Review Essays

Bringing the Peasants Back In:
Agrarian Themes in the Construction and Corrosion of
Statist Historiography in Rwanda

DAVID NEWBURY and CATHARINE NEWBURY

A: Each group tells a different story. They all say "My ancestor was shaykh of the land." . . .

B: Perhaps you could present several versions, each one from a different source, and not say "this one is true; that one is false."

A: No way! . . . If I present different versions, they will accuse me of sowing discord. They will say, "He just wants to make problems." That's a sure thing. Besides, there is truth; there is lying. How can I treat all sources equally? By God, it would bring disaster.

Conversation with a Jordanian local genealogist.¹

After a while the truth of old tales changed. What was true before became false afterward.
A Kuba elder.2
The genocide that began in April 1994 marks an infinitely sad watershed in Rwandan history. It also marks a potential watershed in Rwanda's historiography, for such a rupture with the past raises necessary questions on the ways in which this past will be seen in the future. In the wake of the genocide, many people were drawn to comment on Rwandan society. They began with all good intentions, but too often these observers had little or no background on the country. Their introduction to the society and its people was through the genocide alone, and their intellectual experience often drew heavily on the models of previous genocides elsewhere. So the themes they chose to focus on—ethnicity, violence, and the state—were those most obvious to the genocide itself. But Rwandan history—and its historiography—as well go far beyond the issues raised in the genocide. In fact, we argue, the relation between history and the genocide, as portrayed in many recent accounts, needs to be reversed. Instead of seeing history exclusively through the genocide (as many did, misleadingly referring to "500 years of tribal warfare"), one can only understand the genocide through an understanding of Rwanda's history.	1
Therefore, although this article is not about the genocide, this may be an appropriate time to pause and review some aspects of Rwandan historiography. We especially want to explore an aspect of the historiography by which politics have often been removed from rural life and agricultural practices separated from political life. This disjunction of rural actors and formal politics has been characteristic of many writings on the genocide, which tend to focus on elite politics and central actors alone, but it is also characteristic of broader Rwandan historiography as well. Rather than attempting to provide a new integrated reinterpretation of the Rwandan past, our intent is to inquire into how this separation of peasants from politics occurred, asking how Rwandan history came to be understood and interpreted in the manner that it was. This is an essay in historiography, not history.	2
The dominant vision of Rwanda's history emphasized political homogeneity, ethnic distinctions, and the power of the state. In so doing, it sublimated alternative visions, which expressed regional particularity, diverse forms of identity (based on kin, class, occupation, and friendship networks), and the interaction of local agency with elite policy. We start by noting both the content and the silences of these conversations on history—both what the presentation of Rwanda's history includes and what it omits. We then address the question of why incorporate peasants within an established elite history, and we present various neglected themes that, by their very nature, tend to transcend (and transform) the dominant historiographical patterns. Finally, we discuss several important works in Rwandan rural studies, exploring the ways by which their inclusion in the historiography might lead to broader historical understanding.	3
We have abbreviated many of these discussions, and undoubtedly will have omitted some observers' "favorite" topics and sources.3 But our intent is not to provide a complete inventory of sources on Rwandan history.4 Instead, we seek to raise neglected issues and to provide an intellectual grounding on how such themes might reshape an understanding of Rwandan social dynamics. Such an approach compels one to move from a static, statist vision to a complex consideration of the forms of peasant participation that pervaded the history of this society. What we have tried to do is indicate the possibilities of a rural dimension to conventional history and a historical dimension to rural experience. In short, we seek ways to bring Rwandan peasants into the understanding of politics and politics into the understanding of rural society.	4
To do so, we draw on wider intellectual paradigms. While in this essay we do not present peasants acting, we nonetheless hope to pry open a few cracks in the monolith through which subalterns can find a place in their own history. Our analysis draws on two analytic approaches that have gained some prominence over the last several decades, roughly coinciding with Africa's postcolonial historiography: peasant studies and subaltern studies.	5
Peasant studies became one of the principal themes of postcolonial historiography in Africa for two related reasons: the general absence of peasants from the colonial literature and the development of methodologies that allowed rural subjects to be included in historical inquiry.5 In general, however, the recent features of African peasantries differ from those of other global regions. First, while peasant models from elsewhere tend to privilege tightly knit villages, in Africa peasants often lived dispersed across the land in interacting networks, not residential communities.6 Second, African peasants constituted a substantial proportion of the population (differentiating them from many Latin American models), and they often retained some autonomy in relation to the state (differentiating them from many Asian models). The combination of these two elements meant that, though not unchanged, African peasantries have endured through time well into the postcolonial period, when in other areas they have been more fundamentally transformed.7 Third, historical methods to study peasants in Africa developed mostly through the use of qualitative data (testimony) rather than quantitative data (statistics); the disciplinary gap remained wide between those who sought to understand social process through oral sources and those who sought knowledge through statistical figures.8 Thus, despite significant differences among peasants within Africa, peasantries in Africa differed from peasantries elsewhere in their relation to the state and in the disciplinary approaches by which peasants were studied.	6
A second important postcolonial school of inquiry has been that of subaltern studies, developed principally through the work of South Asian scholars who sought to move beyond elite-centered analysis.9 In exploring the interaction of elite and subaltern classes, these scholars sought to reveal the forms of intellectual dominance of those in power and to portray rural residents as competent historical actors with their own goals, ambitions, resources, and defenses.10 For the most part, however, such approaches were not new to historians and anthropologists working in Africa. From the late 1950s, these analysts had moved beyond the history of colonial actors to inquire into the experiences, initiatives, and exploitation of rural Africans. To account for the importance of local initiatives as well as external power, they privileged voices from below—voices outside the hegemonic script, the dominant intellectual and cultural patterns of thought and behavior. In order to do so, they developed a new set of methods—drawing on oral sources, archaeology, linguistics, ethnography, botanical analysis—and proceeded to submit these unconventional source materials to historical critique through a new set of analytic tools. Historians referred to these paradigms as "social history," to distinguish this approach from earlier histories of political elites and to move beyond the formal institutional histories that dominated colonial historiography.	7
Thus scholars working on Africa have long focused on the recipients of state power. They have drawn on colonial documents by reading critically, accounting for the ideological assertions or the outright fictive reconstructions that pervaded colonialist discourses. They have looked for the dialectics at the intersection of hierarchical power and the moral economy of the oppressed. And they were aware that colonialist discourses could be drawn on not only by the European powerholders but also by local actors in circumstances where these promoted their particular goals. During the colonial period in Rwanda, for example, politically dominant groups adopted colonial hegemonies and drew on colonial force to extend their power well beyond the effective administrative reach of the precolonial monarchy. Analogies with European concepts of feudalism, portraying clientship as a "social glue" that pervaded social intercourse at all levels, gave the appearance of political homogeneity to Rwandan culture; at different times, members of different politico-ethnic factions, both Tutsi monarchists and Hutu counter-elites, had recourse to colonialist paradigms of the racial foundations to social stratification (which we discuss below).11	8
Like subaltern studies, therefore, social history in Africa was counter-hegemonic in its conceptualization, and, like the subaltern practitioners, social historians in Africa "read against the grain" of colonial documentation and explored the silences of history produced by elites. Nonetheless, these studies differed from South Asian applications because the research context—and the sources—differed from South Asia, as did the lived experiences of political actors. Today, social history, like subaltern history, is seen as too complex to be folded into a single rubric. But the principles and goals that animated these fields of scholarly research were shared by analysts in Africa and South Asia. In Rwanda, however, historians were slow to focus their attention on peasants, whose histories were effectively sublimated within the history of the state and seen as part of a society that was culturally homogeneous, even if politically stratified. In most accounts, the image of Rwanda overpowered the ethnographic realities. It is to that image that we now turn.	9
Located near the geographic center of Africa (just south of the equator and not far from the mid-point between the Atlantic and Indian Oceans), Rwanda held the allure, for many observers, of the paradigmatic precolonial African state: centralized, stratified, ethnicized, and "feudal."12 Like all stereotypes, each of these characteristics fit uncomfortably with the historical record. But popular lore is powerful, and European lore postulated for Rwanda an "ancient" history marked by exotic origins. (Egypt, Ethiopia, or an early Judaic "tribe" were all candidates.) Heroic migrations and military conquest added to the allure of this historical drama played out in the highlands of Central Africa—a landscape combining open plains in the east, stunning volcanoes in the west, and, in between, the beauty of a thousand hills in infinite variation, all garnished with the splendor of a luxuriant tropical "natural garden." To European observers, Rwanda was "le pays des milles collines," the "Switzerland of Africa."13	10
But most of all, what attracted outside observers (and produced one of the largest bibliographies of precolonial Africa) was the people. For this tiny, densely packed country of 10,000 square miles (about the size of Vermont), with a current population of more than 7.5 million (compared to Vermont's 600,000), was most renowned for its ethnic configurations, which were heavily idealized in popular lore. From the legacy of early twentieth-century observers, these ethnic groups took on dramatic distinctions: in these views, ethnicity was supposed to pervade all aspects of Rwandan culture, from personal capacities to political stratification. So pervasive was this assumption that, without empirical evidence, the three principal ethnic groups—Tutsi, Hutu, and Twa—were seen as entirely separate: separate in their cultures, histories, and racial origins.	11
The precolonial aristocratic lineages, which were Tutsi, were celebrated by European observers for their physical characteristics—tall and slender, with aquiline noses and fine hair: "black Europeans" some called them.14 Although these characteristics were associated primarily with members of the aristocratic class, they came to be applied to all Tutsi. In fact, however, there are significant physiological variations within this social category, and even at the apogee of the precolonial monarchy, the political classes—those who wielded power—accounted for less than 10 percent of all Tutsi.15 In other words, by the assumptions of "corporate ethnicity," there was a near universal tendency among outsiders to apply the dramatic physical features of the ruling lineages indiscriminately to all Tutsi and to assume that all Tutsi were members of the political elite; the reality was significantly different.	12
Similarly, "Hutu," a social classification that included the vast majority (about 85 percent) of the population, were assumed to be invariably short, sturdy, and dark. And physical features were taken as indicators of invariable intellectual and cultural features: while all Tutsi in this racial matrix were seen as "refined" and "born to rule"—as befit the "governing classes"—all Hutu were seen as "naive," "stalwart," and "easily duped"—as befit the laboring classes, in this sociological-cum-racial wonderland. Nonetheless, here again, as with the Tutsi model, the perceptions of "corporate ethnicity" were inaccurate; "Hutu" were never a homogeneous group, not in historical, social, or cultural terms.16 Despite the differences among Hutu, however, the "dual colonialism" of central court administrators, under the suzerainty of first German and then Belgian rule, brought an awareness of shared exploitation within the expanding powers of the state.17 During the 1950s, within a rapidly changing political climate (in which the colonial power, the Catholic Church, and the United Nations all had important roles), Hutu leaders responded to widespread rural discontent and organized a political movement that eventually overthrew the monarchy and led to the formation of a republic under Grégoire Kayibanda (1962). While not unique, the fact that this "revolution" was one of the few meaningful social transformations to accompany decolonization in Africa added yet another layer to the aura of Rwanda.	13
A third social category, "Twa," was composed of people whom Westerners associated with pygmies; popular imagery portrayed them derisively as "pygmoid": shorter still of stature, dark, and "powerfully built," with short legs and round heads. As a social category, however, the Twa were composed of disparate groups showing diverse characteristics. Some were attached to the royal court (at least one serving as an important notable under the late nineteenth-century king Rwabugiri); others were more on the margins of state-defined society. Twa were often exempted from state obligations, and, though individually renowned for their strong personalities, as a group they were often disregarded as well as disdained in state politics. Because of their numbers (less than 1 percent of the Rwandan population), cultural diversity, and economic marginality, Twa were not often accounted for in political analyses; instead, they were usually relegated to the status of exotic appendages to Rwandan society. Nonetheless, their presence was fully a part of the Rwandan social matrix.18	14
Such were the alluring features of Rwandan social imagery. But postcolonial work on Rwanda has moved beyond the late nineteenth-century racial assumptions and the early twentieth-century whiggish paradigms of earlier representations. By contrast, postcolonial studies stress the social diversity, individual agency, and historical transformations that have been central to Rwandan social processes, in precolonial as well as colonial times. The social groups noted above did not "arrive" as corporate groups, or with their current labels; instead, more recent social identities emerged as part of the larger processes of social flux, individual action, and political power. There was a great deal more individual mobility and interchange than any static model of some collective "Rwandan past" can account for. The state was not "created" by a single culture hero or even by a single group. Power and ethnicity did not coincide originally; they took shape and salience in relationship to each other, not in confrontation with each other. Before the mid-eighteenth century—and in some contexts, long after—region was more important than royalty in defining identity, and ecology more influential than ethnicity in molding people's lives.19 In short, there was much more individual confrontation, social contestation, and local effervescence than normally ascribed to historical process in this region.	15
An example of the modern-day landscape in western Rwanda, showing terracing, crop variety, and hilly terrain. Photograph by Danielle de Lame, courtesy of the Musée Royal de l'Afrique Centrale, Tervuren, Belgium.
In what follows, we will first trace the general contours of precolonial historiography, stressing the local orientation of the early accounts and their transition to an emphasis on royal history. We then examine the broader factors that influenced this transformation, and the way in which the writing of Alexis Kagame came to incorporate the one within the other, sublimating local data to central court perspectives. But postcolonial work came to challenge some of the centrist historiography of the colonial period; our next section surveys some of the available documentation on rural transformations. Finally, we explore more recent trends that have addressed the postcolonial crisis of the rural areas and given voice to rural actors. Still, the rural concerns of local peasants and the centrist agendas of state administrators have largely been kept autonomous in the historiography, we argue, as rural transformations have been viewed as "technical" issues, while centrist policies have been seen as "political" processes, independent of local influence. Rethinking Rwandan historiography, therefore, requires giving central place to the dialogue that existed between these two domains—and challenging the premise that Rwandan history was royal history alone while local agency only became significant when incorporated into the state. Since these assumptions were especially prominent in precolonial historiographies, we turn to a consideration of how these histories were forged and framed.	16
The first histories—oral histories—were local histories; and of course local histories still exist in profusion, though they are often relegated to the shadows of irrelevance.20 During the colonial period, however, what came to be seen as salient was state history, the history of the elites, the features important to outsiders. But history as process can only be understood as the interaction of local agency with external influences—through commercial, ecological, and social contacts with others. Consequently, with the development among Western historians of methodologies of oral history from the 1960s, postcolonial work has come to consider oral sources as essential to the reconstruction of Rwandan history.21 In the process, postcolonial historians have done much to return a sense of diversity, fluidity, and confrontation to history, to broaden the range of social classes and intellectual themes included, and to introduce a much richer understanding of colonial rule and colonial culture to the written historical accounts.22 Yet, despite the breadth of new knowledge and revisionist understanding such approaches allow, local data have been largely marginalized from the conventional historical accounts.	17
The earliest written accounts on the Rwandan monarchy reflected those of the earlier oral accounts: they often stressed the regional roots to historical knowledge and emphasized the complicated combination between the penetration of royal power and the sometimes superficial character of royal rule. From these accounts, it is clear that local dynamics were often local as well as dynamic, and that regional particularities were often more prominent than the broader cultural generalizations extended to the entire society. The German writings make this especially clear; in the early years of colonial rule, for example, Jan Czekanowski underscored the importance of regional particularities, especially contrasting the north and northwest of the country with the royal court norms that characterized the central and southern regions.23 Later historical writings, even through the 1930s, continued to highlight regional cultural characteristics—and the tendencies toward regional autonomy from the court.	18
Père Albert Pagès published the first great classic on Rwandan dynastic history, Un royaume hamite au centre d'Afrique, only in 1933. But because this work represented the fruit of over twenty-five years of observation and involvement with Rwanda, it captured some of the flavor of earlier accounts deeply rooted in local sources.24 Even while this account celebrates the ruling social stratum and the royal line and focuses on Rwandan culture heroes (especially the epic kings Ruganzu Ndori and Kigeri Rwabugiri), the events associated with them are often presented in very localized fashion; their heroism is shown in local—and sometimes humble—events.	19
Rwandan royalty. Center right, Musinga, the king of Rwanda from 1896 to 1931, surrounded by four of his wives; to his right is Kanjogera, the queen mother. On Kanjogera's lower legs, note the ubutega anklets, a product acquired through the Lake Kivu regional trade network. Copyright Africa-Museum Tervuren (Belgium).
Similarly, the two-volume study by Père Louis de Lacger, Ruanda (1939), is an important transition work in this regard. It includes both local and statist viewpoints, although they are kept chronologically separate: local elements are most prevalent in his presentation of the precolonial period, statist perspectives in the colonial section.25 Throughout his presentation of precolonial history, de Lacger stressed the regional differences and historical autonomy of the different areas that eventually became "Rwanda." Presenting the expansion of the Nyiginya central court on the model of the extension of French royal power from an "Ile de France," de Lacger gave full credit to the cultural integrity and political autonomy of the various communities conquered by (and to varying degrees absorbed into) the larger monarchical state—factors omitted in later works.	20
Nonetheless, turning to the colonial period, de Lacger celebrated the "Hamitic" rulers, and he presented a much more homogenized, less conflictual image, focused particularly on the march of Christianity and the preeminence of Catholicism. Not long before de Lacger's book was published, the Belgian administration had deposed the king (Musinga), citing his opposition to "European civilization."26 Shortly thereafter, Musinga's successor Rudahigwa—carefully selected by Belgian authorities in consultation with the principal Catholic prelates—was to dedicate the country to Christ the King. In celebrating this success story of the Catholic Church in Rwanda, de Lacger lost sight of the diversity, heterogeneity, and lines of contestation that marked the earlier sections of his book. But his approach also reflected more fundamental changes in Rwandan historiography at about the time of his research, from full attention to the regional disparities in the earlier works to a concentration on the royal court and the homogenization of Rwandan society in later colonial works.	21
As part of the principles of the Missionnaires d'Afrique, the earliest Catholic order in Rwanda, the missionary priests (also known as the "White Fathers") had been encouraged to study local customs and were required to learn the local language; thus much of the early written historiography was the work of Catholic priests.27 In the early years of European presence in Rwanda, because of their involvement in local issues, the perspective of the White Fathers differed from the views of the secular authorities. From the beginning, German administrative policies extended the power of existing elites tied to the royal court, in both Burundi and Rwanda.28 But German colonial authorities were few: for example, in 1913 in Rwanda (a territory of more than a million people at the time), there were five German administrators compared to forty-one missionaries (thirty-four of them Catholics). So the German colonial administration depended on mission personnel for local contacts. Following World War I, when Belgium took over Rwanda, this civil dependence on ecclesiastics increased dramatically. The Belgians were new to the area, preoccupied first with the war and subsequently with boundary issues with Britain; because of these factors, they felt pressure to assert greater effective administrative presence than Germany had done. In this endeavor, they turned to their Catholic co-religionists—and found them to be willing cohorts.	22
