Rwanda's President, Paul Kagame is often lauded for transforming a ‘devastated nation’ into a ‘model for the developing world’, characterised by national unity, political stability and economic growth (Crisafulli & Redmond Reference Crisafulli and Redmond2012: 109). A consequent image is that of a dramatic break in political leadership and ideology between Rwanda's ethnocentric past and its ‘unified’ present. But rhetorical calls to eradicate ethnic boundaries in the name of national unity are not unique to Rwanda,Footnote 1 nor are they unique to the post-genocide period in Rwanda's history. Through a detailed comparison of political rhetoric in pre-Independence and post-genocide Rwanda, defined by Desrosiers (Reference Desrosiers2014: 199) as ‘public speech strategically constructed and deployed’, this article reveals striking continuities in the rhetoric and ideology of the Tutsi monarchy, chiefs and Indigenous authorities during the late colonial period (1956–1959), and the official discourse of the Rwandan Patriotic Front (RPF). It also shows that unity-centric rhetoric never entirely disappeared. Although the political fortunes of elite Tutsi dramatically reversed with the rise of Hutu nationalism, exiled Tutsi who fled the 1959 uprising for Uganda continued to champion a post-ethnic ideal. Raised in exile, the founding members of the Rwandan Patriotic Front (RPF) created a rebel movement incorporating key facets of this rhetorical repertoire, at first for intellectual reasons and later, following their rise to power, as a legitimising discourse to extend state reach and control.
Viewing rhetoric as a key resilience tactic, this article argues that the RPF, like its predecessors, uses rhetoric in a strategic fashion to shape regime relations with domestic and international audiences in ways that reinforce power. Both then and now, rhetorical appeals to unity and ethnicity denial, based on revisionist references to Rwanda's pre-colonial dynastic heritage, are designed to suppress counter-narratives about the relationship between ethnicity, nationhood and power to stifle political opposition and enforce discipline at home. This same rhetoric is intended to frame domestic policies in ways that resonate with global discourses on nation-building and self-determination to conceal power concentration by a small (Tutsi) elite. But while these rhetorical tactics failed to suppress the rise of Hutu nationalism in the 1950s, ultimately leading to the dismantling of Tutsi power, they serve to strengthen authoritarian resilience in post-genocide Rwanda, highlighting the central role of contextual factors in shaping the effectiveness of rhetoric as a resilience tool. This article argues that shifts in the distribution of power inside Rwanda (namely, the presence/absence of the colonial authority) and changes in the magnitude and scale of anti-Tutsi violence (pre/post-genocide) introduced different challenges and constraints on the political leadership, which helps explain why a similar rhetorical strategy is associated with different political outcomes.
The article qualitatively analyses archival materials collected from the Diplomatic and African Archives in Brussels to map the rhetorical landscape of the late 1950s.Footnote 2 In like manner, it draws on transcribed Presidential speechesFootnote 3 and official government documents,Footnote 4 as well as ethnographic data collected during six months of field research in 2015–2016, to encapsulate and analyse the Government of Rwanda's (GoR) public discourse. The article makes a three-fold contribution to the academic literature. First, it joins a number of historically minded scholars in challenging the ‘presentism’ of existing research on Rwanda, which often produces simplified narratives of post-genocide reconstruction that portray the ruling administration's approach to nation-building as distinct from the ethnocentrism of Rwanda's past. The collective works of Catharine and David Newbury (see e.g. Newbury & Newbury Reference Newbury and Newbury2000), as well as Lemarchand (Reference Lemarchand1970), Reyntjens (Reference Reyntjens1985) and Chrétien (Reference Chrétien and Straus2006) were among the first to emphasise the frequent lack of historiographical awareness of scholarship on the region and take steps to nuance this. Likewise, Vansina (Reference Vansina2004) invited scholars to examine the ‘antecedents to modern Rwanda’ to encourage a rethinking of pre-colonial royal history that dispenses with glorified notions of social harmony, omnipotent kingships and unvanquished militaries. In a more recent contribution, Carney (Reference Carney2014) traces the deterioration of Hutu–Tutsi relations in 1959 to the evolving ethnic discourse of the Catholic Church, and specifically the White Fathers. But while these works push the boundaries of historical knowledge to unearth deeper trends over the longue durée, they do not explicitly connect these antecedents to post-genocide developments. Desrosiers & Thomson's (Reference Desrosiers and Thomson2011) article is a compelling exception, offering a comparison of pre- and post-genocide Rwanda to show continuities in the leadership of former President Juvénal Habyarimana and the incumbent Paul Kagame. Yet, in identifying sharp contrasts between the regimes regarding their ethnic composition, relationship with Rwandan peasants and respective stances on identity, Desrosiers & Thomson (Reference Desrosiers and Thomson2011: 432) inadvertently reinforce the view that the RPF's anti-ethnic posturing is distinct from previous historical periods. This article travels further down the historical chain to first, examine how ethnicity and nationhood were framed in the late colonial period, and second, to challenge the popular narrative that the RPF's contemporary rhetoric represents a clear break from the past.
The article also adds a comparative layer to accounts of nation-building in the post-genocide era by unearthing important continuities in the political elite's strategic use of rhetoric. From a theoretical perspective, it follows a burgeoning comparative literature on state survival and authoritarian resilience in its treatment of rhetorical and symbolic displays of power as central to political processes, rather than ‘epiphenomenal’ (Wedeen Reference Wedeen1998, Reference Wedeen2015: 5; see also e.g. Dunn Reference Dunn2003; Jourde Reference Jourde2005; Desrosiers Reference Desrosiers2014). As a corollary, it presents speeches, manifestos and visionary documents as ‘more than mere words’, showing how they shape constellations of power in political settings where ‘[a]uthoritarianism is never fully attained’ (Desrosiers & Thomson Reference Desrosiers and Thomson2011: 437; Desrosiers Reference Desrosiers2014: 203). It thus contributes to scholarly analyses of the various strategies authoritarian rulers adopt when confronted with leadership constraints. By offering insights into the ways that political elites frame ethnic boundaries through a longitudinal comparison of regime rhetoric on ethnicity and nationhood, the article makes a third contribution, framing rhetoric as an instrument of ethnic control and exclusion. In this way, it seeks to refine standard theories of ethnic control (see e.g. Lustick Reference Lustick1980; Yiftachel Reference Yiftachel2006; McGarry Reference McGarry, Weller and Nobbs2010), which attach no theoretical significance to the discursive dimensions of regime persistence.
