Across different African cosmologies, we have asked the same question: when something goes wrong, where do we look for the cause? Each group has provided a different answer, giving us profound insight into the foundations of their culture and way of life.
Here, we explore the answer that guides the Twa, Hutu and Tutsi of Rwanda and Burundi. These three groups happen to be deeply intertwined, sharing religious traditions, languages and political institutions that bind them in a similar cultural world. In Rwanda, they share the Kinyarwanda language whilst in Burundi, their counterparts share the closely related language of Kirundi. Their traditional religious worlds include belief in Imana and practices concerning the dead. They pray to the same God, fear the same dead and tell a single founding story but with different meanings.
So this entry inverts the usual method. We will find the shared worldview first, the god and the cosmology that Twa, Hutu, and Tutsi hold in common. Then we will turn to the one story that is specifically about the three of them, the charter myth of Gihanga and the pots of milk. Then, we will lay out some of its rival readings, one from each rung of the ladder it describes, and a fourth read into it from Europe. Finally we will tie the whole thing outward, because this milk-pot myth is not a Rwandan peculiarity. It is one node in a family of origin stories that runs across the Great Lakes, through the old world of Bunyoro-Kitara and Ankole. The similarities between these traditions, and the way colonial racial theories subsequently interpreted them, help explain how indigenous accounts of kingship and social hierarchy were recast as stories about supposedly distinct races.
One caution must come before anything else, because it is the argument of the piece in miniature. To write about 'the Twa, the Hutu, and the Tutsi' as three ancient and separate peoples is already to take a side in a quarrel, and it is the side that colonial officials and, later, political actors during the twentieth century wanted everyone to take. The categories were real, but they did not represent three separate races. They developed within an interconnected society in which occupation, wealth, political power, kinship, and social status shaped relations between communities. These identities were historically more complex and, in some circumstances, more negotiable than the fixed racial categories later imposed by colonial administration. They were not, however, simply interchangeable social ranks. Keep that in view as the founding story unfolds, because the stories themselves became part of the struggle over how these communities understood their past.
Imana, the Supreme God
At the top of the shared cosmos sits Imana, the supreme being of Banyarwanda and Barundi. In many accounts of traditional Rwandan and Burundian religion, Imana is described as the creator and ultimate source of life and fortune, often understood as remote and benevolent. He is deemed less a meddler in daily life and more the distant source from which all fortune flows. The character of Imana is written into the very names the people carry. A whole family of Kinyarwanda names ends in -imana, each name representing a small theology: Habyarimana - it is God who gives birth; Hategekimana - God alone rules; Ndagijimana - God is the keeper of what I own; Nsengimana - I pray to God. Fortune, fertility, children, cattle, rain, all are Imana's to give.
A saying is popular among the people of Rwanda. They say Imana keeps a particular home in Rwanda. "Imana yiriwa ahandi igataha i Rwanda" — God spends the day wandering elsewhere but comes back to sleep in Rwanda. It is a proverb that seeks to adopt the dwelling place of a god more than it is theology, but it effectively captures the shape of the God. A highly mobile god, everywhere in principle, but yet turned towards this particular land at the end of the day.
In descriptions of traditional Rwandan religion, Imana is not usually approached through the dense network of direct sacrificial rites associated with some other African high gods. In fact, there are very few direct rites. The creator is often presented as the ultimate guarantor of an ordered world. And exactly as we saw with the Kikuyu and the Shona, a remote creator leaves a gap between heaven and the household, a gap that has to be addressed through relationships closer to everyday life. In Rwanda and Burundi, this gap is addressed through the dead.
Abazimu, the Spirits of the Dead that Bring Misfortune
Ask the older worldview why a child sickens, why the harvest fails, why a household slides into poverty, and they'll point to the abazimu, the spirits of the dead. Just like the Luba, the Twa, Hutu and Tutsis believe that when a person dies, he never really leaves and that his soul can be called to linger around as a protector. Every person, in this understanding, survives death as an umuzimu, keeping the name and temper they had in life and lingering near the places they knew. However, unlike the Luba, the abazimu are prone to resentment, and they can visit illness, barrenness, and ruin upon their descendants, though only upon their own kin. The living, in turn, have an obligation to keep them in good temperament, making offerings and small rites, guterekera, to keep the dead settled and the household safe.
