The Geopolitical Situation in the Great Lakes Area in Light of the Kivu Crisis
|Publication Date||1 February 1997|
|Cite as||WRITENET, The Geopolitical Situation in the Great Lakes Area in Light of the Kivu Crisis, 1 February 1997, available at: http://www.refworld.org/docid/3ae6a6be0.html [accessed 7 December 2013]|
|Disclaimer||This is not a UNHCR publication. UNHCR is not responsible for, nor does it necessarily endorse, its content. Any views expressed are solely those of the author or publisher and do not necessarily reflect those of UNHCR, the United Nations or its Member States.|
The Kivu crisis as such, that is the period between the first active persecutions of the Banyamulenge in late August 1996 and the occupation by Zairian rebels of Goma in mid-November, can be seen as a turning point in the more global crisis which has been endemic in the Great Lakes region since the attack on Rwanda by the exiled Rwandan Patriotic Front (RPF) on 1 October 1990. The effects of the Kivu crisis are massive. The Burundi guerrillas based in Zaire are pushed out, leading both to an intensification of the war in Burundi and to a strengthening of Major Buyoya's regime in Bujumbura. The threat posed by Rwandan refugee camps to their country of origin is removed, although the fate of those refugees who did not return home has been tragically neglected. At the same time a low-level civil war has been imported from Zaire into Rwanda with the massive return of the refugees, and a new area of insecurity and violence has appeared on Uganda's western border near Kasese, with a new guerrilla front operating from Zaire. And last but not least, the very fate of Zaire is now uncertain in the face of a major insurrection in its eastern provinces.
2. THE KIVU CRISIS
Although its outbreak surprised many observers, the Kivu crisis had been building up for several years, even before the 1994 Rwanda explosion. Kivu, both North and South, was a tense region in Zaire ever since the end of one-party rule in 1990, due to the rivalry between the so-called autochtones ("native" tribes) and the Kinyarwanda-speaking groups who were accused of being "foreigners". The choice to have ethnic representation at the Conférence Nationale Souveraine in 1991 only worsened matters.
The situation of the Banyarwanda in Kivu is rather complex. In the North, several "layers" of Banyarwanda people had arrived at different times in recent history. The oldest group were the Banyarwanda families, both Bahutu and Batutsi, who had lived for centuries in the "western marches" (in the European medieval sense of the term) of the old Rwanda kingdom, before the arrival of the Europeans. In addition, there were the mostly Bahutu descendants of Banyarwanda who had been brought in by the colonial authorities in the 1920s and 1930s to supply manpower for the underpopulated Belgian Congo. They were joined by more recent immigrants, exclusively Batutsi, who had fled the massacres of the years 1959-1963 in Rwanda. Together the North Kivu Banyarwanda formed a tightly knit, mutually supportive community without distinction between Tutsi and Hutu that thrived economically and as a result made enemies. Once their protector Barthememy Bisengimana fell from power, a new Citizenship Law, adopted in June 1981, ensured that the Banyarwanda could technically be considered "foreigners". During the following ten years, tensions remained, but did not explode until the fateful Conférence Nationale of 1991, where the "native" tribes attempted to monopolize representation in the Assembly and then to use their position to eliminate the Banyarwanda at first economically and then later physically.
The Rwandan civil war had changed the situation in the Banyarwanda communities of Kivu. Many young Tutsi had enlisted in the RPF, while the Habyarimana regime was very active in North Kivu though a politicized "agricultural cooperative", the Mutuelle Agricole des Virunga (MAGRIVI). When some local tribes, especially the Bahunde and the Bayanga began first attacks on the Banyarwanda in March 1993, the Banyarwanda community was not able to unite in a common front to fight back these attacks. As former Mulele guerrillas and marginal ritualistic groups such as the Mai Mai became involved, the war became multilateral and by late 1993 degenerated into virtual anarchy. Things took a turn for worse when, with the RPF victory in the Rwanda civil war, a large number of highly politicized Bahutu refugees came over the border into the Goma area. Looking for local allies in their attempt to create a territorial base from which to harass the victorious Tutsi government in Kigali, the ex-Forces Armées Rwandaises (FAR) and interahamwe militiamen responsible for the April-July 1994 genocide joined in an alliance with the hard-pressed local Bahutu population. Together they attacked the North Kivu Batutsi and pushed back the Bahunde and Banyanga militias. Many of their adversaries took refuge in Rwanda where the RPF regime gave them asylum.
Meanwhile, the situation also deteriorated in South Kivu, albeit in a much less spectacular way. The local Kinyarwanda-speaking people known as the Banyamulenge5, named after the hill where they first settled in the early years of the nineteenth century, were in fact Batutsi. The Banyamulenge as well were quite successful economically and had developed large farms and ranching estates. Aware of the usefulness of contacts with the Europeans, they had created a local NGO, the Milima Group, which worked in partnership with foreigners, expecially with the Belgians, in the powerful Réseau Zaire. As a result, relations with many of the local tribes (Babembe, Bazira, Bafulero, Bavira) suffered. The Banyamulenge had observed the fate of their fellow tribespeople in North Kivu, where Batutsi were hunted down, killed or forced to flee, and felt the need for protection. Many young Banyamulenge had joined the RPF during the civil war in Rwanda and had subsequently remained as members of the Rwandan Patriotic Army (RPA). When they realized that the same alliance of Zairian administration, local militias, ex-FAR and interahamwe was assembling in the South, they approached the government in Kigali for help. The request proved timely, as the political leadership in Rwanda was looking for means to deal with a number of acute regional political problems.
In particular, Rwandan leaders were concerned with the Bahutu refugees in Kivu, few of whom had returned in the preceding two years as they largely remained under the influence of their former political and military leaders. Moreover, the international community had failed either to move the refugee camps further away from the border or control them militarily, and the ex-FAR had rearmed and was harassing Rwanda's western borders. Given the minority standing of the new regime, this could only lead to increasing repression on part of the RPA. For leaders in Kigali, the possiblity of intervening militarily in the refugee camps, was very tempting indeed, particularly as the refugee camps in the Uvira area of South Kivu were also linked to the support system of Hutu guerilla from Burundi united under the Front de Défense de la Démocratie (FDD) and led by the former Interior Minister of Burundi, Léonard Nyangom. In many ways, the war had become a regional one as the FDD fought in Burundi alongside the ex-FAR and interahamwe militiamen, while the RPA assisted the Burundi army in the northern provinces of Cibitoke, Kayanza and Ngozi.
On 25 July 1996, as the Hutu armed opposition stepped up its attacks in Burundi, Major Pierre Buyoya the former Tutsi President who had lost the June 1993 election after carrying out a remarkable program of political liberalization, regained power in a military coup. The coup marked the final step in a process of progressive erosion of the power-sharing agreement signed in September 1994 between the Tutsi opposition and the post-1993 putsch remnants of the legal FRODEBU government. Under the dual impact of Tutsi internal sabotage and Hutu military violence, the "government" voted in during the June 1993 election had lost almost all real power; "President" Sylvestre Ntibantunganya had no authority over the Tutsi army and was increasingly looked upon as a traitor by his own fellow Bahutu. The power Major Buyoya seized was therefore virtually non-existent, although he gained the attention of the international community, which in the first half of 1996 grew increasingly concerned about a possible repeat in Burundi of the 1994 Rwandan genocide.