The article proceeds in chronological order, analysing regime rhetoric during the late 1950s, tracing the preservation of key discourses among Tutsi in exile, and finally, examining continuities in rhetorical framing in the post-genocide era.
RHETORIC AS A RESILIENCE TOOL IN COLONIAL RWANDA, 1956–1959
The 1950s were a volatile time in Ruanda-Urundi that foreshadowed a looming crisis. In 1946, the newly founded United Nations (UN) placed Rwanda and Burundi under an international tutelle (trusteeship). Comprising members of the Belgian colonial administration, the five permanent members of the UN Security Council and three rotating countries, the responsibilities of the tutelle included examining colonial reports, debating and assessing petitions and carrying out tri-annual visits to Ruanda-Urundi. Following its first two visits, the tutelle emphasised in 1954 the need to accelerate political autonomy, and demanded that Belgium implement a 10-year development plan, which introduced democratic reforms at the local level (Carney Reference Carney2014: 46–7). Alongside this administrative shift, Mwami (King) Mutara III Rudahigwa abolished the vestigial practices of uburetwa (forced labour performed for a Tutsi chief as payment for ‘occupation’ of land) and ubuhake (corvée labour requiring a patron to lend land and cattle to a client in exchange for his service), opening the prospects for democratic elections (Pottier Reference Pottier2002: 13). The first local election by male suffrage took place in 1956. Seventy-five per cent of men voted, and the percentage of Hutu sub-chiefs rose from 50% in 1953 to 66% in 1956. Yet, Tutsi chiefs continued to control 81% of territorial council seats, 57% of colonial administrative positions and 31 of 32 seats in the Conseil Supérieur du Pays (CSP), the highest advisory body of the state (Carney Reference Carney2014: 74). With Tutsi belonging to the two royal clans, Abanyiginya and Abega, controlling 80% of chiefdoms, political office was ‘not so much a Tutsi monopoly of traditional and bureaucratic power as the closed oligarchy’ of noble Tutsi (Linden & Linden Reference Linden and Linden1977: 227).
The official discourse on ethnicity and nationhood evolved with these political developments. Prior to 1956, ethnic tensions were not strongly reflected in Rwanda's public discourse. In fact, the dominant social category of the early 1950s was ‘évolué ’ (evolved men) (Carney Reference Carney2014: 56). Conversant in French and comprised of Tutsi and former Hutu seminaries, the évolués constituted the leaders of a potential middle class with a ‘progressive outlook’ and ‘nascent bourgeois aspirations’ (Linden & Linden Reference Linden and Linden1977: 227). Rather than framing political demands in ethnic terms, they advocated for broader reforms, centring on the ‘moral, intellectual and material progress of Rwanda’ (Les Amitiés, 1951 in Carney Reference Carney2014: 48). But the 1956 legislative election contributed to the breakdown of the évolué movement. With few Hutu candidates appointed to higher positions, disgruntled Hutu leaders increasingly identified as an ethnic counter elite, and prominent Catholic newspapers aided their cause by raising taboo subjects such as the Mwami's ethnic identity and the nature of his accession (Carney Reference Carney2014: 75).
Burgeoning Hutu discontent threatened to uproot Tutsi dominance at home, making it essential for the ruling elite to adjust its rhetorical strategy. The 8th Session of the CSP signalled this shift. Between 13 and 28 June 1956, the CSP held a closed-door meeting to address the question of ‘social relations among the Banyaruanda’. Its members unanimously ‘expressed a wish’ to remove references to ‘Matutsi, Mahutu, and Matwa’ in ‘census booklets, fact sheets and official acts’ – a motion first introduced in the previous CSP deliberations in January of 1956 (Conseil Supérieur du Pays 13.08.1956: 134–5). CSP members claimed that ‘the current situation [was being] provoked by certain personalities that desire to demonstrate that munyaruanda does not exist in reality but that there are groups of races: Batutsi, Bahutu and Batwa’ (Conseil Supérieur du Pays 13.08.1956: 134). This was presented as dangerous and inopportune because it destroyed national unity. In the subsequent CSP session, Conseil members insisted that action be taken to remove racial designations from official documents, demographic surveys, identity cards, court records and all other administrative documents (Conseil Supérieur du Pays 18–21.12.1956: 1). They also requested that the Mwami make a statement to the nation denouncing the exploitation of race for political gain (Linden & Linden Reference Linden and Linden1977: 254). The Mwami's declaration was published in the Catholic newspaper, Temps Nouveaux d'Afrique, in September 1956. Themes of unity, equality and progress graced his address: he famously claimed that ‘Rwanda is home to one homogenous people where rights should be the same for everyone and not a field of disputes between racial and social factions’ (Temps Nouveaux d'Afrique 2.9.1956).
Unity-centric rhetoric was also designed to shape the perceptions of key international actors. In the late 1950s, the UN tutelle's visiting missions represented the most direct medium through which to influence the views of prominent statesmen at the UN: ‘For the first time in history, the inner workings of colonial empires were now subject – in a regular and regulated fashion – to the scrutiny of the broader international community’ (Pearson Reference Pearson2017: 536). Visiting missions generated flurries of protest and reform in colonial countries, and the 1957 tutelle to Ruanda-Urundi was no exception (Carney Reference Carney2014: 79). Rival manifestos greeted the visiting mission, with the CSP publishing the first of these, titled Mise au Point (Statement of Views). The latter was strategically crafted as a resilience tool to reinforce and buttress Tutsi supremacy. By presenting the black-white division as the salient social cleavage; by outlining the principles of an emergent anti-colonial nationalism; by minimising the plight of the Hutu; and by espousing ‘a characteristically mythical re-interpretation of the socio-political structure of Rwanda’, CSP members projected an image of Rwanda as a homogeneous and internally united nation (D'Hertefelt quoted in Mayersen Reference Mayersen2014: 122). By appealing to nascent global discourses on human rights and self-determination, they also sought to court international support for a rapid independence (Conseil Supérieur du Pays 1957: 2). The Hutu counter-elite vigorously protested these claims in their response, ‘Note sur l'aspect social du problème racial indigène au Ruanda’, better known as the ‘Bahutu Manifesto’. The manifesto espoused a number of key priorities that contrasted sharply with the Mise au Point, including: ongoing support for European intervention, without which it claimed Hutu would be ‘subjected to more inhumane exploitation’; vocal condemnation of the Tutsi elite; and support for majoritarian democracy. The manifesto also ‘energetically protested’ the removal of racial indicators from identity cards on the grounds that their eradication would distort measurable indictors of reform (Bahutu Manifesto 1957: 6). While Belgium displayed ambivalence to the Manifesto, conferring tacit approval for enduring Tutsi privilege, the tutelle recognised its disintegrative potential, and reduced the ‘Mise au point’ to a veiled attempt, ‘to strengthen, by a premature autonomy, the changing prerogatives of the dominant class’ (UN 1957: 19; Lemarchand Reference Lemarchand1970: 150). Although its proposed solution to educate and ‘enlighten’ irksome elements of the new and old elite smacked of ‘almost ridiculous optimism’, the mission's report signalled a shift in international perceptions of Rwanda's Indigenous authorities (Wagoner Reference Wagoner1968: 161–2).
