With the institution of the Conseil national de développement (CND) in the early 1980s, the then president of Rwanda, Habyarimana, declared that the country had successfully democratized. Rwanda had achieved its ultimate democratic stride and become a full-fledged democracy. The CND was meant to serve as a legislative body under the Second Republic and to fill the gap left by the suspension of its predecessor in the wake of a coup in 1973. Steps and changes undertaken by Second Republic had constituted, as he declared, “so many essential milestones, on the way to a complete and true democracy,” with the reinstitution of the legislative assembly completing the process.Footnote ¹ According to presidential rhetoric, with the CND now in place, Rwanda also fully epitomized liberal democratic principles, with Rwandans free to revel in their rights and liberties. As Habyarimana further exclaimed, “this innovation will certainly have proved to you that the Rwandan people fully exercise their freedoms.”Footnote ²

This is despite some of the inherent contradictions regarding how the CND was meant to serve its legislative function, as well as with regard to Rwandan democracy and rights. Far from independently, the Conseil was to exercise its function with the president, and the democracy it represented was to be “responsible.” Citizens’ rights and freedoms, which the CND was meant to embody, were to be enjoyed with “discipline.” A single party, the Mouvement révolutionnaire national pour le développement (MRND), was represented within the halls of the CND, mooting any hopes for political alternance. In a similar fashion, under the cover of claims of universal suffrage, presidential elections yielded far from competitive results. In all three Second Republican presidential elections prior to 1990, citizens returned President Habyarimana to power with an overwhelming 99 percent share of votes, casting doubts on just how “popular” – that is, how truly reflexive of citizens’ aspirations – these results were. Habyarimana’s democracy grew in undemocratic soil.

Yet, that Rwanda was a democracy seemed evident to President Habyarimana, at least in rhetoric, a claim to a democratic ideal not unfamiliar to scholars of authoritarian regimes at the time and today. Obvious democracy of the kind Habyarimana spoke of belonged in Rwanda to the realm of political imaginaries, ideology, and strategy. In practice, seemingly democratic strides maintained a complex relationship with authoritarianism.

The ideals of democracy – claims to representativity and freedoms – including forms of democratic performances and institutions have always mattered, even in a noncompetitive political environment. Beyond the rhetoric of presidents, our understanding of the relationship between democracy and authoritarianism has evolved over time in scholarship: from obvious and opposed realities, at least conceptually, to a more complex understanding of the rhetorical deployment, borrowing, masquerading, and hybridity and mixed practices that exist, as some of the more recent debates in the study of authoritarianism have suggested.

Much like this episode around the creation of the CND suggests, this chapter starts on the twinned concepts of democracy and authoritarianism, and how we should understand their relationship, to help set the stage for some of the debates that have animated the comparative study of authoritarian regimes in recent decades. As this chapter shows, authoritarianism has, by convention, been defined in relation to democracy. But starker, and more clear-cut, understandings of the two have more recently made way for a finer look at the practices that lie in the middle of the two ideal-types. Institutions we have commonly come to associate with pluralism, such as parties and forms of electoral competition, have grown in autocratic contexts and even helped bolster authoritarianism, as the more recent debate on authoritarian resilience has explored. This institutional turn can only take us so far in understanding some of the deeper and more granular realities of authoritarianism, however. I further discuss, based on work of a more local or sociological nature, what understanding authoritarianism as a “trajectory” looks like. The chapter ends on a discussion of the challenges and limitations inherent to the study of authoritarianism in Rwanda and beyond.

Conventions Regarding the Study of Authoritarianism

There is indeed a long tradition of scholarship on authoritarianism – even longer than a focus on the most recent debates would suggest – that can help situate the notion of “authoritarian trajectory.” Preeminent among these have been debates around definitions, which have come to the chagrin of many who believe too much time has been spent naming, defining, and “topologizing” notions surrounding the concept of authoritarianism, to the detriment of studying authoritarian realities in and of themselves.

Despite the debates, however, most scholars start from the same point when defining authoritarianism. They have a relational understanding of autocracy, where it is necessarily defined in comparison to democracy, which also explains why authoritarianism is commonly conceptualized as “nondemocracy.”Footnote ³ Indeed, authoritarianism is often treated as a “residual” category encompassing all regimes that do not qualify as democratic.Footnote ⁴ As a result, it is not uncommon to see authors pour more efforts into defining the concept of democracy than into defining authoritarianism itself, though offering a definition for such a history-laden concept as democracy is no simple task.

There has been a tendency for some to adopt a minimal understanding of democracy centered on procedural elements: competitive, free, and fair elections held at regular intervals with a real chance for the alternation of power in the legislative (elected directly) and executive (elected directly or indirectly) branches, as well as meaningful options at the ballot box in terms of multiple candidates and parties with a credible chance of winning.Footnote ⁵ Some scholars insist on a demonstrated capacity for alternation, where parties actually lose elections and step down in accordance with the same rules that had brought them to power,Footnote ⁶ especially in the face of cases in which all the procedural boxes appear to be ticked, but a single party holds on to power for decades. A key example of the latter is the Partido Revolucionario Institucional (PRI) in Mexico prior to the 2000s. Other scholars have opted for more expansive definitions of democracy that are focused on substantive issues such as the absence of violence, accountability, the respect of rights, including key political rights like association and expression, the vibrancy of political mobilization or civil society, or the provision of public goods. These added criteria, however, raise issues regarding normativity, measurability, comparability, and the ability to examine relations across these elements.Footnote ⁷ These are also the “liberal,” more than the “democratic,” side of the historical type that is liberal democracy.

As a residual category, authoritarianism is understood, as Linz seminally put it, as “nondemocratic systems” – those “that share at least one characteristic, that of not being like those we shall describe with our definition of democracy.”Footnote ⁸ That is, authoritarianism is seen as a regime falling short of the democratic ideal, whether minimally – in terms of competition, with no real independent and unco-opted alternate candidates and parties, as well as no alternation of in terms of who is in power – or, more broadly, in terms of the recourse to violence to manage challenges, or of violations of basic individual and associative rights. For example, Gandhi’s minimal definition of dictatorship, which centers on “rulers acquir[ing] power by means other than competitive elections,”Footnote ⁹ resembles Svolik’s slightly more precise definition: a regime that “fails to elect its legislature and executive in free and competitive elections.”Footnote ¹⁰ Levitsky and Way’s conceptualization of authoritarianism goes further than competition; they also identify an unlevel playing field as well as challenges to civil liberties as features of authoritarian regimes.Footnote ¹¹ Few have ventured beyond the democracy–autocracy dichotomy in their conceptual efforts, however, in large part because, as a residual category, authoritarianism brings together an extremely varied and evolving lot of political experiences,Footnote ¹² which cannot be captured by too strict or too refined a definition.

