Cattle, identity and genocide in the African Great Lakes region

Cattle, identity and genocide in the African Great Lakes region Andrew Reid Archaeological research into the origins of the cattle-keeping people of southwestern Uganda has overturned long-held beliefs about the separate origins of the pastoral and agricultural pop­ ulations of the region. Study of indigenous political development and nineteenth-century colonialism shows that the present-day ethnic identities, which fuelled the genocide in Rwanda in 1994, are recent constructions.


I
n 1994 the world was rocked by the genocide in Rwanda.In the space of a few months between 500 000 and 1 000 000 people lost their lives, in the main because of their perceived Tusi (Tutsi) ethnic identity.The outside world tended to look on this carnage not as an individual event but as another exam ple of ethnic conflict endemic to the Afri can continent.Such ethnic rivalries, as that between Tusi and Hutu in Rwanda, were seen as timeless, age-old and inevi table.Participants in the conflict itself rationalized the violence through the belief that the two communities had dis tinct and conflicting origins, the Tusi having come from northeastern Africa, invaded Rwanda and Burundi and sub jugated the indigenous Hutu population.However, archaeological research in the Great Lakes region helps to demonstrate that this perception is untenable and ig nores both long-term processes of political development and recent colonial history.

The origins of cattle keeping
The differences between Tusi and Hutu populations in Rwanda and Burundi are very similar to those between Hima and Iru populations in southwestern Uganda and northwestern Tanzania.One of the main differences that distinguish Tusi and Fig ure 1 Long-horned Ankole cattle near Ntusi, southwestern Uganda.
Hima from other groups is their associa tion with large cattle herds.Character istically, their cattle are of the striking longhorn Ankole breed (Fig. 1).Recent work has begun to reveal something of the origins ofthis cattle keeping.Prior to 1000 AD there is almost no archaeological evidence for cattle in the Great Lakes region.However, the study of comparative linguistics suggests that there would have been significant interactions between stone-tool-using, cattle-and grain-produc ing pastoral populations, originating from the north, and iron-using forest farmers of the west, from as early as 500-300 BC.1 By 500 AD these communities appear to have blended together to form a single hetero geneous society, speaking the Bantu lan guages of the forest farming populations.
Linguistic evidence demonstrates that, from about 800 AD, there was an explosion of cattle-related words, especially relating to elements such as skin colour, skin pat terns and horn shapes.This reflects a shift in socio-economic patterns, in some parts of the region, towards greater emphasis on pastoralism.As anyone fa miliar with pastoral societies will know, pastoralists spend huge amounts of time discussing their stock, a social environment in which new words are invented rapidly.Signifi cantly, all of these new terms definitely have Ban tu language roots, demonstrating that they were not introduced by later unidentified pastoral populations from the north or east.
Our first archaeological evidence of these transformations comes from the site of Ntusi in southwestern Uganda, which was occupied from the eleventh to the fif teenth centuries (Fig. 2) .zThe assemblages of animal bones recovered from Ntusi are dominated by cattle, which constitute as much as 90 per cent of all the animals con sumed at the site.Very large proportions of these animals, presumably mostly bulls, were being slaughtered while they were still immature, a pattern of herd manage ment consistent with large stock numbers.Archaeological survey revealed a series of smaller sites, most of which are cattle enclosures, suggesting that Ntusi was a small chiefdom that sustained political domination over outlying cattle-keeping settlements up to 15 km from the centre.Besides the heavy emphasis on cattle, it is also evident that the cattle-keeping ethos was escalating.A simple division of bone size data into earlier and later material shows that the body size of animals was increasing (Fig. 3), presumably as, with increasing expertise, herdsmen selected for larger animals in their choice of breed ing bulls.
However, the picture at Ntusi is not solely one of cattle.Grinding stones for the preparation of grain, small curved iron grain-harvesting knives and several grain storage pits were encountered in surface collections and during excavations.Two chance finds of charred grain indicate that the crops grown included sorghum and finger millet.Significantly for our under standing of the development of pastoral society, these agricultural elements were recovered in households that have clear pastoral orientation.At one cattle-keeping location, a study of the abrasion damage on the inside of pots suggests that over 30 per cent of vessels were used for preparing staple grains.3

