Press "Enter" to skip to content

Expressive violence and the slow genocide of the Banyamulenge of South Kivu


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.


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, 2010Lendman, 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, 2013Zarni 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, 2019bThakur, 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., 2018Mels 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, 2006Taylor, 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, 2013Ntanyoma, 2019aVandeginste, 2015Weiss 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, 1936Newbury, 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, 2019Larcher, 2018Muhame, 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, 2004Verweijen and Vlassenroot, 2015: 28). Military and political contestation of a Banyamulenge right to traditional chieftaincy has continued ever since (Muchukiwa, 2006Muzuri, 1983Verweijen 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, 2013Ntanyoma, 2019b).

After the 1994 Rwanda genocide, Hutu military and para-military fled into Zaire, shielding behind Rwandan Hutu civilians (Hintjens, 2006; Lemarchand, 1998Nzongola-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, 2000Olsson 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, 2009Stearns, 2011Turner, 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.

The Banyamulenge’s position became more difficult as they were caught between Rwanda and the FARDC, specifically following the 1996 rebellions, when the armed group known as Gumino was formed (Stearns et al., 2013bVlassenroot, 2002). Gumino,1 affiliated to the Banyamulenge, split several times, and was unable to defeat Maimai military power on the ground, or defend Banyamulenge civilians effectively (Ntanyoma, 2019a; Stearns, 2011). The last Gumino fighters had reintegrated into the FARDC in 2010 (Ntanyoma, 2019aStearns et al., 2013b).

Meanwhile, in 1999, the MONUC (Mission de l’Organisation des Nations Unies pour le Congo), was created, and a few years later became MONUSCO (with the addition of ‘la Stabilisation’). With 20,000 troops under Chapter VII provisions, and 3000 soldiers from the African region, MONUSCO has struggled to protect Congolese civilians, and seems unaware of the extent of Banyamulenge vulnerability, though they are displaced close to a MONUSCO base (Beljan, 2014Clark, 2011Kets and De Vries, 2014). Despite a huge budget, UN peace-keepers cannot stabilise Eastern Congo or South Kivu (Kuele and Cepik, 2017). We explain how they could do better, perhaps, if they managed to understand the situation on the ground (see Titeca and Fahey, 2016).

Prunier’s study Africa’s First World War: Congo, the Rwandan Genocide and the Making of a Continental Catastrophe (Prunier, 2009; see also Turner, 2007) shows how the Rwanda genocide revived dormant ideas of race in the region. The second Congolese war ended with a peace agreement in Sun City, in 2003, but violent conflict broke out again in 2015, intensifying from 2017 (Autesserre, 2012: 250). Narratives of violence as rooted in racial and tribal identities, now circulate almost unchallenged in South Kivu.

This third long cycle of warfare involves a range of violent methods, including economic warfare against property (land, housing, cattle). The civilian population has been displaced on an unprecedented scale. In the presence of MONUSCO, regular brutal killings of besieged IDPs, illness and death from starvation, are all producing a slow genocide (Kleinfeld, 2019Ntanyoma, 2019b). Whilst MONUSCO and the UN affiliated organizations recognize the complexity and multiple causes of violence around the Banyamulenge (UNJHRO, 2020), their reports tend to view the conflict through the lens of tribal violence to possibly justify the incapacity to intervene (UNSC, 2020). In some cases, the UN reports highlight events selectively, perhaps to avoid questions being raised about MONUSCO’s mandate. In May 2020, the UN Security Council (UNSC) (2020: 4) described the situation in Kamombo and its neighbourhood at the time of the Makanika attack (a Banyamulenge-affiliated armed group) and their occupation of these localities (May 2020), whilst failing to report on earlier violence against Banyamulenge since May 2019 that had led to this occupation that had led to this occupation in the first place.

Similarly, MONUSCO has failed to protect the property (land and cattle mainly) of displaced Banyamulenge, even those guarded by UN forces in Mikenke IDP site. Thousands of cattle were looted in more than 20 attacks that targeted civilians’ property within meters of the MONUSCO base. More particularly, IDPs in Mikenke were attacked four times as Maimai groups entered the IDP site and killed and wounded individuals supposedly under UN protection. Similar incidents have taken place in Minembwe and also in the Bijombo groupement (Uvira territory). There is no working transport infrastructure in these areas, yet in Bijombo, for example, MONUSCO sheltered Babembe, Bafuliro, and Banyindu civilians whilst leaving displaced Banyamulenge unprotected some 15–20 kms away.

Although MONUSCO’s mandate stresses civilian protection, its limited capacity to understand local dynamics and root causes of violence has undermined this mission (Autesserre, 2012). MONUSCO relies on the National Army and Intelligence and seeks to maintain an aura of political neutrality. By failing to protect civilians living a short distance from the MONUSCO bases, the UN troops have failed to prevent Banyamulenge being subjected, for example in Bijombo, to severe hunger, to attacks with impunity and to deliberate impoverishment by surrounding armed forces. In other words, the neutrality of the UN peacekeeping mission in Eastern DRC is highly questionable.

