Patron-client relationships have been an inextricable part of political, economic, and social life. The logic of clientelism goes to the heart of precolonial domination, colonial maneuvering, the spread of Christianity, local politics and development outcomes, and even family relationships in Rwanda. It is one of the few stable features of a political landscape that has been punctuated by a series of violent regime transitions and the alternation of power between sets of Tutsi and Hutu elites.
Each set of ruling elites attempted to centralize power by eliminating or co-opting all viable rivals, and the basic distribution of power between rulers and the masses did not change significantly over time. The latter remained a rural population to whom the affairs of les évolués were mysterious – their fortunes remained hostage to directives and inducements emanating from above. Commoner citizens, Hutu and Tutsi, lived next to each other for hundreds of years, inter-marrying, lending a helping hand, swearing kinship, drinking together, and attending the same churches. Rural life necessitated interdependence and solidarity at the ground level. Their collective well-being, however, depended on the patronage and protection of authoritarian elites.
Ordinary Rwandans would not, of their own volition, choose to live socially separate lives along ethnic lines – but the ethnic identity of the rulers in place impacted ground-level dynamics. Each set of rulers instituted material disadvantages for the “other” group. The use of clientship as an instrument to extract self-interested submission was widely recognized. The more coercive and unequal varieties of patron-client relationships created a sense of common disadvantage among those caught in their web. The politicization of the identities “Hutu” (the dominated) and “Tutsi” (the dominant) can be directly linked to the institution of clientship as practiced under Tutsi elites of the monarchy in the precolonial and colonial era. Even though 90 percent of Tutsi were commoners enmeshed in the everyday variety of interpersonal relationships with Hutu, the politics of clientelism did create a substantial awareness of group-based identity and a lived experience of its concomitant protections and vulnerabilities. The tables turned in the post-independence era when ordinary Hutu were the beneficiaries (at least in official rhetoric) of Hutu elites in power, and ordinary Tutsi submitted, even consented to the Hutu Republics that wielded over them the discretionary power of life or death.
Each set of ruling elites found it necessary to formulate a legitimating claim to power. For Tutsi elites in the monarchy, it was the doctrine of inherent racial superiority; for Hutu elites of the Hutu Republics, it was the doctrine of popular revolution against a feudal system and rights for the majority race. The RPF derived its moral authority from stopping the genocide and promising to restore the unity it said people had once enjoyed before the colonial powers distorted it, and before Hutu elites eventually destroyed it. Each official script involved a partial reading of history and was used to explain the basis for those elites’ moral claim to power, and justify the manner in which that power was used. These legitimating scripts of moral rights and wrongs were backed by the coercive power of authoritarian elites who demanded absolute political loyalty. In times of crisis, that loyalty could be put to the test by calling for violent mobilization or other forms of participation in defense of the cause. The scripts could become consequential, with life and death implications for members of the “other” group. Not much is known about the extent to which ordinary people really believed or internalized the official stories they were asked to consume in each successive period of rule, but what is clear is that local social ties could rarely withstand the brunt of state-driven coercion, threats, and inducements. During the genocide, as during earlier episodes of violence (and even during the recently concluded period of gacaca justice), there were opportunists and wrongdoers and also some who were righteous among ordinary citizens.
This chapter does not intend to present an exhaustive account of each historical period. It presents only an overview of the pervasiveness and scope of clientelist relationships during successive periods of authoritarian rule. It makes the point that generations of Rwandans have been intimately familiar with the intertwined logics of ethnic politics and authoritarian patronage. They are sensitive to their own position relative to authority not only in the domain of the state, but also in interpersonal relationships. They are able to identify their interest under constrained circumstances and have adapted accordingly, always “know[ing] well what is expected of them” (Reyntjens Reference Reyntjens2011: 12). This chapter covers more than a century of historical time, beginning with precolonial Tutsi ruleFootnote 1 and ending with the military victory of the RPF. The section on precolonial Tutsi rule takes a close look at how clientelism generated power for the patrons and stabilized an unequal distribution of power between rulers and the ruled. It provides insight into the operation of clientelistic relationships as a mechanism of inter-group integration, center-periphery linkage, and governance, as well as the potential for destabilization under changing conditions of competition between alternative patrons. The section on the Hutu Republics continues to demonstrate the widespread prevalence of clientelistic relationships (both social and political in form, involving the exchange of protections and benefits for loyalty and support) but trades a comparably close look at the actual operation of clientelistic relations in political governance for attention to two themes – a discussion of the legitimating foundation of Hutu politics, and a look at the instrumental use of political violence.Footnote 2
This shift helps to lay the historical groundwork for the remainder of the book by identifying the main events and processes that the RPF reinterpreted for mass consumption in its formulation of an official ideology (see Chapter 2), but those messages were received with varying degrees of skepticism (see Chapter 3). This chapter takes a look at scholarly interpretations of those events – of Tutsi rule unsullied by the colonial impact – an era that the RPF defended and ordinary Hutu resented, and of the period of Hutu rule that the RPF attacked and ordinary Hutu defended (in part). In its final pages, this chapter presents an account of the descent into genocide. The scholarly literature suggests that popular mobilization was not the result of indoctrinated sympathy for the politics of genocide, as the RPF has been wont to suggest. Taken together, these sections provide the necessary background to show that RPF elites “misread” the experiences, collective memories, values, and attitudes of the Hutu masses (even when inclined toward charity and rehabilitation), while the latter struggled to connect with the new official narratives, mostly suspected bad faith on the part of the new ruling elites, and accordingly calculated their options under the new dispensation.
This chapter then serves two purposes. First, it highlights Rwandans’ long experience and deep knowledge of the reciprocal uses of clientelistic relationships under conditions of authoritarian rule, and their awareness of the conditions that can destabilize a clientelistic equilibrium. Among Rwandans, a clientelistic response appears to be an intuitively understood, rational, and deeply historical go-to strategy for those seeking control as well as those seeking protections or opportunities for advancement. As such, clientelistic logics are prone to persist because the norm of reciprocity is indeterminate in terms of the specific resource exchanges and mutual obligations it implies – therefore new patrons and different resources involve mostly an adaptation and adjustment to the way clientelism works under new circumstances (Lemarchand Reference Lemarchand1972: 76). Second, this chapter examines the realities – through a scholarly lens – of being on different sides of (1) the patron-client relationship under precolonial Tutsi rule and (2) the violence of the past under Hutu rule. It foreshadows the difficulties this created for good faith understandings between RPF elites and the Hutu masses (that will be fleshed out in subsequent chapters). The RPF regime was understood as a new period of Tutsi rule, and the cement of clientelism turned out to be the more likely resource for integrating rulers and the ruled than any sense of legitimacy. It was also less costly to both parties than the open use of brute force.
