The Mamdani Lecture for Odia at 60

Congo and Sudan: Lessons for Nigeria



Mahmood Mamdani

Columbia University


[Text of talk delivered to the Nigeria Institute for International Affairs on Tuesday, March 16, 2010.]


How should we think of political violence in Africa? This is the question that I want to address today.

When it comes to making sense of political violence in Africa, most of us take the cue from how the international community reports it. By the international community, I mean those who speak in the name of the international community: that is, the corporate media, the international NGOs and UN agencies.

For that international community, Congo is today the paradigm case of senseless violence. The proof is said to be two-fold: first, the sheer numbers of the dead; second, the evident lack of reason for the violence.

Let us begin with the astronomical numbers of those said to have died in the violence. Beginning in 2001, the prestigious New York-based International Rescue Committee carried out multiple surveys of war-related deaths in Congo since the start of the conflict in 1998. According to the IRC, there were 1.7 million war-related dead in Congo from 1998 to 2001. These estimates climbed to a staggering 5.4 million by January, 2008. If correct, these figures would represent about 8 percent of DRC’s current population. Thus the Congo war was termed “Africa’s First World War,” the title of Gerard Prunier’s recent book on the Congo wars.

But how correct is this comparison? About 8.5 million troops were killed in the actual First World War. In contrast, the estimate of those who died of direct violence in the Congo wars is less than a hundred thousand. The balance, over five million, were said to have died of “war-related” causes, such as infectious diseases, malnutrition, disruption in vital supplies, and so on.
Your count of “war-related deaths” depends on your estimate of those who died of the same causes – such as infectious diseases, malnutrition – before the war. In October 2008, two Belgian demographers, André Lambert and Louis Lohlé-Tart, were invited by the European Commission to assess the 2005-06 voter registration process in the DRC. They drew on the data they had gathered, and whatever else they could muster, and wrote a devastating critique of the IRC estimates: they concluded that the excess death toll between 1998 and 2004 was roughly 200,000—which is one-twentieth of the IRC’s 3.9 million excess death estimate for the same period.

Their findings triggered further reviews. WHO commissioned a peer review, which concluded that the IRC “estimates” had been based on “extrapolations” that were “speculative at best”. A WHO-affiliated unit, called Health and Nutrition Tracking Service (HNTS), said the IRC’s estimates of mortality rates before the war had been too low. Another reputable source, the Human Security Report, used a more realistic baseline rate for the period May 2001 to April 2007 and found the results led to a very different conclusion: “The best estimate of the excess death toll shrinks to less than one-third of the IRC’s original figure––from 2.83 million to 0.86 million.”

At this point, then, we have several wildly divergent estimates of war-related dead in Congo: they range from the IRC’s sensational 5.4 million to the Belgian demographers subdued 200,000. Where does the truth lie?

Let me turn to Darfur, where there were equally divergent estimates of war-related deaths. In this context, an audit agency of the U S Government, called the Government Accountability Office (GAO), got together with another reputable U. S. institution, the U. S. Academy of Sciences, and put together a panel of experts. The GAO had compiled six different studies, which ranged from a high estimate of 400,000 dead, coming from Save Darfur-linked researchers, to a low of 50,000 to 70,000 dead, by the World Health Organization. The panel of experts agreed unanimously that Save Darfur estimates were the least reliable, because Save Darfur researchers had extrapolated the figures from a survey of refugees in camps in Chad and had generalized them to the whole of Darfur. The experts said that a more reliable estimate may be closer to 118,142, as calculated by CRED, a WHO-connected research unit in Europe, and not 400,000, a quarter of the Save Darfur figure.

As in Congo, there was a second distortion. WHO reports suggested that many of the deaths in Darfur may not be war-related as much as drought-related, for one reason. WHO estimated that 70-80% had died from the effects of drought and desertification; that these were mainly infants and children who had died mainly from diarrhea and dysentery. More importantly, the drought had preceded the war.

The GAO report was sent to the State Department which agreed with its findings. It then made its way to Congress and then to the GAO’s internet site. Just as the reassessment of the dead in Congo has had little effect on media reports, so did the GAO’s reassessment have little effect, either on media reports or on Save Darfur’s continuing claims, relayed in full page ads in New York Times as well as in subway and bus posters, on the numbers of the dead in Darfur.