The Catholic missionaries were well placed. The strategy of the court had been to situate missions in regions distant from the central court, areas recalcitrant to the authority of the Rwandan Nyiginya dynasty; often, these were areas of outright resistance to court rule.29 In ten years, nine missions were established, seven of them in outlying areas, to varying degrees resistant to court control. Such placement by the king was intended to remove missionaries from any meaningful involvement in the activities of the court, but it also meant that early missionary writings took full account of regional cultural distinctions and local social action. The missions were thus suitable as institutions advancing hegemonic expansion for the court as well as for providing intelligence reports to the administration.30 So the Catholic Church provided many benefits the incoming administration lacked: personnel, familiarity with the languages, cultures, and histories of the different regions, and strategic spread.	23
But the shift from German to Belgian rule after World War I was also a shift from a predominantly Protestant to a primarily Catholic power. To the Catholic missionaries in a period of intense sectarian competition, this represented a welcome change. After all, in the course of religious wars during the 1880s in Buganda (a kingdom on the north shore of Lake Victoria, not far from Rwanda), the White Fathers had been driven out by the Muslim faction. They had taken refuge at the south end of Lake Victoria, converted the exiled king, and returned to take power in Buganda, only to be defeated again by the Protestant faction—which had benefited from the timely arrival of British arms and advice.31 Subsequently, Baganda Catholics had been placed in what the missionaries saw as a secondary status (to Protestants) in the colonial structures of Buganda. Therefore, in Rwanda, the White Fathers were happy to participate in a colonial administration favorable to the creation of a Catholic state; indeed, the guidelines of proselytization for the order were to work through the existing political elite.32 In addition, the Belgian administration eventually found an ambitious leader to work with, one eager to participate in a catalytic role: Père Léon Classe, the vicar-general of the order. The infrastructure for forging an orthodox—or in Pierre Bourdieu's terms, a doxic—history was thus firmly in place: it was composed of an alliance of ecclesiastics and administrators.33	24
However, while these were the internal features to the colonial historical drama, it was an external conflict that, through an alliance of ecclesiastics and administrators, sparked a transformation of the colonial historiography of Rwanda. World War I had a significant impact on Africa. At enormous cost to the local populations, Britain and Belgium had mobilized their colonial subjects to fight the Germans in East Africa. In the case of Rwanda, many men were mobilized as soldiers and porters, but a severe famine in the northwest, resulting directly from the hostilities, took the lives of many as well.34 The political effects following the war were no less significant. After the war, the two victorious European allies sought to divide the territorial spoils whose acquisition had disrupted the lives of so many Africans. Britain sought to annex the areas immediately west of the Kagera River, areas they claimed were particularly well suited for the construction of the Cape-to-Cairo railway.35 They premised their demands on the claim that much of this area had formed the kingdom of Gisaka, until recently politically autonomous of the Nyiginya royal court of Rwanda. Without going into the details of this diplomatic struggle, it is enough to note that such claims directly affected the internal consolidation of colonial rule.36 Most important for our current interests, a shared reaction to the British claims helped consolidate an emerging alliance between the three parties noted above: the Belgian colonial authorities, the central court of the Rwandan monarchy, and the Catholic Church. They countered with the claim that the Rwandan central court armies had conquered Gisaka (a claim contested by many in Gisaka) and therefore had extended its territorial dominion to the Kagera River. Any deviation from this position was considered treason to Belgium, subversion against the king, and (since Britain was seen as a Protestant power) near heresy to the Catholic Church.	25
Whatever their other differences, these three factions agreed on the threat that British claims posed to the integrity of the state. The royal court saw a threat to its ideological and material bases, for the eastern areas in dispute were particularly important to the monarchy, both symbolically and materially. These were the areas where the Nyiginya monarchy claimed its origins. Although in the eighteenth century, the center of the monarchy had moved to the west, vast areas of this region nonetheless continued to serve as the principal pasturage of the royal cattle herds.	26
The church saw this region as important because of the possibility it presented for the intrusion of Protestant missions to a Catholic domain—a possibility they wished ardently to thwart. Although German Protestants had been established in Rwanda since 1907, the Church Missionary Society (CMS), based in Britain, was seen as the greater threat; it was backed by the diplomatic clout of a major colonial power, the financial might of an industrial state with a strong tradition of missionary expansion in Africa, and the resources and experienced personnel of one of the largest and most established missionary orders in the region. In fact, during the British occupation of this area after World War I, a semi-autonomous evangelical faction of the CMS had established a mission—in Catholic eyes, a "beachhead"—in this region.37	27
The Belgian authorities saw such claims to the area west of the Kagera River as a threat for three reasons. It was a slight to their colonial integrity and national pride. Furthermore, they relied on the monarchical institutions (as their administrative grid) and on the Catholic Church (for carrying out social policy); they hardly wanted to offend their allies in the administration of Rwanda. Moreover, the Belgian administration feared that, through the railway scheme, British financial penetration to this area could also draw off labor from this presumed "colonial labor reserve," which had been a central element in colonial thinking on Rwanda and Burundi. Before the late 1920s, Belgian administrators had directed Rwandan labor toward the mines in Katanga and elsewhere in the Belgian Congo; they saw the emigration of Rwandan workers to British territories as detrimental to their own economic aspirations.	28
The primary response to the British claims, therefore, was to assert the hegemony of the royal court in this region. And the principal respondent was Monseigneur Classe. But this debate was not simply a series of diplomatic niceties carried on behind closed doors. It was a debate that raged at the League of Nations, and it was well publicized in both missionary circles and the popular press. In the process, it transformed the nature of Rwandan historiography. Instead of the earlier emphasis on (or at least recognition of) regional autonomy and regional distinctions, the written works now focused almost exclusively on royal ascendancy; this shift promoted a centralized view of "Rwandan" history. In sum, in responding to this perceived external threat, the new "doxic" vision of Rwandan history consolidated the administrative, court, and missionary perspectives into a single secular narrative, one later to be taught in the schools, promulgated in the press, legitimated in academic works, and "confirmed" in diplomatic handbooks.	29
The essential elements of this vision stressed the homogeneity of the society, the power of the monarchy (including ethnic stratification), and the longevity of the kingdom. Rwanda was portrayed as a unitary and enduring society, completely consolidated internally and clearly demarcated from its neighbors. To be sure, both these points contradicted much of the earlier written and oral historiography of this region. Nonetheless, this historical image of a unitary "Rwanda" became as central as its claims to chronological longevity, and there was a connection between the two. In this context, "longevity" was a powerful factor in legitimizing "the state"—the claim to time depth implicitly projected the colonial assumptions of "the traditional monarchy," unchanged, into the distant past. From this, there developed an image of a highly centralized, rigidly stratified, and ancient state. From the 1930s, political issues came to focus not on regions but on royalty, social issues focused not on ecological distinctions but on ethnicity, and historical issues focused not on local initiatives but on external origins.	30
In this schema, the state was seen to have been established 500 years before.38 And while full of conflict (often within the royal line or between the Rwandan monarchy and other competing Tutsi dynasties), Nyiginya dynastic history proceeded undeterred. The only major change was in its inexorable expansion; even internal changes—coups d'état and dynastic shifts—were masked behind the ideology of unbroken continuity.39 Most important, the linkages with the broader population, as expressed through such institutions as clientship and army organization, were portrayed as static and unchanged over 500 years. The Nyiginya state institutions of the late nineteenth century, seen as the "traditional" institutions of Rwanda, were assumed to have been characteristic of the state since its founding at a particular moment in time: the state, it was argued, was created whole.	31
The proponents of this centrist view of history were responding to an external stimulus, the boundary dispute on the Kagera River. As is often the case, the external context came to define the contours of a new historical imagery. This revised internal historical understanding was produced to address external political issues; the facts had to conform not to local evidence but to regional agendas. Like many ideologies, this vision was based on a "truncated empiricism"; "facts" were cited without being placed in context, without consideration of alternative data, without internal critique, and without reference to the contentions of historical process: the events simply "happened." The manipulation of the ambiguous and controversial claims of Nyiginya court presence in areas just west of the Kagera River provides an example of this truncated empiricism, an approach that removed ambiguity and contradiction and denied the possibility of multiple histories. In fact, the Nyiginya lineages that became the political core to the royal dynasty had had influence, varying over time, in the disputed area. At issue was not whether they had a presence there, but what sort of presence, how that had changed over time, and what effect this had had on people's identities. In lived history, such concerns were central; in the context of larger political debates, however, such issues were easily elided.	32
To assert the vision of a consolidated Nyiginya monarchy, the writers advancing the Belgian view against British claims ignored the local sources (which were more ambiguous on such questions) in favor of Nyiginya royal accounts alone. Such historical accounts irrevocably associated with "Rwanda" any region attacked by armies of the Nyiginya court and any areas inhabited by people of "Rwandan culture": there is a powerful irredentist streak to Rwandan royal history.40 One example is in the treatment of the history of Gisaka, the area in question during the dispute with Britain after World War I. Dynastic Rwandan sources claim Gisaka as "Rwandan" from the time of Rujugira (1675–1741 in the "official" royal chronology) or from Ruganzu Bwimba (1312–1345).41 Yet Gisaka was still being fought over throughout the nineteenth century, and even after the "conquest" of the Gisaka royal family in the late nineteenth century, the population remained "rebellious" (that is, independent) well into the colonial period. Nonetheless, colonial histories of Gisaka ignore these "rebellions"; in this fashion, the presentation of local history was restructured to conform to Nyiginya royal accounts.42 One problem evident from this example relates to the treatment of regional diversity. Rather than being seen as an essential dimension to national history, regional histories were seen as opposed to national history.43 In the view of the central court, to focus on a region, and especially to privilege local initiatives, was a threat to the dominant paradigm. Thus "history" was presumed to adhere to nationality; in this case, "Rwandan history" could only belong to "Rwanda," and true historical actors could only be members of the royal court.	33
But these historiographical issues were not limited to Gisaka. Within the area now referred to as Rwanda, one can cite a score of historical polities of varying sizes and characteristics, from extensive kingdoms of the east (such as Gisaka, Ndorwa, Bugesera, Mutara) to the smaller autonomous political organizations of the west. The latter were based on one of two models. The "forest polities" included several thousand people, often a mobile population that engaged in trapping and gathering as well as agriculture and livestock. Indeed, several of these polities—Busozo, Bukunzi, Kingogo, and Bushiru—continued as independent units into the 1920s.44 A second general model differing from that of the Nyiginya dynasty was associated with the densely populated and sedentary social formations in the far north of what is today Rwanda. These were based on a form of segmentary kinship structures, without centralized political functions or territorial ritual focus.45 In sum, within the area now referred to as Rwanda, there were numerous regional variations: one can identify at least nine distinct cultural zones, with their own (politically autonomous) histories, before the eighteenth century, and often with important cultural connections reaching beyond current Rwandan state boundaries. For several, their autonomy continued well into the nineteenth century and, for some, even into colonial times.46 Though not often present in the historiography, therefore, regional differences are part of the ongoing story of Rwandan history.	34
Thus "conventional" history has privileged royal over local sources—or rather, it has obscured and silenced local sources. It also privileged written over oral accounts—or rather, privileged early written accounts (based largely on speculation) over later written accounts (based more often on local narratives). Consequently, it may be worth looking briefly at the development of this conventional history. Aside from the speculation of ancient Greeks, the first written accounts for this general region were by late nineteenth-century European travelers. John Hannington Speke was the earliest; he visited Karagwe and Buganda (east of Rwanda) in 1862. Speke's book included an entire chapter on his vision of the history of the region. A migrant himself, he saw the history as one of migration: "In Abyssinia a pastoral clan from the Asiatic side took the government of Abyssinia from its people and have ruled over them ever since. It may be presumed that there once existed a foreign but compact government in Abyssinia which becoming great and powerful sent out armies on all sides of it."47 Subsequently, one of these "armies" was "lost sight of in the interior of the continent and crossing the Nile close to its source, discovered the rich pasturage of Unyoro [today part of western Uganda] and founded the great kingdom of Kittara, where they lost their religion, forgot their language, extracted their lower incisors like the natives, changed their national name to Wahuma, and no longer remembered the names of Hubshi or Galla."48 Thus Speke's self-defined "theory of ethnology" is apparently a history of lost religion, lost language, lost incisors, and lost identity. (In other words, it lacks any empirical indication of Abyssinian origin.) And with little evidence to go on, Speke concludes: "So much for ethnological conjecture."49 Yet despite the absence of any empirical basis for the historical scenario he proposed, Speke's self-defined "conjecture" provided a convenient paradigm for others uncritically to follow.	35
Later accounts were associated with the establishment of colonial rule. One of the most influential of these was by Sir Harry Johnston. As a British colonial administrator, he had arrived in East Africa in the aftermath of a long civil war in Buganda, in which his administrative predecessor, Frederick Lugard, had played an instrumental role. In the wake of conflict, Johnston saw conquest as a key element to the history of the region, just as Speke, as a migrant, had seen migration as the key to history. But Johnston was also responsible for British negotiations with the elite to emerge from the Buganda civil wars—an elite that had twice overthrown the king. His aim was stability at minimal cost to Britain, and the Buganda Agreement of 1900 had three principal results: the recognition of Britain as the "protective power," the definition of the administrative elite to rule Buganda supervising the infant royal successor, and the introduction of freehold land tenure as the material foundation of the newly established Baganda elite. But in Johnston's eyes, legitimating the claims of this Protestant, pro-British elite also required legitimizing the history of this neo-monarchy. Consequently, Johnston was active in promoting a view of history that validated the new order; he was joined in this by members of the ruling elite, who published many works in Luganda as well as in English.50	36
Twenty years later, a process analogous to that in Buganda was to occur in Rwanda. The development of this aspect of Rwandan historiography, however, needed a third important step—beyond migration and conquest—in this complex history of histories. It was supplied through the work of Père Julien Gorju, who worked in his early years in western Uganda before being appointed bishop to Burundi. He was well placed to transpose Anglophone "ethnological theory" into Francophone intellectual paradigms, and, as he was working within an extremely tightly knit cohort of influential Catholic intellectuals, the regional interpretations he applied to Burundi became influential in Rwandan circles as well. Writing in the 1920s, Gorju worked at two levels: he drew on local traditions of western Uganda, but he fit them firmly within the broad regional framework sketched out by Speke and his successors in Uganda.51 Furthermore, Gorju wrote in French, he wrote with the authority of the church, and he wrote in a period when racial parameters were the principal features of history, culture, and identity, all seen as intertwined. So he was well situated to serve as the linchpin between Anglophone works in Uganda and an emergent Francophone corpus in Central Africa (including Rwanda), presenting a case for a broad history of this "interlacustrine" region of eastern Africa.52 In this history, migration, conquest, and "racial" stratification were all of a piece.53 Social stratification became the indicator of groups of different origins, in a hierarchy established by conquest. Culture internal to an ethnic group was seen as static because it was tied to race. Change, on the other hand, came from outside; cultural contact between two unchanging groups invariably meant conflict. In this region, social stratification was seen as the inevitable result of contact between two irreconcilable groups; it was also the proof of their incompatible racial differences. In other words, this was history in which all the important parameters were external: different racial origins, migration, and conquest were the operative factors that produced social stratification.	37
It is curious that, throughout this intense consolidation of "official history" in Rwanda, there remained an acknowledgment of local integrity. Nonetheless, since official history was based on a written tradition and local history on oral traditions, these were seen as parallel but separate phenomena: the former was "history," the latter was "legend." Only after World War II did these two approaches to history begin to be brought together. In the 1950s, acknowledging that royal ideology was insufficient in dealing with local problems, the Belgian administration assembled regional histories in a small booklet for use by its administrators.54 Interesting in its content, this compilation, entitled Historique et chronologie du Ruanda, is even more important for its admission that useful history is local history—and for acknowledging the diversity of local histories, even while it still tended to privilege local elites. Furthermore, a collection of land surveys conducted in the late 1940s and 1950s demonstrated the enormous institutional variability and individual mobility within this supposedly heavily administered state and stable society.55 So within some colonial circles, at least, local history was recognized; it just was not recognized as important to "History."	38
Even more telling were the official royal court histories, mostly popularized in the voluminous work of Abbé Alexis Kagame. He was an energetic scholar who was both Catholic priest and court intellectual, with each role enhanced by the other: Kagame of the 1940s represented simultaneously the indigenizing church and the modernizing monarchy. To outsiders, he served as the essential link between "tradition and modernity," between "custom and civilization." As ecclesiastic, academic, and court interpreter (and as a prolific writer), Kagame was the single most influential actor in consolidating court historiography—and in turning court history into "Rwandan History."56 But though intensely and uncritically focused on advancing the hegemony of the royal court, Kagame's works still illustrate the significance of local history. Many of his writings exude a mastery of local detail that would challenge the most ardent empiricist; one way he asserted his correctness was in the flood of undocumented "facts" that pervaded his writing. But despite his access to local detail, Kagame virtually neglected larger issues. While local data gave texture to central court history, in Kagame's vision it was nonetheless central court contact that gave significance to local data. So even the plethora of local data seemed to reinforce statist history. In short, during this formative period of Rwandan historiography (1930–1960), alternative histories were not absent; they were just ignored (or absorbed). Postcolonial historiographies unveiled such alternative accounts.	39
From the 1960s, a new methodological feature emerged. To move beyond colonial history in Africa, one needed to move beyond colonials: the history of Africans required understanding the visions of those affected by power as well as the intentions of the powerholders. How was power implemented, how was it received, and what were its effects? These were important issues—equally important to how power was conceived by the powerholders. To pursue these issues meant a return to the history of the local level. This shift differed from colonial hegemonic accounts in two ways. It introduced new topics and themes. More important, it introduced new conceptual understandings of old topics. Clientship, ethnicity, kingship, clanship, and royal ritual were all staples of colonial historiography. But for the most part, they had all been treated as fixed, unchanging, often primordial features of Rwandan society: they were presented as unhistoric in themselves, lacking internal transformation. The new historiography broke open these black boxes and saw these institutions as historical products embedded in multiple historical forces. They were no longer seen as just primordial features of the Rwandan social landscape: they were susceptible to history; they were testimony to change.57	40
By giving voice to an array of social groups with diverse perspectives on these issues, the new methodologies uncovered multiple versions of history. A historian's task was no longer that of selecting one "correct" version of history but of finding ways to account for plurality; empirical evidence was not just identifying the "correct" facts but accounting for and explaining multiple experiences. Institutions appeared different—they were different—when experienced from below. The critical factor was that the vision of social process moved away from the court to include the rural areas and away from elites to include peasants. But the real question here is not, "How are peasants treated in the historiography?" but, "How are peasants seen to participate in the historical debates?" There are two dimensions to this issue. First, in small-scale societies where face-to-face relations are important, history was always present in the public domain, whether in judging the boundaries between fields, tracing the genealogies of cattle clientship, negotiating marriage or divorce, succeeding to the role of head of household, or participating in a religious séance: the local community always included history in its public discourse.	41
On the other hand, local history that diverged from state paradigms was not easily retained in the wider sources. Even in conflicts among the elites at the central court, the hegemony of the state in Rwanda was powerful, and, in challenging the historical interpretations of the powerholders, the stakes were high. An example from court history illustrates this pattern in dramatic terms. It concerns Nkoronko, the son of one king, Gahindiro, and the brother of another, Rwogera. In the mid-nineteenth century, Nkoronko was killed because he was aware that Rwabugiri, Rwogera's successor, was not the legitimate heir to the throne. In fact, Nkoronko was killed on the orders of Rwabugiri. Since Rwabugiri was one of Rwanda's most autocratic warrior-kings, Nkoronko's execution in itself might not be surprising. But the drama went deeper, for in all likelihood Rwabugiri was also Nkoronko's biological son.58 So the power of historical interpretation led not only to eliminating one's rivals, even one of the most powerful of the land, but to patricide. Historical knowledge was a cause of concern in other instances as well. In one case, even ritual leaders of the court were killed for asserting the legal rules of succession; and in many other cases, individuals were either driven off their land, or bound, mutilated, or killed for refusing to accede to claims on their land by incoming nobles.59 In such contexts, power becomes an effective tool for molding historical thinking.	42