National developments add to this fraught local political situation. Following long-delayed 2018 presidential elections, violence against Banyamulenge intensified instead of diminishing. A UN Panel of Experts expressed the fear that battle-lines were being (re)drawn in ways that could yet lead to another ‘African World War’ (RFI, 2019; UN Group of Experts [UNGoE], 2018a). Fighting was especially intense in South Kivu, bordering Rwanda, Burundi and Tanzania (Battory and Vircoulon, 2019Kleinfeld, 2019), particularly in the High Plateau of Minembwe-Uvira, and in the Mid-Plateau areas of Uvira, Fizi and Mwenga. From 2019 to 2020, this ‘Third Congolese war’ intensified further, and has now become an apparently concerted effort to rid the region, and the DRC, of so-called ‘Rwandaphones’.

Representing the Banyamulenge in South Kivu

One distinctive feature of the DRC Constitution is that civic citizenship is grounded solely in membership of an ethnic community, linked with a specific territory (Jackson, 2007Mamdani, 1999). To be a citizen, any Congolese man or woman must belong to one recognised ethnic group or sub-group, and this group must have its own local chieftaincy area and its own territorial representation (Court, 2013Jackson, 2007). Jackson (2007: 90–93) has extensively documented how, in DRC, nationality is generally located in farming land, and that, in contrast to this, many ‘Rwandophones’, including the Banyamulenge, are historically nomadic pastoralists. In this way, ‘Bantu’ Congolese communities, the Babembe, Bafuliro, Bavira and Banyindu in South Kivu, which historically specialised in farming, have sought to claim uncontested control over entire regions, and a monopoly on the status of full Congolese citizens, with their own chieftaincies and government arrangements.

Minembwe Territoire was established in 1998, during the RCD rebel insurgency, when the then Decentralization Minister was an active member of this rebellion and himself also a member of the Banyamulenge community (Lemarchand, 2013: 432; Stearns et al., 2013bVerweijen and Vlassenroot, 2015Vlassenroot, 2002). The creation of Minembwe Territoire was supposed to rehabilitate the Banyamulenge, ending their historical exclusion from traditional chieftaincies in Congo in place since the colonial era (Muchukiwa, 2006: 134; Rukundwa and Republic, 2004Stearns et al., 2013b: 27; Verweijen and Vlassenroot, 2015: 1). In 2013, a ministerial decree created over a hundred municipalities and cities across Congo under the decentralisation provisions. This decree was implemented in late 2018, just before the general elections. With decentralisation measures from 2018, Minembwe was listed as a Rural Municipality, controlled by Banyamulenge representatives.

The Banyamulenge, it is also claimed, do not have the same entitlement to control government office, even at municipal levels (Bøås and Dunn, 2014Jackson, 2006). When Minembwe Rural Municipality was declared a decentralized entity in 2018, the area was inhabited mainly by Banyamulenge people. Their control over any local government whatsoever has been hugely contested (KST, 2020bNtanyoma, 2019b: 19). Within the region, the Banyamulenge community has remained in control of just one traditional entity, the Bijombo groupement, and this too is disputed.

Members of self-styled ‘native’ ethnic communities argue that Banyamulenge are not entitled to manage any traditional entities, since they are not ‘autochthonous’ to the DRC (Muchukiwa, 2006: 111). Within the Minembwe Rural Municipality, entire localities and villages have been destroyed, and razed to remove evidence of where Banyamulenge lived, mostly peacefully, alongside their South Kivu based neighbouring communities. Ignoring two centuries of history or more, the rhetoric of ‘race’ based national identity has become a way of denying the belonging and citizenship of Banyamulenge as ‘people from nowhere’. Although records corroborate that Banyamulenge individuals managed some local administrative entities during the precolonial and colonial periods (Loons, 1933Reyntjens and Marysse, 1996: 15), with their removal from local government positions, the primordial claim of civic nationality of the Banyamulenge is undermined (Mamdani, 1999). Dragged into wars of claims and counter-claims of membership of the Congolese nation/state, expressive violence has dogged the Banyamulenge since the early 2000s (Stearns et al., 2013bVlassenroot, 2002).