We know that data collection is among the first of casualties in any social crisis. When international NGOs confidently continue to put forward figures even after their reliability has been expertly questioned, we have a right to question their motivation. I suggest that this practice reveals more about those making the projections than about the places they write about. Is the deliberate exaggeration of figures part of an NGO fund-raising strategy, or is the motivation more sinister, meaning more political? I frankly am not in a position to tell.

Please do not misunderstand me. My point is not that the numbers who have died from African conflicts are miniscule and not worthy of concern. The number who have died are too many by any yardstick, so much so that there should be no reason to exaggerate them.

The International Community does not just tell us how many have died in African conflicts. The also interpret the violence behind these deaths.

Think of the reports on the violence in Congo. Even as responsible an organization as Human Rights Watch ascribes the deaths to “anarchy”. Anarchy evokes meaningless, gratuitous, repetitious violence. In Africa’s World Wars, Gerard Prunier describes Congo’s “rebels” as “first and foremost armed movements without ideology, without any large civilian constituency, and without any sort of unified cause … more akin to vampires than to soldiers.”

Read a human rights organisation report on violence in Congo or in Darfur, you will find that the structure of the report is basically the same. The report begins with a page or two on history. The bulk of the report is given to documenting atrocities. The real point of the report is to identify perpetrators. The technology of human rights activism is summed up in one phrase: “name and shame”. The report concludes with a set of recommendations. These inevitably call for the perpetrators to be punished.

Here is the problem: contemporary international human rights reports show hardly any interest in the issues that drive the violence. In situations where the violence is not a stand alone event but part of an ongoing cycle, where there is a history of violence in which victims and perpetrators have tended to change sides, it is more important to identify and address the issues that drive the violence than to demonize the latest group of perpetrators. If we are interested in bringing the violence to a stop, we should be interested not just in crime and punishment but, more so, in reform.

To make my point, I would like to focus on one event that seems to recur in African conflicts: ethnic cleansing. The political history of post-independence Congo is marked by ethnic cleansing: in Katanga and Kasai in 1961, and then again in Katanga three decades later, in 1991; and then, in Ituri and in Kivu. The political history of Darfur is also similarly chequered: ethnic cleansing first surfaces in the 1987-89 civil war between the sedentary Fur and nomadic ‘Arab’ tribes and then in the 2003-04 counter-insurgency that overlay the rekindled civil war. A look at the Rift Valley in Kenya or northern Ivory Coast is enough to tell us that Congo and Sudan are not exceptions. Ethnic cleansing has become a central part of political violence in post-colonial Africa.

The events in Congo and Sudan suggest that ethnic cleansing is not anarchical,but methodic. Nor is it the result of sheer conspiracy from above, for the simple reason that violence on such a scale requires the coming together of initiatives from both above and below. It requires joining elite conspiracies to popular organisation. The challenge is for us to understand the popular dimension of this process. We need to understand the historical processes and the institutional practices through which these agencies were shaped.

I would like to trace this historical development step by step through the colonial period. Colonial authorities claimed that Africans have always lived in tribal homelands. At the same time, they told us that Africans have always been on the move, whether as nomads on hoofs or as farmers practicing shifting cultivation – or just running away from perennial wars and slave raids. If even a little of that is true, then there must have been at least some places where ethnic groups were mixed up. How were these places turned into tribal homelands in the colonial period?

The answer is: administrative force. Take the example of Katanga, where King Leopold, and the Société Générale de Belgique, Belgium’s largest corporate concern, partnered with British interests to form Union Miniére du Haut-Katanga (UMHK) in 1906. Their object was to exploit Katanga’s mineral resources. For this, they needed to squeeze labor from hinterland populations. This required a firm administrative grip on rural populations.

