This is particularly important for at least two reasons. First, Rwanda has proven a particularly troubling example of the interplay of political power and violence with collective memory—the concept that when we remember we do so as members of social groups, and that our understandings of the past legitimate social orders in the present (Connerton 1989; Halbwachs 1980). Scholars of Rwanda and the Great Lakes region of Africa more widely have specifically shown how people’s “mental maps of history” (C. Newbury 1998:7) or “mythico-histories” (Malkki 1995) order and reorder particular social and political categories in the region, and have created imagined communities of fear and hatred (Lemarchand 2009:57, 70). Evidence shows how Rwandan political leaders have used competing interpretations of Rwanda’s particularly contentious history as a central tool in solidifying, polarizing, and mobilizing group identities toward violent conflict in the latter half of the twentieth century, including the genocide.
Second, political elites in Rwanda have long controlled and centralized the production of history to justify their own rule, while obscuring their role in doing so (D. Newbury 2009; Vansina 2004). Historian Jan Vansina has argued that the royal court in Rwanda used “historical remembrance” as the “ultimate legitimation” of its rule as far back as 1780, and by the nineteenth and twentieth centuries the royal court was an “institution in charge of controlling the production of history and its representation … an institution of such a wide reach and such a degree of subtlety” that researchers and Rwandans alike became “caught in its cognitive glue” (2004:5, 90–95).6 A decade after the genocide, many outsiders and Rwandans similarly found themselves caught in the “cognitive glue” of the regime’s version of history (Des Forges 1995; Pottier 2002; Reyntjens 2005), while there was “ample evidence that the regime continue[d] to manipulate the historical record for the sake of an official memory” (Lemarchand 2009:105), providing “disinformation” about both the distant past and the period from 1990 through the present (Pottier 2002; Reyntjens 2009:57–58). Denaturalizing Rwanda’s dominant narrative is crucial, thus, to keep us attuned to how contemporary versions of the past are actively produced in discursive, embodied, and material ways, and how they legitimate particular forms of belonging, exclusion, rights, and access in the present.
Historical Context and Points of Debate
For hundreds of years, ancestors of the people who came to be called Hutu and Tutsi in Rwanda lived side by side. They spoke the same language (Kinyarwanda), shared the same traditional religion, participated in the same economic networks, and intermarried to varying degrees. Precolonial Rwanda is most simply understood as a monarchy, ruled by Tutsi kings and chiefs. Lineages and patron-client ties figured heavily in social and political organization.
The nature of social harmony, stratification, and power has been heavily debated, characterized by the Hutu-power narrative as feudal exploitation and by the postgenocide government as harmonious. Much of this debate centers around the characterization of ubuhake, a controversial form of cattle clientship. Ubuhake was a contract between individuals in which the patron (shebuja) gave one or two head of cattle to the client (umugaragu) in usufruct but maintained ownership of the cattle and assured his client of protection. The client had to help his patron whenever needed, and the relationship was hereditary (Maquet 1961:129–142; Vansina 2004:47). The institution had a powerful political role in incorporating Rwandans in a dense network of social ties, yet it was a network that resulted in a social integration based on inequality, in which Hutu had access to cattle and protection, while Tutsi maintained ultimate control over cattle, the symbol of political, economic, and social power (Maquet 1961; Reyntjens 1987:72–73). Anti-Tutsi propaganda in newspapers, radio, and schools from 1959 through 1994 cast ubuhake as a historically deep form of domination and exploitation of all Hutu by all Tutsi in the precolonial and colonial periods. Much scholarship has, however, now shown that while there was indeed deep inequality in precolonial Rwanda, precolonial kings were not simply autocrats who ruled as they pleased (Vansina 2004:66, 85), and patronclient institutions did not merely involve exploitation of subordinate Hutu (C. Newbury 1988:90; Vansina 2004:33).7
Similarly, the peopling of Rwanda—when the peoples referred to as Hutu, Tutsi, and Twa arrived in the territory now called Rwanda—has been long contested. The dominant view was that Tutsi arrived in the region centuries after Hutu, but the details and implications of this view remain contentious. Did it mean Tutsi were destined to rule, or that they were foreigners who should not be allowed to do so? Did it mean Hutu were autochthonous and therefore had rights to govern? There remains debate about the precise amount and origins of biological difference and what the differences mean for interpreting human migrations (Cavalli-Sforza et al. 1994; Excoffier et al. 1987; Mamdani 2001:43–50), but Vansina’s recent analysis suggests that what physical differences existed must go back millennia, not centuries, as posited in the migration histories (Vansina 2004:37–38, 198).