At the initiative of governments in the region, economic sanctions were imposed against Burundi, and the new RPF regime in Kigali came under strong pressure by President Museveni to comply fully with these sanctions. But from the point of view of RPF strongman Major General Paul Kagame, the sanctions had a major disadvantage as they affected one side to the conflict only. While Burundi's regular army was badly affected, the FDD forces were safely ensconced in their Uvira rear bases and suffered little. As a result the military balance began to tip and FDD pressure on Bujumbura mounted dangerously during September 1996.
The Rwandan leadership wanted, however, to avoid an FDD military victory for three reasons:
For Rwanda's Tutsi community, already traumatized by the 1994 genocide, the possibility of another genocide taking place in the neighbouring country was intolerable;
If victorious, Léonard Nyangoma and his allied were likely to offer their country as a base for attempted attacks on Rwanda by the ex-FAR;
Collapse of the Buyoya regime caused by a Hutu military victory would open the way for former President Jean-Baptiste Bagaza, who enjoyed the support of hard-core Tutsi supremacists both in the army and amongst civilians, to take over Tutsi leadership. It was public knowledge that Jean-Baptiste Bagaza's plans in case of military defeat was to ensure ethnic survival through mass exodus of the Tutsi of Burundi to Rwanda. The new RPF regime in Rwanda was, however, extremely concerned about the destabilizing impact of several thousand foreign Batutsi who, having lost everything, would expect help from their "brothers".
The RPF could thus hope to achieve several of its key objectives by allowing the young Banyamulenge soldiers within the RPA return home and take control of Kivu: save its Banyamulenge cousins from persecution; liquidate the FDD bases around Uvira and thus indirectly strengthen the Buyoya regime without having to violate the embargo; and with some luck also succeed in dislodging the refugees and forcing them back into Rwanda, where they could be brought under control. As we now know, the events which unfolded between early September and late November 1996 enabled the Rwandan leadership fully to realize all three objectives. But such an all-out military action in the politically fragile environment of the region was bound to have a number of consequences, some of them unforeseen.
3. SUDDEN DEVELOPMENTS IN THE REFUGEE SITUATION
The RPF leadership politically revived a veteran of the Congolese civil war of the 1960s, Laurent Kabila to anchor the military operations in the local reality of Eastern Zaire. By late October 1996, the almost forgotten guerrilla leader had come to the fore as a major actor in the conflict with controversial but undeniable support from Rwandan leaders.
The key factor, which motivated RPF support for Laurent Kabila and his, at the beginning, rather meagre band of Zairian rebels were the Rwandan Hutu refugees and their increased military aggressiveness. After the fact,a number of voices in the international community lamented the political blindness and inaction which had allowed the situation to degenerate to such a point. But clearly prior to and during the crisis, the international community (with the exception of a few NGOs which had refused to work in the camps) had preferred to ignore the militarization in the refugee camps.
As the number of refugees later became an important bone of contention, when a large part of the refugees returned to Rwanda, this question bears some examination. The most reliable figures available are those provided by UNHCR in late September 1996.
In Zaire, there were five large camps in the area around Goma, which sheltered about 717,000 refugees. Around Bukavu, 23 small camps housed about 397,000 refugees. Around Uvira, the situation was different with only 76,000 Rwandan refugees, but 143,000 refugees from Burundi.
In Tanzania, about 410,000 Rwandans were located in the Ngara area, while another 123,000 were in Karagwe. The political environment in Tanzania was, however, much less critical because the Tanzanian Government, unlike President Mobutu, never tried to use the refugees as pawns in any local or international politics. As a result, Tanzania had managed to assert a fair degree of administrative and even police control over the refugees, and the camps in Tanzania had never developed into military base camps for the ex-FAR and interahamwe. The fighting which took place in Kivu therefore had no equivalent in Tanzania.
When the operations began in earnest in late September 1996, the priority for the Banyamulenge/ADFL forces was to stabilize their control of the hills overlooking Uvira and to neutralize the military potential of the Hutu refugee camps in the lower lands. This was achieved in less than a month.
Soon after, the camps in North Kivu came under attack. Contrary to what was frequently reported at the time, the attackers were neither Banyamulenge nor part of Laurent Kabila's forces they were at that time fully occupied in the South but a mixture of Mai Mai militiamen from the Bahunde and Banyanga tribes and local Batutsi from the Masisi/Walikale/Rutshuru area, all backed by Rwanda. They successfully used force to empty the camps of Katale, Kahindo and Kibumba, most of the refugees fleeing either to the heavily forested slopes of the Virunga mountains or further west to the Mugunga/Sake area, just north of Lake Kivu. Soon, the refugees around Mugunga numbered almost half a million persons. The humanitarian situation deteriorated sharply.
France then vigorously launched the idea of an armed international intervention, but was met with considerable hesitation from other western governments, who feared involvement in an extremely complex politico-military situation. France was also suspected of aiming to gain military control of Kivu under the guise of a "humanitarian" intervention, only to hand it back later to President Mobutu's regime. The U.S., still affected by the "Mogadishu Syndrome", was especially reluctant to commit ground troops to any operation on the African continent.
For the leadership of the former Rwandan government this was a period of panic. The ex-FAR and the interahamwe appeared confused and only carried out a number of poorly coordinated actions to delay the advance of the Banyamulenge/ADFL forces, who were advancing on the western shore of Lake Kivu, while their northern allies kept up pressure on Mugunga and Sake. When Laurent Kabila's forces occupied Goma, the refugees were trapped. During the night of 13 to 14 November, the pro-Rwandan forces opened fire on Mugunga with light artillery. Then they proceeded to attack with infantry and the ex-FAR defences quickly collapsed after some weak fighting.
The consequence was completely unforeseen: the refugees started to walk en masse towards neighbouring Rwanda. Their motivations were unclear. Always optimistic, the international agencies concluded that the main impediment to their return had been the influence of the Hutu militias in the camps. A number of other possible causes are likely, including (certainly) violence by Rwanda's allies and political calculation on the part of the former Rwandan authorities in the camps, who possibly considered that their best option was to return to into Rwanda under the protective shield of the refugee masses and carry on the war from within.
The unexpected mass exodus, half forced, half voluntary, had an immediate impact on the still imprecise plans for a military intervention. Within days an increased reluctance could be detected on part of most governments, except France, and barely a week later the plans were cancelled.
Seizing the opportunity where forceful action would go almost unnoticed in the general commotion, Tanzania declared that all Rwandan refugees, about 530,000, on its soil must leave by the end of the year, and the Tanzanian army quickly moved towards the camps. Seized by panic at the idea of return, and probably also coerced in part by their still-active leadership, several thousand refugees fled in various directions in the hope of reaching the borders of either Kenya or Malawi. The Tanzanian army quickly caught up with them, forced them back towards the camps and from there pushed them on across the border into Rwanda. There was no UNHCR protest and the few aid workers operating in the camps who complained were expelled by the Tanzanian authorities. By Christmas, the camps were emptied.
