This article reflects critically on scholarly engagement with the past and with notions of authority in the African Great Lakes (GL). Its principal concern lies in the ways in which history has been and can be used to shed light on the present. Contemporary studies on the GL typically begin with a historical overview of the region. Whether talking about a ‘chain’ of ethnic violence or about the heritage of historical modes of political authority, the past is an obligatory point of reference. Work on the region, however, is significantly overdetermined by the present or the recent past; the historical summaries provide cursory glances rather than in-depth analyses of the region's past events and trends. The key consequence of this is that concepts such as ‘authority’ and our assumptions about what matters are greatly oversimplified in comparison to the rich past from which they supposedly emerge.
This tendency to simplify the past has long been decried by some of the GL's key observers, historians and non-historians alike (e.g. Schoenbrun Reference Schoenbrun1993; Newbury and Newbury Reference Newbury and Newbury2000; Reyntjens Reference Reyntjens2018). Their work established the common consensus that, in Rwanda, Burundi and neighbouring states, ethnicity, conflict and power (the themes that often draw researchers to the region) had to be understood in their historical context. Yet when it comes to dominant discussions of the present and very recent past, this consensus is reduced to a cursory acknowledgement. History is framed as an anticipatory echo of the crises of genocide and war that unfolded across the region in the 1990s and their consequences today. Put simply, the past is understood through the prism of the present. We believe that challenging this ‘presentist’ approach and rediscovering history beyond the ‘background summary’ opens up new perspectives on how we understand recent and current events in the GL.
Work on political authority is particularly susceptible to a ‘presentist’ lens. While it is commonly acknowledged that political authority in the region has a history, this history is more often cited as a constant than interrogated as a question. Stereotypes of hierarchical domination and obedience are used to understand the ‘roots’ of today's turbulence. But history and authority – and, indeed, histories of authority – are not inexorably bound by the present. Authority, including the authoritarian forms it has taken in the region, takes root in history, and in so doing is never simple (Bayart Reference Bayart2014: 172).
Bridging the disciplines of history and political science, we propose alternative historical lenses to this presentism and its focus on crises and state power that still dominate the study of political authority in the GL. Focusing on Rwanda and Burundi, we argue that political authority has perennially been subject to uncertainty, opportunism and temporary arrangements of power relations that do not fit with the narrative of absolute and hierarchical control that is common with regard to the GL. Political authority carries multiple temporal and ‘spatial/scalar’ dimensions in its forms and practices (Planel Reference Planel2015). To see this, we must widen our historical sensibilities. First, we must go beyond the narrowly defined moments of crisis ubiquitous in conventional renderings of the region's past. Second, we must look beyond the familiar confines of state structures and presidential power circles, recognizing that political authority expresses itself as much in personal and local connections as in national arrangements.
If these are common complaints of historians, in order to bring them into dialogue with contemporary political science we propose two heuristic tools to analyse less presentist, and, indeed, less crisis- and state-centric histories of political authority: trajectories, a concept aimed at capturing flux in the practice of affirming authority; and transactions, meant to help us reaffirm the other ‘scales’ of authority beyond the state. In our opening discussion below, we explore the roots and problems of presentism in the region, introduce the alternative and productive work that expands on these limitations, and suggest how concepts such as trajectories and transactions can help bring historical complexity to contemporary understandings. We then propose two vignettes to illustrate our suggested concepts. We examine ‘trajectories’ evident in the misunderstood murk of the 1973 Rwandan coup d’état, and ‘transactions’ underpinning local languages of praise and petition in mid-century rural Burundi. These vignettes, and the heuristic tools they illustrate, not only help us see these historical moments in a richer manner, but also point us towards the deeper dynamics surrounding political authority in the GL that are overlooked by presentist accounts.
Presentism and authority
In an important critique of historical work in Africa, Richard Reid laments how a rising ‘presentism’ in the discipline has led to a predilection for ‘work on the modern period’ and ‘a clear tendency toward historical foreshortening’ (Reference Reid2011: 135). He is right to object to the neglect of the ‘precolonial’ in most contemporary historical conversations, but the problem of presentism in GL scholarship goes further. ‘Presentism’, especially in its manifestations outside self-consciously historical conversations, may be defined as undue attention to the present and recent past, to the neglect of the larger historical contexts that give rise to them. In some of its most extreme forms, presentism discourages a critical and curious eye regarding the past and how it informs our understanding of the present. It presumes that the past is known and that we can get on with studying the urgent present. Or it leads us to read the past through the lenses of the present, and to assume that the issues that grip the region today are the same as those that preoccupied the region in the past.
Such presentism has coalesced around two key biases in the region: an obsession with crisis, including of an ethnic nature, and an overemphasis on the state (Newbury Reference Newbury2012). This is partially the result of a new wave of scholarly interest in the region owing to the violence it experienced in the 1990s (war and genocide in Rwanda between 1990 and 1994; war in Burundi beginning in 1993; two wars in Zaire/Congo, one in 1996–97 and another beginning in 1998). For those focused on the conflicts of the 1990s, ethnicity is at the core of the violence and states are understood as the engines of bloodshed. As a result, work on the region has become fixated with these themes (Newbury and Newbury Reference Newbury and Newbury2000: 832). Rwanda is the most extreme example of this trend: with genocide defined as the deliberate targeting of a group based on its identity, and so many of our models of genocide built around the state as the driver of the violence, ethnocentric violence and the state are now rarely questioned starting points to understand the country, including its past.
