Preliminary Thoughts on the Congo Crisis*

It is widely believed that the root problem of the African state is the artificial nature of its boundaries; were not these boundaries, after all, first arbitrarily drawn up at the Berlin Conference of 1885–86 and then imposed from the outside? I would like to begin by putting forward two suggestions for consideration. One, all boundaries are more or less artificial. Two, if we want to understand the crisis of the state, an understanding of how power is organized is likely to prove a more illuminating starting point than the nature of boundaries.

Citizenship and the Congolese State

There is a thesis now common in Africanist political science that the state is collapsing in more and more African countries. The Congo is often held up as an example of this. The key problem with this thesis is that it proceeds by making analogies and, in the process, overlooks what is different about the state in Africa. Its starting point is not the state in Africa, particularly the type of state created under colonialism, but the assumption that the African state is an attempt to reproduce the modern state in Europe. Hence the conclusion that the attempt to imitate the original has failed. The difference is understood as evidence of a failure, and is then theorized as a collapse.

State crisis in Africa can be illuminated by Africa’s experience with globalization. From this point of view, Africa has gone through several globalizations, starting from the original diaspora that led to the peopling of the world, to slavery, to colonialism, and to the current globalization whose post-Fordist waves are said to be dissolving the nation-state. My point is that the roots of state crisis in Africa lie not in the current globalization but in an earlier one: colonialism. The key problem with the talk about the “crisis of the state” in Africa is that it misses its colonial genealogy. In doing so, it misses the link between the current problems of African polities and the bifurcated nature of African states—a phenomenon forged in the colonial period.

The state in Africa is a product of a radically different history, a history of conquest. Alien power faced the problem of legitimacy. In [End Page 53] response, the British reformed their mode of rule, first in equatorial Africa in the early part of this century. They called it “indirect rule.” The French followed suit in the 1920s, when they shifted from “assimilation” to “association” as the basis of colonial rule in the African colonies. Belgium effected a similar shift in its African colonies in the 1930s.

It is this reform that begins to explain what is different about the state in Africa. Indirect rule reorganized colonial power as two distinct authorities, each ruling through a different legal regime, one civic and the other ethnic. The basis of civic power was the central state, which expressed its will through civil law. In contrast, the local state was organized as a Native Authority, overseeing the implementation of a customary law. Civil law claimed to speak the universal language of rights, but the regime of rights was applied only to the population of metropolitan origin, described as racially distinct. Natives were portrayed as creatures of habit, rather than as being capable of a rational exercise of freedom. It was said that they needed to be ruled through a different regime, one that would enforce custom. This, however, did not lead to the creation of a single customary law and a single customary regime ruling all natives. Instead, the colonial power claimed that each ethnic group had its own distinctive custom; it thus created a different set of customary laws for each ethnic group and established a separate Native Authority to enforce each set of laws. The final result was a Janus-faced power. Like civic power, native power too was a colonial creation. The difference, however, was that while civic power was racialized, the Native Authority was ethnicized.

This bifurcated form of the state underwent a reform after independence. The reform process varied from one country to another, but one could discern the more radical from the more conservative current. The Congolese reform followed the more conservative variant: while civic power was deracialized, the Native Authority remained ethnicized. In fact, with the withdrawal of the Belgian cercle commander at independence, one could say that the ethnic aspect of the Native Authority got further entrenched.

When Africanists speak of the collapse of the state, they are speaking of the collapse of civic power, not that of the Native Authority. The point is that what holds Congo together is not as much the civic power in Kinshasa and Kisangani and so on, but the hundreds of Native Authorities that control the bulk of the population in the name of enforcing “custom.” For southern Africans, rural Congo is better thought of as a giant federation of Bantustans, a reformed colonial state. 1 [End Page 54]

Citizenship and the Banyamulenge Question

The term Banyamulenge has undergone a radical change in meaning over the past decade. Originally, it referred to only one section of Kinyarwanda-speaking residents of eastern Congo (Kivu), those in South Kivu. They had initially come in as refugees following a major battle of succession in the precolonial Kingdom of Rwanda in the 1880s, and eventually settled in a place called Mulenge; thus the Banyamulenge, the people of Mulenge. More recently, with the crisis of the Mobutu regime and the expulsion of all Kinyarwanda-speaking residents of Congo—starting with the Banyamulenge of South Kivu—the term has come to identify all Kinyarwanda-speaking residents of Congo. In this essay, I shall move between both uses of the term; the context should illuminate the particular use under discussion.

