Print This Page
From: Missile Specialist Gooding-Lord

Small Arms and Conflict Among East African Pastoralist; The Karamoja Security Complex


This Article, written by Dr. Mercy Fekadu Mulugeta of the Institute of Peace and Security Studies (IPSS) evaluates the role that the proliferation of small arms in Africa played in cross-border conflicts and cattle raids amount pastoralists in East Africa. The traditional argument reguarding territorial conflicts among pastoralists holds that the proliferation os small arms has caused increased violent conflicts. Dr. Mulugeta argues that small arms are simply the modern day weapon of choice, but the uptick in violent conflicts is due to the increasingly shrinking geographical space remaining for these pastoralists to use. Muse

 Small arms and conflict among East African pastoralists:

The Karamoja (in)security complex




Scantily dressed cattle herders with very few material possessions, one of which is probably a gun, is the popular picture of the East African pastoralist. But it is not easy to identify the exact role of arms in a community or in battles (Raudzens 1990), and the line between the impact of arms and the impact of armed conflict remains blurred (Ginifer and Turner 2005). Early literature on pastoralist communities suggested that small arms increased death and destruction and contributed significantly to the degeneration of cultural institutions and practices, leading to the commercialization of cattle raiding and the rise of unprecedented violence (Mirzeler and Young 2000Gray et al. 2003). On the other hand, recent literature has increasingly questioned the validity of the ‘established knowledge’ around small arms and cattle raiding (Eaton 2008Knighton 2003). Factors mostly listed as root causes of cattle raiding, one of which is the proliferation of small arms, are not enough to fully explain it (Eaton 2008). Sagawa (2010) argues that accepting the knowledge that automatic weapons have increased violence is a ‘[narrow] attempt to understand the region in a techno-deterministic way’. Recent developments in pastoralist communities could easily be interpreted as the ‘degeneration of culture into chaos’ (Knighton 2003: 431); however, as Knighton (ibid.: 432) notes, ‘raiding and insecurity are a perennial feature, and firearms have been a dimension of warfare for 120 years’.

Following this developing line of discussion, I argue that the study of small arms is instrumental in revealing existing insecurities that are not created but unmasked by the availability of small arms. The study pursues a regional perspective through a case study of the Nyàngatom people in south-western Ethiopia in a way that complements secondary sources on neighbouring pastoralists in the region. It develops the concept of a Karamoja security complex, specific to this setting, to illuminate a trend that cuts across time and national borders.

Methodology and scope

A concentration of pastoralist communities is found in the East African Great Rift Valley that extends from the Red Sea to Mozambique; some of these pastoralists are known by the collective name Ateker or Karamoja cluster, which extends from north-eastern Uganda across the (agro-)pastoralist regions of South Sudan [End Page 739] and Northern Kenya to the lowlands and foothills of south-western Ethiopia, and which includes the agro-pastoralists of the South Omo zone. In this region, the Nyàngatom, and other pastoralists in South Omo and neighbouring countries, have extensive access to small arms. The number of autonomous pastoralist groups (sixteen in South Omo) found within a small land mass (about 22,000 square kilometres in South Omo) (Turton 1991), the relatively peripheral status of the pastoralist communities because of weak ‘state control’, and the existence of ecological, religious and linguistic (Niger-Kordofanian, Afro-Asiatic and Nilo-Saharan) diversity (Schlee 2009) make the region suitable for a study of small arms and their use, and the relationships between pastoralist communities. This situation has made it possible for the Nyàngatom, who have lived under the Ethiopian state since the end of the nineteenth century, to be somewhat autonomous and able to engage in warfare using alliances beyond the state. What follows is a qualitative case study of the Nyàngatom woreda,1primarily based on fieldwork undertaken in July–August 2014 and 2015.2

To analyse my data, I make use of the concept of the regional security complex, but further develop it to propose a Karamoja security complex. This concept emerged in tandem with ‘new regionalism’ theories, which challenged earlier tendencies to view a region as a ‘permanent geographical fact’, instead acknowledging that regions ‘overlap and do not have clear boundaries’ (Appadurai 2000: 7). An un-territorialized and supple understanding of regions and units is useful for sub-state units, and even for states themselves, which, in much of Africa, are not defined as territorial entities along ‘Westphalian’ lines (Cormack 2016). The study of regionalism has recently acknowledged the multi-dimensionality, fluidity and diversity of actors that could potentially form new entities that could be analysed using ‘comparative regionalism’ (Börzel 2011): a field that is ‘both multi-sectoral and sector specific; state and non-state’ and explores ‘multiple forms of organization and designs’ characterized by ‘porous and overlapping regions’ (Söderbaum 2015: 23). Unconventional border paradigms provide a conceptual space of analysis in such understandings of regionalism and are deemed necessary to understand pastoral space and conflicts (cf. Cormack 2016). This framework allows the constellation of pastoralist communities in the Horn of Africa to be studied with ‘better contextual sensitivity and less conceptual rigidity’ (Söderbaum 2015: 23).

New regionalism theory (NRT), as outlined by Hettne and Söderbaum (2000), develops earlier studies (Buzan 1991Buzan et al. 1998) to allow for contextual flexibility of this kind. It defines five levels of region-ness: regional space, regional complex, regional society, regional community, and region-state (Hettne and Söderbaum 2000). A regional space, these authors claim, is ‘a group of people living in a geographically bound community, controlling a certain set of resources and united through a certain set of cultural values and common bonds of social [End Page 740] order forged by history’. Regional complexes are ‘understood as a “conflict formation” or a “regional (in)security complex”, in which the constituent units, as far as their own security is concerned, are dependent on each other, as well as on the overall stability of the regional system’ (Hettne and Söderbaum 2000: 464). Buzan and Wæver (2003) add to this model by proposing a way to study security complexes whose security needs have been tied to each other because of global and local factors. Such a complex is constituted by a group of separate units ‘whose primary security concerns link together sufficiently closely that their . . . securities cannot reasonably be considered apart from one another’ (Buzan and Wæver 2003: 44). This model focuses on intertwined security needs between actors that are geographically close, with the rationale that ‘most threats travel more easily over short distances than over long ones [and so] security interdependence is normally patterned into regionally based clusters: security complexes’ (ibid.: 4). The territoriality of security can be overridden when the actors are not challenged by distance or when the security issue, such as an economic threat, can travel extensive distances easily. Regional security complexes (RSCs) can also be identified as sub-complexes when they possess all the features of a complex but are ’embedded within a larger RSC’ (ibid.: 51).

