In the introduction to an acclaimed article on the silences and misunderstandings of Rwandan historiography, Catharine and David Newbury argue that to understand the cataclysmic event of the genocide in Rwanda in 1994, one should not look at the past through the lens of the genocide, as is often done, but rather analyze the genocide through an understanding of Rwanda’s history.Footnote 1 Unfortunately, although there is excellent research from eminent scholars with long careers in the region who all have placed the genocide and violence of 1994 in a longer historical perspective, actual historical research based on primary sources on the period before the genocide has become increasingly scarce since 1994.Footnote 2 While most contemporary studies do refer to this more historical literature, for many younger scholars, the history of Rwanda starts in 1959.Footnote 3 If the colonial period is discussed, it is most often to address the debate on how ‘Hutu’ and ‘Tutsi’ became distinct groups. Within Rwanda, history is a subject fraught with tension, and the internal historiography of Rwanda is closely scrutinized by the Rwandan government.Footnote 4
Similar tendencies can be observed in the eastern Congo. The historiography of the eastern Congo was relatively rich in the years before the wars of the 1990s, and excellent work was produced by Congolese historians. Today, however, researchers (both insiders and outsiders) predominantly focus on contemporary issues, or on events since the 1990s.Footnote 5 In many – but not all – cases this leads to history being used as a backdrop in the introduction, rather than an analysis of how historical processes continue to have an impact today.
This article argues that scholars and analysts should pay more attention to the way the past is connected to the present within Rwanda and eastern Congo – not only because not doing so can lead to the ‘foreshortening of history’, to borrow a term from Richard Reid, but also because we cannot understand the processes propelling violence fully if we study them within a limited timeframe.Footnote 6 While longer timeframes have been adopted with regards to some of the historical narratives that propped up ethnic violence in the region – for example on the Hamitic hypothesis – this article shows the intertwining of past and present goes far beyond these topics.Footnote 7
It is not a new insight to suggest that representations of the past are often informed by a certain form of presentism, and that they reflect contemporary needs and concerns rather than the past they are supposed to represent.Footnote 8 Yet it is relevant to analyse what this means in a context where historical narratives are regularly used to exclude certain groups of people, and where memory and history have often been forged in the context of traumatic events or violent conflict. In contexts such as Rwanda, where strictly defined historical narratives are a way of dealing with a gruesome past, and renderings of the past have become an important part of the ideological underpinnings of the regime, ‘memory’ and ‘history’ have become a battleground on which state-sanctioned historical narratives and memories clash with ‘counter-memories’.Footnote 9
This discursive battle impacts on the kind of histories that can be written about the region. The politicized nature of historical narratives, and the profound infusion of present-day concerns into historical narratives make the use of oral history difficult. This is not only a concern for ‘professional’ academics, but also for international interveners and local practitioners who frequently include historical information on local communities in reports and recommendations. Often these ‘histories’ are based on scanty documentation and on interviews and focus groups, whose information is taken at face value, thus reproducing a politicized social present. This article claims that the solution is more, rather than less, historical research, and that research in general needs to pay more attention to the historical construction not only of identities, but also of territories.
This article, then, seeks to unravel how the multiple layers of history, memory, and recurring waves of violence became condensed into (ethnicized) mental maps guiding interpretations of both the past and the social present, not only among Congolese, but also among outsiders looking in. The focus is on the relationship between ‘mythico-histories’ pertaining to territory, belonging, and territorial expansion, and their role in the construction of ‘autochthony’ and ‘victimhood’ (for Congo) and for state-building processes (for Rwanda).Footnote 10
While discourse in itself is not sufficient to explain violence, the forms and intensity of violence are deeply affected by the discourse – the mental maps – of the actors. Violence cannot be understood without understanding the discursive concepts that justify it for the actors.Footnote 11 In Kivu, such discourses, based on slippery ethnic categories, are particularly pronounced – and directly associated with violence.
A BRIEF HISTORY OF THE CONFLICT
Before the Rwandan genocide in 1994, violent conflicts in North and South Kivu between different groups remained limited mainly to the local level. It is only with the arrival of Rwandan Hutu refugees in Congo that local conflicts took on regional dimensions. From 1996 the area witnessed five episodes of acutely destructive and disruptive violence associated directly or indirectly with external Rwandan support. The first was a direct incursion as Rwanda sought to close the refugee camps that harboured over one million exiles who had fled the country in the aftermath of the genocide in 1994. From late October 1996 Rwandan troops invaded Congo and attacked the refugee camps. Their stated targets were Interahamwe militia and politicians responsible for the planning and execution of the genocide, who had started to launch attacks on Rwanda from the camps in Congo. However, among these Rwandan refugees were also many innocent men, women, and children, who were forced to flee further into Congo and/or killed during these attacks. In addition to the Rwandan refugees, millions of Congolese were also affected by this invasion – driven from their homes, humiliated in a number of ways, and many killed.
In early 1997, the Rwandan government helped consolidate several Congolese groups opposed to Mobutu into one fighting organization, forming the Alliance des Forces Démocratiques pour la Libération du Zaire (AFDL), in the guise of a ‘local rebellion’. After a long march across Congo, the Rwandan Patriotic Army (RPA) – the army of the Rwandan Patriotic Front (RPF), the dominant force in the post-genocide Rwandan government – and their Congolese allies overthrew Mobutu in May 1997. This episode became known as the First Congo War.
A year later, partly responding to the intense hatred within Congo towards the Rwandans, Kabila tried to expel the latter, provoking a third Rwandan incursion from August 1998. This marked the beginning of the Second Congo War, at one point involving no less than nine African countries. Part of that invasion included the establishment of another ‘cover-group’, the Rassemblement Congolais pour la Démocratie (RCD) and, later still, the Congrès National pour la Défense des Peuples (CNDP) and the Mouvement du 23 Mars (M23). While each of these groups nurtured their own local grievances, they were also armed and/or supported in different degrees by the government of Rwanda. Consequently, they were seen in Congo as proxies for Rwandan incursion – and as a cover for the extraction of mineral and other resources that followed.Footnote 12
As a result, the discourses of violence in Congo became deeply ethnicized. The complication was that the several groups of Kinyarwanda-speakers long resident within Congo – often divided, even antagonistic, among themselves – were seen by many Congolese as Rwandan fifth columns. For many Congolese, the cultural identity of Kinyarwanda-speakers became to be seen exclusively through a political lens, and hence these groups were perceived as inevitably affiliated with the Kigali regime. In response to the multiple external incursions against Congo from the east, the ethnic criteria of violence thus also became directed against all Congolese Kinyarwanda-speakers – some of whom had lived for centuries west of the Virunga Mountains and Lake Kivu (the current political boundary) and had never been included within the political domain of the Rwandan central court or the Rwandan colonial administration.