The Period of Precolonial Tutsi Rule
It was around 1860 that economic and political power began to accumulate in the hands of Tutsi elites.Footnote 3 By 1950, this group (about 10 percent of the Tutsi population in Rwanda) controlled virtually all positions of power and prestige. A vast web of patron-client relationships had enabled their control over the cattle wealth, land, and labor resources in the country. Their consolidation of power began prior to the arrival of the colonial authorities but was facilitated significantly by colonial rule (Germany 1899–1916; Belgium 1916–1962). The cumulative impact of the exactions of Tutsi elites and colonial authorities (particularly the Belgians) created massive rural discontent and paved the way for revolution in 1959.
A nascent set of Hutu elites had attempted to frame the opposition to elite Tutsi rule in terms of a class struggle of all disadvantaged commoner citizens (Hutu and Tutsi), but theirs was not as compellingly stark a frame as that proposed by another group of Hutu elites, in whose view the political problem was one that involved a fundamental clash between the inherent values and interests of two races. The terms “Hutu” and “Tutsi” had been used in oppositional ways for 200 years before the arrival of the Belgians – but the specific idea that “The Hutu” and “The Tutsi” were different racial groups was a colonial construct introduced by them. The racial thesis had already been instrumentalized by the Tutsi nobility to emphasize their inherent superiority and natural right to rule; decades later, the idea of race was embraced by Hutu elites who used their race-based majority to push for the “democratic” right to rule. It is not clear to what extent ordinary Rwandans believed at this time that they were biologically different groups, but the identity “Hutu” had come to be associated with slavery and cumulative disadvantage, and it resonated with vast sections of the peasantry that were tied up in predatory clientelistic links with their Tutsi elite patrons. A political platform promising liberation for the demographic majority had the potential to secure an automatic electoral victory for Hutu elites.
This racially polarized frame performed two important operations. It noted the collusion between the colonial authorities and Tutsi elites but singled out the latter as the more potent threat. It also obscured the ties between elite Tutsi and ordinary Tutsi by arguing that “The Tutsi” had to be overthrown if “The Hutu” wanted freedom from oppression. In so doing, the frame asked in effect that ordinary Hutu should blur the distinctions between their Tutsi friends and neighbors (with whom they had much in common), and their Tutsi elite patrons (whom they resented). Even though there was an outbreak of violence in 1959, ordinary Hutu did not mobilize in large numbers against their Tutsi friends and neighbors. Thousands of Tutsi went into exile and the monarchy was abolished in 1961 – heralding the end of Tutsi rule. Ordinary Hutu voted in local and general elections, laying the groundwork for the transition to the self-proclaimed First Republic in 1961 and then independence the following year. Three decades later, the RPF in power would ridicule the events of 1959 as an “assisted revolution” on account of the lack of popular mobilization. It would attribute the regime change not to the genuine grievances of the peasantry but to the opportunistic collusion between Hutu elites and the colonial authorities. Although Hutu elites remained attached to what they called the “social revolution” of 1959, RPF elites sought to destroy its moral and political significance by calling the events the first act of “genocide” in Rwanda.
Clientship Contracts and Control
The kingdom of Nduga (in what is today central Rwanda) formed in the 1600s and expanded in the next two and half centuries up to the borders of present-day Rwanda. The Tutsi Court’s penetration of the southwest region was accomplished in the last quarter of the 1800s. The Court had a tenuous hold on a number of other regions and struggled against rebellions in the north, west, and east well into the twentieth century. In distant parts of the kingdom, occasional levies and symbolic recognition of the Court was replaced by a denser web of connections as the Tutsi king planted delegates who acted as co-rulers with local kings and began to extract regular dues on land, labor, and cattle.
This process of integration was aided by conquest, marriage ties, and inducements of entering into clientship relations with the Court that promised regional elites more power and wealth than they could hope to gain via the status quo in their regions. They received command posts in the army, personal cattle endowments, management control over royal herds, and by extension, authorization to acquire pasture lands that would support those herds. Regional elites used these resources not only for personal enrichment but also to extend their own influence against rivals from within their territories. To be a client of the king also meant the promise of protection and common defense in the event of raids and incursions by surrounding kingdoms.
The king allowed the use of these herds to remain in the hands of his clients, but he could, at his discretion, claim all of the client’s cattle. This included the offspring of the herds granted by the king, as well as herds not originally granted by him. Since clientship was a hereditary relationship, future generations of the client stood to gain from these associations with the king, but they also ran the permanent risk of losing their assets in cattle at some point. These cattle clientships (called ubuhakeFootnote 4 – a herding metaphor) were known to be in effect as far back as the early 1700s. They implied an unequal contractual relationship “so compelling, in fact, that to accept a contract of this sort was to submit to the King.” At that time, so early in the history of the Tutsi Court, ubuhake became the “first pillar of ... authority over the kingdom. At the outset, the King was probably the only patron and through ubuhake he established an unequal and permanent political alliance between himself and other leaders” (Vansina Reference Vansina2004: 47–48; 181–188).
At the ground level, the main instruments of state power were the chiefs installed by the Tutsi Court. They acted as patrons, providing security guarantees for “Tutsi” cattle-owning lineages in exchange for receiving cows and luxury items as a sign of submission (C. Newbury Reference Newbury1988: 78). For non-cattle-owning farming lineages, defined as “Hutu,” the patron extended the loan of a cow for usufruct use to the farmer client who was expected to reciprocate with labor, gifts, and of course, loyalty.Footnote 5 Weaker actors sought patrons who would protect them from the competing demands of other powerful actors. Farmers sought out herders as their patrons, and lesser herders became clients of more powerful herder patrons. Thus, all “Hutu” had “Tutsi” ubuhake patrons, and less important “Tutsi” were ubuhake clients of more powerful “Tutsi” patrons. These were “integrative alliance(s),” relationships of “instrumental friendship” struck between unequals in which protection was exchanged for political support (74).