Of course, this focus on the 1990s, on ethnicity and on state institutions is understandable. Genocide in Rwanda and interconnected regional wars were a watershed of monumental significance. Yet, as critical as these events were, they have eclipsed a more open kind of engagement with the region's history, becoming the primary reading keys to the GL's past. This is the paradox of presentism: the problem is not a complete absence of history but a reduced and static view of it that begins only from the most urgent contemporary questions. In seeking to understand the 1990s, present-minded scholars paid close attention to historical arguments relating to the colonial ‘imaginations’ of ethnicity and the heritage of royal authority in the region (Ranger Reference Ranger, Ranger and Vaughan1993; Chrétien and Prunier Reference Chrétien and Prunier1989; Lemarchand Reference Lemarchand2009), but this productive engagement then stalled. The tendency to ‘write history backwards’ from present concerns limits our view of what history actually matters (Cooper Reference Cooper2000: 307). We look only for the ‘lessons’ we think we need to know, and end up with stunted answers. Despite ongoing historically minded work on diverse themes across the region (e.g. Burihabwa and Curtis Reference Burihabwa and Curtis2019; Piton Reference Piton2016; Mathys Reference Mathys2017; de Haas Reference de Haas2019; Castryck Reference Castryck2020), history and contemporary analyses have seen only limited exchange wherever presentism stands in the way.
We see this especially in how political authority is studied. In conceptual terms, political authority is a complex and contested notion. Political scientists would define it as the power or right of the state, its representatives and, informally, of other actors to make decisions. But political authority requires its practice, as well as its recognition by those it applies to in order to exist (Cassinelli Reference Cassinelli1961; Arendt Reference Arendt1961: 91–141; Huemer Reference Huemer2013: 5). Turning to ‘authority’ as it may be defined in the GL, we might think first of the notion of ubukuru, the status and nature of primacy, superiority, greatness and of being older, first among others. This suggests an innate quality tied to one's position in a family, political order or social system. Alexis Kagame wrote of how itegeko, an ‘order’ from such a superior (umukuru), imposes on another ‘the obligation, not to choose, but to act in a determined manner’ (Reference Kagame1966: 231). But the superiority that creates such an obligation may be underpinned by the moral and personal qualities of the superior. Truthfulness, good judgement, equanimity and wisdom mark the ideal of ubushingantahe as an ‘organizing principle of social existence’ in Burundi, commanding respect through personal virtue (Ingelaere and Kohlhagen Reference Ingelaere and Kohlhagen2012: 49). Writing about hierarchy, Michel Kayoya (Reference Kayoya1968: 7) emphasized the values of iteka: it might take on the ‘technical sense of order, law, edict’, but it also suggests the qualities that deserve respect, found in the relational terms of the gift, blessing, divine favour, ‘the respect an inferior receives from his superior’. With iteka comes itekane (calmness, peace, security, a specifically good order), the two qualities together serving to legitimize the rightful ‘authority’ of the superior to issue orders that ought to be obeyed.Footnote 1 In return, that which one finds to be necessary (ngombwa) in response to power is derived from a passive form of ‘to want’, or kugomba, a ‘willed wish’ to follow the order (Kagame Reference Kagame1966: 232–3). Authority, in short, may be recognized by innate position, enforced by profound constraint, and wilfully or even enthusiastically endorsed, all the while being held subject to critical moral judgement and therefore to possible revocation.
This complexity, however, has tended to get lost in the ‘flattening out’ of the past that occurred after the events of the 1990s. In place of its ambiguity tied to practice and recognition, authority emerges in many accounts – including in the accounts of authorities in the region themselves – as a history of centralization, domination and obedience.Footnote 2 Authority in these terms is simply the force to coerce and compel. For example, the African Union's (AU's) report on the Rwandan genocide indicated that ‘it seems most likely that it was under Mwami (King) Rwabugiri, the Tutsi who ruled during the late 1800s, that the chief characteristics of modern Rwanda were fixed. From that point, a powerful head of a centralized state provided firm direction to a series of subordinate structures’ (AU 2000: 2.7).Footnote 3 Abbreviated to such terms of historical summary, authority becomes a static, top-down constant of domination, regardless of what histories of Rwabugiri's time might suggest (Newbury Reference Newbury2009; Vansina Reference Vansina2004), let alone the century and a half that has passed since. The AU's description is an extreme example, but stands as an illustration of how presentist uses of the past serve only to confirm the most common ‘canonical conventions’ (Newbury Reference Newbury2012) about the GL: this obsessive focus on crises, the state and elites at the top. The fixed understanding of authority as centralized state power is bound by these conventions.
The obsession with crises is reflected in the formal or informal embrace of ‘critical junctures’, popular with the conflict studies scholars and political scientists who flocked to the region following the violence in the 1990s.Footnote 4 Critical junctures are used to study complex social processes by organizing them around a clear causal narrative and moments believed to capture a specific trend; they cut through an abundance of facts and less ‘exciting’ periods (Capoccia Reference Capoccia, Mahoney and Thelen2015). This outlook on the region has not only overdetermined specific trends related to authority – and especially centralization and hierarchization as vectors of state-led violence – but has also led to an ‘arbitrary and biased periodization’ (Bayart Reference Bayart2014: 175), overdetermining which moments matter. It further confirms the common tendency in the GL to focus unduly on the national level. Presentism's statist bias locates authority rigidly, at the top and in the hands of elites, frequently the ones considered to matter in the making or breaking of authority.Footnote 5 A state-centric vision thus turns into an absolute ‘relationship of alterity’ (Bayart Reference Bayart1984: 145), one of very clear and distinct roles between ‘leaders’ and ‘followers’. In reality, these are, however, arbitrary roles.
