Michael R. Mahoney
In the history of southern Africa, as elsewhere, ethnicity has often been a source of pride and a feeling of belonging, but it has also sometimes been a source of shame and of conflict, ranging from minor disagreements to massacres. Ethnicity has been a factor throughout the history of the region, from the Zulu Kingdom (for example) including and excluding the different peoples it conquered, to white conquerors manipulating ethnicity often quite explicitly and self‐consciously for the purpose of dividing and ruling their African subjects, to postcolonial Africans struggling to find a place for both ethnic self‐assertion and national unity. Since ethnicity is a universal phenomenon, it figures in the history of every southern African country: Botswana, Lesotho, Mozambique, Namibia, South Africa, Swaziland, and Zimbabwe. Ethnicity does not matter only for Africans: it also matters for the region’s residents who are of European, Asian, or mixed descent. Ethnicity overlaps with race, kinship, political affiliation, and a whole host of other social phenomena. It is, at bottom, political, even though it involves some of the most basic aspects of any person’s sense of self, which transcend political differences.
For the purposes of this chapter I shall be defining ethnicity broadly as a sense of identity and belonging to a named group of people based on a belief in that group’s common ancestry, culture, and language. This ethnic identity may be asserted by individuals themselves, or it may be assigned to them by others, or both. Ethnicity is just one possible form of identification (others include gender, class, race, religion, ability, political affiliation, and so on) that any individual might claim at any time, and many people may even assert multiple ethnic identities (Cornell and Hartmann 2007). Researchers who study the phenomenon should be careful not to reify these abstract concepts and should not assume that any identity is a concrete thing that exists beyond acts of identification (Brubaker and Cooper 2000). Ethnicity is usually what people are talking about when they refer to “tribalism” in Africa. Though the terms “tribe” and “tribalism” are commonly used throughout the world, including in Africa, when talking about Africa, I shall follow scholarly convention here by using “ethnicity” instead. This is mainly to avoid the implication that ethnicity in Africa is fundamentally different from ethnicity in Europe or any other part of the world. Above all, whether we are talking about ethnicity or tribalism, the phenomenon itself is not enough to explain anything and must itself be explained (Lowe 1997).
Given the enormity of this topic and of the scholarly literature that has examined it, this chapter will of necessity be selective rather than exhaustive in its approach, and idiosyncratic to boot. I will make many arguments here, but the two main ones are, first, that ethnicity has been a factor throughout southern African history and not just since European contact, and, second, that ethnic groups in southern Africa and elsewhere have always had conflicting tendencies toward both inclusion and exclusion.
The most important scholarly work on this subject to date has been the 1989 collection of essays entitled The Creation of Tribalism in Southern Africa (Vail 1989). Edited by Leroy Vail, the late renowned historian of Mozambique, the book dealt head‐on with a subject that was either taken for granted (and thus left unexamined) by the media, by government, and by society in general, or generally dismissed as a distraction by academics, government officials, and activists more concerned with national or class liberation or nation building.
The contributors to the book were building on the pioneering essay by the Zimbabwean historian Terence Ranger, “The Invention of Tradition in Colonial Africa,” which argued that during the colonial period in Africa traditions were radically reformed, and sometimes even invented outright, by white officials and missionaries and by African chiefs and elders and Christian converts. These changes tended to come at the expense of the colonized in general, and especially chiefs’ subjects, women, and youth (Ranger 1983).
In the introduction to The Creation of Tribalism, Vail synthesized the various contributors’ arguments to come up with a model of the creation of tribalism very similar to Ranger’s. European colonizers, and later white‐dominated governments in settler regimes such as South Africa and Rhodesia, insisted that Africans were above all “tribal” and did everything in their power to emphasize the ethnic differences between themselves. European missionaries contributed to this effort through the educational systems and academic research, which they dominated, including the codification of African languages in dictionaries and grammars. Much of this work was also done by African Christians, who wanted to improve their legitimacy in the eyes of their non‐Christian African neighbors. African chiefs cooperated in order to shore up their authority in the face of the upheavals of colonization and urbanization, as did African elders and men in general, for the same reasons. Even some elder women promoted a vision of ethnicity that was disadvantageous to them as women but accorded higher status to them as elders (Vail 1989).
