains invested in the existence of a purportedly authentic geography, but this supposed homeland is also constituted since the history of its inception by fragmentation. The postcolonial homeland is here scarred by a repeating caesura, constantly threatening to unmake it, and thus requiring repeated re-imaginings of collective boundaries. These are the terms under which nationhood and belonging are navigated. Malkki’s interlocutors in Mishamo camp and Kigoma town, as well as my interlocutor Claude four decades later—despite their different geographical and temporal locations—share this postcolonial historical reality of originary fragmentation. The nation itself is not a stable political entity but fragmented since/with colonization. That is, there exists an idea of what it is and perhaps was in the past as formulated through political struggles and mythico-history, but each political contestation puts this idea itself at stake. This is what the formulation of originary fragmentation entails. It is not only membership in the political community that is fought over but also the content of what constitutes the nation itself: politically, semiotically, and morally.

The way out of originary fragmentation is political, Mahmood Mamdani (2022) has suggested for decades now, though most emphatically in his latest publication, Neither Settler nor Native. Radical postcolonial reform, he writes, requires “decoupling the nation from the state” (Mamdani 2022, 329). For him, to decolonize the political requires “stripping away the nation, or the tribe as nation,” and replacing it with “the mere state,” that is, a “legal sovereign, with equal treatment of citizens” (Mamdani 2022, 328, 330). This is so because “power in nation-states lies always with those who identify with the nation, not with coalitions that assemble through a political process” (Mamdani 2022, 329). Postcolonial states in Africa inherited a legal system both racialized and ethnicized. Decolonial reform, Mamdani writes, succeeded in “deracializing” the civic sphere by applying civil law, once the domain of the colonizers, to decolonized Africans, but it failed to de-ethnicize. This fact constitutes the core of the “crisis of postcolonial citizenship” (Mamdani 2001, 14).

A consequence of this crisis is precisely what Malkki crystallizes in describing the mythico-history of the Mishamo camp as cosmological ordering. Exile becomes essential in this repeated exercise of reordering the cosmological and political boundaries of the nation. From Tanzania, these exiles will, for decades, renegotiate their relationships to nationhood and citizenship vis-à-vis both Burundi and Tanzania. Some will return to Burundi peacefully as individuals and sometimes flee again. Others will attempt to return forcefully with rebel groups formed in exile. And others yet will return generations later in the 2000s because of a peace negotiation treaty organized with the help of the Tanzanian government (Burihabwa and Curtis 2019). This reordering of the nation does not just happen in individual minds and narrations of the displaced. It finds support not just in people who return but also in political parties and movements formed in exile. The Party for the Liberation of the Hutu People in Burundi (PALIPEHUTU) emerged in 1980 in the same Mishamo camp Malkki discusses. The Front for Democracy in Burundi (FRODEBU), the party that brought to power the first Hutu president in Burundi after democratic elections in 1993, was created in Butare, a southern Rwandan town at the border with Burundi in 1980. Key leaders of the main political party in Burundi today, the CNDD-FDD, have also organized for forceful return while in exile (Burihabwa and Curtis 2019).

Thirty years after Mishamo and Kigoma in Tanzania, through yet another wave of displacement, Claude arrived in Rwanda. “I am lucky; this gray beard saved me,” he noted. “It makes me look older.” The primary targets of 2015 were young adult males, the people in the streets of Bujumbura who most ardently protested Nkurunziza’s third presidential run. But, Claude remarked, “Nkurunziza is trying hard to turn this into an ethnic problem. A large majority of the people disappearing are young Tutsi men between twenty and forty years old.” I asked why. “Revenge from the past,” he responded, reiterating the narrative of serialization vis-à-vis political violence. This seriality, as we shall see, also reverberates in narrations of violence in Rwanda.

Rwanda 1994/1959

The 1994 Rwandan genocide against Tutsi is also narrated as having antecedents in earlier episodes of violence against Tutsi, beginning in 1959. As with the events of 1972 in Burundi, those in 1959 Rwanda also remain contested (Onana 2002; Péan 2005). By the late 1950s, the country offered a tense political scene, with different political parties demanding independence from Belgium. Political independence finally became a reality in July 1962. A few years earlier, in November 1959, a wave of anti-Tutsi violence had engulfed the country, targeting primarily the Tutsi local administrative chiefs favored by the Belgians. Thousands of Tutsi fled the country and decades later some of their descendants began preparing for a forceful return under the umbrella of a guerilla army: the Rwandan Patriotic Front, or RPF. Thus, beginning with the first agitations for political sovereignty, the boundaries of the new nation were contested by people inside and outside of it. Fast-forward to 1994: the RPF—the guerilla army formed in exile by descendants of the exiled ‘59ers—ended the genocide against the Tutsi after a four year civil war.