The four key features of an RSC are: (1) a boundary, which differentiates it from its neighbours; (2) an anarchic structure, which means that it must be composed of two or more autonomous units; (3) polarity, which relates to the distribution of power among the units; and (4) social construction, which covers the patterns of amity and enmity among the units (ibid.: 53).

The Karamoja region

New world orders have implications. New theory has effects on the way the world is seen. The conjunction of governmental and academic perspectives may become dangerous for the objects of policy who do not own those perspectives.

How, then, might we use this model to account for conflicts–and the role of small arms–in the Karamoja region? As long as theories continue to shape policy, it is worth searching for a meaningful and useful systematic understanding grounded in the perspectives of the ‘object of policy’. The Horn is best known for its state-level conflicts that have been studied systematically. The Karamoja complex and the groups that constitute it deserve an approach in which their needs and actions are no longer evaluated as resulting from their ‘warlike’ nature (Markakis 2011: 42) or from the presence of small arms alone. In a significant advance on an approach that could be branded as ‘old regionalism’, the Horn of Africa can be recognized as a security complex comprising Ethiopia, Sudan, Djibouti, Eritrea and Somalia, countries whose security needs and ties are closely linked (Buzan and Wæver 2003Berouk 2011). The regional complex model replaces this ‘old’ regionalism with its ‘new’ counterpart, by exploring each unit’s need for security, and their relations with other political entities that are not necessarily state actors.

The Karamoja region has remained unnoticed against the backdrop of the Horn of Africa. The Conflict Early Warning and Response Mechanism (CEWARN) has been producing incident reports of conflicts among pastoralist communities in the Karamoja cluster that demonstrate the viability of the [End Page 741] discussion on (in)security.3 According to the reports, the most frequent violent incidents in the Karamoja cluster are cattle raids. Based on the characterization of regional securities in terms of ‘miniature anarchies’ by Buzan, Wæver and de Wilde (1998: 13), this pattern of interaction could indicate that the Karamoja region is a security complex, where cattle and the maintenance of polarity are the key motivations for armed conflict (Turton 1991Leff 2009Glowacki and Wrangham 2013).

Conflict in the Karamoja (in)security complex

Before presenting my primary data and analysis, it is necessary to analyse conflicts and established patterns of relationships in the region as presented in secondary sources. Nyàngatom historical oral accounts, colonial documents and academic research confirm that neither Nyàngatom conflictual relations with neighbours nor the large numbers of casualties are a post-Kalashnikov (late 1980s) development. Mburu (2003) discusses how the occupation of parts of the Ilemi Triangle in the colonial period by the British East African protectorate and its shifting policies of restricting access and attempting to control the ivory trade created tension between the Turkána, Dassanetch, Nyàngatom and their other neighbours. Gulliver (1952) discussed the same pattern of conflict in his survey of the Turkána. He recorded that the Turkána would fight with all their ‘Turkána’-speaking neighbours, while there was an alliance between the Toposa and the Nyàngatom. Tornay (1981) discusses close relations between the Nyàngatom and the Toposa, since Nyàngatom transhumance occurs in South Sudan, among their friends the Toposa. The factors that encouraged the movement of the Nyàngatom westwards for transhumance could include intermarriage, fear of conflict with the Dassanetch, and pressure on the land along the Omo River (Tornay 1981Turton 1991). Carr (1977) mentions ‘periodic hostile outbreaks’ between the Nyàngatom and the Dassanetch, including incidents in 1948 and 1972. The Nyàngatom report Dassanetch attacks on Nyàngatom kebeles such as Lebere (including a recent incident in 2014) in which cattle were raided and people killed. Knighton (2003) writes that the British noted the existence of raids and recurring violence among the Karamojong, Jie and Turkána as early as 1902, and the Swahili and the Abyssinians joined the skirmishes at about the same time. In recognition of the importance of a historically rooted analysis of the Karamojong, Knighton (2003: 434) further observes:

There is only one year in the recorded history of the Karamojong, 1925, when no homicidal attempts were made by one group on another, but there is both a time for peace by bilateral negotiations and a time for war.

Like all ‘strangers’ (emòit) among the Nyàngatom, the Italians were received with suspicion during their occupation of Ethiopia from 1936 to 1941. Later, the Italians developed amicable relations with the Nyàngatom, giving them guns, ammunition, cartridge belts and other equipment in exchange for cattle.4 The [End Page 742] dislodging of the Italians from Ethiopia involved the British, who arrived among the Nyàngatom from two directions, one through Omorate and the other from the Kibish River. Once the Italians had been removed from the area, the British settled 2 kilometres away from the Ethiopian police station, across the river. This site, now located in Kenyan territory, is where many Nyàngatom now living in Natikar were born and is also where their fathers were buried, qualifying the land as traditional territory, since the Nyàngatom ‘do not leave a land our fathers and grandfathers are buried in’.5 The Nyàngatom still claim this land as their own, and they have complained against the inaction of Ethiopian governments from that time until the present: ‘EPRDF [Ethiopian People’s Revolutionary Democratic Front] is doing the same. They told us not to cross over because they [the EPRDF] are afraid. If they are afraid, it does not mean we are.’6

Another important source on conflicts and small arms is Tornay’s (1979) account of his visit to the Nyàngatom from 1970 to 1976. In 1970, Dassanetch, Turkána and Nyàngatom were in such violent conflict in the Ilemi Triangle that the Kenyan police received orders to make the triangle a conflict-free zone by expelling all three (Tornay 1979: 103), an order that might have been given during the colonial era (Eulenberger 2015). After this period, the Nyàngatom clashed with both the Turkána and the Dassanetch. In an incident described by Tornay (1979: 106), the Nyàngatom–as so often before and since–involved the Toposa in their conflict with the Dassanetch by mounting a joint raid. They brought home some 1,000 head of cattle and 2,000 additional small stock and donkeys. These helped them replenish their herds and buy ammunition from Maji (ibid.: 106).7 Additional clashes included the Dassanetch attack on Nyàngatom in 1972, when at ‘least 204 Nyàngatom’ died, and the Hamar/Kára attack on the Nyàngatom in 1973, when 80 to 100 Nyàngatom died (Tornay 1979de Waal 1991). In addition, after the Turkána were disarmed by the British, Nyàngatom, Toposa and Dassanetch massacred entire villages, killing hundreds of women and children (Eulenberger 2015).

