Constructing identity through commemoration: Kwibuka and the rise of survivor nationalism in post-conflict Rwanda

Abstract

In the years following Rwanda's civil war, the country has remembered those killed in the 1994 genocide with 100 days of official commemoration, known as Kwibuka. The temporary commemoration period is characterised by an explicit acknowledgement and public discussion of ethnic identity, which stands in puzzling contrast to the state's policy of ethnic non-recognition, enforced during the rest of the year in hopes of achieving national homogeneity (Ndi Umunyarwanda). Thus, one observes seemingly diametrically opposed practices of legally erasing identity groups because of their link to conflict and a unique, three month-long saturation of reminders in the form of public speeches, memorial programming and burials, and commemorative signage. A blurring of ‘Tutsi’ with ‘survivor’ and the deliberate passing down of survivor identity to Tutsi youth have created, over time, conditions for a ‘survivor nationalism’, which exacerbates social tensions and risks sustainable peace in the long term.

INTRODUCTION

Every event looks slightly different. Some include a sombre walk through city streets, others begin with Catholic Mass. Others still include burials of bodies newly unearthed by génocidaire confessions. The messaging of each, however, is consistent – dictated by the state, citizens at every level participate in discussions, attend thematic lectures, and perform songs that centre on an annual theme. Together, these events comprise a deliberate, nationwide exercise in genocide commemoration, both creating spaces for post-genocide catharsis and resurrecting pre-genocide identity divides.

Between 2010 and 2017, I travelled to Rwanda seven times. Each of these visits happened to fall at some point during Rwanda's annual one-hundred-day commemoration period, known as Kwibuka,¹ which marks the recognised duration of the country's 1994 genocide. In 2014, I was present for the start of Kwibuka20 – the genocide's 20th anniversary. Coinciding with the official renaming of the genocide² to ‘the 1994 Genocide Against the Tutsi’, the emphasis on ethnicity observed during Kwibuka seems to stand in direct opposition to the government's policy of ‘ethnic non-recognition’ (King & Samii 2017), mandated outside of the commemoration period.

In 2017, I returned for field research from May to July, conducting 38 semi-structured, anonymous interviews, and nine participant observations at commemoration events. I began my research with a focus on the literature around collective memory (Longman & Rutagengwa 2004; Hinan 2010; Longman 2017), state reach (Fujii 2004; Ingelaere 2014; Lyons 2016), trauma (Caruth 1995; Uvin 1997), ‘master narrative' (Jelin 2003), and national identity (Ueno 1999). I suspected that the Rwandan government was using commemoration to repress ethnic mobilisation and political dissent by instrumentalising trauma through state-controlled commemoration. What I found instead was not a passive public, but a society re-normalising ethnic rhetoric. Kwibuka facilitates an increasingly public and political ‘survivor identity’ that is rooted in pre-genocide ethnic recognition and centres survivors as the subjects of (inter)national sympathy. In response to and within the state's construction of Kwibuka, Rwandan publics perform a nationalism which, like the state-constructed commemoration practices, centres survivor-ness.

While policies of ethnic non-recognition and a ubiquitous rallying cry of unity – Ndi Umunyarwanda, meaning ‘I am Rwandan’³ – are meant to promote an all-Rwandan sense of national identity, the opposing experience of Kwibuka actually appears to be encouraging a more nuanced nationalism-within-a-nationalism. For 100 days, the word Tutsi is highly visible; even if one were to avoid the events, there are canvas signs hung at businesses, government ministries, and homes that explicitly name the Tutsi as the victims of the 1994 genocidal violence. National media sources show continual coverage of Tutsi survival, in/justice, reconciliation stories, and the onetime inclusion of ‘moderate Hutus’ among the counted victims has been largely muted. I posit that the centring of survivor identity during Kwibuka – survivor having become interchangeable with Tutsi – has resulted in the advent of a survivor nationalism diametrically opposed to the homogeneity of the rest of the year.

ETHNIC RECOGNITION POLICIES AND FORMATION OF SURVIVOR NATIONALISM

John Sorenson writes that ‘nationalist movements create their own mythologies, organising key incidents, real or invented, and symbols into narrative forms which evoke emotional resonance’ (1991: 313). In pre-genocide Rwandan history, a nationalist insurgency – the Rwandan Patriotic Front (RPF) – was built around a centralised Tutsi identity over many years in exile (Power 2002: 353). Sorenson further posits that strong nationalist movements are framed as a collective struggle against a broad, central concept (such as ethnic identity) and a sense of territorial entitlement above more nuanced grievances. The nationalist ideology espoused by the RPF – now the ruling political party – as it gained power was a mobilising factor for those in exile eager to take up arms to express grievances and retake territory. The contemporary Rwandan state's control of information continues the mythologising work, but the same mythos does work for two different ideologies: the broader Rwandan nationalism which is encouraged year-round, and survivor nationalism which manifests publicly during Kwibuka commemoration.

In post-conflict states, a unifying national identity can lend legitimacy to leadership during transition to an ostensible ‘peacetime’. The RPF's master narrative constructs a very specific post-conflict nationalism that calls for erasure of past identity cleavages and moves Rwanda toward homogeneity. The Ndi Umunyarwanda nationalist present and future occupies a political space alongside the official outlawing of historical ethnic identifiers. During Kwibuka, the outlawing of ethnicity (Hutu, Tutsi and Twa) becomes obscured, confusing the Ndi Umunyarwanda narrative by resurrecting ethnic identity in the daily, public experience.