There remained the problem of those refugees who did not return and who caused some confusion as to their number. At the time of the mass return to Rwanda, the UN information team in Nairobi had published a puzzling assessment, giving a figure of 400,000 returnees with 500,000 remaining in Zaire. This total of 900,000 left roughly 300,000 persons unaccounted for. Later the estimated number of returnees was increased to 500,000-600,000, and the number of those estimated to have remained in Zaire reduced to about 200,000, leaving still about 400,000 "vanished" refugees. According to the latest estimate, based on approximate counts at a variety of locations between Western Kivu and Manyema, the estimate of the number of refugees in Zaire was again increased to 560,000.
The entire episode, the more than two years that were wasted during which the situation was left to deteriorate needlessly, while western governments allowed the idea of intervention to be guided more by political than humanitarian factors, and the confused and at times forced repatriation of Rwandan refugees, could only leave a bad taste in the mouth of the international public.
4. THE IMPACT ON RWANDA
The Rwandan leadership had conceived of the return of the Rwandan refugees to Rwanda as a process for establishing political control; given the excellent administrative and military network of the RPF in the country, the returnees were to be rendered harmless and the threat they posed while outside the borders diminish. This belief soon proved to be wishful thinking. Given the extremely short time span of the return (four days) and the vast number of returnees (over half a million) re-entry control was minimal. Even if weapons could not very easily be taken back in, identities were hard to check.
In many communes, the Hutu returnees massively outnumbered the Batutsi, both the survivors and those who had returned from exile in 1994. Moreover, many Bahutu expressed no repentance whatsoever about the genocide, and some even denied that it ever took place, which led to extremely tense situations in certain regions. The military presence of the RPA was insufficient to keep the hilly countryside under control, particularly in the Gisenyi and Ruhengeri préfectures in the Northwest, an area of the country which had been a bastion of Habyarimana supporters and a hotbed of interahamwe activism and where a majority of the returnees were originally from. There were also frequent conflicts over the control of land and houses in this overpopulated country, where about 140,000 houses had been destroyed during the genocide.
To make matters worse, the genocide trials, which had been delayed for over two years, now suddenly started. Many of the returnees were arrested on a variety of charges, including participation in the genocide, a charge which could lead to the death penalty. While the arrests and trials were obviously necessary, the timing and the style of the proceedings were highly counterproductive. The Rwandan Government had been advised on guidelines to minimize tensions during the unavoidably painful genocide trials. But the reality of the trials was very different. The accused were not allowed representation by lawyers; the charges were unclear; the judges were not very professional; in some cases there were no witnesses or evidence, and when there were, the accused were not allowed any form of cross-questioning. In fact, they were mostly forbidden to speak. At the end of these "trials" the accused were generally condemned to death.
In spite of the poor judiciary standards evidenced, the international community was hardly in a position to criticize, not only because of its total lack of action at the time of the genocide, but also because the international tribunal also encountered major difficulties. It had also been very slow to start up it had finally commenced trials at about the same time as in Rwanda and was wracked by stories of corruption and nepotism, misappropriation of funds, hiring of girlfriends as secretaries, careless handling of evidence and improper use of equipment, especially airplanes, which often seemed reserved for personal use rather than for the work of the tribunal. The problems were different from those encountered in Rwandan courts, and the usually accepted standards of justice were better adhered to, but a report resulting from an inquiry organized by the UN was highly critical. The opportunity to have the genocide trials achieve justice and serve the purpose of national reconciliation appeared wasted from the start.
The lack of international intervention and the mockery of justice evidenced in the trials in Rwanda encouraged the Hutu extremists who had returned hidden in the mass of the refugees to carry out a number of attacks within a small region. On 11 January 1997, sixty armed men robbed the hospital at Kabaya, killing three hospital workers. The next day twelve people were summarily shot by interahamwe in Giciyi. On 13 January, UN Human Rights observers were assaulted and threatened with death. Then, on 18 January, three volunteers with the Spanish medical NGO, Medicos del Mundo (MDM), were deliberately shot dead near Ruhengeri. The killings of foreigners finally drew international attention to the rapidly and continually deteriorating security situation.
In Ruhengeri préfecture, armed men occupied public buildings, beating or even killing civilian government employees. On 21 January, twenty-four people were killed in Nyamugali commune, Ruhengeri préfecture. The Army cordoned off the area and initiated a search-and-destroy operation. Although Vice-President Paul Kagame admitted that eighty persons were killed, NGO sources put the number of deaths in the hundreds. Claude Dusaidi, Political Adviser to General Kagame, put the blame squarely on the returnees. On 2 February, a Canadian priest who had endeavoured to promote cooperation between the Tutsi and Hutu and had assisted the International Tribunal in its enquiries, was shot dead, quite probably by a returnee, while he was celebrating mass. On 4 February, two UN Human Rights observers and three Rwandan travelling companions were killed in an ambush on a road near Cyangugu.
The violence appeared increasingly deliberated even if it did not seem closely coordinated. The underlying political tactics were clear: scare away foreigners, increase frictions between the RPA and the Hutu population and create insecurity on the roads so as to disrupt communications all with the ultimate hope of pitching the ordinary population actively against the Government. The apparent "burundization" of Rwanda was encouraged by the Government's own actions, such as when vice-presidential adviser Claude Dusaidi declared that the regime would "take care of the problem with an iron hand".
5. THE IMPACT ON BURUNDI
Following the 25 July 1996 coup by former President Pierre Buyoya and the subsequent imposition of economic sanctions on the country by its African neighbours (31 July), the situation remained at a stalemate for a few weeks.
President Buyoya was subjected to considerable pressure by extremists on his own side, who threatened to remove him if he negotiated with the "bandits". But he was also hard hit by the embargo. In a gesture of goodwill, he reopened Parliament in mid-September and permitted political parties again to take up activities. The measure was largely symbolic, since most of the FRODEBU MPs had fled, and those who had stayed carried little political weight. Nevertheless the Conseil National pour la Défense de la Démocratie (CNDD) and its armed wing, the Front de Défense de la Démocratie (FDD), tentatively began to move towards negotations. Pierre Buyoya responded cautiously but with an obvious desire to achieve some kind of a diplomatic breakthough in a situation which was becoming increasingly untenable.
However, a few days later the Banyamulenge and ADFL forces launched an attack on the refugee camps and the FDD support bases in Zaire, sending the refugees into headlong flight. CNDD leader Léonard Nyangoma fled to Tanzania, where he met with former President Julius Nyerere in his capacity as Burundi mediator. Although a vague pretence of readiness for talks was maintained by both sides, neither was willing to embark truly on negotiations. The Banyamulenge victory and the FDD debacle had given President Buyoya a certain military breathing space, while Léonard Nyangoma was unwilling to negotiate from a position of weakness.
The entire humanitarian support structure which had provided the FDD with its fall-back positions in Zaire began to collapse, and the Banyamulenge forcibly pushed any Hutu refugees that were caught to the Burundi border. About 40,000 Burundi Hutus eventually returned between October and December, while another 100,000 fled westwards into the interior of Zaire together with large numbers of Rwandan Hutu refugees from the South Kivu camps. In Burundi, the Army initiated killings of returnees as early as the last weeks of October. The killings were systematic in part, but also arose more spontaneously in areas. In some cases, victims were carefully selected and constituted mostly young men who were suspected of being potential FDD recruits. In other cases, killings took place in unplanned reprisals as the Army fought a new upsurge of guerrilla activity in the provinces of Cibitoke, Ngozi and Kayanza.