This challenge is all too apparent in scholarly efforts to organize different authoritarian experiences according to typologies, a long “taxonomic” tradition of comparative authoritarian work that continues to be a core component of the work today.Footnote ¹³ Researchers propose categorizations of different natures. For instance, one strain of scholarship focuses on who rules or who composes the “inner sanctum”Footnote ¹⁴ of power. This includes Geddes’s focus on personalist, military, and single-party regimes,Footnote ¹⁵ which is not unlike Cheibub et al.’s distinction between monarchic, military, and civilian dictatorships.Footnote ¹⁶ Other forms of categorization focus on degrees of coercion or the types of means deployed, such as Schatz’s and others’ “soft” and “hard” distinction further discussed below and explored in relation to Rwanda in Chapter 4.Footnote ¹⁷ Similarly, Alvarez et al. propose a distinction between “exclusionary” and “mobilizing” dictatorships,Footnote ¹⁸ and Levitsky and Way between “full,” in the form of “closed” or “hegemonic,” and “competitive” authoritarianism.Footnote ¹⁹ Any new label or typology is quick, however, to spark debates as to whether it reflects aptly and fully the many different authoritarian forms and captures the boundaries between them. These boundaries are, after all, set to remain “blurry and controversial,” as Diamond argues,Footnote ²⁰ considering the variety that exists in terms of authoritarianism.

Part of the definitional challenge also stems from the fact that authors do not agree on the nature of the democratic–authoritarian distinction. Is it a question of degrees? Do regimes slide toward more democratic or autocratic leanings on a scale of political openness or control? Or are democracy and authoritarianism two incommensurate realities, that is fundamentally completely different types, operating differently and deploying different means? The former position is closer to some of the thinking behind the literature on hybrid regimes or “in-between” types, while the latter is defended by authors such as Svolik, who argues that the notion of “in-between” is problematic: You either have democracy or you do not, since “the difference is decidedly one of kind before it is one of degrees.”Footnote ²¹ Both sides of the debate bring important insights to the table. We know there is variety and, just as importantly, fluidity across regime types, as the work on democratic backsliding has illustrated.Footnote ²² But we also know that, at some point, a threshold of “regime change” is crossed, though with much conceptual and measurement uncertainty around exactly when the line is crossed.Footnote ²³ Once this threshold is crossed, however, we are in clear authoritarian territory, as reflected by the means, practices, and strategies deployed.

Thus, in a move similar to Schatz’s, who chose to eschew semantic debates and to refer instead to these general conventions, this “conventional terrain” is taken to be sufficient to set the conceptual boundaries of this book.Footnote ²⁴ For all the debates about definitions – outside the more ambiguous cases generally found in a zone of grey somewhere between democracy and authoritarianism – most people tend to agree about who/what is authoritarian. Without dismissing the important work of fine-tuning concepts and typologies, we have enough of a conventional sense of basic authoritarian characteristics – such as limited access to power, a lack of competition in determining this access, and a recourse to strategies beyond “conventional” democratic norms and beyond liberal normsFootnote ²⁵ in terms of expected rights and liberties in the face of political power – to know an authoritarian regime when we see one.

Transitions, Resilience, and the Institutional Turn

Definitional debates are just one of the key divides in the literature on authoritarianism. There have also been important shifts in the puzzles addressed by comparative authoritarian scholars. One of these can be traced back to the 1990s/early 2000s. Scholarship in recent decades, and particularly since the end of the third wave of democratization, has predominantly focused on a single question: What accounts for regime transitions, or lack thereof, understood broadly as shifts across regime types in the form of breakdowns of authoritarianism or, more recently, in terms of the interest in “how democracies die”?Footnote ²⁶ Why do some regimes transition toward democracy, away from stable authoritarianism, or vice versa, while others do not?

As became apparent in the 2000s, transition is not necessarily a smooth process and sometimes stalls or reverses. Both the “hybrid” regime literature and the “authoritarian resilience” literature emerged as a response to the transition assumption implicit in third wave literature that studied the “triumphant” expansion of democracy and, in so doing, attributed a dominant role to key “democratizers,” such as the opposition and civil society.Footnote ²⁷ With the growing realization that many democratic transitions simply did not take place, the pendulum swung to a new, largely institutional research agenda focused on explaining this “strong undemocratic undertow,” seen in the mixed results of the third wave.Footnote ²⁸ Whether studying the factors behind endurance versus transition fully moves us beyond the previous “transition paradigm” is, however, another question.Footnote ²⁹

Work on hybrid regimes focuses for its part on “half-way” regimes. The notion of hybrid regimes refers to countries seen as neither democratic nor authoritarian, but melding the characteristics of both; these regimes are stuck in the limbo of transition, not having attained the democratic ideal implicit in the third wave. In other words, hybrid regimes are caught somewhere in the “messy middle,” as Schelder argued, but possibly durably so, as Gagné and Mahé remind us.Footnote ³⁰ Many have, however, found the notion of hybridity “messy,” conceptually and empirically, and rather vague on this state of “in-between.” It is hence dismissed today by many scholars, though not all.Footnote ³¹

Dropping presumptions about the specific direction in which regimes were or are headed, implicit in the labels associated with the hybrid regime scholarship, such as “illiberal” or “flawed democracies” or “partly free,”Footnote ³² a number of scholars have reasserted their interest in authoritarianism per se, highlighting its resilience as the puzzle to explain. Authoritarian resilience or endurance literature focuses on how regimes weathered the latest democratization wave with or without the fundamental change of nature entailed in the notion of hybridity, often by adopting in name or essence some democratic guises. Indeed, scholars of this most recent iteration of comparative authoritarian literature look to understand the meaning of these democratic traits amid clearly authoritarian settings. For some, this combination of traits amounts to a new sub-type of authoritarianism, one not unlike a hybrid regime because it allows meaningful, but biased competition, as captured by the notion of competitive authoritarianism.Footnote ³³ But for others, this competition amounts to a democratic “façade,”Footnote ³⁴ useful to manage a new world environment set against full blown or manifest authoritarianism,Footnote ³⁵ as well as to diffuse internal challenges, not unlike other strategic forms of democratic deployment or controlled liberalization seen in the 1960s and 1970s.

The authoritarian resilience literature, and more broadly what Pepinsky and others have recently coined the “institutional turn” in comparative authoritarianism scholarship,Footnote ³⁶ has made important contributions, including pushing a program of systematic comparative analysis (medium-N or multi-case qualitative analyses) to assess the explanatory power of different factors thought to contribute to authoritarian endurance. And, indeed, this work has yielded a rich take on what differentiates pathways of authoritarian resilience and decay, following the pioneering work of Geddes, who sought to systematize our understanding of democratization and transitions by looking at the characteristics of different types of authoritarian regimes. Setting the bar high in terms of analytical and methodological rigor, Geddes examined data from more than 160 regimes and surmised that what she terms personalist, military, and single-party regimes have different characteristics and dynamics, which lead them to “break down in systematically different ways.”Footnote ³⁷ This insight spurred the interest of a new generation of scholars looking to explain non-transition/transition outcomes rigorously. This interest has however led, as it did in Geddes’s work, to a focus on trajectories across types,Footnote ³⁸ creating a directional perspective on hybridity, backsliding, and the broad outcomes of the authoritarian experience, whether resilience or decay. They have focused less on the relations or trajectories within subtypes – personalist, military, single-party, or otherwise.