At an early stage, clients had some exit options available to them. They had the freedom to move elsewhere or seek out a more benevolent local patron if necessary. Still, there were few checks on the discretionary powers of the king. Smaller herders were sometimes driven to declare their cattle part of the royal herds in an effort to retain at least management control and prevent forcible seizure by the Court. Under colonial rule, as a growing number of powers became concentrated in the hands of fewer chiefs, clients would become increasingly “tied” to their patron and less insured against the latter’s arbitrariness. The rapid growth of population would also leave few places to escape to. Vertical patron-client networks diffused through society as conditions of political and economic insecurity grew in tandem with the rapacity of the chiefs.
Clientship and the Construction of Political Identity
It was in the precolonial era that individualized patron-client relationships began to emerge as a way for vulnerable people to seek a measure of social and economic security. Larger social entities such as lineages and clans diminished in importance and individuals were now vertically linked to more powerful actors and state agents – without the protective buffer of these social formations and kin networks. The common plight of the weak, caught up as they were in particularly humiliating and coercive clientship relations, produced a situation in which the “polarization between Tutsi and Hutu was already at the forefront of peoples’ consciousness” (Vansina Reference Vansina2004: 194).
As early as the 1600s, the terms “Tutsi” and “Hutu” had been used to designate unequal social statuses and opposite social categories. “Tutsi” referred to the wealthier elite among the herders, and later came to qualify only the political elite among this wealthy herder class. Two centuries later, as Court representatives began to be stationed in peripheral regions such as the southwest, the word “Tutsi” in those parts became synonymous with those proximate to state power (C. Newbury Reference Newbury1988: 52).
An individual who belonged to this elite group could be mockingly “othered” as “Hutu” if he indulged in lowly, unbecoming behaviors. In the 1700s, “Hutu” referred to the servants at Court involved in menial labor. Around mid-century, the term “Hutu” was applied to porters and non-combatants who serviced the “Tutsi” warriors in the armies. Vansina notes, “As most non-combatants happened to stem from lineages of farmers, the elite eventually began to call all farmers ‘Hutu’ and to oppose this word to ‘Tutsi,’ now applied to all herders, whether they were of Tutsi origin or not” (Reference Vansina2004: 135). Most humiliating was the demand for unpaid labor services of a menial variety (ubureetwa) that were imposed only on Hutu. Lesser Tutsi who were clients of more powerful Tutsi patrons were confined to more dignified tasks when it came to labor contributions. Introduced in the late nineteenth century at a time of famine, disease, and general hardship in central Rwanda,Footnote 6 ubureetwa demands grated on Hutu, causing several farmer revolts against Tutsi domination (that were all put down by the state). It had started out as dues for farmer tenants who cultivated land that was owned by a patron, but it later turned into a general demand imposed by the state-appointed chief of the land on all farming families within his jurisdiction.Footnote 7 Ubureetwa became an everyday practice of exploitation. Half of a farmer’s time had to be devoted to unpaid services to the chief; the latter also laid claim to a portion of the crops and the family’s food supplies. Farming families complied so as not to lose their fields. The disadvantages experienced by Hutu were something of a structural feature of social and political life before the colonial powers arrived. It was, according to Vansina, an “unbearable oppression” (Reference Vansina2004: 192).
Dwindling Choices for Clients
By the last quarter of the nineteenth century, the growing aristocracy had to be accommodated with endowments of cattle and land. Territorial domains began to be subdivided into smaller portions to placate larger numbers of notables and Court lineages – who in turn ratified ubuhake ties with as many clients as possible and raised extortion levels to make these endowments profitable. From a socioeconomic perspective, downward mobility was more likely than moving upward in the social hierarchy. For farming Hutu, it was difficult to be upwardly mobile, and even the cattle of smaller herders (occupationally Tutsi) could be seized by the king when the latter needed to replenish the royal herds. If this happened, these Tutsi became farmers, to be henceforth known as Hutu.
Exactions upon them had increased, but clients were linked to the fate of the patron they had attached themselves to. Every decade or so, there were mass purges, executions, and political realignments as factions competed for power and influence at the Court. Chiefs and notables who were denounced or disgraced could be killed, and their families, clients, and loyalists massacred as well. In the colonial era, there would be few alternatives for political and economic security other than the institution of clientship. The introduction of administrative changes meant that control over land, cattle, and people began to concentrate in the hands of a few chiefs – driving vulnerable individuals to pre-emptively seek the protection of those “patron-chiefs” despite their increasingly unfavorable terms.
The Colonial Era
The twentieth century opened with the installation of the new King Musinga by way of a bloody but successful coup.Footnote 8 The brief period of German rule (1899–1916) coincided with the first half of his reign. The Germans’ preferred strategy of general non-interference and indirect rule did not produce substantive changes in the way the Court administered the country. In the wake of Germany’s defeat in the First World War, however, Belgium acquired the Rwandan colony and chose to become deeply involved in governance for nearly half a century, from 1916 until independence in 1962.
Competition and Collusion between Alternate Patrons
The Court entered into an arrangement with the Germans in order to secure military protection against the incursions of foreign traders and the advances of the Belgians in neighboring territories. The Court also allowed the White Fathers (les pères blancs)Footnote 9 to open missions in select areas because it was believed that the Fathers were clients of the Germans and could not be refused outright. Sensing an opportunity for protection from the exactions of Court notables, ordinary Hutu sought the patronage of the White Fathers over that of the Germans.
The latter’s firm support for the Court was widely known – their official position openly stated that they would back the Court even if it meant overlooking the grievances of the Hutu population. The Court made every effort to guard against the public impression that the Germans were a rival or more powerful entity than the Court. This was difficult when the Court ended up paying a fine to comply with German directives on certain matters, or when the Court paid lavish tribute to the visiting members of the German royal family. To discourage challengers from seeking German patronage, the Court spread rumors that the Germans were, in fact, clients of the Court. The few taxes imposed by the Germans were also collected by Court agents to convey the impression that the Court was indeed the unrivaled authority in the land.