Victors in conflicts across human history have erased or suppressed innumerable narratives of the state-building process in many ways – through the standardisation of educational materials, for example, or the centralising of state-controlled media.⁴ In the case of Rwanda, the current government came to power as a direct result of the civil war which began in 1990 and is remembered for its culmination (but not end) in the 1994 genocide. A nationalist insurgency – the RPF was made up primarily of Rwandan Tutsis, many of whom had lived in Uganda since childhood – achieved victory over Hutu extremists, but not before civil war and brutal violence had made an indelible mark on Rwanda's physical landscape and national psyche. The total, non-negotiated victory of the Tutsi rebels allowed for power consolidation and a monopoly on political discourse in the ‘new Rwanda’ (Lyons 2016: 177). Since victory, the RPF (which did not change its name upon taking power as a political party⁵) and its leader, General-now-President Paul Kagame, have used a variety of tools to ‘articulate and narrate that they represent a new beginning for their country’ (Lyons 2016: 177). The RPF's successful transition from insurgency to authoritarian regime has subsequently allowed it control over Rwanda's national identity for over two decades.

Since the RPF came to power, much of the nationalist discourse surrounding Rwanda's current political situation has centred on the genocide, the political reconstruction that followed, and the development boom that the country has recently been enjoying. Government discourse reinforces the RPF as having ended the genocide and invokes the spectre of past violence as a reminder that it could happen again.⁶ Likewise, economic and infrastructural progress is attributed to the party's leadership as protecting against the threat of future violence. The Forces Démocratiques de Libération du Rwanda (FDLR) – an active insurgency in neighbouring DRC that formed when génocidaires were chased across the border in 1994 – are invoked as an existential threat to Rwanda during Kwibuka speeches.

In conjunction with its constitutional reform outlawing acknowledgement of ethnic identities, the government has also outlawed ‘genocide ideology and denial’ through Law No. 18/2008. The combination of these two official policies ostensibly erases ethnic identity from the public (and, supposedly, the private) space and implements a cohesive national identity. Elisabeth King and Cyrus Samii, in a 2017 study, provide strong evidence that policies of ethnic non-recognition in states with low ethnic fractionalisation act to discourage the threat of political mobilisation along ethnic cleavages. It stands to reason that the Rwandan government would desire to discourage ethnic recognition, as Rwanda's recent history is rife with negative examples of ethnic mobilisation. Therefore, national solidarity is predicated on a neutral identity mandated by the state.

If Rwandan nationalism as constructed by the state is contingent on ethnic non-recognition, then what work might the selective ethnic recognition that takes place during Kwibuka do for this new Rwandan nationalism? While there is still an identifiable binary (survivors and non-survivors), I argue that survivor communities are much more defined by current commemoration practices than non-survivor communities are. Survivor is becoming defined not by Rwandan antiquity, but by explicitly Tutsi antiquity: a history of violence, marginalisation, discrimination because of economic prosperity, colonial exploitation. Survivor identity in Rwanda is predicated on a single past event – the 1994 genocide – but transforms Tutsi from ‘victim’ to ‘survivor’ by focusing on the future. Therefore, the future of the ‘new Rwanda’ is paradoxically rooted in an understanding of the 1994 genocide as having existed in a vacuum; the greater context of civil war is minimised in both official⁷ and unofficial⁸ histories. Prosperity for survivors then becomes a primary indicator of prosperity for the entire nation.⁹ This focus on survivors’ futures as equating Rwanda's future is exemplified in the communities of students (secondary- and university-level) who now openly claim survivor status despite not having been alive during the genocide.¹⁰

It was extremely difficult to locate non-survivor voices for inclusion in 2017 field interviews; in spite of the ostensible openness around ethnicity during Kwibuka, the taboo around explicitly categorising individuals remains. Furthermore, those who identify as Tutsi during Kwibuka – at least around Kigali, where the majority of this work took place – are disproportionately open about their experiences and more likely to be visibly involved in commemorations. Finally, because most of my interview connections were made by snowballing, survivors sent me to other survivors and rarely indicated that they knew non-survivors personally. The result is analysis that is heavy on the experiences of survivors, without the balance of learning how those who do not get to identify as survivors feel about the apparent divide.

The difficulty of locating non-survivor voices has its own implication: a privileging of one group while socially and politically erasing another¹¹ (under the guise of supposedly erasing all distinction). In this way, the government's culpability for actions taken during a civil war can be undermined and the broader spectrum of violence forgotten. One result of this seems to be that non-Tutsi are being held responsible for survivors’ trauma without having their own war-related trauma positioned within the ‘new Rwanda’ rhetoric. Another result is that non-survivor comes to mean Hutu more generally, which in turn comes to be positioned no longer as an ethnic identity, but instead as a genocide identity: perpetrators. So, those non-survivors who would, under a politics of ethnic recognition, be known as Hutu, become standardised as non-survivors, which in turn implies perpetrators. It is important to consider whether a survivor-centric nationalism is, deliberately or accidentally, exclusionary in its implications. Or perhaps, as suggested by one Hutu parliamentarian, survivor nationalism is an ideology that could be adopted by non-Tutsi identities and, therefore, transcend ethnicity and actually make up the possibility of a ‘new Rwanda’ attainable. I posit that survivor nationalism is growing both organically (community response to atrocity) and inorganically (state-directed commemoration programming) within the larger Rwandan state.

Benedict Anderson's concept of ‘imagined communities’ (1983) relates a shared spatiotemporality as being central to a nationalist identity. Survivor nationalism finds an international rallying cry of Ndi Umunyarwanda claiming a ‘deep, horizontal comradeship’ (Anderson 1983: 6) with ties to the literal nation, which then expand outward into diaspora communities.¹² However, the Ndi Umunyarwanda claim breaks down during Kwibuka, as ethnicity is once again publicly identified both by the self and by the state. The identification of ethnic variation has mapped onto and been reproduced by younger, post-genocide generations that have supposedly been raised in a Rwanda where ethnicity and identity no longer exist. Setting aside, for the moment, the role of private spaces (e.g. the family) in the perpetuation of ethnicity in emerging generations, Kwibuka appears to be the single acknowledgement of ethnicity in the national public space; it plays an essential role in the reformation of ethnic identity, understandings of which are now linked inextricably to the genocide narrative.