Ranger and the many contributors to various aspects of African tradition, including ethnicity, were not timeless, but rather were created by colonial officials, missionaries, African “traditional” elites, and “modernizing” African Christians during the colonial era. As a result, the contributors to Vail (1989) tended to favor an instrumentalist interpretation of ethnicity. Instrumentalists argue that ethnic attachments are constantly being created and recreated and manipulated to serve particular political ends. For decades, a major debate in the study of ethnicity has been that between instrumentalism and primordialism. Primordialism is the belief that ethnic (or nationalist) attachments are very old and fundamental, existing in our very genes and the languages we learn as children, not to mention the culture that infuses our daily lives. Ethnic attachments are thus irrational and apolitical in origin. As many scholars have noted, instrumentalism and primordialism are not necessarily mutually exclusive: primordial attachments create a potential for instrumentalist ethnic mobilization. However, the danger, as far as instrumentalists are concerned, is that a focus on primordialism can obscure the instrumental aspects of politicized ethnicity. Conversely, to focus exclusively on instrumentalism risks reducing ethnic attachment to its political implications, obscuring the reasons for the success of ethnic appeals, and blinding us to the contentions within ethnic groups, not to mention implying that ethnic attachments are necessarily bad, leaving us to wonder why ethnic appeals resonate in the first place (Sisk 1996).
The most successful effort in African studies to transcend some of the limitations of both primordialism and instrumentalism has been the historian John Lonsdale’s theory of political tribalism and moral ethnicity. For Lonsdale, moral ethnicity is the arena of contention defined by primordial ethnic attachments. It determines who may legitimately participate in political disputes within an ethnic group, as well as how those disputes are to be pursued. While moral ethnicity may set a limit on the possible outcomes of political disputes, it does not completely determine those outcomes. Political tribalism is any rhetoric that equates belonging to an ethnic group with support of particular political points of view. If you do not support a certain point of view, then it follows that the legitimacy of your belonging to the group is called into question. The relationship between ethnicity and politics can sometimes be characterized by moral ethnicity, while at other times it may be characterized by political tribalism (Lonsdale 1992). More recently, Lonsdale and his fellow East African historian Derek Peterson have developed the notion of “ethnic patriotism,” the definition of which largely overlaps with that of “political tribalism,” but it better conveys the ethnic pride that is usually used to legitimate certain kinds of ethnopolitical appeals (Lonsdale 2009; Peterson 2012).
One major criticism of Ranger’s and Vail’s approach has been that their research implies that these processes of invention or creation only took place during the colonial period when ethnicity and other aspects of African tradition had long histories of change even before European colonizers arrived on the scene (Atkinson 1994; Guyer 1996). With very few exceptions, such as the Qamu chiefdom of colonial Natal (Lambert 1995) and the Kavirondo “tribe” of colonial Kenya (Southall 1970), most ethnic groups of the colonial and postcolonial eras descend from groups of the same name that existed in precolonial times.
However, though the names of the groups have deep histories, it is less certain that those groups have always been ethnic groups. Some scholars, such as Phillip Bonner, John Wright, and Carolyn Hamilton, have argued that the groups existing before the wave of political consolidation that hit southern Africa in the late eighteenth and early nineteenth centuries were kin or lineage groupings. Others, such as W. D. Hammond‐Tooke, have countered that these were primarily political groupings (Hammond‐Tooke 1991). Lineages (kin groupings) and chiefdoms (political groupings) seem to have been distinct even in oral traditions that speak of the earliest strata of precolonial history, though the distinction might be blurred by the fact that the leadership of a chiefdom was identified by the lineage to which they belonged. Bonner, Wright, and Hamilton are right to emphasize how often people of different lineages from their chiefs developed fictive kinship claims to argue that they were indeed related. However, oral traditions offer numerous examples of people belonging to one lineage but pledging allegiance (Zulu khonza) to a chiefdom of another lineage without making fictive kinship claims, and the terms for “lineage” (Zulu uhlanga) and “chiefdom” (Zulu ubukhosi) are different in most Southern Bantu languages.