AUTOCHTHONY AND MYTHICO-HISTORIES
The Kivu region of the eastern Congo has been home for centuries to dozens of groups defined as ‘ethnic’. Some of these speak Kinyarwanda (the language spoken in Rwanda), hence are referred to as ‘Rwandophones’ or ‘Kinyarwanda-speakers’.Footnote 13 They share variants of a common broader language unit with Rwandophones elsewhere in the region. Today this language group includes members who identify as (or are identified as) Hutu, Tutsi, or Twa. The existence of a group across the current border also speaking Kinyarwanda is one of the reasons why this group is often labelled as ‘non-autochthonous’ by other Congolese. In a move to stress their autochthony, and to emphasize the difference with ‘Rwanda’, Congolese Kinyarwanda-speaking communities use toponyms to refer to their community, such as Banyabwisha, Banyamasisi, Banyamulenge, etc.Footnote 14
The idea of ‘autochthony’ refers to a presupposed ‘natural’ relationship between ‘people’ and the ‘soil’, and thus implies claims to territory. In Africa, it is often directly related to ideas about national citizenship.Footnote 15 In Kivu, discourses on ‘autochthony’ are crucial to understanding modes of mobilization that include or exclude people from certain rights or access to certain resources. There, ‘proving’ autochthonous status is often pivotal to people's livelihoods, and has implications for getting access to land, or finding a job.Footnote 16 Being considered as autochthonous has in the past also been important to claim legal citizenship. During moments of increased tension, being considered non-autochthonous can lead, and has led, to persecution and violence.
Autochthony discourses are highly elastic, are ill-defined, and are often superficial. Shifting from one ‘other’ to the next is possible, without these autochthony discourses losing credibility or impact.Footnote 17 The malleability of autochthonous categories points to the volatility of these discourses delineating who is in and who is out.Footnote 18 Hutu are thus sometimes included, and at other times excluded as part of the ‘autochthonous’ population of Kivu and the Democratic Republic of Congo (DRC) in general. Moreover, Kinyarwanda-speakers themselves are also internally divided. Many Hutu in Goma, for example, consider themselves as ‘more autochthonous’ than Tutsi.Footnote 19 Because of the volatility of these autochthony discourses, they are very easily manipulated and mobilized.Footnote 20 Often autochthony discourses are embedded in wider discourses about identity in the Great Lakes region, such as the Hamitic hypothesis.Footnote 21 This hypothesis, a colonial construct that presupposed a hierarchy in African races, held that Tutsi origins lay outside of the Great Lakes region. While this discourse played an important role during the genocide in Rwanda, it also recurs in autochthony discourses in the DRC.Footnote 22
According to Bayart et al., claims to ‘autochthony’ are not only characterized by their ‘slipperiness’, but also by their strong emotional appeal and their dubious historical basis.Footnote 23 The Hamitic hypothesis is an example of a discourse with a strong emotional appeal but without much historical validity. It could be considered as a ‘mythico-history’, a term coined by Liisa Malkki when conducting research among Hutu refugees in Tanzania. These refugees’ renderings of the past:
went far beyond merely recording events. It represented, not only a description of the past, nor even merely an evaluation of the past, but a subversive recasting of it in fundamentally moral terms. In this sense, it cannot be accurately described as either history or myth. It was what can be called a mythico-history. … [T]he refugees’ historical narratives comprised a set of moral and cosmological ordering stories: stories which classify the world according to certain principles, thereby simultaneously creating it.Footnote 24
Therefore, the notion of ‘mythico-histories’ is not only useful in understanding how these autochthony discourses are constructed, but also in understanding the nature of historical narratives in the region in general. We will now deal with the mythico-history of a ‘greater Rwanda’, focusing on how mythico-histories connect past and present, and how they relate to discourses on conflict and territory.
THE ‘GREATER RWANDA’: A MYTHICO-HISTORY IN AND ABOUT RWANDA
Focusing on the mythico-history of ‘a greater Rwanda’ shows that it not only bolstered Rwanda's meddling in the eastern DRC, but also that it is important to contemporary state-building processes within Rwanda. The idea of a ‘Greater Rwanda’ is based on two related premises: (1) by imposing borders, the Europeans divided one presupposed homogenous group (the Banyarwanda) that existed before the imposition of the borders; and (2) through the imposition of these artificial borders, a large part of the territory that it had acquired through conquest was amputated from the precolonial Rwandan kingdom.