CONSTRUCTING SURVIVOR NATIONALISM

Explicit Tutsi identity is lost within a more general history of Rwanda – a history that has been neutralised by policies of ethnic non-recognition. Tutsi nationalism would be unlikely to surface in post-genocide Rwanda because Tutsi identity no longer exists in the public space. Survivor identity, however, is rooted directly in the political reality of the genocide; its genesis was in 1994, which situates it very strongly and deliberately as proof of Rwandan resilience. Participants in commemoration programming frequently remind their audience that Kagame's troops liberated the country from unspeakable violence. This acts as a focal point for contemporary Rwanda's understanding of the civil war as a one-sided issue of historic Tutsi grievance and persecution, and an active call for respect and protection of survivors who still face social marginalisation, post-traumatic stress, and violent retaliation by those who would ‘finish what they started’.¹³ During Kwibuka, survivors actually become a metaphor for the Rwandan nation: violently victimised but committed to resilience and a prosperous future.

Rwanda defines itself as forward-oriented; on a rhetorical level it refuses to be defined by the horrors of its genocide. A photography exhibit in July 2017 showcased portraits of Rwandans in their late teens and early twenties. Most of the subjects were not alive during the genocide, but all had lost family members to the violence. The artist was clear about his message: young Rwandans want their country to be known for more than just the genocide, want to move into a new future, one which is not defined by massacres. Yet in spite of this, the entire exhibit centralised the survivor identity of its subjects. This is consistent with the tone of many of my interviews: survivors, especially youth, are proud to claim survivor status while concurrently being unable to move past the original horror that this identity is linked to.

There is a sense of collective grievance and vulnerability among survivors, and it seems that survivors confirm and affirm each other. While understandable, this also casts the Tutsi population – including the Tutsi RPF – as innocent against the Hutu – who are frequently linked back to the Interahamwe and its contemporary manifestation, the FDLR – offenders. My 2017 interviews revealed that certain characteristics are assumed about survivors. There is first an assumed morality, which manifests often in a belief that ‘a survivor could never do [something immoral] to another survivor’.¹⁴ This assumption was invoked often in discussions of genocide denial/ideology convictions; I would try to glean whether my interviewee believed that a survivor could perpetrate an act of violence against (A) another survivor, or (B) a non-survivor. Responses relating to the former were consistent with the aforementioned quote; such questions were met with confusion followed quickly by attempts to explain the sense of social cohesion among survivors and the fact that this camaraderie would never be disrupted by intra-group violence. Responses relating to B, on the other hand, emphasised the vulnerability of survivors and interviewees would carefully emphasise sentiments like ‘it is not survivors who are violent. Survivors only receive violence.’¹⁵

In considering the work done by survivor identity in post-conflict Rwanda, I considered whether another classification – instead of nationalism – might be more appropriate to discuss the survivor identity I was observing.¹⁶ However, it is the political nature of this identity – as daily life in Rwanda ‘is itself politicised’¹⁷ – that political identity's connection to state discourse, and the Rwandan state's deliberate construction of a broader Rwandan national identity that together situate current manifestations of survivor identity as a nationalism. Ostensibly, Rwandan nationalism is predicated on ethnic non-recognition. But it is precisely that policy of non-recognition – and the reality that the punishments for violating that policy are disproportionately applied to non-survivors – that creates conditions for survivors to be elevated in commemoration discourse. Survivors are their own imagined community, both within and exclusive from the state-constructed national identity. Outside of the Kwibuka period, survivor identity blends as a part of the wider population – the entire country has survived an atrocity and moved on from it together. During Kwibuka, however, survivors exist on the correct side of a cavernous identity divide.

Some of the people I spoke to who identified openly as survivors and discussed genocide-related trauma were not actually in the country in 1994. Many were in university abroad, exiled in Uganda, or staying with family in Europe. Trauma takes myriad forms, and not being present for the actual slaughter does not by any means mean that a Rwandan cannot be deeply affected by violence against country, fellow citizens, family members, or neighbours. However, this broad identifying with survivor, a word quite particular to experience, typically taken on by those who have had an action – physical or abstract – exacted upon them directly, feels out of place for the repatriated who were not present in Rwanda in 1994. In this case, as in so many, language matters, and a more nuanced distinction would be necessary to distinguish those who lived through the genocide first-hand and those who lived through the period during which the genocide took place, but at a distance. The Venn diagram of these experiences overlaps not on survivor as the word is commonly used outside of Rwanda, but on Tutsi.

Tutsi nationalism would not do the work that survivor nationalism does because Tutsi identity is officially the ‘old Rwanda’. As King & Samii (2017) indicate, it is in the Rwandan government's best interest to implement a policy of ethnic non-recognition, as Rwandan ethnic identity is polarised rather than highly fractionalised and the minority group holds political power and implements the policy. By forgoing Ndi Umunyarwanda for 100 days a year and instead framing national rhetoric around survivor or non-survivor status during that time, the unique social privilege that comes with survivor status is encouraged, perpetuated and even imagined by the Rwandan government, which thereby weakens its own policy against ethnic identity.