This situation was enough to trigger renewed military mobilization by neighbouring communities, with Maimai challenging the right of Banyamulenge people to control even a single municipality in DRC (Radio Okapi, 2020b). Rural Minembwe ended up being the most hotly contested political seat in the country. In Fizi territory, which was numerically dominated by the Babembe people, Babembe political leaders have contested Banyamulenges’ status as Congolese nationals since at least the 1960s. From the 1980s through to the 2006 general elections, as reported on Radio Okapi, there were demonstrations and protests in Bukavu and Uvira, with support from veteran politicians like Anzuluni Bembe (Radio Okapi, 2006aRukundwa and Republic, 2004: 136; Vlassenroot, 2013: 34). The aim of such protests was simple: to dispute the right of Banyamulenge people to control any part of the DRC government (see also Radio Okapi, 2006b).

Since early 2019, the area around Minembwe has become the epicentre of violent contestation of the right of the Banyamulenge community of South Kivu to be represented politically, and hence to be Congolese citizens. In 2018, the Decentralization Minister, a Banyamulenge, was accused of manipulation, even though the Decentralisation Act was signed in 2013, before he became Minister (Mfundu, 2020Radio Okapi, 2020b).

Challenges to the Banyamulenge as ‘not Congolese’ are not new; in 1995, the Vangu Mambweni Commission under President Mobutu demanded that those of ‘Rwandan descent’ be expelled from Zaire by the end of the year. The Commission alleged Banyamulenge were sent by Rwanda to enslave autochthonous Congolese people and steal their land (Emizet, 2000Huening, 2013Prunier, 2009). These massacres placed Banyamulenge people in a very vulnerable position, convincing Banyamulenge fighters to join the RPA (Rwandan Patriotic Army) and fight alongside Rwandan troops. In the longer term, the price of this decision was that the Banyamulenge as a whole, and not just armed soldiers, would come to be wrongly associated with the regime in Kigali, as ‘invaders’ and ‘outsiders’, seeking to dominate those who claim to be autochthonous Congolese citizens. All this reveals how narratives of race have been used – and continue to be used – to reinforce divisions among the Congolese population, notably around election times.

Despite their mostly good relations with neighbouring communities in South Kivu, during the pre-war era, the Congolese citizenship of the Banyamulenge, and other ‘Tutsi’ or ‘Rwandan’ people has constantly been brought into question, in local social media and by local politicians from neighbouring ‘Bantu’ communities (Huening, 2015). Maimai groups affiliated to Babembe, Bafuliro, Banyindu, and Bavira communities consider themselves the only ‘autochthonous’ Congolese, openly challenging the right of their neighbours, the Banyamulenge, to remain in the Congo (Geschiere and Jackson, 2006Mathys, 2017Sanders, 1969Verweijen, 2015). Within nativist and also ‘liberationist’ Congolese political discourse, Banyamulenge are an enemy within, a sort of fifth column of Rwanda, whatever their actual affinities and alliances today. Nzongola-Ntalaja (2002) exemplifies this view on the Left of Banyamulenge soldiers as Rwandan proxies. The Banyamulenge community did not take part in the Sovereign National Conference of 1990–1992 (Conférence Nationale Souveraine) for this reason, since it vocalised hostility to their role under Mobutu, as expressed in the report of the Vangu Mambweni Commission, referred to earlier.

In these ways, citizenship questions help explain the weakened position of the Banyamulenge in recent decades (Court, 2013Jackson, 2007). Both proximate and underlying causes of recurrent cycles of irregular warfare in Eastern DRC, and in South Kivu, are complex and historically specific.

Within the race narrative, which we consider in the next section, there is little room to acknowledge the present and growing vulnerability of Banyamulenge civilians to slow attrition through hunger, lack of proper shelter, medical care and security. All local communities in South Kivu can, with good reason, claim to have been victimized in the past, given patterns of warfare since 1994. Widespread grievances mean that armed groups and civilians tend to disregard the suffering, deaths and displacement of civilians from other communities, even sometimes from their own (Stearns, 2011). Massacres and destruction of property continue to fuel resentments at past atrocities, and all communities have reason to feel invaded, within the wider security void of South Kivu. What makes this dangerous for the Banyamulenge in particular, it is now suggested, are the narratives of ‘race’ and ‘tribe’.

‘Race’ politics in South Kivu

Local grievances among ‘Bantu’ in South Kivu centre on a supposed struggle between ‘autochthonous’ Congolese and Rwandan ‘immigrants’ (Bøås and Dunn, 2014Jackson, 2006). Fighting over the right of representation in local government is, for the Banyamulenge, a matter of community survival, and the foundation of a collective claim to citizenship, as explained. Losing both farming land and cattle to armed Maimai groups thus spells a significant loss of entitlement, and is proving life-threatening for Banyamulenge civilians collectively. Political representation as the basis for citizenship as an authentic Congolese ‘ethnic’ group matters as much as land, cattle and other material essentials for sustaining life. Indeed, political membership of a territorially embedded community remains the only possible basis for a claim to citizenship, land and resources in DRC, even today (Huening, 2013Jackson, 2006Ntanyoma, 2019a).