This rigid approach to authority has been criticized by long-time observers of the region (de Lame Reference de Lame2005; Newbury and Newbury Reference Newbury and Newbury2000). It is also being increasingly challenged by work on the grass-roots, local, lived or subjective experience of power (Bayart Reference Bayart1984; Ingelaere Reference Ingelaere2014: 215; Planel Reference Planel2015: 14; Thomson Reference Thomson2013), and by perspectives on ‘hybrid governance’ and ‘negotiated statehood’ in the region and elsewhere (e.g. Raeymaekers et al. Reference Raeymaekers, Menkhaus and Vlassenroot2008; Hagmann and Péclard Reference Hagmann and Péclard2010).Footnote 6 At the same time, the historical dimension necessary to escape the presentism of this conversation is still largely lacking. Responsibility for this lies not only on the side of the social scientists. Despite their work exploring far more complex histories of precolonial and colonial experiences (especially in Burundi – e.g. Mworoha Reference Mworoha1977; Chrétien Reference Chrétien1993; Deslaurier Reference Deslaurier2002; Hatungimana Reference Hatungimana2005), historians may have inadvertently contributed to the constraining periodization of ‘critical junctures’ and ‘fixed’ understandings of authority by comparatively neglecting postcolonial decades in their work (Ellis Reference Ellis2002). Postcolonial, ‘non-critical’ history – that is, beyond critical junctures – is something that we have only been able to glimpse in rare work blending history and anthropology (e.g. de Lame Reference de Lame1996).
In short, we need richer understandings of ‘authority’, just as the Kinyarwanda and Kirundi vocabularies discussed above would suggest. Further, we need to find ways to bring subtler perspectives on the past into dialogue with the more critical contemporary approaches to the present. Seeking a vocabulary to facilitate this dialogue, we therefore propose two alternative lenses to understand histories of authority in all their fluidity and diversity: the notions of trajectories and transactions.
Seen as a trajectory, the past is never a neat succession of breaking points. Instead, it is made of non-deterministic, equivocal decisions, moments and trends. It shifts and reverses in unexpected ways, both over a long course and around some of the junctures on which we might be tempted to focus. A trajectory may arc over centuries (Mathys Reference Mathys2017; Reyntjens Reference Reyntjens2018) or decades (Burihabwa and Curtis Reference Burihabwa and Curtis2019), but it might equally be considered in shorter temporal scales, when one might perceive transformations in relationships of authority bound by days or months. As an analytical frame, ‘trajectories’ encourage us to reflect on ‘context effects’, a term borrowed from psychology to stress that actions and occurrences, even small ones, have a role to play in the emergence of patterns and trends. Pointing towards the everyday fabric of authority, its ‘equivocal and reversible’ nature (Bayart et al. Reference Bayart, Mbembe and Toulabor2008: 144), such trajectories both help us see the contradictions and complexity that matter beyond the moments of ‘crisis’ and provide the very real context of those ‘critical moments’ themselves.
As a complementary framework, the notion of transactions shifts the emphasis of our argument but not its substance. It looks to authority as an ongoing set of relationships seeking to claim, obtain or reaffirm control. Transactions are a space of mutual negotiation. Through dynamics of exchange, ‘bridging’ and settlement, political authority is practised and understood through myriad interpersonal relationships, acts and expressions, not only at the national level of ‘big man’ politics, but just as importantly at the local level. If ‘trajectories’ bring to light the multiple temporalities of authority, transactions emphasize the multiple social and spatial scales that relationships of authority cut across. The ‘trans’ prefix speaks to this manner of working ‘across’ circumstances and possibilities, ‘bridging’ one's local and quotidian circumstances with macro scales, actions and discourses. Transactions speak of this very basic dialectic that plays out in all manifestations of authority, both substantiating the reality of hierarchical domination and revealing the greater textures of authority that emerge from all aspects of life.
In order to illustrate how we see these concepts mediating past and present concerns, we apply them in two brief historical vignettes. First, we consider the rapid shifts in trajectories of authority to be found on the sidelines of a commonly cited, and rather deterministically interpreted, ‘critical juncture’: the 1973 Rwandan coup d’état. Then, we discuss the transactions of claims-making and critique within local politics during Burundi's first decade of independence, seeing how authority was shaped in the languages of praise and petition. Misunderstood ‘critical moments’ and passed-over spaces of quiet exchange, these different pasts matter, both in their own right and in the ways in which they can be used to help us rethink our narrow terms of authority in the present.
Trajectories and transactions of political authority
Rwanda: trajectory and interstice of the violence and 5 July 1973 coup
Rwanda is a paradox regarding authority. It is at once seen in scholarship as a regional paragon of authoritarian power and control (e.g. Reyntjens Reference Reyntjens2018; Twagiramungu Reference Twagiramungu2014), yet it is regularly shaken to the core, including by brutal ethnocentric violence. Indeed, for many analysts, the most extreme episode of violence the country experienced, the 1994 genocide, could occur only because of the Rwandan state's strong ability to command Rwandans to kill (e.g. Prunier Reference Prunier1995: 76; Uvin Reference Uvin1998: 22).Footnote 7 The genocide is depicted as the climax of decades of pogroms against the Tutsi (1959–61, 1963–64, 1973, 1990 and onwards), but also of parallel authoritarian consolidation, especially under the Second Republic (1973–94). But with Rwanda seen as such a successful and controlling authoritarian state, what explains the regular crises? How do we reconcile this focus on strength and on purportedly but little-questioned authority with a susceptibility to episodic and catastrophic unravelling? The coup that occurred the night of 4–5 July 1973 can help unpack this paradox. It suggests an ambiguous political trajectory, illustrative of the rich, ongoing competition for authority at the heart of Rwandan political life.