MEMORY, COMMEMORATION AND MEMORIALISATION IN RWANDA

When Elizabeth Jelin speaks of the selective ‘master narrative’ created through the manipulation of collective historical memory, she warns against agents of the state creating official, hegemonic histories which both erase ‘errors and missteps by those who are defined as heroes’ and selectively centralise national identity and social cohesion (Jelin 2003: 27–9). In conflict-afflicted environments, such master narratives can be a tool for maintaining a delicate peace. Memory is used by states, individuals, institutions and other collectives to narrate the past, conceive futures, and make sense of the present. Memory reinforces group belonging, but when linked to a collective trauma it can have a subduing effect of reinforcing a status quo (McGrattan & Hopkins 2017). Annually in Rwanda, ordinary citizens publicly experience individual and collective trauma, which is invoked through testimonies, skits, speeches, and community gatherings during Kwibuka.¹⁸

The creation of Rwanda's Commission Nationale de Lutte Contre le Génocide (CNLG) in 2008 coincided with the genocide's renaming. At its onset, CNLG ‘was responsible for all matters related to the history and memory of the genocide, including monitoring the implementation of the law on genocide memorials that aimed to define their content and their form. This was the first legislative intervention on the subject, marking the state's tighter control on memorial spaces’ (Dumas & Korman 2011: 21). Memorialisation, commemoration and ethnic non/recognition policy have thus been intertwined for many years.

For the purposes of this article, memorialisation will refer to static memorial spaces (such as massacre sites that have been turned into museums) while commemoration will refer to active, participatory ways of remembering (such as Kwibuka events). Most genocide commemoration events in Rwanda are held during the Kwibuka period, while memorial sites are in use year-round and primarily entertain tourists. I asked each of my Rwandan interviewees whether they visit memorial sites individually or outside of the Kwibuka period; every single one answered that it would be highly unusual to do either, though a few members of student survivor organisations indicated that they sometimes take large groups of fellow student survivors before or after Kwibuka in order to introduce new members to the memorials.

Rwanda's memorials exist primarily because of the 1994 genocide. A 2011 study by Dumas and Korman reveals that prior to 1994, there was only one memorial site in Rwanda. As the newly unified government built out ministries and other bureaucratic infrastructure, a Memorial Commission was created and mandated to identify localities which had experienced the highest number of Rwandans massacred between 7 April and 3 July. At first, memorialisation was ethnically neutral, with a single exception. Dumas & Korman (2011: 19–20) describe:

Visitors to Mugonero, in western Rwanda, will be struck by the curious war memorial that was erected there. It is curious because to our knowledge, no equivalent exists in Rwanda. At the entrance of the Adventist complex, which includes a church and a very large hospital, is a cement block on top of which sit a machete and a nail-studded club – which the killers referred to as … ‘no mercy to the enemy’ – flanking two fists breaking their chains. In addition to a sculpture displaying such instruments of violence, the frontispiece of the adjoining building, in which the bones of victims are preserved, bears an equally striking inscription: … ‘Memorial for the innocent Tutsis swept away by the April genocide, killed like animals under the eyes of a State run by cruel and despicable men. We will remember you forever.’ Until 2009, this was the only memorial that made direct reference to the identity of the victims as Tutsis.

In 2003, the Rwandan government adopted its official policy of ethnic non-recognition (King & Samii 2017). If ethnic recognition refers to the ‘formal identification of ethnic groups by name in constitutions or political settlements’, then non-recognition is understood to be the opposite, primarily implemented, the report finds, by minority-led governments in post-ethnic conflict settings, with concerns about facilitation of ‘interethnic comparisons’ and the re-entrenchment of ‘ethnicity as a political cleavage’ (King & Samii 2017: 2). The policy of ethnic non-recognition is in line with the 2008 law against genocide denial and ideology, and genocide memorial sites, though more politicised than they were between 1994 and 2008, do not appear to play a significant enough role in the daily lives of Rwandan citizens to have much of an influence on ethnic understandings. It is the active practice of commemoration that stands in puzzling opposition to the constitutional policy of ethnic non-recognition.

This commemoration is Rwanda's most measurable example of the country's shifting identity politics. Survivor nationalism in contemporary Rwanda is predicated on solidarity and communal identity forged by those who survived the genocide. This is, in one sense, an expected response to atrocity and war as people who lived together through extreme trauma can collectively contend with emotional, physical and political concerns that have to be addressed in the aftermath. There is an empowerment in the term ‘survivor’ – it implies an agency over the narrative, a movement beyond victimisation, and a reclamation of power. But younger generations taking on survivor status indicates that it has become a constructed identity rather than a simple empirical fact.

Bert Ingelaere points to the government in contemporary Rwanda as the centre from which not simply policy, but knowledge itself is ‘actively construed, managed, and controlled’ (Ingelaere 2010: 41). Observable Kwibuka commemoration parallels this. While the Kagame administration has, in recent years, outsourced commemoration planning to various individuals and commissions as well as encouraged smaller, more localised commemorations toward the periphery of the Rwandan state, all Kwibuka themes, public lectures and programming are first created and approved by the central state. These programmes follow with the standardised national narrative and are characterised as essential to Rwanda's progress; in 2017, one of the public lecture themes was ‘building the nation’.¹⁹ Public remembrance in Rwanda continues to be ‘regarded as a duty and one of the main elements on the road to reconciliation’ (Róg 2014).

