CHAPTER 5

THE SPREAD OF THE WAR

Apart from attacks in West Nile and Lango since the early 1990s, the LRA conflict has been felt most intensely in the districts of Gulu, Kitgum and Pader. The year 2003, however, saw what appears to have been a deliberate and well coordinated spread of the war, to the east in particular. In May 2003, the LRA ambushed a bus along Pakwach-Karuma road in Gulu District. 207 This was followed on 15 June 2003 by simultaneous attacks in the districts of Lira, Apac and Katakwi, in what some in Teso have called their “September 11th ”. 208 Three days later, the LRA attacked Adjumani town for the first time in fourteen years. 209 These attacks were soon followed by others of greater frequency and intensity in the districts of Kaberamaido and Katakwi in Teso region. On 24 June, the LRA attacked Soroti town, resulting in at least 200 deaths and the abduction of hundreds more children, including 100 schoolgirls. 210 A group of IDPs in Lira described this sequence of events: “The attacks were gradual. We knew the rebels were camped at a certain place, but they did not attack for a long time, so we waited . We told the UPDF but they did not respond. So we waited , and suddenly places were attacked and because they didn’t protect us, we had to run.” 211 Thus, LRA attacks have not only continued with intensity across the Acholi sub-region, but have also spread to several other areas.

On 26 June, it was reported that more than 2 000 veterans of the defunct Teso-based rebel Uganda People’s Army (UPA) led by Musa Ecweru, then Resident District Commissioner (RDC) of Kasese, and local MP John Eresu, had joined the UPDF to fight the rebels in Teso. 212 This development was later opposed by a group of northern parliamentarians, who viewed the deployment of paramilitary groups as unconstitutional. 213

Explaining the extension of the conflict

The movement of the LRA into Teso and Lango appears to have occurred for several reasons. First, many believe that the LRA were running short of food and other logistical supplies, having thoroughly looted Gulu, Kitgum and Pader for the past 17 years. With more than 80% of the population in these districts displaced in IDP camps, most of the land currently lies fallow. Given the fact that the LRA relies chiefly on plundering agricultural stocks to resupply its food needs, and that many planting seasons have been missed in these districts, there are few crops left to loot. Teso, on the other hand, is known as a fertile farming region that supplies beef, chicken and potatoes to other regions of Uganda . It was thus a key target area to attack when foodstuffs were in short supply in the usual operational area.

Second, the LRA appears to have believed it could gain support from these areas, particularly among former rebels from Teso and Tororo/Busia. As an official in Soroti commented ,

[Kony] thought that since we had a rebellion here, he could find potential allies. They thought they could then proceed to fight the government until Kampala . They wanted to spread their area of influence. The UPDF is taking over Gulu and Kitgum, so they came here. But the people of Teso have not joined them; we have the Arrows. 214

Our interviews reveal that the LRA first spent several days in both areas without attacking. According to local residents in Obalanga (Katakwi District, part of Teso region), the rebels were initially very friendly – playing football, watching videos, and generally interacting freely with the population. 215 Informants also reported that the LRA came with a list of names of former UPA rebels who had fought against the government from 1987 to 1992. They wanted to know the locations of these ex-fighters so as to activate them to fight the “dictatorial” Museveni government. A similar thing appears to have occurred in Tororo and Busia in September and October 2003, with the LRA allegedly sending six scouts in search of fighters from the former 9 October Movement. 216 Kony appears to be following in the footsteps of UPDA fighters from Gulu, who in 1988 attempted to join UPA rebels in Teso against the government. 217 The move to Lango (Lira and Apac districts) came later in September, and appears to have been in part a reaction to being pushed back from Teso. 218

The drive to activate former fighters in the east may have been an attempt by the LRA to reorganise itself in a time of uncertainty, given the potential implications of the Sudan peace process, which brought the LRA’s key supply line increasingly under threat. In the past, Kony had shown his ability to adapt to changing circumstances. For instance, during the 1994 negotiations he evidently took advantage of a lull in the fighting to abduct more children and go to Sudan for arms and ammunition. 219 With reports that the LRA command structure had been reshuffled in recent weeks, 220 something similar may have been taking place.