Far from aligning themselves with Rwanda, the Banyamulenge of South Kivu have, for some time, sought to distance themselves from the neighbouring country. Ten or twenty years ago, the Babembe, Bafuliro, Banyindu, and Bavira were indeed victims of ‘invading Banyarwanda’ armed groups. Whilst this portrayal may be credible for the late 1990s and early 2000s, today the immediate threat has passed. Despite this shift, perceptions of ‘race’ identities remain fixed in the mould of those years. Now it is autochthonous ‘Bantu’ who take up arms and threaten the very survival of the Banyamulenge people. Yet perceptions of threat and danger have not caught up with changing realities; all Kinyarwanda speakers tend to be targeted and lumped together as the ‘Trojan Horse’ of Rwanda (Nzongola-Ntalaja, 2002Turner, 2007). Such stigmatising stereotypes are popularised through ‘hate speech’ on radio, in social media and in public political pronouncements (Genocide Watch and Ntanyoma, 2021Huening, 2013Weiss and Carayannis, 2004: 123).

The terms ‘Bantu’ and ‘Hamite’ originated in 19th century imperial race science. The colonial ideology of a racial hierarchy, first introduced by European educators and administrators popularised the Hamitic hypothesis in the Great Lakes region, especially in those regions under Belgian control. As Sanders noted back in 1969, referring to the term, Hamitic: ‘The word still exists, endowed with a mythical meaning; it endures through time and history, and, like a chameleon, changes its colour to reflect the changing light. As the word became flesh, it engendered many problems of scholarship’ (Sanders, 1969: 531). In the DRC it has changed its colour to become a way to justify a slow genocide through attrition.

The Hamitic hypothesis remains embedded in mutual distrust between ‘Bantu’ and ‘Hamitic’ peoples, defined in a classificatory system that essentialises ‘racial’ differences. According to this historical narrative dating from John Hanning Speke, all Rwandans or Tutsis in the Great Lakes region were late-comers, invaders who had usurped Hutu kingdoms, imposing their own ‘alien’ rule over Hutu peasants. The humiliations of overlordship and feudal subservience created a sense of humiliation among ‘Bantu peoples’ that ‘nurtures a profound thirst for redress and vengeance on the part of the [allegedly] defavourized group’ (Taylor, 2001: 57). Eurocentric and profoundly racist, the Hamitic hypothesis is still widely expressed in the popular political culture of South Kivu and even elsewhere in the African Great Lakes region (Hintjens, 2001). In this sense, in South Kivu, it can be reasoned that: ‘the process of decolonization has not been completed … [and] perhaps … we witness the lingering effects of colonialism in the form of the Hamitic hypothesis’ (Taylor, 2001: 55).

Regional leaders, journalists and priests, Maimai militia leaders, and politicians, have all labelled the Banyamulenge as ‘invaders’ in South Kivu, or worse (Turner, 2007: 92). Whilst their neighbours, the Babembe, Bafuliro, Banyindu and Bavira view themselves as ‘autochthonous’, as full DRC citizens, this is not true of how Banyamulenge are viewed (Court, 2013Ntanyoma, 2019aStearns, 2011). The Hamitic hypothesis of ‘Tutsi’ invaders creates a racial stereotype very difficult to contest and deconstruct, with dramatic consequences for Banyamulenge civilians now confined to tiny areas in Minembwe, besieged and attacked on a regular basis (KST, 2019aNtanyoma, 2019bRFI, 2019).

Other, relatively neglected framings around root causes of warfare in South Kivu, related to land, citizenship and identity politics are drowned out by dominant narratives around ‘racial’ identities. During a public meeting in Kinshasa, the 2018 presidential candidate Martin Fayulu claimed that Minembwe rural municipality was a ruse to annex part of Congo on behalf of Rwanda.2 In Baraka, in South Kivu, on 17 January 2020, a young militant from the Congolese Lamuka opposition coalition, whipped the crowd up into a hate-filled frenzy, giving Banyamulenge just 48 hours to leave the DRC. He said those unwilling to leave voluntarily should be forced out (KST, 2020b), adding that local people who assisted ‘Rwandophones’ would be considered enemies. One self-proclaimed General, William Amuri Yakutumba, gained local political support through being openly hostile to Congolese of ‘Rwandan descent’, claiming former President Joseph Kabila was hostage to ‘foreign powers’ who ‘oppressed’ ‘native’ Congolese. In a similar vein, Christian Sita Mampuya, representative of the Congolese community in South Africa, suggests the Banyamulenge are not Congolese, but refugees from Rwanda who seek to take over other communities’ lands (South African Broadcasting Corporation [SABC], 2019).3 Similar narratives characterize Congolese social media, where violence by autochthonous Congolese against Rwandaphone newcomers or those long established in DRC, is justified by othering those attacked.