The first premise confounds the cultural continuum of Kinyarwanda-speaking people, sharing cultural traits, with the existence of a political realm that incorporated these Kinyarwanda-speaking people; the second premise equates conquest or presence at a certain time with long-lasting occupation and rights to territory.Footnote 25 Apart from the political and ideological dimensions of this idea of a ‘greater Rwanda’ the historical record does not support this mythico-history. If anything, one could easily argue that Rwanda was smaller, and not larger at the end of the nineteenth century (especially after the demise of mwami (king) Rwabugiri in 1895), or at least that not all regions nowadays included within Rwanda were under control of the centre at that time.Footnote 26
It is necessary to scrutinize the genesis of the idea about the alleged amputation of Rwandan territory. It often comes up in daily speech in Rwanda and it is actively propagated in Rwandan state-sponsored media such as the New Times:Footnote 27
Berlin Germany, 1884–5. Rwandans may never know what was discussed in the Berlin Conference. … But they know that, whatever the method of decision, at the end of it they came out the ‘tinier’ for it. From a vast country that covered swathes of eastern Congo, southern Uganda and north-western Tanganyika, Rwanda became the tiny hill of Central Africa. There is no doubt that having a strong, centrally-organized administration had something to do with it. No one wants a strong, bothersome influence near when they set out to ‘civilize’ a region.Footnote 28
More importantly perhaps is the impact it has on perceptions of the involvement of Rwanda in the conflicts in Kivu and the potentiality of these discourses for aggravating tensions towards the Kinyarwanda-speaking population in the DRC.Footnote 29
The most telling example of these Rwandan claims was made on 3 October 1996.Footnote 30 Pasteur Bizimungu (then president) gathered an international audience in the middle of the first RPA attacks on Congo (then Zaïre), and just before the RPA launched the first attacks on the Hutu refugee camps in North and South Kivu.Footnote 31 He discussed the boundaries of the precolonial Rwandan kingdom and claimed that Rwanda had been ‘dismembered’ by the colonial borders. Although he did not state it explicitly, the claim that without the imposition of colonial rule, large parts of the eastern Congo would have been Rwandan anyway seems to have been a way to legitimize the initial stages of Rwanda's involvement in Congo and the meddling that was to come.Footnote 32 Bizimungu supported his claims with maps of precolonial Rwanda, based on Alexis Kagame's work, showing that the borders of Rwanda's precolonial kingdom stretched far into the eastern DRC, as far as Lake Edward.Footnote 33
Catholic priest and historian Alexis Kagame was born in to a family of biru, official ideologues appointed by the royal court of Rwanda to preserve traditions, which heavily influenced his historical outlook.Footnote 34 His visions of the expansion of Rwanda were ideological projections propping the court's political and territorial claims rather than (always) describing actual historical realities. As Catharine and David Newbury have argued, official royal court histories and royal Rwandan history both have a powerful ‘irredentist streak’.Footnote 35
Kagame projected an image of a unified Rwanda long before it showed centralizing tendencies and confused temporary conquests and evanescent occupation (which did take place) with actual and long-lasting annexation (which did not always take place). That this vision about a ‘Greater Rwanda’ as propagated by Kagame was a court-sanctioned vision can be illustrated by mwami Musinga's (1896–1931) discourse in a letter written to the Germans in which he promised his support during the First World War.
[D]o you think … that I love the Belgians so much more than you, or the British more than the Germans? Look, the Belgians came to Ishangi and killed several Banyarwanda and took their cattle, they took part of my territory, notably Bwisha, Gisigari, Bunyungu, Kameronse [today in DRC] and Bugoyi. That has been taken from me, and the possessions I had there, and they also have taken away other parts of territory that I do not remember anymore, they prohibited the population of these places to bring me anything. … If you fight against them [the Belgians], it is in my interest.Footnote 36
It seems that the trickling down of this idea into the rank-and-file of Rwandan society is something that occurred later, and that even people associated with the monarchy did not always adhere to this vision of the past. Even for mwami Rwabugiri's reign (1867–95), the heyday of precolonial Rwandan expansion, the idea of actual territorial gain outside of the current Rwandan borders is contested. Rwabugiri's reign was marked by multiple military campaigns against regions in what is now the eastern Congo and he managed to occupy some of these regions for a limited period.
However, this temporary occupation never led to full integration into the Rwandan kingdom.Footnote 37 Narrators in the Ibitéekerezo, a body of historical narratives collected between 1958 and 1960, refute the idea that Rwabugiri's conquests led to the actual expansion of the Rwandan kingdom.Footnote 38 One narrator, Nganguure, claimed: ‘All the foreign countries that Rwabugiri had conquered, they all reconstituted themselves. No stranger remained in the country. They reconstituted themselves instantly.’Footnote 39 Birasenyeri, a lieutenant of Rwabugiri involved in his military campaigns expressed a similar idea: ‘Of all the fleeting conquests in Bunyabungo [Bushi] and Idjwi only the memory remains.’Footnote 40
Most likely, the idea of ‘a greater Rwanda’ mobilized by those connected to the monarchy began its wider propagation with Alexis Kagame's work. During Rwanda's Second Republic (1973–94) it seems that Kagame's work remained the main source for the history of Rwanda. In the manuals for history education during this period, the idea of le Rwanda ancien stretching into the actual territories of the DRC and Uganda was present.Footnote 41 Even Ferdinand Nahimana, staunch critic of Kagame's version of the past on other topics, seems to have accepted Kagame's version of the geographical extent of the expansion of Rwanda.Footnote 42 He wrote:
The kingdom of Rwanda was constituted, before the colonial intrusion in a true state (véritable entité étatique), with solid structures of government, which were exercised over a rather homogeneous population, especially because of the language, Kinyarwanda. Rwanda had known borders, easily visible. Thus, towards the west, the kingdom encompassed the whole of Lake Kivu, islands included. The colonial objectives of … Leopold II, interfered while the kingdom tried to annex, successfully, the territory populated by Havu, Andani [sic] and Hunde … thus the whole region situated to the west of Lake Kivu and the Rusizi river. Also in the west … the Rwandan kingdom already had a stable border which ended in Lake Rwicanzige, the actual Lake Edward.Footnote 43
These ideas of a greater Rwanda have persisted to the present day. If anything, they have become more closely related to the nation-building and reconciliation project that the RPF has instigated. As Pottier and Reyntjens both point out, the (re)writing of the history of Rwanda was (and still is) not a mere by-product of the RPF's rule, but has been a priority of the government and is an important ideological foundation of their regime.Footnote 44
The main claim of this officially sanctioned history is that before colonialism ‘[a]ll Rwandans were living together, and speaking the same language, they had the same culture and were loving each other.’ This unity between the three groups constituting ‘Rwandans’ was first eroded by German and Belgian colonial practices and later destroyed by Belgian colonialism.Footnote 45 This mythico-history – which denies regional differences, inter- and intra-ethnic fighting and unequal political integration – also has consequences for the perception of territory and borders. The idea of living together, sharing cultural traits and speaking the same language echoes nineteenth-century discourses on the European nation-state.
Indeed, the idea of Rwanda as a precolonial nation-state has been picked up by historians in Rwanda, who have argued that a unified Rwandan nation of Kinyarwanda-speakers already existed in the nineteenth century, even if the historical record reveals that the situation was more complex.Footnote 46 Frank Rusagara vented similar assertions in Resilience of a Nation.Footnote 47 He advocated for the campaign ndi umunyarwanda (‘I am Rwandan’), which promotes the government's version of ‘reconciliation’ and the idea of ‘Rwandanicity’, in Europe.Footnote 48
Likewise, during itorero trainings – a form of ‘civic education’ meant to instil ‘Rwandanicity’ in Rwandan subjects – one of the key messages about Rwanda's presupposed precolonial golden age is that the territory of the precolonial kingdom was in perpetual expansion, and surpassed its current borders. To emphasize the greatness of ancient Rwanda, maps including parts of the DRC, Uganda, Burundi, and Tanzania into Rwanda are sometimes shown to the itorero participants.Footnote 49
Although it is difficult to assess their impact, political discourses have mobilized the history of the expansion of the Rwandan kingdom to legitimize military and political interventions in the DRC.Footnote 50 Today this ‘mythico-histoy’ also buttresses and is related to discourses about colonial wrongdoing that led to territorial losses – which is also very present in explanations of how ‘Rwandans’ came to be divided – and an aspiration to a certain grandeur that Rwanda has lost.Footnote 51 The evolution of this idea of a ‘Greater Rwanda’ shows that it is not new, but that it is being repackaged for different reasons.