The Court used German backing to quell internal challenges from rival notables and rebels. The Germans also acted as intermediary during the Court’s tense confrontations with the White Fathers who had begun to wield substantial powers in their mission areas. The Fathers had backed certain military challengers to the Court. They had also emerged as a source of patronage for Hutu in their areas who took religious lessons and attended the missions as a way to avoid the customary obligations owed to notables and patrons with whom they had ubuhake ties. They used the mediation of the Fathers to resolve disputes with their patrons. Quickly realizing that “a chief who has a cow has control over the man who has borrowed the cow” (Des Forges Reference Des Forges2011: 62), the Fathers entered into ubuhake patron-client relationships lending out the cattle they had purchased or rustled together by conducting raids on the hills. They distributed tracts of land that were part of the vast mission areas and revoked those grants when disobeyed. They distributed other material incentives: the prospect of an education at the seminary, and salaried jobs at the missions. In most cases of conflict, German intervention helped the Court and the White Fathers find face-saving compromises that did not significantly hurt the interests of either party. If the Court was forced to back down on any particular issue, the Fathers made it known that the Court was on its way to becoming their client.
These dynamics continued under Belgian rule. Catharine Newbury has noted that the Court and the Belgians had a shared interest in extraction and the maintenance of power, and each made use of the other (Reference Newbury1988: 117). They were separate entities nonetheless and this created some opportunities for local actors to leverage the presence of the Belgians as an alternative patron. On various occasions, they sought the intervention of the Belgians against the excesses of Tutsi elites. Belgian administrators were forced to address such issues even if they did not deal with them very seriously. The White Fathers also moved closer to the Court at this time. Unlike the German Protestant administrators, the Belgians were keen to have the Court agree to mass instruction and conversion to Catholicism. The missionaries were eager that the Court check the influence of Protestant denominations that had arrived in Rwanda. The Court, for its part, needed the missionaries to stop their (the missionaries’) patronage of dissidents. As notables formally began Christian instruction, their clients followed in their footsteps as well. The mass acceptance of Christianity was driven by clientelist logics (Des Forges Reference Des Forges2011: 214).
If the Belgians checked or side-stepped the Court, it created suspicions that the Europeans intended to establish themselves as the ultimate power in Rwanda. King Musinga’s efforts to contain the Belgians backfired, and in a few years, they engineered his ouster, replacing him with his son, Rudahigwa, who was more amenable to collaborative rule. Despite tensions and occasional confrontations, it was not until the very end of colonial rule that the Belgians abandoned their support for Tutsi rule. Thus it was for almost half a century that the Belgians and the Court collaborated in mutual interest. This collusion was fruitful for both parties in the short and medium run – but unleashed changes at the ground level that would ultimately lead to regime change.
The Provision of Moral and Material Resources for Tutsi Rule
The Belgians introduced the “Hamitic myth” according to which the Tutsi were a distinct race, genetically closer to the Europeans than the African Bantu stock. They were believed to have originated somewhere in North Africa – a region spatially proximate to civilized Europe. A theory of physiognomy went hand in hand with a theory of inherent values and abilities. Tutsi were supposed to be taller and lighter-skinned than Hutu. They were also supposed to be inherently gifted with superior intelligence, higher moral virtues, and ruling skills – no wonder then that they had secured their domination on Rwandan territory. A great deal of colonial “scientific” effort went into defining a set of physical features and moral virtues for these “races.” Hutu were categorized as short, stocky, and darker skinned with blunt features – the markers of an indigenous and allegedly inferior gene pool that lacked refinement and intelligence. Hutu were thought to be naïve and child-like, less reserved with their emotions and uniquely suited to hard labor and peasants’ work.
It was not surprising that the Tutsi nobility embraced this view of inherent superiority and wove it into courtly rituals and oral histories. Court ritualists invented traditions that made the Court and its control over Rwandan dominions appear timeless and natural (D. Newbury Reference Newbury2002.). Hitherto a marker of social status, economic and political power (or lack of it), the identities “Tutsi” and “Hutu” underwent a dramatic metamorphosis. As newly race-based identities, they were now to be inscribed on identity cards, transmitted across generations by blood, and inherited down the paternal line. The ethnic census and the ethnic identity card were institutional innovations that reinforced the fixed race-based categories and transformed them into a hard structural feature of Rwandan demography. In terms of the race-based head count, Hutu emerged as the majority group, representing more than two-thirds of the general population.
In the spheres of education, economy, and politics, the colluding patrons – the Court, the Belgians, and even the Catholic Church – pursued actions that produced the “collective subordination of this ethnically defined group” (C. Newbury Reference Newbury1988: 116). In Catholic schools, Tutsi children were taught additional subjects, such as French language instruction and record keeping, congruent with their status as the ruling class. Hutu children received an education that prepared them for labor-oriented jobs in mines and industry. In any case, Hutu children were often unable to attend school because of the heavy corvée burden in their areas. The Belgian-funded school up and running in 1932 – the Groupe Scolaire d’Astrida – mandated for many years a minimum height requirement as a filtering mechanism for incoming students, and sought to equip Tutsi elites with professional qualifications for the modern world. Only a small number of Hutu were allowed to enroll.
Increasingly Coercive (Less Reciprocal) Patron-Client Relationships
In an effort to increase revenue extraction, the Belgians undertook administrative changes that concentrated powers in the hands of a small number of hill and provincial chiefs. Given their limited oversight of the chiefs, the latter were able to act arbitrarily, adding their own demands to the existing burden of taxes and diverting those resources for their personal ends. Ironically the pressures and uncertainty this created drove people to seek these same individuals as their patrons. They hoped to mitigate the chiefs’ predatory behavior with a measure of protection that was customarily expected from patrons toward their clients. However, since the chief and the patron were one and the same individual, there were few avenues left to appeal abuses.Footnote 10
While ubureetwa (unremunerated labor) demands continued to be imposed on all Hutu men and represented mainly a form of humiliating service,Footnote 11 it was ubuhake (cattle clientship) that was the main mechanism of control. If individuals owned personal cattle and needed access to grazing areas, they could be driven to seek ubuhake patrons from among chiefs or sub-chiefs who controlled that land. Tutsi “patron-chiefs” used these ties to undermine possession of their clients’ personally owned cattle (C. Newbury Reference Newbury1988: 140–141). Clients symbolically surrendered their own herds to avoid their forcible seizure. With this preemptive move, clients retained use rights to their personal or lineage cattle and simultaneously secured the protection of the patron. If a client was loaned a cow, it could mean that the patron had appropriated the right to seize the client’s personal herd when he needed it, or he could now ask the client to perform a range of menial services from physically transporting him on a hammock to taking a beating in his place (since this patron could be a client for someone even more powerful). In general, the least reciprocal and most exploitative relationships were those in which the greatest power differential existed between patron and client.