When the remnants of the genocidal army of the former Rwandan Republic fled into Congo, they were dominated by Hutu Power ideology. They made local allies among the Congolese communities who were similarly hostile to anyone defined as Tutsi. Through associating with Maimai, the FDLR (Front Démocratique pour la Libération du Rwanda), ALIR and other armed groups perpetuated the Hutu Power ideology, to justify the slaughter of innocent civilians, including Banyamulenge. Today the ideology of essentialised racial differences is used to justify land grabbing, and the looting of cattle. The logic is substantially the same as that used during the genocide, a colonial ideology of ‘Bantu’ as ‘autochthonous’, and ‘Hamites’ as recent, pastoral immigrants.

Meanwhile, a second narrative parallels the first, and refers to ‘tribal’ or ‘ethnic’ roots of violence in the Kivus and across Eastern DRC. This narrative is common in international media circles, in scholarly analyses and NGO accounts of warfare in DRC. The Hamitic hypothesis is implicit rather than explicit in such outsider narratives. From this perspective, ‘autochthon tribes [are] … innocent lambs, savagely attacked by evil “Banyarwanda predators” over the course of the past two decades or more’ (Prunier, 2009: 53).

At best, all sides in the conflicts, all ‘tribes’, are viewed as equally guilty of armed aggression, and civilians are being killed in cycles of retaliation for earlier civilian deaths (Prunier, 2009: 53). Unfortunately, Prunier also notes, this kind of tribal narrative is shared by both ‘Kivu civil society and its European NGO allies’ (Prunier, 2009: 53). Since 2017, hostility between Rwanda and Burundi has increased, and the complexity of cross-cutting ‘tribal’ or ‘race’ alliances has intensified, becoming close to untraceable. Currently the Maimai’s ideology is closely connected with the narrative of fighting ‘foreign invaders’, collapsing Rwanda with former local proxies, the Banyamulenge military, even though they are no longer allied (Verweijen, 2015). In this vein, Jackson (2007: 488) suggests that:

Most ordinary Congolese still hold the Banyamulenge to blame for the great suffering that the country has undergone since 1998, seeing them as a ‘Trojan Horse’ for Rwandan rapine and irredentism against the national territory, contributing to the enormous present polarization around the ‘nationality question’.

In the Maimai narrative, the struggle against Banyamulenge is framed to sound similar to other liberation struggles to resist land grabbing and invasion by dominant outsiders (Eltringham, 2006Hintjens, 2006). Dressed in anti-colonial clothes, inherited from the national liberation language of the early Simba uprisings at the time of independence, like that uprising, now too, Maimai have ended up targeting their Banyamulenge neighbours, and victimising their fellow Congolese (Ntanyoma, 2019a: 35). ‘Race’ and ‘liberation’ narratives can produce a toxic, deceptive mix which outsiders may confuse with ‘tribal’ war.

The ‘First African World War’ of 1998 to 2003, pitted local and regional actors against one another, and placed them in alliances on an extended range of battlefronts. The DRC’s neighbouring countries, notably Rwanda and Uganda, but also Angola and Zimbabwe, played critical direct and indirect military roles in the war. Vogel and Stearns (2018) and Ntanyoma (2014) argue that, by arming and training insurgencies like the CNDP or M23 during the First African World War, regional states, especially Rwanda, became deeply enmeshed in the politics of violence through proxies in South Kivu. The roots of expressive violence in this present phase of warfare lie partly in that crucial period, and in the memories of past atrocities against civilians.

Not wanting to hear or see: Slow genocide

International media and press coverage of this third period of warfare in South Kivu has been minimal, despite the alarming scale of violence, destruction, displacement and death. The UN and FARDC seem unable or unwilling to protect civilians (UNGoE, 2018a2019). The KST is the only agency consistently collecting data on violent attacks and displacement, since: ‘ … the UN Office for Humanitarian Coordination (OCHA) suspended the publication of national totals of internally displaced people … due to pressure from the government’ (KST, 2019b: 5). KST estimated that between 1 June 2017 to 26 June 2019, there were ‘6,555 victims of violent deaths, mass rapes, abductions and kidnappings for ransom, as well as destruction of property and political repression’ (KST, 2019b: 6). As also noted:

In general, armed violence is concentrated in the densely populated highlands in the far east of the Kivu provinces, close to the Ugandan, Rwandan, and Burundian borders. One can draw a line between Eringeti town in the far north of North Kivu and Baraka in the far south of South Kivu that encompasses a large majority of all violence in the two provinces’. (KST, 2019b: 6)

One major concentration of armed attacks and civilian casualties is in the border areas around North and South Kivu, where ‘the Congolese army has not secured the most important thoroughfares, posing a serious obstacle for trade and travel in the region’ (KST, 2019b: 8). This makes key roads in South Kivu insecure (where they are passable). The COVID-19 lockdown cut links with the outside world. Although in the past: ‘ … many prefer to travel by airplane, boat, or not at all’ (KST, 2019b: 8), lockdown ensured that Banyamulenge were stuck in their small enclaves as the COVID-19 pandemic spreads. There is no option to leave for most Banyamulenge civilians trapped in Minembwe. There is no safely accessible airport.