As indicated by the “institutional” label, in delving into these “regime trajectories” or “outcomes,” two notions too often blurred in the literature, this newer generation of scholars has rediscovered institutions. This rediscovery was warranted in response to what some saw as an overemphasis on actors, especially key actors of change, in the transition literature. The turn toward institutions was also meant to address the quick dismissal in previous work of “pseudo-democratic institutions” as artifices having no real function or purpose. On the latter in particular, work by Brownlee and by Gandhi has stressed how parties and legislative chambers can play a real role in managing elite competition in authoritarian settings, much like elections can, according to Gandhi and Lust-Okar.Footnote ³⁹ Gandhi, for example, has helped to broaden our understanding of co-optation among elites. She has shown the importance of other incentives, besides financial or power-driven incentives, such as policy concessions in this context. Like Svolik,Footnote ⁴⁰ she has also insisted on the role parties and institutions play in confirming a regime’s commitment to the power-sharing co-optation schemes it employs to manage elites. This institutional focus has brought a renewed sense of the very real game of push and pull that results from the effort to co-opt potential challengers within or at the door of “inner sanctums” and within the “selectorate,” those who can make or break a leader, and produce a “winning coalition.”Footnote ⁴¹

Versus a Deeper Authoritarian Politics

In addressing the perceived flaws in previous literature, the pendulum may have swung too far, according to a growing number of critical voices. Decried by Svolik as suffering from a tendency to study facets of authoritarianism “in isolation,”Footnote ⁴² recent literature has been described in more dramatic terms by Pepinsky, who believes that the institutional turn leads the study of authoritarianism down the road of “superficiality” and “narrowness.” For all its focus on institutions, Pepinsky argues, this scholarship is centered on systematically observable “surface politics,” to the great neglect of the “deep politics” of authoritarianism.Footnote ⁴³ In his words, “authoritarian regimes do many things besides grow/stagnate and survive and collapse,”Footnote ⁴⁴ but we mostly get only macro and institutional glimmers of these other things. And, while Pepinsky hints at what some of these deep politics may be, such as the recourse to local-level violence, identity-related dynamics, state–bureaucracy relations, and co-optation strategies directed specifically at citizens, and so on,Footnote ⁴⁵ we are left, after the recent decades of institutionalist literature, with only a vague sense of why these deep politics matter to our understanding of authoritarianism and with few insights into how to begin studying them. This lack is notable even though a number of Francophone authors, such as Bayart, Hibou, and Rowell, as well as others beyond the Francophone world, such as Fu, Shue, and Wedeen, have over the decades showed the value of this deep dive through their work.Footnote ⁴⁶ It is also worth pointing out that this literature, especially Francophone literature, has often drawn on African cases to develop its deeper look at authoritarianism.

Indeed, the current institutional agenda leads us to take a narrow look at who (actors) and what (tools or means) populates authoritarian politics. The recent literature has proved quite coherent in its focus on elites. The interest in national-level elites should come as no surprise, since elites, and in particular intra- and inter-elite competition (depending on the scholar), are generally perceived to be the main source of threat to political arrangements and hence of change or lack thereof in authoritarian regimes. When not effectively managed by institutions, elites are seen as a key element in instability or breakdown.Footnote ⁴⁷ Thus, breaking with the problematic image of the lone autocrat, this literature stresses the very lively and dynamic circle of power at the top of authoritarian regimes, within which elites vie for resources, influence, and power to secure cooperation and dissuade challengers.Footnote ⁴⁸

But our understanding of these elite negotiations remains rather limited. Take, for instance, Brownlee’s definition of “elite” as “national-level agenda setters, figures who wield regular and substantial influence over a country’s political system.”Footnote ⁴⁹ He marks a clear distinction between “elites,” the “ins” in the system, and “opposition figures,” the “outs.” This distinction obscures the fact that to rise as “opposition figures” actors often already need to wield some degree of influence – that is, they need to already be some kind of “elite.” Elite pacts and power-sharing instances are also commonly conceived as single-iteration games. Co-optation is treated as an open/shut process, in which the deal is done once the elites buy in and integrate the system, as long as resources keep flowing. This conceptualization keeps the focus on unsettled times of change or possible change. Outside these unsettled periods, agents are assumed “in” and the institutional bias prevailsFootnote ⁵⁰. This portrait is rather static, as opposed to a more nuanced understanding in which “in” and “out” are understood as subject to continual negotiation, with some elites choosing to navigate back and forth depending on the situation or the agenda. Brownlee acknowledges the implicit rigidity of his conception: the in/out line may be blurrier, and agents may cross it regularlyFootnote ⁵¹.

Our rather rigid take on elites may in part stem from the motives ascribed to them, and especially the implicit or explicit maximization assumptions that are at the heart of many analyses. The presumed driver of elite action tends to be opportunism – the desire to maintain or attain power, influence, or resources, for example.Footnote ⁵² The corollary is the common notion that, once “in,” agents do not work to undermine the system they benefit from, at least until a possibility of greater benefit appears or the system itself threatens benefits. In other words, the literature upholds the notion that “ins” do not bite the hand that feeds. The prospect of gains or threats to gains therefore becomes a mechanism to explain pathways of endurance or transition. These assumptions with regard to the nature of elite motives remain theoretical, however, and more importantly, they provide us with little ground to understand the mixed or ambiguous motives and more dynamic positioning of actors in reality. This is especially the case with regard to settled circumstances, where situations may prove more fluid than at times of crisis or at radical breaks that require clear in/out decisions from elites.

While the institutional turn leads to a focus on the national level and the regime’s elite supporters and challengers, comparative authoritarianism scholars have had much less to say about the rest of society in authoritarian settings. With few exceptions, recent scholarship says little about local-level elites and, in particular, about those tasked to be the system’s local face and its implementers.Footnote ⁵³ There is little sense of their role and whether they even matter in and for the authoritarian apparatus. Considering the lack of attention to these local elites, one could conclude that they do not matter much, at least for understanding pathways of resilience or transition.

