Slow genocide of the Banyamulenge of South Kivu

Rukumbuzi Delphin Ntanyoma and Helen Hintjens

International Institute of Social Studies, Erasmus University, The Hague, The Netherlands

First Published on May 8, 2021

A MONUSCO delegation in Fizi, one of the territories affected by recent violences, March 16 2019 (MONUSCO/Jacob de Lange)


Recent warfare in Eastern DRC, especially since 2015, is marked by violence inspired by ‘race’ narratives. Identity politics around ‘race’ is used to legitimise ‘expressive’ or reprisal-oriented violence against ‘Hamitic’ or ‘Tutsi’ minorities. The case of the Banyamulenge of South Kivu is examined in this article. Following Autesserre, we show that one-dimensional narratives – in this case of ‘race’ – tend to over-simplify the dynamics of political violence. Anti-Hamitic racism is derived from colonial ideas around race hierarchies, and has resulted in systematic killings of Banyamulenge civilians in what resembles a ‘slow genocide’. Expressive violence has, in turn, produced a lack of concern for the plight of Banyamulenge civilians among the military, humanitarians, media, scholars and NGOs. Given armed alliances between local Maimai forces, Burundian and Rwandan opposition and the DRC army, such ‘race’ narratives cruelly legitimise violence against civilians from ‘Tutsi’ communities, associated by neighbouring communities with Rwanda. Resultant displacement, starvation and killing of Banyamulenge civilians in this context amount to an on-going, slow-moving genocide. As the COVID-19 crisis unrolls, the decolonisation of identity politics in Eastern DRC, and in South Kivu in particular still seems very remote.


South Kivu, Banyamulenge, identity politics, colonial, warfare, expressive violence, race, tribal, DRC, citizenship


The world is simultaneously moving in opposite directions. (Rosenau, 2003: 12)

South Kivu Province, in the High Plateau of Uvira-Minembwe, is facing an underreported, misunderstood and disturbing humanitarian crisis. The scapegoating of specific minorities, informed by colonial-era ideas about ‘race’ and ‘tribe’, is being instrumentalised as part of what Kalyvas terms ‘expressive violence’ against civilians (Kalyvas, 2006). In South Kivu, such forms of violence are enacted against Banyamulenge with some 150,000 civilians besieged as IDPs (Internally Displaced Persons) in Minembwe since January 2019, as well as smaller numbers in Bijombo and Mikenke, Banyamulenge are sitting targets for attacks. Violence against this minority is often justified as revenge for the past war crimes of Banyamulenge troops.

Banyamulenge civilian IDPs have, for some years, been targets of what can be seen as a slow-moving genocide (see Ntanyoma, 2019b). Slow genocide can be defined as ‘the emotional and physical harm done to survivors of violence over time…[and] emotional and physical harm resulting from witnessing or participating in violence and the continuing experiences of living in unsafe and violent communities’ (Cottam et al., 2006: 2). According to the International Crisis Group (ICG), in 2004, the situation in Darfur resembled ‘Rwanda in slow motion’ (ICG, 2004). ‘Slow genocide’ is also used to classify the long, drawn-out destruction of Palestinians by Israel (Barghouti, 2010; Lendman, 2010). Similarly, the situations of the Rohingya in Myanmar, and of West Papuans in Indonesia have been defined as ‘slow-burning’ or ‘slow-motion’ genocides (Elmslie and Webb-Gannon, 2013; Zarni and Cowley, 2014: 682). Just as Rohingya are referred to as Bangladeshis in order to deny their claims to full Myanmar citizenship, so too, Banyamulenge are referred to as Rwandans, which serves to discredit their claims to be recognised as full and equal Congolese citizens.

It is well to remember that genocide, as Sheri Rosenberg reminds us, may not consist of one spectacular set of events (although it can do), but can take the form of a gradual process of ‘genocide by attrition’, with the slow death of a people and their culture over time (Rosenberg, 2012: 18–19). This insight applies to the Banyamulenge community in South Kivu, and yet to declare the killing and starving of a community a slow genocide, even publicly, does not guarantee a strong response will come that will stop the killers in their tracks, as demonstrated in the case of Darfur 20 years ago, when declaring what was taking place a genocide did not produce the expected results (Straus, 2005).