I suggest that we distinguish an ethnic group – a group that speaks the same language – from a tribe, a group defined by a common territory. I am suggesting that we view “tribalization” as a colonial administrative project. In Katanga, a series of decrees were passed, in 1906, 1910, and then 1933, requiring that each ‘tribe’ be identified, separated, and resettled, each in its own ‘homeland,’ each supervised by its own Native Authority. One District Commissioner complained: “Batshioko, Lunda and even Baluba are totally jumbled and it will be very difficult to organize them into separate chefferies.” The separation was accomplished between 1925 and 1930. “Customary” chiefs were charged with supplying designated quotas of labor and food, at first to the mines and, later, to European farms and administrators.

Like Katanga, Ituri too was the site of lucrative gold mines, Kilo and Moto, to which King Leopold’s men were lured as early as 1903. As in Katanga and Kasai, so in Ituri, colonial pacification began with a policy they called “regroupement”. Over nearly two decades, from 1916-17 to 1930s, the authorities separated the predominantly pastoral (Hema) from the predominantly agricultural (Lendu) populations, herding each into its own homeland (territoire) supervised by its own tribal authority (chefferires).

A census tagged every villager as a tribesman or woman, as a ‘native’ of a particular tribal homeland. Those living far from their “natural” leaders were targeted as “runaways” from tribal homelands. When the Lendu moved away from drought struck areas in the mid-30s, one District Commissioner wrote that force would be necessary “to maintain the regroupement which was under threat”.

Each tribal homeland was run by a Native Authority. This power was not elected; nor was it appointed from all those who lived in the administrative unit. The Native Authority was appointed only from the “tribe” said to be indigenous to the land. Non-indigenous groups were required to pay tribute to “indigenous” chiefs in the Native Authority. The Native Authority system politicized ethnic identity gave the name “tribe” to this politicized group. As a political identity, tribe became the basis of systematic discrimination between groups: only tribes officially acknowledged as “indigenous” were entitled to “customary” rights, which included the right of access to land and the right of participation in local governance. This system of discrimination was sanctified as “customary” and was enforced by law.

The colonial system rested on a dual system of institutionalized discrimination: race in urban areas, and tribe in the countryside. Whereas racial discrimination was justified as reflecting a civilizational hierarchy between colonizers and colonized, tribal discrimination was said to recognize cultural difference between natives.

What is the connection between the system of power I have just outlined, the Native Authority system, and ethnic cleansing? So long as the tribal system of power continued to discriminate between ethnic groups, all institutions came to bear a tribal imprint. Recruitment for the mines or the civil service or the army was driven by tribal identity. Not only competition but also resistance developed along tribal lines.

Let us return to Katanga. Labor migration gave rise to a triangular relationship within Katanga, with each group classified as indigenous or not. The interesting thing is that there were two ethnic groups but three different classifications in the “tribal” system. The first were the Lunda, said to be indigenous to Katanga. Then came the Luba immigrants from Kasai, who were divided into two groups. Those who had moved to Katanga before colonialism were considered ‘indigenous’ and were identified as Luba-Katanga. In contrast, those who had arrived during the colonial period as labor migrants were tagged as not indigenous and were known as the Luba-Kasai.

All three groups organized as separate political parties. Alongside, there was a fourth party, representing Belgian settlers in Katanga. In the mines of Katanga, the Belgians confronted the Luba, organized in militant unions. With the development of militant nationalism, Belgians promoted an alliance of the settler party with the indigenous Lunda, known as the alliance of “civilizers” and “authentic Katangans”. The alliance first targeted the Luba-Kasai, and then all the Luba.

It is the logic that the identity of the native was tribal, not only when it came to exercizing power but also when it came to resisting power. It is the logic that fed the dynamic to secession and, ultimately, ethnic cleansing. As the alliance between “civilizers” and “authentic natives” gelled, the colonial establishment – the church, the state and business – took to backing “nativist” tribal movements, in both Katanga and Kasai. With Belgian support, each mounted a separate drive for secession, first in Katanga (11 July 1960) and then in South Kasai (8 August 1960). Branded “aliens” in both places, the Luba became the first target of ethnic cleansing in both South Kasai and Katanga.

This is the context in which the first major political crisis in Congo’s history unfolded. The new government responded to the secession in Katanga by sending the army to suppress it. On their way to Katanga, troops of the Congolese National Army were ordered to put down the South Kasai secession. They went on a rampage, slaughtering civilians.