Tornay discusses how the bond between the Nyàngatom and the Kára started to deteriorate in 1968. According to his records, small clashes continued in 1971, with one or two dying on each side, and the allies of the Kára, the Hamar, joined the counter-attack (Tornay 1979: 103). At about the same time, in 1972, when the Ethiopian government was pushing for a peace agreement with the Dassanetch, the Nyàngatom concluded a peace agreement with the Kára. Tornay argues that this took place ‘voluntarily, although under some economic pressure’ (ibid.). This peace did not last long, since the Nyàngatom were starving and sorghum theft was the main way of acquiring food from the Kára. Finally, the Kára and the Hamar raided the Nyàngatom and the Nyàngatom later avenged the casualties.

The Nyàngatom were the most vulnerable group and they suffered the majority of attacks by the Dassanetch during this period. The ‘Yellow Guns’, then under the leadership of Luguti (the Nyàngatom leader or ‘king’ of the time), took [End Page 743] revenge slowly and tactfully.8 The conflict in Sudan was so intense that many Nyàngatom acquired automatic guns and even portable rocket launchers easily from the Sudanese. After more than ten years, in 1988, Luguti gave his blessing for the youth to avenge the Dassanetch attacks of the 1970s. Many Dassanetch died and the Nyàngatom raided many cattle. Coming back from the raid, the Nyàngatom had to go through Kenya, luring the Turkána and the Kenyan police. Both were enraged about the number of casualties and the chaos created by the Nyàngatom. The Kenyan police approached the Nyàngatom the following day, whereupon the Nyàngatom killed every one of them, leaving the driver to return alone.9 The next day the Kenyan army was granted permission by the Ethiopian government to bomb the Nyàngatom.10 Alex de Waal reports this incident in Evil Days: thirty years of war and famine in Ethiopia (1991); the incident led to the death of about 700 Nyàngatom and five Nyàngatom villages were partly destroyed. The apparent agreement of the Kenyan and the Ethiopian governments was a result of their consensus on the need to penalize the Nyàngatom for ‘destabilizing’ the region. The actions of the Kenyan government further strengthened the Nyàngatom’s enmity with ngímòi (strangers) and consequently with their traditional enemies, the Turkána.

This Nyàngatom attack on the Dassanetch in 1988 is a case often cited to show the deadly nature of raids after the introduction of automatic rifles. Sagawa (2010) argued that this particular incident took place immediately after the introduction of such weapons and that it is not representative. However, the 1970s had seen equally lethal raids and the enmity that resulted in the incident was rooted in historical relations rather than in the presence of guns. For comparison, one should also look at an incident that took place more than a decade later on 23 February 2003, when a Nyàngatom raiding party attacked the Hamar. The Hamar killed more than ninety Nyàngatom men and lost only three to five people, crediting their triumph to the old-fashioned rifles they used during the battle; to them, these models were easier to use for accurate targeting.11 In another incident, the Nyàngatom killed 600 to 800 people (10 per cent of the Mùrsi) in 1987. Alvarsson (1989: 68) reports that this incident was fatal to the Nyàngatom as well, because ‘[t]he military order was soon transformed into an unordered and undisciplined row. At least eight Nyàngatom warriors were killed from behind by their own forces during the phase of wild shooting.’

Armaments in the Karamoja (in)security complex

We advise our young boys saying, ‘You should know your own people so you would not shoot at them . . . but if the enemy comes, kill.’12

These are words Nyàngatom elders use to educate a boy whose responsibility has increased enormously because his father has bought him a gun. Nyàngatom start [End Page 744] to carry guns when their parents and society consider them mature and responsible enough.13 The elders explain: ‘We do not give arms to children because he might, by mistake, kill a family member. When the child is ready and is old enough, the father gives him a gun.’ The Nyàngatom boy has to be able to distinguish an enemy from family. Both parents and the whole society play a big role in helping boys grow up to know their enemies through stories and songs.

Stories contain lessons about the Nyàngatom or the friends of the Nyàngatom (the Toposa, who are their maternal cousins, according to oral history) and about the killing of the enemy (the Turkána). The enemy is an existential threat. It threatens to take one’s life, the lives of family and friends, and the means of livelihood: livestock. Naturally, such existential threats force the potential victim to take extreme measures against the enemy (Williams 2008). Nyàngatom pastoralists commission their young men to counter these existential and ‘life-determining’ threats (Booth 2007).

The Nyàngatom are trained to know their own. As mentioned above, the Toposa are close friends of the Nyàngatom while the Turkána are just enemies, like all non-Ateker groups except the Omo Murle and the Koégu. The Omo Murle are a small offshoot of the Murle of South Sudan and were assimilated into the Nyàngatom ethnos as the seventh territorial section (Ngarìch) (Tornay 1981Gebre 2014). To the Nyàngatom, these two groups are not befriended but assimilated.14 With the inclusion of these groups in the woreda15 the Nyàngatom have named after themselves, the bond relationships and assimilation become further institutionalized. According to the Nyàngatom woreda police,16 crime or conflict within the Nyàngatom ethnos take place without guns. For instance, as an informant17 explained, if a Nyàngatom boy abducts a girl who is also a Nyàngatom, her family will go to him to seek retribution with sticks, not guns. Police officers who have had a chance to work in other areas assert that the woreda should be congratulated for being one of the safest places in Ethiopia, if it were not for the conflict with its neighbours. The Nyàngatom know their own in the same way they know their enemy, and, as Gebre (2014: 57) explains, ‘intra-Nyàngatom raiding is unknown and unthinkable’. There is no gun violence within the twenty kebeles18 in the Nyàngatom woreda. Intra-Nyàngatom relationships are governed by cultural and traditional mechanisms that regulate people’s interactions. The Nyàngatom believe that the family of [End Page 745] the person who kills a fellow Nyàngatom will die of a curse associated with this unpropitious action.19 While the vital importance of the taboo on lethal violence and intra-ethnic predation for the well-being of the community is explicitly recognized among Ateker pastoralists, traditional beliefs like this strengthen it substantially. Eulenberger (2013: 78) elaborates the importance of such beliefs: ‘This aspect marks the most crucial point of differentiation, that is, the contrast between the realm of [relative] permanence of this taboo (ethnos and allied groups) and the realm of its periodical and unpredictable inversion into the mode of nearly unrestrained enmity.’ Killing a Toposa, Omo Murle or Koégu falls under the same category of socially prohibited acts. In reality, Nyàngatom armed youth not only refrain from raiding ‘their own people’ but also protect them and their property. The guns in the hands of the youth provide security for all Nyàngatom. The following interview with a Nyàngatom elder illustrates this:

