Dr. Augustin Iyamuremye’s report: Impunity foreseen in 1993

The genocidal violence that overwhelmed northwestern Rwanda between late December 1992 and early January 1993 did not erupt unexpectedly, nor did it occur in a vacuum of ignorance. It unfolded in full view of the state, announced in advance through speeches, rumors, administrative meetings, and coded language that had already acquired a deadly political meaning. What happened in the communes of Kayove, Ramba, Gaseke—and Rutsiro was neither spontaneous nor unplanned. It was the predictable outcome of a political climate—purposefully cultivated by those in authority, where dehumanizing rhetoric against Tutsi, administrative complicity, and strategic inaction combined to produce what survivors later experienced as organized mayhem. The report submitted on 6 January 1993 by Dr. Augustin Iyamuremye, then Head of the Intelligence Service, to Prime Minister Dismas Nsengiyaremye stands as a shattering indictment of that complicity. It documents, with dates, places, names, and internal acknowledgments, how much was known, by whom, and how little was done. Written in the sober language of administrative correspondence, it details insecurity, targeted attacks, organized looting, and the systematic persecution of Tutsi populations in the communes of Ramba and Gaseke and Kayove in Gisenyi Prefecture; and Rutsiro commune in Kibuye Prefecture. The report does not speculate; it records. It does not exaggerate; it documents. Yet its significance lies not only in what it says, but in what followed—or more accurately, what did not follow. The president was informed. The chain of command was alerted. No meaningful action was taken. In such circumstances, inaction ceases to be neutral. The events in the report are revisited in granular detail because details matter. Dates matter because they reveal sequencing and intent. Place names also matter, because violence is always local before it becomes national. Names of officials matter because impunity is not abstract; it is administered by identifiable men occupying specific offices. On 15 November 1992, President Juvénal Habyarimana addressed crowds in Ruhengeri, reiterating themes of internal enemies and betrayal. One week later, on 22 November 1992, Léon Mugesera would push this logic to its grotesque extreme in Kabaya. His call to send Tutsi “back to Ethiopia via the Nyabarongo (river)” was not an isolated outburst but the crystallization of a discourse already circulating at the highest levels of power. Mugesera named enemies, identified targets, and predicted future violence with chilling clarity. When such rhetoric goes unpunished—indeed, when it is echoed and normalized—it becomes instruction. These speeches were not rhetorical overindulgences on the margins of power; they were signals emitted from its very center. The two speeches did not merely inflame passions; they supplied a vocabulary, a logic, and a sense of permission. Derogatory terminologies such as inyangarwanda—or haters of Rwanda, inyenzi or cockroaches, and “the enemy” were used interchangeably to strip Tutsi of citizenship and humanity, collapsing civilians into combatants and neighbors into targets. In this atmosphere, local authorities did not need explicit written orders to act. They understood the signal. Words spoken at the center radiated outward, shaping expectations in neighboring communes and sectors. Violence became thinkable, then permissible, then routine. Against this background, administrative decisions such as the organization of an Umuganda on 28 December 1992 in the Gishwati forest, ostensibly for security purposes, took on a far more sinister meaning. What follows is a detailed reconstruction and analysis of how these events unfolded, how warnings were issued and ignored, how state agents acknowledged risks yet proceeded anyway, and how the absence of decisive intervention amounted to endorsement. The purpose is not merely historical. It is moral and political: to demonstrate that mass violence is always preceded by knowledge, and that silence in the presence of such knowledge is never neutral. The where and when of the Violence According to Iyamuremye’s report, the first outbreak of violence occurred on 28 December 1992, the very day scheduled for an improvised Umuganda in the Gishwati forest. That morning, ethnic disturbances began in Gihira cell, Kayove Sector, Kayove Commune, in Gisenyi Prefecture. Within hours, the violence spread beyond administrative boundaries, reaching Kirwa, Karuruma, and Ruhanga cells in Bwiza Sector of Rutsiro Commune, also in Kibuye Prefecture. The same day, attacks were reported in Rwili Sector of Gaseke Commune in Gisenyi Prefecture. The simultaneity of these outbreaks, across different communes and prefectures, already undermines any claim of spontaneous local disputes. On 29 December 1992, the violence intensified and expanded. New incidents were recorded in Boneza and Busanza Sectors of Kayove Commune, in Bayi Sector of Ramba Commune, in Magaba Sector of