Meanwhile, in South Kivu, Maimai attacks on Banyamulenge civilians, given the community’s stigmatised status, are ignored in the media, or framed as ‘inter-ethnic’ violence rather than one-sided killings. State failure to protect all persecuted minorities in Eastern DRC is evident. Far from promoting dialogue or a military solution, Forces Capitals Armees de la Republique Democratique du Congo (FARDC) are reported to turn a blind eye to attacks by Maimai rebels on Banyamulenge, and even recently, said to have assisted the Maimai (Stearns, 2011: 126). Amid tensions in Minembwe, Kivu Security Tracker (KST) has remarked that ‘FARDC members had given ammunition to the Maimai groups two days before’ a Maimai attack (KST, 2020a). Not surprisingly, Vogel and Stearns (2018: 6) classify the FARDC as belligerents in the third phase of warfare since 2015. In 2020 and 2021, with COVID-19, come further risks of intensified victimization of minority communities, such as the Banyamulenge and others, under cover of lockdown.

Local armed groups were active in South Kivu in the 1960s during the Mulelist rebellions (Stearns et al., 2013a). They were then largely dormant until just after the Rwanda genocide which ended in July 1994, when a million Rwandans, and a heavily armed retreating force, fled into then-Zaire. Open warfare intensified from 1996 to 1998, involving regime change, becoming a general war involving regional states, from 1998 to 2003 ‘Africa’s first World War’ (Prunier, 2009). Most recently, warfare again shifted its nature, from 2015 onwards, to take more expressive forms, involving both reprisals ‘…conducted against those who had nothing to do with the prior act’ and uncoordinated acts of revenge, where ‘…the attack is aimed indiscriminately’ at a specific ethnic group, in this case the Banyamulenge (Boyle, 2010: 190).

During the first and second Congo wars (1994–1998 and 1998–2003), Banyamulenge fighters were closely associated with Rwandan military and political intervention in Zaire, and later in DRC. Banyamulenge forces often formed a proxy force for Rwanda. Today, with this alliance long gone, perceptions persist that Banyamulenge are allied to, and loyal to, Rwanda. National and global events have ‘frozen’ the Banyamulenge in this role. On the pretext that they are not loyal Congolese, but intent on subversion, the Banyamulenge have been systematically marginalised and persecuted. The Mission de l’Organisation des Nations Unies pour la Stabilisation en République Démocratique du Congo (MONUSCO) has little chance of maintaining peace, whilst its main partner is the FARDC. MONUSCO soldiers appear not to have noticed the acute humanitarian tragedy unfolding near their base in Minembwe. International humanitarian agencies completely ignore the crisis in South Kivu and fail to respond to human need in Minembwe and other centres where IDPs are concentrated. Instead of a focus on civilians, the main priority of those working for ‘peace’, has been to bring armed factions together for dialogue. Caught in a prolonged and crippling state of siege since March 2019, Banyamulenge are now exposed to the threat of being wiped out.

Maimai commonly define their enemies in terms of ‘race’, whilst international organisations, NGOs and researchers, tend to blame ‘tribal conflict’, assuming that South Kivu and indeed much of Eastern DRC, is divided by ‘age-old’ ethnic divisions. Banyamulenge are viewed as just one party in complex patterns of a many-sided violence among different communities (Ntanyoma, 2019b; Thakur, 2010). Since the colonial era, narrative framings around racial and ethnic violence have intensified, especially with globalisation (Rosenau, 2003: 289). Narratives of ‘race’ and ‘tribe’ tend to obscure how current violence against the Banyamulenge, as both asymmetrical and expressive, takes the form of systematic persecution of a particular identity group. Woven through this article is the quest for some solutions in the face of the slow and seemingly deliberate genocide of a stigmatised minority (Ndahinda, 2013). As Rosenberg points out, we should not ‘ … lose sight of the fact that genocide is a fluid and complex social phenomenon, not a static term’, and not confined to its purely legal definition (Rosenberg, 2012: 17). The slow pace of killings does not preclude the conclusion of a systematic destruction of ‘a people’ taking place daily, and invisible to most external actors, including to most genocide scholars.