The preeminence of state authority over local knowledge is present even in recent times at such "mundane" levels as local production, where an agricultural officer often becomes a hegemonic authority. Several factors account for this. First, such agents often serve more as representatives of the state than as advocates for agriculturalists: their job is to enforce state directives in agricultural production. But they often have much less direct knowledge of their field (or others' fields) than those they are advising, for such officers are often chosen more on the basis of educational levels than agricultural experience. Furthermore, they are invariably males. Although women are the major rural producers, and often know more about local conditions than men (including the agricultural officer), relations between men and women are frequently strained in rural areas—male agricultural officers relate to women producers with difficulty. Finally, even for male cultivators, respect for an agricultural officer often marks this relationship as one of deference rather than dialogue. Consequently, an agricultural officer may arrive and ask to talk to the male head of household, yet the critical advice does not get to the actual (female) producers, nor does the critical knowledge of local conditions get to the agricultural officer. As a result, not only does "state agriculture" become a coercive field, but much local knowledge (local variations of crops, soils, pests, labor practices) is lost, in the name of standardizing and "rationalizing" agriculture.60 Even the field of agrarian history, therefore, is subject to the state and often neglected by the state; in an analogous fashion, the same is true for intellectual fields such as "history," also seen as the prerogative of the state and its authorities.	43
But there are ways to assert alternative views. One is through orality, often indirectly by allusion, occasionally by direct narrative. Another is through local newspapers or articles in local journals; during the 1950s in Rwanda, for example, a Rwanda-language newspaper, Kinyamateka, published alternative histories divergent from those of the dynastic oral traditions. They were extremely important in affecting political attitudes and mobilizing political action in the late 1950s, because of the inherent power of the articles themselves and because they articulated widespread, preexisting perceptions among rural Rwandans.61 For more recent periods in Rwanda, radio programs sometimes served this purpose of representing perspectives that differed from those of the powerful. For example, in the 1980s, one of the most popular programs on Radio Rwanda consisted of interviews with youth and "people in the hills." The opinions that aired on these shows were often critical of the state. Such outspoken views reflected more general discontent and contributed powerfully to the process of undermining the legitimacy of the government in the rural areas during the late 1980s. When handled sensitively and critically, these are the kinds of sources that are needed to transcend state hegemonies. They are the kind of "subaltern voices" that oral historiography has always dealt with—when it has moved beyond the oral narrative of the royal court. And it is the kind of knowledge that we must presume has always been present but to which historians have not always lent a ready ear.	44
Nonetheless, the question remains: do these alternative visions change our understanding of social process, or are they simply different ways of addressing the same topics? The answer is both. Where dynastic histories stressed the expansion of the state and the glories of the court, for example, local accounts stressed the brutal character of the internal state processes and the inequities of the system. Both perspectives addressed royal power, but one celebrated the power of the state while the other stressed the insecurity of the peasant. However, these divergent perspectives represent more than different constituencies being interested in different things and seeing different outcomes: these debates relate to divergent definitions of the state and represent differing understandings of state power. To trace power relations meant that one had to move beyond the conventional definition of state politics and address neglected themes of Rwandan rural history.	45
Postcolonial writings generally recognized the narrow character of earlier work and sought to move beyond court perspectives. But in inquiring into the world of those outside the privileged elites, writings on Rwandan rural areas diverged, taking one of two pathways. Anthropologists, political scientists, and historians often applied the new methods and the new interest in rural peoples to questions relating to colonial or precolonial issues. They thus focused on the past, reassessing earlier presentations.62 On the other hand, many "development" studies addressed postcolonial rural life, but these were technical studies and tended to extract rural areas from politics. In these perspectives the rural crisis was generated by rural realities—ignorance, population pressure, soil degradation: the state entered into these issues only to "solve" them. Thus these separate schools reproduced some of the same characteristics of the earlier historiography; they separated rural life from central politics, but they did so within a context marked by chronological segregation. Those who looked at peasants within a political matrix did so only with reference to the past, while those who looked at rural areas in the postcolonial period did so without accounting for power relationships.63 Those who looked at the state in rural areas saw the rural residents as objects or victims, as recipients of state power, not as subjects, or actors in their own right.64 One school removed peasants from power, displacing the focus of their work to earlier periods; another removed power from rural residents, by examining the technical aspects of the rural crisis without politics.65 All parties addressed issues of conventional politics and the state—but some saw the state as a causal factor of class formation, while others saw it as a palliative to existing social differentiation.	46
Relations of power, however, were more complex than could be measured through policy issues alone. To move beyond such debates over the peasantry and the state, therefore, different themes needed to be explored. These have usually been neglected in general accounts but were often present in an unintended fashion in the technical accounts referred to earlier. Though many could be illustrated, five themes in particular seem important to Rwandan social process but have not been well represented in the official historiography: the uneven representation of regions; the congruence of the state boundaries with the historical unit; labor; gender; and ecology. As with rural historiography, these have not always been absent, but they have been sublimated, by their absorption into state-defined visions of history. To be sure, they are all interrelated. But disentangling them from state parameters allows "hidden histories" to emerge. The examples below help illustrate the pattern by which these new intellectual configurations help conceptualize the diverse forms of historical agency, previously veiled within the elegant trappings of central court discourse.	47
First, as noted earlier, hegemonic historiographies emphasized central court narratives at the expense of local narratives—despite the "subliminal" presence of such data in several sources. After Rwanda's independence (in 1962), some sought to redress the lack of local voices through the use of case studies; if peasants needed representation, their specific circumstances and goals needed clear articulation. Nonetheless, little was done to draw out the common themes or comparative processes among such local narratives. A second feature, related to, but distinct from, the first theme, concerns the insular nature of Rwandan history. In conventional historical inquiry, official state contours marked the only arena of historical action; political boundaries had come to define the historical unit. But such conventions largely neglected the flow of resources, commodities, people, ideas, ritual concepts, religious practices, friendship networks, and identities across the "inviolate" boundaries of the state—boundaries sometimes established, as we have seen, through the combined efforts of the Belgians, the Catholic Church, and the court. In fact, members of most of the nine subcultures mentioned earlier often maintained broad ties outside the colonial boundaries that were often as intense as those within the current state parameters. Moreover, these ties—in religion, marriage, commerce, and productive activities—helped define the regional characteristics of these subcultures, characteristics often mocked by or written out of Rwandan history.66 Nonetheless, if official Rwandan historiography—or national hagiography—has arbitrarily cut Rwanda off from neighboring areas and treated "Rwanda" as a historical island, it behooves historians to reexamine the record and to explore these broader linkages.67 To do so would represent more accurately the lived experience of Rwandan non-elites.	48
In addition to regional and cross-border dimensions to the history of the people of Rwanda, the remaining three related themes have been neglected as well. One is the issue of labor—curiously so, since one of the attractions of Rwanda to the colonial strategists was as a convenient labor pool for the mines in Katanga, the European plantations in Kivu, and the production of cash crops from Rwanda. But what was justification for Belgian acquisition of colonial territory became invisible in Belgian representation of their rule. Rather than publicizing labor policy as part of the civilizing mission, colonial representation shrouded such policies with the veil of "traditional obligation."68 Official historiographies—premised on the presence of an atavistic "feudal state"—emphasized clientship, especially ubuhake cattle clientship, as the essential mode of mobilizing labor. Rather than a form of forced labor, ubuhake was presented as based on "the premise of inequality," as a voluntary, benign institution—in which a client freely offered his labor in return for the usufruct of a cow; colonial apologists, reflecting royal court ideology, claimed that this incorporated the poor and bound Rwanda into a single social unit, transcending ethnic differences.69 To structural-functionalist anthropologists, the beauty of this interpretation was that, as portrayed in this form, clientship appeared to resolve a fundamental contradiction in colonial thinking: that Rwanda was seen, on the one hand, as a single enduring, coherent society and, on the other, as a "caste" society, characterized by permanent, unchanging, cultural and racial differences. This portrayal reinforced the distinctions between ethnic categories while binding them together. In other words, the power of clientship preserved simultaneously two competing concepts of "tribe"—one applied to the "culture" as a whole, the other applied to an "ethnic" group alone. Within the anthropological discourse on ubuhake, these two coexisted.	49
Research since Rwanda's independence, however, has clearly established the ideological nature of such a vision.70 These studies show that clientship was frequently driven by considerations of power: the "protection" received by the weak was often protection against the outright expropriation of their existing cattle—and sometimes of their personal safety as well. The "voluntary" aspect of this relationship, therefore, often consisted of a choice between losing one's life or losing control over one's labor. Furthermore, clientship evolved among many complex forms and was contextually defined: practices changed according to changing power relationships. Whereas formerly, cattle clientship was seen as the social glue connecting different classes and ethnic groups, the new data showed clientship to be a much more diverse phenomenon, with cattle clientship less important than land clientship in many areas.	50
Several factors emerged from postcolonial research that challenged the orthodox view of ubuhake, the iconic form of cattle clientship.71 Ubuhake clientship was infrequent in the precolonial period, not widespread—and certainly not universal—as the conventional understanding asserted. For example, at the end of the nineteenth century, at the height of precolonial state institutions, only 8 percent of family heads were involved in ubuhake ties at any time in their lives, according to a careful study carried out in south-central Rwanda, near the cultural heartland of the kingdom. Far from being universal, ubuhake was of very limited extent. Furthermore, according to the precepts of "modernization theory," during colonial rule, as individuals are released from the responsibilities placed on them by archaic institutions, "feudal" institutions such as clientship would be expected to diminish in their extent. Yet the data suggest something else: over the first generation of colonial rule, ubuhake ties in this sample actually increased, almost doubling the percentage of family heads involved; therefore, clientship appears to have been extended and reinforced by colonial power. Finally, rather than serving to link Hutu and Tutsi in "voluntary" association, and thus binding ethnic groups, clientship ties in this area were primarily contracted among Tutsi with political positions: in empirical terms, ubuhake appears more important as an alliance among political elites than as a mechanism that brought all people into the political system. In light of these data, the conventional view of cattle clientship as a pervasive and voluntary form of "social glue" is untenable. It presumes as universal within Rwandan society an acceptance of the ideology of the central court—the "premise of inequality," whereby clients willingly accepted subservience. We now know, instead, that such subservience commonly resulted from the routine use of political power on the part of the elites.	51
The ideology of clientship was so powerful that other forms of labor control were often not explored, even while recognized.72 Yet labor relations extended far beyond clientship forms; they were tied to state power, to land access, to mobility factors, to local power relations, to the politics of kinship, and to power relations within the residential unit: drought, commercial opportunities, marriage relations, and health all had an effect on labor. So did religion. Religious units sometimes had shared work patterns, and religious celebration was often seen as withdrawal of labor. One such incident occurred with the emergence in the mid-1920s of a cult for young women on the shores of Lake Mohazi. Although described as a process of celebrating the spirits of the lake, and as a means of assuring the later fertility of the women involved, it was opposed by many of the chiefs, since it had the effect of withdrawing labor from the chiefs' fields.73 Labor was gendered in other ways as well. Men were drawn away on forced labor for porterage (before the 1930s), for work on communal fields, for required cultivation, and by recruitment for mines and plantations. These various forms of male labor requisition of course meant added burdens on women for domestic production.74	52
Forced labor in colonial Rwanda. Church Missionary Society. Copyright Africa-Museum Tervuren (Belgium).
Yet Rwandan historiography has virtually ignored gender issues—a fourth neglected theme of the literature.75 An illustrative case is coffee production, the preeminent colonial crop that by the end of the 1950s represented more than 70 percent of the export earnings for Ruanda-Urundi.76 Of the work that went into coffee production, 70 percent fell to women, by one postcolonial estimate.77 But labor was gendered by exclusion as well as by inclusive demands: many Rwandan men left to work in the Congo and Uganda, sometimes for long periods of time; in their absence, all the burdens of agricultural work, including required crop cultivation and sometimes even corvée labor, fell to the wives left behind.78 Yet domestic labor relations are one of the most "naturalized" of social functions; because these gender relations are portrayed as "normal," the emphasis on state power obscures the political nature of such contradictory power contexts and economic fields of force.79 "Official" histories of Rwanda have left labor relations to languish behind the veils of clientship, "development," and the domestic domain, and historians have failed to bring gender into labor issues. Furthermore, just as labor is tied to gender issues, it is also tied to the question of borders and the concept of "insular Rwanda." Large numbers of Rwandans were involved in quasi-capitalist labor relations, over long periods of time, in Uganda, Tanganyika, and the Congo.80 But this occurred outside the boundaries of the Rwandan state; thus, though wage labor was in many cases a formative experience for Rwandan men, it tended to be excluded from "Rwandan" history because it was experienced outside the territorial domain. Once again, Rwandan history was misleadingly limited to the state boundaries.	53
Finally, the theme of ecology is as neglected as it is important in Rwandan history, for many of the same reasons noted above: the belief that Rwanda was a "homogeneous country," that regions were not important, that state initiatives were the preeminent feature of Rwandan history, that the history of the country was embedded in the history of the royal court, and that colonial and postcolonial "development" progressively brought under control the capricious natural forces of the peasant's world. Yet a great deal of colonial documentation was given over to such capricious forces, and, contrary to the official discourse, the major bulwark against ecological disaster was not the state—which over time extracted resources and expropriated pasture—but the knowledge and abilities of the peasants themselves. For the people of Rwanda, one defense against natural disasters was mobility—their access to different internal ecologies and their willingness to migrate across international boundaries, despite the vigorous attempts of the state to control such movement.81 Official histories from the colonial period celebrated the required cultivation of "famine crops" (such as manioc), and indeed, these were important.82 However, these same histories also systematically effaced (or misinterpreted) rural initiatives addressing ecological crises, since such initiatives (flight, cultivating marshes, killing livestock) were often seen as a threat to administrative control—especially by local (Rwandan) administrative elites; on the other hand, required cultivation, especially forced cultivation, could directly benefit these elites, who sometimes redirected such labor to their own benefit.	54
Understanding peasant ecology demands much more than describing the suffering experienced under dramatic crises or detailing the application of exogenous technology to rural production. Yet these were the major ways it entered colonial discourse. In Rwanda as elsewhere in Africa, colonial documentation privileged technical subjects such as climatological conditions, the suitability of different crops, and the carrying capacity of the land.83 The dominant view of rural areas assumed that poverty derived from problems of climate, soils, illness, and "la mentalité paysanne," with rural producers characterized as ignorant or indolent. Therefore, solutions to poverty were also posed as technical—terracing, culling cattle, draining swamps for cultivation, or requiring the cultivation of "famine crops"—with little discussion of confrontation, conflict, or struggle in the countryside; politics and power were almost entirely absent.84	55
To be sure, under some circumstances, rural producers recognized the benefits of communal labor.85 Nonetheless, resistance formed around the forced implementation of communal labor by chiefs or agents who often bent these colonial demands to their own advantage. Moreover, colonial demands often diverted labor from consumption crop production. Issues relating to struggles over the "commons" were also important.86 These included access to pasture, deciding whose cattle were to be culled, rights to the resources of communal forests, and claims to farmland formed from swamps drained with communal labor. The peasants saw these as political impositions, not technical innovations, and many of these requirements worked against rural production patterns; what the state saw as simply "technical" issues, therefore, rural producers saw as eminently political. Technical rural studies often recognized many of these issues; what they failed to account for were the peasants who silently lined the roads in the late 1950s at the time of UN Trusteeship Council Visiting Missions, holding placards that read "Our Fields, If You Please!" (NOS CHAMPS S.V.P.!). Theirs were the political voices suppressed by the "anti-politics machine" of statistical tables.87 And their placards—indeed, the entire rural effervescence associated with the "Rwandan Revolution"—prompted historians to revisit the available documentation, rereading it in a way that challenged the power of centrist history.	56
Awaiting a UN delegation in the late 1950s, Rwandans protest in front of terraced fields. The placard on the left reads, "Our Fields, If You Please!" (NOS CHAMPS, S.V.P.!). In the center, the placard states, "Down with Tutsi colonialism. Democracy first, independence will follow. No immediate independence." (À BAS LE COLONIALISME TUTSI. DÉMOCRATIE D'ABORD, INDÉPENDANCE VIENDRA. À BAS L'INDÉPENDANCE IMMÉDIATE.) The placard to the right says, "Long live Rwanda, long live Belgium, long live the ONU." (VIVE LE RWANDA! VIVE LA BELGIQUE! VEVE [sic] L'ONU!) Copyright Africa-Museum Tervuren (Belgium).
Many of the postcolonial sources mentioned earlier were valuable not only for their critique of colonial historiography but also for opening possibilities for a new historiography—one that is inclusive of rural dynamics as well as state initiatives.88 The lacunae of the historiography stemmed from the neglect of voices and sources that did not conform to the hegemonic vision. Such sources did indeed exist (as we discuss below), but the possibilities of exploring rural agency were not always seized upon. Therefore, the challenge to historians today is to draw from these sources an understanding of rural conditions, reading against the grain of statist assumptions and assessing the interaction of local initiatives with state policy. But to date, that has seldom been attempted. Because colonial documentation on rural areas was often of a technical, scientific nature, such sources have been almost entirely marginalized when drawn on by historians; at best, only bits and pieces have been utilized, and even then, this has been done in such a way as not to threaten the hegemonic contours of the dominant state paradigms. Consequently, in portraying rural issues, centrist assumptions were retained in a manner analogous to the way Alexis Kagame had mediated between local histories and central court hegemony in the context of political history.	57
However, recent studies of rural Rwanda have given greater prominence to local agency and to the particularities of rural communities.89 Transitional in this regard was Philippe Leurquin's inquiry into rural living standards in Rwanda and Burundi during the mid-1950s.90 Nonetheless, while this study broke through the separation of court and countryside, Leurquin's approach still preserved the characteristics of colonial discourse; as a "scientist," he avoided direct discussion of politics and rural class relations. Yet the empirical content of his study raised important questions and provided valuable data on rural poverty. For example, establishing that the principal variables in wealth were urban/rural, this work challenged colonial assumptions of ethnicity: Leurquin's data showed no significant differences between rural Hutu and Tutsi in income and access to food.91 Tutsi households had slightly more access to cattle than Hutu, in part because of more frequent Tutsi ties to an urban population, and thus greater off-farm resources. But other differences were minor. No longer, concluded Leurquin, could one assume that all Tutsi were wealthy cattleowners, and all Hutu poor agriculturalists. To be sure, writing in the 1950s, Leurquin still operated within a mental framework of viewing Hutu and Tutsi as distinct but internally homogeneous groups—a thought pattern itself a product of colonial ideology. But this colonialist assumption in his work only lent greater credibility to his statistical findings, which cut against the grain of such thought paradigms; Leurquin's data disproved the centrist assumptions that royal aristocrats represented all Tutsi.	58