What is the difference between a boy with a gun and the one without?

A boy with arms will not be afraid to take his cattle to places that are far.

So, the difference is in the way a person with arms and the one without shepherd their cattle?

No, the one without a gun is friends with the one with a gun. The two of them tend their cattle together.

When a Nyàngatom is looking for a gun, he commonly starts by asking around among the Nyàngatom in Ethiopia for someone looking for a buyer.20 If this is not successful, he (and his friends) could make a trip to the Nyàngatom of the Naita area on the border with South Sudan, or to the Toposa living there. For the Toposa, acquiring guns is not difficult. War has lingered in South Sudan for almost forty years, and the trade in arms and ammunition is still booming and penetrates even the remotest areas. During the second Sudanese civil war, many Toposa and Nyàngatom joined one of the competing factions, mostly the SPLA, in order to receive initial training and a gun, which they would then run off with as soon as possible.

When first bought, cleansing ceremonies are held for guns.21 A gun passed on from father to son does not need such cleansing. However, guns bought or plundered during a war are suspected of harbouring curses that might be passed on to the new owner.22 To avoid such misfortune, the Nyàngatom slaughter a goat and cover the gun and the new owner with the animal’s dung. This is indicative of the cultural value small arms hold and is clearly a way of coping with a different form of insecurity: that is, curses that are transferable with material gifts. Through this ceremony, the boy is protected from possible curses and bad luck. He also takes this opportunity to make sure that everyone is aware that he now possesses a gun. Family members take turns firing the gun, making noise and drawing [End Page 746] attention to it. Fathers use the opportunity to bless the boy and feast with neighbours with what is left of the goat.23 Similar practices are found among neighbouring pastoralists. Sagawa (2010: 95) explains that, upon purchasing a gun, the Dassanetch sit around the gun ‘located at the centre of the house and say “kill the enemy”‘.

Alongside human enemies, wild animals are another commonly mentioned threat against which the possession of small arms can be deployed. In this context, being free from the fear of the two threats is considered security. Although other inevitable threats such as drought are confounding realities, the Nyàngatom are certain that they can provide security for their family and livelihood against threats imposed by wild animals and human enemies using guns. At least in principle, the Nyàngatom will only kill a person who has come to kill or raid. Therefore, a person who comes in peace, even if they are an enemy, is left untouched. However, if a Nyàngatom man decides to kill someone, he must drive him away from Nyàngatom territory before doing so.24

Conceptualizing enmity in the Karamoja (in)security complex

Turkána, Kára, Dassanetch, Hamar, Mùrsi–all these are enemies. [Singular: emòit; plural: ngímòi,25 also translated as ‘stranger’]

(Nyàngatom pastoralist, 2014)26

Everyone except the Toposa is our enemy. Even you are an enemy–but you are a type that does not come to kill. The Turkána and our other neighbours are the kind of enemy that kills.

(Nyàngatom civil servant, 2014)27

In most wars, the enemy is not an abstract figure: a soldier knows one when he or she sees one. Soldiers in all parts of the world are trained to identify the enemy; for example, a soldier in another part of the world stated: ‘I never felt guilty about killing people who deserved to die. In my eyes, they deserve to die because they are the enemy. I am trained to think that way.’28

The Súri, according to Abbink (2009: 37), have different categories of neighbours, among them the category of ‘real enemies’, which includes the Toposa and Nyàngatom, whom they respect despite their traditional enmity. Gebre (2014) observes that:

based on the existing historical, religious, cultural and linguistic backgrounds, one might predict that the Nyàngatom would maintain [peaceful] forms of relationships with their fellow Ateker neighbours in Kenya (Turkána) and South Sudan (Toposa) [with which [End Page 747] they share their language and most of their culture] and yet a different one with the non-Ateker groups in Ethiopia.

But this is not the case. The social sanctions that apply to killing fellow Nyàngatom also apply to the Omo Murle and the Koégu, who are not Ateker and not even Nilotes but, originally, spoke totally different languages. However, their otherness is not completely forgotten. For example, in a conflict between the Kára and the Nyàngatom, the Kára avenged their dead when they ‘massacred the inhabitants of two villages of the Nyàngatom Ngaric section’ (Girke 2008: 196). Alvarsson (1989: 71) explains how the ‘massacre’ was not a real humiliation to the Nyàngatom because the people killed were ‘only the Murle’.

The Nyàngatom word ‘emòit‘ is used for all non-Nyàngatom except the Toposa, the Omo Murle and Koégu. As a stranger to the land of the Nyàngatom, the researcher is referred to as emòit, but not all ngímòi are the same:29 one kind is the enemy who comes to kill; another is the stranger who comes in peace. However, as long as both are ngímòi, crossing over from one category of emòit to another requires the slightest act of aggression. Emòit does not necessarily refer to a member of a group with which the Nyàngatom are currently fighting. When a group is labelled ngímòi, a short break in the pattern of conflict does not change their status.30 Their status is not one that people acquire through their acts; it is one they have depending on whether or not they are born a Nyàngatom31 or a Toposa.