Above the family dead stands a wider shield, the cult of kubandwa, a religion built around a company of hero-spirits, the imandwa, and led by the legendary hunter Ryangombe, known in Burundi as Kiranga (Ryangombe and Kiranga should not automatically be treated as identical in every local account). Ryangombe, the story goes, was killed by a buffalo and passed into the spirit world. In accounts recorded by ethnographers, his otherworldly home is associated with Mount Karisimbi, an extinct volcano in the Virunga range. Through initiation, ecstatic dance, and possession, ordinary people can bind themselves to him and win his protection against the afflictions the abazimu can send.
In the tradition, anyone can be a part of this cult. The dying Ryangombe had commanded that all his people be initiated into his cult, Twa, Hutu, and Tutsi alike, to the exception of the king. The one great popular religion of the region was therefore, by its own charter, blind to the social ranks around it. A Tutsi herder, a Hutu cultivator, and a Twa potter could kneel at the same rite, take the same new spirit-name, and enter the same fictive family of the initiated. Yet shared initiation did not necessarily mean that all distinctions disappeared during worship. Luc de Heusch's study of the cult notes that Hutu and Tutsi remained socially differentiated even within the religious institution.
Set the principles of the religion that bound them all together beside everything that would later happen to them and the reality becomes unbearable. The cult offered a form of religious belonging that was not confined to the ruling dynasty or to one occupational category. The deepest shared institution of these people was one that deliberately dissolved the very distinctions that would one day be used to set them killing one another. There were local variants, among them the northern cult of the spirit Nyabingi, later a banner of anti-colonial revolt. But the pattern held: one God above, one company of the dead below, and one cult that gathered the three groups into a single congregation.
The Founding Charter: Gihanga and the Pots of Milk
While cosmology unites the three groups, the founding myth separates them. The central figure in this myth goes by the name Gihanga, a name that comes from guhanga — to create or to found. Gihanga is the culture-hero and first true king, remembered as the one who brought cattle, fire, ironworking, and the sacred royal drum, and from whom, through his son Kanyarwanda, the Nyiginya dynasty of Rwandan kings claimed descent. In the older layers of the tradition he is a founder of kingship and of clans, not a divider of races.
The story that does the dividing is a short and pointed one, and its cast varies by narrator. The milk-pot story exists in several versions, and its characters should not be collapsed into a single genealogy. Several recorded versions identify a father who is sometimes Gihanga, sometimes Kigwa, the one who fell from the sky, sometimes a figure named Kazi, sometimes Imana himself. He has three sons: Gatwa, Gahutu, and Gatutsi, the eponymous ancestors of the Twa, Hutu, and Tutsi. To choose his heir, the father gives each son a pot of milk to guard through the night. In the morning the reckoning comes. Gatwa has drunk his milk. Gahutu has fallen asleep and spilled his. Only Gatutsi has kept watch and kept his pot full. And so the father ordains the shape of the world: Gatutsi shall rule and hold the cattle, Gahutu shall serve his brother and may gain cattle only through service, and Gatwa shall be cast to the margins of society. A gentler variant tells it as a distribution of gifts, in which Imana hands out livelihoods and the late-arriving Twa receive only the hunt, the Hutu receive the hoe, and the Tutsi receive the herd. Either way the message is the same. Social hierarchy is not an accident and not an injustice. It is the ancestors' own doing.
Notice what kind of story this is. The Dogon and the Kongo told us how death entered the world; the Kikuyu told us how a people received a land. This myth is not really about the cosmos or about suffering at all. It is a charter, a story whose whole purpose is to explain and justify a social order, to make a hierarchy feel as old and as settled as creation.
The Twa, Hutu and Tutsi Give Different Meanings to the Same Origin Story
In the Tutsi royal interpretation, the version preserved and elaborated in the dynastic tradition of the court, the myth offers a justification for kingship and pastoral authority. The ruling lineage and the social order associated with it are represented as the outcome of a founding judgment: the son who succeeds in the test receives authority, while the others are assigned subordinate positions. The story turns a political arrangement into a sacred one and gives that arrangement the appearance of ancient legitimacy. The surviving court traditions served dynastic interests, although it would be difficult to establish that the milk-pot story originated exclusively among the royal court or that all Tutsi communities interpreted it in the same way.