Today, one would be hard-pressed to find academic debates surrounding the 1973 coup.Footnote 8 According to the academic canon, the story of the coup is rather flat: republican authorities found themselves overwhelmed by popular anger expressed in a wave of violence against the Tutsi minority in February–March 1973, itself inspired by violence against Hutu in neighbouring Burundi the previous year. Or, looking to benefit from the violence in Burundi, Rwandan authorities chose to stoke ethnocentric violence against the Tutsi in Rwanda to prop up their popularity. Either way, within months, a largely northern-controlled army overthrew a government dominated by politicians from the centre and south, intending to ensure a return to stability it claimed was threatened by First Republican authorities and fostering the end of the ‘southern’ domination of national civilian institutions. Accounts therefore draw a clear and direct link between the violence that shook the country in March–February 1973 and the coup in July of the same year. The First Republic is seen as stable, until its last days when it teetered on the brink, only to be replaced by a new republic quickly reaffirming the Rwandan proclivity towards centralization and control from the top (e.g. AU 2000: 3.18; Mamdani Reference Mamdani2000: 137–8; Kimonyo Reference Kimonyo2008: 59–60). Revisiting the crucial months leading up to the coup, however, suggests a less obvious and linear series of events.
We can begin with the conventional starting point: the violence in Rwanda in the months prior to the coup, although some of its early manifestations had taken place in late 1972. This violence is largely depicted as ethnocentric in nature as it targeted the Tutsi population. But it also drew on deep frustrations among youth, the educated and rural Rwandans regarding the lack of change a decade after the First Republic's ‘social revolution’, which had promised deep social and political transformations. The ‘social revolution’ had begun as a movement to decry Tutsi dominance of political and socio-economic opportunities in the country, but it was not simply about ethnicity. It succeeded in reversing who controlled the country – the Hutu were now overrepresented in institutions – but stopped short of dismantling the system of discrimination it battled. Instead, the First Republic that followed the ‘social revolution’ reinforced divisions between the north and south/centre of the country, and increasingly between urban elites and peasants. The violence began as expulsions and violence against Tutsi in schools and at university, soon followed by additional purges in the private and public sectors. The movement eventually spread into rural areas, where new groups (privileged Hutu and wealthy foreigners, for example) also became the targets.
For many, this violent episode also served to exacerbate frustrations within the regime, especially of a regional nature, pitting the north against the south/centre (Kimonyo Reference Kimonyo2008: 59–60; Straus Reference Straus2006: 190–1). As a Belgian official at the time indicated, ‘One could easily sense during these events a lack of unanimity on the part of Rwandan authorities, facilitated by the fact that the president had a strong tendency, since his health became deficient, to let things go [laisser faire].’Footnote 9 By the end of March 1973, the violence had receded.
While academics’ only major disagreement regarding 1973 pertains to the role played by national authorities – were they overwhelmed or culpable? – the court martial set up to judge First Republican authorities was unequivocal: it found that Grégoire Kayibanda, the deposed president, and his accomplices had demonstrated a clear intent to kill.Footnote 10 The official communiqué announcing the coup on Radio Rwanda on 5 July stated that ‘the country had been about to fall into the abyss’, while the new president, Juvénal Habyarimana, claimed in the following days that the country had risked a blood bath or even genocide on the part of First Republic authorities.Footnote 11 But the coup occurred months after the violence ended. Why did the judgement and proclamations following the coup seem to suggest future violence? Were they necessarily talking about the same violence as the one that had occurred earlier in 1973? In our eagerness to draw a causal link between the violence and the coup, the crucial months in between and the power plays they gave rise to have been neglected.
Few are aware of or acknowledge another coup attempt during these in-between months, in April 1973. This was more than a rumour: many in Kigali described strange behaviour on the part of ministers and certain officers on 4 April, in a context of palpable apprehension across the country and within the military.Footnote 12 The April plot seems to have unfolded around a long meeting involving half a dozen key ministers. Ultimately, Habyarimana and a faction loyal to Kayibanda chose not to take part in the coup attempt, hence foiling the plot. Rumours of an imminent coup resurfaced again in May, making the notion of a plot in the making a secret de Polichinelle or open secret in Rwanda, where political circles seemed keenly aware that something was brewing.Footnote 13 Kayibanda himself alluded to a coup on different occasions, including during his ‘message de pacification’ (message of peace) following the violence in March.Footnote 14 Similarly, days before the successful overthrow of the First Republican regime, on the eleventh anniversary of Rwanda's independence, the president went off script, targeting his ministers, reminding them that they owed him everything, ‘their big bellies and Mercedes’, and daring coup plotters to attempt one right there and then.Footnote 15 For the French ambassador, this fit of rage was ‘telling of the deterioration of the climate within circles of power’.Footnote 16
Turmoil among Rwandan elites derived not only from the violence in February–March 1973, but also from rarely discussed decisions on the part of First Republican authorities during the months in between the violence and coup. These include a decision to merge the police and the Garde nationale (the equivalent of the armed forces), which observers saw as a means to break up problematic factions within the security sector, and potentially ensure greater control over a fringe of malcontents within the police force.Footnote 17 In parallel, these months also gave rise to tensions and rumours within the army, including regarding the dismissal of Habyarimana, then Minister of Defence, and another key officer, Alexis Kanyarengwe, possibly in connection with the aborted coup in April. The rumours turned out to be false, although a number of officers were moved to civilian duties. Years after the coup, Habyarimana still referred to these transfers as an affront to armed personnel, showing the extent of the frustration within the security sector regarding this decision. Indeed, all these rumours and decisions on the part of the authorities reflected and also fed tensions within the security sector – and, just as importantly, between officers and civilian authorities.