Nonetheless, some scholars have acknowledged the importance of broader state–society relations. Authors such as Svolik and Gandhi include populations in the theoretical dilemmas or challenges that they argue authoritarian leaders face.Footnote ⁵⁴ Popular or mass-led overthrow, though much less frequent than elite challenges, remains a possibility in authoritarian settings. There is therefore a sense that the “control” of populations, and consequently their “compliance” or “acquiescence” is necessary.Footnote ⁵⁵ However, because a bottom-up challenge is considered a much less likely threat than one from elites, authors more commonly talk of the means deployed to bring populations into the fold, that is, they rarely question regimes’ ability to attain this control and compliance. The reigning impression is that coercion and co-optation in the right dose are generally effective on authoritarian publics, and scholars consequently neglect discussions of how these strategies operate and whether they succeed, especially in comparison to the attention given to the same mechanisms in relation to elites.Footnote ⁵⁶

The turn to institutions has brought some attention to the role institutions or organizations play with regard to publics. Svolik, for example, proposes an interesting take on single parties’ ability to recruit and co-opt key segments of the population, to proselytize, and to monitor behavior at the base.Footnote ⁵⁷ Several authors make similar arguments about parties’ ability to penetrate and hence control society. However, little attention is given to how the process unfolds and the degree to which this penetration is achieved. Levitsky and Way partly address this issue, proposing scope and cohesion as means to assess a state’s coercive capacity and a party’s strength. But the ability to measure these characteristics without an assessment from the base – that is, empirically gauging how a party’s efforts are received by the authoritarian public – is questionable.Footnote ⁵⁸ As a result, institutional work has a decidedly one-sided outlook on state–society relations: We hear from the regime’s side, but we get very little in the sense of how the population engages with the state. As Geddes commented as the institutional research agenda was beginning to make headway:

For the most part, recent scholarship has largely remained at the level of abstractions. Empirical investigation has proved rare in recent decades and tended to stay confined to historical or anthropological/area studies endeavors.Footnote ⁶⁰ Theoretical reflections on society’s relations with the state in authoritarian settings have proved just as rare, the emphasis largely remaining on those who have ways to remove leaders from office – that is elites. This is an unfortunate and easy dismissal of politics beyond the national level, especially in terms of getting at the grain of authoritarianism.

This narrow focus with regard to actors may be, in turn, affecting our understanding of authoritarian means or “tools,” to borrow Schatz’s tool kit imagery.Footnote ⁶¹ With attention focused on opportunistic actors at the top, discussion of authoritarian tools has centered overwhelmingly on co-optation as the means deployed to manage elites. Some, like Art, lament this focus,Footnote ⁶² arguing that we should get back to studying the coercive tools discovered through authors like Hannah Arendt: violence, repression, and intimidation. But a return to studies of coercion may be just the first step. The authoritarian state’s engagement with citizens is largely taken to rest on a limited tool kit centered on aggression, repression, or patronage. In other words, the universe of authoritarian practice largely ends at coercion and co-optation.

A number of authors who have worked to build a bottom-up understanding of authoritarianism, or have at least included an emphasis on the citizens’ side of authoritarian state–society relations, have proposed that we adopt a more encompassing view of authoritarian means. They have insisted on the role of normative, rhetorical, and symbolic strategies too often dismissed as slogans and political bluster. In the same vein, recent scholarship on new technologies is studying the power of the “digital word” or the power to control it, including in authoritarian settings.Footnote ⁶³

Norms, language, rituals, and symbolism all play a role in trying to achieve compliance, by enacting state power, expressing expectations, and striving to institute self-discipline and pre-emptionFootnote ⁶⁴ among publics, especially in states looking for alternatives to the “moral hazard of coercion”Footnote ⁶⁵ or the costs of co-optation. Co-optation at the heart of recent institutional work and coercion matter, and they matter a lot in terms of managing potentially disturbing relations. But once we broaden our sense of which relations matter in authoritarian settings, we also need to broaden our outlook on what constitutes authoritarian strategies. This broadening is what I propose, seeking to move beyond the predominant institutional focus to reinvent our interest in “authoritarian trajectories.”

Arguing in Favor of a Focus on Trajectories

Rwandans often enrich their conversations with references to the many proverbs that exist in their culture(s).Footnote ⁶⁶ One proverb that has fast been gaining in popularity teaches us that “the dancers change, but the dance remains the same” (Ababyinnyi barahindutse ariko imbyino iracyari ya yindi). It popped up in interviews for this book, but also finds its way into some – including my – writing on Rwanda.Footnote ⁶⁷ This proverb can be used, and is indeed used, to stress continuity. Despite events and circumstances, some deeper trends remain: There is a set and fixed choreography, which is reproduced whoever performs it. It should come as no surprise, then, that the proverb has been used as an allegory for the continued authoritarianism Rwandans have experienced since independence, and arguably even prior to independence; changes in the identity of the authorities, including dramatic events such as the 1959 “Social Revolution” and the 1994 genocide, have never been true breaks in the way the political game is played.

The proverb can be interpreted this way at one level, but at another level it speaks of dancing, of motion and interaction: A dance commonly exists as movement, and its choreography involves the dancers moving in relation with each other, their public, and their environment. Behind the continuities, the structures, the longue durée trends, the proverb thus also speaks of the movement that makes up what we recognize as the dance. This interpretation too can serve as an allegory for authoritarianism. At its most basic, before we seize the recognizable characteristics of the choreography that allow us to differentiate it apart from other forms of dance (this is ballet, this is tango, this is breakdance), characteristics that also allow us to tell it apart from other forms of authoritarianism (this is a military regime, this is a personalist regime, this is a single-party regime), authoritarianism is in essence movement. It is a lively, dynamic engagement and set of relations. Admittedly, in the world of dance, a dancer can sometimes stand on the stage alone. Dictators for their part, despite some of our imagery regarding authoritarianism, never govern alone.

I argue in favor of a focus on authoritarian trajectories and in favor of subverting the notion of “trajectory” to emphasize this dynamic nature of authoritarianism. Focusing on authoritarian trajectories means focusing on “dancing” itself, as opposed to its outcomes and what they tell us about the form of dancing we are seeing, or its shining moments. It also means focusing on dancing as a broader movement involving all dancers and, why not, the public who may support the performance through its attention and gaze, who may be beating the beat and the rhythm for the dancers, or only feigning interest. It means focusing on all dancers, as opposed to looking exclusively or primarily at the principal dancer and the corps de ballet, the inner sanctum.

But the notion of authoritarian trajectory is not a new concept in and of itself. The phrase is sometimes used offhandedly, especially in the media, to signal a regime’s turn to “harder” forms of authoritarianism, its descent toward more stringent and coercive forms of power. This use signals movement, to be sure, but only in one direction. In its more formal use in comparative authoritarian literature, the notion can largely be attributed to Levitsky and Way and their discussion of pathways or outcomes and the determinants of changes in regime types.Footnote ⁶⁸ This use captures movement, though understood primarily as a shift across types of regimes or away from or toward resilience, as has tended to be the case with scholarship more broadly. And so, the use of “authoritarian trajectory” has so far proved limited, only hinting at dynamism. The concept, as currently used, has even come to embody the limits and blind spots of recent literature, which focuses on Pepinsky’s “surface politics”Footnote ⁶⁹ and understands authoritarian trajectories as open/shut movements (from soft to hard, stable to unstable) or shifts from state to state, with little emphasis on the transition process itself or dynamism within each state. The lack of movement in these conceptions of trajectories across types offers some irony: The focus is on where we get to, as opposed to the process of getting there, on the destination rather than on the road.

