Whatever its basis in a specific urban geography, the antagonism between Indian and African became generalized through the work of discourse. Outside of structures of representation, racial groups do not exist as collective agents: individuals and factions claim to embody racial or national totalities as a strategy for mobilizing sections of populations. Consequently, racial conflict can coexist—indeed, it always coexists—with social formations and relationships organized according to other categories and logics. If an Indian-owned infrastructure instantiated the hierarchy of “brown over black,” this hierarchy was haphazard and rested on local foundations whose dynamics varied considerably.9 Alongside depersonalized (and depersonalizing) interactions in buses and shops, more familiar and sometimes intimate relationships developed: between doctors and patients, landlords and tenants, friends, neighbors, coworkers, and lovers. When younger activists from the ANC and Indian Congress began to grapple with the question of African-Indian political cooperation in the mid-1940s, they confronted a polarized terrain of distinct organizations and identities, which they knew from their own lives did not express the complexity of race in Natal. As later chapters will show, this dilemma would lead them to rethink the relationship between race, political organization, and African nationalism—but only after considerable suffering and loss.
STEREOTYPES, DISCOURSES, AND DESCRIPTIONS
In sources relating to the question of African-Indian social dynamics (newspapers, memoirs, interviews, and government reports), three distinct forms of racial language appear: stereotypes, generalizing discourses, and specific descriptions. All three of these modes are grounded in a racial consciousness that, however much it might incorporate an individual or group’s lived experiences, translates social relations into idealized images and, in turn, shapes the field of personal interactions in racial terms. Because these types of language presuppose the existence of coherent groups (“the Indian” and “the African”), they should not be used as direct evidence of an underlying reality: racial language organizes and shapes social life in ways that require historical analysis and explanation. As Sander Gilman argues, stereotypes operate at the level of fantasy.10 They are images that have broken free from their original context through assimilation into collective identities.11 Because they are a means of self-definition in terms of an “outsider,” they can function in contexts where there is no empirical correspondence between the stereotype and its putative object. A stereotype generally takes the form of a core association—for example, “white South Africans are racist”—accompanied by a cluster of related attributes that are invoked selectively and shift overtime. In most cases, multiple stereotypes exist regarding the same group, allowing racial fantasies to shift between differing, and often contradictory, characterizations without any underlying logic. For a historian, stereotypes serve largely as evidence of the imaginations, desires, and resentments of the individual or group deploying the image, rather than the actual content of racialized interactions.12
In contrast, racial discourses are the product of specific socioeconomic contexts. They incorporate stereotypes into narratives that correspond, in however partial and distorted a fashion, with existing social realities. Racial discourses are common stories told about the relationship between groups, although one protagonist (usually the group of the narrator) can remain implicit and therefore invisible in the account. Such stories often relate to specific kinds of spaces—the everyday theaters where the drama of interaction unfolds—or geographic areas. They provide relatively fixed scripts that influence the expectations and behavior of individuals within these sites while molding broader perceptions through their circulation as rumors, jokes, and urban legends.13 Since racial discourses are organized around the existence of an imagined binary, their basic structure assumes one of two forms: cooperation or conflict, friendship or war. In a very real sense, they form part of the material infrastructure that reproduces social interactions in a racialized form. In turn, the authors of specific descriptions seek to confirm or contest an established racial discourse. These sources recount events (an unusual or exemplary encounter, for instance) in the concrete language of personal experience. Descriptions often contain elements or details that subvert the generalizing terms of racial discourse while employing the same stereotypes. In such cases, specific descriptions do not necessarily provide evidence of a competing discourse. Nor should they serve as the basis for alternative sociological generalization regarding “race relations.” They relate events, relationships, or interactions that confirm or disrupt a standard script even as the author continues to perceive the participants and overall context in racial terms.
The most striking aspect of stereotypes regarding Indians was their uniformity. The same basic image appears across much of southern and eastern Africa, despite the varying character of the colonial policies and racial dynamics of different states, regions, and cities. Moreover, the core attributes of this image remained generally stable even as concrete social relationships and their underlying political economies transformed significantly. The “Indian” was synonymous with the “merchant.” The principle traits of this figure expressed popular resentments over the reputed practices of shop owners: “dishonest,” “crafty,” and “exploitative” were common epithets. Correspondingly, this language presupposed an undifferentiated victim of the Indian’s machinations, the African, who assumed the opposite characteristics. These were the common, even universal, terms of struggle between moral economies based on different forms of wealth: the honor-based ethos of an agrarian society and a mercantile diaspora’s profit-driven reckoning of value.14 The Indian was not just ethnically foreign, but embodied the increasing power of an alien mode of calculating and distributing wealth.15 One informant told Leo Kuper in the 1950s: “We cannot compete with Indians in business. Far from it. I don’t think we’ll ever pitch up to their understanding. Where merchants work it out for us, it’s alright.”16 As a figure, the merchant gave phenomenological immediacy to market forces that were capricious and otherwise invisible.
Whatever its origins, once this association was fixed as a stereotype, an enormous range of human behavior—from the thrift of working-class housewives to the international diplomacy of Krishna Menon and Nehru—became legible in terms of a common racial essence. The resulting image lent itself to paranoid and conspiratorial readings of social relationships. If Indians pursued their economic interests through duplicitous means, then acts of friendship, altruism, or solidarity (the distinction scarcely registered) masked their true intentions and therefore, paradoxically, were the most “Indian” forms of deception. Later, the same logic would transform the Indian into a highly visible embodiment of postcolonial (and postapartheid) corruption.17 By projecting threatening qualities onto the other, the merchant stereotype neutralized the internal conflicts and heterogeneities of urbanizing African society, and therefore provided an alibi for the divisions that troubled political, religious, or cultural nationalist claims to represent a coherent moral community. If we somehow resemble the corruption of the outsider, according to this argument, it is due to the outsider’s corrupting influence among us. Even proximity (spatial as well as social) could be recast as a form of infiltration, and therefore served as a confirmation of the Indian’s devious nature. This reiteration of distance reflected the most important attribute of the merchant stereotype. No matter how long his or her family had lived in South Africa, the Indian would always remain a foreigner. In popular culture, this image often appeared alongside signifiers of cultural difference: the smell of curry, the sari of the woman shopkeeper, and the intonations of South African Indian English. Several observers of mid-century Durban noted that these associations echoed core tropes of anti-Semitism. The Indian was, according to a common saying, the Jew of Africa.18
Because of its near ubiquitous presence in interviews and sources written by Africans, the merchant stereotype produces something of a false surface. Circulating between editorials and racist jokes, between political speeches and township gossip, this image encouraged many contemporaries (and some later accounts) to postulate a general hostility to Indians and a bifurcated social landscape.19 Underneath this surface, the reality was considerably more complex. It is important to remember that racial language is, after all, a form of language: a contextually specific act of speech or writing directed toward an audience. The same words could mask opposed attitudes and intentions. In the years following the Durban Riots, H. I. E. Dhlomo and Ngubane wrote a series of articles analyzing the complexity of African-Indian