This is not unique for the Nyàngatom people. Sagawa (2010) identifies a similar pattern in the relations of the Dassanetch and their neighbours. Of the six groups that live around the Dassanetch, one is referred to as ‘gaal Kunno‘, meaning ‘our people’, while all the others are the enemy–’Kiz‘. This does not mean that the Dassanetch engage in constant violence with the Kiz; they even make friends with and marry from the groups they consider their enemy. However, violent conflict with these groups is always possible and contrasts with the intolerability of killing fellow Dassanetch, a cultural taboo. Guns owned by the Dassanetch are a threat to the Kiz and guns owned by the Nyàngatom a potential threat to ngímòi, but neither the Nyàngatom nor the Dassanetch have descended into unprecedented violence as a result of gun ownership, and intra-ethnic violence remains a taboo.

The Hamar present a similar case by referring to ‘Amhara32 with the same word they use for their warring enemy neighbours: gal. Such enmity is consistent with the features of ‘simple warfare’;33 according to Glowacki and Wrangham (2013: 444), one characteristic of such warfare is that ‘violence is directed towards others not because of that person’s individual identity but because of their membership in a particular group’. So in these pastoralist societies, even if there are bond [End Page 748] friendships between people from different communities, including communities fighting each other, the ethnic identity of the person is key to the attitudes, rules and choices shaping their interaction. Abbink (1994: 6) writes about such friendships: ‘These individuals remain “Mùrsi” or “Bodi” or “Hamar” or “Súri”, i.e. are judged and dealt with on the basis of their collective identity. Individual behaviours do not count in the final instance.’

This case helps to illustrate two important facts: one is that the sharp change in intra-ethnic relations after the introduction of automatic weapons (as in the case of the Súri (Abbink 2009)) is not characteristic of all pastoralist communities; the second is that even if there is a visible difference between the pre- and post-Kalashnikov era, one cannot necessarily attribute the changes to this one factor only. In contrast to Nyàngatom society, among the Súri, intra-ethnic killing using guns and other forms of disintegration have reached a level seen as ‘more problematic’ by many Súri than the conflict with their neighbours (Abbink 2009: 47).

The Turkána are currently involved in continuous violent clashes with the Nyàngatom, especially after the incidents of the late 1980s.34 Without doubt, the Nyàngatom and the Turkána are now fighting enemies, but the claim that this came about only after–and because of–the introduction of firearms is not borne out by the available reconstructions of Ateker history (Tornay 2001Knighton 2005). Factors such as the availability of pasture and food shortages can have a substantial impact on decisions regarding peace or war, but they do not determine who is targeted. The who has been established through decades, or even centuries, of cooperation and alliance, and there are set standards regarding the relations the Nyàngatom maintain with their neighbours–relations for which the RSC is an appropriate characterization.

The enemy can be distinguished by their type of attire and physical traits.35 In this case, the enemy comprises the Turkána, Kára, Dassanetch, Hamar, Mùrsi, Súri and all others who pose threats to the Nyàngatom’s physical and economic security, for example by raiding livestock and obstructing access to pasture. In these situations, arms are important instruments for the protection of priceless assets. The Nyàngatom still feel the need to keep arms for protection even from neighbouring communities with whom they are not currently in a state of violent conflict (like the Kára), because latent hostility remains operative even if there is no current violence.

Bound in tradition and geography

Regional reports on community and ethnic conflicts produced by the Federal Police Commission36 from September 2012 to September 2014 show a similar [End Page 749] trend to that described above. Although the data were not fully accessible, what I could find shows that there have been five major violent incidents between the Nyàngatom and their neighbours–two with the Dassanetch, another two with the Turkána and one with the Súri.

The Nyàngatom and their neighbours still fall in and out of conflicts without breaking the trend. For instance, the Kára lived side by side with the Nyàngatom by the Omo River within living memory (Girke 2008: 198). The Nyàngatom now cultivate the land, while the Kára continue to claim it.37 Tension continues because the Nyàngatom dismiss the claim as invalid and the Kára insist that the Nyàngatom should be confined to Nákuá, a region along the Kibish River where their grandfathers resided in the 1970s and early 1980s. The territorial claim appears to be a plausible cause of conflict, but early accounts indicate that the two groups had conflicts even before this particular territory became an issue. Although the last violent incident occurred in 2008, the conflict is still alive in its latent form.38 For the Nyàngatom, territorial rivalry is perceived as a problem they have with the Turkána, as it is the latter who have displaced them from important areas in recent decades; while for the Kára, territorial rivalry is an issue they have with the Nyàngatom. According to the Kára, they have lost the western bank of the Omo to the Nyàngatom because of the Nyàngatom’s increased need for agricultural land, which could be explained by the combined effect of several factors: population growth (six-fold since the introduction of food relief, modern healthcare and assault rifles in the 1980s–i.e. from the previously relatively constant level of 5,000 to 30,000 today (Tornay 19932009Eulenberger 2015)); overgrazing of the Nákuá pastures as a result of development interventions by the Swedish Pentecostal mission (Tornay 2009: 82); and the loss of practically all former western grazing lands and cultivation sites due to Turkána and Kenyan exclusion policies (Gulliver 1951Tornay 19932009Lamphear 1992).

Compared with the relationship they have with neighbours such as the Kára, the most dangerous and relentless enemies are the Turkána. Compared with other kebeles in the woreda, the kebeles around Kibish39 (bordering the Turkána in Kenya) are those most affected by the continuous violence. When travelling to several parts of the woreda, I observed that there is a constant risk in the area of killings, theft, raids and violent clashes. There are more men visibly carrying arms, increased awareness of and vigilance towards movements (of both humans and animals), and the sound of gunshots is extremely frequent compared with the situation in other kebeles.