Some would argue that borrowing a concept already so clearly connoted is a dangerous strategy. But the point here is not to argue in favor of “deep” trajectories over “surface” ones. Transitions or trajectories away and toward authoritarianism matter. But subverting the notion of authoritarian trajectory puts the spotlight on two alternate levels of analysis: the middle-range or regime level, especially in terms of the everyday or regular making and undoing of authoritarian power, and the micro or individual interaction level. It allows us to see authoritarianism in a more encompassing manner than the national and elite level focus has allowed. On the one hand, this redeployment of the notion of authoritarian trajectory stresses a regime’s path, conceptualizing authoritarianism as a process instead of as an “achieved” state or broad direction or outcome: describing the practice, navigation, and management of an authoritarian regime from the vantage point of the ordinary interactions that produce or are constitutive of it. On the other hand, it calls on us to understand a regime as a composite of multiple individual trajectories, all along the chain of relations that form the authoritarian system, from the more acknowledged and studied national elite trajectories, to local ones that encompass elite and non-elite alike.

From a middle-range standpoint, this approach means understanding authoritarianism as a dynamic process both in unsettled and settled times. Authoritarian governance is rarely, if ever, free of negotiation and challenge, even outside of times that could be categorized as “survival threatening” or “terminal” crises. Indeed, authoritarian power is never an achieved state; rather it is a relational process,Footnote ⁷⁰ regularly practiced in the face of challengers and populations that need to be brought or kept in the fold.

Among national-level elites, contrary to the notion of a buy-in, which seems to set up a dichotomy between settled “in” elites and unsettled “out” elites, vie-in from both “ins” and “outs” is continual and dynamic. Both co-opted and unco-opted elites continue to exert demands of various forms on the system, and to work with it, work around it, or even take it on. The boundary between “in” and “out” must be understood as fluid, and engagement across this line of in/supportive behavior and out/undermining behavior as much more common than a focus on determinants of endurance or change would suggest.

And, while elite’s continual vying-in may not always be what breaks the system – although more extreme forms of vying-in generally have a key role in doing so – it nonetheless ensures that the trajectory of authoritarianism is not straightforward or unidirectional, but made up of ebbs and flows, as the regime engages and seeks to manage this vying-in. Indeed, while the notion of trajectory has been too often used to imply a straight line toward a specific outcome or state, the notion of trajectory itself shuns such a straightforward understanding. Trajectories in the human world are rarely linear. Capturing a regime’s authoritarian trajectory is about looking more closely at the specific shifts the system makes over time as it tries to deal with the different forms of appeals and challenges eating away at its groundings and stability. Authoritarianism cannot be fully captured by looking at snapshots or “end shots”; rather, a full understanding requires a careful look at regimes’ origins and dynamic evolution over time, in response to both supposed trivial and critical challenges. Indeed, a key element of this understanding is the recognition that critical challenges are often born of an accumulation of contradictions, built over time from more ordinary forms of challenges.

Regimes may oscillate between types or borrow institutions or means from other types, including from democratic forms of governance, in these “within” shifts as opposed to radical shifts. Though they are not often the straw that breaks the camel’s back for a regime, these shifts can tell us a great deal about authoritarian strategies and means and the versatility, for better or worse, of authoritarian regimes, as they manage to stay afloat through both daily and critical challenges. Actors in authoritarian regimes may not always be effective, but they are adaptive and deploy strategies and means with more variety and strategy and in a much more quotidian and processual manner than the focus on violence or patronage or achieved states suggests.

In other words, a focus on authoritarian trajectories at the regime level entails focusing on the “concret du comment,”Footnote ⁷¹ the concrete and granular “how” of authoritarianism:

• Recognizing authoritarianism as a dynamic, ongoing process – a Sisyphean undertaking to attain stability, control, and reach, including in day-to-day relations,

• Which results from regularized vying-in across levels and the regime’s engagement with this vying-in,

• And which entails shifts within specific political arrangements that do not radically alter the nature of the regime,

• Suggesting a more varied and adaptive deployment of strategies and means.

But this also entails looking at autocratic regimes from a micro- or individual-level standpoint. The concept of authoritarian trajectory entails understanding the authoritarian process as the arch-trajectory of individual trajectories, further stressing the flux and the possibly contradictory currents within the arch-trajectory. This authoritarian process is not a black box, or simply an institution-focused process, but the composite of individual trajectories even if, from a standard constructivist standpoint, these trajectories are necessarily influenced by the structures and institutions within which they unfold. From this perspective, authoritarianism is a living and lived system, created, expressed, challenged, and engaged with by actors who seek to trace their own trajectory within the system. If the sum becomes greater than its parts, it is nonetheless agents who, in seeking their course and in managing their existence, make the choices and adopt the behaviors that create the system or challenge it, but who also, by choosing to navigate it, to vie-in, to carve out their space in it, reproduce or unmake it.

This understanding is especially important because it calls on us to understand authoritarianism as a system that rests on individuals for its deployment along an “implementation chain,”Footnote ⁷² with national political and administrative elites at the top. But the system is also very much reliant on local elites, political and otherwise, to act as its “implementers” at the base of the system, and local populations to serve as “recipients” and “adopters” of the system, though rarely in a straightforward fashion.

As antennas and implementers of the regime, local authorities can interpret and deploy authoritarianism with some variation, engaging what they recognize to be the “field of play.”Footnote ⁷³ And, like national-level elites, local authorities also know the system and how, to varying degrees, to manage it to trace their own trajectory within it, carving out their own space for action, operating with and for the system, but sometimes circumventing it or corroding it in the process. Considering the agency involved in deploying the authoritarian system locally, we should not assume the system is monolithic. The face of the regime projected by local-level authorities often takes on a local color and character. In other words, authoritarianism may manifest itself differently locally, especially when the system is softer.

In turn, this means that, across the same authoritarian system, varying space is available for populations to manage their individual trajectories. Reflecting on how we tend to think of individuals caught in dictatorships, Schedler explains that “[i]n the absence of individual autonomy and freedom, popular attitudes are always suspected as the products of authoritarian manipulation.”Footnote ⁷⁴ However, this rather stark understanding of the authoritarian state, and its ability to reach deep into individual lives to produce compliance, is not helpful in understanding local individual trajectories and what they entail for the broader system. Locals, in managing their lives, learn when it is rewarding to operate within and when to operate at the periphery of the system. But, for the most part, they operate within the system, managing their existence within it and its opportunities and constraints. They do not, however, operate simply in either/or compliant/resistant ways, or, as some may argue, in calculating ways, strategically looking to insert themselves in power networks and hierarchies. Rather, engagement is more ambiguous, as individuals work to manage under the circumstances, doing what it takes and focusing on their daily struggles as opposed to simply choosing to buy-in or opt out of the authoritarian system, or simply being co-opted or resisting, as the figures in the literature too often leave us with. Local trajectories should be conceived to be as dynamic as we take the arch-trajectory to be.