Nzongola Ntalaje, the Congolese political historian, has argued that Lumumba committed his “first major political blunder” when “as the number one national leader, instead of seeking to heal the rift in a bitter inter-ethnic conflict”, he chose to side with one group against another. Thus he provided political enemies the opportunity to corner him politically and eliminate him physically. Immediately, Dag Hammerksjold, the UN Secretary General, accused Lumumba of being responsible for “genocide”. That same day, 5 September, Kasa-Vubu dismissed Lumumba as Prime Minister.

The Native Authority system continues to drive the crisis of citizenship in Congo today. Driving a wedge between two politically defined groups – indigenous and not indigenous – it continues to fuel the dynamic leading to ethnic cleansing. The prophetic round of ethnic cleansing began at independence, with Katanga as he paradigm case, but it was repeated on a more dramatic scale in 1992-93, again in Katanga, then in Ituri in the conflict between Hema and Lendu, and finally in Kivu.

The political crisis in Congo is today at its most extreme in eastern Congo, in Kivu, where the Native Authority system pits ‘indigenous’ tribes against the Banyaruanda minority. The Banyaruanda are the speakers of the language Kinyarwanda, identified with the historical kingdom of Ruanda. Along with speakers of Kirundi, its sister language, they number roughly 40 million. Mainly resident in Ruanda and Burundi, Uganda, Tanzania and Eastern Congo, the Banyaruanda comprise the largest language group in the region, if we exclude the trans-ethnic speakers of Kiswahili. Since most Banyaruanda live outside Rwanda, they face the crisis of citizenship in its most acute form: wherever the mode of governance is defined by the Native Authority system, the Banyaruanda are defined as ‘non-indigenous’ outsiders. Even if born where they live, they remain without ‘customary’ rights, whether to land or to appointment in a local authority.

As with Lunda migrants into Katanga, the Banyaruanda in Kivu were also divided between those who came before Belgian colonization and those who came after. The former were considered indigenous, but not the latter.

Denied ‘customary’ access to land, the Banyaruanda took to purchasing land as private property, and refused to pay tribute to customary chiefs. The result was a contest between two different notions of rights: on one side the right of the citizen, and on the other “customary” right. The political dilemma became acute where the Banyaruada became a majority, as they did in Masisi and Rutshuru by 1958. The conflict between the Banyaruanda and the indigenous groups broke out in 1963 and turned into a wider contest: immigrants demanded ‘democracy’ and the indigenous groups called for ‘custom’ to be upheld. Known as La Guerre du Kinyarwanda, this conflict lasted two years.

So sensitive was the citizenship status of the Banyaruanda minority at independence that the Roundtable Conference in Brussels was unable to fix its juridical status. The Fundamental Law the conference passed left the citizenship status of the minority unresolved; instead, it called on the Congolese people to settle the issue at a future date. The 1964 constitution famously declared that only a person with an ancestor who was “a member of a tribe or part of a tribe established in the Congo before 18 October 1908” would qualify as a citizen of Congo. The consequence was to bar all colonial era labor migrants from citizenship.

The most important dimension of the citizenship problem in Kivu is the failure of the indigenous majority to work out a political principle that would accommodate developments on the ground and extend citizenship to migrants, thereby constantly redefining the political community. The consequence of this failure was that migrants into Kvu increasingly sought protection from an outside power, initially Mobutu and then, after Mobutu, neighboring Rwanda.

Even the internal opposition, a gathering of over 100 political parties and over 400 civic groups, which came together as the Sovereign National Conference in Kisangani in 1991-92, failed to address this question. In contrast, Mobutu tried, as in 1972, when his reform extended citizenship to those who had immigrated to Congo during the colonial period. Called upon to think of a principle other than ethnic descent as the basis of citizenship in Congo, the CNS failed.

It is this failure that ultimately led to the collapse of the CNS. The failure was tragic because the CNS began as a spectacular success. Its proceedings were televised throughout urban Congo. The effect was to inspire further initiatives. There was a mushrooming of civic organisations, thickening the texture of the internal political opposition. The Sovereign National Conference was closed abruptly on 6 December 1992 when it was ready to deal with two of the most politically sensitive dossiers: one on ill-acquired goods and the other on political assassinations. The failure of the CNS to reopen pointed beyond the regime’s strength to internal weakness of the opposition.