By 1957–58, Catholic authorities ceded control of Church-sponsored channels of communication to Hutu politicians, and Rwanda's ethnic divisions rose to the top of its political agenda (Lemarchand Reference Lemarchand1970: 106–7). Kinyamateka, the most popular vernacular newspaper, ‘became the official mouthpiece of the Hutu movement’, while Temps Nouveaux d'Afrique, a European newspaper published in Bujumbura under the auspices of the White Fathers (Roman Catholic, ‘Society of the Missionaries of Africa’), brought international attention to Hutu demands for democratic reforms (Lemarchand Reference Lemarchand1970: 108–9). Even so, the Tutsi elite did its best to ignore rising tensions. The 25th anniversary of the Mwami's accession prompted a lavish three-day celebration in which the colonial Resident of Rwanda, a CSP representative and the Mwami gave speeches. No one mentioned the Hutu–Tutsi problem (Perraudin Reference Perraudin2003: 143). Instead, the affair degenerated into a ‘boisterous celebration of the monarchy and traditional Rwandan culture’, replete with tributes to Belgian administrators and Tutsi chiefs. Prompting a heated reaction from certain politicians and seminarians, the Mwami was eventually compelled to address the elephant in the room (Carney Reference Carney2014: 84). In March 1958, he established a Special Commission to study ‘the Social Aspect of the Mututsi-Muhutu’, composed of six signatories of the Bahutu Manifesto and eight members of the CSP. Gathering twice in April 1958 and again in June, the transcripts of the Special Commission highlight stark differences in the political rhetoric of CSP members and Hutu representatives.
Catholic Hutu leaders, Calliope Mulindahabi and Joseph Habyarimana Gitera, both signatories of the Bahutu Manifesto, demanded to know, ‘why the three ethnic groups do not enjoy the same advantages in the administrative domain of the country?’. Both advocated for the inclusion of Hutu representatives in the CSP to offset the Tutsi-dominant ‘feudal system of courtesan culture and favouritism’ (Comité de l’étude 1958: 5). In response, CSP members made use of the language of anti-ethnic and anti-colonial, pan-Rwandan nationalism, arguing that disparities in political power would erode as educational opportunities improved, Rwanda democratised and colonial Belgium withdrew. For example, Tutsi Chief, Pierre Mungalurire dismissed Gitera's claim that the CSP was ‘constituted only by Tutsis’ (Deuxième séance 1958: 5), accusing Gitera of pursuing ‘division and separation’, a ‘dangerous’ combination that ‘could create a regrettable precedent for the future of our country’ (Deuxième séance 1958: 3). He urged fellow committee members to reject ethnic representation on the grounds that it would produce ‘clanic compartments’, harmful to unity (Deuxième séance 1958: 3). Tutsi Chief, Michel Kayihura voiced similar concerns about the threat of racial divisions to national unity, claiming: ‘The nations of today want to remove the borders between them … and Mr. Gitera, with his system, wants to create borders in one and the same country; it's evident that this system is contrary to the progressive ideas not just of our nation, but of modern and civilised nations’ (Comité de l’étude 1958: 20). During the second session, Tutsi reformer, Prosper Bwanakweri, focused on the nature of ethnic identities, insisting that the Hutu–Tutsi divide was a social rather than a racial distinction. He thus encouraged Rwandans to avoid speaking in collective racial terms given that the majority of Tutsi were also poor, simple men (Deuxième séance 1958: 13).
In the end, the commission put forth modest recommendations to enhance Hutu representation in education and political administration, but on the Mwami's advice, the CSP rejected them all (Carney Reference Carney2014: 90–1). Rudahigwa's accompanying statement insisted on the fabrication of the ‘Hutu–Tutsi problem’ by Hutu polemicists and foreign influencers. ‘The enemies of this country shall not succeed’, he proclaimed: ‘the entire country has come together to seek out the bad tree that has produced the bad fruits of division. When it is found, it will be cut down, uprooted and burnt’ (Temps Nouveaux d'Afrique 6.7.1958: 5). By projecting an image of Rwanda as a place of social harmony and mobility and by stressing themes of national unity, patriotism and collective strength while denying ethnic distinctions, the Mwami sought to ‘court – at least rhetorically broader constituencies and alliances’ among modern and traditional sectors of Rwandan society (Desrosiers Reference Desrosiers2014: 209). By depicting Hutu leaders as ‘enemies of the state’, moreover, he aspired (unsuccessfully) to delegitimise Hutu demands for political reform and contain the threat the counter movement posed to the Tutsi nobility.
Between 1957 and 1959, political parties began to reflect the deep ideological divisions between the established elite and the emergent Hutu counter-elite. Hutu leader and later President of Rwanda (1962–73), Grégoire Kayibanda founded the Mouvement Social Muhutu (MSM) in June 1957 (renamed PARMEHUTU in 1959), which embraced a pro-European, pro-clerical and anti-monarchist stance, and called for the immediate advancement of the Hutu ‘race-class’ in all aspects of social, political and economic life, coupled with the destruction of ‘the existing feudal system instituted by the Tutsi’ (Comité de Parti Nationale Parmehutu 1959). Fundamental differences of opinion about whether to direct demands against all Tutsi or the high aristocracy drove a wedge between the leadership (Lemarchand Reference Lemarchand1970: 151). In November 1957, the now rabidly anti-Tutsi, Joseph Gitera, launched MSM's rival, the Association pour la promotion sociale de la masse (APROSOMA), which championed an explicitly xenophobic rhetoric (Carney Reference Carney2014: 84). Ironically, APROSOMA would become the more moderate of the Hutu parties, espousing a pan-ethnic ideal (Newbury Reference Newbury1998: 13).