A focus on authoritarian trajectories at the micro-level thus entails

• Recognizing authoritarianism as a composite trajectory,

• Which entails that its implementation across different levels cuts across local trajectories,

• Acknowledging that local elites’ engagement with and deployment of authoritarianism may be varied and individualized,

• Which implies that the authoritarian system may take different forms locally,

• Suggesting the existence of space for local populations to deploy ambiguous engagements with the system, as they manage their existence within it.

From the regime-level to the individual-level, these are elements of the authoritarian trajectory we need to rediscover to see the more granular process of making authoritarianism, and how, in some of these very elements, we also find impediments to regime consolidation and the control authoritarianism seeks.

Methodological Conundrums of Studying Authoritarianism in Rwanda

Shifting our focus to this richer understanding of authoritarianism as a trajectory is not without methodological implications. There are very real questions surrounding our ability to study authoritarian realities, not to mention very real ethical dilemmas. While not all is dissimulation – or political for that matter – it is nonetheless sensible to heed Schedler’s advice not to operate based on a what you see is what you get (WYSIWYG) rule in authoritarian settings.Footnote ⁷⁵ Any study of authoritarianism should begin with a healthy dose of skepticism regarding our ability to fully capture the struggles and dissension within a regime, especially among the reclusive inner sanctum. Similarly, there is no straightforward way to get at the sense of threat felt by authorities and surrounding elites,Footnote ⁷⁶ knowing we are not in leaders’ heads and should raise doubts about the “public face” of a regime. And how can regime reach be measured beyond the capital and beyond the smokescreen of regime propaganda, to account for the population’s reception, that is, for how people really feel about their leaders, knowing trust may be an issue in an environment that forbids or shuns criticism?

These questions and an altogether different approach to the study of authoritarianism have been at the heart of what some are calling “political ethnography,”Footnote ⁷⁷ which brings an “ethnographic sensibility” to the study of authoritarian powerFootnote ⁷⁸. This follows a similar, prior intellectual move on the part of social scientists, and anthropologists in particular, to look at political violence from an ethnographic lens, which allowed us to rediscover some of its more ordinary or day-to-day manifestations, as practices and meanings, and some of its neglected actors.Footnote ⁷⁹ In the same vein, political ethnographic work calls on us to get at meanings, a broader set of practices, and other “unidentified” or “unseen” political objects,Footnote ⁸⁰ even those that are not readily observable or neatly measurable as, for example, institutions or structures. We need to reconstruct a “meaning environment,” with its interpretations and engagements, as well as its rich social and sociological relations, through the meticulous piecing together of cues and material to capture authoritarian moments or senses, to the extent that we can. More concretely, capturing authoritarian trajectories requires producing a good approximation of the socio-political universe of an authoritarian regime as a lived system, including from a subjectiveFootnote ⁸¹ or experiential standpoint.

This open, more ethnographic approach to authoritarian – and post-violence – realities has already made headway in recent scholarship on history, memory and how it is made manifest under Rwanda’s current authoritarian regime, which has stressed the need to go beyond surface realities. Through careful, and often blended or mixed strategies, scholars like Ingelaere, Fujii, Jessee, Mwambari, Ndushabandi, Purdekova, and Thomson, to name a few, have looked for the voices behind the dominant state narrative of the genocide, how it occurred and how it is to be remembered and memorialized.Footnote ⁸² This work has helped decenter the focus away from the unifying narrative imposed by the government to the plurality of perspectives and lived experience beyond the official script. Never in absolute contrast to one another, these state and unscripted realities exist in both tension and conjunction, which calls on us to forego rigid categories.Footnote ⁸³ This work has helped inform the overarching argument of the book, but also my thoughts and approach on studying highly politicised settings, as I explore here in the context of the First and Second Rwandan Republics.

Though we may never find ourselves in the heads of those promoting or living in authoritarian settings or we may never be perfectly able to capture the complex middle ground of the official and beyond, through “creative practices”Footnote ⁸⁴ framed by historically inclined methods and an ethnographic sensitivity, we can aim to capture authoritarian trajectories. These qualifiers are meant to stress the need for blendedFootnote ⁸⁵ methods and foci, more than purist understandings of historical methods, for example, not always keen on political analysis, or ethnography, with strict expectations concerning embeddedness and “full participant observation.”Footnote ⁸⁶ In other words, although no single perfect means exists to capture authoritarianism from a granular standpoint, blending approaches is a good strategy to capture the ebbs and flows of authoritarian trajectories and different levels of perspective, including the perspective of authoritarian publics.

Researching the First and Second Rwandan Republics raised some of the challenges of studying authoritarian trajectories, including from a historical standpoint. As much as I adopted blended “creative practices,” including document (archival and other) analysis and interviews, each method presented its specific challenges, compounded by the implications of my own position with regard to the subject of the book and research.Footnote ⁸⁷

The multisite archival work I undertook for the project focused on key collections found in Rwanda (National Archives and the Rwandan collection at the University of Rwanda), Belgium (Diplomatic Archives-African Archives and Diplomatic Archives), Canada (Order of Preachers-Dominican Archives), and France (Diplomatic Archives). These collections were used to develop a better understanding of the country’s political and administrative structures at the time, and regime policies and practices, including capture some of the more symbolic components of authoritarianism that were deployed at the time through material producing the “public face” of the regime (speeches, reports). They also helped me map key events, trends, and shifts over the course of the period, and assess, at least from the standpoint of the top, the regimes’ degree of “achievedness.” Studied closely, this material also suggested a sense of the system, as lived by national authorities, including whether elites in power felt threatened or not.

Surveying this material was nonetheless complicated by the state of Rwandan archival collections, with few government report series, and particularly ministerial reports series, having survived in their entirety. Two additional complicating factors were missing documents in other archives – sometimes to the great surprise of local archivists – and different archives’ limitation periods, restricting access to more recent documents. In addition, the now sensitive nature of Rwanda-related material, especially considering the genocide, translated into a clear reluctance to provide some documents, especially those concerning security. As a result, I often had to piece together material found in the archives to fill the gaps. This included, for example, looking at some of the remaining reports on political or development trends in the country, collating speeches on key reforms, studying external observers’ discussion of the stability or popularity of the regime or its main figures, building a basic record of authorities’ visits to the countryside, or assembling an impressionistic sense of the distance – physical, social, and economic – between Kigali and Butare (the Rwandan intellectual hub at the time) and the rest of the country through observers’ sometimes quirkyFootnote ⁸⁸ descriptions of their travels beyond the “big cities” or more important locales.

As can be imagined, the work also incorporated a strong dose of triangulation across all sources, as well as reflexivity regarding the structures and agendas behind the material collated, since each archival source introduced in its own way new forms of biases. As Nyabola so rightly pointed out, “[a]rchives are not neutral; they’re sites for contestation and projections of power. An archive is a living thing in which what is explicit and what is silent are equally important.”Footnote ⁸⁹ This equally applies to all archives produced by the different branches or components of a state, from ministries on their own functioning to its representation abroad. I relied significantly on Rwandan ministries’ reports and on diplomatic communiqués not because I believed in their absolute veracity, but to help piece together and convey the universe they belonged to.