Leurquin's data on coffee production also made it possible to link rural economics to larger historiographical issues. Noting the huge expansion of coffee cultivation (some 62 million coffee trees had been planted in Ruanda-Urundi by the end of 1957), Leurquin argued that coffee production brought favorable returns to peasants. But even while praising the role of the state as an "entrepreneur," he deplored the high social cost of the way coffee cultivation had been expanded—by force.92 Leurquin's study did not directly relate either to the history of agricultural production or to the rural politics of the 1950s, and thus he did not detail the constraints imposed along with coffee. But he raised the issue. Coffee growers were compelled to tend their trees according to prescribed rules and procedures—pruning, spraying, and mulching the trees at fixed intervals. More important, interplanting food crops among coffee trees was prohibited. By documenting the nature of state intervention in rural areas, Leurquin's writing unintentionally tied the technical aspect of state involvement to peasant political protest, which emerged in the late 1950s. From this technical document, there emerged an interpretation of political causality that transcended purely ethnic explanations.	59
Later work explored more directly the connection between forced production and rural politicization. On the basis of research carried out in the late 1970s, Learthen Dorsey argued that the colonial state penetrated into rural Rwandan society primarily through its policies of forced cultivation, including those for cash crops (predominantly coffee).93 Rather than responding to the crisis of the rural areas, this argument suggests, the state was in fact part of this crisis. And although Dorsey omits agency and does not discuss peasant political action in depth, his work lays the foundation for understanding the grievances that led to such action. In its timing and significance, state penetration differed by region, and the effects of these policies varied by gender.94 Nonetheless, such an analysis places the state directly in the production process.	60
The data derived from Leurquin and Dorsey suggest that ethnicity was not the primary causal factor of rural class differences. Data from other regional studies in Rwanda also subvert these earlier assumptions and reverse the cause-effect relationships. They indicate that, rather than resulting from "ethnic" differences, rural class differentiation preceded and served as a catalyst for ethnic identity. Economic insecurities, therefore, contributed to political mobilization along ethnic lines. Incorporating rural experiences and "bringing the peasants back in" to thinking on Rwandan history leads historians to read more broadly and thus to transcend the flawed conventions of statist historiography. In short, it helps provide more accurate historical explanation.	61
The 1950s were a turbulent time in Rwanda, and this turbulence forms another significant gap in the historiography—even more so as many post-genocide studies seemed to revert to colonial assumptions that the principal, indeed the only, factor of change was colonial power. They therefore tend to attribute the revolution of 1959–1961 exclusively to Belgian conniving. To be sure, Belgian policies were important. But they did not act in a vacuum; they were often responses to internal peasant grievances. By ignoring the rural experiences of late colonial rule and by effacing peasant political initiatives, such histories align themselves with a long pattern of earlier historiography. In doing so, they distort the historical record, for the late 1950s was a time of political ferment. Hutu activists and Tutsi progressives—those who sought alliances across ethnic lines—called for changes in the exploitative practices of the central court authorities. They advocated measures to address problems of poverty, inequality, insecure access to land, inadequate opportunities for education, and issues facing youth.95 While Leurquin and Dorsey do not address these conflicts and debates directly, their studies nonetheless provide material that helps explain why the rural areas witnessed so much unrest during this period. Works like these provided an understanding of neglected rural realities; recent analyses have built on these insights, exploring the synergies linking material culture, political economy, and postcolonial crises.	62
For studies focusing on the postcolonial period, scholarly discourses on rural Rwandans shifted from conventional historiographical concerns to "development studies." The former had often privileged political features at the expense of economic features; the latter often privileged economic factors to the exclusion of political aspects. These temporal differences mirrored the sectoral distinctions of earlier writings. There, rural studies omitted politics, and political studies omitted rural actors, privileging agency associated with the central court. But exploring the roots of Rwanda's postcolonial crisis forces an examination of rural politics and a reconsideration of colonial and precolonial relationships to include local agency as well as central court actors.	63
For these purposes, technical studies of the type Leurquin produced, though valuable, were timid. They were poor substitutes for ecological and agrarian understanding. More recent work has set out a new model, relating ecology to political constraints, conscious human activities, and the physical properties of a natural world itself in flux.96 Jennifer Olson's study of ecological processes in southern Rwanda provides an example.97 Drawing on a "regional political ecology" framework, her work describes the intricacies of the agricultural system not in terms of abstractions, models, or norms but through specific economic activities and political relationships representative of many regions of Rwanda. She summarizes agrarian processes in terms of energy transfer between three fields of activity: uncropped fields served as pasturage, manure from cattle served as fertilizer in the cropped fields closer to the homestead, and the remains of food preparation went to the banana groves immediately surrounding the residence. In short, cattle and cultigens went together, as nutrients and energy were moved progressively from outlying areas to the homestead.	64
A mixed economy was essential to rural production.98 But these features were susceptible to a variety of political forces. Policies of the European colonial state affected both labor and land issues. In many cases, Rwandan authorities, delegated by the central court but acting under colonial protection, influenced access to land, ownership of cattle, and domestic decisions over labor. And the policies of the postcolonial state continued to make farming families dependent on off-farm income in fragile economic circumstances.99	65
Colonial policies treated ecological issues as a series of discrete problems extracted from the larger agrarian context, and thus failed to address larger ecological relationships. When "overgrazing" was identified as a problem, draconian measures were taken to reduce cattle. When "idle" land was put into production or claimed by the state, this reduced other forms of land use: fallow, pasture, and sources of cut grass; when colonial officials assumed that rural people were underemployed, labor was drawn off for colonial projects. However, such policies were often carried out without regard to local equity. Each of these policies (culling cattle, reducing green manure, and extracting labor) threatened rural capacities to continue the energy transfers on which agrarian production was dependent.	66
Of these elements, labor demands were crucial. As we have indicated, labor was drawn off at a high rate, for the colonial state system in Rwanda was based on coerced labor: for porterage, road construction, plantation or mine recruitment, "communal" (often chiefs') fields, "traditional" work for chiefly authorities, and required cultivation. One spectacular demand for labor was terracing—by 1960, the colonial government estimated that 439,000 kilometers of erosion ditches (terracing) had been constructed, enough terracing in this tiny country to circle the globe eleven times; Leurquin estimated that, under the optimistic assumption that each person dug 10 meters of ditch a day, work associated with constructing and maintaining these ditches accounted for over 55 million work days. Other colonial policies—each one justified on its own technical terms—also intervened in the agricultural system. Burning fields was outlawed, complicating the maintenance of pasturage. Forced crops meant that cultivation spread to steeper slopes, aggravating erosion. Reforestation entailed planting eucalyptus trees, which, though quick to mature, were deleterious to the productive system: they draw off water, and the toxicity of their roots makes them unsuitable for intercropping. Productivity declined for many reasons, each "inconsequential" in itself—as cattle became scarce, labor was conscripted, fallow land decreased, new lands opened on steeper slopes, and mobility was reduced so that people no longer had access to valley or forest resources.	67
For the Gikongoro region, the contradictions of such policies became apparent during the famine of 1989–1990. A government report noted that, of those affected, 40 percent were in female-headed households (twice their proportion in the local population), 96 percent had farms of less than half a hectare, and 74 percent had less than one-quarter hectare.100 Therefore, the report concluded, the famine was beyond the control of the government. To the peasants, however, increasing land inequalities and rising food costs were indeed a result of government policies. If poverty increased vulnerability to famine, then the government was deeply implicated. In peasants' eyes, the state was denying its own responsibility in creating the preconditions for this disaster. This was another example of the contrast between statist perspectives and the views of those affected by state policy.	68
Not until two decades after independence did academic researchers (Rwandan or expatriate) give much attention to peasants in postcolonial Rwanda.101 To be sure, before the 1980s, technical reports by government agencies and donor-funded development projects provided information on the rural sector. But these tended to be focused on technical concerns rather than broader production processes within the rural political economy.102 During the second half of the 1980s, however, Rwandan academics and intellectuals began to write openly about the difficult conditions faced by rural cultivators, the exploitative practices of merchants who purchased coffee and food from peasants, and the inadequacy of government programs to provide alternative markets for peasant production. Many such analyses were published in the Rwandan journal Dialogue, and these articles stimulated a lively debate among Rwandan intellectuals, sometimes joined by expatriates with long experience in the country.103 By writing on such topics, Rwandan intellectuals identified the void in the earlier literature. They thus reminded others that Rwandan historiography is not some isolated domain. Instead, Rwanda's history over the 1990s reminds us that the political nature of writing history is intimately intertwined with the politics of the present.	69
However, behind the historiographical debate loomed economic crisis. Even as the government cut back on personnel and reduced social spending during the 1980s, the politically well-connected minority continued to build villas and lived a lifestyle modeled on that of the international jet set. Critics of such conspicuous consumption in the face of widespread poverty directed attention to a range of important issues. Prominent among them were the legal and cultural constraints to women's economic advancement, the excesses perpetrated by "moniteurs agricoles," and the general lack of peasant voices in the formulation and implementation of rural policies. Externally funded projects, which had proliferated during the Second Republic, also came under scrutiny. Rather than empowering rural producers, such programs, it was argued, benefited a small stratum of educated intermediaries, employees of the projects.104	70
Implicitly (but sometimes openly), these writings criticized the ideology of "planned liberalism," the approach to economic development adopted from the mid-1970s by President Juvénal Habyarimana's regime. This approach assumed that peasant living standards could be improved even while encouraging an unrestrained private sector and permitting rapid class differentiation. Yet the major sources of income for this privileged sector came from among peasant communities. Control over land, labor, transport, and local commerce—all tied to political power constructs—were major sources of "wealth-generation" for the acquiring classes. In a 1985 article, Jean Rumiya, at the time a professor of history at the National University of Rwanda, analyzed the contradictory effects of planned liberalism. He acknowledged that the government could claim some positive results: Kigali, the capital, had grown significantly; paved roads had improved the transport system; the number of automobiles had increased; and the country was able to procure sufficient amounts of gasoline, construction materials, and food products. But few benefits from this growth reached the rural masses. And changes in values were creating a crisis in rural society. In the past, a man was considered wealthy if able to farm two hectares of land, own a bicycle and a radio, build a durable house, and provide education for his children. But in the mid-1980s,	71
with the appearance of fortunes acquired in the corridors of the administration and through trade . . . the symbols of wealth have changed: villa, automobile, bank account, salary, etc. . . . The prototype of the wealthy man is no longer the well-off peasant, respected by his peers, or the government official who returns to his hill, but the city dweller, preferably in Kigali, whose living standard attains an international level in the areas of leisure, transport, or lodging. This paradise has a strong attraction for youth . . . But in this type of competition, there are only a few winners.105
By arguing that government policies in postcolonial Rwanda had intensified rather than reduced social inequalities, Rumiya implicitly called into question an important pillar of the government's claim to legitimacy: such glaring inequalities belied the egalitarian ideals of the Rwandan Revolution of 1959–1962. A series of academic studies published in the second half of the 1980s proceeded to make this critique explicit. Writing in 1985 with reference to the First Republic, for example, Filip Reyntjens remarked: "Contrary to the egalitarian principles of the revolutionaries, the disparity between the revenues of politicians and peasants [after independence] was, and still is today, very great." In his view, the disparity was probably no less than it had been under the prerevolutionary government. Nonetheless, such a situation "for a regime that came to power on the basis of support from the rural areas [was] serious."106	72
Regional inequalities in the distribution of government development funds served as another focus of these critiques. Under Habyarimana's regime, Gitarama and Butare Prefectures in the center of the country (the political centers of the government overthrown by Habyarimana) saw the least government support; with about 20 percent of the country's population, they received about 1 percent of government funding. Furthermore, local administrators and the commercial class structures hindered peasant voices and the capacity of rural people to organize. André Guichaoua's research noted these trends and showed the extent to which strategic areas of economic and social life in the countryside (commercial outlets, agricultural supplies, consumer goods, credit, transport) were dominated by people outside the rural milieu (traders, absentee landlords, civil servants, and military personnel). The concentration of power and the polarization of wealth benefited only the elites in the capital.107	73
In a well-grounded economic analysis, Fernand Bézy issued a particularly scathing indictment. Citing the results of a government survey published in the mid-1980s, which estimated annual per capita expenditures for consumption among Rwanda's peasants at less than $150 (U.S.), he condemned the "pauperization of peasants."108 Writing in a prophetic vein in 1990, he argued that the problems were not just economic but political, with a risk of serious social conflict if current conditions continued. New approaches were needed, he concluded, to ensure food security for rural dwellers, protection from merchant exploitation, and the establishment of many small, labor-intensive industries in different regions of the country to provide employment and produce basic essentials. But achieving such a program would require nothing less than a reordering of current political structures, a "transformation of the society."109 His advice went unheeded.	74
Critiques by Rwandan intellectuals and studies such as those of Guichaoua and Bézy delineated important facets of Rwanda's political economy in the 1980s. Combining economic analysis with sensitivity to power and politics, they showed how government policies, highly placed politicians, and others tied to the powerful were implicated in the reproduction of inequalities that permeated Rwandan society. But they were generalizing. A fuller portrait required more specific analysis of culture and consciousness, and a narrower focus on particular localities, to bring peasants back in to an understanding of the political and social processes of the state. A 1996 study by Danielle de Lame did just this, and demonstrated the value of such an approach. Focusing on a rural community in western Rwanda at the end of the 1980s, de Lame broadened the earlier concerns of Leurquin and others to include politics, gender, class relations, culture, conflict, and consciousness; she thus examined the local ramifications of the rural immiseration noted by analysts such as Bézy. By situating material process within the context of cultural values, she showed that dignity for rural dwellers is more than just a question of economic factors. Through the study of a single locale, her work revealed the strategies by which rural people tried to sustain culturally prescribed obligations, produce adequate food, and obtain an income sufficient to assure their children's well-being and their own participation in the community.110 Sometimes, these requirements are very simple: salt, soap, decent clothing, and a new hoe; and even these were beyond the reach of some. But de Lame did not portray rural dwellers in Rwanda as an undifferentiated mass; delineating areas of both commonality and cleavage, she showed that rural immiseration did not affect all people equally. Access to products such as banana beer, for example, was unevenly distributed. Yet such access determined whether a peasant family could participate in customary circuits of exchange, marking processes of social inclusion and exclusion.111	75
While we cannot do justice here to the complexity of de Lame's comprehensive study, the conclusions are important. Despite the abundant literature on Rwanda, there are very few carefully grounded empirical studies that focus on the defined local community. De Lame's study does this. But it also does more, by showing how local dynamics were linked to larger issues. Society was in crisis, she suggested, because of the multiple tensions structuring rural social relationships, threatening cultural reproduction for many households.112 Children sought income outside the home rather than working in parents' fields; young men and women found it increasingly difficult to acquire the requisite resources needed to marry; and money, education, or a salaried job (the paths to social mobility) were beyond the reach of most. These conditions had significant political ramifications. From 1988, de Lame observed many signs of the population distancing itself from the government—refusing to attend meetings called by the bourgmèstre, not showing up for the hated communal labor, cutting down coffee trees. Many people openly expressed their class consciousness, using the word umukungu (rich person), for example, as a scathingly pejorative label for people they resented.113	76
Other recent work also illustrates the importance of addressing rural issues. Michelle Wagner analyzed the social preconditions to the political struggles that ravaged a community in southern Rwanda.114 Catherine André and Jean-Philippe Platteau have shown the complex interaction of ecological factors with political violence in a community in northwestern Rwanda.115 Timothy Longman delineated the diverse roles of churches and church leadership within the local politics leading to the genocide.116 Focusing on rural producers, these studies all show the disastrous consequences of sharp changes in the economy of Rwanda at the end of the 1980s: the precipitous decline of the international commodity price for coffee, the famine that wracked central and southern Rwanda, the abrupt withdrawal of social services by the state, and the rapid erosion of legitimacy of the regime in power. Such were the rural conditions that underlay the genocide;117 they were neglected by historians, for the most part. But the genocide forces us to take notice; rural patterns of change provide an entry point for rethinking Rwandan history and Rwandan historiographies as well.	77
This survey has had three goals. One was to provide familiarity with the conventional historiography of Rwanda—and with some of its lacunae and limitations. Another was to introduce work on rural areas and peasant realities, focusing on issues of inequality, agricultural productivity, and the crisis of reproduction. And a third was to suggest openings for future investigation, by which the literature on rural areas could be more effectively accommodated within historical paradigms. In the early years of the twentieth century, rural areas had been important to historians of Rwanda, and history used to be important to understanding rural areas. But today, these two are assumed to operate in different intellectual universes; peasants are barely visible in historians' scenarios, and history is barely audible within the discourses on rural society. We propose a reunion of these partners, too long separated: providing a rural dimension to conventional history and a historical dimension to rural studies, bringing politics into the understanding of peasants. "Social history" needs social actors as well as history, and it needs history as well as social actors.	78
In a country where 90 percent and more of the population is composed of rural producers, a focus on the peasantry brings people back into what has developed over the past sixty years as a history of institutions and ethnic categories. An analysis based on social agency would conform more closely to the empirical record. "Bringing the peasants back in" opens awareness to a range of new themes. It allows for cultural particularities and individual experience within Rwandan history. It encourages dialogue with other historiographies. It reinserts a critical dimension to understanding the nature of conflict—beyond simply "age-old" ethnic hostility and "tribal warfare."	79
However, considering broader social themes does not mean avoiding the topics that have served as the core of Rwandan historiography until now. If past history has focused exclusively on elites, the response is not to focus exclusively on peasants. Instead, one needs to incorporate peasants—to break down the separation of peasants from elites, not to reinforce such dichotomies. We are not arguing for the study of peasants in isolation but for the study of a Rwandan history that includes peasants. Bringing the peasants back in is not additive to Rwandan history, therefore, but transformative of Rwandan history. It forces us to revisit the conventional themes with a new vision.	80
David Newbury taught for several years at the Institut Supérieur Pédagoque in Bukavu, Zaire, and currently teaches history at the University of North Carolina, Chapel Hill. He has published Kings and Clans: Ijwi Island and the Lake Kivu Rift, 1780–1840 (1991). He also edited African Historiographies: What History for Which Africa? (1986), with Bogumil Jewsiewicki, and Paths towards the Past (1994), with Robert W. Harms, Joseph C. Miller, and Michelle Wagner. His publications include many articles on the politics of social identity and state power in precolonial and colonial Central Africa and on the historical roots of the recent crises there.      Catharine Newbury teaches in the Political Science Department and the Department of African and Afro-American Studies at the University of North Carolina, Chapel Hill. She is president-elect of the African Studies Association. Her earlier research focused on the historical development of ethnicity in Rwanda and on women and agrarian change in Central Africa. She is currently studying the roots of the politics of violence in this region. The author of The Cohesion of Oppression: Clientship and Ethnicity in Rwanda, 1860–1960 (1988), a study of the interplay between state building and the growth of ethnic cleavages during colonial rule, she has also written for journals such as Cahiers d'études africaines, Africa, Comparative Politics, the Canadian Journal of African Studies, the African Studies Review, and Africa Today. Forthcoming is an interdisciplinary collection of papers she co-edited with Mamadou Diouf and Pearl T. Robinson, Transitions in Africa: Creating Political Space.