The institutionalisation of Hutu consciousness in the form of two well organised political parties posed an ominous threat to the colonial privilege of elite Tutsi, and ‘brought Tutsi unity under pressure’ (Mamdani Reference Mamdani2001: 119). This splintering of Tutsi politics crystallised into two rival political parties. As the name implies, the predominantly but not wholly Tutsi and monarchist party, Union National Rwandaise (UNAR) adopted a similar tactical approach to the CSP. Emerging in the immediate aftermath of the mysterious death of Rudahigwa in July 1959, UNAR produced a natural alliance between the Tutsi aristocracy, Tutsi chiefs and Muslim traders, many of whom believed the Belgian authority assassinated Mutara, ‘to replace an obdurate Rwandan king with a more compliant pro-Western figure’ (Carney Reference Carney2014: 106). Wedded to Tutsi supremacy at home and ‘deliberately playing up the sacredness and invincibility of the monarchy’ (Lemarchand Reference Lemarchand1970: 159), UNAR charted a fervently anti-colonial and anti-Catholic programme. It accused the Belgian administration and European missionaries of creating the Hutu–Tutsi divide, ‘exploit[ing] our country [and] destroy[ing] our ancestral customs’ to delay Rwanda's rightful independence (Rukeba, quoted in Lemarchard 1970: 159). This rhetorical strategy allowed UNAR to cultivate a counter-image of itself as the only party able to ‘restore these customs, to shake off the yoke of Belgian colonialism to reconquer Rwanda's independence’ (Rukeba, quoted in Lemarchard 1970: 159). UNAR disguised its ethnic favouritism by nominally appointing the outspoken Hutu leader, François Rukeba as President. It also outwardly rejected racialised rhetoric in favour of the discourse of national unity. ‘Unity among all Banyarwandans’ was presented as the foundation of progress. ‘Chauvinism and clanism’, created by colonialists, were the enemy of both UNAR, the Mwami and by extension, the Rwandan nation (Comité de Parti Nationale UNAR 1959). In its circular, distributed three days after the party's creation, UNAR partisans chanted: ‘Children of Rwanda! Subjects of Kigeri, rise up! Let us unite our strengths! Do not let the blood of Rwanda be spilled in vain. There are no Tutsi, Hutu, or Twa. We are all brothers! We are all descendants of Kinyarwanda!’ (quoted in Lemarchand Reference Lemarchand1970: 161).
In contrast, the Rassemblement Démocratique Ruandais (RADER) adopted a reformist stance on the issue of Tutsi ascendance and a soft line against Belgian colonial occupation (Lemarchand Reference Lemarchand1970: 120–1). Led by CSP member and noted Tutsi reformer, Chief Bwanakweli, RADER strove ‘to work toward the realisation of a social, economic, political and cultural order based on authentic democracy and harmony among the constituent groups of Rwanda’ (quoted in Lemarchand Reference Lemarchand1970: 160). It received the bulk of its support from university students who championed progressive ideas (Lemarchand Reference Lemarchand1970: 160).
By the autumn of 1959, when all four parties mobilised to contest UN-sponsored elections, Rwanda ‘was a land of tensions, rumours, and troubles’ (Codere quoted in Mayersen Reference Mayersen2014: 123). Unable to quash the rise of Hutu nationalism with its rallying cries for unity and ethnic non-recognition, UNAR turned to violence and intimidation of its political opponents (Wagoner Reference Wagoner1968: 186). It gravely miscalculated. When news broke that a band of UNAR militants attacked a Hutu sub-chief, violence spread throughout the country ‘with the spontaneity and intensity of a tropical thunderstorm’ (Linden & Linden Reference Linden and Linden1977: 267). Popular discontent among the Hutu masses ‘was given a specific focus … and a new mythology … associated with republican ideals’ – and directed against Tutsi privilege and monarchical rule (Lemarchand Reference Lemarchand1970: 168). The Belgian administration played a decisive role in neutralising resistance to the revolutionary movement, signalling a dramatic shift in allegiance. In reality, Belgium's about-face was more gradual than much of the historical commentary suggests, for even as violence destabilised the country, UNAR continued to reduce Hutu discontent to a recalcitrant minority of ‘hotheads’ and championed a virulent anti-Belgian and anti-clerical ideology, driving the Belgians to favour the Hutu parties (Straus Reference Straus2006: 183). As thousands of Tutsi fled the country, Belgium recognised UNAR's shrinking support base and forged ahead with Rwanda's first communal elections in July 1960, against the recommendations of the UN tutelle. PARMEHUTU achieved a landslide victory, and shortly thereafter, Hutu elites orchestrated the ‘Coup d’état de Gitarama’, proclaiming the abolition of the monarchy and the creation of a republic with clandestine support from the Belgian administration (Lemarchand Reference Lemarchand1970: 109).
REVIVAL OR SURVIVAL? THE EMERGENCE OF THE RPF IN EXILE
The instantiation of Rwanda's First Republic (1962–73) marked a seismic shift in the distribution of power inside the country, giving rise to a rarely questioned, ‘fiercely ethnocentric portrait of the Kayibanda years’, which obscures the continuation of powerful counter-narratives that circulated both inside and outside the country (Desrosiers Reference Desrosiers2014: 201). Indeed, as the centre of Tutsi politics shifted from home to exile, the same rhetorical frames the Tutsi elite espoused in the late 1950s continued to circulate among refugees and their descendants and were eventually embraced by the founding members of the RPF.