Researchers, and historians in particular, have long warned of the limitations of archives. They are often incomplete, as I found in Rwanda and Europe, and yet overwhelming in terms of the amount of information they contain and the ability to process it in a representative manner. But the incomplete nature of archives needs to be understood at a deeper level. Archives are always a reflection of the power relations of their time, and of what and who they take to matter. What becomes an archive always comes down to what a state and society deemed worth to preserve and what was not. Archives are therefore a space of selective, privileged, powerful, or “stately” voices, simultaneously overshadowing others, sometimes in the form of a violent act of rejection or neglect.Footnote ⁹⁰ What gets archived is also rarely written as research material. It was written to respond to a specific need, incentive, or situation at the time or to keep a record. As researchers come to this material decades and even more later, archives never fit neatly their research questions. Rediscovering them generally comes in the shape of selecting what is judged relevant, “making archives fit” or making them make sense in a context that is not the one they were born of. As such, researchers always come at archival material with interpretations and bends, that is to say their biases and the limitation their ideas constitute. Mbembe quite trenchantly explains that archives “have no meaning outside of the subjective experience of those individuals who, at a given moment, come to use them. It is this subjective experience that places limits on the supposed power of the archives, revealing their uselessness and their residual superfluous nature.”Footnote ⁹¹ Archives are the embodiment of structural power, but when rediscovered by those who use them they are at the whim of their subjective reading. In other words, archives should not be taken at face value or as “proof,”Footnote ⁹² unless understood as epitomizing the temporary system of belief of the powerful at the time, and the interpretation of archives by researchers should never be taken as “objective.” And so, in piecing together the universe they represent, it is also essential to keep these deeper limitations in mind.

This is why I looked for recurrences in discussion of specific events in the different archives I mined, trying to establish the historical record, rather than presume statements were accurate, though this work always leads to some interpretations. When archival material seemed relevant but of a more uncertain nature or more impressionistic, I signal it in the text explicitly (indicating, for example, “apparently” or “purportedly”). The main strategy I employed, however, was to compare and contrast three main sets of archival documents, Belgian, French, and Rwandan, to triangulate the information they conveyed. This is not to say that I addressed all gaps and biases, but this look at both Rwandans’ and key external observers’ reading of the situation produced a rich, contrasted interpretation and even dialogue across sources (for example, around a key security crisis in 1963, engaging both the fears of Rwandan authorities in terms of their reputation and their appeals for help, while also reading about how Belgium, largely the target of these appeals, interpreted the situation and decided on giving help).

But more importantly, I complemented archival material with alternative sets of documents, used for similar purposes. These other documents included newspapers and journals published at the time, specifically Rwanda’s Kinyamateka, from 1958 to 1990, and Dialogue, from its creation in the mid-1960s to 1985. In addition, the research examined memoirs or the retrospective – introspective? – writing of several privileged observers, such as some of the last Belgians to serve in the Ruanda-Urundi administration,Footnote ⁹³ or, for subsequent decades, expatriates, development workers, and long-time residents. I included, as well, the writings of Rwandans who lived during the period in Rwanda, reflecting on what they observed. In the case of all three categories, the work also often served as a vehicle for the author’s political convictions and opinions, something both acknowledged and factored into the analysis of this type of material. Many would shy away from this type of material, which they would judge to be biased or to be simply opinion. I believe that the stances and debates they reflect, however, even if opiniated, are part and parcel of the trends studied in the book, not detached from them. Few of us fully think outside the box. These actors and their commentaries were shaped by and shaped the period. As a result, this type of contribution has much to add to our understanding of “camps” and perceptions at the time. Finally, I used scholarly writing produced at the time, though relatively sparse, as a “mapping” tool for trends and events, as well as “observer” material. A good number of the scholars writing at the time conveyed, through their work, their broader impressions of the country, even if those perspectives were not always couched as straightforward political commentary.

Getting at Lived Authoritarianism

None of these different sources are perfectly accurate, but together they provided a strong base on which to build a sense of regime structures, policies, key challenges, and the regime’s ability to handle these challenges and extend control, hence a base from which to assess the broader regime trajectory. Nonetheless, they fell short in terms of weighing in on the issue of local state reach or influence among Rwandans and their local experience of the system. With the exception of rare academic scholarship focused on assessing the regime’s presence in Rwandans’ lives,Footnote ⁹⁴ observers’ privileged positions – think of the ambassador more commonly in contact with elites in the capital than with a peasant in a remote location – often led them to neglect the more local perspective, or not to venture beyond the realm of very specific circles at all.Footnote ⁹⁵ To gain a stronger experiential sense of regime presence and influence in Rwandans’ lives, a series of interviews (fifty-one) were carried out. The bulk of these interviews were conducted in 2015, with an interpreter for the most part to allow participants to express themselves in Kinyarwanda, their mother tongue and often the only language they spoke. These interviews were intended to get a sense, from Rwandans themselves, of the importance of the state, regime, and authorities in people’s lives and of how people engaged with institutions and authorities. The interviews were conducted in the North (Rubavu, formerly Gisenyi) and Center-South (Muhanga, formerly Gitarama), the two key centers of power during the period (the Center and South of the country under the First Republic; the North under the Second Republic), as well as in the East (Ngoma, formerly Kibungo) and West (Karongi, formerly Kibuye), which were, and continue to be to some extent, areas more removed from power and power struggles. Two sectors were visited in each district, an urban one and a rural one.Footnote ⁹⁶ Kigali was purposefully skipped, since it is home to the growing class of socio-economic and political elites, with a non-negligible portion of its inhabitants having moved from abroad after the genocide. In a country where, according to the World Bank, 83 percent of the population remained rural in 2018, Kigali, much like many modern capitals, is an outlier.Footnote ⁹⁷ The focus of the project was on gauging ordinary Rwandans’ perspectives; Kigali seemed the least likely place to find them.

Participants lived in Rwanda under the First and Second Republics, or at least under the Second Republic. Interviewees were at least fifty-five and over at the time of the research, which guaranteed they were at least thirteen years old when the Second Republic was instituted and hence old enough to have a decent sense of political trends and events at the time and in subsequent years. Though some participants reported having spent some years outside the country, often to flee violence in the case of Tutsi, for example, all nonetheless lived in Rwanda for a significant portion of at least one of the two republics in Rwanda. Participants were required not to have been or to be political or administrative elites in national level institutions, though some played a role at local administrative levels, and in particular the cellule, secteur, and commune level, including responsables, a conseiller, and a bourgmestre. (For administrative levels, see Figure 1.1).

Figure 1.1 Pre-genocide administrative levels below national authorities (following the 1974 administrative reform)

Note to Figure 1.1: *The cellule was not an official level at the time of major administrative reforms in 1974 (Décret-loi 26/09/1974 portant organisation communale), though authorities regularly referred to it, including in official documents. Ministère du Plan (1974). Sous-préfectures were eventually added to the structure in 1981–1982, though they were suggested in both the 1962 and 1978 constitutions.