Another explanation for the war’s spread was that it was an attempt to punish the Iteso for their continued support of the NRM. As one informant in Soroti said, “They say the Iteso are the strongest supporters of the Movement in the north, so they must be punished for it.” 221 Others claimed that it was a deliberate attempt by Kony to prove that, after Operation Iron Fist, the LRA was still a force to be reckoned with, giving the lie to claims that the LRA was about to be finished off once and for all. For instance, Museveni wrote in a letter to the New Vision in August 2002, “You can be sure this conflict will be over, latest by February, when the grass will have been burnt, if it goes that far.” 222 As a religious leader said, “Kony wants to prove that he is alive and well after Operation Iron Fist, which is supposed to have finished him. So to show that it was a failure, the best way is to spread.” 223

Some interviewees believed that the LRA attacked Soroti because they were seeking revenge for Acholi UNLA soldiers who were killed in Teso in 1986: “During the withdrawal of UNLA, as they were running north, they were intercepted in Teso and killed. This was in 1986. The Iteso pretended to entertain them, but they killed them at night. Some people are coming to revenge these killings.” 224 Others speculated that ex-UPA rebels who had joined the LRA in the 1980s, invited the LRA to enter their district: “We have heard of the invitation sent by the rebels to come here by many rebels, especially the ex-UPA ones who did not surrender. These ones are with the rebels and they are the ones directing them in this region.” 225

Whatever the reasons for the LRA extending its geographical focus, it has radically changed perceptions of the conflict. The extent to which the conflict is being seen increasingly within a national historical context is symptomatic of this change of opinion. In addition, and in response to such wider interpretations of the war, there has been an increase in pressure for the war to end both by those directly affected by the war, and by those increasingly aware of its protracted nature.

Rising tensions: UPDF, local militias and the LRA

Regardless of the reasons for the LRA’s extension of its geographical focus, the impact has been horrific. More than 358 000 people have been displaced in Teso and Lango since the eastward move, 226 so it is hardly surprising that informants expressed considerable anger and fear at the presence of the LRA in their districts. Such feelings have been translated into frustration and resentment towards a UPDF they view as unable to protect them from rebel attacks. The UPDF’s military strategy is seen by many citizens on the ground to be wanting in the face of the rebel group’s brutal and effective guerrilla tactics. In this respect, attitudes are similar to those in Gulu, Kitgum and Pader, and the following comment summarises the prevailing attitude towards the UPDF:

The UPDF only fight during the day. If they keep doing this, they will never end this war. They only fight during the day and even then, only when they’re attacked. That’s why this war has gone on for seventeen years. And that’s why we in Teso are very critical. There also seems to have been a big problem with the commanders. Many of them are used to soft life. We’ve refused that in Teso. 227

Local militias

Following the LRA’s continued presence in Soroti, and the UPDF’s initial inability to cope with this incursion, the Arrow Group was quickly formed to mobilise the local population against the rebels. Indeed, the rapid deployment of Arrow forces appears to have been instrumental in halting the spread of the rebels further east. According to a government insider, the group intended to move further east to Kumi, Mbale and Tororo and to begin ambushes along the road to Kenya near Tororo in order to “provoke anger, make the population think the situation is out of hand, and give the appearance that the government is not in control.” 228 In the words of an Arrow commander,

The UPDF had two main shortcomings when it came to Teso. First, they didn’t speak the language, so they weren’t getting precise information. Second, this area has very flat land – there are no physical features to guide you, so they didn’t know the terrain. So it is our mission to plug this gap. We picked young men who had knowledge about warfare, and we use them for intelligence gathering and guidance. This is where the name Arrow comes in – the Arrow flies as if it knows where it’s going. It’s a codename for precision. This was our original mission, to help guide the UPDF. 229

The Amuka, or “Rhino” group was then mobilised in Lira, but to less effect. Many informants believed that the Rhinos had not been as successful as the Arrow Group because of fewer supplies and weaker support. As one ex-Rhino female recruit said, “The Amuka boys are very few, and many who had volunteered are leaving because they don’t have guns, and you can’t fight rebels with stones.” 230 Although the Arrow group has made similar complaints, there was a clear perception that the Arrows had been more effective in resisting the LRA. In addition, there was concern within Lira that mobilising local defence would increase the killing of civilians:

They’re a militia to supplement government forces to fight against LRA. The government has to assure compensation. Being militia is a risk, people think. Not everybody supports them. 231 I don’t think [Amuka] is good. Because this LRA, if they arrest you, they kill you because they say this place is full of Amuka. 232

Rather than operate as independent militias, the Arrow and Rhino groups have thus far been deployed with UPDF units as “zonal forces” to protect civilians in rural areas. Although there are clear issues of capacity, the Arrow and Rhino groups offer potential benefits. First, their detailed knowledge of the local terrain, particularly in flatland Teso, is helpful in tracking down LRA rebels. Second, their intelligence-gathering capacity is also boosted because of their local language skills and very high levels of trust with the local populations – key factors that have reduced UPDF effectiveness in Gulu and Kitgum. As one of the Arrow commanders commented , “The first thing in this kind of situation is you need the population on your side. You can have all the technology you want, but whoever has the numbers, the population and the organisational skill will prevail.” 233 Some of the benefits of community militias in other conflicts strengthen this argument. The Kamajors or Civil Defence Force in Sierra Leone, for example, provided intelligence and local knowledge that were key factors in defeating the RUF in that country’s civil war. 234

