Africa's World War: Congo, the Rwandan Genocide, and the Making of a Continental Catastrophe

by Gerard Prunier

  Late on the afternoon of November 13 Rwandese mortars opened fire at close range on the Mugunga camp. They fired off and on all night, and the infantry attacked the next morning. The ex-FAR fought for a few hours and then withdrew westward in the early afternoon, leaving one million refugees in a state of confusion and panic. They then split up into three groups, the main one heading east toward Rwanda, and two smaller ones heading west toward Walikale. In west Africa this was seen as another defeat; Ivory Coast president Henri Konan Bédié declared, “We [francophone Africans] feel enraged because we are powerless to confront this situation.”54 By the morning of Friday, November 15, hundreds of thousands of refugees were crossing east into Rwanda, the same border they had crossed west into Zaire when fleeing the RPF twenty-six months before.

  In the meantime, totally unnoticed in the general commotion caused by the explosion of the camps and the massive return of so many refugees, another event was taking place, which would eventually bring Uganda into the war: the ADF guerrillas, which the Sudanese Secret Service had been preparing for months, had been forced into emergency action on November 13 because, after the taking of Ishasha by the Congolese rebels, they had felt that their rear base at Kasindi in Zaire had now come under threat.55 Museveni phoned Mobutu (who seemed genuinely surprised) and told him he would enter Zaire in hot pursuit in case of a renewed attack.56 Trying to turn necessity into opportunity, ADF leader Ssentamu Kayiira issued a communiqué saying that his men were fighting “to reintroduce multi party politics in Uganda, stop Museveni’s nepotism giving all the juicy jobs to Westerners [meaning people from Ankole and Kigezi, not Europeans] and re-establish cordial relations with Uganda’s neighbours.”57 There seemed to be a slight problem of political coherence among the rebels: while the Tabliq component of ADF was preaching through its mosque circuit that it was fighting for an Islamic state with Jamal Mukulu as its president, Bakonjo prisoners told UPDF interrogators that they were fighting because they had been promised that they would finally get their autonomous kingdom.58 In any case, Museveni had made a decision: he would have to go into Zaire to get the ADF, and since the rebels were moving north it seemed to be a good time. The internationalization of the war was now an unavoidable fact.

  The refugee exodus.

  After mid-November 1996 the Rwandese refugees were the object of a raging controversy centering on how many of them there were. How many were there in the camps at the time of the attack? Various head counts had been conducted by UNHCR during the prior few months, giving the following results:

  Goma Area

  Bukavu Area

  Uvira Area

  (counted Mar. 31,1996)

  (counted Aug. 30, 1996)

  (counted Sept. 6,1996)

  Rwandese: 715,991

  Rwandese: 305,499

  Rwandese: 75,948

  Burundians: 2,000

  Burundians: 2,000

  Burundians: 145,518

  Source: UNHCR Regional Special Envoy’s Office, Kigali, September 26, 1996.59

  Now the question was, what happened to them once the shooting started? The Burundian refugees were the first ones to be hit and to flee. But over half of them were caught by the Rwandese-rebel forces and herded toward the Burundi border. These (numbering 77,000) were coyly put by UNHCR in the category of “spontaneous returns.”60 Another 23,000 slowly trickled back over a period of eight months, until June 1997, while an estimated 10,000 managed to make their way independently to Kigoma in Tanzania. Taking into account the very uncertain counting methods of returnees at the border, this leaves a gap of between 25,000 and 40,000 Burundian refugees disappeared, presumed dead.