CONSTRUCTING KWIBUKA

The aforementioned commission that oversees Kwibuka, CNLG, reports directly to government executives and begins planning every year in February. A theme is proposed and publicisation of that theme strategised, but all final decisions are left to higher-ups in the Kagame administration.²⁰ Though employment by CNLG is not contingent on survivor identity, all staff members I encountered in 2017 had a personal survivor narrative. Recently, the Rwandan state has begun to deliberately decentralise the commemoration period. Most of my interviewees noted the shift (after Kwibuka20 in 2014) from larger, state-performed commemoration events to more intimate, community-based events. One government official told me that ‘it does not make sense to only have speeches at the national level’ because there is a need to build ‘consensus within communities about the genocide against the Tutsi’.²¹ The same official said that she believes survivors’ stories do not get told except during Kwibuka; for her, the first week is for everyone, for reconciliation, but the rest of the 100 days, when commemoration is more localised, is for survivors to speak and feel heard. Another interviewee told me angrily that ‘perpetrators had their chance to speak at the gacaca²² courts. Kwibuka is the survivors’ turn to speak.’²³ Decentralisation of commemoration can provide a safe space for processing genocide experiences in villages, neighbourhoods and university campuses. However, the state retains total control over all commemoration materials and programming. Localised community events can differ slightly in their general presentation, but thematic content remains aligned with the government-approved programming. After the first week, local events mostly take place in keeping with the anniversaries of massacres; if a village experienced genocidal violence on 5 May, then that village will hold its village-level commemoration on or around 5 May.²⁴

Presumably in the interest of maximising participation and attention, Rwandan school terms always break at the beginning of Kwibuka. Commemoration events are held in Kinyarwanda; many describe the resurgence of Rwanda's langue maternelle as having its origins in the Ndi Umunyarwanda programme. Using this language as the standard for commemoration reinforces the uniqueness of the Rwandan case and encourages participation at all levels of society (since often French and English are limited for populations on the social margins). Every event I observed in 2017 followed the same basic format with the same basic components: speeches that follow the year's theme,²⁵ one or two survivor testimonies, a moment of silence for all those killed in the genocide, the presence of a guest of honour, and performances of Kwibuka-themed songs. Some events are preceded by a Catholic Mass, and some others include skits; one commemoration event I attended in Byumba featured a skit depicting the trauma that the children of genocide survivors experience and how they deal with it, including a graphic recreation of an attempted suicide. The skit was performed by secondary school students, none of whom were alive in 1994. These components focus on Tutsi identity and the resilience of Tutsi as survivors, and consistently include references to President Kagame and the RPF both having delivered Rwanda from destruction and as the reasons for quotidian peace and security. State-trained trauma counsellors scan the crowd for survivors experiencing post-traumatic stress disorder (PTSD) attacks.

Kwibuka is multi-layered. The first week is described as the most intense, corresponding with the history of the genocide; the rate of genocide-related deaths between 7 and 14 April 1994 is widely acknowledged to have been the highest of the 100-day period. Commemoration during this week is nationwide and a strict expectation of respect is enforced. While formal punishment for non-Kwibuka activities – such as watching sports matches on television or socialising in bars – during the first week is said to have waned in recent years,²⁶ the national mood remains palpably sombre.²⁷ One evening in a Kigali bar, a non-Rwandan friend recounted that he had noticed helicopters sweeping spotlights over Kigali the first night of Kwibuka, two months earlier. Two Rwandan friends responded to this by saying that the government sends helicopters out to look for bars that break the rules and show soccer matches, to which a Canadian responded, surprised, saying he had also seen the spotlights but assumed they were just part of the ‘celebration’.

Government officials and Kwibuka planning authorities repeatedly told me that after the first week, things return to normal. While it is true that most people's daily lives for the majority of the 100 days do not include commemoration activities, there are regular reminders of Kwibuka and ethnicity throughout. Radio programmes play Kwibuka songs, and the national broadcasting agency runs continuous coverage on commemoration events, new graves (more on this below) and interviews with survivors on television.²⁸ Commemoration signage with messages pertaining to the year's theme goes up at businesses, NGOs, ministries and sometimes even private homes (outside of Kigali, some hand-paint signs in the style of the professionally printed banners). The phrasing of these signs is strictly controlled by CNLG; every year, a few iterations of the year's approved banner are designed and individuals, groups or businesses that want to print a banner can request the design. Signs that do not follow CNLG's template are removed when discovered.²⁹ These signs are nearly ubiquitous between 7 April and 3 July and often include references to Tutsi identity now that the genocide has been officially renamed. These shifts in and out of the politics of ethnic non-recognition are abrupt; visual and aural reminders of ethnicity and violence appear overnight on 7 April and disappear just as quickly overnight on 3 July.

CENTRALISING SURVIVORS, DECENTRALISING TESTIMONIES

Consistent with Ingelaere's (2010) observations of knowledge construction in post-genocide Rwanda, my preliminary discourse analysis indicated that Kwibuka programming is publicised from the centre of society – the government and social elites – outward to the periphery – ordinary citizens. The language used in Kwibuka programming has changed over time, becoming more centred around a consolidated state narrative which replaces Tutsi identity with survivor identity, centralises survivor identity, and as a result marginalises non-survivor identities. Several of my interviewees informed me that it is in vogue to claim either Tutsi or part-Tutsi status because to do so is understood to carry a certain privilege and security during Kwibuka. It is unclear whether this practice extends to claims of survivorhood more generally.

Multiple interviews and observations suggested that voluntary non-participation implies a certain label: perpetrators. A Rwanda Broadcasting Agency executive informed me in no uncertain terms that ‘it is responsible to be [at commemoration]’, and that overall ‘the state is looking out for everyone's best interest’.³⁰ This is inconsistent with interviewees who expressed that Kwibuka is primarily for survivors, indicating that this issue does not have consensus on a local level. The disconnect may impact the way that decentralised programming develops over time.