At Kajamakin, one of the Kibish kebeles, situated less than 1 kilometre from where the Kibish River flows seasonally, people do not go to the river without a gun. Every morning the armed men in the community accompany the women to the waterholes at the river. There they water the goats in small flocks because it is too dangerous to take them all at once. This is the sort of occasion on [End Page 750] which police officers will accompany people to the river; luxuries such as bathing are unheard of.40

The portrayal of the Turkána as the aggressors is attributed specifically to what Admasu (2014) called ‘Kenya’s northward expansion’, which occurred during the colonial period and after the bombing in 1988. In addition to the northward push of the Nyàngatom by the Turkána, Tornay discusses their northward push by the Dassanetch. Tornay (1979) presents the shift of the ethnic frontier zone from the 1920s to the 1970s; during this period of fifty years, the Nyàngatom moved approximately 40 kilometres from Leere (near Kelem) to Natikar (Kibish), initially a transhumance grazing site from which the Nyàngatom and the Dassanetch had jointly displaced the Turkána earlier. Abbink (2009: 37) describes a corresponding shift of Súri territory northwards and eastwards by the combined Nyàngatom and Toposa, who ‘wreak havoc among Súri cattle herds and have forced them to migrate to the north-east’. The Naita Mountains (Shulugui and T’amudir), which were once the ancestral land of the Súri and, according to their oral tradition, the area where they first emerged, are now 50 to 60 kilometres south of where they reside today. Inter alia, the relationship of the Nyàngatom and the Súri today is marked by active violence whenever they encounter each other in the Tirga Mountains. According to Tornay (2009: 82), the Nyàngatom and the Toposa were helped by the SPLM (which also had Nyàngatom and Toposa members) in militarily deposing the Súri from their territory. Dassanetch–Turkána relations show similar dynamics: the Turkána were able to occupy all the traditional Dassanetch grazing lands west of the border, not least due to the support they received from the British and Kenyan governments, including the supply of arms and ammunition to Turkána pastoralists (Carr 1977Almagor 1979: 128, 143).

Turton (1991) observes that neither the Nyàngatom nor their neighbours think of their current territories as the places they historically occupied. They have all come here from somewhere else and have shifted settlements and moved northwards within living memory of most people. Despite this, grazing and cultivable land on the riverbanks is one of the resources pastoralists fight over. As Turton (ibid.: 256) himself points out: ‘Flood retreat cultivation is relatively reliable, since it depends on the heavy “summer” rains which fall over the highland catchment area of the Omo . . . [while] shifting cultivation is highly unreliable, because of the low and erratic local rainfall.’

In answering the question of whether group-level benefits motivate individuals to participate in warfare, Glowacki and Wrangham (2013)observe that, unlike other communities who plunder other materials, for pastoralists the raiding of cattle could be a sufficient motivation for warfare. An individual in desperate circumstances pays fifteen or twenty cows to buy a gun to protect his or her property and family from the enemy. When the person becomes economically unstable, he sells back the gun in exchange for cattle to feed his or her family; by being cashed in during an emergency, the gun serves as a kind of rainy day fund. Most of what the Nyàngatom do ceremonially and economically with guns shows that the Nyàngatom’s true culturally valued treasures are their grazing lands and cattle rather than guns. As one moves closer to the Kenyan border, where the [End Page 751] Nyàngatom and the Turkána clash continually, the demand for arms increases tremendously. The Nyàngatom often sell their guns in time of need or sell an extra gun that comes into their possession as the spoils of war, but they do not trade arms to meet basic consumption needs. Whenever their cattle die because of drought or disease, they sell their arms to replenish them, again putting their money where their heart is–in their cattle. Small arms in Nyàngatom do not necessarily represent ‘cultural relevance’ or serve as a ‘status symbol’, whereas the cattle the arms are meant to protect are the very essence of a pastoralist’s identity.

Although the local police41 and community elders mention territory as the root cause of conflict, cattle raiding is the most frequent form of violence both today and over the known history of the groups concerned. The act of cattle raiding is commonly justified by notions of revenge. When a person loses cattle, or simply wants more, he is bound to remember that once–either recently or in the more distant past–he or his family was raided by Turkána. He considers it a debt that he wants to collect because he is now in need.42 The Dassanetch call this debt eu (Sagawa 2010: 91), which can refer to a debt resulting from a gift such as livestock or from the death of a person. Ateker languages talk of atachàkin–’to pay back’, ‘to reattribute’ and ‘to balance accounts’, both in regard to favours owed to someone and to the reciprocation of harm.

Conclusion: a Karamoja security complex

The facts discussed above demonstrate that pastoralists in this part of the Horn of Africa present an intriguing case of multifaceted security interdependence. The Nyàngatom and their neighbours can be analysed as being composed of units constituting a Karamoja security complex (KSC), a regional construction that is more fluid and flexible than Buzan and Wæver’s regional security complex (2003). Emerging here is a model more cognizant of the complexity of regions as social constructs but still possessing core characteristics of the RSC. The primary characteristic of a regional complex is the ‘boundary, which differentiates the RSC from its neighbours’ (ibid.: 53). While certain geographic and socio-cultural characteristics are shared by pastoralist groups in the Horn, the boundary is maintained through practices that uphold the social construction of identities and norms, distinguishing those groups from other cultural communities. The boundary of the territory claimed by the Nyàngatom and their neighbours surpasses the effectively controlled territory and includes lands that are imagined and claimed (for reasons of ancestral connection and cultural attachment). The KSC is therefore a phenomenon whose boundaries will shift as a result not only of population movement, but also of changes in identity, culture, historical narratives and beliefs. These aspects have proved mutable in the Karamoja cluster; an example would be the incorporation of the Koégu and the Omo Murle in the traditional territorial sections of the Nyàngatom.

Two other important characteristics are the ‘anarchic structure’ and ‘polarity of power’ among components making up the KSC. The autonomy and polarity [End Page 752] of pastoralist communities before the introduction of the modern state system persisted in much of postcolonial Kenya, Sudan, South Sudan, Uganda and modern-day Ethiopia. This is partly due to the nature of statehood in the region, which for a long time has been characterized by low state presence and the weak regulatory impact of modern institutions on the life of local populations. These conditions allowed traditional systems of social organization and regulation to remain largely intact in many indigenous communities. The Nyàngatom are a case in point. In other communities, such as the Súri, these systems were seriously weakened but have not totally collapsed. Besides their formal status in the local government of the Ethiopian federal structure, these pastoralist communities make their own choices and alliances and govern their territories in an independent–even if negotiated–manner (Tornay 197919932009Girke 2008Abbink 2009Eulenberger 201320142015). Some of the autonomous units constituting the KSC are found within the same state (for example, the Nyàngatom, Dassanetch, Súri and Hamar), demonstrating that, like the regional and international system (Wendt 1999: 2), individual states have spaces where law and institutions have very limited capacity with certain anarchic characters. Comparatively, some multi-state regions in the world have less anarchy: in other words, they have achieved greater ‘stateness’ or ‘nationness’ than places such as the Lower Omo, where the Nyàngatom are found.