The Belgians clearly understood the “submission” value of ubureetwa and the “stabilization” function of ubuhake institutions. The decision to continue ubureetwa was justified thus: “the principle of workdays owed by the Hutu to a notable was an expression of the latter’s obedience,” while a 1938 report warned that the abolition of ubuhake would destabilize chiefly authority and would be “dangerous for the peace” (C. Newbury Reference Newbury1988: 141, fn. 76; 145, fn. 86). It was the scrutiny of the UN Trusteeship Council that served as the impetus for the abolition of ubureetwa in 1949. Money payments replaced unremunerated labor obligations. The Belgian administration acted with the consent of King Rudahigwa to announce the abolition of ubuhake in 1954. A small section of notables even acknowledged that it was a form of “slavery.” Chiefly control over pasture land continued however, as did work for the chiefs (on an informal basis) in many places until the late 1950s. These incremental efforts at reform could not stem the demand for regime change.
The Hutu “Social Revolution” (1959–1961)
A nascent Hutu elite educated in Catholic mission schools had emerged on the national landscape. They tapped into the “resentments of the dispossessed” and demanded nothing less than a systemic transformation (C. Newbury Reference Newbury1988: 116).
The “Manifesto of the Bahutu”Footnote 12 (1957) claimed to speak in the name of everyone who was Hutu. It cited irreconcilable differences between “The Tutsi” and “The Hutu” races. The problem was understood as the Tutsi “political, economic, and social monopoly which, given the de facto discrimination in education, ends up being a cultural monopoly, to the great despair of the Bahutu who see themselves condemned to remain forever subaltern manual laborers” (see Nkundabagenzi Reference Newbury and Newbury1961). Like Tutsi elites who had instrumentally adopted the racial difference thesis because it had served to justify domination, these Hutu elites also appropriated the racial thesis and fashioned it to their political ends. With talk of Belgian withdrawal and the fervor of nationalism sweeping the colonized world, the “ethnic temptation” (Chrétien Reference Chrétien, Amselle and M’Bokolo1985), that is, the electoral implications of the Hutu majority vote, was too great to resist.
Grégoire Kayibanda and his party PARMEHUTU (Parti du Mouvement de l’Emancipation Hutu) transmitted their message at the ground level with the help of a small cadre of grassroots Hutu leaders. They were local mission-educated school teachers, catechists, traders, and contract employees of the Belgians. The literate read to the illiterate the writings of Hutu radicals from publications such as Kinyamateka, the Catholic weekly. Peasants stood silently in their fields holding placards protesting “Tutsi colonialism” to a visiting UN delegation (Newbury and Newbury Reference Newbury and Newbury2000). Aware of international pressure on the colonial administration to initiate democratic reforms, Hutu elites now appealed for the recognition of their claims. Tutsi conservatives, however, refused to recognize there was a “Hutu-Tutsi problem.” They were unwilling to concede more than social and economic reform, and dispensed with the Belgians on the crucial matter of appointing a successor to King Rudahigwa (who had died in Bujumbura). When violence broke out in November, the Belgians intervened in favor of the Hutu revolution.
The first strikeFootnote 13 was launched by a group of Tutsi youth against a Hutu sub-chief – in retaliation, an uprising spread rapidly through rural areas, particularly those that were PARMEHUTU strongholds. The violence was aimed mainly at Tutsi chiefs, but spilled over into torching and looting the homes of ordinary Tutsi as well. Small groups of Hutu militants perpetrated the violence; the general population protested for the most part by simply refusing to obey Tutsi authorities. To their Tutsi friends and neighbors, they offered sanctuary or escorted them to safety across the border.
An armed effort to suppress the uprising was launched by the Court and the conservative Tutsi monarchist party UNAR (Union Nationale Rwandaise), but the Belgian Governor enlisted support from Belgian paratroopers and Force Publique soldiers in the Congo to intervene in order to protect Hutu leaders and restore order. As ordinary Tutsi as well as Tutsi chiefs and sub-chiefs fled Rwanda for neighboring countries, the Belgian authorities reversed the structure of power by appointing Hutu in interim administrative positions. The monarchy was rejected in a nationwide referendum. Independence on July 1, 1962 marked the end of Tutsi rule in Rwanda.
The Uses and Abuses of the Ideals of the Social Revolution
For Hutu elites – moderate and radical – the social revolution represented the ideal of “freedom from enslavement” (Uvin Reference Uvin1998: 26). It became the focal point of Hutu elites’ emotional attachment and political conviction – a cause worthy of any self-respecting Hutu whose political identity had become indexed to a history of oppression. But the events of 1959 were a series of uprisings more than an organized revolution. It had been successful because of colonial intervention and did not ultimately lead to fundamental changes in political values or governance. For these reasons, Prunier writes, “although the 1959 revolution was a fake, it was nevertheless a fake based on truth. The feelings, the revolutionary hope for a better life, the willingness of the masses to organize were all present” (Reference Prunier1995: 347).
Unfortunately, the social revolution produced in its wake a new set of victims – the newly dispossessed Tutsi living in exile. It legitimated Hutu rule but could not in the end serve as a practical check against the excesses of the new regime. Rwanda transitioned into authoritarian rule only thinly veiled by its claim to democracy by definition (the rulers represented the majority ethnic group). The Hutu Republics institutionalized discrimination against Tutsi living within the country. The repression and exploitation of the Hutu masses (the supposed beneficiaries of the new regime) created another round of rural discontent – strategically deflected by extremists toward “The Tutsi” enemy in the context of economic crisis, domestic political changes, and an invasion from the north in the early 1990s. When in crisis mode, radical Hutu elites usually called for the “defense of the revolution.” This was a moral obligation, a civic duty, and a call to support government policy in order to combat dangers the depth and urgency of which the population was unable to assess on its own. As the civil war escalated, Hutu extremists made a case for genocide as the ultimate defense of the gains of the revolution.