KST suggests a major reason warfare in South Kivu today is not adequately researched, reported on or in the consciousness of policy makers, is a lack of adequate and timely data. They explain:

Until recently, the largest such datasets – Uppsala Conflict Data Program/Peace Research Institute Oslo (UCD/PRIO), Correlates of War (CoW), and Armed Conflict Location and Events Dataset (ACLED) – mainly sourced violence from local and international media reports. This created a problem of underreporting violence, as much of the violence happens a long way from these correspondents and may not be deemed newsworthy … In addition, these media do not have the means to triangulate or confirm violence, and are therefore more prone to misreporting incidents. The United Nations … sources its data largely from governments. (KST, 2019b: 4)

Implicitly, some deaths are more newsworthy than others. Maimai militias claim Banyamulenge are ‘Rwandan-backed foreigners’, traitors in their midst (Jackson, 2006: 104; UNGoE, 2018b: 10). Violations of housing, land and property rights (HLP), especially significant in South Kivu today, are rarely reported or mentioned by researchers. Mass displacement of hundreds of villages forced some 150,000 Banyamulenge to flee to a few tiny areas. Mass displacement has been accompanied by killings, torture and attacks on the displaced, and tens of thousands of cattle looted, all in the past two years, often with the aim of impressive cruelty (Simons and Chifu, 2018). Health and school facilities have been decimated. Yet no humanitarian agency is helping the IDP Banyamulenge population in South Kivu (Bashi, 2019). This is not a two-sided ‘tribal’ conflict; it has become one minority’s struggle for survival.

The massive violations of HLP rights should be a central concern. As Leckie notes: ‘ … arbitrary acquisition of housing, land, and property through land-grabbing is commonplace in post-conflict environments and requires the attention of UN peace operations’ (Leckie, 2009: 9). HLP rights violations should be a focus since they lead to great suffering, hunger and even death. As with other stigmatised minorities worldwide, as Huggins explains:

The confiscation or outright destruction of housing, land, and property titles; local housing and property cadastres; property registries; and other official records giving proof of ownership, occupancy, tenancy, and other residential rights accompanies most conflicts and complicates the implementation of restitution and other remedial measures. (Huggins, cited in Leckie, 2009: 9)

Huggins also notes that HLP rights were not included, for example, in the Arusha Peace Accords signed just before the Rwanda genocide of Tutsi in 1994, in retrospect a major oversight (Huggins, 2009: 204).

From 2015, Burundian groups supported by Kigali crossed the DRC border and established bases in Kiryama. Besides the Front National de Liberation (FNL) of the self-proclaimed General Nzabampema Aloys (Levine and Nagar, 2015Nkurunziza and Anyadike, 2016). This armed group is linked to Burundi’s 2015 electoral protests (Verweijen, 2015). Attacks on civilians intensified, in 2017 to 2018 in the region of Bijombo and later in 2019 around Minembwe. The second Burundian group is Forces Républicaines du Burundi (FOREBU) linked to General Niyombare Godefroid. It first occupied this region around October 2016. FOREBU was later renamed Forces Populaires du Burundi (FPB) (UNGoE, 2018a). Meanwhile, Résistance pour un Etat de Droit au Burundi (Red-Tabara), has been based in Kabembwe, in Fizi territory near Burundian refugee camps in Lusenda. Clashes between these groups have been reported, for example between Red-Tabara and the FPB, probably related to personal power struggles rather than ideological differences. The key problem is that several Burundian groups have allied with various local Maimai militias as proxies, whose only point of agreement is their hostility to the Banyamulenge in the region. Maimai and Biloze-Bishambuke (akin to Maimai) armed groups set the local agenda, even whilst they act as proxies for Burundian rebels (Radio Okapi, 2019). The Maimai’s justificatory tropes remain very simple; irreconcilable ‘racial’ differences with those they term ‘Rwandaphones’.

In reaction to this, local Banyamulenge self-defence groups, known as Twirwaneho-Gumino, have been revived. Twirwaneho means something like ‘Let us fend for ourselves’. According to Verweijen and Brabant (2017), the Banyamulenge’s Twirwaneho can be equated with Maimai groups, since their primary mission has been to protect property. However, the similarities end there. Twirwaneho self-defence groups were organised only after Gumino collapsed, by local Banyamulenge youth who feared that, since they were now surrounded, civilians would continue to be attacked (Verweijen and Brabant, 2017: 18).