The opposition was faced with the dilemma of managing internal tensions within its own ranks, the very challenge that had undermined Lumumba’s position in 1961. The more it failed to manage these tensions, the more the opposition fragmented. The worst outcome was in Katanga, where events were sadly reminiscent of 1960. As in 1960, the Lunda and the Luba had organized under separate political roofs, the former in the Katanga-based party, called Union des Fédéralists et des Républicains Indépendants (UFERI), led by a relative of Tshombe, Jean Nguza Karl-i-Bond (Nguza), and the latter under UPDS, led by Etienne Tshisekedi, himself a Luba from Kasai. Initially, UFERI joined UPDS to form Union Sacré, the most important opposition bloc in CNS. But this was an inter-ethnic unity of two separate ethnically-based organizations. Intent on splitting the opposition, Mobutu first appointed Tshisekedi as Prime Minister and then replaced him with Nguza. It was Nguza who closed the CNS in 1992 on Mobutu’s orders. On its heels followed the second episode of ethnic cleansing in Katanga, on a scale much larger than in 1960. This time, over a million Kasaians expelled from Katanga.

Let me turn to Darfur and Sudan. Since I have traced this history in detail in my recent book, Saviors and Survivors: Darfur, Politics and the War on Terror, it should be sufficient to underline key developments in this lecture. The first was the creation of tribal homelands under British colonialism.

The British faced several crises during their centuries-long imperial venture. The most serious of these was in mid-19th century when two revolts, the 1857 Uprising in India and the Morant Rebellion in Jamaica, rocked the empire at its two extremes. The next great crisis was the Mahdiyya in Sudan. When the British returned to defeat the Mahdiyya and colonize Sudan, they were determined to fragment the colony as effectively as possible. Thus began the program of “tribalization”, beginning with the creation of tribal homelands. From the very outset, this was a political program. It favored British allies against those who had joined the Mahdiyya, and then it favored settled over nomadic groups, since the former were easier to control. So the colonial power created “tribal homelands” – called hakuras – for peasant groups, and smaller ones for cattle nomads who were semi-sedentary, but none for the wholly sedentary camel nomads.

This did not seem to matter much until drought and desertification hit the region. Studies by the United Nations Education Program, released a few years ago, show that the southern rim of the Sahara expanded nearly a hundred kilometers over four decades, from the mid-1940s to the mid-1980s, pushing the northern nomadic tribes south in search of better land. The result was a classic ecological conflict between nomads and peasants over the best land in an ecological disaster zone.

I was a consultant for a year for the Darfur-Darfur Dialogue and Consultation, a unit created by the African Union after the Abuja negotiations. The DDDC carried out a research on the dynamics of the conflict. Its findings showed that the conflict had spread over two axis: a north-south axis that pit nomadic against peasant tribes, and an east-west axis in the south that pit two kinds of nomadic tribes – those with homelands and those without – against one another. The media focused exclusively on the north-south axis of the conflict which it portrayed as one between ‘Arab’ and ‘African’ tribes.

There are two problems with this portrayal. The first is that the driving force of the conflict was not ethnic identity but the search for land in an ecological crisis. Whoever controlled the land would survive the crisis, whoever lost control over land would parish. But because land had been defined as “tribal homeland,” the fight for land turned into a fight between tribes. This is how most observers on the ground understood the logic that fed the growing brutality in the conflict. As in Kivu in eastern Congo, the conflict in Darfur pit two notions of rights against one another: the camel nomads claimed the right of the citizen to settle anywhere in the country, whereas peasant groups laid claim to tribal “customary” rights to land and local governance.

Second, there is no single history of “Arabs” of Sudan. In particular, the history of “Arabs” of northern Sudan and that of Arabs of Darfur is radically different. Whereas the “Arabs” of northern Sudan include immigrants from the Arab world, the “Arabs” of Darfur include immigrants from West Africa, mainly the Fulani who are known as the Fallata and identify as an “Arab” tribe in South Darfur. There is also a difference in the relationship of Arabs to the slave trade in different parts of Sudan: the slave trade in the Sultanate of Funj in the North was driven by an “Arab” elite, but in the Sultanate of Fur was driven by a Fur elite. The result was that whereas most former slaves in the north tended to identify as “Arab”, most former slavers in Darfur tend to identify as Fur.