Under RPF rule, moderate Hutu elites tried to salvage the moral defense of the social revolution from the indefensible violence that had been unleashed in its name. But RPF elites would not concede that ground. They sought to jettison this legitimating foundation of Hutu politics by positioning the events of 1959 as the first of a series of acts of “genocide” against the Tutsi.
Power and Patronage under Hutu Rule
The Republics (1962–1973; 1973–1991) were single-party authoritarian regimes that sought the loyalty of the masses by proclaiming Hutu as the preferred beneficiary of their patronage. Given the mutually beneficial premise of clientelistic relationships, Hutu who gazed upon the new regime persuaded themselves that “they, the humble peasants from the hills, somehow shared in that power ... suitably thankful to their ... superiors ... who helped them” (Prunier Reference Prunier1995: 81). Quotas were introduced for Tutsi in education and public employment. Under the Second Republic, for instance, there was not a single Tutsi mayor or governor in local or provincial administration. There was only one Tutsi officer in the army, two Tutsi parliamentarians, and one Tutsi cabinet minister (75). Lands left behind by Tutsi refugees were redistributed to Hutu peasants. Hutu at the local level understood the significance of these changes. University administrators, teachers, and students organized themselves into “comités de salut public” to appeal to the newly defined public good and oust their Tutsi colleagues and fellow students (Vidal Reference Vidal1991).
No effort was spared to remind the country that power was to be used in defense of a moral cause – the correction of the historical oppression of “The Hutu.” Glaring excesses were justified by moral arguments. The state oversaw the massacre of hundreds of Tutsi civilians living within the country. This was retaliation for the armed incursions of small groups of Tutsi refugees in the 1960s (the inyenzi attacks).Footnote 14 President Kayibanda signaled the resolve of the new republic to defend “The Hutu.” In a genocidal threat, he proclaimed that the “whole Tutsi race will be wiped out” in the event of continued attempts to restore Tutsi rule (Erny Reference Erny1994: 62–63). In neighboring Burundi, the massacres of thousands of Hutu by the Tutsi-dominated regime served to make the point that the survival of the Hutu state could not be taken for granted.
In 1973, Kayibanda’s military chief Juvénal Habyarimana undertook a coup d’état that he justified as a “moral revolution.” He claimed that the country had strayed from the ideals of the 1959 social revolution. His official speeches were littered with references to the danger of allegedly Tutsi qualities: “the spirit of intrigue and feudal mentality,” the “harm” done to society of “the one who refuses to work.” On the one hand, he was the “top shebuja [patron] in the land” (Prunier Reference Prunier1995: 86) and celebrated the value of “manual labor,” pandering to “The Hutu=The hardworking peasant” as the moral standard-bearer of the Rwandan state (Verwimp Reference Verwimp2003: 29). On the other hand, his policies reduced the peasant population to “extreme pauperization” and the extreme “reduction of life chances” (Uvin Reference Uvin1998: 3).
Tutsi civilians, for their part, quickly learned to be submissive to the will of the authoritarian Hutu Republics. There were no political parties or pressure groups to represent their interests. Of those that had survived the massacres in the 1960s, about 40–70 percent (140,000–250,000 Tutsi) went into exile at this time. Those who remained were loyal to their political “patron” President Habyarimana for stopping the massacres of Tutsi and calling for interethnic reconciliation. There were benefits to political quiescence. Quotas were not implemented to the letter when it came to positions in the public sector (these excluded top-level portfolios). In non-state sectors, such as business and NGO-led development projects, there were large numbers of Tutsi in employment. The numbers of Tutsi students at the National University had dropped from 90 percent at Independence to 40 percent by 1970 (Mamdani Reference Mamdani2001: 135); still, the census figures for 1991 showed that Tutsi were on the whole better educated than their Hutu counterparts (Verwimp Reference Verwimp2003: 24). The quotas, then, worked less as a tool of direct discrimination, than as “administrative proof” that “the state was watching out for the interests of the majority Hutu” (Uvin Reference Uvin1998: 35–36). Of course, the unrivaled patron could arbitrarily withdraw these allowances. In times of crises, politicians demonstrated their commitment to the defense of Hutu interests by insisting that the quotas be applied strictly.
Rwandans were intimately familiar with the politics of patronage and selective benefits. Both Republics were founded on clientelistic logics according to which administrative functions were distributed, and state functionaries owed loyalty not to the citizenry but to those who had appointed them (Gasana Reference Gasana and Guichaoua1994). If President Kayibanda, who hailed from the south, favored the southern region in terms of public investment, and preferred southern Hutu for state jobs and so forth, then President Habyarimana did the same for the north. One-third of the most important government positions, and almost all of the top positions in the army and security apparatus were held by people from Gisenyi, Habyarimana’s home province in the north (Reyntjens Reference Reyntjens1994: 222). Almost 90 percent of all public investment was directed to the provinces in the north and west that had historically low populations of Tutsi (Verwimp Reference Verwimp2003: 23). In the north, even land-based clientelist structures (legacies from an era prior to the establishment of Tutsi Court control in the region) returned in “full force” (De Lame Reference De Lame2005: 60).
At the ground level, clientelism continued to be an integral part of the everyday rural experience. Although cattle clientship (ubuhake) had been abolished in 1954, in some places, the poor continued to search for a local patron and “beg for a cow,” not for the economic wealth that the cow represented, but in order to record the relationship and compel the patron to make good on the promise of protection the relationship entailed (Gravel Reference Gravel1967: 328). Ordinary Hutu without connections to wealthy or powerful patrons remained cut off from the stream of benefits that flowed into local communities from above. The government directed to specific regions the flow of foreign aid for agricultural development – particular ministers and administrators then allocated contracts, benefits, and jobs to political allies and family members. More than two-thirds of rural development project funds would go toward salaries, cars, and infrastructure for rural development projects. Project assets such as cars were used for personal ends; investment funds and land designated for project purposes were allocated to local powerbrokers – traders, political cadres, military men, and rural administrators who were close to urban elites (Uvin Reference Uvin1998: 123–124). The decade of the 1980s was closing with “feelings of failure and disillusionment.... The fallout from these projects (was that) ... profits went to the protégés of the regime” (De Lame Reference De Lame2005: 64).Footnote 15The peasants understood full well that one had to attach himself to a more powerful actor in order to survive and advance. Informal patron-client ties persisted in everyday life – only this time the patrons were Hutu (Lemarchand Reference Lemarchand and Lemarchand1977: 89; Chrétien Reference Chrétien, Amselle and M’Bokolo1985: 156). In her study of small-scale brick and tile makers in the 1980s, Jefremovas noted how connections to politicians were needed to secure access to clay land and to get contracts to sell their products. Even kin networks and gender relationships involved clientelist ties. Women attached themselves to influential men in order to gain a measure of control over resources and the economic surplus. A son could look upon his father as a patron in the expectation that his obedience and labor for the family would be rewarded with an inheritance in the form of a piece of land. If such an expectation was spurned, the son might enter into a clientelist arrangement with someone else in order to get hold of farming implements, such as a hoe (Reference Jefremovas2002: 85, 106).