With the balance of power almost entirely with the Maimai and against Twirwaneho-Gumino, cattle theft, land seizures and destruction of housing mean that Twirwaneho-Gumino have been completely unable to protect Banyamulenge property. They cannot therefore be equated with Maimai after all. Their aim is now to prevent genocide of the civilian IDPs in Minembwe and Bijombo. Hundreds of villages have been obliterated in just two to three years, and rural settlements formerly inhabited by Banyamulenge people now lie devastated and in ruins (RFI, 2019). Houses were burned, but also demolished, just to ensure the ‘invaders’ would not return. Given this, to equate Maimai and Twirwaneho seems disingenuous.

Recent sources point to on-going crossings of Rwandan Defence Forces (RDF) at the DRC border to fight Rwandan opposition rebels on Congolese soil (BBC, 2015; KST, 2019c; Rolley, 2020a2020b). Burundian or Rwandan involvement is reported in most recent clashes, sometimes reportedly involving troops from both countries (Nichols, 2016UNGoE, 20162018a). As the governments of these two countries use ‘proxy war’ partners in the DRC, they seek to avoid direct confrontations inside Rwanda or Burundi (Vogel and Stearns, 2018).

Self-defence mechanisms like the Twirwaneho or others such as the Maimai, rebound and ultimately produce negative effects in the form of retaliations for past attacks, on innocent civilians of all sides. Whenever one group, such as the Banyamulenge, feel threatened and fear armed attacks, they organize their own self-defence forces to curb the threat, and the existence of this armed group is then taken to justify other armed actors in the region taking precautionary measures, in the form of attacks on civilians. The complexity of violence enters the stage of criminalization and reproduces collective victimization based on identity (Kalyvas, 2006Stearns, 2011). Such preventive attacks by all sides end up harming mainly civilians. One recent study of the situation in the Ruzizi Plain, in the Mutarule-Bwegera region, concluded that the violence continued because: ‘The fragile and militarized socio-political context, where everyone relies heavily on his militia … [and so] long as the state cannot guarantee individuals’ security and that of their properties, different actors can interfere within the fragile environment to reach their hidden agenda’ (Ntanyoma, 2014).

Besides the retaliation and forms of criminalization of violence between armed groups affiliated to ethnic communities, the partiality of key political and civil society actors in DRC tends to portray Banyamulenge affiliated groups as stronger and more cruel than others. For instance, in recent attacks and counterattacks, Banyamulenge affiliated groups were accused of having launched an attack in Kipupu locality (South Kivu) that allegedly killed 220 civilians. Some actors claimed being eyewitness including parliamentary members who confirmed to media to having counted the dead bodies (Mwamba, 2020a, 2020b; Radio Okapi, 2020aRFI, 2020). Specifically, in such tragic incidents, the Banyamulenge are mostly situated in a ‘no-man’s land’ between Rwanda and DRC (Maswana, 2020). The accusations of hundreds killed are still referred to by political actors, social and mainstream media. However, it turned out after investigation pushed by the UN peacekeeping mission in DRC that around 15 people were killed (Basimike, 2020UNSC, 2020: 4). In the case of Kipupu, UNJHRO (2020: 10) remarked that ‘[s]everal local sources indicated that the civilian population had reportedly left the village of Kipupu during the month of March 2020’. While each death in this context adds prejudice to the existing fragile socio-cultural and security context, mobilization across the country in relation to the Kipupu incident might indicate how violent incidents that involve Banyamulenge affiliated self-defence or armed groups can be manipulated for the interest of harming this minority group. As we have tried to show throughout this article, the criminalization of armed violence can blur the specific vulnerability of any minority targeted for slow genocide, in this case the Banyamulenge.

The international media have paid more attention to protests and electoral processes in the DRC than to the selective impacts of this third major phase of warfare (Battory and Vircoulon, 2019Kleinfeld, 2019). In a context of volatile violent conflict, the mechanisms that aim to prevent threats from other armed groups may reproduce the threats and even intensify the risks of attacks on civilians (Stearns, 2011: 107). This is understandable when one considers the everyday security vacuum in South Kivu and Eastern DRC more generally.