Participants had diverse socio-economic profiles; the main differences emerged from the urban-rural divide built into the recruitment process. Most participants in urban centers held low-paying jobs, working as caretakers, or street vendors, or operated small businesses, for instance selling second-hand clothes. They had little education or training; the few who had more education (high school and beyond) were teachers or leaders of local religious communities, for example. In rural areas, most participants identified as peasants, ranging from landless peasants working in the fields of others to wealthier peasants who owned cattle and grew products to sell. A significant proportion of rural participants described themselves, however, as subsistence farmers, farming for their family’s needs. Similarly, Longman describes peasant farmers as “characteristic of Rwandan social structure, liv[ing] not in villages but in isolated family dwellings surrounded by banana groves and fields of manioc, sorghum, and beans.”Footnote ⁹⁸ Like in urban centers, a few rural participants had had the opportunity to study. Those who had been more educated generally worked as teachers or religious leaders. With the exception of this local intelligentsia, most of the interviewees, in both urban and rural settings, lived within a short distance of where they were born. The majority of participants’ precarious situation is representative of most Rwandans’ existence at the time, living in a country categorized as one of the world’s poorest in the 1960s and 1970s. Seven out of the fifty-one interviewees were women. Participants were not asked to identify along ethnic lines, but ethnic identity, surmised from answers, reflected common, but rarely problematized estimates among the general population, with a slight overrepresentation of Tutsi and no Twa participants.Footnote ⁹⁹

Participants were recruited using different strategies, including through preexisting Rwandan contacts, mostly personal and nonacademic, the Rwandan branch of an international organization, local administrative authorities, local religious authorities, and informal encounters. Snowball sampling contributed additional interviews. Considering the authoritarian nature of politics in Rwanda, overreliance on local authorities was avoided, although authorities often furnished lists of potential participants or made suggestions from which some participants were drawn. Participants took part in a semi-structured interview lasting from thirty minutes to over three and a half hours. With few exceptions, interviews were conducted with the help of an interpreter and vetted afterward by a second interpreter/transcriber. A few interviews were conducted in French. Though interviews were the formal part of the research, to these should also be added many other informal conversations I have had since I first began traveling to Rwanda in 2004, both in Rwanda and among diaspora groups abroad, with members who left at the end of the period (late 1980s) or after the genocide.

Some of the key issues that arose over the course of the formal interviews included securing access to participants, while nonetheless ensuring their privacy and gaining their trust amid a highly authoritarian setting, as well as respondents’ ability to recollect accurately trends and events that took place well over twenty-five years ago. Rwanda is not unique when it comes to the two sets of issues.

First, after publications and reports critical of the current Rwandan government – in particular, its human rights record and authoritarian stance – the regime has become wary of academic research and has developed as a result a number of means to control research conducted in Rwanda beyond “innocuous” topics of research, including a national-level ethics process, sometimes with a strong political undertone, requests for multiple institutional authorizations, and mandatory local affiliation, among other procedural hurdles.Footnote ¹⁰⁰ But the main challenge was ensuring the privacy of participants and gaining their trust in the face of local authorities’ incentive to “keep tabs” on who does what and on outsiders’ presence and activities in the area. These important concerns are common to research conducted in challenging settings, including authoritarian ones,Footnote ¹⁰¹ and can only be addressed through sensitivity and reflexivity. Addressing them successfully requires an awareness of the environment and potential changes in it and an ability to balance the need for some form of acknowledgment and support from authorities with the need to protect – and communicate – research independence, especially to participants, as well as ensuring the ability to protect this independence. This rests in significant part on trusting participants’ ability to develop good strategies to navigate research constraints, empowering respondents to make decisions about interview locations and times, for example, or the management of interview topics. Finally, it rests on adaptability, especially of methods, such as recruitment processes and questionnaires within the bounds of a set, but flexible research protocol.

Another set of issues centered on participants’ ability to recall clearly events and trends that occurred decades ago. While age and health sometimes factored into participants ability to recall the period, the biggest challenge proved to be the legacy of the genocide and its influence on interpretations of the First and Second Republics. The genocide, and the war that preceded it, was a watershed for the country; unsurprisingly it is a prominent reading key for Rwandans that affects how they recall preceding periods.

There are a number of reasons for this. To start, it is one of most dramatic events participants have lived through, which means it probably colors how they recall institutions and authorities in previous decades. A clear break between pre-, during, and post-genocide was indeed not always evident for participants. Some participants could not help returning to the genocide in conversations. This tendency could also be a result of the official “sensitization” to history people have undergone, under many guises, which insists on drawing a clear, direct line between the First and Second Republics and the genocide. The current Rwandan government has a very clear and negative public view of the previous regimes – and European colonizers – as direct producers of the structures that led to ethnocentrism and the genocide. Under this ideological vision, speaking about 1960–1990 necessarily amounts to speaking about the setup to the genocide.

In addition, timelines often got blurred, including across the First and Second Republics, and at times even with the “Third” Republic. Sometimes, the blurring was purposeful. With social, political, and even legal limitations on what can be said about the genocide and the current regime, some participants appeared to welcome the opportunity to speak about the genocide in an unscripted manner or about the “Third” Republic, under the guise of speaking about the older republics. More often, however, people simply got somewhat mixed up considering similarities across political regimes and the length of time between the two republics and the interviews. It is also a result of the fact that many participants saw changes in Kigali as having little import in their lives. Although they often recalled key dates, such as the 1973 coup or broad generic trends – “Habyarimana stood for development,” for example – participants were not always as clear on specifics, such as secondary events or political figures at the national or prefectural level. These gaps further stressed the need for other types of sources for the “mapping” component of the research.

These are also not original challenges. Historians or historically inclined scholars regularly face and address these issues by remaining sensitive to their participants’ state of mind, interrupting the interview if needed or breaking it into shorter sessions, working with flexible interview tools to adapt to the flow, pace, or order of recollections, reemphasizing timelines, and (re)prompting when needed. It also stresses the importance of triangulating with alternative material and preexisting knowledge of the period.

Overall, while all material proved imperfect, each with its own gaps and biases, in combination the sum proved greater than its parts and helped translate the granularity and layers of complex relations at the heart of Rwanda’s authoritarian trajectories. In seeking to understand authoritarianism, it is also to this deeper look at events, trends, and actors that we must turn to find new ways of understanding both endurance and change in authoritarian settings, which have been at the heart of the most recent debates in the comparative study of authoritarianism. These types of “within” and more fine-grained dynamics have largely been neglected. This oversight is not uncommon, however. And it is one shared by many researchers commenting on Rwanda. Scholarship focused on political trends in Rwanda, when looking to the decades prior to the genocide, has tended to promote an “achieved” perspective, except for major moments of change and especially moments of ethnic violence. For the rest, when it comes to the in-between, few have judged it interesting enough to focus on, as the next chapter explores.