Gaseke Commune, and in Remera cell of Gitebe Sector, Rutsiro Commune. The report emphasizes that “everywhere, these were attacks launched by groups of Hutu who burned or destroyed the houses of Tutsi, especially the Abagogwe, slaughtered or looted their cattle, and seized their belongings and food reserves.” This is not vague language. It establishes perpetrators as “groups of Hutu”, victims as “Tutsi, especially Abagogwe”; and acts involving arson, looting and destruction of livelihoods. These acts are not random vandalism or crimes. It was a targeted economic and social annihilation. They correspond exactly to what international law later recognized as methods of ethnic destruction. By 30 December 1992, the human toll had become undeniable. The report records two deaths in Magaba Sector, Gaseke Commune, on that date, and one death in Rwili Sector on 28 December, also in Gaseke. These figures, the author notes, were provisional, as full assessments were difficult due to the sporadic yet widespread nature of the attacks. In Kayove alone, the first day saw five houses burned in Kayove Sector, damage affecting seven households in Boneza Sector, and two households in Busanza Sector. In Gaseke, in addition to the deaths, numerous houses were burned and several cows looted. In Ramba, cattle theft predominated, while in Rutsiro, particularly in Bwiza Sector, clashes resulted in three injured persons, the looting of approximately ten cows and one hundred goats in Gitebe Sector, and the destruction of at least one house. The Umuganda trap The report traces the genesis of these events to 23 December 1992, when Bourgmestre Isidore Maburakindi of Kayove Commune toured the commune to mobilize the population for the Umuganda scheduled for 28 December in the Gishwati forest. During a meeting at the Gihira market in Kayove Sector, attended by many Abagogwe residents due to their proximity, Maburakindi stated that “all men would go to help the soldiers from Gisenyi search the Gishwati forest to look for ‘inyangarwanda who might be hiding there.’” When asked later, Maburakindi admitted having used this language, justifying the operation by citing insecurity in the forest in October 1992, when four unidentified corpses had been discovered. No investigations about these deaths followed. Those corpses, the Iyamuremye’s report notes, had been buried by the commune, and the matter reported to the Prefectural Security Council, which decided on a forest search. Crucially, Maburakindi added that the Council, fearing that the population might engage in looting or massacres of Tutsi—as had occurred during a previous Gishwati operation in 1991—decided that soldiers should participate in the Umuganda to “effectively supervise” the population. This admission is one of the most damning elements of the report: the authorities anticipated violence, anticipated massacres, and yet proceeded with a mobilization that predictably produced exactly what they claimed to fear. Following the Gihira meeting, rumors spread quickly that the Umuganda was a cover for massacring Tutsi Abagogwe. After this meeting, a rumor spread, reported verbatim by the intelligence service. There were Hutu residents who openly celebrated, declaring “abategetsi batanze ubunani.” In English translation, the statement was: “The government authorities have given us a New Year gift.” This was not symbolic. It was material. The “gift” was the permission to loot Tutsi property—cows, crops, houses, and food reserves—just days before the New Year. The report specifies that Hutu were rejoicing because they anticipated economic gain from violence. The reference to “abategetsi batanze ubunani” reveals another fundamental dimension of the violence: material reward. The looting of Tutsi cattle and crops was not incidental; it was integral. By allowing perpetrators to enrich themselves, authorities transformed violence into opportunity. Participation promised not only ideological satisfaction but tangible gain. This dynamic deepened popular involvement. Ordinary people, drawn in by the prospect of loot and reassured by official indifference, became complicit. Violence spread not because everyone was ideologically radical, but because the costs were low and the benefits immediate. “Abategetsi batanze ubunani” is a crucial insight: genocidal violence was incentivized, not merely ideologized. Dispossession was a motive and not a byproduct. On Sunday, 27 December 1992, three Abagogwe emissaries from Kayove—namely—Kidonga, Nyagasaza, and Buhirike—went to see Bourgmestre Maburakindi to express their fears, and explicitly questioned the planned Umuganda. Maburakindi later confirmed that the meeting took place, claiming he dismissed the concerns as fabricated rumors. This establishes foreknowledge beyond doubt. Yet, when Abagogwe who had fled to the Rutsiro communal office stated, in Maburakindi’s presence, that they had never seen their emissaries again, his response was disturbingly elusive: “If you have not seen them again, perhaps they fled... fled