In the postindependence period, the bifurcated state has become the basis of a bifurcated citizenship. The colonial state enforced a dual political identity. It made a distinction in law between those indigenous and those not. It went further by drawing a racial distinction among the nonindigenous and an ethnic distinction among the indigenous. It thus imposed a racialized civic identity on the former and an ethnicized native identity on the latter. The postcolonial state deracialized the civic identity; civic citizenship stopped recognizing any difference based on race or place of origin. But it continued to reproduce the native identity as ethnic. The result has been a double, or bifurcated, citizenship: one civic, the other ethnic. Civic citizenship is a consequence of membership of the central state; it is specified in the constitution and is the basis of rights. These are mainly individual rights, in the political and civil realm. In contrast, ethnic citizenship is a consequence of membership in the Native Authority; it is the source of a different category of rights, mainly social and economic. Furthermore, these rights are not accessed individually but by virtue of membership in an ethnic community. The key socioeconomic right, it is worth mentioning here, is the right to use land as a source of livelihood. Herein lies the material basis of ethnic belonging, particularly for the ethnic poor.

The political consequence of this bifurcated citizenship is that while everyone is a civic citizen, a citizen of the state called Congo, not everyone has an ethnic citizenship. Since civic citizenship has been deracialized, everyone—indigenous or not—is a citizen of Congo. But only those indigenous have a Native Authority and, as a consequence, an ethnic citizenship. Because they do not have a Native Authority of their own, immigrants considered nonindigenous are excluded from ethnic citizenship. The immediate practical consequence of this is that nonindigenous citizens [End Page 55] are denied “customary” access to land since they do not have their own Native Authority.

The Native Authority in Congo is three-tiered. It is the chief of the second tier who controls access to land. The Banyamulenge in Kivu Province have their own chief at the first tier, but they are treated as ethnic strangers at the second tier. The 1981 law accepted the Banyamulenge as civic citizens, but not as ethnic citizens with the right to their own Native Authority. It is worth noting that the Banyamulenge identity—as that of other immigrants from Rwanda, like the Banyamasisi and the Banyaruchuru—is territorial, not ethnic: the Banya-Mulenge refers to those of the place called Mulenge. This group identity is, in turn, like a geological deposit, layered, with each layer signifying a different history. Starting from those who were there when the borders of colonial Congo were first demarcated, the identity Banyamulenge includes every wave of immigrants to Mulenge, including those who came in the wake of the genocide of 1994. They are all Banyamulenge. The irony of a common identification for all Rwandese-speaking persons resident in a single place, regardless of when they got there, is that the depth of claim of those longest resident is obscured by the shallowness of the claim of the latest wave of immigrants. It is common to hear civil society organizations in Kivu Province complain thus: you can’t tell who is who and when they got here; they all claim to be Banyamulenge, even those who got here only yesterday. The consequence is that, in native eyes, Banyamulenge becomes a collective identification of those nonindigenous.

This question is not unique to Congo. It is a dilemma that arises wherever there are substantial numbers of immigrants and where the state inherited from colonialism draws a structural distinction between two kinds of citizens: those indigenous, and those not.

One can see this in the case of Uganda, historically another neighbor of Rwanda with a substantial number of Kinyaruanda-speaking immigrants. It is the state oppression of all Banyaruanda—whether they were born in Uganda or not, and whether they were civic citizens or not—by the Obote II government that led to Banyaruanda youth joining the National Resistance Army (NRA) led by Yoweri Museveni. It is estimated that as many as four thousand of the roughly sixteen thousand NRA guerrillas who marched into Kampala in January 1986 were Banyaruanda.