Kayibanda's early years were turbulent and violent as corruption and ineptitude fuelled popular discontent and unleashed regional divisions between Hutu in the south and central regions and Hutu in the north of the country. Kayibanda also encountered resistance from organised Tutsi opponents and militants in exile, who fled ethnically charged violence in Rwanda, ‘at first in an almost continuous if irregular stream between 1959–1964’, and then again in a more limited fashion between 1972–73 (Prunier Reference Prunier1995: 61). Most of the former aristocratic elite, including prominent members of the royal family and UNAR, fled to Uganda. Many settled in refugee camps in the Mbarara region with no idea that they would remain in exile for nearly 30 years (Jones Reference Jones, Campioni and Noack2012: 239). Capitalising on their ‘burgeoning sense of grievance’ (Lemarchand Reference Lemarchand1970: 172), UNAR created a government-in-exile with François Rukeba, the original President of UNAR, as Prime Minister and Pierre Mungalurire, former member of the CSP and Special Commission, as the Minister of Finance. The government-in-exile sought to destabilise the fledging Hutu administration by forging key alliances with regional and international actors, including the Mouvement National Congolais in the Democratic Republic of the Congo, the People's Republic of China and other communist countries in the UN Trusteeship. Rather than showing open ethnocentrism toward Hutu and the Hutu regime, however, UNAR continued with its moderate and strategic rhetoric, projecting an image of pre-colonial Rwanda as a place of cultural unity and social mobility, tracing the roots of ethnic discord to colonial Belgium and the Catholic Church, and condemning the ‘racial’ privileges of Europeans in the colony (Mamdani Reference Mamdani2001: 120).
Virtually all Tutsi refugees were converted into stalwart supporters of UNAR by the early 1960s. As time went on, however, the exiled UNAR leadership succumbed to political infighting and factionalism between ‘Mutaristes’ (supporters of the monarchy), and ‘progressivists’ (proponents of Republicanism) (Lemarchand Reference Lemarchand1970: 199–205). Restrictions imposed on the refugee leadership by Milton Obote's government in Uganda further handicapped their capacity to execute a successful counter-revolution (Lemarchand Reference Lemarchand1970: 207–9). As memories faded, ‘Rwanda slowly became a mythical country’, representing an idyllic ‘land of milk and honey’ for the younger generation raised in exile and influenced by UNAR ideology (Prunier Reference Prunier1995: 66). Despite their geographic dispersion, members of the Tutsi diaspora remained in contact through social clubs and cultural associations (Prunier Reference Prunier1995: 66). Perhaps the most significant Tutsi diaspora movement in Uganda, called ‘the Rwandese Refugee Welfare Foundation’, was created in 1979 and renamed the ‘Rwandese National Alliance for Unity’ (RANU) the following year. RANUs intellectual fathers called for the creation of a socialist state in Rwanda that would enable refugees to return home. They admired the struggles of Mao Tse Tung and Che Guevara, as well as pan-African visionaries such as Kwame Nkrumah and Thomas Sankara (Reed Reference Reed1996: 484; Waugh Reference Waugh2004: 37). But the association of RANU with UNAR is also strong, with notable parallels in their rhetorical calls for unity, self-determination and anti-ethnic, pan-Rwandan nationalism. The curious choice of the acronym ‘RANU’, which is UNAR spelled backwards, further suggests an attempt to preserve UNARist ideology whilst camouflaging its external appearance (Nyirubugara Reference Nyirubugara2013: 118). Several of RANU's founders claim the association is merely coincidental, but regardless of the name's origins, UNAR's political platform probably influenced the convictions of RANU's founding fathers.
RANU surfaced in response to rising xenophobia and anti-Rwandan sentiment in Uganda, which accompanied the final years of Idi Amin's presidency, and escalated once Milton Obote took power for the second time in 1980, after which he labelled Rwandan refugees ‘foreigners’ and ‘natural allies’ of Yoweri Museveni, his main political foe (Mamdani Reference Mamdani2001: 167–8; Waugh Reference Waugh2004: 36–7). RANU's leaders were therefore rendered toothless shortly after its inception, driving the movement underground into neighbouring Kenya. Obote's open hostility and organised repression of Rwandan refugees in western Uganda, most notably in 1982 (see e.g. Mamdani Reference Mamdani2001: 168–9), prompted many angry young Rwandan men to swell the ranks of RANU in the mid 1980s, as well as Museveni's National Resistance Army (NRA). By the time it captured Kampala in 1986, the NRA had grown to 14,000 soldiers, 4000 of which were Rwandan-born, many of whom secured key posts in government, the army and business after Museveni ascended the presidency. Faced with growing anti-Tutsi sentiment, however, Museveni began to purge his administration of Rwandans, including his Deputy Chief of Military Intelligence, Paul Kagame. In this changing political climate, RANU's original socialist ideology was no longer palatable to its new members who were generally Catholic, socially conservative and not highly educated. The ‘second generation of Tutsi exiles’ therefore laid down a different strategy than their antecedents, thoroughly refining the revolutionary project from RANU's creation in 1980 to its transformation in 1987 into a politico-militant ‘Front’ dedicated to the return of exiles to Rwanda by force if necessary (Prunier Reference Prunier1995: 71–3; Lemarchand Reference Lemarchand2009: 38).
The change in name was accompanied by a second politically loaded name ‘Inkotanyi’, which has monarchist and therefore Tutsi overtones, stemming from the ruthless and bellicose Mwami KIGERI IV Rwabugiri, who ruled under the epithet, ‘Inkotanyi cyane’ (the Redoubtable Fighter) (Nyirubugara Reference Nyirubugara2013: 71, supra. 62). Curiously, RANU was decidedly anti-monarchist from the outset, in part because it feared the influence of the exiled Mwami KIGERI V Ndahindurwa (half-brother to the aforementioned Rudahigwa), who opposed the RPF's politico-militant ambitions. The use of the ‘Inkotanyi’ label therefore suggests that the leadership either was not aware of this historical reference, or more likely that it sought to preserve the monarchy's symbolism while disassociating from its contemporary guise. It thus adopted a highly tactical coded message, harkening back to Rwanda's ‘golden age’ to draw a historical continuation between the pre-colonial royal court and the RPF rebels in exile (Nyirubugara Reference Nyirubugara2013: 118–19).
Three years later, the Rwandan Patriotic Army, the armed faction of the Front, invaded Rwanda from its base in Uganda, igniting a four-year long civil war between the RPF and Habyarimana's armed forces. Under pressure by 1993 to negotiate a democratic solution, Habyarimana alienated and emboldened an increasingly powerful extremist faction within his regime. His assassination on 6 April 1994 provided the immediate catalyst for the outbreak of violence, as Hutu Power extremists took control of the state and began systematically murdering political opponents within 24 hours of the President's death. In the following days, violence spread throughout central and southern Rwanda, with extremist Hutu militias, soldiers and ordinary citizens indiscriminately targeting Tutsi civilians, as well as Hutu and Twa who attempted to stop the violence, in a frenzy of genocidal killing, which claimed the lives of an estimated 500,000 to 800,000 Rwandans – overwhelmingly Tutsi – in 100 days. In July 1994, the RPF took control of Rwanda and brought a swift end to the genocide (Jessee Reference Jessee2017: 10).
