1994 was the year the African intelligentsia lost its innocence.footnote1 Literally in the same week, South Africa began a historic journey to reconciliation and the genocide in Rwanda became an epic tragedy. If we had been told a few years earlier by some clairvoyant that there would soon be a genocide in one country and a reconciliation in the other, I wonder how many of us would have guessed their identities right.

I had been to South Africa, but not to Rwanda. Being a Ugandan, I had closely followed the march of the rpf to Kigali, and had wondered about the role my country played in the unfolding tragedy, and about the responsibility of Ugandan intellectuals. Eventually, I did go to Kigali, and then to Gitarama, to the university in Butare, to Ntarama, and finally to Nyamata. It was a short visit, during July of 1995, and would be followed by several others. During it, I slept little and thought much. I call what follows my reflections after the visit; I am aware that the time I spent in Rwanda was short, the cross-section of people I met narrow, and my preparation for the visit inadequate. Even so, I have chosen to err on the side of boldness. If the distance and detachment of the outsider can often be a source of ignorance, I can only hope that these may sometimes also prove to be a source of insight.

I began with my own research into the colonial state in equatorial Africa, a context from which to underline the distinctive aspect of Rwandan history.footnote2 I had understood the colonial state as a bifurcated apparatus under which crystallized two distinct forms of authority: the first a civil power speaking the language of rights, the second a ‘customary’ power speaking the language of tradition. While civil power claimed to protect rights by enforcing civil law in civil society, the claim of customary power was to enforce custom. The bifurcated state was described by a duality in legal theory which separated the civil law governing relations of non-‘natives’, both among themselves and with ‘natives’, while ‘customary’ law kept ‘natives’ in their place. This bifurcated power reproduced a dual identity. Civil power was racialized; both the non-‘native’ beneficiaries of civil rights and the urbanized ‘natives’ denied rights were imprinted with a racial identity. Customary power, by contrast, was ethnicized or tribalized. The Native Authority defined and enforced an official version of custom as customary law. The stranger under the Native Authority was not a racial but an ethnic stranger: customary power reproduced an ethnic identify.

In the African colonial experience, custom was broadly defined to include not only personal law but land. To the extent that land and labour were defined as customary, they remained beyond the scope of the market: force was the only way to drive them out of the customary realm. Briefly, the regime of custom tapped into authoritarian possibilities in culture. Custom became the language of domination, masking a regime of force. When force was outlawed in British colonies after the First World War, and in French colonies after the Second, this applied only to civil power. Native Authorities continued to exercise their right to use force, so long as that force was customary—and force used by Native Authorities was, of course, considered customary. The tautology was complete. Colonial rule consolidated, not the force of tradition, but the tradition of force.

Rwanda, however, was distinctive. While most colonies were organized as polyglot formations—with a central civil authority and a constellation of district-based customary authorities, each with a distinct ethnic identity—in Rwanda district authorities did not correspond to ethnicpowers. Instead, the Batutsi were like a layer of cream spread over the entire society, administering a subject peasantry, the Bahutu. To under stand this difference, we need to grasp the distinctive nature of the process of state formation in Rwanda.

My grasp of post-genocide Rwanda has been shaped by a consideration which has three main aspects: first, Bahutu and Batutsi (the prefixes are ‘mu’ for singular, ‘ba’ for plural) are bipolar identities reproduced by a form of the state that institutionalized them as such: there cannot be one without the other. In this relationship, first shaped by the pre-colonial Rwandan state but fully crystallized by the colonial state, Batutsi came to be identified with power and Bahutu with subjecthood. Second, despite its limitations, the ‘social’ revolution of 1959, which turned this relation upside-down, was a subaltern revolt, akin to Mau Mau in Kenya and the Zanzibar Revolution of 1963. So ‘Hutu Power’, the ideological springboard of that revolt, was a subaltern but truncated consciousness, shaped and limited by the institutional context of the indirect-rule state. Without understanding this, it is difficult to grasp the link between proponents of ‘Hutu Power’ and the subaltern Bahutu masses in the genocide—for while this power was extreme, it was not marginal. The foot-soldiers it mobilized understood the genocide as a defence of the social gains of 1959. They referred to the killing as ‘work’, and indeed communal work, a ‘customary’ obligation, organized and directed by ‘customary’ authorities in the state. Finally, while I have no ‘solution’ to the problem of Rwanda, I do claim that the search for one must incorporate three insights: that failure to ensure that those responsible for the genocide are brought to justice will only multiply instances of revenge; that a broad-based regime in a tense post-civil-war situation must include all political tendencies willing to renounce violence—even ‘Hutu Power’; that the construction of a broad base from the level of the elite can only be a stop-gap measure unless it is reinforced by institutional reform of the conquest state from below.

Who are the Bahutu and the Batutsi? This is a question that came into every discussion I had. While the Bahutu opposition had been a target of the massacre, this was a political matter; the Batutsi faced true genocide—the attempt to eliminate them as a people. Given this single fact, that illuminates the tragedy of Rwanda for the world at large, I was non-plussed to be told over and again by leading people in the rpf: ‘We speak the same language, have the same culture, and live on the same hills; we are the same people.’ But in casual conversation and out on the street, some of the same individuals would readily identify Muhutu and Mututsi. Sometimes by physical appearance, but in a place like Ntarama in the southern lowlands, where Batutsi made up as many as a third of the population, and where there had been many intermarriages—a third of the Batutsi daughters were married to Bahutu, I was told—this was hardly a reliable method. During the killings, people were asked to produce their identity cards.