---

LOCKSS system has permission to collect, preserve, and serve this Archival Unit

Content in the History Cooperative database is intended for personal, noncommercial use only. You may not reproduce, publish, distribute, transmit, participate in the transfer or sale of, modify, create derivative works from, display, or in any way exploit the History Cooperative database in whole or in part without the written permission of the copyright holder.

 

Notes

We wish to thank our colleagues in the 1997–1998 Program in Agrarian Studies at Yale University and in particular its director, James C. Scott, for their invaluable comments on an earlier version of this article.

¹ Andrew J. Shyrock, "Popular Genealogical Nationalism: History Writing and Identity among the Balqa Tribes of Jordan," Comparative Studies in Society and History. 37 (1995): 334.

² Quoted in Jan Vansina, Children of Woot: A History of the Kuba Peoples (Madison, Wis., 1978), 19.

³ Rwanda provides one of the most extensive bibliographies of any region of Africa of comparable size; the authoritative bibliography of published written works (to 1980) includes over 5,500 entries in the social sciences alone: Marcel d'Hertefelt and Danielle de Lame, Société, culture, et histoire du Rwanda: Encyclopédie bibliographique, 2 vols. (Tervuren, 1987).

⁴ Preferring to concentrate on more foundational sources, we have omitted from explicit consideration several recent historical surveys: Roger Heremans, Introduction à l'histoire du Rwanda (Kigali, 1988); Jean Rumiya, Le Rwanda sous le régime du mandat Belge (Paris, 1992); Ferdinand Nahimana, Le Rwanda: Emergence d'un état (Paris, 1993); Jean-Népomucène Nkurikiyimfura, Le gros bétail et la société rwandaise: Evolution historique des XII^e–XIV^e siècles à 1958 (Paris, 1994); Antoine Nyagahene, Histoire et peuplement: Ethnicités, clans, et lignages dans le Rwanda ancien et contemporain (Paris, 1997); Bernard Lugan, Histoire du Rwanda (Paris, 1997); C. M. Overdulve, Rwanda: Un peuple avec une histoire (Paris, 1997).

⁵ For a comprehensive overview of studies on peasants in Africa, see Allen Isaacman, "Peasants and Rural Protest in Africa," African Studies Review 33 (1990): 1–120 (also reproduced in Cooper et al. cited below). Among many other works on African peasantries, see Martin Klein, Peasants in Africa (Beverly Hills, Calif., 1980); Stephen Bunker, Peasants against the State: The Politics of Market Control in Bugisu, Uganda, 1900–1983 (Chicago, 1987); Jonathan Barker, Rural Communities under Stress: Peasant Farmers and the State in Africa (Cambridge, 1989); Steven Feierman, Peasant Intellectuals: Anthropology and History in Tanzania (Madison, Wis., 1990); Sara Berry, No Condition Is Permanent: The Social Dynamics of Agrarian Change in Sub-Saharan Africa (Madison, 1993); Colin Bundy, The Rise and Fall of the South African Peasantry (Berkeley, Calif., 1979); William Beinart and Colin Bundy, Hidden Struggles in Rural South Africa (Berkeley, 1987); Charles van Onselen, The Seed Is Mine: The Life of Kas Maine, a South African Sharecropper, 1894–1985 (New York, 1996). For general studies of peasants and peasants in areas outside Africa, see Frederick Cooper, Allen F. Isaacman, Florencia E. Mallon, William Roseberry, and Steve J. Stern, Confronting Historical Paradigms: Peasants, Labor, and the Capitalist World System in Africa and Latin America (Madison, 1993); James C. Scott, The Moral Economy of the Peasant (New Haven, Conn., 1976); Scott, Weapons of the Weak: Everyday Forms of Peasant Resistance (New Haven, 1985); William Roseberry, Anthropologies and Histories: Essays on Culture, History, and Political Economy (New Brunswick, N.J., 1984); Florencia E. Mallon, Peasant and Nation: The Making of Postcolonial Mexico and Peru (Berkeley, 1995); Steve J. Stern, Resistance, Rebellion and Consciousness in the Andean Peasant World (Madison, 1987).

⁶ This point has been made convincingly by many observers; for one statement, see Elizabeth Colson, "The Impact of the Colonial Period on the Definition of Land Rights," in Victor Turner, ed., Colonialism in Africa, 1870–1960, Vol. 3: Profiles of Change: African Society and Colonial Rule, (Cambridge, 1971), 193–215.

⁷ For a Latin American perspective on such issues—differing from that adopted here—see Michael Kearney, Reconceptualizing the Peasantry: Anthropology in Global Perspective (Boulder, Colo., 1996).

⁸ While there have been some exceptions (notably in Kenyan studies), in general the legacy of Chayanov has not been prominent in African studies, nor have African examples been represented in the premier journal of the field, the Journal of Peasant Studies, in proportion to their historical importance. The significance of Chayanov's studies for the Journal of Peasant Studies was reconfirmed in the symposium "Aleksandr Chayanov and Russian Berlin," Frank Bourgholtzer, ed., Journal of Peasant Studies 26 (1999). For Chayanov's work, see Daniel Thorner, Basile Kerblay, and R. E. F. Smith, eds., A. V. Chayanov on the Theory of Peasant Economy (Madison, Wis., 1986).

⁹ For an introduction, see Ranajit Guha and Gayatry Chakravorty Spivak, eds., Selected Subaltern Studies (London, 1988); Ranajit Guha, ed., A Subaltern Studies Reader, 1986–1995 (Minneapolis, 1997); and the journal Subaltern Studies. The term "Subaltern Studies" derives from a concern with the issue of how elite classes came to define the culture, including the vision of history, of the non-elite classes of society. See Guha, "Introduction," A Subaltern Studies Reader, xiv–xvi. The first attempt to probe these questions systematically was Antonio Gramsci, Selections from the Prison Notebooks of Antonio Gramsci, Quintin Hoare and Geoffrey Nowell Smith, eds. (New York, 1971). While Gramsci's entire project to some extent dealt with the problems of defining and deciphering subaltern histories, these issues were most explicitly addressed in the chapter "Notes on Italian History: History of Subaltern Classes; Methodological Criteria," 52–55. Gramsci of course was mostly concerned with class dynamics in an industrializing Italy; nevertheless, his observations provide useful methodological insights for the study of peasants in colonial and postcolonial contexts. In particular, it is worth noting Gramsci's acknowledgment of the challenges of such analysis: "the history of subaltern groups is necessarily fragmented and episodic . . . Subaltern groups are always subject to the activity of ruling groups, even when they rebel and rise up . . . , even when they appear triumphant the subaltern groups are merely anxious to defend themselves." And finally he warns us (p. 55): "This kind of history can only be dealt with monographically and each monograph requires an immense quantity of material which is often hard to collect." Further discussion on these issues is found in the three contributions to the AHR Forum cited in note 10, and (among many others) in David Arnold, "Gramsci and Peasant Subalternity in India," Journal of Peasant Studies 112 (1989): 155–77; and T. J. Jackson Lears, "The Concept of Cultural Hegemony: Problems and Possibilities," AHR 91 (June 1985): 567–93.

¹⁰ The most explicit application of the subaltern studies paradigms to Africa is Frederick Cooper, "Conflict and Connection: Rethinking African Historiography," AHR 99 (December 1994): 1516–45. This was part of a larger forum on subaltern studies, including Gyan Prakash, "Subaltern Studies as Postcolonial Criticism" on South Asian historiography, and Florencia Mallon, "The Promise and Dilemma of Subaltern Studies: Perspectives from Latin American History," AHR 99 (December 1994). In a recent analysis of the history of Nyabingi healers in northern Rwanda and southwestern Uganda, Steven Feierman notes that, in many cases, a concern with hegemonic power has led analysts to depreciate and undervalue the continuing importance of local initiatives. He proposes an approach to historical reconstruction that can unveil subaltern discourses and practices rendered "invisible" by the changes of colonialism and capitalism: Feierman, "Colonizers, Scholars, and the Creation of Invisible Histories," in Victoria E. Bonnell and Lynn Hunt, eds., Beyond the Cultural Turn: New Directions in the Study of Society and Culture (Berkeley, Calif., 1999), 182–216.

¹¹ In the 1950s, Rwanda's population was classified into three principal social categories: Hutu constituted about 85 percent of the population, Tutsi about 14 percent, and Twa less than 1 percent. Strong racial, cultural, and historical stereotypes were associated with each of these categories, stereotypes we discuss below. Although these categories existed from precolonial times, their political salience changed during colonial rule, with the establishment of a clear ethnic hierarchy in the minds of Belgian administrators. Early accounts clearly show great regional variability and temporal flexibility among these categories, yet these images of "ethnicity" came to be accepted even by many within Rwanda, and by the end of colonial rule ethnic stratification had become a strong feature of Rwandan social reality. For a discussion of this process for one region of Rwanda, see Catharine Newbury, The Cohesion of Oppression: Clientship and Ethnicity in Rwanda, 1860–1960 (New York, 1988).

¹² Many works portray Rwanda in these terms, but the most prominent among them is Jacques-Jean Maquet, Le système des relations sociales dans le Ruanda ancien (Tervuren, 1954). The intellectual premise of this study was constructed in the late 1940s; for commentary on the research strategies and assumptions behind this work, see Alexis Kagame, Le code des institutions politiques de l'ancien Rwanda (Brussels, 1952), 12. Kagame notes that Maquet drew up his list of informants at the central court and designed his questionnaire with Kagame's guidance. It is therefore not surprising that Kagame noted that Maquet's results "validated" his own. (Maquet, however, omits these details in his discussion of method). Translated into English as The Premise of Inequality in Ruanda: A Study of Political Relations in a Central African Kingdom (London, 1961) at a time of the efflorescence of British structural-functional anthropology in Africa, and with the imprimatur of the International African Institute, Maquet's work became one of the iconic works of Anglophone scholarship in Africa. Even fifty years after the research on which this book was based, many today still uncritically accept this as a valid source for Rwandan society and history. That is analogous to scholars of the 1960s drawing on intellectual frameworks dating from before World War I; such a lapse would be inconceivable in most fields, since the methods, data, and conceptual understanding of African research have seen as profound a transformation over the last fifty years as in the previous fifty. For a critique of Maquet, see C. Newbury, Cohesion of Oppression, chap. 1. On Kagame's broader role in the construction of a hegemonic historiography, see notes 38 and 56 below.

¹³ Not all general commentaries on Rwanda share the popular imagery noted here, but most concentrate on ethnicity, kingship, conflict, and clientship as fixed historical categories. In doing so, they ignore much of the research of the last thirty years, which presents important new contours to historical understanding; for an early overview, see Chubaka Bishikwabo and David Newbury, "Recent Research in the Area of Lake Kivu: Rwanda and Zaire," History in Africa 7 (1980): 23–35. Especially prominent among the general histories of Rwanda were Alexis Kagame, Un abrégé de l'ethnohistoire du Rwanda, vol. 1 (Butare, 1972); and Kagame, Un abrégé de l'histoire du Rwanda, vol. 2 (Butare, 1975).

¹⁴ The idealized description of Rwandan aristocratic classes frequently drew on racial terminology. For a notorious example of such glorified imagery, see Paul del Perugia, Les derniers rois mages (Paris, 1978). In a manner typical of the early works on Rwandan history, Louis de Lacger describes Tutsi in the chapter "Les Batutsi, une branche des couches kouschites, ethiopiens, ou hamites." Absent any empirical data, he—like so many others of his day—conjoined race, culture, and history, drawing on selective physiological traits to postulate a sweeping historical narrative encompassing thousands of miles and thousands of years. De Lacger notes that the Greeks, Romans, and Arabs all described these "Hamites," and continues: "Arriving in Rwanda from upper Egypt or the Abyssinian highlands, one recognizes [the Tutsi] immediately. One had already seen these men of tall stature, . . . with slender bodies and long lanky limbs, well-proportioned features, and a noble, serious, and haughty demeanor. These are the brothers of the Nubians [in Sudan], the Galla [Somalia], the Danakil [Eritrea]. They have Caucasian features and resemble the Semites of Asia Minor. But their coloring is black, sometimes bronze or olive; their hair is curly . . . Before becoming black, these men were bronze in color." He then traces their emigration from Kush on the upper Nile to Egypt and then Ethiopia. Subsequently, he notes the presence of a "continuous pastoral avenue" from Ethiopia to Lake Kivu, which allowed these people to migrate to the Great Lakes region "without having to leave their ecological milieu." De Lacger, Le Ruanda (1939; Kabgayi, 1961), 56–57.