RHETORIC AS A RESILIENCE TOOL IN POST-GENOCIDE RWANDA
Seismic shifts in Rwanda's politics occurred during the immediate aftermath of the genocide. By August 1995, the RPF had concentrated power into the hands of its immediate constituency, privileging Anglophone Tutsi returnees from Uganda who fled Rwanda during the social revolution of 1959 or during episodes of violence against Tutsi in the post-independence period (Thomson Reference Thomson2013: 105). Alluding to these realities in a radio broadcast before stepping down in 2000, the nominal Hutu President Pasteur Bizimungu described himself as a ‘Hutu de service’ who served ‘encadré par des fidèles de Kagame’ (quoted in Zorbas Reference Zorbas2004: 45). By the end of the 1990s, RPF loyalists filled approximately 70% of government posts and the military and intelligence services were almost exclusively controlled by ‘a minority of a Tutsi minority’, echoing the oligarchic system in colonial Rwanda (Reyntjens Reference Reyntjens2015: 21).
To suppress popular misgivings about what Reyntjens (Reference Reyntjens2015) calls the ‘Tutsification’ of power and the foreign origins of the RPF, the regime espouses a coherent and consistent political rhetoric on national unity, ‘ethnic amnesia’ and historical revisionism, which mirrors the rhetorical strategy of the Tutsi elite in the 1950s, and UNAR in exile (Lemarchand Reference Lemarchand1994: 30). Whereas the political conditions in the late 1950s limited the popular appeal of integrationist rhetoric among the Hutu masses, key changes in the socio-political landscape concerning the distribution of power (colonial/post-colonial) and the magnitude and scale of anti-Tutsi violence (pre/post-genocide) have strengthened the pertinence of these rhetorical themes in the post-genocide era. The remainder of this section explores how the post-genocide regime instrumentalises Rwanda's history to shape state-society relations in ways that ferret out counter-narratives about the relationship between ethnicity, nationhood and power, while simultaneously exploiting global discourses on civic nationalism and self-determination to build a positive external image of Rwanda that disguises the concentration of power along ethnic lines.
Unlike the Indigenous authorities in the late 1950s and UNAR in exile, the GoR outwardly adopts an anti-monarchist stance, fearing the lingering power of monarchical attachments inside Rwanda. Yet, frequent references to Rwanda's pre-colonial (Tutsi) dynasties suggest the regime has preserved and repackaged key elements of monarchist symbolism, albeit in disguised form, akin to RANU and the RPF in exile. The official rhetoric fashions the Nyiginya Kingdom, which ruled Rwanda for 200 years prior to European arrival, as Rwanda's ‘golden age’ and the high point of its socio-political development. Commanded by a formidable series of kings (Mwami) and their elite warriors (Intore, ‘chosen ones’), regime rhetoric emphasises Rwanda's military fortitude at the height of Nyiginya power, while also highlighting the central role of its warrior kings in fostering ‘strong unity’ and harmonious relations among ‘Hutu’, ‘Tutsi’ and ‘Twa’ (GoR 1999: 6). By painting the entire colonial and post-colonial periods as ‘dark ages’, the regime implies that Rwandans lost their true identity and adopted the divisive practices of their (neo)colonial oppressors. Couched in the vocabulary of tradition and cultural authenticity, the regime echoes UNAR's party platform in 1959, emphasising Rwanda's ‘cultural roots’ (RoR 2000: 4), and drawing from its ‘centuries-old history the positive values which characterised our ancestors’ (RoR 2003: art. 8 of preamble) to ‘[m]ake cultural heritage a determining factor for national identity and pride’ (RoR 2008: 19). Officially characterised as a ‘cultural school’ (RoR 2015) and ‘a holistic training system’ (Ntirenganya Reference Ntirenganya2021: 1), the GoR's Itorero programme is the primary vehicle through which citizens engage with Rwanda's pre-colonial history and culture. According to Sundberg (Reference Sundberg2016), the programme links ‘model citizenship’ to the Intore ethos found in Rwanda's ancient heritage as a military powerhouse. As such, it establishes a ‘historical continuation’ between the pre-colonial royal court, the exiled RPF rebels and the current state leadership (Sundberg Reference Sundberg2016: 84). In this imaginary, the rebels, who invaded Rwanda from Uganda in 1990, are depicted as modern incarnations of Intore who fought a heroic ‘liberation struggle’ to rid Rwanda of its destructive political leadership. As the current leader of the RPF, the state and the military, and the former commander of the ‘liberation armed struggle’, President Kagame is fashioned as the ‘hero leader’ and ‘Father of Rwanda’, personifying its ancient kings and epitomising its contemporary citizen ideal (see Sundberg Reference Sundberg2016Reference Sundberg2016: 63–98).
Rather than representing an ethnically neutral conception of state and society, however, ethnicity runs deep within this narrative, for the association of the Rwandan monarchy, the Intore and the RPF with elite Tutsi remains strong in the popular consciousness. Keen to mask this association, the GoR follows the Tutsi administration in the late colonial era and promotes the revisionist notion that national unity is ‘a traditional value which must be reasserted, reinforced and taught to all Rwandans’ (GoR quoted in Desrosiers & Thomson Reference Desrosiers and Thomson2011: 144). The Ndi Umunyarwanda (‘I am Rwandan’) programme, targeted at Rwandan youth, champions a homogenising conception of Rwandan identity, encapsulated by the near-ubiquitous Itorero slogan: ‘one people, one vision, one nation, one team’ (quoted in Sundberg Reference Sundberg2016: 80). The flipside of this process is to present ethnicity as a sinister foreign construct linked to the racist ideology of the Belgian colonial authority and the Catholic Church. The GoR delegitimises ethnic identity on this basis, echoing the CSP's counsel in 1958. The 2001 Organic Law, for example, criminalises ‘the use of any speech, written statement or action based on ethnicity’, and the 2008 Law ‘Relating to the Punishment of the Crime of Genocide Ideology’ defines the latter in sweeping terms to render ethnic labelling illegal with severe punitive consequences, including possible imprisonment of up to 25 years and fines ranging from 200,000 to 5 million RWF (approximately US$5100) (art. 4, 6).