  The problem of the Rwandese is much more complicated. Just to give an idea of the order of magnitude of the uncertainty, in early 1997 the United Nations boasted about a precise 726,164 refugees having “self-repatriated” in mid-November 1996,61 but by early 1997 it had revised that estimation to a prudent 600,000.62 In fact, nobody really knew. The human torrent that had crossed the border away from the North Kivu camp disaster on Friday, November 15, was estimated at first in the evening at “around 100,000 persons.”63 The next day UNHCR admitted that “the registration system at the border [was] breaking down.” Figures started to get extremely subjective, depending on the personal preferences of who was issuing them. Ray Wilkinson, the UNHCR spokesman, declared on November 15, “The 700,000 refugees of North Kivu are all heading home,”64 but the next day he thought that only 350,000 had crossed the border; at the same time, Eric Mercier of Médecins Sans Frontières was putting the figure at 500,000.65 On November 18 the UN offered further details about its statistical procedures: the flow, starting on the 14th, was evaluated at between 10,000 and 12,000 people per hour. It started in the afternoon and kept going evenly until the evening of the 18th. Counting on a ten-hour-a-day flow (the barrier was closed at night) the figures came to the following:

  Friday, November 15:

  60,000

  Saturday, November 16:

  120,000

  Sunday, November 17:

  120,000 (full day, but the flow is slowing down)

  Monday, November 18:

  40,000 (same as above)

  Tuesday, November 19:

  5,000 (the flow had become very thin)

  Wednesday, November 20:

  5,000

  Thursday, November 21:

  5,000

  Friday, November 22:

  5,000

  Saturday, November 23:

  5,000

  Sunday, November 24:

  5,00066

  Monday, November 25:

  1,016, the first precise head count. The rush was over.

  So what do we have? A grand total of 371,000 returnees out of a base population of over 800,000 in North Kivu at the beginning of the exodus.”67 In South Kivu it was even worse: very few refugees had crossed the Cyangugu barrier in spite of Rehabilitation Minister Patrick Mazimpaka talking about “massive daily crossings.”68 In fact, numbers now became pawns in the diplomatic tug-of-war about whether or not there should still be a multinational intervention force and what its purpose should be. Michela Wrong summed it up perfectly:

  Until recently many Zairian politicians believed they would be saved by the arrival of a multinational force with a humanitarian agenda of feeding one million Hutu refugees. By stopping the rebel operations such an intervention could allow the Army to recapture lost ground and shore up the tottering regime. But yesterday’s apparent return of hundreds of thousands of refugees to Rwanda has suddenly changed the picture.69

  The Americans seized the moment. On November 18 a U.S. State Department spokesman declared, “The mass return of nearly half a million Rwandese refugees from Zaire obliges the U.S. to review its plans in the African Great Lakes crisis.” In case the point was not clear enough, U.S. Defense Secretary William Perry added, “We might not go. We are not the Salvation Army.70 The intervention momentum was lost. Canadian Gen. Maurice Barril, who was supposed to take command of the MNF, declared the situation to be “unclear, with remaining refugee estimates varying from 100,000 to 500,000… . It is necessary to be better informed about conditions on the ground to study the military choices which could be made.”71 The numbers game continued raging. “The RPF regime denies the existence of the 700,000 refugees left in Zaire,” declared the Rwandese opposition group Forces de Resistance de la Démocratie.72 while General Kagame said at a press conference, “The majority of the refugees have returned… . There are only a few scattered refugees remaining in Eastern Zaire… . International agencies are inflating the numbers of those left behind for their own purposes.”73 Sylvana Foa, the UN spokeswoman, declared that the United Nations believed there were still 746,000 refugees in Zaire and that the problem was not resolved. Boutros Ghali’s preference for high figures was probably not entirely independent from his pro-French position. U.S. Ambassador Robert Gribben in Kigali did not agree with those numbers, writing, “Most of the refugees still in Zaire
are either Zairian or Burundians and the number of Rwandese refugees are nothing like those put forward by aid agencies… . They are in the twenties of thousands rather than these vast numbers.” This infuriated the NGO Refugees International, which asked for the recall of the ambassador. Overflights and satellite pictures gave a different, more sober, story.74 There were very large numbers of Rwandese refugees left in eastern Zaire, although they were probably somewhat less than the 700,000 UN figure. There was one relatively small group west of Masisi, another larger group north of Sake, enormous numbers heading west of Bukavu toward Shabunda, and smaller numbers heading south around Fizi. All in all, the total number was probably between 600,000 and 650,000.