The KSC discussed here also has a ‘clear pattern of amity and enmity among the units’ (Buzan and Wæver 2003Buzan et al. 1998: 13). This pattern is not ‘permanent’ but is ‘durable’ (Buzan et al. 1998). Knighton (2003: 436), after analysing the historical relations of the Karamojong, concludes that: ‘The technological qualities of different weapons, whether spears or guns, are a factor in the outcome of battles, but they do not nullify the pattern.’ In some cases, the units in the KSC do not border each other directly but the relations within the complex are so intense that the security situation of one has a strong bearing on the situation of another, and ‘their security problems cannot be analysed separately’ (Buzan et al. 1998Buzan and Wæver 2003).

The main security issues binding these pastoralists seem to be actual and perceived existential threats directed by their enemies against their livelihood (mostly cattle and sometimes sorghum) and their territory (for grazing and flood-retreat agriculture). Over the years, the work of pioneering researchers such as Rada and Neville Dyson-Hudson, who state that ‘territorial conquest does not appear to be the objective of fighting’ (1980: 46), has guided research into ecological approaches. Among those who strongly challenge such claims is Greiner (2013). The security complexes have intertwined the security needs of actors that are geographically close (Buzan and Wæver 2003: 4); thus, the factor that is most significant in the above analysis is not the importance of territory but the relatively permanent character of East African pastoralist societies and their boundedness within an increasingly shrinking geographic space that makes them especially vulnerable to threats that travel easily and effectively over shorter distances, such as the threat of small arms. Small arms are not the source of the hostile relationships maintained by groups such as the Nyàngatom with their enemies: they are ‘tools’ in the hands of a man with insecurities. An emerging area of cooperation among the Horn states relates to small arms control; sub-regional instruments including the Nairobi Protocol (2004) have emerged along with the Regional Centre on Small Arms (RECSA), [End Page 753] which was established to coordinate implementation.43 It is essential that these initiatives operate with the understanding that units in the region have interwoven security concerns and relations that have been constructed over more than a century.

Mercy Fekadu MulugetaMercy Fekadu Mulugeta PhD is a lecturer and researcher, currently working on an African-centred solutions (AfSol) research and publication project at the Institute for Peace and Security Studies (IPSS), Addis Ababa University. Her research interests include small arms, security governance in border and pastoral lands and state building. Email:


Part of the data collection for this research was financially supported by the German Gesellschaft für Internationale Zusammenarbeit (GIZ). I also appreciate the valuable and constructive comments from the three anonymous reviewers and from Immo Eulenberger.


Abbink, J. (1994) ‘”Tribal” violence, peace-making and ethnography: a comment on Prof. I. Strecker’s paper “War and Peace in South Omo”‘ in Z. Bahru (ed.), Proceedings of the Eleventh International Conference of Ethiopian Studies, Addis Ababa, April 1–6 1991. Addis Ababa: Institute of Ethiopian Studies, Addis Ababa University.

Abbink, J. (2009) ‘The fate of the Suri: conflict and group tension on the southwest Ethiopian frontier’ in G. Schlee and E. E. Watson (eds), Changing Identifications and Alliances in North-East Africa. Vol. 1: Ethiopia and Kenya. New York NY: Berghahn.

Admasu, L. (2014) ‘The complex inter-ethnic interactions along the Ilemi Triangle: the case of Nyàngatom, Toposa and Turkána’. MA thesis, Addis Ababa University.

Almagor, U. (1979) ‘Institutionalizing a fringe periphery: Dassanetch–Amhara relations’ in D. L. Donham and W. James (eds), The Southern Marches of Imperial Ethiopia. Cambridge: Cambridge University Press.

Alvarsson, J. A. (1989) Starvation and Peace or Food and War? Aspects of armed conflict in the Lower Omo Valley, Ethiopia. Uppsala: Uppsala University.

Appadurai, A. (2000) Globalization and Area Studies: the future of a false opposition. Amsterdam: Centre for Asian Studies Amsterdam.

Berouk Mesfin (2011) ‘The Horn of Africa security complex’ in R. Sharamo and Berouk Mesfin (eds), Regional Security in the Post-Cold War Horn of Africa. Pretoria: Institute for Security Studies.

Booth, K. (2007) Theory of World Security. Cambridge: Cambridge University Press.

Börzel, T. A. (2011) ‘When Europeanization hits limited statehood: the Western Balkans as a test case for the transformative power of Europe’. KFG Working Paper 30. Berlin: Freie Universität Berlin.

Buzan, B. (1991) People, State and Fear: an agenda for international security studies in the post-Cold War era. Colchester: ECPR Press. [End Page 754] 

Buzan, B. and O. Wæver (2003) Regions and Powers: the structure of international security. Cambridge: Cambridge University Press.

Buzan, B., O. Wæver and J. de Wilde (1998) Security: a new framework for analysis. Boulder CO: Lynne Rienner.

Carr, J. C. (1977) ‘Pastoralism in crisis: the Dassanech and their Ethiopian lands’. Research Paper 180. Chicago IL: Department of Geography, University of Chicago.

Cormack, Z. (2016) ‘Borders are galaxies: interpreting contestations over local administrative boundaries in South Sudan’, Africa 86 (3): 504–27.

de Waal, A. (1991) Evil Days: thirty years of war and famine in Ethiopia. New York NY, Washington DC and London: Human Rights Watch.

Dyson-Hudson, R. and N. Dyson-Hudson (1980) ‘Nomadic pastoralism’, Annual Review of Anthropology 9: 15–61.

Eaton, D. (2008) ‘The business of peace: raiding and peace work along the Kenya–Uganda border (Part 1)’, African Affairs 107 (426): 89–110.