Paradoxically, the intensity of fighting increased, as most of the guerrillas fleeing Zaire reentered northern Burundi only in an effort to reach the eastern border with Tanzania, hoping to find a safe haven there. Distinctions between civilians and guerrilla, and between settled peasants, internally displaced persons and returning refugees, gradually began to disappear. While the fighting largely pitted Tutsi against Hutu, in some cases Hutu fought Hutu, as guerrillas took by force what they could not get by political sympathy. Killings of civilians were sporadic, and to make matters worse supporters of the FDD and the Hutu People's Liberation Party (Parti pour la libération du peuple hutu - PALIPEHUTU) began to fight each other in the refugee camps in Tanzania.
The result of this renewed violence in the North was an increase in the number of refugees fleeing to Tanzania, at the same time as the Tanzanian authorities tried to repatriate (at times forcibly) Burundi refugees. Approximately 156,000 refugees fled Burundi during November and December 1996 and during the first three weeks of January 1997, joining the 150,000 or so "older" refugees, and making a total of over 300,000 Burundi refugees in Tanzania.
As violence ravaged the northern part of of Burundi, sanctions continued to bite into the economy. Prices rose between 60 per cent and 150 per cent, and most medicines and many essential commodities (kerosene, cooking oil, sugar, flour) were priced out of the purchasing power of ordinary people. As fuel was priced out of the market, trade in food also declined. However, the local population had come to depend on food which was traded commercially to survive during the civil war, when the fighting had forced many away from their own fields. Increased transport costs put this "market" food beyond their reach. Thanks to a steep increase in smuggling and black market operations, all products remained available, but at prices that only the rich could afford. The military and political elite of the country remained thus insulated against the effects of the embargo, although they argued its effects on the poor to have sanctions lifted. However, although there is a certain softening in the attitude of some European countries such as Germany, at the time of writing, the pressure for lifting the embargo has not been strong enough to achieve results.
In an attempt to break the deadlock in national politics, President Buyoya called for a "national debate" on the future of the country, although without success. The July 1996 coup had pushed what remained of the legalist branch of FRODEBU into the arms of the CNDD, and the Tutsi-dominated government was left without any legitimate negotiation partner. While the Kivu crisis and its strong effect on the CNDD/FDD helped Pierre Buyoya domestically, it strengthened his position mainly vis-à-vis his radical Tutsi rivals rather than his Hutu adversaries. The success of the the Kivu campaign (even if the winning was done by others) allowed him to feel sufficiently certain of his position to arrest his rival Jean-Baptiste Bagaza and a number of radical officers and Tutsi militia leaders. If that gave him a new margin for manoeuvre in relation to the Tutsi hard-liners, it did not help him much with the Hutu. Even the most moderate of FRODEBU members are unlikely to be content with anything less than the end of Army rule, while radicals push simply for the annihilation of Tutsi.
Faced with increased activity by the rebels and with no political solution in sight, the Army resorted to the same anti-guerrilla tactic pioneered by the British Government against its Boer enemies during the South African campaign of 1900 and used also by the French in Algeria and the Americans in Vietnam. Population concentration camps were established and all civilians guarded heavily within their perimeters ostensibly "for their own safety", although the aim was to cut off mobile guerilla groups from access to food and cover. The strategy is militarily efficient but catastrophic in political and human terms catastrophic. Generally, in the long run it has been an important cause in losing wars, since the terrible hardships it imposes on the peasant populations turn them against the rulers more efficiently than any subversive propaganda could.
According to a recent UN report, the camps in Burundi are spread throughout the North (in the provinces of Cibitoke, Kayanza, Ngozi, Karuzi, Muyinga and Kirundo) and hold at least 200,000 people. Many of these locations have insufficient access to good water and to food, and the situation is unlikely to improve in the near future. It is such that even the Burundi mediator, Julius Nyerere, an experienced and unemotional statesman, has said that he has the feeling of being able to achieve nothing and of being only "an umbrella covering massacres".
6. THE IMPACT ON UGANDA
The Kivu crisis has also had its impact on Uganda, although to a much lesser degree than either in Rwanda or Burundi. President Yoweri Museveni's links with the Tutsi-led RPF during the civil war in Rwanda and his support for the RPF-dominated government in Rwanda since has firmly placed him in the "American camp". However, given the global geopolitical alignments in the region and the Sudan question which is at the heart of the whole problem, this alignment has earned him the enmity of the regime in Zaire. In mid-November, as the Rwandan refugees poured back into Rwanda, Uganda was surprised to find itself under sudden attack from Zaire. The Ugandan army retaliated quickly, pushing into Zaire in hot pursuit of the attackers.
The guerrilla "movement" which attacked Uganda was a new and rather strange entity calling itself Allied Democratic Forces (ADF) for public relations purposes. Three groups form the ADF. They include Rwandan Hutu radicals who were former members of the interahamwe or FAR and had fled the attacks against the camps in the Goma area, moving northwards along the Rutshuru-Lubero-Butembo-Beni road. They joined a second older group of members of the tabligh Islamic sect who had benefited from the help of the Southern command of the Sudanese Army based in Juba. The tabligh sect, while it originated in Pakistan where it is usually considered moderate and even quietist, had developed in Uganda in the last ten years, taking on a radical Islamist tinge, and there had been many clashes with the police over its illegal occupations of Muslim places of worship. In April 1996, a number of tabligh militants from a variety of ethnic backgrounds had started a small guerrilla group in southern Bunyoro. Beaten by the Ugandan army, they had retreated to the Beni area of Eastern Zaire, where Sudan had later contacted them to help prepare their armed return. The third group, Bakonjo and Baamba members of the so-called National Army for the Liberation of Uganda (NALU), who lived on the slopes of Mount Ruwenzori and combined a grudge against Uganda with a tradition of guerrilla fighting in almost inaccessible terrain, also joined under the same Zairian-Sudanese sponsorship.
This unlikely blend of former Hutu militiamen, Islamic fundamentalists and disaffected tribesmen had thus been formed into a "guerrilla movement" of sorts, although it never issued any manifesto or programme and its few pronouncements betrayed its foreign origin, such as when it "warned President Yoweri Museveni against his expansionism and destabilization of neighbouring countries". But if the origins of the ADF movement were shadowy, its attacks in the Kasese area were very real. Since the Ugandans did not expect attacks from the West, they were at first quite successful. However, as we will see, the subsequent Ugandan counter-attack and pursuit into Zaire caused many diplomatic problems.
7. THE IMPACT ON EASTERN ZAIRE AFTER MID-NOVEMBER 1996
We will deal only briefly with this aspect of the issue as it will otherwise draw us into the more general question of the situation in Zaire, which is beyond the scope of this study. The Kivu crisis, beyond the impact it had on the neighbouring countries, had also a profound impact on Kivu itself, both North and South, and on the neighbouring provinces of Upper Zaire to the north and Shaba to the southwest.
Laurent Kabila, the ADFL leader who had emerged from total obscurity to world prominence in a few weeks thanks to his support by both Rwanda and Uganda, seized the opportunities provided by his quick victory combined with the state of near chaos in Zaire. He started taking on new recruits to bolster his modest force and pushed his troops north to the Watsa-Isiro area near the Sudan border, south along the shore of Lake Tanganyika all the way down to Moba and westwards to Lubutu and Punia, on the road to the regional capital of Kisangani. His actions raised question as to his real ambitions and the possibility that he might want to assume power over the whole of Zaire; the degree of support the ADFL leader was receiving from Rwanda and Uganda and their ambitions with respect to Zaire also became suspect.