  In all the hubbub about refugees, the Zairian rebels had been almost forgotten. But at a meeting held in Goma on November 20 André Kisase Ngandu declared, “Now we have to think for ourselves… . Our main problems are how to use taxes, a fair administration and freedom for all of Zaire.” He was increasingly vocal and he recruited easily. To the young men who joined him he was clearly saying that the main problem they were going to have was how to deal with the very heavy “protection” they were getting from the Rwandese. Those who joined him were usually those who wanted to bring Mobutu down but who did not trust Kigali. As a result Kigali considered him with growing distrust.75 This did not go unnoticed by Kabila, who took exactly the opposite tack, behaving in a very subservient way toward “the tall ones,” as everybody called the Tutsi. Kabila even acquired at the time the ironic nickname Ndiyo Bwana (“yes sir” in Swahili) because it was his usual answer to whatever the Rwandese were telling him. But the time was not yet ripe for the Rwandese to activate him. First they had to finish dealing with the refugee question.

  Rumors were now beginning to filter out indicating that the RPA and its local auxiliaries were killing the refugees left behind in Zaire. In mid-November, near the temporary camp at Chimanga, Zairian peasants found 310 bodies and buried them.76 The Dutch paper De Standaard published interviews and excerpts from the diary kept by a refugee since October 20. He described how the Rwandese army was hunting down the refugees, systematically killing all the males between the ages of ten and fifty, and how he had personally seen heaps of bodies he estimated at 120 or 150 people. He added that the Interahamwe had also killed an estimated 500 people in Mugunga on the last day, after one of their numbers had been murdered and after they understood that most refugees were too tired and too dispirited to risk another dash in the jungle.77

  But the full extent of the massacres was not to be understood until the publication several months later of a report compiled during those terrible days in Kivu.78 This report was questioned at the time because it was anonymous. Its author was Father Laurent Balas, a French priest who had lived in Kivu for the past six years and who spoke Swahili. The reason he initially testified anonymously was that he was still living in Goma and so his life would have been in danger had his name been made public.79 Contrary to some allegations made at the time, Father Balas did not harbor any anti-Tutsi feelings. In late October 1996 he had saved forty-two Tutsi who had taken refuge in his house when the Bukavu anti-Banyamulenge pogroms extended to the northern capital, and he had nearly been lynched while trying to take seven Tutsi women across the border to Gisenyi.80 His testimony was simply a spontaneous reaction to the new horror he now had to witness and which the international media, to his pained surprise, did not report on at the time. He was keen to point out that the refugees were the victims of a deliberate murder campaign and not simply of the fighting:

  If it was only for the fighting the Hutu refugees would have no more reason to flee than the local Zairian peasants. What they are fleeing are the massacres perpetrated by the Tutsi “rebels.” . . . The “rebel” discourse is to say that all the refugees who have not gone back to Rwanda are génocidaires… . But UNHCR estimates that only about 7% of the refugees took part in the genocide which means that many innocents have not gone back… . Calling the refugees génocidaires . . . is similar to the inyenzi (“cockroach”) name used by the interahamwe themselves during the 1994 genocide so that they could kill with a lighter heart… . There are mass graves everywhere; but they are carefully concealed and looking for them is extremely dangerous. [There follows a description of several mass burial sites and of random dumping grounds for corpses.] . . . On 24th December the “rebels” captured two young Congolese Hutu and used them as guides to take them to refugee locations. One of the guides commented: Waliwauwa wote wakimbizi wale, wote kabisa, hata moja aliyepona meaning: “They killed them all these refugees, everyone of them, not a single one survived.” . . . Tens of thousands have been murdered and hundreds of thousands will be made to die of exhaustion, of sickness, of hunger… . Will that be an end to the problem of the 1994 génocidaires? This is far from sure. The interahamwe and ex-FAR who committed most of the murders are young and strong. They run fast and they have weapons… . The refugees who are getting murdered are not the killers; they are families who cannot run faster than the children with them and who are fleeing and dying as a group.

  Father Balas regretted that the media maintained an almost complete blackout on the massacres even after he contacted the Paris daily Le Figaro and Radio France Internationale. The international community’s attitude toward the problem was cavalier at best. For example, when Maurice Barril declared that there were no refugees left in eastern Zaire, he based his declaration on having spent half a day around Sake in December 1996, driving on the open road in a Rwandese army jeep.