We discovered, however, that support for the militias is mixed. On the one hand, the vast majority of people in Teso view the Arrow Group as a major resource for defence against a rebel army creating havoc in their lives. Three IDPs echoed a widespread community sentiment: “The Arrow Boys are our saviours, because the UPDF sleep. We really appreciate them.” “Arrows have been killing the LRA commanders. When we hear the name Arrow Boys, we feel saved.” “They work so hard even without food and money and without bigger weapons.” 235 However, the increase in reprisals from the LRA has created a more ambivalent attitude towards the local defence mechanisms. According to a local government official in Lira, “On 29 September, the local population reported the presence of the LRA to the LDUs who flushed them out. Then Kony sent a message to the locals that since they had refused to support them, they would come back and commit genocide on the people.” 236 Once again, Kony appears to have become very angry with the population for not supporting him, and brutal attacks against civilians soon began in a repeat of the tragic “Bow and Arrow” saga of 1991–1992. 237

Anger towards the LRA, increased ethnic tensions

Thus there is huge animosity towards the LRA in both Teso and Lango, as was the case in Gulu, Kitgum and Pader. As a rebel force, it is judged purely on its actions: the terror tactics being employed have left little room for an understanding of why Kony is fighting. Instead, the impression was one of people in a state of shock, suddenly confronted by an enemy that had previously kept its distance.

Although the LRA may have made some initial attempts to draw individuals into its ranks, as in the past it has continued to rely solely on forced recruitment. In the words of an NGO worker:

They are not interested in the local people’s support because they can always abduct. They don’t recruit those willing to join them because they don’t want government spies. If they recruited people who went there voluntarily, there would be infiltration. Children are easily brainwashed and if they get lost, they can’t easily find their way back home. 238

However, the fear that has been generated has been translated into a more general anger towards the Acholi people, as Kony himself is an Acholi. At one end of the spectrum were those who equated Kony with all Acholi people. As a head teacher said, “We feel that the Acholis have come to destabilise the innocent and development in Teso without any genuine cause.” 239 One secondary school girl alleged, “Most Acholi students turn into rebels during holidays in order to get enough money for fees and pocket money. Kony is fighting because the Acholi are generally rebellious and they like fighting.” 240

Many informants, when questioned further, acknowledged that this was not necessarily the case: “You will hear people saying all Acholi or all people from the north are supporting the rebels, which is not true.” 241 Indeed, there was recognition of the fact that there are non-Acholi elements within the LRA. 242 However, such levels of understanding have not stopped anti-Acholi sentiment from beginning to take root within the districts under recent LRA attack. A religious leader described the confused mixture of responses in this way:

Until June 15 2003 , this was a war of the Acholi and the Teso did not know what was going on. The initial reaction was of anger, and both religious and political leaders made serious attacks about the Acholi, which was not good. People do not equate LRA to Acholi, but there’s certainly a lot of anger among all the people. Someone in the village who speaks Acholi may be mob killed. The rebels are sending spies pretending to be mad, but they were found to be spies and killed and they had to use other tricks. They killed two women dressed in the Buganda gomesi [traditional dress] who went to hospital and looked at the bodies of the rebels. The people were furious. The Acholi are happy now because the war has spread from their districts because it has now left them in peace. 243

In particular, there was frequent reference to the fact that the LRA had survived for so long because the Acholi people were supporting Kony, evidenced by the fact that they were not organising themselves into effective self-defence units. The following comments illustrate this attitude: “The fundamental question is why they haven’t mobilised an Arrow group”; 244 “The only problem I have with the Acholi is that they have not come out openly to condemn Kony”; 245 “We went to Gulu … as a delegation from Soroti. We wanted to gauge the feelings of Acholi. The assessment confirmed what we were hearing that these people were supporting Kony.” 246 Such comments were made despite the well-known fact that previous self-defence mobilisation in Acholi in 1991 and 1992 resulted in vicious rebel reprisals.