Fighting is still taking place on several fronts in Eastern Zaire and the political situation in Kinshasa remains fluid. If one thing appears certain, it is that the status quo of the Mobutu regime has practically no chance of survival. Whether Laurent Kabila succeeds in taking power in Kinshasa or whether some other outcome prevails, Mobutism as the system should meet the same fate as the Cold War which provided its supportive rationale.
The involvement of Rwanda and Uganda in the war has, however, been greatly exaggerated, with "thousands" of Ugandan and/or Rwandan troops described as taking part in the fighting in such improbable places as Fizi and Kalemie, where neither country has any strategic interests. Both Rwandan and Ugandan troops have indeed taken part in the fighting, particularly to establish buffer zones adjacent to their borders, although such a zone remains to be established at the border between Upper Zaire and Uganda. It is also likely that some Rwandans and Ugandans stayed on with the ADFL as specialists in radio transmission or artillery, seconded by their governments or as private mercenaries. Still, the bulk of the ADFL forces are Zairians of various ethnic backgrounds.
Another consequence of the Kivu crisis in Zaire is the humanitarian catastrophe facing the refugees from Rwanda and Burundi who were formerly settled in South Kivu and the Zairians fleeing westward to escape the fighting. As described earlier, the number of refugees remaining in Zaire was contested in international circles for quite some time following the massive refugee returns of mid-November 1996 and were an important element in the diplomatic dispute between France, which pushed for an an international intervention, and the U.S. which was against it. Eventually their existence was admitted again and their numbers even exaggerated to a degree. The situation of yhe refugees remains tragic. They have lost most of their "nuisance value" and have been largely displaced from television screens partly because of the difficulty of access but also the dimishing interest of the world's media for the consequences of the crisis. They therefore tend to be overlooked or at least given very low priority on the international agenda.
The question which we briefly alluded to above remains whether the Kivu crisis can be considered the result of growing anarchy à la Robert Kaplan, or the result of outside interference, especially in the form of a Franco-American confrontation, or more simply the product of regional geopolitical dynamics. The first two explanations rather grandiose and superficially attractive though they may be are nevertheless likely to be false and the most probable explanation seems to be the more humble regional one.
Firstly, in spite of their violence, the Kivu crisis and the subsequent developments cannot be considered anarchical any more than the successive European wars which occurred from the mid fourteenth century onward. The series of conflicts in the Great Lakes region are interlocked much in the same way as were the Central, Western and Danubian European conflicts from the end of the Middle Ages through the two world wars more recently (which were basically European wars, at least initially). These conflicts reflect the nature of the nation-state, with its political form, geographical boundaries and ethnic composition, its guiding principles and its relations with neighbouring nation-states. At the end of the colonial period, many problems pertaining to ethnicity, nationality and the state were left unresolved in Africa, just as they were unresolved in early modern Europe at the time of the religious wars of the sixteenth century. It was these conflicts which indirectly provided the political and cultural framework for the nation-states of Europe in the seventeenth century which prevail to this day (even if the validity of the nation-state is in question in part owing to the development of supra-national entities). Africa two hundred years ago was a gigantic ethnic and cultural jigsaw puzzle, a continent where the nation-state had almost no existence. Giving birth to real nation-states as opposed to the paper constructions left by colonization will not be an easy or painless process. Still, the conflicts occurring there at present should not be considered to be more anarchical than those which took place during the construction of the nation-states in Europe (or in Asia).
Would it be then that the crisis was caused by outside interference? Superficially attractive, this theory does not stand up very well to the test of facts. International actors have no strategic interests in Kivu. Although the mineral wealth of Kivu is often cited, this does not make sense for a number of reasons: firstly, although substantial, the gold mines of Kilo Moto near Bunia and the small diamond mines around Bafwasende and Isiro are not "prime material" in mining terms; secondly, these mines were already in the hands of large multinational corporations, which had no need of a war to control them. Neither was it a question of rivalry and and expulsion of the multinationals; rebel ALDF-leader Laurent Kabila on 26 December 1996 requested that they return and restart the exploitation of the sites which had been abandoned. This type of explanation has its roots in an outdated view of imperialism which dates back to Rudolf Hilferding and Vladimir Ilyich Lenin. Today's global economy has, however, rendered classical imperialism obsolete. The nation-state does not exercise control over transnational capitalism and therefore physical control of territories becomes irrelevant. The large companies have the financial control of capital flows and the technological capacities, without which ground control is useless. In short, Kivu is only of marginal interest to the global world order, and there is no need to fight about it, as it can be more easily controlled through the stock market.
If we dismiss the strategic motive as having disappeared with the end of the Cold War, then why would in this case France and the U.S. engage in "war by proxy"? The answer is largely cultural and has its roots in France's colonial legacy which has given rise to a complex nexus of pride, ideas on language and culture, dreams of former grandeur and memories of a past "civilizing mission". French leaders are extremely concerned about the possible disintegration of Zaire and interpret Rwandan support for Laurent Kabila's ADFL as part of a plan to dismember Zaire and thereafter all of French-speaking Africa with the support of the U.S. This understanding of events feeds on a more general domestic malaise related to economic globalization and a feeling of "cultural invasion" by U.S. lifestyles. But evidence of a U.S. plot is very meagre, to say the least, and the U.S. has tried to reassure its nervous ally by issuing a stern warning to Uganda, Rwanda and Burundi not to intervene in the Zairian civil war.
This leaves us with the most likely cause, a struggle around the definition and control of the nation-state in the area. Uganda suffered an intermittent civil war for similar reasons between 1966 and 1986. Rwanda has been in a state of crisis even prior to its independence and Burundi since 1965. As for Zaire, after the trauma of the 1960-1965 civil war, it was left to crumble slowly under a corrupt and remote dictatorship. None of these countries can be said to have a clearly defined polity, an accepted basis for citizenship, popularly recognized borders or a global form of national consensus, even on a few very limited national issues. But some are closer to such a goal than others. Medium-sized Uganda, for example, is from that point of view much closer to some degree of national integration than either giant Zaire or tiny Rwanda.
In the case of the more unstable political entities, their very existence could be at stake and border realignments are not to be ruled out at some point in the future. The Rwandan genocide has opened a Pandora's box of ethnic, geographical, national and regional conflicts of considerable magnitude. Prudent diplomacy can attenuate those, resolute humanitarian action can take care of the victims, but the notion that somehow goodwill or even force (which the international community does not appear ready to use) can stop the process is not sustainable. Things will evolve in their own way, at their own speed, according to their own logic. For better or for worse, Africa has now come of age, and that situation implies the right to make one's own decisions, including violent ones.
The views expressed in the papers are those of the authors and are not necessarily those of UNHCR.