The toxic power of mass violence first infested the Congo under Leopold II, when: ‘ … a quasi-genocide … durably traumatized the population’ (Prunier, 2009: 76). In Congo, capitalism relied, not on ‘free’ workers or a ‘free market’, but on forced labour and divide-and-rule extractive policies of violent predation (Nzongola-Ntalaja, 2002). After independence, neo-colonial predation under President Mobutu, who was installed by the West, was followed by long years of war at the turn of the millennium. The extreme cross-generational trauma of Congolese civilians resulting from these historical and more recent mass atrocities leaves scars that can produce recurring cycles of attack and revenge (Maeresera et al., 2018; Mels et al., 2010). As Mobutu’s regime collapsed, the toxic realities of the Rwandan genocide, inspired by racial thinking, spilled over into Zaire, intensifying ‘race’ as a vector of violence and conflict in Eastern Congo end sentence at Congo.

Since 1994, ‘violence of the so-called Congolese conflict … was the product of unsettled questions that the Rwandese genocide had brushed raw’ (Prunier, 2009: xxxii). The so-called ‘Hamitic hypothesis’ is one of the most poisonous legacies of colonialism, disseminated through school textbooks, the clergy, and the administration, both before and after independence, in Congo and the Great Lakes region. As a ‘hypothesis’, this posits racial’ differences between ‘Bantus’ and ‘Hamites’ as at the root of violent conflict in the Great Lakes region and beyond (Eltringham, 2006; Taylor, 2001). This imperial, 19th century notion of ‘race’ remains alive and well in South Kivu today. Globally imposed reforms, from political decentralisation to multi-party elections have tended to reinforce this polarisation between communities, intensifying hatred and violence along identity lines (Rosenau, 2003: 289, 338). Such is the case for the Banyamulenge of South Kivu.

Historically, Banyamulenge fighters were seen as perpetrators of violence rather than victims of violence (Ndahinda, 2013; Ntanyoma, 2019a; Vandeginste, 2015; Weiss and Carayannis, 2004: 133). When the Banyamulenge assisted the Rwandan army in replacing Mobutu with President Laurent Kabila, they were punished by having some of their land dispossessed specifically in Katanga province and South Kivu (Ntanyoma, 2019a: 75; Stearns et al., 2013b: 18–19). These trends have intensified, especially since 2015. ‘Bantu’ communities, feeling themselves victims of Banyamulenge and Rwandan fighters, coalesced into a broad military front. Some have become perpetrators, seemingly dedicated to wiping out the Banyamulenge. One wonders: ‘Is there a layering of identity that can occur when persons are subject to forms of mass trauma and violence, and then participate in such actions themselves?’ (Davey, 2019: 5). If so, then the reverse may also be true: that those once considered archetypal perpetrators, namely the Banyamulenge, can later become victims of atrocities that threaten their very survival.

Together, one of us a Congolese national, concerned with explaining armed violence in Eastern DRC, the other a half-Belgian Westerner specializing in the African Great Lakes region, we analyse why the Banyamulenge have experienced rapid take-over of their land, homes, and cattle, resulting in a state of siege, alongside the massacre of displaced civilians (Ntanyoma, 2019b; Radio France International [RFI], 2019). Some information used was gathered during doctoral fieldwork by the main author when he was in South Kivu between November 2018 and May 2019.