Ironically, the Banyaruanda question became a major social—and political—question in Uganda as individuals of Rwandese origin occupied prominent state positions under the NRA. The social question was connected to land and became a major public issue when Banyaruanda squatters confronted Baganda ranchers in Mawogola county in Masaka District. The squatters laid a claim to land and utilized their majority status in the county to press it home as a democratic demand. Ranchers [End Page 56] countered by questioning whether a nonindigenous group can have land rights as do natives. As the land question was translated into a nationality question (specifically, the Banyaruanda question), attention focused countrywide on the prominent position of individuals of Rwandese origin in the hierarchy of the NRA and that of the state. The social question triggered a political crisis. This was in the late 1980s. The important point to understand is that the key impetus behind the decision of the Rwanda Patriotic Front (RPF) to cross the border from Uganda into Rwanda in 1990 was not a political crisis in Rwanda, but one in Uganda.

How is one to come to terms with the conflict between those the Native Authority identifies as ethnically indigenous and thus with a claim to customary rights, and those it brands as ethnically nonindigenous and thus lacking in any such customary claim? We can identify two different solutions from recent developments in the region. The first solution is to create a separate Native Authority for those branded as nonindigenous, such as the Banyamulenge in Kivu Province. This was the solution promoted by the Rwandese army in Congo, after Kabila came to power. But it was also a solution that was very unpopular in Congo, especially among those living in Kivu. From their point of view, this solution meant that the land over which the Native Authority would be created would be the land that would be appropriated by the Banyamulenge. Not surprisingly, the solution advanced by the Rwandese army exacerbated ethnic tension in Kivu Province.

An alternative to changing the boundaries of existing Native Authorities to create a new Native Authority is to reform the very nature of power organized as the Native Authority in the local state. This was the solution arrived at in practice by the NRA during its guerrilla war in the Luwero Triangle. It was also the solution endorsed as policy by Uganda’s National Commission of Inquiry into Local Government System, set up in 1986. This solution was to democratize the local state by dismantling the system of chiefship, by turning the chief into an administrative officer supervised by an executive committee elected by a village assembly of all adults resident in the village, whether indigenous or not. It thus redefined the basis of customary rights from ethnicity to locality (territoriality). While this was a better solution than multiplying the number of Native Authorities—and thus multiplying the problem itself—it did not do away with the problem. As subsequent developments showed, it would not hold without a political alliance of working people, of both those indigenous and those nonindigenous, cementing the fissure the conquest state had introduced into the local population. As this alliance began to erode toward the late 1980s, the Ugandan solution also began to unravel.

The lesson of the Ugandan experience is that the reform of the state, and therefore of citizenship laws, will not be an automatic consequence of [End Page 57] elections and majority rule. It will require a combination of an enlightened leadership with the organizational capacity and will to undertake a protracted education of the population, both those indigenous (the majority) and those not (the minority). This lesson is confirmed by the Congolese experience. For it is worth bearing in mind that while the Mobutu-sponsored 1981 law granted civic citizenship to the Banyamulenge, the National Conference of democratic forces in Congo opposed this law when it convened a decade later, in 1991.

Kivu Province and Its Link with Rwanda

Kivu Province is where losers in Rwanda traditionally end up; and it is in Kivu that they prepare to return to power in Rwanda. That, at least, is how conventional wisdom in Goma and Bukavu has it. This, no doubt, introduces a double tension in Kivu, both internal and external, the former within Kivu society and the latter between Kivu and the power in Rwanda. It is also a tension that has tended to grow in intensity as the refugee and exile population has grown in size. No wonder Kivu found it difficult to contain this pressure in the aftermath of the Genocide of 1994. Then, over a million Rwandese refugees streamed into Kivu, both north and south, and set themselves up in camps. The former state army (FAR) and the militia that was active in the Genocide (Intrahamwe) continued to be supplied militarily by the French and controlled the camps, while international NGOs, mostly American-funded, fed them.