 On the origins of crisis in North Kivu see: United Nations, Department of Humanitarian Affairs, Situation Report on Masisi, North Kivu, Zaire (Nairobi, 26 February 1996); Fédération Internationale des Droits de l'Homme, Forcés de fuir: la violence contre les Tutsi au Zaire (Paris, July 1996); Dialogue, "La guerre de Masisi", Special Number, No 192 (August-September 1996); Doctors Without Borders USA, Ethnic War in Eastern Zaire, Masisis 1994-1996 (New York, November 1996); Amnesty International, Zaire: Violent Persecution by State and Armed Groups (London, November 1996).To the best of our knowledge there is no literature on South Kivu concerning recent political problems and ethnic tensions.
 In Bantu languages, prefixes considerable change the meaning of a word. Using -nyarwanda, for example, a Munyarwanda means one person, Banyarwanda include several persons or an entire tribe, while Kinyarwanda is the language. In in the same way, for -rundi there is a Murundi, the Barundi and Kirundi.
 Barthememy Bisengimana, a Tutsi immigrant in charge of the President's office in Kinshasa, was the second most powerful man in the country next to President Mobutu during a period of ten years.
 Both are minority tribes, while the Banande form the majority. Their marginal status made them all the more desirous to eliminate the Banyarwanda whom they considered to be "foreigners".
 On their origins, see Monique Chajmowicz, "Kivu: les Banyamulenge enfin à l'honneur", Politique Africaine (December 1996), pp 115-120.
 Although a few Bahutu client families had moved to South Kivu with their Batutsi shebuja (patrons), with time they had become "tutsified".
 One of the reasons for the persecution of the Banyarwanda in Kivu was the attempt by the political forces around President Mobutu to build a clientele in the traditionally rebellious East by humouring the local tribes on the one hand and courting the support of the Bahutu refugees from Rwanda, who formed a ready constituency, on the other.
 Several interviews by the author with members of South Kivu civil society. London, February 1996.
 For a good assessment of the situation at that time, see United Nations, Security Council, International Commission of Inquiry for Rwanda, Third Report of the International Commission of Inquiry for Rwanda to the UN Security Council (New York, 28 October 1996).
 See Le Monde diplomatique, Colette Braeckman, "Hantise du génocide au Burundi", March 1996. Although the political situation was - and remains - practically intractable, there was never any danger of a real genocide in Burundi, for the simple reason that neither the Army nor the Hutu rebels had the practical means of committing one. The danger was of course that one side might gain a definite advantage and then be free to act as it wished.
 Jean-Baptiste Bagaza had arranged for the leak of an internal memo from his political party, the National Recovery Party (Parti pour le redressement national - PARENA), in which he outlined his views on the question.
 On the war in the 1960s, the best reference work is Catherine Coquery-Vidrovitch, Alain Forest and Herbert Weiss (eds), Rébellions-révolution au Zaire (1963-1965) (Paris: L'Harmattan, 1987), 2 vols. On Laurent Kabila's later chequered career see Le Soir, Colette Braeckman, "Le Kivu s'interroge sur ses nouveaux maitres", 6 November 1996; Libération, Claude Wauthier and Stephen Smith, "Le passé retrouvé de Laurent Kabila", 7 January 1997.
 On the question of Rwanda's involvement in the Kivu conflict, see Le Soir, Colette Braeckman, "Le Zaire entend prouver l'agression rwandaise", 29 October 1996; Libération, Marie-Laure Colson and Stephen Smith, "L'Intervention des militaires rwandais", 31 October 1996; Le Figaro, Pierre Prié, "Zaire/Rwanda: une guerre éclair planifiée par Kigali", 1 November 1996; United Nations, Integrated Regional Information Network (IRIN), Bulletin, No. 4 (2 November 1996); Reuter, Chris MacGreal, "Zaire: Rebels Without a United Cause", 15 November 1996.
 When Laurent Kabila's movement, the Alliance des Forces Démocratiques pour la Libération du Congo-Zaire (ADFL), started operating around Uvira in late September 1996, it included only about 1,500 men, most of whom were Banyamulenge.
 See for example The Washington Post, Stephen Buckley, "In Zaire, Good Intentions Yield Tragic Effects", 10 November 1996.
 Typical of that blindness was the attitude expressed in the White Fathers' review Dialogue in its Special Number on "Les réfugiés rwandais en Afrique Centrale", No. 191 (June-July 1996), where Father Guy Theunis wrote (pp 35-36): "The main authority in the camps is that of the local administration, whether Zairian or Tanzanian, helped by the UNHCR." In over 120 pages, he includes not a word either on the political control of the former government over the refugees or the military activities of the ex-FAR. The impression given in this otherwise respectable publication is of refugee camps that are like large summer camps, poor of course, but run by well-meaning Christians who spent most of their time working hard to promote community development. This smooth unrealistic façade was the general attitude of almost all bodies dealing with the Rwandan refugees between the summer of 1994 and October 1996.
 The Rwandan refugees from the southern camps fled mostly westwards deeper into Zaire and remained there at the time of writing (mid-February 1997). A minority walked past the lake and to join their northern brethren near Goma. The Burundi refugees were divided evenly with about half returning to Burundi, often under heavily armed escort of the Banyamulenge forces, and half joining the Rwandans in their westward flight.
 The factual account of the situation on the ground which follows is pieced together from interviews with a number of actors and observers, both Rwandan and foreign. It is of course not authoritative but probably represents a close approximation to what happened.
 Libération, Stephen Smith, "Non-assistance à un million de réfugiés", 4 November 1996.
 Libération, "Zaire: le forcing français", 7 November 1996.
 See for example The Guardian, Martin Woollacott, "A Minefield Beneath UN Boots", 9 November 1996 or The Sunday Times, Andrew Roberts, "Zaire: A Problem for France Alone", 10 November 1996.
 See The Economist, "Someone Else's Doing, Someone Else's Problem", 9 November 1996; Le Monde, Afsane Bassir Pour, "Washington freine toujours l'envoi d'une force multinationale au Zaire", 11 November 1996.
 See United Nations, Integrated Regional Information Network (IRIN), Bulletin, Nos 25 27 (15 - 18 November 1996); The Financial Times, Anthony Goldman, "Exodus Averts Disaster: Aid Workers Left Stunned as Rwandans Flee Home", 16/17 November 1996; Le Monde, Dominique Le Guilledoux, "Les réfugiés Hutu du Zaire regagnent en masse le Rwanda", 18 November 1996.
 Oxford Analytical Daily Brief, 16 December 1996 (electronic format).
 See Reuter, "Zairians Say 300 Refugees Killed by Tutsi Rebels", 25 November 1996; Charles Corey, Proofs of Atrocities in Eastern Zaire are Multiplying (Washington: USIA, 3 December 1996); Le Monde, "Plusieurs charniers contenant les corps de réfugiés Hutu ont été découverts dans l'Est du Zaire", 8/9 December 1996.
 Le Monde, "Le reflux des réfugiés remet en question les objectifs de l'intervention au Zaire", 19 November 1996; The Independent, John Lichfield and Christopher Bellamy, "West Struck by Second Thoughts on Zaire Force", 20 November 1996.
 The Independent, Imre Karacs, "Zaire Military Option Ruled Out", 25 November 1996.
 United Nations, Integrated Regional Information Network (IRIN), Bulletin, No. 52 (9 December 1996).
 International Herald Tribune, James MacKinley, "Tanzania Forces Refugees into Camps", 14/15 December 1996; Financial Times, Michela Wrong, "Army Drives Hutu Masses into Rwanda: Refugee Camps Closed in Acts of Forced Repatriation", 16 December 1996.