In addition to these legal measures, the GoR relies on rhetorical and symbolic devices, including vision and policy statements, presidential speeches and government programmes to project an image of unity as distinct from the divisionism of Rwanda's recent past. Forming the backbone of the state's development enterprise, Rwanda's ‘Vision 2020’ document draws this contrast in stark terms: ‘In general our country has been badly governed during most of the last century. Both the colonial and post-colonial regimes ruled Rwanda on the basis of favouritism and prejudice. As a consequence, Rwanda is one of the poorest countries in the world and we, the people of Rwanda, were until recently, strongly divided’ (RoR 2000: 2, emphasis mine). Kagame's 2003 ‘Inaugural Address’ echoes similar themes of historic disunity, insidious foreign interference and a newly restored national consciousness: ‘In the aftermath of the genocide’, Kagame decries, ‘the so-called experts on Rwanda erroneously preached separate homelands for different sections of society. But we established a Government of National Unity to build on the achievements of earlier periods in our history, when Rwanda was one and united’ (quoted in Vandeginste & Huyse Reference Vandeginste, Huyse, Marysse and Reyntjens2005: 121, supranote 46). Likewise, his 2010 address emphasises the importance of ‘national cohesion’, and of ‘reviving the cultural bonds on which [Rwandan] society was founded … to restore a dignified nation’. Using similar language to the CSP in 1958, Kagame praises the Rwandan people for rejecting the demands of other nations to organise Rwandan politics along ethnic lines, and for recognising that ‘these so-called ethnic divisions are not inherently Rwandan classifications, but rather colonial constructs’.
Every Rwandan, even in the most remote communities, is conversant in this official discourse, ‘if only because all are subjected to a constant barrage of its rhetorical iterations’ (Wedeen Reference Wedeen1998: 504). This generalised familiarity has instilled an impressive degree of self-censorship on the rural population, particularly concerning matters of ethnicity and national identity (see e.g. Straus & Waldorf Reference Straus and Waldorf2011; Reyntjens Reference Reyntjens2013, Reference Reyntjens2015; Thomson Reference Thomson2013). However, this does not imply passive internalisation on the part of ordinary Rwandans. On the contrary, Rwandans find creative ways of enacting a ‘quiet agency’ (Grant Reference Grant2015) to articulate critiques of power through ‘everyday acts’ of ‘subtle, indirect and non-confrontational’ resistance (Thomson Reference Thomson2013: 129; also see e.g. Breed Reference Breed2013/2014; Mwambari Reference Mwambari2019; Waldorf Reference Waldorf2006). Nonetheless, this widespread superficial adherence to the regime's ‘guidelines for politically acceptable speech and behaviour’ is a testament to the GoR's extraordinary capacity to regulate public conduct through speech (Wedeen Reference Wedeen1998: 520). By controlling ‘the symbolic context within which struggles over the meaning of the nation, of selfhood, and of … political power’ take place, the regime entrenches ‘norms of order and obedience’, which function as an ‘effective sub-system of coercive control’ (Wedeen Reference Wedeen1998: 520; Desrosiers & Thomson Reference Desrosiers and Thomson2011: 430–1).
Contextual factors help explain why a similar tactic failed to curb the rise of Hutu nationalism in the 1950s, ultimately leading to a shift in the balance of power inside Rwanda. In particular, the GoR's official discourse on the genocide, centring on the annual commemoration period known as Kwibuka,Footnote 5 stands in tension with its ‘policy of ethnic non-recognition’, and reflects a strategic choice on the part of the leadership to instrumentalise the genocide for political gain: an option that was not available to elite Tutsi in the 1950s (Baldwin Reference Baldwin2019: 355). With Tutsi privilege embedded into the fabric of the Belgian colonial system, the native administration had no such reference on which to establish collective victimhood, which raised questions about elite demands to eradicate ethnicity in the absence of radical change to the institutional framework of the Rwandan state. In contrast, the blurring of Tutsi identity with ‘survivor’ identity in the post-genocide era has resulted in a state-sanctioned version of what Baldwin (Reference Baldwin2019) calls, ‘survivor nationalism’, associated with an ‘explicitly Tutsi … history of violence, marginalisation, [and] discrimination’ (358). As a corollary, the decision to change the official name of ‘the genocide’ to ‘the genocide against the Tutsi’ in a 2008 constitutional amendment inadvertently re-inscribes ethnicity onto Rwanda's national narrative, thereby lumping all Hutu into the reviled residual category of ‘non-survivor’. The fact that only Tutsi were recipients of ‘justice’ during the gacaca trials – the transitional justice mechanism used to try genocide suspects – is further evidence of ethnicised guilt and innocence (see e.g. Longman Reference Longman2017: 91–134). Hutu and Twa experiences of hardship, sacrifice and war-related suffering, often at the hands of the RPF, are erased from the national memory, implying that all Hutu who opposed the genocide were killed, and all living Hutu were either active participants in the killings, passive bystanders, or if born in the post-1994 era, culpable by association (Mamdani Reference Mamdani2001: 267). This dual strategy of denying ethnicity while elevating survivor status disguises and simultaneously entrenches the strong ethnic and regional bias of the RPF regime. Aforementioned legislation on ‘divisionism’, ‘sectarianism’ and ‘genocide ideology’ serves as ‘coded language’ to convey the consequences of challenging the official genocide narrative.