Now stand at the bottom, with the Twa. In versions that place Gatwa at the bottom of the hierarchy, he fails the test or receives the least desirable inheritance. The Twa, who are in fact the region's oldest inhabitants, forest hunter-gatherers and master potters who were there long before either farming or cattle arrived, appear in their own founding story only as the son who could not restrain his appetite and forfeited his place. The myth does not merely rank them low. It writes them out of the central contest between farmer and herder entirely, and that erasure has had a long afterlife. The Twa remain, to this day, the smallest and most marginalised community in both countries.
Between them stands the Hutu reading, and it is the one that has changed over time. It would be too simple to say that Hutu communities accepted the myth's verdict unchanged for centuries. The surviving evidence does not establish a single, uniform response across all periods and communities. But in the twentieth century, as a Hutu counter-elite emerged, the story was turned over. In the reading that powered the Bahutu Manifesto of 1957 and the 'Hutu revolution' at the end of that decade, the Tutsi were no longer the rightful heirs of the vigil but foreign conquerors who had arrived from elsewhere and subjugated the indigenous majority. The same characters, the same pots of milk, and an inverted moral: the herder was not the chosen son but the usurper, and the farmer's servitude was not deserved but imposed.
Then came the fourth reading. When Europeans arrived they brought a theory called the Hamitic hypothesis, the pseudo-scientific conviction that anything sophisticated in Black Africa must have been introduced by a lighter, quasi-Caucasian race from the north or the Horn of Africa. They looked at the Tutsi aristocracy and its cattle and its kings and decided they had found their Hamites, a superior race of natural rulers, essentially different in blood from the Hutu they governed. The Belgian administration then did the fatal thing. It took a category that had been fluid and made it fixed, issuing identity cards in the 1930s that stamped every person as Hutu, Tutsi, or Twa for life. A social rank became a racial fact, printed on a card and inherited without appeal.
The scholars who have studied these traditions closely, among them Jean-Pierre Chrétien, have examined how indigenous accounts of origin were subsequently interpreted through colonial racial theories. The founding stories themselves are primarily concerned with kingship, social position, and the legitimacy of political authority; they do not, by themselves, establish that the Tutsi were a biologically distinct people who invaded Rwanda from the north. The narrative of foreign racial origin was reinforced by the Hamitic hypothesis and colonial interpretations of the region's history.
The internal evidence also complicates the idea of immutable racial categories. In some historical contexts, wealth and cattle ownership could facilitate upward social mobility, a practice known as the kwihutura. A Hutu who accumulated enough cattle could, in effect, become Tutsi, and a Tutsi who lost his herds could slide the other way. This did not mean that Hutu and Tutsi identities were freely interchangeable whenever a person gained or lost cattle. Social recognition, political relationships, and historical circumstances also mattered. The roughly eighteen clans, the ubwoko, cut clean across all three categories: Hutu, Tutsi, and Twa. Members of different categories could share clan names, totems and associated traditions. This is evidence of a deeply interconnected social world, although shared clan affiliation does not prove that the three categories were merely occupational ranks or that the distinctions between them were insignificant.
Related Stories in the Great Lakes
The milk-pot story has parallels elsewhere in the Great Lakes, but the versions need to be kept distinct. In Ankole, a recorded tradition tells of Ruhanga, the creator, and three figures named Kakama, Kahima, and Kairu. Their test involves keeping pots of milk safe through the night. The brothers' conduct determines the roles assigned to them: Kakama is associated with kingship, Kahima with cattle, and Kairu with cultivation. The story connects these figures to the Bakama (the ruling kings), the Bahima pastoralists, and the Bairu cultivators.
The details differ between tellings. In one account, Kakama succeeds because Kahima gives him some of his milk. In another, both Kakama and Kairu spill their milk, but Kahima shares his with Kakama. The important thing is that the story is not simply about one brother passing a test while the others fail. In the latter version, Kakama's position is also linked to the help he receives from Kahima.