This narrative of a ‘greater Rwanda’ has also been instrumentalized across the border in the eastern Congo, albeit in a different fashion. There, it feeds into historical narratives about expansionist tendencies of Rwandans in general and Tutsi in particular, and has been turned into a lens through which more recent aggressions by Rwanda, or rebellions with support from Rwanda, are being read. In turn, this framework is mobilized to legitimize violence against Rwandophones or Congolese Tutsi who, in this vision, are either seen as an extension of these Rwandan plans or considered as Rwandan tout court.
CONTESTED CITIZENSHIP AND THE COLONIAL TOOLBOX
Kinyarwanda-speakers in Congo are not a homogenous group. They have very different backgrounds and arrived during different time periods in what is present-day Congo. Most researchers distinguish four different groups:
-
(1) ‘Hutu’ and ‘Tutsi’ who settled in Kivu long before the colonial period (for example, the Banyamulenge in South Kivu, Hutu and Tutsi of Bwisha around Rutshuru, …).
-
(2) ‘Banyarwanda immigrants' (or their descendants) who were settled in Kivu by the colonial administration from 1937 onwards to serve as a labour pool for the plantation economy.
-
(3) The ‘fifty niners’, Tutsi refugees who fled Rwanda because of events in Rwanda between 1959 and 1963.
-
(4) Hutu refugees who fled Rwanda in the aftermath of the 1994 genocide.Footnote 52
Throughout Congolese history the citizenship status of Rwandophones has been contested. The date of arrival has been important to establish legal citizenship in Congo. During the transition to independence, the right of ‘Banyarwanda immigrants’ was not solved.Footnote 53 Only in 1964–5 was a law passed that clearly stated that those who ‘had at least one ancestor member of a tribe or part of a tribe established on Congolese territory before 1908’, were considered of Congolese nationality.Footnote 54 The law of 26 March 1971 defined citizenship as: ‘The people originating from Ruanda-Urundi established in Congo on the date of June 30 1960 have acquired the Congolese nationality on said date.’Footnote 55 Barely a year later, in 1972 this law was slightly altered: Congolese nationality was again conferred to those belonging to a corporate ‘Congolese ethnicity’. Nevertheless, those from Ruanda-Urundi living in Congo on January 1 1950 were still eligible for Congolese nationality.Footnote 56 In 1981 the law granting Congolese nationality to people from Ruanda-Urundi was again revoked, and pushed the date to which an ethnic community had to be established in Congo back to 1885.Footnote 57
The current law gives Congolese nationality to all those individuals belonging to ethnic groups whose people and territory constituted the Congo in 1960.Footnote 58 The law does not exclude all Kinyarwanda-speakers from access to citizenship. The first and second group of Kinyarwanda-speakers have, in theory, no problem claiming Congolese nationality. However, the problem lies in its interpretation.Footnote 59 Many people still believe that the cut-off date for nationality should be the date of the Berlin Conference in 1885 (as it was in the law between 1981 and 2004), or are not aware that the date has been changed. At the same time, many Kinyarwanda-speakers, and especially Congolese Tutsi, do not feel reassured by the current law because their citizenship is still often contested in local contexts and their citizenship has been switched on and off opportunistically for political reasons in the past.Footnote 60
More problematic is the ambiguous language of the law, which vaguely defines ethnic groups as having not only ‘people’, but also ‘territory’. In practice, being considered as a ‘people’ with a ‘territory’ is often dependent on or equated with having had a ‘customary’ organization recognized by the colonial administration. Consequently, in North Kivu, depending on the point of view and interpretation, this makes it easier to deny the second wave of Kinyarwanda-speakers’ (those who arrived during the colonial period, and their offspring) legitimate claims to nationality and thus citizenship. While between 1937 and 1958 they had a chieftaincy of their own (the Banyarwanda-chefferie in Gishari, Masisi highlands), this chieftaincy was abolished in 1958, effectively blocking these second-wave Kinyarwanda-speakers from access to land and political authority on the local level in Masisi Territory.Footnote 61
To understand contemporary struggles around land and political power at the local level in North Kivu, the creation of chieftaincies by the colonial administration is crucial. Through the introduction of these chieftaincies, cultural communities with varying degrees of political centralization and political coherence in what are nowadays the North and South Kivu provinces were (re)organized into hierarchical and centralized political entities in neatly defined territories. This idea of mapping ‘culture’ and ‘political power’ in a one-to-one relationship to a neatly defined territory was then telescoped into the past to give these re-invented political entities legitimacy. Most importantly, this ‘territorialisation of identity’ that was a result of this administrative reorganization influenced and continues to profoundly influence the access of these communities to the colonial and postcolonial state.Footnote 62
The idea that cultural groups must coincide with a hierarchical and centralized political organization which in turn is closely related to well-delineated territory became integrated in the discursive framework of ‘autochthony’ and has inscribed itself in the discussion about ‘who belongs where’. Denying an ‘autochthonous’ status based on whether certain groups had certain forms of ‘customary’ organization under colonial rule became a factor in political struggles and struggles over resources.Footnote 63
The language of ‘autochthony’, which juxtaposed an ‘autochthonous’ Congolese population against those ‘coming from Rwanda’ (the migrants who came to Congo in the context of the resettlement scheme for Rwandan labour), had already started in the 1950s among agents of the colonial administration. See, for example, an excerpt from a letter of the Provincial Governor of Kivu in 1956, which was a reaction to an article written by a certain abbé Kajiga.Footnote 64
We must do everything to keep the indigenous authority on different levels in the hands of the autochthones [sic], we cannot in any way encourage the study of Kinyarwanda and we must favour the use of local languages and Swahili. In the schools the use of Kinyarwanda needs to be constrained, the [?] need to be always in preference attributed to autochthones [sic], the immigration and installation of new Watusi (Tutsi) coming from Rwanda should be avoided as much as possible (autant que faire se peut), finally we must … monitor that the natives are mingled to the maximum with the Rwandans already installed. …. I insist you make all agents who are interested in the problem of the coexistence of the Banyarwanda and our population read and reflect on Kajiga's study so they realize (s'impregenent) the insidious dangers of this Rwandan irredentism and so we develop with all the tact and caution necessary the attitude to defeat it.Footnote 65
Notwithstanding the different historical context of this citation (the reactions of the colonial administration were shaped by fears concerning local conflicts and ‘Rwandan irredentism’ rather than shaped by years of armed conflict involving Rwanda, directly or indirectly), the gist of this discourse resembles present-day exclusionary discourses in Congo.Footnote 66 Particularly, the emphasis on the juxtaposition between so-called ‘autochthones’ and Rwandans, and the suspicions that Kinyarwanda-speakers – especially Tutsi – are somehow prone to territorial expansion are simmering through.Footnote 67 It is known that in Francophone Africa the discourse of ‘autochthony’ had colonial roots, but it seems that this was also the case in the Belgian colonial sphere.Footnote 68 Similar discourses have often been reiterated during and after the Congo wars, and are still heard today. Such perceptions were, in the past, but also more recently, possibly fuelled by historical narratives in Rwanda of which we have seen examples above.