There were also important changes in the economic sector. Munyarugerero speaks of the decision in June to create a National Trade Office (Office national du commerce or ONACO) to ensure national control over external trade (Reference Munyarugerero2003: 133, 148). This decision was also meant to afford greater control over segments of the economy with large foreign involvement, such as businesspeople of Asian origin. But it stoked significant opposition in Rwanda. The Belgian ambassador even cited the creation of ONACO as a key factor behind the July coup.Footnote 18 Illustrative of how important the ONACO decision was to Rwandan political life, the first act Second Republican authorities undertook once in power was to suspend the new institution.Footnote 19 Habyarimana, as the new president, also insisted that he was ‘favourable to a large involvement of foreign capital in the development of Rwanda and … resolute in taking measures needed to calm the worries of investors’.Footnote 20
Finally, though just as importantly, these also proved to be key months for Rwandan political institutions. A successful modification of the Rwandan constitution was ushered through in early May to allow Kayibanda to run for president by abolishing previous term limits. According to the modifications, the president could remain in post regardless of his or her age or the number of terms served. While the modifications were not a surprise, since Kayibanda had been re-elected as head of his party on 28 April, they consecrated increased attempts to control institutions by an ever-shrinking clique around him. As a result, they fed the frustrations of the politically aspirational (youth, northerners) regarding the democratization promised by the Rwandan revolution, some of the very aspirations and frustrations expressed, through violence, during the February–March incidents.
These decisions can be interpreted as a means to control possible nodes of dissidence on the part of a regime whose legitimacy had already been put to the test during the violence earlier that year. But they were rooted in a much larger political context where challengers nipped at the regime's hold on power. They were also in and of themselves important irritants within this context. Each set of decisions targeted groups with a strong potential for disruption, including the army, parts of the economic core of the country, and the politically aspirational class. It is therefore not surprising to see some of these groups, such as the military or foreign businesspeople, associated with coup rumours circulating at the time.Footnote 21 Furthermore, all of these decisions were taken ahead of presidential and legislative elections planned for September 1973, which may have contributed to the scramble to tighten control or challenge it. Although cancelled following the coup, the elections played another key role: in early July 1973, most of the Rwandan political class were away from Kigali in their constituencies for voter registration, making it an opportune moment for a political takeover.Footnote 22
Yet in spite of a context rich in anti-government sentiments, a previous coup attempt and general coup rumours, nothing predetermined the events of early July. Accounts of the night of 4–5 July even suggest a good amount of spontaneity.Footnote 23 According to two key embassies in Rwanda, the night began with the governor of the National Bank, Jean Birara (a northerner), being called to meet the president. When Birara arrived at the presidency, Kayibanda was accompanied by approximately ten political figures from the centre and south. Birara was shown a letter addressed to Habyarimana in which northern officers complained of being harassed by southerners. Birara was then insulted, and fled. Next, Habyarimana was called to meet the president.Footnote 24 During the meeting, tensions flared and Habyarimana also fled. He alerted the Garde nationale and proceeded to arrest those who remained at the presidency, as well as other political figures present at the meeting. In some versions, including Habyarimana's, a list of 500 opponents to the regime meant to be eliminated was also found during the arrests, suggesting a larger plan on the part of First Republican authorities to clean house and eliminate dissidents. Within hours, the Garde nationale took control of the different ministries and institutions, and a Comité pour la paix et l'unité nationale (Committee for Peace and National Unity or CPUN) composed of ten military officers, as well as Habyarimana, was formed and announced the coup over the radio.
The timing of the events, with Kigali empty of its political class, made it an opportunity to precipitate a power play. But this could have been as much a strategy on the part of the incumbent president and ‘certain ultras’ among powerholdersFootnote 25 looking to neutralize or deter potential challengers as a strategy on the part of the coup instigators. After all, President Kayibanda and his ‘ultras’ had called the meeting with Habyarimana. To add to the sense of uncertainty, it is intriguing to note that on at least two occasions following the coup, the new authorities in Kigali seemed unsure of what to do with Kayibanda.Footnote 26 Habyarimana apparently even asked the Belgian ambassador if keeping Kayibanda as president would ensure the new government's international recognition. And contrary to common knowledge, many First Republican authorities, including ministers, remained free for weeks. According to the French chargé d'affaires, by 1 August, only ten of Kayibanda's closest allies had been arrested, while many others (ministers, MPs, party representatives) remained free, their only constraint being the confiscation of their official car.Footnote 27 Many more were eventually arrested and imprisoned under dire conditions, but in the early weeks their fates remained uncertain.