Current scholarship argues that the categories of Hutu and Tutsi were primarily socioeconomic, referring to elites or to control over wealth (particularly cattle) and power (C. Newbury 1988:11–12). As far back as the seventeenth century, the term Tutsi “referred mostly to a social class among herders, a political elite,” while the term Hutu applied to everyone else, including poor Tutsi and foreigners (Vansina 2004:37, 134–135).8 The boundaries of the ethnic categories had some flexibility, so people could move across statuses over generations as their families gained or lost wealth (D. Newbury 1998:84–86, 2009; Twagilimana 2003:55). Further, ethnicity was only one form of identity, and other forms, such as region, class, lineage (a corporate descent group), and clan (a social descent group), were often more significant (d’Hertefelt 1971; C. Newbury 1978:17, 1988; D. Newbury 1998:83, 2009). The term for ethnic group, ubwoko, also translates as clan. Each clan was made up of people from all three putative “ethnicities,” and clans changed over time, operating as strategic alliances between lineages (C. Newbury 1988:96, 1980b; Vansina 2004:34–35, 198).
Colonial rule arrived in Rwanda when the first German colonial officer arrived in 1897. Two decades of German rule influenced primarily the royal court without interfering substantially in internal affairs. When Belgians took over Rwanda during World War I, in 1916, they extended their influence deeper than that of the Germans, such that Belgian rule had more extensive social and economic impacts. The colonial state contributed to formalizing and legitimizing the Hutu-Tutsi distinction, reifying formerly fluid boundaries, in its efforts to make more legible the existing complex ethnic configurations (Lemarchand 2009:9; Scott 1998). As part of the standardization of the administration in the late 1920s, the Belgians introduced identity cards that marked ethnicity (Des Forges 1995; Twagilimana 2003:55). The reification of ethnic groups was linked to European ideas of biological race, and it took an essentialized and hierarchical view of Hutu and Tutsi, which contemporary scholarship has completely debunked. It built on the Hamitic Hypothesis, put forth by explorer John Hanning Speke in 1863. Speke proposed that a race of tall, sharp-featured people who had Caucasian origins and were superior to the native Negro had introduced the cultures and civilizations of Central Africa.9 Anthropologists converted Speke’s conjectures into scientific truths with regard to African peoples in ensuing years (e.g., Seligman 1930; Westermann 1949 [1934]).10
Through the twentieth century, scholars writing on Rwanda continued to emphasize that there were three separate groups in Rwanda (Hutu, Tutsi, and Twa), and continued to imply that their current form had existed relatively unchanged since Tutsi arrived in the tenth century (D. Newbury and C. Newbury 2000:836; Twagilimana 2003:53). For example, anthropologist Jacques Maquet described Rwandans’ prevailing stereotypes of each other in the 1950s as follows: “Tutsi were said to be intelligent (in the sense of astute in political intrigues), capable of command, refined, courageous, and cruel; Hutu, hardworking, not very clever, extrovert, irascible, unmannerly, obedient, physically strong” (Maquet 1961:164).11 Explanatory models privileging biology were perhaps particularly seductive and enduring in the absence of other clear markers of difference between the groups.12 Even as belief in the scientific validity of race began to weaken and scholars moved from biological toward other sociocultural means of identifying groups, Hutu and Tutsi did not fit easily into