Jefremovas noted the personal and often arbitrary nature of the clientelistic relationship: “everything depends on the will and affection of the superior.... [I]f he wishes, he will take back what he has already given in order to give it to someone he likes better. An inferior, whether serf ... or wife ... can only protest in vain” (Reference Jefremovas2002: 66, citing Albert Reference Albert and Paulme1971). In a richly detailed study of rural life and social relations in the late 1980s, De Lame observed various forms of clientship ties including some approximating the old ubuhake cattle-based transactions that Rwandans had once used to forge alliances for mutual benefit. Despite rituals of conviviality that people were reluctant to give up, De Lame astutely noted that clientship only remained a source of “fragile cohesion.” It “stage[d] a relationship,” “masking unequal exchanges....making them palatable” (Reference De Lame2005: 376–382, 465).
The Beginnings of Rural Unrest
Long before the 1990s, Lemarchand had noted that clientelistic networks, exchanges, and tacit understandings underpinning ruling elites could become dysfunctional with the growing disaffection of the “client” masses and the introduction of competing political “patrons” (Reference Lemarchand1972: 81–82). Indeed, as the broader political and economic environment changed, the single-party MRND (Mouvement Révolutionaire National pour le Développement) regime discovered it could not count on the undivided loyalty of the Hutu masses.
The peasant experience had become one of unrelenting coercion, deprivation, and humiliation at the hands of those who wielded power. They suffered a massive loss of income as the world price of coffee (a smallholder crop) crashed. Rwandan peasants could hardly feed themselves from their tiny parcels of land; the youth were unable to search for a better livelihood because of restrictions on migration. Fertility rates dropped in anticipation of a bleak future with uncertain chances of survival. Although peasants were highly taxed, the government invested no more than 5 percent of capital expenditures into agriculture (Uvin Reference Uvin1998: 56, 149). Every effort was made to extract the maximum possible resources from peasant labor – there were fines and punishments in case of non-compliance.
At this time, international pressure for liberalization at the end of the Cold War combined with domestic opposition from southern Hutu politicians. Multiple parties were now allowed to operate and President Habyarimana’s MRND party lost its chokehold as the only viable patron. By 1992, a coalition-based transitional government was in place with representatives drawn from parties whose leaders had defected from the MRND. As soon as peasants sensed the implications of these developments for MRND control at the ground level, participation in the hated umuganda (system of obligatory but unremunerated communal labor for one day a week) fell drastically and had to be abandoned as policy. It had long stood as a symbol of ruling elites’ “discretionary power over labor” (Verwimp Reference Verwimp2003: 22). Rwanda was experiencing drought and famine conditions when the world price of coffee crashed. In severe distress, peasants directed their frustrations against the state. In many places, they surreptitiously destroyed development project infrastructure, uprooted coffee plants (banned by law), and threw stones at elites who drove by in their luxurious vehicles (Longman Reference Longman1995).
The RPF’s invasion on October 1, 1990 provided an opportunity to reassert control by turning the restive population against a new target. Extremist wings of political parties formed their own militias, and by late 1993, weapons had been distributed to arm local “self-defense” groups. In many places, massacres of Tutsi civilians were organized by local leaders acting on the order of higher authorities.
Public officials and the extremist media raised the specter of a “Tutsi threat” by obliterating the crucial distinction between Tutsi civilians living inside the country and the Tutsi rebels fighting from the outside. Tutsi civilians were accused of working as accomplices of the RPF and the image of a “two headed snake” was conjured up to convey the unified nature of the threat. In December 1991, a secret military commission report in Rwanda officially defined the ‘‘main enemy’’ as ‘‘Tutsi inside or outside the country, extremist and nostalgic for power, who have never recognized and will never recognize the realities of the 1959 social revolution and who wish to reconquer power by all means necessary, including arms” (ICTR Prosecutor v. Ferdinand Nahimana et al., 2003: 36, paragraph 112). An intergenerational line of continuity was drawn between Tutsi guerrillas of the 1960s and their descendants raised in exile who spearheaded the RPF. According to the March 1993 issue of the extremist magazine Kangura, “The history of Rwanda shows us clearly that a Tutsi stays always exactly the same, that he has never changed.... Who could tell the difference between the inyenzi who attacked in October 1990 and those of the 1960s. They are all linked ... their evilness is the same” (HRW 1999: 73).
They also advocated for a stark social and economic separation of the two groups. As early as December 1990, the Kangura had published the “Hutu Ten Commandments” that called on all patriotic Hutu to “stop having mercy” and cease with immediate effect every personal and professional relationship they shared with Tutsi. All those of mixed heritage (ibimanyi) were suspect as well. A Hutu father and Tutsi mother would mean that their children were Hutu, but according to extremist logic, those children would still carry biological Tutsi traits, those deeper qualities that were gene-based, that made them a threat to the integrity of the Hutu state. They were the enemy “in disguise,” dangerous because they were able to pass themselves off as Hutu. The idea that an enemy lurked among them that needed to be unmasked created an overwhelming sense of paranoid urgency (Eltringham Reference Eltringham2004).
In some ways, the biological argument made the extremist rhetoric easy to comprehend: “The Tutsi enemy” was every individual of the Tutsi race. However, accepting this argument as just and reasonable was a different matter. Ordinary Hutu were being asked to brutally sever the everyday ties that bound ordinary Hutu and Tutsi together.