In sum, decisions by the colonial administration about how they chose to administer the population are mobilized today to exclude Kinyarwanda-speakers (in the example below the Banyamulenge) from citizenship. The absence of a ‘customary’ organization recognized by the colonial authorities is simply equated with not having been there, and thus discredits claims to Congolese citizenship and the rights that go with it for these people. This excerpt from a declaration of the Mai-Mai of Fizi in the context of the 2008 Goma Peace Conference is quite telling in this regard:Footnote 69
They [the Banyamulenge] say they have arrived in Congo two centuries ago, this is not true, it is common knowledge (notoriété publique) that the Belgian colonizers organized Groupements according to the different ethnic and cultural groups. For example, the Basikasingu living in Fizi Territory because they did not share the same culture as the Babembe had their own Groupement in the Lulenge sector … If they [the Banyamulenge] were there, how could the Belgians have forgotten them? All the chiefs of the Groupements received palata (royal insignia). If they are Congolese, let them show one.Footnote 70
Nevertheless, even having had a ‘customary’ authority recognized by the Belgian colonial administration is not a guarantee. Different perceptions of the group of Kinyarwanda-speakers in Rutshuru Territory show again how ‘slippery’ this category of ‘autochthony’ really is. During the colonial era, this group was granted its own chieftaincy led by a Hutu mwami. For many, these people (sometimes also called Banyabwisha)Footnote 71 are considered ‘autochthonous’ Congolese.Footnote 72 Others still try to cast doubt on the ‘autochthony’ of the Kinyarwanda-speakers living in this Bwisha chefferie, framing them as migrants arriving after the delineation of the colonial borders.Footnote 73 Often, such narratives on autochthony also divulge a rather positivistic preoccupation with historical ‘hard proof’ and ‘documents’, preferably originating in the colonial past. The reference to the palata above is a case in point, but there are also other examples.
The ‘guerre Kanyarwanda’ (1962–5), is the first time that ‘autochthons’ openly and violently clashed with Rwandophones, and especially those of the second wave (the immigrés) in Masisi, Rutshuru, and Walikale (North Kivu). While this war emerged in the context of the discussion of provincial administrative boundaries in Kivu, it was also about access to power on the local and provincial level. It is said that during this war one of the first moves of the Hunde was to burn the archives of the population registers of all administrative localities. This enabled them to qualify all Kinyarwanda-speakers as ‘strangers’ or ‘refugees’.Footnote 74 In later periods, the use of colonial ethnographic maps to ‘prove’ that there were no Kinyarwanda-speakers in Kivu before 1945 is another striking example.Footnote 75
Thus, the ‘colonial toolbox’ has been an important building block in constructing Rwandophones as non-autochthones. The ‘guerre Kanyarwanda’ (and the wars between 1993–4 that are not addressed here) are examples of how autochthony discourses could contribute to violence, even when not directly causing it. These first rounds of localized violence foreshadowed the more widespread and massive outbursts of violence that would be triggered by Rwandan invasions from 1996 onwards, and which are discussed in the remainder of this article.
COLLECTIVE MEMORIES AND THE POLITICS OF VICTIMHOOD
Autochthony discourses not only juxtapose who is in and who is out, they also often position ‘victims’ against ‘aggressors’. Therefore, victimization discourses can be used to legitimize the use of violence, framing it as a form of self-defence.Footnote 76 Such narratives of victimization often make strong historical references. It is not uncommon to see ‘memories’ and ‘countermemories’ justifying atrocities or ‘pre-emptive’ attacks that are legitimized based on historical wrongdoings.Footnote 77
The Bunyungu royal family is the lakeside branch of one of the Hunde royal families in Congo. One of the historical songs sung at the Bunyungu Court relates how a Rwandan mwami decapitated a member of the Bunyungu royal family. According to the song, the decapitated nobleman rebelled against the Rwandan mwami.Footnote 78 After the latter had thrown the nobleman's head into Lake Kivu, it crossed Lake Kivu while singing, and went ashore somewhere between Sake and Minova.Footnote 79 According to the memories attached to the story about this head, at the place where it went ashore, there is now a gigantic ficus tree, which remains untouched to this day.Footnote 80 After a performance of the song, almost all the men present were visibly touched. When asked why, they said that it was because one of their own was killed by another mwami from a neighbouring country, and that it tells the history of their people facing hardship caused by a neighbouring state.Footnote 81
Their response to the song shows that history is very much alive in this region. Oral traditions do make mention of a Hunde prince having been killed by a Rwandan mwami, although it seems to be unclear who exactly.Footnote 82 While it would be tempting to see this as just a historical tale, the image of a beheaded victim seeking revenge is also a ‘core cliché’ of various oral epics in the region.Footnote 83 In this case however, the story was reinterpreted against the backdrop of Rwandan ‘aggression’ since 1996. In the context of the consecutive wars mentioned in the introduction, many atrocities and human rights violations were committed by Rwandan and Rwandan-backed forces, but also by others, between 1996 and 2003.Footnote 84 It is in this sense that the song can also be considered to be playing a role in the construction of Hunde, and by extension, Congolese victimhood at the hand of Rwanda.