Such allegations of collaboration have meant that the spread of the war, perceived by many as an “Acholi” war, has created antagonism along ethnic lines. On the one hand the conflict is seen to have become “national” since spreading geographically beyond Gulu, Kitgum and Pader, but on the other, it has become increasingly articulated along ethnic lines. For instance, this is how an Arrow boy interpreted the war:

You know, these days people are fighting against the tribe. The Iteso alone, Karamojong alone etc. For us, we don’t have a problem. If the Acholi stop their habit of taking our children, we don’t have a problem. We will not go outside Teso looking for Kony. 247

Indeed, the rise of the Arrow and Rhino groups is potentially worrying in this context, as wider ethnic clashes among different northern and eastern groups are now a potential risk. Relations between the Iteso (the overwhelming majority of the residents of Soroti, Kaberamaido and Katakwi), the Acholi (Gulu, Kitgum and Pader) and the Langi (Lira and Apac) have been peaceful over the past several years. However, differences and stereotypes are easy to re-mobilise, particularly because of a lack of interaction and communication – very few people travel to and from the war-torn areas of Gulu, Kitgum and Pader. These perceptions have surfaced as a result of the spread of the war, as some people confirmed their stereotypes of Acholi as “a warlike people” because of the violence of the LRA in their region. As a teacher in Soroti observed, “Before this war, Acholis and Iteso were good friends. But now, there’s tension. The RDC of Kaberamaido is an Acholi, and he was attacked because he left a lot of loopholes for the LRA to come. The Deputy RDC of Katakwi, also an Acholi, has run away.” 248

In trying to defend their communities, some citizens have already killed people in “mob justice” attacks simply because they were speaking the Acholi version of Luo – thereby immediately equating them with the LRA. 249 Although conciliatory public statements by politicians and religious leaders have eased the situation, there is little doubt that the animosities brewing in the Teso and Lira sub-regions could have serious ramifications for stability within the area.

Future implications

In addition to fears of tensions developing along ethnic lines, there were other concerns about the longer-term implications of arming local militias. In particular, there are a number of ways in which informants believed armed local defence units could use their weapons in the future. First, many informants expressed their fear that individuals could take advantage of being armed to settle personal scores or engage in criminal activities. This fear was expressed during a discussion with a group of IDPs: “We fear that some of them might become thieves and start robbing people with the guns they have.” 250 One local analyst added, “People are fearful of thuggery if they [Arrow Boys] keep their weapons after the LRA. There is a fear that some Arrow Boys will settle scores on family and individual levels.” 251

Second, numerous interviewees commented that the local defence units could, in future, be used to protect livestock against continuing attacks by Karamojong cattle-rustlers. As one IDP argued, “When this war ends, we should go and disarm the Karamojong ourselves because the government has cheated us about disarming them.” 252 One informant even used the word “Konymojong”, to express the wider problems confronting civilians in the Teso sub-region. As a journalist argued, “Those [Arrow Boys] in Katakwi, where there have been Karamojong cattle raids, they will remain permanently armed.” 253 Given the impact of Karamajong raids on the population, and the inadequacy of defence against these, it is hardly surprising that civilians see increased defence capacity as a means to future security.

Third, fears that the Arrow and Rhino groups could turn their weapons against the government have already been raised in many circles. 254 Teso and Lango are majority opposition areas, and between 65% and 80% of the Arrow Boys are former members of the Obote II UPC army, many of whom joined the anti-Movement UPA rebellion from 1987 to 1992. 255 Some informants believe that the fighters may therefore place significant demands on the government, become local vigilantes, or even rise up against the Movement in the event of a crisis. As one journalist commented , “The Arrow Boys are a big force of unemployed young people – they are not so educated and very energetic. If you promise them too much, there could be a lot of discontent.” 256 The government has been careful to integrate the groups into UPDF command structures so that the two will work hand-in-hand. However, rumours were already abounding in Soroti about tensions between Arrow and UPDF forces about minor issues such as who actually killed the rebels. While we were unable to investigate such claims, it is possible that these tensions could escalate into something larger in the future.

Fourth, some suspect that the Rhino and Arrow groups may be used to intimidate the opposition before the 2006 elections. Supporters of such an argument believe that the Rhino Group is much too closely assimilated into the UPDF, and that the Arrow leaders are too entrenched in the Movement to rise up against it. As one journalist said, “Once a rebel, always a rebel, but [Arrow leader] Musa Ecweru has a lot of ambition and has been a Movement supporter for a long time already. Let’s put it this way, after the war, Mukula will not be a junior minister, and Ecweru will not be an RDC.” 257 There is a suspicion that they will therefore use the militias to help the government win the elections and reap the political rewards afterwards. As one local analyst commented, “The Arrow Group is in reality a force to quell opposition groups in the lead-up to the 2006 election.” 258 While this remains pure speculation at present, such a development would not be without precedent: similar sub-state militias were used by regimes before elections in Kenya, Zimbabwe and Rwanda. 259 Thus the arming of paramilitaries, as well as having immediate consequences, could affect the stability of eastern and northern Uganda even if the LRA disappears.