In power since 1994, the RPF-led government and its state institutions have been centrally preoccupied with the narration of the genocide and the country’s history. In 2011, the National Commission for Unity and Reconciliation (NURC) brought together researchers and historians from the national university to rewrite the history of Rwanda since its inception. The resulting volume describes the genocide against Tutsi in 1994 as the “result of a culture of institutionalized crime based on ethnic racism beginning in 1959.”⁴ In this rewriting, the past also serves as prologue, and the violence of 1959 proves key as the moment of expulsion and the beginning of the project of return. Many state celebrations express this conception of the past. On July 1, 2012, as Rwanda celebrated fifty years of independence from Belgium, thousands lined up before sunrise to enter Amahoro Stadium in Kigali for festivities. The official end of the genocide, when RPF troops took over Kigali in 1994, took place on July 4, and on this day in 2012, the two historical moments were joined. People had gathered to celebrate Imyaka 50 Y’Ubwigenge (fifty years of independence) and Imyaka 18 yo Kwibohora (eighteen years of collective self-liberation). Kwibohora loosely translates as “untying oneself,” in the sense of freeing oneself politically; July 4 is hence liberation day. Ubwigenge means “independence.” Ubwigenge is a noun, and it is passive vis-à-vis Kwigenga (to be independent), the active formulation. At the semiotic level, the choice of these two words indicates the different levels of significance accorded to both events. Kwibohora, or self-liberation, is agentive. The word kwibohora denotes a purposeful action by the self and the collective to escape bondage. July 4, 1994, is marked as the true decolonial moment that creates a new moral order—unlike 1962.

Another formulation was Imyaka 18 Twibohoye (eighteen years since liberating ourselves). Independence from Belgium is inserted into a past prior to real liberation. All events before July 1994, still rife with “ethnic divisions,” do not make for “real independence,” but are rather prologue to genocide.⁵ Today, outside the country, other political groups are yet again contesting the meaning of these resignifications and redrawing the political and moral contours of what Rwanda 1959 and 1994 mean.

In and Out of National Ordering

Claude had chosen to live far from Muzehe’s town and outside the Mahama camp. Kigali offered a wider array of choices given his childhood friends in the city. He missed being in Burundi, however, and often insisted that there was no place like home. He wanted above all to go back to Bujumbura. But he was also clear that “there is no other country we could have fled into,” both claiming a collective “we” and marking an affinity to Rwanda, his place in exile. Pressed to define this “we,” Claude seemed surprised. I should have understood. He gave me a long-winded explanation on the ethnic composition of groups in the region, their kinship relations, and origins. He insisted that what propelled him and the other Burundians to flee was not an ethnic but “a political problem,” an unwanted president taking revenge on his opponents by manipulating pre-existing ethnic antagonisms. The choice of Rwanda, however, was related to what he just told me about kinship relations. Unsurprisingly, Claude’s imagined kinship relations mapped onto the racialized colonial ethnic categories. Without saying it explicitly, he indicated that people like him—read Tutsi—felt welcome in Rwanda because “their kin” were now in power. The ascension to power of the predominantly Tutsi RPF, Mahmood Mamdani (2001, 261) writes, has brought forth a “diasporic state” whose main preoccupation is an “overwhelming sense of moral responsibility for the very survival of all remaining Tutsi, globally.” This logic is in part what Claude’s explanation reveals.