In a similar vein, there seems to be a historical trope linked to the alleged cruelty of Rwabugiri specifically, or Rwandan bami in general. One of the men present during the performance of the song had spoken before on the topic of Rwabugiri: ‘They said he was a colonizer, even while drinking beer, he drank it sitting on the knees of others. He tortured people, instead of putting his staff in the ground, he put it in a foot.’Footnote 85
‘Collective memories’ of precolonial ‘Rwandan’ wrongdoings are not limited to the example above. I collected historical narratives about Rwabugiri among other groups as well, and it seems that all the lakeside people share such memories.Footnote 86 However, although these collective memories are based on actual historical events – Rwabugiri's campaigns affected most lakeside communities and Idjwi – it is difficult to discern in how far these collective memories have been influenced by more recent events (such as the Rwandan invasions mentioned above); the instrumentalization of such narratives by politicians; and the influence of discourses circulating in the run-up to and during the 1994 genocide (see below).Footnote 87 The statement on Rwabugiri above, for example, echoes this tract circulating in Bukavu around October 2000:
People of South Kivu, following the barbarous crimes committed in Kavumu, Makobola, Burhinyi, Mwenga and Bunyakiri, massacres against our peaceful population of Bukavu are already being prepared by Kagame, Museveni and Buyoya … History does not contradict us. The terrible atrocities committed shortly before the beginning of the 20th century by the Tutsi kings prove sufficiently to what extent you are descended from Cain. Just imagine: A Tutsi king, every time he wanted to stand up, had to lean on a spear that was plunged into the leg of a Hutu subject. The point was very sharp and covered with poison. What cruelty!Footnote 88
In this tract, suffering is given more time-depth by referring to a distant past, while the distant past is used as a warning for what might come.
Apart from this discursive strategy, it is possible that the discourses evolving in Rwanda between 1990 and 1994 also had an impact on collective memories and on discourses circulating within the Congolese context. The song Bene Sebahinzi, by notorious musician Simon Bikindi, and one of the songs regularly returning on Radio Télévision Libre des Mille Collines (RTLM)’s playlist, narrates the suffering of the Hutu population at the hand of the (Tutsi) monarchy. It sings the histories of conquest under the (Tutsi) monarchy and the extermination and defeat of small-scale polities and their kings in western Rwanda. The song also reminded the audience of the cruelty of the monarchy by telling how Kalinga (the royal drum), was decorated with the genitals of those defeated kings.Footnote 89 While Bene Sebahinzi was more than once the soundtrack of killing-sprees during the genocide in 1994, this trope of royal cruelty was repeated during RTLM's emissions, and was also reproduced through other media such as Kangura, one of the newspapers stoking virulent hatred against the Tutsi and politically-moderate Hutu in the run-up to the genocide.Footnote 90
Calls to violence rousing the Rwandan Hutu population against the Tutsi were in 24.4 per cent accompanied by allegations that the RPA were not Rwandan and part of a conspiracy to establish a Hima-Tutsi-monarchy in the region, while 14.14 per cent of the statements referred to the cruelty of the monarchy (but also armed insurgencies in the 1960s).Footnote 91 Kangura addressed audiences that went beyond Rwanda, and also tried to reach out to Hutu in the eastern Congo, as in Kangura 6 (1990) with the Appel à la conscience des Bahutu (see also below), while RTLM sometimes also referred to Tutsi infiltrating from Congo (then Zaïre).Footnote 92
It is difficult to unravel how specific narratives have travelled within the region around Lake Kivu. Moreover, certain tropes and clichés are probably part of a deeper cultural lexicon shared by different groups within this region, but have been given new meanings in the pre-genocidal context in Rwanda, and as result of the wars in Congo since 1996. Thus, narratives about the cruelty of Rwabugiri (or other ‘Rwandan’ kings) were already part of ‘collective memory’ among different groups at the western shores of Lake Kivu, but they no longer bear the same meanings as they did in the past – and even in the past their meanings were not fixed. In turn, these discourses have been used to legitimize violence against Congolese Tutsi.
TERRITORY AND THE INSTRUMENTALIZATION OF VICTIMHOOD
The imagery associated with the Hamitic hypothesis mentioned above is one discourse through which Congolese Tutsi became framed as ‘outsiders’. The ‘colonial toolbox’ also played an important role. But politics, and especially their role during the Congo Wars (1996–8/1998–2003) and in the CNDP and M23 rebellions, have also shaped the conflation of Congolese Tutsi with all Tutsi, and of Congolese Kinyarwanda-speakers with Rwanda in the eyes of many other Congolese.Footnote 93
Above we saw that the rights to Congolese nationality for Rwandophones has been ‘switched on and off’ whenever political leaders – and especially Mobutu – felt that playing out ‘ethnicities’ against each other would suit them.Footnote 94 In South Kivu, the changes in the 1981 nationality law (which put the cut-off date at 1885) had barred Banyamulenge from participating in the elections in 1982 and 1987, as their adversaries argued that they had arrived only after the establishment of the Congo Free State in 1885. This furthered already existing tensions between the Banyamulenge and other communities in South Kivu.Footnote 95
At the beginning of the 1990s, the transition to multi-partyism led to fierce competition between political elites in South Kivu. One member of these political elites, the vice-president of the National Assembly and member of the Bembe community, Anzuluni, was the instigator of several anti-Banyamulenge campaigns. The arrival of refugees from Burundi and Rwanda fleeing (anticipated) massacres in 1992–3 only increased these tensions. Anzuluni increasingly used an anti-Banyamulenge rhetoric in public, which led to widespread harassment of the Banyamulenge.Footnote 96 Speeches such as these bolstered already existing anti-Tutsi sentiments, identified Banyamulenge as the enemy, and portrayed (pro-active) self-defence as legitimate.Footnote 97
Meanwhile, the recruitment of many Banyamulenge by the RPF, who had started their campaign against Kigali in the early 1990s, reiterated the identification of the Banyamulenge with Rwandan Tutsi. In the summer of 1996, this led to growing anti-Tutsi sentiments when these early recruits started to return to South Kivu, and tit-for-tat killings between the Banyamulenge and other communities became common currency, while cross-border tensions also began to mount.Footnote 98