¹⁵ On the reasoning behind these figures, see Helen Codere, The Biography of an African Society: Rwanda, 1900–1960 (Tervuren, 1973), 70.

¹⁶ For a consideration of regional differences, see below. For sources, see notes 23–26 and 43–45.

¹⁷ For an elaboration of the concepts of dual colonialism and local agency under colonial rule, see C. Newbury, Cohesion of Oppression; and Jean-Baptiste Rushatsi, "Monarchie nyiginya et pouvoirs européens: Le nord et le nord-ouest du Rwanda sous un double colonisation, 1894–1916" (PhD dissertation, Université Catholique de Louvain, 1994).

¹⁸ For commentaries on the Twa, see José Kagabo and Vincent Mudandagizi, "Complaintes des gens de l'argile: Les Twa du Rwanda," Cahiers d'études africaines 14 (1974): 75–87; de Lacger, Le Ruanda; Jean-Claude Desmarais, "Les Twa du Rwanda: Du social au biologique," L'informateur 8 (1975): 62–64. Classic sources on the Twa include Paul Schebesta, Les Pygmées du Congo belge (Brussels, 1952); Peter Schumacher, "Die expedition des P. P. Schumacher zu den Zentralafrikanischen Kivu-Pygmäen in Ruanda," Anthropos 22 (1927): 289–90, 530–49; 23 (1928): 395–435.

¹⁹ These themes have been considered in greater detail in David Newbury, "The Invention of Rwanda: The Alchemy of Ethnicity" (paper presented at the African Studies Association, Orlando, 1995); David Newbury, "Rwanda and Burundi: A Precolonial Overview" (paper presented to the conference on Central African History, Oxford University, July 1995); and David Newbury, Kings and Clans: Ijwi Island and the Lake Kivu Rift, 1780–1840 (Madison, Wis., 1991).

²⁰ Some exceptions include Pierre Smith, Le récit populaire (Paris, 1975); Luc de Heusch, Rois nés d'un coeur de vache (Paris, 1982); Codere, Biography; and especially the massive collection of transcribed texts of Jan Vansina, "Ibiteekerezo: Historical Narratives from Rwanda" (Chicago, Center for Research Libraries).

²¹ The point of departure was Jan Vansina, L'évolution du royaume Rwanda des origines à 1900 (Brussels, 1962); the methodological premises were first laid out in Vansina, De la tradition orale: Essai de méthode historique (Tervuren, 1961). For other studies drawing extensively on oral sources, see (among other examples) Claudine Vidal, "Economie de la société féodale rwandaise," Cahiers d'études africaines 14 (1974): 52–75 (on labor and ethnicity); Alison Des Forges, "Defeat Is the Only Bad News: Rwanda under Musinga, 1906–1931" (PhD dissertation, Yale University, 1972), on colonial rule and court politics; C. Newbury, Cohesion of Oppression, on clientship, colonialism, ethnicity, and the state; Jim Freedman, Nyabingi: The Social History of an African Divinity (Tervuren, 1984), on local religious practices; Codere, Biography, on biographies, including women's biographies; and Danielle de Lame, Une colline entre mille ou le calme avant la tempête: Transformations et blocages du Rwanda rural (Tervuren, 1996), a comprehensive cultural study of a local community, with a focus on political economy. For a general overview and assessment of social research over the first twenty years of postcolonial Rwanda, see Bishikwabo and Newbury, "Recent Research." On the development and debates of oral historiography in Africa, David Newbury, "Contradictions at the Heart of the Canon: Oral Historiography in Africa 1960–1980" (paper presented to the conference on Oral Historiography in Africa, Bellagio, February 1997).

²² Surveys of colonial Rwanda that emphasize these themes include René Lemarchand, Rwanda and Burundi (London, 1970); Ian Linden with Jane Linden, Church and Revolution in Rwanda (Manchester, 1977); Filip Reyntjens, Pouvoir et droit au Rwanda (Tervuren, 1985); Rumiya, Le Rwanda.

²³ Jan Czekanowski, Forschungen im Nil-Kongo-Zwischengebiet (Leipzig, 1917). Such perspectives were also present in the important work of Peter Schumacher, "P. Schumachers Aufzeichnungen: Ruanda," Micro-Bibliotecha Anthropos, vol. 28A, 1958, Anthropos-Institut, St. Augustin bei Bonn, Germany. Also see Richard Kandt, Caput Nili (Berlin, 1921); on Kandt in Rwanda, see Reinhart Bindseil, Ruanda und Deutschland seit den Tagen Richard Kandts (Berlin, 1987).

²⁴ Albert Pagès, Au Ruanda, sur les bords du lac Kivu (Congo belge): Un royaume hamite au centre de l'Afrique (Brussels, 1933).

²⁵ De Lacger, Ruanda.

²⁶ On the deposition of Musinga, see de Lacger, Ruanda, 525–42; Des Forges, "Defeat Is the Only Bad News," chaps. 8–9, esp. 349–54; Ferdinand Nahimana, Le blanc est arrivé: Le roi est parti (Kigali, 1987); Rumiya, Le Rwanda, 169–89. The Catholic prelate of the day, Mgr. Léon Classe, had a powerful influence on the decision to depose Musinga, who was rusticated in part for his "anti-Christian" and "amoralistic" views; de Lacger, Ruanda, 525.

²⁷ "Les Missionnaires d'Afrique" is the formal title of the order. But members of the order are commonly referred to as the "White Fathers," after the long white cassock that characterizes their dress code, modeled on the Muslim attire of North Africa where the order was founded. On the development of the Missionnaires d'Afrique and on mission guidelines, see François Renault, Le Cardinal Lavigerie: L'église, l'Afrique, et la France (Paris, 1992). We lack a comprehensive history of the Missionnaires d'Afrique, but on the early writings of the White Fathers in Rwanda, see R. Heremans, A. Bart, and F. Bart, "Agriculture et paysage Rwandais à travers des sources missionnaires (1900–1950)," Cultures et développement 14 (1982): 3–39; Roger Heremans and Emmanuel Ntezimana, eds., Journal de la Mission de Save (Rwanda), 1899–1905 (Ruhengeri, 1987). On broader aspects, see François Renault, Lavigerie, l'esclavage, et l'Europe, 2 vols. (Paris, 1971); and Roger Heremans, L'éducation dans les missions des Pères Blancs en Afrique centrale, 1879–1914 (Brussels, 1983).

²⁸ German policy was first forged in Burundi. But the attitudes adopted there were applicable with even greater force to Rwanda, because in the late nineteenth century Rwanda was much more centralized and the royal court had much more power within its central sphere of influence than was true for Burundi. For Burundi, see Joseph Gahama, Le Burundi sous administration Belge (Paris, 1984); Hans Meyer, Les Barundi (Paris, 1984); and Pierre Ryckmans, Une page d'histoire coloniale: L'occupation allemande dans l'Urundi (Brussels, 1953), 6: "Our policy must aim at upholding the authority of the chiefs so that they will be convinced that their well-being . . . depends on their loyalty to the German government." On Rwanda, see Des Forges, "Defeat Is the Only Bad News"; Rumiya, Le Rwanda; Nahimana, Le blanc est arrivé.

²⁹ For example, of the five Catholic missions established between 1900 and 1903, four were in areas of opposition to Nyiginya royalty: in addition to Save, not far from the royal court (1900), later missions included Zaza (Gisaka, in the southeast, 1900); Nyundo (Bugoyi, in the northwest, 1901); Rwaza (Mulera in the extreme north, 1903); and Mibirizi (Bukunzi, in the extreme southwest, 1903). Similarly, three of the next four missions were located outside the areas of firmly established court dominance. Because of the conditions of their founding and the requirement of the order that the missionaries keep a daily dairy, these accounts often serve as rich sources for local history; see Heremans and Ntezimana, Mission de Save; Heremans, Bart, and Bart, "Agriculture et paysages"; and P. Schumacher, "Aufzeichnungen: Ruanda."

³⁰ Alison Des Forges, "Kings without Crowns: The White Fathers in Rwanda," in Daniel McCall, Norman Bennett, and Jeffrey Butler, eds., Eastern Africa History (New York, 1975), 176–207; Des Forges, "Defeat Is the Only Bad News"; Linden and Linden, Church and Revolution. While the perceptive quality and the interest in local people and events varied with each observer, the diary of daily events kept at each parish nonetheless provided a valuable window on local events during the early colonial period. Copies are available at the White Fathers archives in Rome. For a published example, see Heremans and Ntezimana, Mission de Save. Also, when used critically and evaluated in context, published works by missionaries such as Albert Pagès, Louis de Lacger, Léon Delmas, Félix Dufays, Marcel Pauwels, and Peter Schumacher provide ample testimony to the value of such observations.

³¹ For the religious wars of Buganda at the end of the nineteenth century, see Michael Wright, Buganda in the Heroic Age (Nairobi, 1971); C. C. Wrigley, "The Christian Revolution in Buganda," Comparative Studies in Society and History 2 (1959): 33–48; Christopher Wrigley, Kingship and State: The Buganda Dynasty (Cambridge, 1996); John M. Gray, "The Year of the Three Kings in Buganda, 1888–89," Uganda Journal 14 (1950): 15–52; M. S. M. Semakula Kiwanuka, A History of Buganda to 1900 (London, 1971); D. A. Low, "Conversion, Revolution and the New Regime in Buganda, 1860–1900," in Low, Buganda in Modern History (London, 1971), 13–54; Low, Religion and Society in Buganda, 1875–1900 (Kampala, n.d. [1957]); John Rowe, Lugard at Kampala (Kampala, 1969); Michael Twaddle, Kakungulu and the Creation of Uganda (London, 1993).

³² For histories of the Catholic Church in Rwanda, see Linden and Linden, Church and Revolution; Roger Heremans, "L'arrivée des Pères Blancs au Rwanda," Dialogue, no. 57 (1976): 72–83; Heremans, L'éducation dans les missions des Pères Blancs; Gamaliel Mbonimana, "L'instauration d'un royaume chrétien au Rwanda (1900–1931)" (PhD dissertation, Louvain la Neuve, 1981); Gamaliel Mbonimana, "Christianisation indirecte et cristallisation des clivages ethniques au Rwanda (1925–1931)," Enquêtes et documents d'histoire africaine 3 (1978): 125–63; Justin Kalibwami, Le catholicisme et la société rwandaise 1900–1962 (Paris, 1991); Paul Rutayisire, La christianisation du Rwanda (1900–1945): Méthode missionnaire et politique selon Mgr. Léon Classe (Fribourg, 1987).

³³ Pierre Bourdieu, Towards a Theory of Praxis (Cambridge, 1976). More than an "orthodox" history—which implicitly accepts the possibilities of alternative viewpoints—Bourdieu's concept of "doxic" hegemony presumes the delegitimation (or absence) of meaningful alternative visions. It is in this sense that we see the power of the hegemonic view of Rwandan history, which came to combine the Rwandan political elite with the ecclesiastical authorities. Other viewpoints were dismissed as "politically motivated" or simply "local" histories, and therefore not worthy of consideration; the history of rural areas, and of peasants, was simply not considered important.

³⁴ Bernard Lugan, "Causes et effets de la famine 'Rumanura' au Rwanda, 1916–1918," Canadian Journal of African Studies 10 (1976): 347–56; Rumiya, Le Rwanda, 59–74; Nahimana, Le blanc est arrivé, 121–28.

³⁵ For a discussion of this dispute, see Rumiya, Le Rwanda, 81–131. This had not been the only boundary dispute in question. Areas in the west (Ijwi Island) and in the north (Bufumbiro, originally referred to by some as "British Rwanda") were also contested, as parts of the precolonial dynastic domain. William Roger Louis, Ruanda-Urundi: 1884–1919 (Oxford, 1963), esp. 41–92; Elizabeth Hopkins, "The International Boundary as a Factor in the Extension of Colonial Control," in Daniel F. McCall, Norman R. Bennett, and Jeffrey Butler, eds., Eastern African History (New York, 1969), 208–45; Donald Denoon, ed., A History of Kigezi in South-West Uganda (Kampala, 1972).

³⁶ Louis, Ruanda-Urundi; David Newbury, "The Rwakayihura Famine of 1928–1929: A Nexus of Colonial Rule in Rwanda," in Histoire sociale de l'Afrique de l'est (Bujumbura, 1991), 269–85.

³⁷ David Newbury, "Augustinian Models in Rwanda: Religious Movements and Political Transformation," Svensk Missionstidskrift 3 (1995): 16–34; Patricia St. John, Breath of Life (London, 1971); Joe E. Church, Quest for the Highest (Exeter, 1981); John V. Taylor, The Growth of the Church in Buganda (London, 1958).

³⁸ The starting point for the chronology of Rwandan dynastic history is Alexis Kagame, La notion de génération appliquée à la généalogie dynastique et à l'histoire du Rwanda des X^è–XI^è siècles à nos jours (Brussels, 1959). For elaboration, see Kagame, Inganji Karinga (Butare, 1959 [1943; 1947]); Un abrégé, vol. 1; and "La chronologie au Burundi dans les genres littéraires de l'ancien Rwanda," Etudes rwandaises 12 (1979): 1–30. The most important reassessments of Kagame's proposals include Vansina, L'évolution; Keith Rennie, "The Precolonial Kingdom of Rwanda: A Reinterpretation," Transafrican Journal of History 2 (1972): 11–53; and Jean-Népomucène Nkurikiyimfura, "La revision d'une chronologie: Le cas du royaume du Rwanda," in Claude-Hélène Perrot, Gilbert Gonnin, and Ferdinand Nahimana, eds., Sources orales de l'histoire de l'Afrique (Paris, 1989). For a broader critique of the official chronology—and of the dynastic genealogy—see David Newbury, "Trick Cyclists? Recontextualizing Rwandan Dynastic Chronology," History in Africa 21 (1994): 191–217.

³⁹ Vansina, L'évolution, chap. 3, makes this point convincingly; see also D. Newbury, "Trick Cyclists."

⁴⁰ Vansina, L'évolution, 74–76; D. Newbury, "Alchemy of Ethnicity." For a more recent case of such claims, see David Newbury, "Irredentist Rwanda: Ethnic and Territorial Frontiers in Central Africa," Africa Today 44 (1997): 211–22.

⁴¹ Kagame, Un abrégé, 1: 58–59, 138–39.

⁴² A. d'Arianoff, L'histoire des Bagesera, souverains du Gisaka (Brussels, 1952); but for commentary on how central court perspectives defined d'Arianoff's presentation, see also D. Newbury, "Trick Cyclists," 193–95.

⁴³ An early anthropological work to focus on a region was Marcel d'Hertefelt, "Huwelijk, famille en aanverwantschap bij de Reera van noordwesterlijk Rwaanda," Zaire 13 (1959): 115–47, 243–85. Other early regional works include Marcel Pauwels, "Le Bushiru et son muhinza ou roitelet Hutu," Annali del Pontificio Museo Missionario Etnologico 31 (1967): 205–322, and many other of his publications; Félix Dufays and Vincent de Moor, Au Kinyaga: Les enchaînés (Brussels, 1938); d'Arianoff, L'histoire des Bagesera; Pagès, Un royaume hamite, 634–701.

⁴⁴ On these polities, see Pauwels, "Le Bushiru"; Ferdinand Nahimana, "Les bami ou roitelets Hutu du corridor Nyabarongo-Mukungwa," Etudes rwandaises 12 (1979): 1–25; Nahimana, Le Rwanda; Emmanuel Ntezimana, "Coutumes et traditions des royaumes hutu du Bukunzi et du Busozo," Etudes rwandaises 13 (1980): 15–39; and Ntezimana, "L'arrivée des Européens au Kinyaga et la fin des royaumes hutu du Bukunzi et du Busozo," Etudes rwandaises 13 (1980): 1–29. For discussion of the historical similarities of these polities with political organizations west of the Rift Valley (in what is now eastern Congo), see D. Newbury, Kings and Clans, chap. 2; and Richard Sigwalt, "Early Rwanda History: The Contribution of Comparative Ethnography," History in Africa 2 (1975): 137–46. For a discussion of royal court (Nyiginya) attitudes toward the cultures of this area, see David Newbury, "Bunyabungo: The Western Rwandan Frontier, 1750–1850," in Igor Kopytoff, ed., The African Frontier (Bloomington, Ind., 1986), 162–93.

⁴⁵ D'Hertefelt, "Huwelijk, famille en aanverwantschap"; Jim Freedman, "Principles of Relationship in Rwandan Kiga Society" (PhD dissertation, Princeton University, 1974); May Edel, The Chiga of Western Uganda (New York, 1957); Denoon, History of Kigezi; Paul Ngologaza, Kigezi and Its People (Kampala, 1969); Benoni Turyahikayo-Rugyema, "The History of the Bakiga of Southwestern Uganda and Northern Rwanda (PhD dissertation, University of Michigan, 1974); Elizabeth Hopkins, "The Nyabingi Cult of Southwestern Uganda," in Robert Rotberg and Ali Mazrui, eds., Protest and Power in Black Africa (New York, 1970), 258–336; Hopkins, "The Ethnography of Conquest," in Christine Ward Gailey, ed., Dialectical Anthropology: Essays in Honor of Stanley Diamond, Vol. 1, Civilization in Crisis: Anthropological Perspectives (Gainesville, Fla., 1992), 137–65. For other examples of identities and dynamics that reach beyond state structures, see David Newbury, "Lake Kivu Regional Trade in the Nineteenth Century," Journal des Africanistes 50 (1980): 6–31; Antoine Nyagahene, "Les activités économiques et commerciales du Kinyaga dans le seconde partie du XIX^è siècle" (Thèse de Licence, UNR-Butare, 1979); Njiga Birhakaheka and Nsibula Kirhero, "Nyangezi dans ses relations commerciales avec le Rwanda, le Burundi, et le Bufulero," Etudes rwandaises 14 (1981): 89–98; Luc de Heusch, Le Rwanda et la civilisation interlacustre (Brussels, 1966); Iris Berger, Religion and Resistance: East African Kingdoms in the Precolonial Period (Tervuren, 1981); Berger, "Fertility as Power: Spirit Mediums, Priestesses, and the Precolonial State in Interlacustrine East Africa," in David Anderson and Douglas Johnson, eds., Revealing Prophets: Prophecy in Eastern African History (London, 1995), 65–82; Freedman, Nyabingi; Feierman, "Colonizer, Scholars, and the Creation of Invisible Histories"; Jean-Pierre Chrétien, "La fermeture du Burundi et du Rwanda aux commerçants de l'extérieur (1905–1908)," in Catherine Coquery-Vidrovitch and Alain Forrest, eds., Entreprises et Entrepreneurs en Afrique, vol. 2 (Paris, 1983), 25–47.