The insertion of a million plus refugees in camps that were armed and resourced from the outside had a devastating effect on civilian life in Kivu. It led to the dollarization of the economy and to the militarization of ordinary life. The Intrahamwe roamed the countryside, often collaborating with the Congolese army. In response many of the Native Authorities created their own militia. These are the Mayi Mayi. 2 The anatomy of political life in Kivu began to resemble that in Rwanda. As in Rwanda, where every political party had come to have its own militia by the Genocide of 1994, so in Kivu every Native Authority began to acquire its own militia in the post-Genocide period.

The Mayi Mayi joined the First Rebellion in Congo, the rebellion against Mobutu, but opposed the rebellion when it came to power. Why? They joined it when the rebellion targeted the Intrahamwe and the allied Congolese state army. And they opposed it when they saw the rebellion turn into the spearhead of a Rwandese-installed government. On the one hand, the Rwandese army began to resemble an army of occupation, its commander even being formally appointed the commander of the Congolese National Army. On the other, this same army began to actively [End Page 58] support the demand by the Banyamulenge that they be given a separate Native Authority in south Kivu.

Militarization spread two tendencies in Kivu and in Congo, as it had in Rwanda. First, the link forged between militarization and genocidal tendencies inside Rwanda spread across its borders. The First Rebellion led to an indiscriminate slaughter of Intrahamwe and of unarmed Hutu refugees. Those responsible for that slaughter were part of the forces that opposed a UN inquiry into the matter. They remain a part of the military forces of the Second Rebellion. The Second Rebellion, in turn, evoked from the Kabila government an invitation to the general population in Kivu to slaughter indiscriminately not only invading forces from Rwanda but also the Banyamulenge in the rebellion, and even any ordinary Tutsi civilian. We need to keep in mind that genocidal tendencies are present on both sides of the conflict, that of the government and that of the rebellion.

Second, the militarization of politics has reduced all credible politics to armed politics. The result is to marginalize all civil society-based politics. Once again, this tendency has become consolidated in Rwanda. Beginning with a marginalization of the Hutu opposition autonomous of both Hutu Power and the RPF, the tendency of state politics in Rwanda has been to demonize all politics autonomous of the RPF—regardless of its political character—as “genocidaire.” It is a tendency strong in both the Kabila government in Congo and in a section of the political leadership of the Second Rebellion. The tendency to reduce all credible politics to armed politics is also present on both sides, that of the government and of the rebels.

The Second Rebellion

We must begin by rejecting two tendencies, one that paints the rebellion as entirely a home-grown affair, and the other that would make us believe that it is wholly a foreign invasion. The rebellion is characterized by both internal and external factors. I shall begin with the internal factor.

It is clear that the political organ of the rebellion—the Rally for Congolese Democracy (CDM)—is a hastily put together affair. As such, it lacks cohesion. We can identify at least three distinct and even opposed tendencies in it. The first is that identified with its chairperson, Wamba-dia-Wamba. The second is identified with the Banyamulenge group, closely allied to Rwanda. And the third is identified with the ex-Mobutists. In this coalition, the balance of forces is clearly in favor of those with military forces, being the pro-Rwanda Banyamulenge group and the ex-Mobutists. These also represent the core of the militarist tendency in the rebellion. [End Page 59]

Let me illustrate my point with two examples. In a recent CNN report on the war, the military commander in Goma stated—in the presence of Wamba-dia-Wamba—that he and his military colleagues would act against Wamba’s political group if they became dissatisfied with their rule. In a recent interview with De Standard in Belgium, Wamba explained that the political wing lagged behind the military system because it was formed after the military operation. He added that he hoped that those in favor of political liberation of the people would eventually gain advantage in the movement. Such honesty and transparency is rare to come by in politics. Wamba-dia-Wamba is a fine scholar and a person of great integrity. While he is formally the political leader of the rebellion, I suggest we will make better sense of both the rebellion and the tendency identified with Wamba if we understand it as opposed to the dominant political tendency in the rebellion.