Pastoralism in the later kingdoms
This emphasis on the coexistence of pastoral and agricultural activities within the same household becomes significant when we move on to look at the place of pastoralism in the states that developed after the fifteenth century.An exclusive class of pastoralists emerged in several kingdoms, including Rwanda, Nkore and Karagwe (Fig. 4).These pastoralists were distinct from other East African pastoral ists, such as the Maasai, not only because _j DOI: http://dx.doi.org/10.5334/ai.0412------------------6; ;';0:;---------------------:7;'; ;0----------------------;  of their association with states but also for their almost complete concentration on cattle.Goats were totally shunned and only one or two sheep would be kept with the herd.The pastoralists separated them selves from all aspects of agriculture and preferred to consume only milk, meat and occasionally blood.In the event that agri cultural produce (other than beer) was eaten, a 24-hour fast was observed before resuming milk consumption.In this ex clusively pastoral world, the position of women, deprived of the opportunity to cultivate, was greatly undermined.As with the breeding of the herds, human re production was a carefully observed busi ness, with very distinct perceptions of the social ties and human features that were regarded as desirable.Through the twin processes of selection from a restricted gene pool and the strict dietary regime, pastoralists themselves became physically distinct from agricultural populations.In their activities they also distinguished themselves by dress codes, domestic architecture, pastimes, other aspects of material culture and, of course, economic practice.simply move their herds to another state.The position of the king was therefore dependent on his ability to mediate the demands of his cattle owners.As a con sequence, royal rituals featured heavily pastoral imagery, and the functions of the state and of the pastoral elite became increasingly intertwined.In Rwanda and Nkore the king was responsible for the organization of schools at the royal court where children of the pastoral elite were instructed in the ways of the state and subsequently formed regiments for the governance and protection of the kingdom.
Perhaps the most significant single image that associated the state with cattle are the unique iron cows from the royal regalia ofKaragwe (Fig. 5).They are part of a range of metal artefacts said to have been made in the early nineteenth century.4Their production and use indicate that the kings were not just super-pastoralists.The Karagwe regalia is general!y agreed to have been made by Ndagara, the king who ruled in the early nineteenth century.In Karagwe The presence of increasingly large herds of cattle, the opportunities this provided for expanding political influence through patron-client and other systems of cattle exchange, and the need for arbitration on the often limited supplies of pasture and water, created conditions where central ized authority became increasingly sig nificant.The kings of the states with large cattle herds were held responsible for the wellbeing of the herds that resided under their authority.If water resources or pasture became depleted, if the animals became sick or infertile, or if security against raiding by other states could not be guaranteed, the owners of the cattle would  and in Burundi the king was regarded as a smith, a decidedly non-pastoral activity.Other elements in the Karagwe regalia include several smiths' anvils (often referred to as "hammer/anvils") and one exceptional item, an anvil with four crudely embossed horns, the design of which combines iron working and pasto ralism in one object (Fig. 6).These regalia were all regularly incorporated into state ritual.Every month the king was required to perform the New Moon ceremonies sur rounded by all his regalia, which ensured the successful re-emergence of the Moon and hence the fecundity of the kingdom's cattle and the fertility of its crops.There fore , the king mediated between the con flicting interests of cattle and agriculture, and agriculture and its related activities were as important as pastoralism to the functioning of the state.Indeed, although in many areas pastoralists did dominate and subjugate agricultural populations, there were also large areas where inde pendent agricultural populations existed, with their own chiefs and with their own small numbers of cattle.
These situations existed in the nine teenth century in what are now Rwanda, Burundi, southwestern Uganda and north western Tanzania.In the mid-nineteenth century there must have been tensions be tween populations relating to the unequal division of wealth, but there are no indi cations of the kind of slaughter that has plagued Rwanda and Burundi over the past 40 years.To explain these conflicts we need instead to consider events after 1850 and the colonial encounter.

The construction of colonial identities
All the archaeological and linguistic evi dence indicates a long-term development of cattle breeding and pastoral identities within the Great Lakes region.This is very different from the opinion of the first Europeans who encountered these socie ties.For instance, the explorer John Ban ning Speke, passing through Karagwe in 1862 on his way to discovering the source ofthe Nile, noted that nobody knew where the Hima came fr om, everybody just assumed it was not the Great Lakes because their appearance and behaviour was unusual.Speke decided that they were Ethiopian of "semi-Shem Hamitic" origin.5This is important because at the outset of European contact, Tu si and Hima could not identify their homeland, but, given their exclusive form of elite pasto ralism and the ideologies needed to sus tain it, it is hardly surprising that they professed to have exotic origins.
It is also important to note the broader context in which European writers were recording their ideas.In the late nine teenth century Europeans were heavily influenced by concepts of race.The states of the Great Lakes challenged derogatory racial beliefs about African intellect and ability.Consequently, theories had to be generated which convincingly explained away the presence of states and social sophistication at the heart of the African continent.In addition, European writers were increasingly involved in creating physical structures and ideologies to effect imperialist goals.The key to British policy was indirect rule, using the existing king doms to conduct local administration while the authorities of the British Protec torate oversaw general issues.Existing elites not only had to be materially sup ported, but their position also had to be ideologically bolstered.Thus, there was considerable use of the idea of invading Hamites, descendants of the biblical Ham, associating this mythical race with Hima and Tusi.It was through Hamites that "the totally savage Negro received ... all the civilization he possessed before the com ing of the white man".6 With Europeans such as Sir Harry John ston suggesting that "superior races com ing from the more arid countries of southern Abyssinia and Galaland have continually leavened the mass of ugly Negroes pullulating in the richly endowed countries between and around the Nile lakes"/ elites naturally chose to identify themselves as Hamites, depicting them se! ves as a superior race and distancing themselves from the ordinary population.