⁴⁶ Any firm enumeration of "cultural zones" in this region, of course, is an arbitrary exercise, as these shade into each other in a complex pattern of overlapping regional and local identities; for analytic purposes, however, while one can argue over the lines of demarcation, the general point of such differentiation remains important. Regional studies that privilege local dialogue with state power include C. Newbury, Cohesion of Oppression; de Lame, Une colline; Pierre Gravel, Remera: A Community in Eastern Rwanda (The Hague, 1968); Jean-Pierre Chrétien, "La révolte de Ndungutse (1912): Forces traditionnelles et pression coloniale au Rwanda allemand," Revue française d'Outre-Mer 59 (1972): 645–80; Freedman, "Principles"; Alison Des Forges, "The Drum Is Greater Than the Shout: The 1912 Rebellion in Northern Rwanda," in Donald Crummey, ed., Banditry, Rebellion and Social Protest in Africa (Portsmouth, N.H., 1986), 311–31; Nahimana, Le Rwanda; Rushatsi, "Monarchie nyiginya et pouvoirs européens."

⁴⁷ John Hannington Speke, Journal of the Discovery of the Source of the Nile (New York, 1868), 242. Speke had also previously traveled in the Horn of Africa and was accompanied by a Somali aide-de-camp. In a sense, then, Speke's historical scenario mirrored his own travels.

⁴⁸ Speke, Journal, 242.

⁴⁹ Speke, Journal, 243.

⁵⁰ Harry Johnston, The Uganda Protectorate (London, 1902). (Johnston was also an intellectual polymath, the author of influential works on linguistics, botany, and zoology, as well as history.) Foremost among the Baganda elites constructing a hegemonic royal history was Apolo Kagwa, for forty years the Katakiro (or prime minister) of Buganda; among other works, see Ssekabaka by'e Buganda (Kampala, 1901), translated as The Kings of Buganda (Nairobi, 1971). John Rowe and Michael Twaddle have explored aspects of Ganda "subaltern historiography"—histories excluded from the mainstream writings: Rowe, "Myth, Memoir, and Moral Admonition: Luganda Historical Writing, 1893–1969," Uganda Journal 33 (1969): 17–40, 217–19; Rowe, "Erieza Kintu's Sultani Anatoloka: A Nineteenth Century History Memoir from Buganda," History in Africa 20 (1993): 313–19; Rowe, "The Anti-Establishment Voice in Twentieth Century Buganda: Or, the Power of the Printing Press and How the Ruling Elite Lost Their Monopoly over It," Makerere University History Department Seminar Paper (August 1991); Twaddle and Rowe, War, Religion, and Revolution in Uganda: Six Dissident Discourses (East Lansing, forthcoming); Twaddle, "On Ganda Historiography," History in Africa 2 (1976): 85–101. In addition, see Wrigley, Kingship and State.

⁵¹ Julien Gorju, Entre le Victoria, l'Albert, et l'Edouard (Rennes, 1920); later, for application specifically to Burundi and Rwanda, see Gorju, Face au royaume hamite du Ruanda, le royaume frère du Burundi (Brussels, 1938).

⁵² Among Africanists, "interlacustrine" is the conventional reference to the region between Lakes Albert, Edward, Kivu, and Tanganyika on the west and Lake Victoria on the east. The societies within this geographical domain form a distinct cultural zone and share several important characteristics: centralized kingship, stratified social organization, and mixed agrarian-pastoral economies. Hence they are collectively referred to as the "interlacustrine" cultures.

⁵³ David Newbury, "Bushi and the Historians," History in Africa 5 (1978): 131–51.

⁵⁴ Historique et chronologie du Ruanda [Kabgayi, 1956].

⁵⁵ Ivan Reisdorff, "Enquêtes foncières au Ruanda," unpublished manuscript (Butare, 1952).

⁵⁶ On Kagame, see Jean-Pierre Chrétien, "Confronting the Unequal Exchange between the Oral and the Written," in Bogumil Jewsiewicki and David Newbury, eds., African Historiographies: What History for Which Africa? (Newbury Park, Calif., 1986), 75–90, esp. 84–87; Claudine Vidal, "Alexis Kagame entre mémoire et histoire," History in Africa 15 (1988): 493–504; Vidal, "Alexis Kagame," in Vidal, Sociologie des Passions (Paris, 1991), 45–61; and D. Newbury, "Trick Cyclists."

⁵⁷ On clientship: C. Newbury, Cohesion of Oppression; and Catharine Newbury, "Ubureetwa and Thangata: Catalysts to Peasant Political Consciousness in Rwanda and Malawi," Canadian Journal of African Studies 14 (1980): 97–111; Claudine Vidal, "Le Rwanda des anthropologues ou le fétischisme de la vache," Cahiers d'études africaines 9 (1969): 384–401; Nkurikiyimfura, Le gros bétail. On ethnicity, see Catharine Newbury, "Ethnicity in Rwanda: The Case of Kinyaga," Africa 48 (1978): 17–28; Claudine Vidal, "Situations ethniques au Rwanda," in Jean-Luc Amselle and Elikia M'bokolo, eds., Au coeur de l'ethnie (Paris, 1985), 167–84. On kingship, D. Newbury, Kings and Clans; Richard Sigwalt, "The Early History of Bushi" (PhD dissertation, University of Wisconsin, 1975). On clanship, Marcel d'Hertefelt, Les clans du Rwanda ancien (Tervuren, 1971); David Newbury, "The Clans of Rwanda: An Historical Hypothesis," Africa 50 (1980): 389–403; Catharine Newbury, "Deux Lignages au Kinyaga," Cahiers d'études africaines 14 (1974): 26–38. On royal ritual, Marcel d'Hertefelt and André Coupez, La royauté sacrée de l'ancien Rwanda (Tervuren, 1964); David Newbury, "What Role Has Kingship? An Analysis of the Umuganura Ritual of Rwanda," Africa-Tervuren 27 (1981): 89–101; Richard Sigwalt, "Early Rwanda History: The Contribution of Comparative Ethnography," History in Africa 2 (1975): 137–46.

⁵⁸ Kagame, Un abrégé, 1: 212–14; Un abrégé, 2: 2, 28–35; Alexis Kagame, Les milices du Rwanda précolonial (Brussels, 1963), 155–57. In fact, the story was much more complicated. Rwabugiri had been born to Nkoronko's wife—even though the official sources claim it was through her liaison with Rwogera, Nkoronko's brother and Rwabugiri's predecessor as king of Rwanda. Rwogera later proposed an exchange of sons with Nkoronko, and Rwabugiri (then named Sezisoni), Rwogera's adopted son (and supposed biological offspring?), then succeeded to the throne. Over the course of his reign, Rwabugiri not only condemned to execution his uncle (and biological father?) but his mother and his wife (the mother of his designated successor and co-regnant) as well; in addition, he had his cousin blinded. And these were not the only cases where members of the royal family were killed or maimed for knowing too much history. Shrewd actors attended carefully to hegemonic history.

⁵⁹ Among many examples, see Czekanowski, Forschungen; Albert Pagès, "Notes sur le régime des biens dans le Province du Bugoyi," Congo 19 (1938): 392–433; C. Newbury, Cohesion of Oppression, chap. 7.

⁶⁰ Johan Pottier, "The Politics of Famine Prevention: Ecology, Regional Production, and Food Complementarity in Western Rwanda," African Affairs 85 (1986): 207–37; Johan Pottier, "'Three's a Crowd': Knowledge, Ignorance and Power in the Context of Urban Agriculture in Rwanda," Africa 59 (1989): 461–77; Pottier, "Taking Stock: Food Marketing Reform in Rwanda, 1982–89," African Affairs 92 (1993): 5–30. For analogous situations elsewhere, see Kathleen Staudt, "Agricultural Productivity Gaps: A Case Study of Male Preference in Government Policy Implementation," Development and Change 9 (1978): 439–58; Kathleen Staudt, "Women Farmers and Inequities in Agricultural Services," in Edna Bay, ed., Women and Work in Africa (Boulder, Colo., 1981), 207–24.

⁶¹ Emmanuel Ntezimana, "Kinyamateka, Temps Nouveaux d'Afrique: L'évolution socio-politique du Rwanda (1954–1959)," Etudes rwandaises, Special Issue (March 1978): 1–29.

⁶² Helen Codere, "Power in Rwanda," Anthropologica 4 (1962): 45–85; Lemarchand, Rwanda and Burundi; Vidal, "Le Rwanda des anthropologues"; Vansina, L'évolution; C. Newbury, Cohesion of Oppression; Nahimana, Le Rwanda. Bishikwabo and Newbury, "Recent Research," summarizes many of these works from the early postcolonial period.

⁶³ This conclusion is shared by Danielle de Lame, "Instants Retrouvés: Rwanda, regards neufs au fil du temps," in Patrick Wymeersch, ed., Liber Amicorum Marcel d'Hertefelt (Brussels, 1993), 115–31.

⁶⁴ For one example, see République du Rwanda, Résultats de l'enquête nationale agricole 1984, vol. 1 (Kigali, 1985). Exceptions that explore the changing relations between peasants and the state include André Guichaoua, Destins paysans et politiques agraires en Afrique centrale, Vol. 1: L'ordre paysan des hautes terres centrales du Burundi et du Rwanda (Paris, 1989); Catharine Newbury, "Rwanda: Recent Debates over Governance and Rural Development," in Governance and Politics in Africa, Goran Hyden and Michael Bratton, eds. (Boulder, Colo., 1992), 193–219; de Lame, Une colline; Timothy P. Longman, "Christianity and Crisis in Rwanda: Religion, Civil Society, Democratization, and Decline" (PhD dissertation, University of Wisconsin, 1995); Claudine Vidal, "Questions sur le rôle des paysans durant le génocide des rwandais tutsi," Cahiers d'études africaines 38 (1998): 331–45. Also sensitive to rural transformations, though this is not the main focus of his study, is Filip Reyntjens, L'Afrique des Grands Lacs en crise: Rwanda, Burundi, 1988–1994 (Paris, 1994).

⁶⁵ For a critique of how development agencies ignored power dynamics in Rwanda, see René Lemarchand, "The World Bank in Rwanda: The Case of the Office de Valorisation Agricole et Pastorale du Mutara (OVAPAM)" (Bloomington, Ind., 1982); Jean-Pierre Godding, "Foreign Aid as an Obstacle to Development: The Case of Rwanda's Rural Development Projects," in Michael Lipton, ed., International Perspectives on Rural Development (Sussex, 1984); Jean-Pierre Godding, "Grands projets et développement communal," Dialogue 134 (1989): 3–15; Peter Uvin, Aiding Violence: The Development Enterprise in Rwanda (West Hartford, Conn., 1998). On the broader issues involved, see James Ferguson, The Anti-Politics Machine: "Development," Depoliticization, and Bureaucratic Power in Lesotho (Cambridge, 1990); and Arturo Escobar, Encountering Development: The Making and Unmaking of the Third World (Princeton, N.J., 1995).

⁶⁶ For precolonial examples of such interactions from western Rwanda and their neglect (or ridicule) in conventional historiography, see D. Newbury, "Lake Kivu Trade"; D. Newbury, "Bunyabungo"; Pagès, Un royaume hamite, 543–701; Nyagahene, "Les attitudes économiques et commerciales du Kinyaga"; Birhakaheka and Kirhero, "Nyangezi"; Chubaka Bishikwabo, "Histoire d'un état Shi en Afrique des Grands Lacs: Kaziba au Zaire" (PhD dissertation, Louvain la Neuve, 1982).

⁶⁷ In part, the perpetuation of this insular character of official historiography comes from the colonial administrative traditions and postcolonial academic constructs. On the one hand, a language barrier separated Rwandan historians from East African scholarship, and the gap between Anglophone and Francophone traditions was reflected in different intellectual traditions, as well as simply in language. Furthermore, to the west, despite Belgian over-rule in each territory, as a League of Nations mandated territory and later a UN Trust Territory, Rwanda was administratively distinct from the Congo; indeed, Belgian administrators in Rwanda were not posted to the Congo. Consequently, Rwandan administrators and government personnel (Africans as well as expatriates) came to see themselves as culturally distinct from areas to their west (in the Congo) as well as to their east and north (Tanganyika and Uganda: British territories). Such perceptions drew attention away from the strong and lasting cultural connections between peoples within Rwandan state boundaries and others to their west, north, and east; colonial emphasis on the autonomy of Rwanda was reflected in the historiography as well.

⁶⁸ C. Newbury, "Ubureetwa and Thangata."

⁶⁹ De Lacger, Le Ruanda, 50–55; Kagame, Abrégé, 1: 28–30; and especially Maquet, Premise of Inequality, 129–73. Maquet's interpretations were widely diffused to Anglophone anthropologists and others through the writings of Lucy Mair. See Mair, Primitive Government (Harmondsworth, 1962); Mair, African Societies (Cambridge, 1974), 166–81; and Mair, African Kingdoms (Oxford, 1977).

⁷⁰ The assumptions of court hegemony about clientship were challenged first by Codere, "Power in Rwanda." Codere's critique was then followed up by Vidal, "Le Rwanda des anthropologues ou le fétichisme de la vache"; and Claudine Vidal, "Economie de la société féodale Rwandaise," Cahiers d'études africaines 14 (1974): 52–74. Other studies that, like Vidal's, show clientship as ultimately embedded in political process and that trace the evolution of multiple forms of clientship and their interrelationships over time include C. Newbury, "Deux lineages"; C. Newbury, Cohesion of Oppression, 73–147; Gravel, Remera, 170–85; Joseph Rwabukumba and Vincent Mudandagizi, "Les formes historiques de la dépendance personnelle dans l'Etat rwandais," Cahiers d'études africaines 14 (1974): 6–25; Nkurikiyimfura, Le gros bétail. Some forms of voluntary clientship and friendship ties were present, of course, but they were not the only forms—indeed, not the predominant forms in Rwanda during the nineteenth and twentieth centuries. In particular, it is useful to distinguish ubuhake forms from ubugabire forms of clientship found in Burundi and elsewhere in the region; see, for example, Albert A. Trouwborst, "L'organisation politique et l'accord de clientèle au Burundi," Anthropologica, n.s., 4 (1962): 9–43; Adrien Ndikuriyo, "Contrats de bétail, contrats de clientèle, et pouvoir politique dans le Bututsi à la fin du XIX^è siècle," Etudes d'histoire africaine 7 (1975): 59–76.

⁷¹ The case study from which these conclusions emerge is found in Jean-François Saucier, "The Patron-Client Relationship in Traditional and Contemporary Rwanda" (PhD dissertation, Columbia University, 1974). For an elaboration, see de Lame, Une colline, 223–25, 236–41. De Lame makes the important point that in Rwandan culture (now as in the past), there are diverse ways to obtain cattle outside the institution of ubuhake. She also shows (pp. 226–27) that in the local area she studied in the late 1980s, more than 50 percent of Hutu households had one or more cows, while only about 33 percent of Tutsi households had cattle. Those with the largest number of cattle (five to nine) were the "modern elites"—merchants or people with a salary.

⁷² Some works that refer to labor issues include Codere, Biography; Philippe Leurquin, Le niveau de vie des populations rurales du Ruanda-Urundi (Paris, 1960); Audrey Richards, Economic Development and Tribal Change: A Study of Immigrant Labor in Uganda (Cambridge, 1956); Belgium, Gouvernement Belge, Rapport présenté par le gouvernement Belge au Conseil de la Sociétés des Nations au sujet de l'administration du Ruanda-Urundi, 1924–1937 (Geneva, 1925–38); Belgium, Gouvernement Belge, Ministère des Colonies/Ministère du Congo Belge et du Ruanda-Urundi, Rapport présenté par le Gouvernement Belge au conseil de Tutelle des Nations Unies sur l'administration du Ruanda-Urundi, 1948–1959 (Brussels, 1949–60). A literary treatment of labor is found in Saverio Nayigiziki, Mes transes à trente ans, Vol. 2: De mal en pis: Histoire vécue, melée de roman (Astrida [Butare], 1955). For a scholarly study that did address labor issues, see Vidal, "Economie de la société féodale."

⁷³ Des Forges, "Defeat Is the Only Bad News," 340–42.

⁷⁴ Many testimonies in Codere, Biography, illustrate this. Also, C. Newbury, Cohesion of Oppression, 169. Our own interviews confirm this for eastern Rwanda.

⁷⁵ Some exceptions on postcolonial agrarian issues include Pierre Kabagabo, "Les programmes agricoles et la contribution de la femme," Session d'étude et d'information des cadres de l'action sociale tenue à Butare du 10 au 26 janvier 1966 (Kigali, 1966); Charles Ntakirutimana, "La femme rwandaise: Purgatoire ou libération?" Le diaspason 11 (1976): 63–68; Charles Ntampaka, "L'égalité de l'homme et de la femme dans la société rwandaise," Le diaspason 11 (1977): 51–57; Odette Ubonabenshi, "Femme rurale et développement du pays," Dialogue 133 (1989): 25–36.

⁷⁶ Gouvernement Belge, Ministère des Colonies, l'Office de l'information et des relations publiques pour le Congo belge et le Ruanda-Urundi, Le Ruanda-Urundi (Brussels, 1959), 253.