Bunyoro has its own origin traditions, which should not be treated as interchangeable with the Ankole account. In one account of Bunyoro's creation tradition, Ruhanga arrives with his brother Nkya Mba, also called Kantu. After Ruhanga departs, Nkya Mba remains, and his sons are Kairu, Kahuma, and Kakama. Ruhanga sets the brothers tests, and Kakama becomes the Omukama, or king; Kahuma becomes his herdsman, while Kairu is associated with evil after his disinheritance. This version gives a mythical charter for kingship and social inequality, but its characters and outcome differ from the Ankole account.
These traditions share a broad narrative structure: a founding authority tests related figures and uses the results to explain political rule and social roles. Their similarities are significant, but they do not prove that the stories have one common origin, that one was borrowed directly from the other, or that Ankole and Bunyoro had identical systems of social organization.
In Ankole, the distinction between Bahima pastoralists and Bairu cultivators had social and political consequences. Nevertheless, these categories and their relationship to wealth, status, and mobility varied across historical contexts. The story should be read as a traditional explanation of social order, not as a complete description of how every person lived or how status operated in every period.
The Societal Impact of the Myth
It is necessary to say clearly, and rather soberly, where these myths and the meanings each group gave them have led. Burundi in 1972 and Rwanda in 1994 saw mass killing on an enormous scale. The number killed in Burundi in 1972 remains disputed, with estimates varying substantially. In Rwanda, the 1994 genocide against the Tutsi unfolded over roughly 100 days and killed an estimated 800,000 to one million people. One would wonder what led to these killings when the myth, as the ancestors told it, did not decree strife. Rather, it described ranks within one people who shared a god, a language, a company of the dead, and a cult that seated them side by side.
The killings were orchestrated when racialisation was somehow birthed from the story — the century-long work of turning a fluid social order into a fixed one; of teaching people that the son who kept his milk and the son who spilled his belonged to different species of human. The genocide was not the myth coming true. It was the myth being falsified, its own logic of shared descent overwritten by an imported logic of race. However, the genocide against the Tutsi in 1994 emerged from a much wider modern political history that included colonial legacies, postcolonial exclusion, extremist propaganda, the civil war, state and military organisation, and deliberate decisions by political and military actors.
So far, this is the hardest lesson our series on African cosmologies has encountered. An origin story is never neutral furniture. It is read from wherever you stand in the order it describes, and it can be re-authored by whoever holds the power to do so, including outsiders who do not believe it at all. The shared cosmology reveals important cultural connections between Twa, Hutu, and Tutsi, while the charter stories reveal ways in which social hierarchy could be explained and justified. Their later racialisation shows how older social distinctions could be given new meanings under colonial rule and subsequent political struggles. The history therefore is not simply one of a single people being taught to become three races. It is a history of interconnected communities whose social distinctions were reinterpreted, hardened, politicised, and, at particular moments, turned into grounds for exclusion and mass violence.
A Note on Sources
Primary and dynastic tradition. The classical royal myths survive chiefly through the Rwandan court traditions, the ubwiru ritual code and the ubucurabwenge dynastic genealogies, recorded and interpreted above all by Abbé Alexis Kagame; readers should note these were codified late, under the Nyiginya court, and carry that court's interests. Jan Vansina's Antecedents to Modern Rwanda: The Nyiginya Kingdom reconstructs the political history behind them.
Secondary and scholarly. Jean-Pierre Chrétien's The Great Lakes of Africa: Two Thousand Years of History is the essential regional synthesis and the source for the argument that the origin myths concerned clan and royal authority rather than racial provenance. Mahmood Mamdani's When Victims Become Killers analyzes the colonial manufacture of racialized identity. Nigel Eltringham's work traces the Hamitic hypothesis and its role on the road to 1994, including the parallel racial rereading of the Bachwezi legend. On the kubandwa cult, see Luc de Heusch and the ethnographic literature on Ryangombe and Kiranga. Figures for the 1972 Burundi killings and the 1994 genocide against the Tutsi follow the standard historical estimates.