The article first introduces expressive warfare in South Kivu, which involves elements of proxy warfare using both informational weapons and physical violence. The article identifies two common narratives (one on ‘race’, the other on ‘tribe’) used to explain and dismiss the recurring violence against Banyamulenge in South Kivu. Both serve to ‘justify’ attacks on Banyamulenge civilians as self-defence by ‘autochthonous’ Congolese armed groups. The result is that local and international organisations are equally unable to recognise the slow genocide taking place in South Kivu today since, almost by definition, the Banyamulenge are framed as an aggressive, militarised ‘tribe’ or ‘race’ (Ndahinda, 2013: 485).

Elements of Autesserre’s narrative analysis framework help to account for the ways violence by local armed actors is framed, and warnings of genocide (ignored), by international actors. In her study, Autesserre (2012) identified a number of over-simplified accounts related to ‘why there is war’ in DRC, ‘who the bad actors are’ and ‘what interventions are needed’ (Autesserre, 2012; see also Autesserre, 2009). Conventional explanations of warfare centre on mineral resource extraction, and on sexual violence, in what is otherwise largely a forgotten conflict (Autesserre, 2012: 204). The commonly proposed solution is to strengthen the Congolese state, reinforcing the security and justice institutions in particular. The ‘resource curse’ does no doubt affect local, regional and global dynamics of violence in Eastern DRC and South Kivu. Certainly, sexual violence should be punished, and gender dimensions of violence should not be ignored. Victims need to be offered physical and emotional healing and some measure of justice (as recognised by Nobel prize winner Dr Denis Mukwege). Yet, empirically, during the present phase of warfare: ‘There is no systematic correlation between violence and mining areas. Only 20 percent of violent incidents occur within 20 kilometres of a mine, and only 3 percent occur within 2 kilometres’ (KST, 2019b: 9). It is interesting to reflect on the gender aspect of identity politics in the DRC, whereby:

 … the dehumanizing nature of many of the images and stereotypes of women and how they are used to justify violence toward women are very similar to the dynamics used by in-groups to justify violence against out-groups in larger conflicts. (Cottam et al., 2006: 4)

Despite this, the ‘conflict minerals’ discourse remains widely used, meaning that other forms of violence, including land grabbing, selective withdrawal of national citizenship, and promoting hate-based identity politics on social media, are all ignored. Yet these elements ‘outside the frame’ are (the ones that need) more urgent attention.

Expressive violence in warfare in South Kivu

Available data suggest ‘current conditions for the populations of the Eastern Congo remain among the worst in Africa … [and] HDI [the Human Development Index of the UNDP] went down from 167 to 187 …  – lowest in the world between 2006 and 2010’ (Autesserre, 2012: 203). And this situation has likely deteriorated since 2010. This can partly help to explain why the persecution of Banyamulenge people has not been attended to, or taken sufficiently seriously, by local NGOs, humanitarian actors and the international community. The scale of suffering in DRC is world-record breaking. All of Eastern DRC’s communities have experienced repeated violence and war, forced labour and abuse. In this context, the temptation is to look for someone, or something, to blame for these woes. The most convenient target happens to be the largest category of victims of genocide in the past – the Rwandans, and the Tutsi in particular- with whom the Banyamulenge are wrongly identified ‘as if they were the same’.

During the most recent phase of warfare in Eastern DRC, civilians are targeted by forms of ‘expressive’ and ‘asymmetric’ violence, sometimes based on ascribed ethnic, racial or tribal identities. The term ‘expressive violence’ implies ‘ … the discursive, symbolic, ritualistic, and generally non-instrumental character of violence’ (Kalyvas, 2006: 24), which aims to inflict pain and destroy the symbols of the ‘enemy’, focussing on ‘identity and sectarianism’. Although such expressive violence may produce some material benefits, these are not necessarily the sole or even the main motivation, which may be something akin to revenge (Kalyvas, 2006: 24). Instrumentally, expressive violence ‘can be used to exterminate a group or to control it’ even though it is not strategic in the usual sense (Kalyvas, 2006: 26). In South Kivu, we show that there has been a shift from a complex, multi-sided armed conflict, to a slow genocide, and a campaign of attrition, against unpopular minorities. The Banyamulenge are not alone in being framed by many other Congolese as ‘Tutsis from Rwanda’, in spite of the strong historical evidence suggesting they have lived inside the present boundaries of the Congo since at least the early 1800 s (Hiernaux, 1965: 377; Moeller, 1936; Newbury, 2009: 265; Weis, 1959).