I observed this perception of Kwibuka being primarily for survivors at many of the events I participated in, even though Tutsis remain a minority in the country and all communities are expected to hold their own events. Recognising this, one might expect that programming would include multiple voices or versions of trauma. But the expectation of participation is complicated, as evidenced by one commemoration gathering that I attended not far outside of Kigali. The testimonies portion of this programme was open call – anyone who wanted to speak about their experience or name people lost in the genocide as a tribute was invited to stand in front of a crowd of approximately 200 people and do so. One survivor stood and, rather than recount his personal genocide experience, informed us proudly that a few minutes earlier he had gone to a nearby bar that was playing music to tell them to close. He spoke fervently and grew agitated, saying that survivors can forgive, but perpetrators should also be at commemoration to apologise and testify in order for there to really be reconciliation. An implication here is that everyone who was at the bar instead of attending the non-compulsory commemoration event was considered a perpetrator because of non-participation. As he spoke, participants around me offered their interpretation of the events, all of which supported the man speaking and decried the ‘perpetrators’ who had been at the bar disrespecting the village's ‘survivors’. The resulting discourse affirmed that Kwibuka participation is for survivors – excluding a large subset of the population – but paradoxically insisted attention and respect be given to survivors by everyone else on community commemoration days.

The move to community-based discussions during Kwibuka week to supplement the larger Kigali events, the incorporation of individual villages’ commemorations held on the anniversary of their respective massacres, and hundreds of special commemorations held for and by government agencies, NGOs, universities and secondary schools perform an image of national solidarity within which all Rwandans can openly recognise their past and look toward their collective future. Government officials and other elites tout the decentralisation of commemoration events as a net-positive, a sign of progress, of power reaching the hands of the people, a continuing reconciliation. However, as evidenced by the aforementioned bar incident, there is an illusion of choice and inclusion with these community-based events.³¹ Along with this, a recorded rise in accusations of genocide ideology and denial and individual survivors’ attitudes toward commemoration participation lead me to conclude that choosing not to participate in Kwibuka is perceived to be highly suspect, though localised community-level events are still meant primarily for survivors. Survivors are assumed to grieve publicly as a homogenous whole during a predetermined period of time. Those who do not participate run the risk – and it is a risk – of being assumed to be either perpetrators or families of perpetrators.

SURVIVOR FUTURES

Perhaps the most puzzling component of Kwibuka commemoration is youth involvement. When I first arrived in Kigali in 2017, I inquired after the trauma counsellors that I knew were present at commemoration events and suspected were state-trained. What I learned is that the vast majority of those trauma counsellors are university students and members of ‘survivor organisations’. While a number of survivor organisations exist, I was primarily interested in the Association des Etudiants et Eleves Rescapés du Genocide (AERG), as they have chapters countrywide and are highly visible during the commemoration period. I spoke with 19 university-age students who are trained as trauma counsellors about their involvement with their survivor organisation and ‘artificial family’.

Artificial families are constructed communities for survivor youth whose biological families were killed in the genocide; they forge significant bonds and care for each other deeply. They also receive financial support from the government and their universities, and during Kwibuka, students who are members of survivor organisations are given passes for missing classes or assignments because they are attending Kwibuka events.³² The AERG students I interviewed were all fiercely involved in Kwibuka, facilitating events, giving speeches, campaigning for the re-election of Paul Kagame (to take place later that summer) and volunteering as trauma counsellors all over Rwanda. I asked every interviewee whether there were any members of AERG they knew of who were related to perpetrators or had family in prison, trying to ascertain whether Hutu youth were welcomed in. In response to this question, all students were clear to state that all are welcome in survivor organisation membership – ‘even you, an American’, I was told repeatedly. In spite of this welcoming policy, only one student said he knew of a non-survivor in his respective survivor organisation.

The metadata – such as volunteered information that fell beyond the scope of the questions I was asking or interviewees’ tones and decisions on where interviews would take place – of my 2017 interviews was significant (Fujii 2010). I did not ask about personal experiences of the genocide nor about ethnic affiliation,³³ yet interviewees frequently offered up personal anecdotes from 1994 and 18 out of 19 student survivors self-identified as Tutsi. Many interviewees also parroted much of the state-constructed rhetoric around the Kwibuka theme which, far from being unreliable information in its conformity, reinforced my hypothesis that the state's narrative and dissemination of information is hugely influential for ordinary citizens, and is affecting the way they both comprehend and communicate their own experiences with Kwibuka, reconciliation policies and new identity politics. While the gravity of vicarious and generational trauma (Caruth 1995; Jelin 2003) that the youth contingent experiences should not be diminished, interviews with youth ‘survivors’ indicate that many of them see their role in commemoration as carrying the responsibility of continuing the survivor narrative for generations to come. As one student asked me, ‘if “survivor” means only those who were alive during the genocide, then how could survivor communities continue after that generation [alive in 1994] is gone?’³⁴

ANOMALIES ACROSS KWIBUKA

Beyond the break in the policy of ethnic non-recognition during Kwibuka and widespread youth involvement, other significant anomalies stood out as my research went on. Many changes to the status quo happen consistently during the 100 days but do not necessarily raise red flags on individual incidental levels. Only when one begins to accumulate anecdotal evidence does the incredible range of anomalous observations become clear. These anomalies point to significant threats to individuals’ mental and physical health.