 All Africa Press Service [Nairobi], "Tanzania Expels Two Catholic Priests", 21 December 1996.
 United Nations, Integrated Regional Information Network (IRIN), Bulletin, No. 28 (17/18 November 1996).
 United Nations, Integrated Regional Information Network (IRIN), Bulletin, No. 38 (24/25 November 1996). According to eyewitnesses to the refugee crossings at Gisenyi on 14 to 18 November, the crowd was so dense and moved so steadily that it was impossible to undertake a head count an on-the-spot. All attempts at later counting were haphazard and incomplete.
 Libération, 16 January 1997.
 United Nations, Integrated Regional Information Network (IRIN), Bulletin, No 97 (8/10 February 1997).
 See Le Courrier International, "L'hypocrisie humanitaire", Special Number, No. 316 (21-27 November 1996) or the article by former Action Internationale Contre la Faim (AICF) director Sylvie Brunel, "Indécente communauté internationale" in Le Monde, 3 January 1997.
 The Times, Sam Kiley, "Rwanda Gripped by Fear as Hutus Return from Zaire", 26 November 1996.
 The Daily Telegraph, Alec Russell, "Hutus Return to Reclaim Homes from Tutsi", 27 November 1996; The Economist, "Welcome Home", 7 December 1996.
 Le Monde, "Les premiers procès du génocide s'ouvrent au Rwanda", 28 December 1996.
 Le Monde, "Plus de 5,000 réfugiés arrétés à leur retour au Rwanda", 9 January 1997.
 See Amnesty International, Memorandum to the Rwandese Government: Amnesty International's Concerns and Recommendations for Fair Trials in Rwanda (London, March 1996).
 The Economist, "Justice for Genocide, Rwanda Style", 11 January 1997. The sentences were not in fact carried, which alleviated the situation somewhat, although this failure added an element of weakness and confusion. The only exception was Frodwald Karamira who was condemned to death in a trial that more closely approximated international standards. See Libération, Jean-Philippe Ceppi, "Karamira, génocideur équitablement jugé", 16 February 1997.
 Le Monde, "Le fonctionnement du tribunal international pour le Rwanda doit être amélioré", 11 January 1997; Libération, Pierre Haski, "Faux départ pour le tribunal sur le Rwanda", 17 January 1997; United Nations, Integrated Regional Information Network (IRIN), Bulletin, No. 99 (13 February 1997).
 Le Monde, Afsané Bassir Pour, "Un rapport de l'ONU met en cause le tribunal pour le Rwanda", 14 February 1997.
 See United Nations, Integrated Regional Information Network (IRIN), Bulletin, No. 82 (18-20 January 1997).
 United Nations, Integrated Regional Information Network (IRIN), Bulletin, No. 86 (24 January 1997).
 United Nations, Integrated Regional Information Network (IRIN), Bulletin, No 92 (1-3 February 1997); Libération, "Rwanda", 10 February 1997.
 United Nations, Integrated Regional Information Network (IRIN), Bulletin, No. 94 (5 February 1997).
 Libération, Rwanda", 10 February 1997.
 Africa Confidential, "Balance of Terror", 6 September 1996.
 Interview with a Burundi Member of Parliament, Paris, October 1996. The threat came mostly from former President Jean-Baptiste Bagaza, who had close links with both the Army and the Tutsi extremist militias.
 Libération, "Parlement et partis restaurés au Burundi", 13 September 1996.
 BBC Summary of World Broadcasts, "Hutu Rebels Reportedly Ready to Hold Conditional Talks", quoting France 2 TV, 18 September 1996. The reason for this new willingness to talk was the pressure which was beginning to mount on the FDD rear bases from the Banyamulenge fighters.
 Adonia Ayebare, "Burundi Leader May Cut Short Military Rule", The East African, 23-29 September 1996; BBC Summary of World Broadcasts, "Major Buyoya Calls for Dialogue with Armed Groups", quoting Radio Burundi, 27 September 1996.
 BBC Summary of World Broadcasts, "20,000 Hutu from Burundi Flee Camps Following Attack", quoting Radio France Internationale, 13 October 1996.
 Reuters, 5 November 1996.
 Le Monde, Isabelle Vichniac, "Des émissaires de l'ONU font état de tris, de disparitions et de massacres", 9 November 1996; Le Soir, "Massacre au Burundi", 23-24 November 1996; The Times, Sam Kiley, "Amnesty Accuses Burundi Army of Massacring Hutus", 22 November 1996.
 The Guardian, Chris MacGreal, "Rebels Stoke Burundi War", 2 December 1996; The Economist, "The Other War in Central Africa", 14 December 1996.
 See Le Monde, "l'Armée du Burundi reconnait avoir tué 126 Hutu à leur retour de Tanzanie", 14 January 1997. The article shows how the fighting started in the Katale camp in Tanzania between refugees who supported of the FDD and followers of the Hutu People's Liberation Party (Parti pour la libération du peuple hutu (PALIPEHUTU), a radical Hutu ethnic party. The Army ended by shooting indiscriminately at the crowd since it could not control it as it crossed the border.
 United Nations, Integrated Regional Information Network (IRIN), Bulletin, No. 84 (22 January 1997).
 Agence France Presse, 25 November 1996.
 Reuters [Rome], "FAO Says Sanctions Are Biting Hard", 5 December 1996.
 Le Monde, Isabelle Vichniac, "Les effets de l'embargo pourraient entrainer une famine au Burundi", 3 December 1996.
 Several interviews with Burundi citizens and with Europeans returning from Burundi, Paris, October to December 1996. See also François Misser, "Burundi Sanctions Hit the Poor", The New African (December 1996).
 Although the embargo was initially decreed (and enacted) by the countries of the region (Kenya, Tanzania and Uganda, with the reluctant agreement of Rwanda), European Union countries also supported it. Because of its effects on the regional configuration post-Kivu crisis (i.e. a situation where the embargo hits the poor but is unlikely to drive either the FDD or President Buyoya to the negotiation table), this support is being questioned by some.
 Le Monde, "Le débat sur la crise au Burundi boudé par les partis politiques", 29 January 1997.
 Libération, "Vague d'arrestation au Burundi", 20 January 1997.
 This difference in objectives (among other things) has been a constant source of disagreement between CNDD/FDD, which has a moderate wing, and PALIPEHUTU which stands for strict Hutu Power. It was one of the causes of the fighting in the camps in Zaire early in 1996 and was again at the heart of the intra-Hutu clashes in Tanzania in January 1997.
 See United States Information Agency [Geneva], 31 January 1997 and Le Monde, 31 January 1997, quoting a recent UN report as saying that, in Karuzi Province alone, twenty-four locations hold over 100,000 people.
 Le Monde, "Le médiateur dénonce l'impasse au Burundi", 5 February 1997.
 This support should not be exaggerated: when Ugandan and Rwandan interests were in conflict, Yoweri Museveni would quickly choose the first, as for instance when he forced General Kagame in August 1996 to support the embargo against Burundi, because he put a higher value on his good relations with Kenya and Tanzania, who had promoted the embargo, than with his tiny neighbours to the south (interview with a Ugandan cabinet civil servant, Paris, October 1996). Also, contrary to many highly unrealistic ideas about a "Tutsi Empire", President Museveni's support did not extend to President Buyoya and Tutsi supremacy in Burundi. See Reuter, "Uganda's Museveni Wants Burundi Leader Removed", 20 November 1996.