The Civil War and Genocide
We know little about how these messages were received and processed by ordinary Hutu. We do know that they participated in the thousands in varying capacities.
Scott Straus has argued that “without war... genocide would not have happened” (Reference Straus2006: 226). War created an atmosphere of uncertainty, even panic, and helped extremist leaders make a case for genocide in self-defense. War’s state of emergency created the context for the hardliners to succeed at the elite level. In turn, they made sure that only those local leaders remained in power who subscribed to the hardliners’ position. The president’s assassination on the night of April 6, 1994 was a result of the war and political impasse. It allowed Hutu hardliners to blame it on the ultimate “provocation” by the RPF and its Tutsi civilian accomplices. Genocide was de facto state policy as long as the hardliners were in control of the state (224–239). Lemarchand noted that the Arusha peace accords in August 1993 had secured disproportionate gains for the RPF and needed to be scuttled by hardline Hutu elites. The mass killings of Tutsi civilians was, in this scheme of things, a “rational” strategy because it would immediately elicit the collapse of any prior agreement and foreclose the possibility of compromise with the RPF (Reference Lemarchand1995: 10). Alan Kuperman has argued that the RPF invasion was a “provocation” for which genocide was “retaliation” by the Hutu regime. His emphasis is on the strategic interactions between warring actors. He notes that except for the Holocaust, most cases of mass killing since the Second World War were triggered by groups that violently challenged the authority of the state and invited retaliation that paved the way for the group’s own demise. In the Rwandan case, the RPF attacked first, and fought on despite being aware that their attempt to seize control of the state put Tutsi civilians at massive risk. Ultimately, “when the rebels finally were poised to seize control of the country, the Hutu regime retaliated by killing more than three-quarters of Rwanda’s domestic Tutsi population in barely three months. Ironically, the Tutsi rebels then defeated the Hutu government but gained control of a country whose Tutsi population had been devastated” (2004: 61).
The logic of self-defense had over the years unfailingly translated into disproportionate retaliatory actions against Tutsi civilians for acts of aggression initiated by Tutsi rebels. In 1992, Léon Mugesera, a senior MRND official from the northern province of Gisenyi, had announced in a public speech: “If someone gives you a slap, give them two in return, two fatal ones” (Article 19 1996: 18). The argument now morphed from a solely defensive second strike into a preemptive first strike in self-defense. Extremist elites stoked fears of renewed Tutsi oppression. They escalated feelings of panic by hinting at Tutsi plans for the systematic killings of Hutu. Mugesera taunted his audience, “Are you really just waiting blissfully to be massacred?” (Article 19 1996: 19) He warned them: “Know that the person whose throat you do not cut now will be the one who will cut yours” (HRW 1999: 86).
To establish killing as a necessary war-fighting tactic, extremist elites manipulated the decision-making process of rational individuals. They cut off ordinary Hutu from every source of alternative information. Their monopoly of the discursive space was used to bombard their captive audience day and night with urgent descriptions of the rebels and their “accomplices” closing in on them. Attacks were staged in communities to serve as “evidence” of the truth content of these warnings (Fujii Reference Fujii2004).
This was masterful manipulation of a fearful population in wartime. These were people who had long been denied the power to make free and fully informed choices under authoritarian rule. They were malnourished and psychologically damaged from years of poverty and servility, and had recently absorbed severe economic shocks. Now faced with an existential crisis, they turned with apparent willingness to scapegoat the target identified by the state (Uvin Reference Uvin1998: 66). Given Rwanda’s long history of state-enforced popular mobilization, groups of ordinary citizens were used to turning out and performing state-mandated tasks. Vertical structures of state penetrated deep into the local level, making it extraordinarily difficult for individuals to evade the ever-watchful eyes of the state. Non-compliers could be identified and punished (Straus Reference Straus2006).
The accumulated body of work on the Rwandan genocide suggests the roles of coercion, state enforcement, opportunism, and fear – so far as the vast majority of ordinary Hutu perpetrators are concerned. For these “captive participants” (Verwimp Reference Verwimp2003: n. 262, citing Gupta Reference Gupta1990), the absence of eliminationist anti-Tutsi sentiments is a significant finding. Straus, for instance, finds “strong evidence of inter-ethnic cooperation before genocide.” Yet he also points to “collective ethnic categorization,” a mechanism by which “perpetrators substituted the Tutsi category for the individuals they were attacking” even as they recognized (in hindsight) that their targets had not been directly involved in the war. Although this substitution had occurred under conditions of fear and uncertainty, it was made possible in the first place because of the preexisting mental category of “Tutsi=enemy” (Reference Straus2006: 9). This was a “script,” always available for radical Hutu elites to activate and adapt to their needs. How it was interpreted and implemented on the ground, however, depended on local dynamics. Fujii (Reference Fujii2009) explained how the same individuals decided to save certain Tutsi and target others depending on the social networks they were part of and the situations that they encountered at different times.
Conclusion
Not all Hutu were direct participants in the genocide. The majority were bystanders who failed to stand up to state-sponsored genocide and struggled to make sense of the enormity of the devastation they had witnessed around them. There were also extraordinarily courageous and resourceful individuals who had provided refuge and carried out risky rescue efforts because of which many Tutsi survived. Then there were Hutu victims – thousands were killed by the RPF and thousands of others were killed by the génocidaires.
When the RPF won, it organized on October 1, 1994 a “triumphalist celebration” to commemorate the fourth anniversary of their attack on Rwanda (Newbury and Newbury Reference Newbury and Newbury1999: 293). It was unbecoming given the immediate context of the horrors that had been perpetrated, but they believed that the day marked the culmination of the first phase of the liberation of Rwanda. The Hutu masses did not necessarily think so. The legacies of history kicked in as they tried to make sense of the maneuvering and emerging policy positions of the RPF. On the one hand, they feared the return of elite Tutsi rule and were inclined toward distrust from the very outset. On the other hand, as the RPF rapidly emerged as a hegemonic party able to eliminate or contain every source of opposition, there were few options other than to cultivate its goodwill. Given the long history of clientelism in social, economic, and political spheres, vulnerable individuals implicitly understood that they needed to solicit the protection and goodwill of this unrivaled patron.Footnote 16