Revisiting the contingencies of these few months of history helps us make sense of the Rwandan paradox of control and crisis, where we see authority as constantly in the making. The simple sequencing of ethnocentric violence, a resultant crisis of authority, and the exertion of a new, more secure claim to power is a retrospective account that served the new regime. But it limits our understanding of how the coup came to be. Rwanda's history/histories of authority has and have always been a more complex trajectory – and indeed equivocal and reversible, in Bayart's words. But this is not only in the sense of dramatic moments of change, which can throw things ‘off course’. We need to understand the events of 1973 in the broader context of ongoing challenges to authority-making in Rwanda, some extending well beyond the violence earlier that year. In turn, the events of 1973, including at the neglected interstice of the violence and the coup, also show the ongoing potentialities of authority-making. Actors from across the political field vied for authority, yet never achieved it absolutely.
None of what unfolded during these months was sui generis: that is, an exceptional set of dynamics completely breaking with what took place during periods of ‘normalcy’. Rwandan political life always saw competition, including through the mobilization of affinities beyond ethnicity and the state, such as ties of clientship, professional solidarity, and so on (Des Forges Reference Des Forges2011; Vansina Reference Vansina2004).Footnote 28 As a French observer at the time put it: ‘Rwandan political life is a dense, incessant confrontation.’Footnote 29 Once we expand our rigid crisis focus, where we too commonly expect authority to be born and challenged, we find beyond it and at its interstice the denser stuff of authority-making and unmaking. It sometimes coalesces into the deeper moments of transformations we take as critical junctures, but it is by nature ongoing and indeterminate. It is to this stuff, to authority as a constant ambiguous trajectory, that we need to turn to understand the paradox: Rwanda teetered throughout post-independence and not just at the critical juncture constituted by 1973, with coercion (the violence, the coup) only the more extreme manifestation of Rwanda's ‘incessant confrontation’ for authority.
Burundi: transactions of praise and parenthood
From these trajectories of uncertainty at the summit of a state, we move to the hills of northern Burundi to consider alternative histories of authority in the linguistic and social transactions that make up the vast substrate of our dominant narratives. This shift in perspective, attention and tone is, after all, precisely what we need to disrupt the ‘flattening out’ of histories in the region. The transactions discussed here drive us to see the state – where presentism tends to confine political authority – in its larger social context of local authority, community and imagination (cf., for example, Leonardi Reference Leonardi2013). Claims and appeals to authority within and without the state, in sharing a language, imagery, memory and even emotion, weave a powerful web of constraint and critique that exposes the complexity of the idea of authority itself.
During the early to mid-twentieth century, as part of a long-evolving art of praise poetry in Burundi, songs of joy and celebration typified the language of address to a patron. ‘Let us extol our parent!’ announced one such song, transcribed anonymously from Maramvya (Rodegem Reference Rodegem1973: 200); ‘Long live the prince! … Turn your noble eyes towards us! Beloved parent, behold your children. … Come, let us dance in joy. Reign!’ These may appear to reinforce regional stereotypes regarding subjection to authoritarian power. But the adulation of authority brought with it the opportunity for claim-making. ‘I love you, you who have given unto me!’ exclaimed another song, unapologetically declaring the self-interest that drove this performance of love (Rodegem Reference Rodegem1973: 204). Authority was constructed in the gift transaction. In return for the gifts of aid and protection, alongside material gifts of tribute and taxation, authority was returned to a patron by a respectful, even grateful, dependant.
The transactional dimension of this language goes beyond such strategic exchange, however. Linguistic skill demonstrated the ability to speak across different forms and scales of authority, bridging the affinities and possibilities in social and political life. This is most notable in the language of parenthood,Footnote 30 a theme shared across the decades of political talk and far from limited to Burundi (Schatzberg Reference Schatzberg2001). ‘All benefactors are called umuvyeyi, parent,’ noted Albert (Reference Albert and Paulme1963: 187) at the time of independence. Twenty years later, under military rule, Ndimurukundo (Reference Ndimurukundo1981: 228) described the ‘striking correspondence’ of Kirundi terminologies and techniques that naturalized an alignment between ‘the law of the father, the fear of the chief, the notion of incontestable authority’. Most recently, following former president Nkurunziza's adulation as Imboneza yamaho (translated in international media somewhat liberally as the ‘eternal supreme guide’), his party attempted to deflect comparisons with Kim Il Sung and Jean-Bédel Bokassa by explaining: ‘He is our elder, our father. No one can compare to him within the CNDD-FDD’ (Sikuyavuga Reference Sikuyavuga2018).
As Burundi suddenly faced the prospect of independence in 1960–61, the profound uncertainty of the moment exposed the possibilities of this language of parenthood. On the one hand, rival politicians of all stripes competed to hail the king, Mwambutsa, as Sebarundi, the Father of the Barundi, to appeal to an uncertain yet solidly monarchist electorate (Deslaurier Reference Deslaurier2002). But when turning to appeal to this electorate directly, the new politicians claimed parenthood for themselves. ‘We are your parents, who wish you the peace of God,’ declared a political tract distributed during the electoral contest in 1960.Footnote 31 It blessed its readership, wishing them ‘cattle and little children [ibibondo]’ just as ‘our ancestors so often blessed their children’. Yet blessings came with a warning, ‘impanuro ya kivyeyi’, a piece of parental admonishing advice: ‘Do not elect Protestants, do not elect Swahili[-speakers, i.e. Muslims].’ Alongside claim-making from below (you are my parent, and you love me as one), democratic politics now marshalled the same terms to claim authority from above (I am your parent, and I love you as one).