Likewise, shifts in the external context help explain why the RPF's rhetoric facilitates the development of strong bilateral relations even though a similar rhetorical strategy failed to mobilise international support for elite Tutsi in the late 1950s. The early leadership underestimated international sympathy for the plight of the Hutu majority, and misjudged Belgium's influence on political developments inside Rwanda. In post-genocide Rwanda, the RPF confronts different external pressures. Aid dependent, and eager to build strategic linkages with powerful international actors, it espouses a ‘civic’ model of nation-building to appeal to donor countries and aid workers alike, (re)emphasising themes of unity, ethnicity denial and pan-Rwandan identity. But while Kagame avoided condemnation of his foreign and domestic politics for nearly 15 years, the regime has come under increasing scrutiny since 2009 (Mann & Berry Reference Mann and Berry2015: 129). As noted by Mann & Berry (Reference Mann and Berry2015), Freedom House (2019) ranks Rwanda as ‘Not Free’ and the UN Group of Experts in the Congo implicates the GoR in the M23 rebel movement in Eastern Kivu (OHCHR 2010). The 2014 BBC documentary on ‘Rwanda's Untold Story’ (Conroy Reference Conroy2014) also accuses the RPF of committing a litany of human rights abuses. In response to these accusations, the GoR has not changed its rhetorical strategy. Instead, it uses what Reyntjens (Reference Reyntjens2015) calls its ‘genocide credit’ to remind the international community of the RPF's role in stopping the genocide in the midst of international inaction, thus endowing the GoR with ‘a nearly unassailable moral high ground’ (Reyntjens Reference Reyntjens2015: 27). Akin to its domestic politics, the ‘Tutsi genocide’ has become the most powerful historical reference in the GoR's rhetorical arsenal, for it allows the regime to present (Tutsi) ‘survivors’ as the only legitimate ‘subjects of (inter)national sympathy’, while ‘invok[ing] the spectre of past violence as a reminder that it could happen again’ (Baldwin Reference Baldwin2019: 356, 358). By framing itself as the political force that liberated Rwanda, the regime also defines the RPF in strategic terms as the only authority capable of preventing Rwanda's relapse. Carefully crafted, Kagame's heavily publicised commemoration addresses function as important communication devices for prominent world leaders, statesmen and foreign dignitaries. On the 20th anniversary of the genocide of Tutsi in 2014, for example, he traced the historical roots of the genocide to external interference in Rwanda's political affairs:
The people who planned and carried out the Genocide were Rwandans, but the history and root causes go beyond this country … [to] external actors and institutions … The most devastating legacy of European control of Rwanda was the transformation of social distinctions into so-called ‘races’ … This was the beginning of the genocide against the Tutsi, as we saw it twenty years ago. With the full participation of Belgian officials and Catholic institutions, this invented history was made the only basis of political organisation, as if there was no other way to govern and develop society. The result was a country perpetually on the verge of genocide.
The subsequent theatrical performance, ‘Shadows of Memory’, performed at Amahoro Stadium during the 2014 commemoration, also executed important political work both inside Rwanda and on the international stage (Azeda 11.4.Reference Azeda2014). This graphic re-enactment of Rwanda's history, culminating in genocide, begins with the scattering of performers in all directions to represent the destructive impact of the colonial power on Rwanda's national unity. Trolleying rapidly through the period of ‘dehumanisation’ and ‘bad govern[ance]’ said to characterise the post-Independence era, colonisers swap their safari hats for blue berets, exiting the stadium just as the killing begins (6:45). Hundreds of performers convulse in a terrifying portrayal of mass annihilation (8:14–10:20). Against this backdrop, tens of RPF soldiers enter the stadium to a rapturous applause and run the width of the track in perfect unison. They awaken and ‘save’ the listless bodies and sing in perfect harmony as the nation is made whole again (10:21–23:40). Perched on his executive chair, President Kagame observes the performance with his characteristic measured stoicism. Stern, proud and humourless, he is portrayed as the archetype of the ‘new’ African leader, with the vision and determination to move Rwanda beyond genocide to a country that ‘can hold its head high in the international arena’ (Minister of Local Government, quoted in Desrosiers & Thomson Reference Desrosiers and Thomson2011: 38).
CONCLUSION
Rwanda has created a powerful and savvy communication machine, which embeds its president and prominent members of the RPF in ‘a cult of personality’ (Wedeen Reference Wedeen2015: 4). Although he lacks the emphatic charisma of many exemplary leaders, Kagame is presented as a ‘Spartan’ (Soudan Reference Soudan2015: iv), ‘exacting’ (Crisafulli & Redmond Reference Crisafulli and Redmond2012: 3) and ‘heroic’ (Fox 2013) leader who is ‘beyond ethnicity’ and able to guide the country in the right direction towards a prosperous future. This fetishising of Kagame is an effective resilience technique, which forms an integral component of the RPF's broader rhetorical strategy. Kagame is made to appear different from his predecessors, with his approach to nation-building represented as historically ‘exceptional’. Challenging this narrative, this article unearths striking rhetorical continuities between pre-Independence and post-genocide Rwanda. Both then and now, leaders deploy rhetoric that exaggerates pre-colonial social mobility, emphasises national unity and presents ethnicity as inherently divisive. In both historical periods, moreover, themes of national unity and ‘ethnic amnesia’ are instrumentalised to suppress dissident ideas about the relationship between ethnicity, nationhood and power in an attempt to thwart domestic opposition to regime elites. This same rhetoric positions the political elite as the architects of progress to manage international perceptions and conceal the ‘Tutsification’ of power (Reyntjens Reference Reyntjens2015: 21).
But exploring these trends over the longue durée makes a broader contribution. So often, the region is viewed as violent, volatile and chaotic, prompting scholars, practitioners and journalists to ‘read history forward’ through the lenses of tribalism, ethnicity and genocide, resulting in a ‘loss of historical texture’ (Desrosiers Reference Desrosiers2014: 200). The perspective offered here sheds light on the strategic underpinnings of regime rhetoric as a power and control management device and dispels myths of post-genocide ‘exceptionalism’. It also invites scholars to become more attuned to the causal complexity of authoritarian persistence, particularly when ethnicity is tightly linked to the expression of political power. Rather than conceptualising ‘authoritarianism’ in systemic terms as a ‘regime type’ or ‘residual category’ into which ‘non-democracies’ fall, this article follows Desrosiers (Reference Desrosiers2014: 203) in viewing authoritarian systems as dynamic and ‘dialectic’ systems involving ongoing tension between ‘forces of control and forces of opposition and resistance’. Rhetoric is modelled as one of several under-theorised resilience strategies that regimes adopt to manage complex challenges and constraints with varying degrees of success. Indeed, by contextualising and historicising regime rhetoric, the analysis demonstrates that similar representations can yield dramatically different results depending on the logic of the situation and the particular confluence of domestic and international variables. Subsequent research should look beyond Rwanda to the wider region to answer broader questions about how regimes construct identity discourses, who communicates, and how elites react to emergent counter-narratives. More research is also needed to understand the moments of choice and change that occur in different institutional settings to identify the conditions that produce subtle and sometimes dramatic deviations from established rhetorical paths.