Claude’s we also illustrates the co-existence of multiple forms of political identification: 1) a we that anchors him in a purported homeland and a specific geography (Burundi, like the camp refugees in Tanzania decades prior); 2) a we that claims a politic exceeding this same geography, tying him to a political network in Rwanda today (his childhood friends once exiled in Burundi who have returned to Rwanda; 3) and finally a we deeply familiar with exile (like Muzehe and his family, Malkki’s interlocutors, and countless others). In other words, there is in Claude’s reference to a we a collective of political allies in geographies beyond Burundi. The first level of signification for this referent is imagined ancestral kinship certainly. The mythico-history in Malkki’s work and Claude’s own ethno-historical account of the region indicate this much. This collective is conceived as racial because it maps onto the racialized categories of Hutu and Tutsi. A second layer of this we is the diasporic orientation of the state both in Rwanda and in Burundi, as evidenced by the history of political organizing in exile in both countries. This layer secures nation-state interests by insisting on return and state power. But it also simultaneously, in its diasporic inclination, holds onto an ethno-imaginary beyond the nation-state, one rooted in a mix of an attachment to a homeland and this history of recurrent exile as cardinal features of postcolonial history. In all layers—the we as kinship and the we as national and diasporic—the experience of exile and exclusion from the postcolonial nation prove central. Home here is connected to movement and emplacement, to destination and return, to ethnos and citizenship. This recurrence of movement complicates the fetishism of rootedness as a cardinal feature of ethnic being.

These attachments in and beyond the nation signal a type of politic insistent on the normative call of nation-ness as linked to a geographically bound, political, and collective self. Yet they also gesture to something exceeding it. The national referent is filled by different ethno-racial passions, internal and diasporic ones that exceed the boundaries of nations. The recurrence of collective flight creates these other kinds of attachment moving alongside nationalist projects. Put differently, Claude is Burundian, and that means either fleeing or growing up in the shadow of prior generations who have done so. But Claude also feels part of a collective tied to Rwanda today. This is so certainly because of imagined ethnic bonds—inheritance of colonial law—but also because those in power now in Rwanda understand exile, having fled themselves and returned forcefully to power. Similarly, the late president Nkurunziza in Burundi was imagined as governing based on residual resentment from the 1972 mass exodus of Burundian Hutus by targeting Tutsi families in 2015 and turning political opposition to his presidency into a Hutu/Tutsi problem.

The academic literature about 1972 or 1994 usually focuses on internal dynamics within one country. That is, the historical and political circumstances within contemporary national boundaries that led to political violence and exile (Chrétien and Dupaquier 2007; Lemarchand 2004). Regional analyses, despite the move beyond national boundaries, also focus on nationalist imaginaries by critiquing what they read as expansionist ethno-nationalism. For Rwanda, for example, it becomes a matter of expanding the power of the current RPF government into the region, such as the Democratic Republic of Congo (Prunier 2011). This focus on national boundaries as epistemological units is in some ways trapped in the same logic that critics of methodological nationalism have noted (Getachew 2019; Wilder 2015). For them, to imagine the ultimate outcome of decolonization in 1960s Africa only as national independence and state sovereignty means to miss an important aspect of liberation movements: their “worldmaking” practices, their aspirations to transform the world after empire into a just world from Africa to the Caribbean (Getachew 2019). To be sure, what is stake in Rwanda and Burundi today is in no way close to ambitions of creating a world of political equals and/or other forms of political communities that exceed the nation-state form. In fact, quite the opposite: there are strong political attachments to nationalist imaginaries. Claude misses his “home,” and that is and might always be Burundi, not his place of exile, Rwanda.

Nonetheless, methodological nationalism misses these simultaneous diasporic and nationalists’ aspirations co-existing side by side that result from specific historical conditions and embodied experiences: repeated exile for collectives based on claims to a “national” homeland. This, I have tried to argue, in fact illustrates that the nation is fragmented—and has been since colonization. This originary fragmentation is what the postcolonial governments have repeatedly tried to remediate with rigid enclosures and claims of autochthony. This is another enduring legacy of the worlds engendered by colonialism. The postcolonial evokes autochthony while also being fragmented and exilic. Access to rights for some and exclusion for others are linked to discourses about rightful inhabitants. At the same time, repeated displacements have created forms of political stakes that exceed these same rigid autochthonous boundaries. Malkki’s work shows us one side of this: in exile in Tanzania, the camp refugees narrate their return into Burundi as their rightful return to a homeland. As Burundian Hutu, they belong rightfully to Burundi. The nation emanates from them as autochthonous subjects. Claude’s exile results from resentment about the 1972 violence, which results in the displacement of Malkki’s interlocutors. Burundi for him is also home, however, a home whose origins he narrates as emanating from people like him first—read Tutsi. Yet—and this is the crucial point—his relocation to Rwanda is made possible by Rwandan friends who, because they were once exiled in Burundi, understand his plight and help.