It is, however, in the daily transactions of small-scale power that we grasp the full ambiguity of such authority. For the months and years that followed independence in 1962, conventional historical ‘background summaries’ skip straight to a failed coup and massacre in 1965, the country's ‘first ethnic crisis’ (Mariro Reference Mariro2005). But if we look away from this crisis, we can see how popular petitions in the rural margins displayed a mastery of the language of parenthood on a far finer grain than decolonization politicians had ever attempted. By the late colonial period, people in Burundi had begun using the written petition as a means of soliciting the power of the state, or of appealing to the global oversight of the United Nations. But it was with independence that this technology of appeal came into its own, as marginal rural actors, graduating from a brief and intense anti-colonial struggle, looked critically upon the authorities in charge of the new state.
The years following independence were, if anything, more uncertain than those that preceded it. The relationship between the newly independent state and its people was fraught with nervous suspicion, compounded by fear of the ‘Hutu Republic’ in Rwanda. While broadly sharing its neighbour's ethnic categories, Burundi's political consensus at independence had largely insisted on monarchist solidarity, and so both Tutsi aristocrats and Hutu royalists had something to fear in the revolutionary developments to the north. Nowhere was this suspicion more critical than on the border between the two states. In this tense atmosphere, in 1964, a group of local political actors in the commune of Mparamirundi wrote a letter to the governor of their province entitled ‘ukukenguruka’, an expression of gratitude towards a superior.Footnote 32 Addressing ‘Muvyeyi gouverneur’, their ‘parent governor’, the petitioners identified themselves as ‘your children’, followed by their names and signatures. Infantilization was not patronizing, but a tool of self-writing in addressing a patron. Celebrating the intimate bond of children and parents, the letter overflowed with praise. The writers wished to ‘converse with’, ‘thank’, ‘render homage’, to the governor, to ‘sing [his] praises’. Suddenly, a break: ‘O …,’ they interjected, seemingly evoking a stuttering emotional intensity that was stereotypically (or strategically) to be expected of a subordinate before a superior (Albert Reference Albert1964: 42). They then embarked on a string of wishes for the governor's own peace and prosperity.
Praise of authority was, however, scarcely disinterested. Despite the expressed ‘gratitude’, the men who wished their ‘parent governor’ such happiness were writing to elicit a gift. They were at odds with their local administrator (bourgmestre), accusing him of cowardice and incompetence, of betraying the country to Rwanda – an often fatal accusation against a border official. The bourgmestre had backed the wrong party under the colonial administration, attempting to launder his reputation by defecting to the victorious ruling party at independence, yet now found himself deeply resented by the long-term party loyalists he was supposed to represent (Russell Reference Russell2019: 174–82). The loving children spoke viciously of their rejected local authority, who had seemingly failed in his duty to provide itekane – good order and security – along the border. They demanded that the governor send them a ‘guide’ worthy of leading them, not one who would fall back ‘like a man meeting a lion in the wild’.
Petitions recognize in their form the authority that they address. Their common style is of deference, expressing a complaint while assuring the established authority that authority itself is not challenged. But they also reflect a belief among the petitioners that the authority to which they appeal can indeed be moved in their favour, however uncertain this outcome may be (van Voss Reference van Voss2001). They make authority stable, but not implacable. Petitioning thus played on hierarchy, rather than simply reacting to it. The emotional appeal to support from above, confirming the recipient's authority, was turned against the (rejected) authority of other, lesser superiors. Yet the play went further, curiously falling short of the absolute sovereignty such parental language otherwise implied. ‘We are honoured to come to express our thanks before your noble eyes and before the power that you have received from His Majesty, Mwami [King] Mwambutsa IV,’ some of the same petitioners wrote to the governor another day, still frustrated by the persistence of their resented bourgmestre.Footnote 33 The petition made full use of the rich Kirundi registers of authority, using words reserved for contexts of nobility and power. Even the eyes of the governor were not conventional ‘amaso’ but ‘inyonga’, a special word reserved for a princely gaze. But the passing reference to the king put the authority of their flattered governor in context: he, too, was subject to the sovereign, from whom he received his power. The skilful petitioners ended with a call to obligation: ‘We ask you as children, may you also give us aid,’ they declared. ‘May you receive us like a parent, since indeed that is what you are.’Footnote 34
Both declarations of power and petitions of supplicants in these uncertain days thus dwelt on the intimacy of family. But domesticity means far more than comfort. Parental obligations expose the parent as much to rebuke as to praise. As Schatzberg notes, when a presidential ‘Father of the Nation’ fails to live up to the paternal metaphor, opposition to him becomes ‘thinkable’, even necessary (Reference Schatzberg2001: 34). Ultimately, the language of parenthood might even carry the most profound indictment of failure in authority's rightful obligations. ‘When a child cries, give him what he wants so that he will be quiet,’ another group of petitioners wrote to government ministers in striking anger against the same detested bourgmestre.Footnote 35 They saved their most vicious sting to the last moment: in appeal to ‘bavyeyi wacu’, ‘our parents’, they signed themselves as ‘imfuvyi’, the ‘orphans’. The orphans sought not only to elicit but to demand pity, without shame, and condemned the failure of their ‘parents’ as neglect. Authority was endorsed and indicted at the same time.