In North Kivu as well, at least from the beginning of the 1990s, many Tutsi joined first the RPF, and later the AFDL. They were motivated by multiple factors, among them the shaky status of their citizenship, the deterioration of relationships between Congolese Tutsi and other communities, and intercommunal violence between 1992 and 1994.Footnote 99 The armed force invading Congo in October 1996, officially starting the First Congo war, was heavily dominated by Rwandan Tutsi, but also by Banyamulenge and Congolese Tutsi. This reinforced the identification between ‘Tutsi’ and ‘Rwanda’.Footnote 100
In North Kivu, the violence of 1992–4 and threats posed by the armed gangs hiding among Rwandan Hutu refugees had already led to the emergence of local militia.Footnote 101 Yet, it was the invasion of the AFDL that led to the mushrooming in rural areas in North and South Kivu of armed groups operating under the label of Mai-Mai. The shared language of these Mai-Mai groups was ‘autochthony’ and local defence against the ‘Rwandan’ invaders.Footnote 102 The fickleness of the political alliances of these Mai-Mai groupings is legendary, and opportunistic motives often trumped ideological ones when it came to those with and against whom they fought. Nevertheless, these Mai-Mai, at several stages supported by Kinshasa, mounted local resistance against the Rwandan invaders, or against those armed groups backed by Rwanda.Footnote 103
This was the case for resistance against the RCD (1998–2003), already mentioned above, but also for those groups that emerged after the end of the Second War (1998–2003), and that had roots in the previous movements, such as the CNDP (2005–9) and M23 (2012–13). The latter movements are generally considered as ‘Tutsi’, although they were not supported by all Congolese Tutsi. They thrived on a mixture of grassroots grievances, while also being used by military and political elites for political and economic objectives, and profiting from state weakness.Footnote 104
In 2012 and 2013 the M23 was one of the most powerful armed movements in the eastern Congo.Footnote 105 They violently clashed several times with the Forces Armées de la République Démocratique du Congo (FARDC), causing massive population displacements. In November 2012 they managed to take Goma. It is thus no surprise that they attracted the most international attention, appeared most in official discourses of the Congolese government, and were most present in popular discourses in the urban centres of Goma and Bukavu.Footnote 106 Moreover, the Rwandan backing of M23 rekindled simmering tensions between Rwanda and Congo.
Often, in official discourses – as was the case for its predecessor the CNDP – M23 was framed as an extension of Rwandan military, political, and economic interests, and secessionist tendencies were imputed to M23 and by extension to Rwanda. Richard Muyej, then Congolese Minister of the Interior was quoted in September 2012: ‘M23 is another name for Rwanda. It's all part of Rwanda's Machiavellian destabilisation plan of the east.’Footnote 107 Muyej's quote is yet another example of a belief of sustained Rwandan involvement in Kivu, which is also widespread among the urban population of Goma, Bukavu, and its hinterland.Footnote 108 At the same time, now as much as in the past, narratives such as these have been central to propping the Congolese government's legitimacy in the light of increasing criticism from the population on the government's impotence to provide a basic sense of security.Footnote 109
In many cases, popular narratives were (and are) not limited to commenting solely on Rwanda's economic and political interests, but were also seen as ‘territorial’ in nature: as a threat to the DRC's sovereignty. The territorial aspirations assigned to Rwanda were framed in a wider context of presumed conspiracies aiming for the ‘balkanization’ of the DRC. In these ‘balkanization’ theories, Rwanda is sometimes seen as an agent or collaborator of ‘the west’, which is perceived as solely interested in plundering the riches of the Congo.Footnote 110 Allegations of Rwanda exploiting Congolese resources are not wholly unfounded, as the exploitation of Congo's resources during the Second Congo War was organized by Kigali's Congo Desk.Footnote 111
When Goma was taken by M23 in November 2012, people in Bukavu compared Kabila to Gorbachev. In doing so they referred to Gorbachev's introduction of glasnost and perestroika that ultimately led to the disintegration of the Soviet Union. Others referred to supposed plans by Rwanda to erect a République des Volcans, a kind of independent Kivu led by Rwanda, which would give Rwanda access to Kivu's resources.Footnote 112 In May 2013, an op-ed titled ‘Tutsi, return home’ appeared in Kinshasa's most popular newspaper Le Potentiel:
AFDL in 1996, RCD in 1998, UPC in 2002, CNDP in 2007, M23 in 2012 … all these bloody episodes have in common the protection of the Congolese Tutsi community – a community transplanted from Rwanda to Congo by Belgium during the colonial period. … The deduction thus made keeps the following logic: The Tutsi community, at least its elite, has never been ready for integration. This supports the thesis of the search for a Tutsiland which would be transplanted unto, it speaks for itself, Kivu. … The Kivus, wounded in the flesh by the transplantation of a Tutsiland, feel entitled to recuperate by every means their territory.Footnote 113
Discourses about a ‘Tutsi-land’ or about alleged Rwandan (or often ‘Tutsi’) plans to colonize Kivu were not new during the M23 crisis. During the war in Masisi in 1992–4 Kinshasa newspapers referred to plans for the establishment of a ‘Republic of the Great Lakes’.Footnote 114
These discourses also seem to have cross-border connections, as they (re)appeared in the context of the war in Rwanda that began with the attack of the RPF on the north of Rwanda in 1990, possibly reinforcing, or reinforced by, discourses across the border. In December 1990, Kangura, one of the media outlets stoking hate against Tutsi in the run-up to the genocide, published a piece aimed at ‘Bahutu who are abroad’ that alluded to ‘a Tutsi plan’ to colonize Kivu. The plan would have been ‘discovered’ in North Kivu in 1962 (in the context of the ‘guerre Kanyarwanda’), but is possibly a fabrication, dating back to the 1980s.Footnote 115
These discourses about ‘Rwandan’ or ‘Tutsi’ plans to ‘colonize’ Kivu are also projected back in time. In the eastern DRC, Rwabugiri's campaigns are used to provide ‘proof’ of the presumed inherent and primordial war-mongering and irredentist nature of Rwandans (and especially Tutsi), especially in the DRC. This is, ironically, a cynical reversal of the argument that equates Rwabugiri's ephemeral occupation with actual annexation in Rwanda (see above), and might possibly also be influenced by discourses on a ‘greater Rwanda’. One of the most telling examples is a passage in the report of the 1995 national ‘Vangu-commission’, a commission established in the wake of the ‘Resolution on nationality’ that denied Congolese citizenship to all Rwandan immigrants. The report explained the causes of destabilization in North and South Kivu as follows:
Towards the end of the 19th century, a Rwandan King Rwabugiri tried several times to attack Buhavu, Bushi and Buhunde. His attempts resulted in his death on the Kivu Lake in a Havu ambuscade in 1895. This information is very important to understand the nostalgic attitudes of the Rwandan Hutu or Tutsi who believed that a launched but failed attack (montée et manquée) creates a motive for territorial claims, and hope to relaunch the military expeditions against the geographical region of the Bahavu, Bashi and Bahunde … .Footnote 116