The creation of Rwanda and Burundi
This situation of renegotiated power for the elites continued in Uganda throughout the Protectorate.The situation in Rwanda, Burundi and Tanzania was rather different as a result of the First World War.At the end of this conflict, German East Africa was divided between Britain and Belgium.It is this division that was to have dire consequences because it created two small nation states that roughly corresponded to pre-colonial political boundaries, in which pre-existing tensions could be exacer bated.It is worth examining the creation of such small territories, because it reveals the arrogance and callousness of European nations and suggests that the recent loss of life in Rwanda and Burundi could have been avoided.
All parties involved in the Treaty of Versailles in 1919, Belgium included, agreed that Britain should have all of Ger man East Africa.However, as a result of Belgium's intervention in the East African campaign during the First World War, it controlled about a quarter of German East Africa and used this as a bargaining point in trying to acquire an appropriate repara tion for its contribution to the campaign.The compromise, readily recognized by all parties, was for Belgium to hand over con trol of the territory it had taken to Britain, for Portugal to give Belgium control of the mouth of the Congo, and for Britain to compensate Portugal with southern por tions of German East Africa.Unfortu nately, this did not happen because of British and Belgian belief in the intrac tability of the Portuguese.In the words of the British representative, not adopting this compromise was "to free myself fr om the trouble of a laborious negotiation with Portugal"a The other crucial fa ctor was the determination of the British to restrict the territory awarded to Belgium to Rwanda and Burundi itself.If the Belgians had been allowed a larger territory, with access to Lake Victoria (Fig. 4), it is likely that it would have been sufficiently heter ogeneous to prevent the two communities from competing directly with each other for power.It would also have been more financially viable, particularly because it would have provided access to the East African ports via the lake and the Uganda Railway.Instead the Kagera River (Fig. 4) became the eastern boundary of the Bel gian territory, because it was along its valley that Britain intended to route that grandest of colonial fo llies, the Cape to Cairo railway.The idea of the railway had already been scrapped by 1923, only four years after the Treaty of Versailles, but by then the borders were considered to be permanent.Thus, Belgium inherited a colony, Ruanda-Urundi, which it neither wanted nor could develop, given its com mitment to the Belgian Congo.

Colonialism and identity
In Ruanda-Urundi the Belgians sought to harness existing political structures by introducing a policy of indirect rule.In so doing, royal authority was extended to areas that had previously been autono mous.Following what the Belgians per ceived to be the enshrined order of society, only Tusi went to secondary schools, or attended seminaries, or joined the colonial administration.Much has been made of the difference in physical appearance between Tusi and Hutu, but, fa ced by their practical inability to distinguish between the two, the Belgian authorities conducted a census and issued identity cards that defined Tusi as those who possessed 10 or more cattle and Hutu those who had fewer.The resultant identity documents have been used ever since to define ethnicity, ethnic identity being inherited from the original census participants, and indeed these identities were used at roadblocks in 1994 to determine who was spared and who was murdered.9The creation of small states meant that antagonism between Tusi and Hutu, exacerbated by colonial favour towards the elite minority, was transformed into direct and violent com petition for power in Rwanda and Burundi.This did not happen in either Uganda or Tanzania, because Hima-Iru tensions were marginalized by their pres ence in a much larger nation, their rivalry subsumed by the need to compete with other regions and populations.
It is ironic that it was recent events around Ntusi that triggered the present cri sis in Central Africa: genocide in Rwanda, a military crackdown in Burundi, and Af rican nations openly fighting each other on Congolese territory.10In the late 1980s in Uganda, the new regime encouraged squatting on land that was under-used.Pastoralists in the Ntusi area, both Hima and Tusi, the latter mainly refugees who fled Rwanda in 1961, quickly moved onto under-utilized ranch land.When land was finally re-allocated, the government chose to ignore Tusi claims.Rather than recog nize individuals as citizens with equal land rights throughout the nation, the gov ernment chose to fall back on colonial land policies of indirect rule, which associated land with particular ethnicities.Faced with this clear rejection, Tusi quickly organized themselves and began the inva sion of Rwanda.The ensuing conflicts in Rwanda, Burundi and the Democratic Re public of Congo are a direct consequence.

The future?
Archaeological research at the settlement of Ntusi and in the territory of Karagwe clearly demonstrates how socio-political formations were constructed over the course of the past thousand years.Identi ties and ethnicities were further redefined, and were imbued with greater signifi cance, by colonial encounters.The kill ings in Rwanda and Burundi were not the inevitable product of incompatible ethnic ity, but the result of identity constructed in the colonial period and the consequence of the inappropriate subdivision of the African continent by European nations.Evidently the genocide could have been avoided.More importantly, the archaeo logical evidence suggests that new and less confrontational interpretations of the past can in the future be used to generate more appropriate ideologies and constructions

Figure 2
Figure 2 Th e settlement of Ntusi, south western Ug anda, showing the distribution of archaeological fe atures.

; 8 0Figure 3
Figure 3 Increase in size of Ntusi cattle based on measurements of fi rst phalanges (toe bones).

Figure 4
Figure 4 Pre-colonial kingdoms associ ated with cattle in the Great Lakes region.

Figure 5
Figure 5 Iron cow (approximately 20cm high) fr om the royal regalia of the kingdom of Karagwe.

Figure 6
Figure 6 Ceremonial iron anvil with embossed horns fr om the Karagwe regalia (scale bar intervals 10cm).