The petitioners knew how to use the language of paternalist authority just as well as their superiors. Cutting across scales of local and national authority, rural petitioners rejected one while challenging the other. Bridging parental and political authority, both the petitioners and their political superiors entwined love and obligation with claim and evaluation; they kept the political relationship in a constant, fluctuating state, bound by reassuring yet incomplete terms of affection. But the local saga of the border bourgmestre was not some footnote to the dominant narrative of the early 1960s in Burundi, underscoring the consolidation of Tutsi rule, or a precursor to the ‘first ethnic crisis’ of 1965. The state was in denial and wanted to hear nothing of ethnic politics, even while its ministers pursued such violent and divisive ends. The language of family explicitly precluded the expression of ethnic difference within that family, and in the petitions cited here it was deployed by Hutu activists to attack a Hutu bourgmestre, accusing him of sympathy with republican Rwanda. In such transactions the details matter. But the form and ambiguity of such languages of authority also speak of far greater possibilities that continue to exist today. Looking to such histories, we can begin to see the scope of authority transactions that exist within and beyond the critical ideologies of ethnicity or absolute hierarchy, and indeed beyond the focus on the state common to discussions of authority in the GL.
Political authority, past and present
The vignettes presented here start a conversation on what is to be gained from unsettling common presentist lenses with regard to political authority in the GL. Our exploration of Rwanda served to illustrate our concept of trajectory, and how authority-making was subject to uncertainty and ‘context effects’: that is, how small and perhaps neglected interactions also matter to trends and events. Our Burundi vignette turned our attention to the transactional nature of local forms of praise and petitioning, beyond the apparent reproduction of a seemingly unquestioned authority. To better illustrate the heuristic value of our concepts, we ascribed a single concept to each case. But trajectories and transactions of authority do not exist in isolation. Rwanda's trajectory of authority may also show how different actors pursued variegated transactions over their position within the political system, choosing to enact or challenge First Republican authority in deeds or in words. Similarly, looking at the level of local transactional language in Burundi helps us understand how authority is equally the product of multiple trajectories: petitioners consolidated authority locally through ambivalent loyalty to the state, while the authority of the state on a national scale seemed to be disintegrating. Trajectories and transactions are the start of a question about the different ways in which we might see the meanings of authority through history, not set categories or rigid concepts with which to label a new paradigm.
Instead of the singular, achieved character of authoritarianism with which we are often presented, especially with regard to Africa, we choose to understand authority as perceived and pursued in a Sisyphean manner. Authority is a project rather than an acquis, because it is ambiguous and relational. It is, therefore, essential to view authority through and with history, albeit for many different reasons. Perhaps, as in the internal machinations of a secretive elite or the semi-privacy of written petitions, the past simply offers a more accessible means of reading an elusive face of authority that today may be hidden behind political sensitivities and evidential constraints. But, in other respects, the threads of these earlier political actions, and the engagements and aspirations to which they gave rise, are performed as part of the making and unmaking of authority by contemporary actors. President Nkurunziza's ‘fatherhood’ to his party, or the contingencies of personal ambition and opportunism around state power throughout the region, may draw directly on knowledge and inspiration from the past, or our understanding of them may simply be informed by this richer scope of past experience. This is how the past comes to form the composite that is the present: not only through cause and effect, or event and repetition, but through memory, interpretation, possible precedents and abandoned alternatives.
Challenging presentism is not only directed at GL scholars. It matters beyond the GL to recognize that the past is not only a prologue but is a very lively part of the present. Turning to history to understand complex issues ought not to be just about excavating roots or establishing foundations. ‘History does not offer lessons,’ notes Cooper (Reference Cooper2000: 312). ‘But it does suggest possibilities.’ It should alert us as much to precedent as to difference, change, and paths not taken. The point is not just to rewrite the ‘historical background summary’ but to find a place to talk about history in the meat of contemporary debate. Nor are we advocating a form of disciplinary imperialism, where guild historians dominate the creation and mobilization of historical knowledge. An engagement with questions such as political authority from a historical perspective ought to be embraced from many alternative perspectives, and must resist collapsing the answer back to simple terms. Our concepts of trajectories and transactions are meant as tools to unsettle rigidities and to question rather than establish a new orthodoxy of the fluid or the local. Scholars of African history have always strived to show the complexity of some of the continent's most important moments. We should embrace this richness to better reflect on the significant contingency and granular exchanges behind the rigid arcs of recurrent violence and dictatorial enshrinement we too often draw.
Of course, in pursuing this position we need not jettison the premises of domination and hierarchy, the importance of crises, or the profound role of the state in its various forms and functionalities. These issues are central reading keys in our region of interest, as well as in many other parts of Africa, for valid reasons. But we need to challenge our assumptions regarding these canonical themes and how they are understood to bridge the past and present. We need to think about how flux in the form of trajectories and agency as captured by transactions around authority interact with and change our understanding of perennial factors of political life. Neither of our lenses on authority exhausts or dictates the shape of the debate, just as neither is limited to the times or places we have chosen to explore here. But, together, they point to the renewed substantive and subtle possibilities for debate we are pursuing, where the past is liberated from a presentist cage and enriches our understanding of the actual present.
Acknowledgements
We wish to thank David Newbury, Philippe Lagassé, René Lemarchand, Emmanuel Ndikumana, Filip Reyntjens and Susan Thomson for their insights and helpful comments on earlier drafts. Additional thanks go to the Social Sciences and Humanities Research Council of Canada (SSHRC) and the University of Ottawa for financial support.