The elite in northern Sudan today is mainly “Arab”, but the elite in Darfur is not. The Arabs of Darfur are in fact its least privileged group, with the lowest levels of income and education and the least representation in the state. If Darfur is marginalized in Sudan, the “Arabs” of Darfur are doubly marginalized.

Let me return to my main point: the Native Authority system and how it generates the dynamics that tends to lead to ethnic cleansing in times of political crisis, and the need to reform it.

When he faced the internal opposition in Congo, and its spectacular success as witnessed in the launching of the Sovereign National Conference in Kisangani in 1991, Mobutu launched a counter-offensive, disguised as a reform of the state. This is when he advanced a new federal principle for Congo. Geopolitique, as he called it, was an attempt to elevate ‘nativism’, hitherto the basis of organization of the Native Authority, into a principle for the reorganization of the central government. Having already passed a resolution that every aspirant to Congolese citizenship demonstrate an ancestral connection with Congo prior to the Berlin Conference, Mobutu now declared that new heads of Ministerial Departments represent their ‘native’ provinces. By calling for regional quotas as the basis for recruitment at the center, Geopolitique further entrenched indigeneity as a principle and institutionalized ethnic competition. Mobutu then went on to demand that “delegates [to the CNS] represent only provinces to which they could be considered autochthon”.



The CNS lost the political battle the day it succumbed to this demand. In Ituri this logic was picked up and used by the Hema against Nande competitors. When a 1995 decree declared all Kinyarwanda- speakers as foreigners, the momentum of ethnic cleansing shifted from Katanga to Kivu. On 7 October 1996, the governor of South Kivu ordered all Banyamulenge to leave the country within a week, or else they would be interned in camps and eliminated.

I have often wondered whether Nigeria’s post-civil war constitution did not emulate the substance of Mobutu’s “geopolitique”, particularly in its inclusion of the “federal character” clause, requiring that key federal institutions reflect the federal character of Nigeria. As I understand this requirement, the key federal institutions are three: the federal army, the federal civil service, and federal universities. For these institutions to reflect federal character, enrollment is driven by a state-based quota system whereby the quota for each state is in proportion to its share of the federal population. Finally, the right to compete for this quota does not belong to all those who live in a state, but only to those who can claim to be “indigenous” to the state in question, meaning that not only they but also their father be born in that state.

It is possible that this provision was adopted as a form of affirmative action for those parts of the country which had lagged behind in educational and social development during the colonial period and that its purpose was to ensure them fair representation in key federal institutions, one proportional to their weight in the population. The question I have in mind does not concern motive, but consequence. My question is: have the unintended consequences of this provision – its costs – come to outweigh its intended benefits for Nigeria?

The federal character principle has extended the colonial principal of Native Authority to key institutions in the federal state. Its unintended effect has been to turn federal citizenship into an extension of ethnically-defined membership of Native Authorities, thereby eroding it. By dividing Nigerian citizens into “indigines” and “non-indigines” – not of Nigeria but of individual states – for purposes of participation in national institutions, it has disenfranchised a growing number of Nigerian citizens, those who do not live in the states where they were born.

That Nigeria is increasingly integrated into a global economy, and has been the subject of market reforms, has intensified the contradiction between the market and the state as currently organized in Nigeria. The tendency of the market economy is to move more and more strata of the population away from the locality where they were born. This includes both rich and poor Nigerians: on the one hand, businessmen, industrialists, and professionals, and on the other, unemployed workers and landless peasants. The state system, in contrast, disenfranchises precisely those who move. The state system penalizes precisely those the economy dynamizes. The least dynamic sectors of the population respond to this situation by calling for a defense of their “customary” rights, and the most dynamic rally around the principle of a “national” citizenship. One lesson of Congo and Sudan is that it may be time to rethink the legacy of both the colonial past and the reforms you undertook to end the civil war.