The fact is that both the rebellion and the government are internally contradictory. The dominant tendency in both is characterized by a militarized form of politics. The question we need to ask is how to demilitarize politics? As a starting point, I suggest it requires making politics more inclusive, particularly by recognizing the legitimacy of unarmed opposition. To recognize the political limits of both the government and the rebellion is to recognize the political opposition to Mobutu that gelled as the Sovereign National Conference in 1990–91. Comprising over four hundred civic groups and over a hundred political groups, this opposition thrived in the period from 1990 to 1996. It was neither a part of the First Rebellion (against Mobutu) nor of the Second Rebellion (against Kabila). It is this unarmed opposition, particularly the democratic sector within it, that will be key to demilitarizing Congolese politics by making it more inclusive.

We can now turn to the external factor in the rebellion. There is need to oppose the external invasion without denying the existence of an internal opposition. It is worth noting that all regimes in the region—and this includes, in particular, Congo, Rwanda, Uganda, Sudan, and Angola—have a habit of insisting that their internal problems are generated by external involvement. Of all, however, Congo falls in a special category. For, unlike all others, Congo is the object of direct foreign invasion, not just indirect foreign interference. By this, I am referring to the invasion that began with the entry of Ugandan and Rwandese forces, which was then countered with forces from Zimbabwe, Angola, Namibia, and others. The foreign invasion seems to have given Kabila a second lease on life. If you want an analogy, think of how Hitler’s invasion transformed Stalin into a national hero. Increasingly isolated at home by late 1997, it is clear that Kabila now enjoys growing popularity as he wears the mantle of national independence. [End Page 60]

The result is that internal reform is more difficult today than it was before the foreign invasion. This is why the first precondition to internal reform is that all foreign forces leave Congo. And yet, we know that foreign forces are unlikely to leave Congo without the acceptance of a broad agenda of internal political reform. Such an agenda will need both international credibility and international support. To marshal both that credibility and that support, any agenda for internal reform will need to recognize the legitimacy of all internal political forces, whether in or out of government, whether armed or not.


Foreign invasion cannot give us democracy as a turnkey project. This was true of Uganda in 1979. It was true of Congo in 1997, and it remains true of Congo in 1999. Many problems ascribed to Kabila would have been faced by any government put in power by foreign forces. It is better to face up to this fact, no matter how things turn out in the present conflict in Congo. Even if our friend Wamba-dia-Wamba, or another person of equal integrity and democratic persuasion, should turn out to be the head of the next government, this single political fact will not go away.

The irony of the Congo crisis is that the government claims to stand for the national question, while the rebellion highlights the democratic question. Our dilemma, and that of Wamba-dia-Wamba, is that any political force that hopes to realize its democratic aspirations will first have to establish its nationalist credentials.

The lesson of Congo is that Africa needs to reassert and strengthen two principles. The first is the defense of territorial integrity and sovereignty in the face of militarism and the associated tendency to export revolution. This means making a clear distinction between the right of peoples to negotiate and to redefine sovereignty, and the obligation of states to respect existing definitions of sovereignty. The second is to oppose militarism in politics, as a first step to democratization. It is my view that this twofold commitment can provide us with the basis for dealing with the deeper issues—citizenship and state reform—that the Congo crisis has brought to the surface.

Mahmood Mamdani

Mahmood Mamdani is currently the A. C. Jordan Professor of African Studies and director of the Centre for African Studies at the University of Cape Town. His most recent book is Citizen and Subject: Contemporary Africa and the Legacy of Late Colonialism (Princeton University Press, 1996).


* This essay is a revised version of a talk delivered at the workshop on the Congo, SAPES Trust, Harare, 23 September 1998.

1. Bantustans were Native (Bantu) Authorities under apartheid.

2. Mayi is the Kivu pronunciation of Maji, the Kiswahili word for water. Mayi Mayi refers to the claim that ritually treated water can render its user immune from externally induced bodily harm.