My 2017 interviewees confirmed a perceived increase in violence toward survivors during Kwibuka. Tutsi interviewees expressed unanimously that they feel more unsafe during commemoration than they do the rest of the year, though the degree to which they felt unsafe was deeply subjective and thus difficult to standardise and measure. CNLG collects governmental data on accusations and convictions of genocide denial and ideology during the 100 days; these data indicate an increase of approximately 60% in both accusations and convictions of genocide denial and ideology, and the actual numbers are increasing most years.³⁵ So while it is not clear whether physical violence actually increases, an increase in accusations points to a perception of increased violence, which is significant for perceptions of security. The same data showed that year by year, the number of accusations and convictions is increasing.³⁶ Interviewees also referred to a nationwide ‘heightened sensitivity’³⁷ that exists during Kwibuka. The available data make it difficult to say for sure, but it is plausible that this heightened sensitivity could be influencing perceptions of violence. If this is the case, then perhaps actual violent acts against survivors motivated by genocide denial or ideology do not occur as frequently as they appear to. The accusations and convictions tell the more urgent story.³⁸

Many interviewees had anecdotes of violence against themselves, family or community members, or property during Kwibuka. Livelihoods are often targeted: banana trees are destroyed overnight, the udder of a cow is cut off. One interviewee with whom I spent three separate days confided that during Kwibuka several years ago she returned from Kigali to her home village, where she kept two goats. When she arrived at home, her goats had been tortured with sharpened sticks but left alive for her to find. The image of her goats hurt in such a way was traumatic in and of itself, but this woman then endured the added trauma of having to kill them herself in order to end their suffering. Others told similar stories of livestock being hurt but not killed so that survivors have to commit the killing act themselves.

Another trend during Kwibuka is the disrespect of memorial spaces and threats that echo the dehumanisation of Tutsi as ‘cockroaches’ that took place pre-genocide. During Kwibuka, chicken bones are sometimes mailed to memorials and mass grave sites with messages like ‘bury these with the others because it will not be long before we come for the rest of you’.³⁹ These symbolic actions dehumanise the victims buried there as well as living survivors, who understandably see such actions as threats of future violence against them.⁴⁰ This creates an apparent dual effect on survivors: they rely on the communality of Kwibuka programming, and yet understand commemoration as a time during which they are going to be significantly more vulnerable than they are during the rest of the year.

Officials working for survivor organisations as well as archivists at the Kigali Genocide Memorial Centre informed me that there is an increase in prisoner confessions during Kwibuka. The results of confessions are discoveries of bodies, which are then returned to the community they originally came from and buried with great ceremony, often during that community's Kwibuka commemoration. Survivors frequently refer to this as ‘finding [their] families’. Unburying the graves seems to be pursued without concern for existing infrastructure; just two weeks before my arrival in 2017, a downtown Kigali petrol station was razed to uncover 15 bodies.⁴¹ I was also told that often these found bodies are on the property of Hutu families who may or may not have owned the property all the way back to 1994 or be aware of the burial sites for any reason. Such properties, too, are razed,⁴² but no one was able to tell me whether families whose homes are destroyed are compensated or re-housed. These uncovered bodies can lead to the creation of new (sometimes temporary) memorial sites or are re-buried at existing memorial sites, which become active sites of commemoration during this limited period. In this way, memorial sites are inextricably linked to Kwibuka; Kwibuka maintains its power, in part, by contributing newness annually.

When asked why she believes confessions during Kwibuka are so much higher than the rest of the year, one archivist specialising in genocide documentation expressed surprise: ‘I've never thought about it, but I suppose you're right that it doesn't make sense. Someone should do a research project about that.’⁴³ I later interviewed a prominent Hutu parliamentarian who works with imprisoned génocidaires and asked him the reason for the increase. He simply smiled, informed me that there are no incentives (such as reduced prison sentences) and credited ‘the spirit of Kwibuka’ for moving the hearts of those imprisoned.⁴⁴ The Rwandan prison system, like so many in the world, has come under significant criticism in recent years, making it difficult to believe that prisoners are delaying giving confessions until they are ‘moved’ to do so.⁴⁵

HUTU IDENTITY IN THE ‘NEW RWANDA’

During my fieldwork, I had trouble locating Hutus or perpetrators to speak to, in part because survivor networks are incredibly robust in and around Kigali, so they are an easier population to access. Additionally, there are limitations on ethnic recognition even during Kwibuka; because so many informants told me that Kwibuka is for survivors, it felt inappropriate to ask to be put in touch with non-survivors and especially to explicitly ask for Hutu contacts. In fact, it would be taboo to ask anyone but a very close friend about Hutus, commemoration period or otherwise. Therefore, this study can only relate the effect of Kwibuka on Hutu communities in a way that resembles relief sculpture more than a constructed model of reality – chipping away at negative space in survivor experiences to guess at their opposites.

As survivor stands in for Tutsi, there are recognisable euphemisms for Hutu used in contemporary Rwanda, though these were more varied and less frequently invoked. During Kwibuka events, ‘Tutsi’ and ‘survivor’ are much more common than Hutu or any of its replacement terms. When I did hear them, euphemisms for Hutu included ‘those who killed’, ‘perpetrators’, ‘those who are now in prison’ and ‘offenders’. Because these euphemisms were used with far more regularity than the word Hutu itself, I coded my language accordingly when asking student survivors about non-survivor participation in artificial families and survivor organisations.

For years, rhetoric around the Rwandan genocide recognised moderate Hutus – those refusing to kill, actively protecting Tutsi, being targeted for appearing Tutsi or sharing Tutsi heritage – as victims. The suggestion now, with the official renaming of the genocide and the clear linkage between ‘survivor’ and Tutsi, is that only Tutsi can have survived or been victimised in the genocide. This presumption risks negating the targeting of those so-called moderate Hutus and not only reinforces the characterisation of Tutsi identity as undisputed survivorhood, but also implies that to be Hutu is to be a perpetrator. While one continues to see mention of moderate Hutus at larger memorial sites, that identity is downplayed and seems to be not much more than an afterthought in Kwibuka speeches and events.

This is not new; other scholarly observations affirm that the post genocide division – survivor meaning Tutsi and perpetrator meaning Hutu – has morphed since the 1990s but is rooted in the same binary (Mamdani 2002). While the state narrative, which is most public (and, perhaps, strongest) during Kwibuka says only the former explicitly, the latter is strongly implied and that implication carries certain assumptions. The earlier-cited Hutu parliamentarian informed me that he believes that all Hutu, independent of personal politics or 1994 locality, must actively take responsibility for what ‘their people’ have done to the Tutsi of Rwanda.⁴⁶ His statement came in response to my asking whether he thinks perpetrators should have a role in commemoration.