⁷⁷ United Nations Development Programme, Rural Women's Participation in Development, Evaluation Study no. 3 (New York, 1980), 81. Similarly for tea production, see Joseph Laure, Des vivres ou du thé? L'alimentation et les conditions de vie de familles rwandaises (Paris, 1986).

⁷⁸ On women forced to do corvée in the absence of their husbands, see C. Newbury, Cohesion of Oppression, 176–77; our interviews elsewhere in Rwanda confirm this pattern, as infrequent but not unknown. On the prevalence of work abroad, see also Codere, Biography.

⁷⁹ For an analogous case in Tanzania of state power reinforcing gender hierarchy, see Marjorie Mbilinyi, "Runaway Wives in Colonial Tanganyika—Forced Labor and Forced Marriage in Rungwe District, 1919–1961," International Journal of the Sociology of Law 16 (1988): 1–29.

⁸⁰ Richards, Economic Development and Tribal Change; Jean-Pierre Chrétien, "Des sédentaires devenues migrants: Les motifs des départs des Barundi et des Rwandais vers l'Uganda," Cultures et développement 10 (1978): 71–101; D. Newbury, "Rwakayihura Famine"; C. Newbury, Cohesion of Oppression, chap. 8; Jim Freedman, "East African Peasants and Capitalist Development: The Kiga of Northern Rwanda," in David H. Turner and Gavin Smith, eds., Challenging Anthropology (New York, 1979), 245–60.

⁸¹ On rural responses to famine: Reisdorff, "Enquêtes foncières"; Chrétien, "Des sédentaires devenus migrants"; D. Newbury, "Rwakayihura Famine"; Anne Cornet, "Histoire d'une famine: Rwanda 1924–1930," Enquêtes et documents africains 13 (1996); Leurquin, Le niveau de vie, 31–32.

⁸² A revealing commentary by Jean Paul Harroy (the highest administrative official of Ruanda-Urundi during the late 1950s) glorifies the role of the state in reconfiguring rural production. See Bonaventure Habimana and Jean-Paul Harroy, "Instauration et abrogation des cultures vivrières obligatoires au Rwanda," Civilisations 30 (1980): 33–66. But Leurquin, who worked in rural areas at the very time of Harroy's administration, notes that in one "experiment," it was announced that a community was allowed to plant what they wished; immediately, the fields of manioc (required cultivation) disappeared, after, he added, "30 years of 'cultures éducatives'" (i.e., forced crops). Le niveau de vie, 72.

⁸³ See E. Everaerts, "Monographie agricole du Ruanda-Urundi," Bulletin agricole du Congo Belge 30 (1939): 343–96, 581–663; and the annual reports of the Belgian government to the League of Nations and the United Nations. The viewpoints are epitomized in the report by the Belgian Ministère des Colonies, Plan décennal pour le développement économique et social du Ruanda-Urundi (Brussels, 1951).

⁸⁴ In this regard, Belgian agrarian policies in colonial Rwanda foreshadowed the kind of "development" strategies analyzed by James Ferguson in The Anti-Politics Machine, in which local knowledge is excised in the name of technical expertise. A particularly telling critique of state-centered planning based on "high-modernist ideology" is found in James C. Scott, Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed (New Haven, Conn., 1998). For additional recent critiques of development discourses, see Frederick Cooper and Randall Packard, eds., International Development and the Social Sciences: Essays on the History and Politics of Knowledge (Berkeley, Calif., 1997); Escobar, Encountering Development; Jonathan Crush, ed., Power of Development (New York, 1995). On Rwanda, see Lemarchand, "World Bank in Rwanda."

⁸⁵ Leurquin notes that, when it was seen to be "in their proper interest," communities had often undertaken unpaid collective work projects, such as constructing irrigation networks ("over considerable distances, traversing ravines in hollowed-out logs"), building bridges (sometimes using bricks made by public labor), and draining swamps. Le niveau de vie, 74. It would appear, therefore, that it was the forced nature of the work (and the question of to whom benefits accrued), not the work itself, that was opposed.

⁸⁶ The classic statement is Garrett Hardin, "The Tragedy of the Commons," Science 162 (1968): 1243–48. However, for a devastating critique of such thinking applied to Africa, see Pauline Peters, Dividing the Commons: Politics, Policy and Culture in Botswana (Charlottesville, Va., 1994). For an introduction to a wider literature on the struggle for the commons, see preeminently E. P. Thompson, Whigs and Hunters (New York, 1975).

⁸⁷ For Tanzania, similar reactions are brought out in Feierman, Peasant Intellectuals. For southern Africa, see, for example, William Beinart, "The Politics of Colonial Conservation," Journal of Southern African Studies 15 (1989): 143–62; and William Beinart, "Soil Erosion, Conservationism and Ideas about Development: A Southern African Exploration, 1900–1960," Journal of Southern African Studies 11 (1984): 52–83; Ian Phimister, "Discourse and the Discipline of Historical Context: Conservationism and Ideas about Development in Southern Rhodesia, 1930–1950," Journal of Southern African Studies 12 (1986): 263–75; Richard Grove, "Scottish Missionaries, Evangelical Discourses and the Origins of Conservation Thinking in Southern Africa, 1820–1900," Journal of Southern African Studies 15 (1989): 163–87.

⁸⁸ These approaches were indicated in Bishikwabo and Newbury, "Recent Research."

⁸⁹ Gravel, Remera; C. Newbury, Cohesion of Oppression; D. Newbury, "Rwakayihura Famine"; de Lame, Une colline.

⁹⁰ Leurquin based his analyses on data collected from 1,240 households in central and southwest Rwanda and north-central Burundi over 12 months (1955–1956). Because of the difficulty of measuring "production," he focused his study on household budgets and food consumed: Leurquin, Le niveau de vie, 112, 131–34, 203–05, 209.

⁹¹ Leurquin, Le niveau de vie, 204.

⁹² Leurquin, Le niveau de vie, 75.

⁹³ Learthen Dorsey, "The Rwandan Colonial Economy, 1916–1941" (PhD dissertation, Michigan State University, 1983).

⁹⁴ For differential tax rates by region during the early colonial period, see C. Newbury, Cohesion of Oppression, 172–74. Leurquin, Le niveau de vie, 251, notes that under colonial requirements, many men's tasks had been reduced, but women's tasks had increased. This was general to colonial work patterns in Africa. On gender and production patterns elsewhere in Africa, see (from an extensive literature) Janet Bujra, "Urging Women to Redouble Their Efforts: Class, Gender, and Capitalist Transformation in Africa," in Claire Robertson and Iris Berger, eds., Women and Class in Africa (New York, 1986), 117–40; Jane Parpart, "Women and the State in Africa," in Donald Rothchild and Naomi Chazan, eds., The Precarious Balance (Boulder, Colo., 1988), 208–30; Jeanne Koopman, "Women and the Rural Economy: Past, Present, and Future," in Margaret Jean Hay and Sharon Stichter, eds., African Women South of the Sahara, 2d edn. (New York, 1995), 3–22; Iris Berger and E. Frances White, eds., Women in Sub-Saharan Africa: Restoring Women to History (Bloomington, Ind., 1999); Marcia Wright, "Technology, Marriage and Women's Work in the History of Maize Growers in Mazabuka, Zambia: A Reconnaissance," Journal of Southern African Studies 10 (1983): 71–85; Maud Shimwaayi Muntemba, "Women and Agricultural Change in the Railway Region of Zambia: Dispossession and Counterstrategies, 1930–1970," in Edna Bay, ed., Women and Work in Africa (Boulder, 1982), 83–103. See also the contributions to the symposium "Women, Family, State, and Economy in Africa," Bolanle Awe, Susan Geiger, and Nina Mba, eds., Signs 16 (1991).

⁹⁵ C. Newbury, Cohesion of Oppression, and de Lame, Une colline, 45–79, detail the long-term evolution of rural grievances underlying the eventual transformations of 1959–1962. These aspects are brought out in other works as well: Lemarchand, Rwanda and Burundi; Donat Murego, La révolution rwandaise (Louvain, 1975); Linden and Linden, Church and Revolution; Reyntjens, Pouvoir et droit.

⁹⁶ There is a rapidly growing field of environmental history in Africa. Among many works, see for West Africa: James Fairhead and Melissa Leach, Misreading the African Landscape (London, 1993). For Central Africa: Robert Harms, Games against Nature (Cambridge, 1987). For East Africa: Helge Kjekshus, Ecology Control and Economic Development in East African History (Berkeley, Calif., 1977); and James Giblin, The Politics of Environmental Control in Northeastern Tanzania, 1840–1940 (Philadelphia, 1992). For southern Africa: William Beinart and Peter Coates, Environment and History (London, 1995); Peters, Dividing the Commons; Elias Mandala, Work and Control in a Peasant Economy: A History of the Lower Tchiri Valley of Malawi (Madison, Wis., 1990); Terence Ranger, Voices from the Rocks: Nature, Culture and History from the Matopos Hills (Bloomington, Ind., 1999); and Megan Vaughan, The Story of an African Famine (Cambridge, 1987). For Africa more generally, see James C. McCann, Green Land, Brown Land, Black Land (Portsmouth, N.H., 1999). Outside Africa, important early works of this type include William Cronon, Changes in the Land: Indians, Colonists, and the Ecology of New England (New York, 1983); Donald Worster, Nature's Economy: A History of Ecological Ideas (Cambridge, 1977); Donald Worster, ed., The Ends of the Earth: Perspectives on Modern Environmental History (Cambridge, 1988); David Arnold, The Problem of Nature (Oxford, 1995); Richard Grove, Green Imperialism: Colonial Expansion, Tropical Island Edens, and the Origins of Environmentalism, 1600–1860 (Cambridge, 1995); Alfred W. Crosby, Ecological Imperialism: The Biological Expansion of Europe (Cambridge, 1986).

⁹⁷ Jennifer Olson, "Farmer Responses to Land Degradation in Gikongoro, Rwanda" (PhD dissertation, Michigan State University, 1994), 80. Although Olson's work was carried out in southern Rwanda, her conclusions are independently confirmed for northwest Rwanda: Catherine André and Jean-Philippe Platteau, "Land Relations under Unbearable Stress: Rwanda Caught in the Malthusian Trap," Journal of Economic Behavior and Organization 34 (1998): 1–47. Elsewhere in the region, similar conclusions emerge. For Bwisha, eastern Democratic Republic of Congo: James Fairhead, "Fields of Struggle: Towards a Social History of Farming Knowledge and Practice in a Bwisha Community, Kivu, Zaire" (PhD dissertation, University of London, SOAS, 1990). For Burundi: Barnabé Ndarishikanye, "Des identités ethnico-politiques forgées dans la violence," Canadian Journal of African Studies (in press); for Buhaya: Priscilla Reining, "Social Factors in Food Production in an East African Peasant Society: The Haya," in Peter McLoughlin, ed., African Food Production Systems (Baltimore, 1970), 41–91; and Marion Pratt, "Women Who Eat Men's Money: Ecology, Culture, Gender Relations, and the Fishing Economy on the Western Shore of Lake Victoria" (PhD dissertation, State University of New York, Binghamton, 1995).

⁹⁸ Cattle have been present in Rwanda for more than 2,000 years and cultivation for more than 1,500—long preceding the attributed migrations of contemporary social categories. David Schoenbrun, A Green Place, a Good Place: Agrarian Change, Gender, and Social Identity in the Great Lakes Region (Portsmouth, N.H., 1998); David Schoenbrun, "We Are What We Eat: Ancient Agriculture between the Great Lakes," Journal of African History 34 (1993): 1–31.

⁹⁹ Pottier, "Three's a Crowd"; Olson, "Farmer Responses," 125.

¹⁰⁰ Olson, "Farmer Responses," 141.

¹⁰¹ For some exceptions, see Vidal, "Economie de la société féodale"; Victor Silvestre, "Différentiations socio-économiques dans une société à vocation égalitaire: Masaka dans le paysannat de l'Icyanya," Cahiers d'études africaines 14 (1974): 104–69; Lydia Meschi, "Evolution des structures foncières au Rwanda: Le cas d'un lignage Hutu," Cahiers d'études africaines 14 (1974): 39–51; Freedman, "East African Peasants."

¹⁰² See, for example, République du Rwanda, Résultats de l'enquête nationale agricole 1984 (Kigali, 1984).

¹⁰³ Some examples include Bernard Itangishaka, "Pour la défense du revenu du paysan," Dialogue 130 (1988): 26–36; Ubonabenshi, "Femme rurale et développement du pays"; Emmanuel Ntezimana, "Principes essentiels et conditions préalables à la démocratie," Dialogue 144 (1990): 33–49; Godding, "Grands projets et développement communal"; Pottier, "Politics of Famine Prevention"; Pottier, "Taking Stock"; Joachim Voss, "L'amélioration de la culture du haricot sur la base d'un diagnostic des contraintes de production, des pratiques et des potentiels des agriculteurs," in Augustin Nkundabashaka and Joachim Voss, eds., Les projets de développement rural (Butare, 1987), 37–46. As editor of Kinyamateka (the preeminent newspaper of Rwanda), the late André Sibomana contributed in important ways to these discussions; a peasant syndicate, Ingabo, was also catalyzed by these concerns. For an analysis of the debate over rural policies in Rwanda during the 1980s, see C. Newbury, "Recent Debates." On Sibomana, see André Sibomana, Gardons espoir pour le Rwanda (Paris, 1997); Hervé Deguine, Rwanda: Enquête sur la mort d'André Sibomana (Paris, 1998).

¹⁰⁴ André Sibomana was particularly eloquent on such issues; see C. Newbury, "Recent Debates," 208–12; on the effects of external aid projects, see Godding, "Foreign Aid as an Obstacle to Development"; Godding, "Grands projets et développement communal"; Lemarchand, "World Bank in Rwanda"; A. Hanssen, Le désenchantement de la coopération: Enquête au pays des mille coopérants (Paris, 1989).

¹⁰⁵ Jean Rumiya, "Ruanda d'hier, Rwanda d'aujourd'hui," Vivant Univers 357 (May–June 1985): 8.

¹⁰⁶ Reyntjens, Pouvoir et droit, 515.

¹⁰⁷ Guichaoua, Destins paysans, 1: 187.

¹⁰⁸ Fernand Bézy, Rwanda, 1962–1989: Bilan socio-économique d'un régime (Louvain, 1990), 32.

¹⁰⁹ Bézy, Rwanda, 54–56.

¹¹⁰ De Lame, Une colline, esp. 297 and following.

¹¹¹ Vidal, "Questions sur le rôle des paysans," 339, also notes de Lame's attention to the economic diversity among peasants in Rwanda.

¹¹² On the character of increasing rural impoverishment elsewhere in Rwanda, see André and Platteau, "Land Relations under Unbearable Stress"; David Newbury, "Ecology and the Politics of Genocide: Rwanda 1994," Cultural Survival Quarterly 22 (1999): 32–36; Olson, "Farmer Responses," chap. 5.

¹¹³ De Lame, Une colline, 68.

¹¹⁴ Michelle Wagner, "All the Bourgmèstre's Men: Making Sense of Genocide in Rwanda," Africa Today 45 (1998): 25–36.

¹¹⁵ André and Platteau, "Land Relations under Unbearable Stress." For another recent case study, see François Migeotte, Une colline rwandaise à travers ses pratiques d'élévage (Tervuren, 1997).

¹¹⁶ Longman, "Christianity and Crisis in Rwanda"; Timothy Longman, "Genocide and Socio-Political Change: Massacres in Two Rwandan Villages," Issue 23 (1995): 18–21; Longman, "Empowering the Weak and Protecting the Powerful: The Contradictory Nature of Churches in Central Africa," African Studies Review 41 (1998): 49–72. Some of these issues regarding the roles of peasants in the genocide are discussed in Vidal, "Questions sur le rôle des paysans"; Danielle de Lame, "Ces annees-là, les communautés rurales," intro. to François Migeotte, Une colline rwandaise, 9–23; de Lame, "Le génocide rwandais et le vaste monde, les liens de sang," in Stefaan Marysse and Filip Reyntjens, eds., L'Afrique des Grands Lacs, Annuaire 1996–97 (Paris, 1997), 157–77. For further scholarly analyses of the genocide and its ramifications, see (from a vast literature) Alison Des Forges, "Leave None to Tell the Story": Genocide in Rwanda (New York, 1999); André Guichaoua, ed., Les crises politiques au Burundi et au Rwanda, 1993–1994, 2d edn. (Lille, 1995); René Lemarchand, "Genocide in the Great Lakes: Which Genocide? Whose Genocide?" African Studies Review 41 (1998): 3–16; Uvin, Aiding Violence; Gérard Prunier, The Rwanda Crisis: History of a Genocide, 2d edn. (New York, 1997); Villia Jefremovas, "Contested Identities: Power and the Fictions of Ethnicity, Ethnography, and History in Rwanda," Anthropologica 39 (1997): 91–104; Reyntjens, L'Afrique des Grands Lacs en crise; Filip Reyntjens, Trois jours qui ont fait basculer l'histoire (Brussels, 1995); Jean-Claude Willame, Aux sources de l'hécatombe rwandaise (Brussels, 1995).

¹¹⁷ In several publications, we have tried to suggest that one cannot understand the genocide without understanding the political economy of Rwanda over the 1980s and 1990s, especially the complex social processes in operation in the rural areas. Catharine Newbury, "Background to Genocide in Rwanda," Issue 23 (1995): 12–17; David Newbury, "Understanding Genocide," African Studies Review 41 (1998): 73–97; D. Newbury, "Ecology and the Politics of Genocide"; Catharine Newbury and David Newbury, "Death and Demos: Failed Democratization in Rwanda" (paper presented at the SSRC conference "Transitions in Africa: Violence and the Politics of Participation," Niamey, Niger, June 1996); Catharine Newbury and David Newbury, "A Catholic Mass in Kigali: Contested Views of Ethnicity and the Genocide in Rwanda," Canadian Journal of African Studies (in press).