elsewhere you do not know.” This statement is revealing for two reasons: One, it shows indifference to the fate of people who had just warned of imminent attacks. Two, it suggests a normalization of disappearance in a context where criminal violence was already underway. That same Sunday afternoon, Dr. Iyamuremye reported, the military unit tasked with supervising the Umuganda arrived in Kayove from Gisenyi. That evening: “The Bagogwe began entrusting their cattle to Hutu friends and fleeing.” They fled to the Rutsiro communal office. Flight, and not protection—became the only rational survival strategy. People do not flee civic work. They flee planned attacks. State failure or complicity? By 29 December 1992, 124 Abagogwe—38 from Kayove and 86 from Rutsiro—were sheltering at the Rutsiro communal office. Those from Rutsiro only returned home or sought refuge with neighbors on 30 December, after the Prefect of Kibuye personally visited Rutsiro to guarantee their safety, accompanied by a military deployment. The contrast between this intervention and the inertia elsewhere is blatant. Meanwhile, on 30 December 1992, thirty-seven Abagogwe from Bayi Sector in Ramba Commune gathered at the home of Sebukufi in Remera cell, Gitebe Sector, Rutsiro Commune. That morning, attackers from Ramba launched an assault, looting and destroying the house of Munyanganizi while the Abagogwe watched from a short distance, powerless. A second attack followed on 31 December around 2:00 p.m., this time involving an estimated three hundred assailants. Only the combined resistance of the Abagogwe and Hutu residents of Gitebe Sector managed to repel them, leaving many wounded. Iyamuremye concludes that the relatively limited loss of life was due not to effective state protection but to the foresight of the Abagogwe, who regrouped and fled before the violence erupted. He emphasizes that attacks were carried out by groups that sometimes encountered resistance from both Hutu and Tutsi civilians, highlighting that genocide is always a political project imposed on society, never an inevitable expression of it. The report raises profound questions about timing and intent. Why, after two months without incidents in Gishwati forest, was a search suddenly organized with only five days’ notice? Why involve the civilian population in a security operation, knowing full well the risks, instead of deploying professional forces alone? The author notes that any supposed malefactors would have had ample time to flee. What did not flee were civilians marked as enemies. When a Head of State is informed of targeted violence against a specific group and chooses not to act, cannot be interpreted as ignorance. In law, in ethics, and in political reality, silence after notification constitutes endorsement. That inaction functions as authorization, reinforcing the belief among perpetrators that they are protected. It was very clear. No emergency directives were issued. No administrators were suspended. Killers were not punished. Looted property was not restored. No perpetrators were publicly arrested or condemned. This was not political paralysis. It was acquiescence. The limited number of deaths recorded was not due to state protection, but because: victims fled in advance, and in some areas, Hutu neighbors resisted attackers. Here, survival depended on chance and solidarity, not law. What happened in Kayove, Ramba, Gaseke, and Rutsiro demonstrated how language, administration, and impunity interact to produce mass violence long before its most catastrophic expression. The names of Gihira, Kirwa, Karuruma, Ruhanga, Boneza, Busanza, Bayi, Magaba, Rwili, Remera, and Gitebe must be spoken because they are not footnotes; they are early chapters in a story that culminated in 1994. The names of officials—Isidore Maburakindi, Leon Mugesera, Juvenal Habyarimana—must be remembered not as constructs but as actors whose words and silences shaped reality. What this report by Iyamuremye exposes, is not merely failure but choice. Authorities knew. They anticipated violence. They warned of it internally. Yet they proceeded, delayed, minimized, or ignored. Killers enjoyed impunity not because the state was absent, but because it was present in the wrong way—mobilizing, signaling, and then looking away. To remember these details is to resist the erasure that impunity depends upon. It is to affirm that genocide does not begin with killing, but with permission, and that permission is always traceable to those who had the power to say no and chose not to. The December 1992 violence documented by Iyamuremye was not an inconsistency. It was a rehearsal, conducted under conditions of forewarning, ideological incitement, material incentive, and administrative complicity. By January 1993: The language of extermination had been normalized. The mechanisms of genocidal mobilization were tested. The boundaries of impunity were set and confirmed. And, this is just one report!

Opinions

Dr. Augustin Iyamuremye’s report: Impunity foreseen in 1993