The opposition party ODM was born out of popular protests against Kibaki’s proposed constitutional reforms in 2005 that did little to limit executive power or reform “Kenya’s ‘top heavy’ political system.”3 Headed by a “pentagon” of five leaders with prominent Luo politician Raila Odinga at the helm, ODM hitched its campaign onto deeply historical arguments, telling histories of lost sovereignty, prophesizing interethnic cooperation, and envisioning a devolved, federated state.4 Their strategy was a regional one: each of the five leaders in the pentagon represented a different regional polity, and their platform promised a more equitable “sharing of the cake” across the country. Debates during the campaign highlighted the multiple social inequalities, historical grievances, and regional disparities in development that cut across simplified “tribal” allegiances.5 The two months of postelection violence left over one thousand people dead and seven hundred thousand internally displaced and ended with a compromised coalition government, with Kibaki continuing as president and Odinga instated in the newly created position of prime minister. Ethnic arguments certainly played a role in voting patterns and in the subsequent violence, but to label the two-month conflict that inaugurated 2008 as merely the cyclical bloodletting of “age-old tribal hatreds” was to misread not only the deep social anxieties and urgent political questions at stake in these elections but also the complex histories of the making and unmaking of political communities in eastern Africa.
Ahead of the 2009 census, newspapers, blogs, and other popular forums filled with debates over the nature of identity in Africa. In an article for the Standard, Otuma Ongalo pondered his census choices:
I’m a son of a Luo man from Nyanza and Mnyala woman from Kakamega who migrated from Ugunja and established a home among the predominantly Bukusu in Bungoma in early 1960s. . . . Part of my family born before mid 1960s speaks fluent Luo, Kinyala and Kibukusu. . . . I’m married to an Isukha woman but our children cannot speak Kisukha, Kinyala, Luo or Kibukusu. So, when enumerators turn up in our homestead they will find a nation, not a tribe.6
Mama Ida, wife of Prime Minister Raila Odinga, drew laughs when she responded to the census question with, “I’m half-Luo, half-Luhya—you know, 50–50!”7 With no provision for such mixed families, Ida Odinga was enumerated as a Luo, the community of her husband.8 One columnist jokingly pictured these plural families as products of a distinct moral economy: “exchanging goats for the tribe.”9 In western Kenya, as across the country, long histories of migration, intermarriage, and interethnic exchange complicated bureaucratic efforts to align people into neatly ordered columns of sanctified and unchanging tribal groupings. As Ongola’s response eloquently captured, the grammar of identity had a history and a geography that made sense of the plurality and multiplicity of identities in postcolonial Africa.
In response to petitions from “minority communities” in this minority nation, respondents for the 2009 census would be “at liberty to break away from the broader Luhya, Kalenjin, Mijikenda, Swahili or Kenyan Somali and identify with one or the other of the numerous sub-groupings under them.”10 The groups targeted in this process of disaggregation at first glance appeared quite diverse, and yet they shared a common history of ethnic imagination at odds with contemporary understandings of ethnogenesis. All extend across national and colonial borders; all have recent histories of confederal alliances among diverse communities; and all have been at the forefront of federalist debates and secessionist campaigns in postcolonial Kenya.11 Of the Mijikenda, a confederation of nine communities along Kenya’s coast, Justin Willis has argued that their modern articulation of ethnicity was “a history, and an identity, of recent origin; a truth whose ambiguity is constantly reflected in historical presentation.”12 It was this ambiguity, this plurality and dissent within territorially unbound ethnic bodies, that government officials in Kenya painted as potentially subversive and destructive to national stability.
The census enumerator never arrived at Perus’s door. Despite a projected success of 97 percent of households across the country, delays, errors, and confusion marred the census in the Western Province, returning only 79 percent of its recorded household data.13 Perus’s dilemma, though, revealed a deeper history of the multiplicity and ambiguities of belonging in postcolonial Africa.
In all these recent political developments in Kenya—the 2005 constitutional referendum, the 2007 elections, the 2009 census, and again in the 2013 elections—one question plagued officials and repeatedly grabbed headlines: how will the Luyia answer?14 With the second-highest voting numbers in the country, the Luyia represent a powerful regional power broker and crucial factor in any electoral contest. However, the political plurality of Luyia constituents consistently defied the tenets of political tribalism, never historically voting as a bloc. Despite representing the most hotly contested, diversely populated, and historically oppositional province during the 2007 Kenya elections, the province did not witness the levels of violence and retribution that spread around its borders. Even with one of their “own,” in the form of Musalia Mudavadi, vying for the presidential seat in 2013, Luyia voters continued their tradition of defiantly plural politics, splitting the vote yet again between multiple candidates and parties.15
The political landscape of western Kenya is “slippery” terrain: “just as the Luhya’s favourite food, ‘mrere’ is slippery, so too is anyone that banks on the Luhya to vote in bulk for them.”16 Many have asked whether they were indeed a proper “tribe”: in the words of one Kenyan columnist, the Luyia were “created by the colonial administration some time in the 1940s. They did not exist before then and have no history as a ‘tribe.’”17 And this questioning of Luyia “credentials” as a tribe extends far beyond Kenya’s popular and political discourses: mentors, advisers, and colleagues at academic conferences and in casual conversations have similarly asked me, “but are the Luyia really a tribe?”18
These questions formed the original stimulus for this book, asking how and to what ends the Luyia community developed this seemingly plural and unpredictable ethnic identity. Indeed, the “Luyia” did not exist as a discrete ethnic appellation before the 1930s; precolonially, and well into the colonial era, they were instead multiple discrete and distinct political communities that defied ethnic categorization and crossed environmental, linguistic, and colonial boundaries. Nevertheless, in 1948, after four decades of British rule in Kenya, a previously unknown ethnic name suddenly appeared atop the first official census. From nonexistence, the “Luyia” appeared with 653,774 enumerated and named constituents. By the 2009 census, Luyia numbers had risen to 5.3 million, the second-largest ethnic affiliation in the country. Despite their recent and self-conscious history of ethnogenesis, Luyia elders interviewed throughout my research described their community within the idiom of “tribe”: to be Luyia was to be “of the same blood,” to “gather together,” to “speak the same language.”19
This story, then, begins in this slippery terrain, in the undulating landscape of what would become western Kenya. From the shores of Lake Victoria to the foothills of Mount Elgon, the immense ecological and topographical variety of this compact region invited a complex mix of African settlers from divergent migratory routes and linguistic backgrounds. While European geographers and administrators mapped a singular, neat territory to order this complex landscape, its diverse inhabitants proved resistant to would-be state builders. With the discovery of gold in their lands in the 1930s, a territorial crisis prompted local political thinkers to imagine, for the first time, an enlarged ethnic polity in western Kenya.
The threats to land and local