 Given the complexity of this question and the limited space available for this essay, we will not go into the details of the geopolitical alignments extending beyond the Great Lakes area. For a quick preliminary treatment of the question, see Le Monde diplomatique, Gérard Prunier, "Le Soudan au centre d'une guerre régionale", February 1997, and the conclusion of this essay.
 La Libre Belgique, Marie-France Cros, "Heurts à la frontière Zaïro-Ougandaise", 19 November 1996.
 Libération, "Le Zaire dénonce des incursions ougandaises", 1 December 1996.
 The following account is pieced together from interviews with a number of informers (French, Zairians, Ugandans, Rwandans and Sudanese) who will need to remain anonymous.
 NALU was in fact the latest modern incarnation of the old Rwenzururu Movement. See Kirsten Alnaes, "The Songs of the Rwenzururu Rebellion" in P.H. Gulliver (ed), Tradition and Transition in East Africa (London: Routledge and Kegan Paul, 1969), pp 243-272; Martin Doornbos, "Kumanya and Rwenzururu: Two Responses to Ethnic Inequality" in Robert Rotberg and Ali Mazrui (eds), Power and Protest in Black Africa (New York, 1970). This rebellion started in the 1950s and was aimed against the British, to protest the inclusion of the Bakonjo tribe in the Toro district, where its members felt they suffered discrimination. This administrative situation did not change with independence, and the Bakonjo continued fighting against all the successive governments in Uganda the last forty years.
 United Nations, Integrated Regional Information Network (IRIN), Bulletin, No 92 (1-3 February 1997). Guerrilla movements are usually more concerned about the internal politics of their own countries than about preventive diplomacy.
 There were already two other guerrilla movements sponsored by Sudan operating in Uganda, the Lord's Resistance Army (LRA) and the West Nile Bank Liberation Front (WNBLF). But both were fighting in the North.
 Although Burundi is usually also alleged to be a supporter of Laruent Kabila's forces by foreign commentators, including the U.S. State Department (see Le Monde, "Washington met en garde l'Ouganda, le Rwanda et le Burundi", 7 February 1997), the conclusion is based more on logic (Burundi "should" support Kabila) than on factual evidence. In fact, Burundi is hardly able to support itself, far less an outside force or entity.
 See Africa Confidential, "Zaire Starts to Crack", 1 November 1996; Le Monde, Frédéric Fritscher, "La rébellion du Zaïre risque d'accélérer l'éclatement du Zaire", 27 November 1996; Libération, Stephen Smith, "L'armada des mercenaires au Zaïre", 24 January 1997; Le Monde, Thomas Sotinel, "La débacle de l'armée zaïroise explique la progression des rebelles dans l'Est du pays", 16-17 February 1997.
 As a still obscure rebel in Eastern Zaire, he had briefly occupied Moba in 1984, and was ejected by French-led Zairian paratroopers.
 Libération, Stephen Smith, "Zaïre: les rebelles à l'assaut", 10 February 1997.
 See Le Nouvel Observateur, Laurent Bijard, "Kabila ira-t-il trop loin ?", 13-19 February 1997; The Guardian, Howard French, "Zaire's Mercenaries Prepare for Last Stand", 13 February 1997.
 François Misser, "Mobutu's Darkest Hour", The New African (February 1997).
 Le Monde, "Des troupes ougandaises et rwandaises seraient entrées au Zaire", 31 January 1997.
 See Africa Confidential, "Outside Agents", 31 January 1997. Also several interviews by the author with NGO personnel, journalists and civil servants of various nationalities, Paris, December 1996 - January 1997.
 The percentage of Kinyarwanda speakers was very high during the Kivu crisis proper (i.e. between late August and mid-November 1996), although it has been declining ever since. A French journalist familiar with the area who spent a few days with ADFL forces described the tribal mix as "one third people from Eastern Zaire with a large number of Bashi, one third Zairians from the diaspora in East Africa, who have come to join him, and one third Kinyarwanda speakers, who are difficult to break down into Munyamulenge, Tutsi from Masisi and 'true' Rwandese". This approximate ethnic breakdown was later confirmed to the author by a Zairian researcher living abroad who followed the ADFL development closely since its inception.
 Libération, Stephen Smith, "Rescapés de la jungle zaïroise", 22 January 1996.
 The Guardian, Chris MacGreal, "Officials Play Numbers with Missing Refugees", 25 November 1996.
 Le Monde, Philippe Lemaitre, "Cinq cent mille réfugiés souffrent de la faim dans l'Est du Zaïre", 4 February 1997.
 See Amnesty International, Great Lakes Region Refugees Victims of Current Crisis (London, 7 November 1996) and UN Secretary General Kofi Annan's demand on 14 February 1997 that belligerents at least not fight near or in the Tingi Tingi camp, where 150,000 refugees were trying to survive (Le Monde, 16-17 February 1997), particularly as most of the actual fighting was done by ex-FAR soldiers, who were mixed with the refugees, while the Forces Armées Zaïroises (FAZ) had run from the ADFL forces rather than fight,
 Robert Kaplan, "The Coming Anarchy", The Atlantic Monthly (February 1994), pp 44-76.
 It is interesting to note that the pre-colonial kingdoms of the Great Lakes historically had already seen existence as a dense cluster of quasi-national states, which is rather unusual in Africa. Many of the so-called "ethnic" conflicts in East Africa since independence are caused by a failure to operate a transition from the old pre-colonial state structures to a modern nation-state in countries where the concept nevertheless had real cultural roots.
 Attempts to create a nation-state are not always successful, as the example of Yugoslavia between 1918 and 1990 shows.
 Africa Confidential, "Zaire: Mining for Trouble", 13 December 1996.
 Canada-based Banro-Cluff Mining in South Kivu and U.S.-dominated Mindev-Barrick in North Kivu. See La Lettre Afrique Energies, No. 311 (4 December 1996) and No. 313 (8 January 1997).
 See Wolfgang J. Mommsen, Theories of Imperialism (London: Weidenfeld and Nicolson, 1980).
 The expression is used in The Economist, "Fashoda Revisited", 8 February 1997.
 For a discussion of this French syndrome see Gérard Prunier, The Rwanda Crisis (1959-1994): History of a Genocide (London: Hurst, 1995), pp. 102-107.
 Le Monde, Frédéric Fritscher, "La question du Zaïre oppose toujours Paris et Washington", 2-3 February 1997. During the Kivu crisis, this author was in frequent contact with high-ranking French civil servants dealing with the problem. Their views were at times contradictory on tactical matters but largely unanimous as to the reality of a U.S. determination to undermine French influence in Africa. All saw the Kivu crisis as a clear example of that determination.
 The Economist, "Great Lakes of Blood", 2 November 1996.
 The remarkable personality of the recently nominated UN Special Representative, Mohamed Sahnoun, might conceivably have such an effect.
 This coming of age is particularly evident in the changing role of the White mercenaries in Zaire. In 1964-1965, they made the difference in the fighting; today, they carry hardly any significance.