For international researchers or policy-makers to comprehend or operate within the local context, they would need to acknowledge that: ‘[f]rames shape our views on what counts as a problem (for example, the illegal exploitation of mining resources) and what does not (for instance, land conflicts)’ (Autesserre, 2012: 206). For the media, perennial pressures: ‘ … to find a story that fits in a few pages, or can be told in a few minutes, and that their audience can easily understand and remember’, obscure insights into specific local dynamics that generate expressive violence aimed at particular ethnically or racially defined minorities (Autesserre, 2012: 207). Beyond over-simplified narratives told by local armed actors, politicians, NGOs, scholars and policy-makers, in South Kivu time-lags in information and policy design mean policies designed in one phase of the war are applied later, in a changed situation. This produces perverse consequences. For example:

 … labelling the Congo a ‘post-conflict’ situation instead of a ‘war’ made a specific set of policies and procedures, such as elections seem natural and appropriate while it made another set of strategies (such as work on local conflicts) seem inappropriate and illegitimate. (Autesserre, 2009: 254)

As armed groups proliferate, each seeks allies with actors both within and external to the region, and DRC’s security vacuum extends the battlefield into every home, village, town and business. Destroying people’s livelihoods has become a weapon of choice in the war of all against everyone. With the social sphere and economy criminalised, violence remains intrinsic to the primary ‘mode of production’ and extraction in DRC (Stearns et al., 2013cVerweijen and Brabant, 2017). An expanding security vacuum sucks in neighbouring states, their opposition movements, allied to local armed actors who continue settling scores through proxy warfare, attacking innocent civilians in the process, for example, the Banyamulenge in South Kivu.

Sadly, besides the Banyamulenge themselves and their close friends and supporters, almost no major agencies seem to be alarmed at the intensity of this unpopular minority’s suffering. It is vital that psychic damage to civilians through direct and indirect means – from destruction of housing, property and looting of cattle, to targeted and terror-inspiring attacks on vulnerable civilians, should be addressed and punished. Property, including cattle, should be returned, through legal means. Roads and infrastructure should be constructed to ensure that once the political commitment is found, humanitarian relief can actually reach the Banyamulenge IDPs.

Racialised narratives need to be challenged and replaced with narratives of wider belonging and more inclusive Congolese citizenship. FARDC and the Congolese state should provide economic, social and physical security for all civilians in South Kivu, not just some. Instead, we see armed forces fomenting insecurity whilst seeking their own material and instrumental benefit (Baaz and Verweijen, 2013: 575; Vlassenroot and Raeymaekers, 2009).

Although MONUSCO operations are gradually being scaled down, the UN troops have not left the DRC. Yet so long as their presence depends on cooperation with FARDC, there is little chance of their being in a position to maintain peace or citizens’ security in South Kivu or Minembwe. Humanitarian tragedies took place even near MONUSCO’s base in Minembwe. Trapped in a prolonged, crippling state of siege since March 2019, Banyamulenge IDPs are threatened daily, and are slowly but surely being destroyed as a people, the core of any definition of genocide. As for other peoples similarly maligned, like the Rohingya, the Banyamulenge IDPs cannot presently return home or recover their lands, livestock, homes and livelihoods (Zarni and Cowley, 2014). Their future as Congolese citizens looks bleak at present. They are becoming ‘citizens of nowhere’. The weapons par excellence of slow genocide include deliberate, targeted starvation and widespread hate speech (Genocide Watch and Ntanyoma, 2021). Both need to be addressed and taken more seriously in identifying cases of slow genocide. To imagine a future of peaceful and secure coexistence, the present crisis should first be seen for what it is: historic grievances and justifiable resentments of many Congolese citizens, projected onto the Banyamulenge, making them now one of Eastern DRC’s most vulnerable minorities.

This past year, Black Lives Matter protests have reverberated globally, and their significance can be felt also in DRC, including in South Kivu. Black Congolese lives should matter, whether one is Bafuliro, Babembe or Banyamulenge. No Congolese civilians should be regarded as ‘enemies within’, as foreigners in their own land, or as expendable. Recent confrontations, involving myriad rebels from across international borders, and local armed groups, has involved expressive violence, taking the form of slow, unrelenting genocide. The legacy of an especially violent colonial rupture is that outmoded theories of ‘race’ continue to be powerfully instrumentalised by DRC politicians. Even if attention were paid to the slow genocide taking place in South Kivu, active humanitarian relief and protection of civilians will not be enough to eradicate deeply held beliefs about ‘Bantu’ and ‘Hamitic’ racial identities as incompatible. Replacing race narratives with narratives of a shared, inclusive Congolese identity may be difficult. The priority may be to move towards a political settlement that at least protects all civilians’ safety, valuing their lives. So long as the community’s precarious situation goes unacknowledged, the expropriation of Banyamulenge land, housing, cattle and political representation will continue, and their slow genocide will drag on. To forge pathways towards the peaceful coexistence of all who live in South Kivu, all Congolese lives would need to matter. Indeed, all Black lives will need to matter in DRC for any kind of lasting peace. In conclusion, what is needed most in future is not a return to the South Kivu ‘normality’, but new ways of radically rethinking inter-community relations and moving from dynamics of violence to a wider and more inclusive sense of national belonging.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Be First to Comment

Leave a Reply

Your email address will not be published. Required fields are marked *