  Many refugees had returned to Rwanda willy-nilly; many others were dead; but there were still several hundreds of thousands in Tingi-Tingi, near Lobutu, in Shabunda, or scattered here and there in the Virunga forests. They were to be part of the next episode, as the “rebels” were now regrouping and preparing for their next move.

  The long walk into Kinshasa

  War and diplomacy.

  Confused discussions about the relevance of the MNF sputtered on inconclusively at the Stuttgart meeting on November 25, 1996. Meanwhile the United States (or rather, some elements thereof) had swung into action. On that same day Peter Whaley, acting deputy chief of mission of the U.S. Embassy in Kigali, crossed the border to meet Kabila in Goma. Both men later refused to make any comment about their meeting.

  The U.S. military involvement in the region had started right after the genocide, when a contingent of sixty American soldiers arrived in Kigali on July 31, 1994. U.S. officials were greatly impressed by Kagame’s leadership and characterized him as “a brilliant commander, able to think outside the box.”81 In early 1995 the U.S. army started a training program for the RPA. It was a large program that brought RPA officers to the United States as well as U.S. army personnel to Rwanda. The fact that the Department of Defense Joint Combined Exchange Training Program did not require congressional approval did make matters simpler when the department wanted to help the Rwandese army.

  When asked about the program during a December 1996 congressional hearing, Ambassador Richard Bogosian said that the training “dealt almost exclusively with the human rights end of the spectrum as distinct from purely military operations.” Considering the massive human rights violations committed by the RPA that I described in the first two chapters of this book, that part of the training must not have been very successful. As for the military aspect, since the RPA was probably the African army with the best experience in unconventional warfare on the whole continent, it should have been able to train the Americans and not the other way around. In fact, under the respectable technical guise of “training,” this was largely a psychopolitical covert relationship in which the Rwandese RPF managed to hook the Americans. Half through guilt, half through admiration, some segments of the U.S. Department of Defense slowly slipped deeper and deeper into cooperation with the RPA, probably feeling that they were doing what some State Department diplomats really wanted them to do but did not have the guts to take responsibility for.82

  The U.S. army
and the CIA opened several communications monitoring stations in Uganda, first on Galangala Island in the Ssese archipelago on Lake Victoria, then at Nsamizi Hill near Entebbe, and finally in Fort Portal.83 During 1995–1996 large U.S. Air Force transport planes (C-141s and C-5s) landed very frequently at the airport in Kigali, somewhat less often at Entebbe. When questions were asked, the answer was always the same: they were “carrying aid for the genocide victims.”84 That aid must have been very heavy. But it seems to have answered needs quite different from those of the genocide survivors, who never reported any particular largesse on the part of the Americans. At that time the RPA acquired a large quantity of excellent communication equipment and many nonlethal military supplies (vehicles, boots, medicines). Once the war started it seems that these supplies were supplemented by secondhand former Warsaw Pact weapons and ammunition which were either flown directly to Goma or air-dropped at convenient points along the advancing AFDL lines. The U.S. Air Force was by then using slower and more rugged C-130s for these operations.85 Apart from this direct logistical support to the AFDL, Washington operated a multipurpose anti-Mobutu machine which ranged from the half-humanitarian, half-military support given by the International Rescue Committee, long rumored to have been an NGO close to sensitive segments of the U.S. administration, to the soothing testimony given on the question of the missing refugees by Assistant Secretary of State to the Bureau of Population, Refugees and Migration Phyllis Oakley in December 1996. After some American arm-twisting at the UN the U.S. Ronco Consulting Corporation got a large de-mining contract in Rwanda to remove more mines than had ever been laid during the war. This had the advantage of legitimizing the impressive U.S. military air traffic since “supplies” were needed. It was an impressive performance which was completely different in style from the heavy-handed U.S. interventions during the cold war. It was stealthy, light, and indirect, with the one remaining superpower on earth easily running circles around a frustrated French diplomacy still caught up in the inefficient old web of its questionable Franco-African friendships.86