Similarly, a tract from the Alliance des Patriotes pour un Congo Libre et Souverain (APCLS), (one of the biggest Mai-Mai groups in North Kivu), written in 2012 by a team of self-proclaimed ‘patriotic Congolese researchers’ at the height of M23 activity states:
Since its existence on earth, the Hunde population has been the victim of humiliation, frustration, discrimination and attempts at extermination, the consequences of repeated aggression of external forces on Congolese territory and especially Hunde territory. In origin, it is always Rwanda of which their intention is to conquer Kivu. … In fact, before the partition of Africa in 1885, the old Bami (Tutsi Rugangu) [sic] of Rwanda often tried to expand their kingdom through incursions in some parts of the Hunde kingdom.Footnote 117
Perhaps statements uttered by the notorious Congolese CNDP rebel leader Laurent Nkunda may have also reinforced this idea. In 2009 Nkunda allegedly stated:
Rwanda is a neighbouring country of mine, of which the national language, the culture and the ethnic composition are identical to my collectivité-chefferie of origin, Bwisha. With this country [Rwanda], as with Burundi, my country has shared a colonial past for 35 years … under the name of the Belgian Congo and Ruanda-Urundi. If there had been no colonization, and thus no creation in Africa of totally new and artificial territorial entities, it is sure there would be no Congo today. But, there would still be Bwisha, transvulcanic province of the ancient Rwanda. In consequence, the ‘Banyabwisha’ would not be Congolese, but Banyarwanda, exactly as before the ‘Anglo-German Belgian Convention of 1910’ relative to the eastern borders of the Belgian Congo.Footnote 118
In other cases Laurent Nkunda also used ‘victimhood discourses’, very real Tutsi grievances, to underpin and to legitimize his rebellion. The following excerpt from an interview in 2007 is exemplary:
Where I am at, I am not defending the Tutsi. I am pleading their cause. There is a Tutsi cause which has never been defended in this Republic. There are people who do not know if they are Congolese or Rwandese. For years, they have been navigating between Rwanda and Congo. No Congolese official ever tried to weigh in on the situation. I tell myself we need to plead for these people so they are not excluded from the Republic. I have had the bad luck to be Tutsi like those people. Still, I am doing important work. I have convinced many other officials, civil and military – who are not from the Tutsi community – of the necessity of pleading their cause. I have sensitized them about the work I am doing, which is not limited solely to defending the Tutsi, but above all because they (ceux-ci) do not have an advocate (défenseur).Footnote 119
As the examples above indicated, ‘victimization discourses’ and mythico-histories often overlap. Both mobilize a distinct historical component to reconfigure the present. Such discourses reframe reality, in order to make certain actions seem more acceptable. By using the past as a warning for the future, these discourses ‘whitewash’ violence by portraying it as self-defence. ‘Autochthonous’ Congolese and Kinyarwanda-speakers – often also Congolese – or ‘Congolese’ and ‘Rwandans’, are seemingly constructed as ‘natural’ enemies by telescoping hostility into a distant past.Footnote 120
While in the cases portrayed above, ‘victimhood’ discourses construct Congolese identity and ‘autochthony’ as a shared experience of being aggressed by Rwanda, such discourses are also used by other groups.Footnote 121 The statements from Nkunda are a case in point, and in Rwanda as well, partial ‘politics of victimhood’ legitimize rule.Footnote 122
CONCLUSION
In 2013, René Lemarchand wrote about the borderland this article is addressing: ‘Connecting the dots between past and present is nowhere more fraught than where history is a violently contested terrain, where claims to citizenship are heavily determined by ideological constructions, and the tendrils of violence rooted in long ago events.’Footnote 123 This article has demonstrated that the past and its interpretations influence the present, and the present profoundly influences representations of the past. In trying to ‘connect the dots’ between past and present, one risks seeing the past as a function of the present which carries the peril of ‘naturalizing’ contemporary antagonisms.
In a sense, history seems to be all but dead in this region. When it comes to oral history, the rule of thumb is that oral histories say more about the social present than they say about the past they are supposed to refer to. In this case, it seems that both the region's recent past and its present-day context has made it extremely difficult to use oral history. The past and its representations have been constantly manipulated to cater to political needs. While this is in part a continuation of older patterns – both the Hamitic hypothesis, and the manipulation of ethnic identities in Rwanda are exemplary – the cataclysmic events in both Rwanda and Congo, have only enlarged the gap between partial and politicized historical discourse and careful historical analysis.
The answer is not to eschew historical research or oral history, but rather, I would argue, the opposite. We should ‘bring history back in’ and pay more attention to the way ‘identities’ and ‘territories’ acquired the meanings they have today, as they are both products of history. Historical narratives also have a history, and are the result of historical processes shaping these narratives, as much as these narratives have and are shaping historical processes. Therefore, it is not enough to only study them in the context in which they emerge. They are multi-layered, and even if they often refer to a shared cultural or historical lexicon, similar discourses might have entirely different meanings in different historical contexts.
Consequently, more attention on the historical trajectories that shape these narratives is necessary. Ignoring contingency in the production of not only ‘territories’ and ‘identities’ but also the discourses underlying these risks reproducing a politicized social present and ‘naturalizing’ differences and antagonisms by giving them more time-depth. This is a danger not only for insiders, but also for outsiders looking in – academics as much as interveners.
In 1998, only a few years after the genocide in Rwanda and amid the Great Lakes’ regional crisis, the late Jan Vansina wrote that when historiographies are hard to find, people easily resort to what is available or circulating freely. Often, such work is biased and/or lacks rigorous historical analysis and reliability. He argued that to prevent such distorted and politicized accounts, historical research of both the recent and more distant past is necessary, and called for histories going beyond the dominant and politicized frameworks of interpretation.Footnote 124 Today this is even truer than it was two decades ago. Let's bring history back in.