It may be useful to view ‘violent non-state actors as one part of the wider and full-spectrum (economic, socio-political, informational) [of] coercive efforts’ (Rauta, 2019: 3). Some features of expressive, hybrid warfare include how it: ‘ … targets the vulnerabilities of a society and system while deliberately exploiting ambiguity to avoid detection’ (Qureshi, 2020: 174). The present stage of expressive violence in South Kivu involves direct physical violence alongside social media hate speech, intelligence and media campaigns by local and national politicians, and targeted physical and symbolic violence of a brutality that communicates a distinct sense of fear and threat to Banyamulenge civilians (Simons and Chifu, 2018).

The situation around the Minembwe-Bijombo High Plateau involves a web of actors, from strictly local leaders and armed groups, to international NGOs and diplomats, the African Union, EU and UN agencies, and regional and central government. In 2015, clashes in Burundi against a Presidential third term coincided with pressure for presidential elections in DRC. After parliamentarians voted to allow Paul Kagame to stay in office, both Rwanda and Burundi started to back proxy armed actors in Eastern DRC. The Banyamulenge felt increasingly abandoned as a lost cause, in the midst of what became the ‘Third Congolese War’ (Battory and Vircoulon, 2019; Larcher, 2018; Muhame, 2019). Other regional countries such as Uganda, and Tanzania are also implicated in backing armed actors in irregular warfare on their behalf, elsewhere in Eastern DRC (Uwiringiyimana, 2019). Some journalists interpreted President Kagame’s 2019 New Year speech as suggesting military and diplomatic collusion between Uganda and Burundi, without naming these two countries, in efforts to overthrow him (Fröhlich, 2019). From late May to November 2019, Rwanda–Uganda relations deteriorated, following alleged shootings at the border (Biryabarema, 2019).

Expressive violence, as described in Kalyvas (2006: 25) is ‘killing for killing’s sake’, and does not differentiate civilians from armed actors. In Bijombo Groupement and Minembwe Rural Municipality, armed attacks sought to wrest control of Bijombo Groupement from the Banyamulenge, first established in 1969. It took two years before the traditional chief of Bijombo Groupement, a Banyamulenge, could take his place (Rukundwa and Republic, 2004; Verweijen and Vlassenroot, 2015: 28). Military and political contestation of a Banyamulenge right to traditional chieftaincy has continued ever since (Muchukiwa, 2006; Muzuri, 1983; Verweijen and Vlassenroot, 2015). The Banyamulenge are the only community in South Kivu who now appear to be entirely ‘on their own’ in their efforts to ensure they have political representation and secure the right to full Congolese citizenship. Some features of hybrid warfare, namely that it: ‘ … targets the vulnerabilities of a society and system while deliberately exploiting. While Kalyvas (2006) warns about the predominance of expressive over other motivations, he argues that ‘(d)educing motive from behavior is a bad idea, as is replacing evidence with politically motivated classifications … ’ and insists that ‘particular act may be consistent with several motives’ (Kalyvas, 2006: 24). Even though the article underscores the warning from Kalyvas, it also stresses that violence targeting Banyamulenge generally displays a sense of expressive violence. Verweijen (2015: 174) underlines that indiscriminately attacking the Banyamulenge increases the popularity of perpetrators (see also Alida, 2017). Whether violence is justified as revenge, peer pressure or collective imaginaries, it builds on racial and physical traits to dehumanize and define who is to be killed (Stearns, 2011: 194; Turner, 2007: 92). For instance, Stearns et al. (2013a: 43) have remarked that:

On 4 October 2011, a vehicle from the humanitarian organization Eben-Ezer was ambushed in Kalungwe village, near Fizi town. After the passengers were separated according to their ethnic origins, seven Banyamulenge were killed, while those of other ethnicities were set free. (See also Verweijen, 2015: 175)

Most local Congolese framings refer to Banyamulenge as ‘Tutsis’, as ‘Rwandaphones’, or as ‘Hamites’. For scholars, NGOs and MONUSCO, such racial terms are translated into a more familiar narrative around the theme of ‘tribal clashes’, in which all sides are said to be equally guilty of murderous and irrational violence (United Nations Joint Human Rights Office [UNJHRO], 2020). Neither the ‘race’, nor the ‘tribal clashes’ narrative acknowledges the distinct vulnerability of Banyamulenge civilians, faced with multiple armed enemies who now completely surround them. Serious attacks on Banyamulenge housing, land and property (HLP), resulting in their internal displacement and dispossession, along with killings of civilians are producing the ‘slow genocide’ of Banyamulenge in South Kivu that we have referred to. Both HLP attacks and massacres of civilians are defined in law as war crimes.

Framing Banyamulenge as those ‘who deserve punishment’, is a form of informational warfare, which by ‘creating alternative false realities’, mainly through media and the internet, obscure the real dynamics of one-sided warfare and expressive violence (Simons and Chifu, 2018). Historically, Banyamulenge were framed as non-native to South Kivu and DRC. What is new, however, is the intensity of attacks on this minority, now displaced by force, and literally besieged since early 2019. At present, no major organisation provides support or protection for this beleaguered community, so powerless have the Banyamulenge become through dispossession and stigmatisation (Ndahinda, 2013; Ntanyoma, 2019b).

After the 1994 Rwanda genocide, Hutu military and para-military fled into Zaire, shielding behind Rwandan Hutu civilians (Hintjens, 2006; Lemarchand, 1998; Nzongola-Ntalaja, 2002). This defeated army regrouped in refugee camps, launching further attacks for several years on Rwandan Tutsi from across the border (Stearns, 2011). RPA soldiers and their proxy allies who, at that time, included Banyamulenge fighters, retaliated against Hutu genocidaires. As defeated FAR (Forces Armées Rwandaises) troops were pushed deeper into (then) Zaire, civilian hostages were forced along. Ngolet, citing the almost mythical Garreton Report of 1996, claims:

 … he [Garreton] noted that in Uvira and other territories conquered by the Banyamulenge, ADFL forces attacked refugee camps, killing and displacing numerous people. Garreton charged that these forces committed grave human rights violations by forcefully expelling the refugees to their country of origin, knowing full well that as Hutus they would face persecution and even death. (Ngolet, 2011: 4)

The RPA did not defeat genocidal Hutu Power armed groups in Eastern DRC, but did kill many thousands, perhaps hundreds of thousands, of Hutu civilians (Emizet, 2000; Olsson and Fors, 2004). Prunier speculates on the RPF’s military modus operandi, for whom fighting in Uganda from the 1980s onwards meant: ‘ … they knew only the gun and the gun had worked well for them in the past … [moreover] … Their self-confidence was strong, their political vision embryonic and they had a limited but efficient bag of tricks to deal with the international community’ (Prunier, 2009: 22). At first, the international community considered RPA intervention in Eastern DRC as the illegal, but understandable pursuit of armed genocidal criminals. Later, however, this changed.

Through 1996–1998, the RPF, and Rwanda’s proxies in Eastern DRC, including the Banyamulenge military, aimed for regime change (Ntanyoma, 2019a). Rwanda’s presence in Eastern Zaire/DRC encouraged other countries from around the wider Great Lakes region to intervene, transforming the second Congolese war (1998–2003) into ‘Africa’s First World War’ (Prunier, 2009; Stearns, 2011; Turner, 2007). Complex alliances emerged around renewed plunder of mineral resources and mass violence, including sexual violence, against civilians, reinforcing hostility towards Tutsi Rwandaphones as presumed outsiders, in both North and South Kivu.