Eulenberger, I. (2013) ‘Pastoralists, conflicts and politics: aspects of South Sudan’s Kenyan frontier’ in C. Vaughan, M. Schomerus and L. de Vries (eds), The Borderlands of South Sudan: authority and identity in contemporary and historical perspectives. New York NY: Palgrave Macmillan.

Eulenberger, I. (2014) ‘Gifts, guns and govvermen: South Sudan and its southeast’ in S. Calkins, E. Ille and R. Rottenburg (eds), Emerging Orders in the Sudans. Bamenda: Langaa RPCIG.

Eulenberger, I. (2015) ‘A brief history of the Ilemi region’. Unpublished manuscript.

Gebre Yntiso (2014) ‘The Nyàngatom circle of trust criteria for ethnic inclusion and exclusion’ in S. Epple (ed.), Creating and Crossing Boundaries in Ethiopia: dynamics of social categorization and differentiation. Zurich: LIT Verlag.

Ginifer, J. and M. Turner (2005) ‘SALW and development programmes: issues and priorities’. Background paper prepared for the Project ‘European Action on Small Arms and Light Weapons and Explosive Remnants of War’. Geneva: United Nations Institute for Disarmament Research (UNIDIR).

Girke, F. (2008) ‘The Kara–Nyàngatom war of 2006–2007: dynamics of escalating violence in the tribal zone’ in E. M. Bruchhaus and M. M. Sommer (eds), Hot Spot Horn of Africa Revisited: approaches to make sense of conflict. Berlin: LIT Verlag.

Glowacki, L. and R. W. Wrangham (2013) ‘The role of rewards in motivating participation in simple warfare’, Human Nature 24 (4): 444–60.

Gray, S., M. Sundal, B. Wiebusch, M. A. Little, P. W. Leslie and I. L. Pike (2003) ‘Cattle raiding, cultural survival, and adaptability of East African pastoralists’, Current Anthropology 44 (S5): S3–S30.

Greiner, C. (2013) ‘Guns, land, and votes: cattle rustling and the politics of boundary (re)making in Northern Kenya’, African Affairs 112 (447): 216–37.

Gulliver, P. H. (1951) A Preliminary Survey of the Turkána: a report compiled for the government of Kenya. Cape Town: University of Cape Town.

Gulliver, P. H. (1952) ‘Bell-oxen and ox names among the Jie’, Uganda Journal 16: 72–5.

Hettne, B. and F. Söderbaum (2000) ‘Theorising the rise of regionness’, New Political Economy 5 (3): 457–72. [End Page 755] 

Knighton, B. (2003) ‘The state as raider among the Karamojong: “Where there are no guns, they use the threat of guns”‘, Africa 73 (3): 427–55.

Knighton, B. (2005) The Vitality of Karamojong Religion: dying tradition or living faith? Aldershot: Ashgate.

Lamphear, J. (1992) Scattering Time: Turkána responses to colonial rule. Oxford: Clarendon Press.

Leff, J. (2009) ‘Pastoralists at war: violence and security in the Kenya–Sudan–Uganda border region’, International Journal of Conflict and Violence 3 (2): 188–203.

Markakis, J. (2011) Ethiopia: the last two frontiers. Woodbridge: Boydell & Brewer.

Mburu, N. (2003) ‘Delimitation of the elastic Ilemi Triangle: pastoral conflicts and official indifference in the Horn of Africa’, African Studies Quarterly 7 (1): 15–37.

Mirzeler, M. and C. Young (2000) ‘Pastoral politics in the northeast periphery in Uganda: AK-47 as change agent’, Journal of Modern African Studies 38 (3): 407–29.

Raudzens, G. (1990) ‘War-winning weapons: the measurement of technological determinism in military history’, Journal of Military History 54 (4): 403–34.

Sagawa, T. (2010) ‘Automatic rifles and social order amongst the Dassanetch of conflict-ridden East Africa’, Nomadic Peoples 14 (1): 87–109.

Schlee, G. (2009) ‘Changing alliances among the Boran, Garre and Gabra in Northern Kenya and Southern Ethiopia’ in G. Schlee and E. E. Watson (eds), Changing Identifications and Alliances in North-East Africa. Vol. 1: Ethiopia and Kenya. New York NY: Berghahn.

Söderbaum, F. (2015) ‘Early, old, new and comparative regionalism: the scholarly development of the field’. KFG Working Paper 64. Berlin: Freie Universität Berlin.

Tornay, S. (1979) ‘Armed conflicts in the Lower Omo Valley, 1970–1976: an analysis from within Nyàngatom society in warfare among East African herders’, Senri Ethnological Studies Osaka 3: 97–116.

Tornay, S. (1981) ‘The Nyàngatom: an outline of their ecology and social organization’ in M. L. Bender (ed.), Peoples and Cultures of the Ethio-Sudan Borderlands. East Lansing MI: Michigan State University Press.

Tornay, S. (1993) ‘More chances on the fringe of state: the growing power of the Nyàngatom’ in T. Tvedt (ed.) Conflicts in the Horn of Africa. Uppsala: Uppsala University.

Tornay, S. (2001) Les Fusils jaunes: générations et politique en pays Nyàngatom (Éthiopie). Nanterre: Société d’Ethnologie.

Tornay, S. (2009) ‘Modernization in the Lower Omo Valley and adjacent marches of Eastern Equatoria, Sudan: 1991–2000’ in G. Schlee and E. E. Watson (eds), Changing Identifications and Alliances in North-East Africa. Vol. 1: Ethiopia and Kenya. New York NY: Berghahn.

Turton, D. (1991) ‘Warfare, vulnerability and survival: a case from south-western Ethiopia’, Disasters 15 (3): 254–64.

Wendt, A. (1999) Social Theory of International Politics. Cambridge: Cambridge University Press. [End Page 756] 

Wepundi, M., E. Nthiga, E. Kabuu, R. Murray and A. Alvazzi del Frate (2012) Availability of Small Arms and Perceptions of Security in Kenya: an assessment. Geneva: Small Arms Survey.

Williams, P. (ed.) (2008) Security Studies: an introduction. New York NY: Routledge. [End Page 757] 

Leave a Reply