Interestingly, this is in line with a 2013 speech by President Kagame who, addressing an audience of Rwandan youth, suggested that Hutu relatives of génocidaires should apologise to Tutsi communities for violent acts that they themselves did not commit (Ssuuna 2013). Such a strategy – with all Hutu accepting responsibility for the actions of their ethnic group which is not even acknowledged as existing anymore – necessitates ‘former’ Hutu once again laying claim to an ethnic identity, publicly outing themselves as being linked to a socially undesirable community of perpetrators. It likewise perpetuates a national duality in which outlawed ethnicities continue by taking on their simplified, political roles rooted in a genocide within a civil war.

The contemporary Rwandan inclination to present a simplified recent history (Mamdani 2002; King 2010) has limited the scope of Rwandan peacebuilding, particularly as it risks diminishing the history of the full civil war in collective political memory. A reductionist approach to identity-based conflict has led to a reductionist characterisation of those identities associated with that conflict. In the aftermath of the Rwandan genocide and civil war, the word ‘survivor’ is not applied to Hutu, even if they were present during (and not complicit in) the genocide. Mamdani attributes this to an assumption that ‘every living Hutu was either an active participant or a passive onlooker in the genocide … the dilemma is that to be Hutu in contemporary Rwanda is to be presumed a perpetrator’ (Mamdani 2002: 267). My own research affirms that this observation continues to be true in spite of an official, constitutional policy of ethnic non-recognition adopted one year after that statement was published.

CONCLUSION

In January 2018, an American friend returned to Rwanda to visit the village he lived in for two years. While there, he learned that two people had recently been killed nearby. While very few concrete details were known, one suspicion was that these people were murdered because of their Tutsi identity. One of the victims was a young adult; the second was an older woman. The latter fits a narrative I was repeatedly given while conducting my fieldwork: physical violence against Tutsi is unique to the generation that lived through the 1994 genocide. Multiple interviewees informed me that the younger generation of Rwandans – the generation that is now joining survivor organisations and seeking training as trauma counsellors – do not suffer from these same biases or violent inclinations. Likewise, this narrative says that continuing grievances against older Tutsi, while certainly rooted in ethnic divisions and age-old, macro-level grievances, is contained within that generation. However, the CNLG data that were shared with me on cases of genocide denial and ideology convictions included many accused who were born after 1994; even still, interviewees repeatedly and emphatically told me that these acts of violence do not affect the youth.⁴⁷

Coming back to the village, the death of the young man is another challenge to such a siloed perception of who is and is not affected by ethnically motivated violence in Rwanda. While the details of this boy's murder are largely hearsay and rumour-based, rumour in Rwanda often reveals a great deal of truth (Fujii 2010). Details provided to the foreigner who heard of the youth's death were ambiguous as to the identity of the killers, but it was believed that he was killed simply for being Tutsi. This stands as a direct challenge to the perception that the post-genocide generation is not experiencing harm and raises concerns that the cycle of violence is not only continuing, but may be expanding.

In eschewing an all-Rwandan nationalism for 100 days of the year by framing national rhetoric around survivor status, the central government's involvement politicises the collective claim to survivor identity and imbues it with the scope to be considered a brand of nationalism. This research shows that survivor nationalism relies on four conditions: (1) the genocide must exist in a historical vacuum, without a fuller recognition of the conditions of civil war; (2) ethnic recognition needs to be annually revived for a limited, controlled time; (3) Rwandans who were not alive during the genocide but nevertheless identify as survivors need to participate in Kwibuka programming; (4) the population must be publicly reminded both that the RPF is the reason for peace and prosperity and that Tutsi were victimised in 1994. These conditions have not come about overnight – they have instead been the result of gradual shifts, both cultural and political. There was no post-genocide frenzy for memorial spaces; rather, policies and practices have been deliberately formulated and implemented across decades. New spatial and contextual components are being added to Kwibuka every year: new bodies are grieved, new memorial spaces are designated, new commemoration events are held at those memorial spaces. Kwibuka is highly mutable.

It is worth noting that survivors do not necessarily enjoy immunity from the state. One woman I interviewed moved us from the patio to the living room almost immediately because what appeared to me to be a quiet suburb with no foot traffic on a weekday afternoon was, to her, a ‘neighbourhood where ministers and government people live’.⁴⁸ She is a survivor, well-respected in her community, is married to a military officer, and, as far as I could tell, had nothing controversial to say. Yet she was nervous enough about being overheard by a passing government official that she physically moved our interview (and our entire lunch spread) indoors. I share this to illustrate that, while survivors experience a certain prioritisation, negative consequences of the state's social control remain cause for concern even to those the government claims need explicit protection.

Interviewees consistently informed me that Kwibuka is for survivors, for Tutsi. Some stated that Hutus had their turn to speak in the gacaca courts, while others confidently claimed that perpetrators are not interested in participating and sometimes actively work to disrupt Kwibuka programmes. These exclusionary perceptions could prove destabilising in a nation still not so far out from large-scale violent conflict. As Elisabeth King points out, ‘the exclusion of certain memories of violence is unlikely to lead to meaningful peacebuilding in Rwanda … Grievances surrounding unacknowledged, or unsettled, historical memories are likely to increase in intensity with time’ (King 2010: 303–4). Survivor nationalism resulting from policies of ethnic non/recognition is doing this exclusionary work at all levels of Rwandan society and poses an existential threat to hope for peace in the long term.