Regardless of whether chiefdoms were lineage groupings or political groupings, they were not ethnic groups per se. This is not, however, to say that they had no ethnic content whatsoever. The sense of common ancestry that is characteristic of lineages, and of real and fictive kinship claims of chiefly families and at least some of their subjects, is also one of the defining characteristics of ethnicity. When it comes to cultural and linguistic ties, the Southern Bantu languages have a prefix (the cognates se‐ in the Sotho‐Tswana languages and isi‐ in the Nguni languages) that refers to the culture and/or language of a people. That this prefix was often attached to the names of precolonial polities suggests that those polities were also culturally and linguistically distinctive.
The relationship between lineage groupings and ethnic groupings has long been gendered. Vail (1989: 15) quotes a Tswana proverb, “women have no tribe,” to show that ethnicity has largely been a men’s affair throughout African history, for various reasons. Most southern African societies are exogamous, patrilineal, and patrilocal, meaning that women tend to be outsiders in the communities in which they live. Men have also tended to monopolize long‐distance travel, putting themselves more often in situations where ethnic distinctions become salient. Elizabeth MacGonagle’s (2007) study of the Ndau, who straddle the border between Zimbabwe and Mozambique, challenges this depiction somewhat by showing how active women were in the promotion of ethnic identity through the cultural education of children and through their roles as healers, spirit mediums, and occasionally chiefs. At the same time, when the Ndau were ruled for a time by the Gaza Nguni, men tended to assimilate to the Nguni identity, while women promoted the Ndau identity (MacGonagle 2007). From the late eighteenth to the early nineteenth centuries, in Portuguese‐ruled Mozambique, the Chikunda ethnicity emerged among men in the exclusively male warrior slave class created by the Portuguese (Isaacman and Isaacman 2004). In the recent past of the Tonga, further south on the border between South Africa and Mozambique, migrant labor has caused men to tend to assimilate to the neighboring Zulu identity, while women have tended to resist this and instead cling to their old Tonga identity (Webster 1991). By contrast, in Manyika society, also on the border between Zimbabwe and Mozambique, lineage membership gave more power to men (though it gave some power to women as well), while power was more evenly distributed between men and women in the broader realm of ethnicity (Schmidt 2015).
The muddled relationship between ethnicity, lineage, and political affiliation has meant that even recent scholarship on ethnicity in precolonial southern Africa has occasionally reached contradictory conclusions about the very existence of ethnicity during this early era. MacGonagle’s (2007) study of the Ndau shows quite clearly that the process of social construction of ethnicity that Ranger and Vail and others described so effectively for the colonial period was also going on during the precolonial era. For example, the Ndau language is a classic example of the aphorism related to the linguist Max Weinreich by an anonymous interlocutor: “A language is a dialect with an army and navy” (Bright 1997: 469). The line between language and dialect is very blurry and has more to do with social and political realities than with linguistic ones. Is Ndau an independent language in its own right or merely a dialect of Shona? Weinreich’s aphorism might be given an African gloss: “A language is a dialect with its own king.” MacGonagle shows how Ndau ideas of their own linguistic and cultural unity and distinctiveness from the (other) Shona go back to the precolonial era. The same goes for the manipulation of kinship identity, which united various clans under the royal lineage and downplayed links with lineages beyond the Ndau kingdom. Both the linguistic and kin elements of the distinctive Ndau ethnic identity were maintained by the Ndau during periods when they were ruled by outsiders, such as the Mutapa and Gaza Nguni states (MacGonagle 2007).
Paul Landau’s (2010) similarly longue durée study of South Africa, conversely, argues that precolonial polities in the region were not ethnic groups, or even kin groups for that matter. Whereas “ethnic group” and “tribe” tend to connote a certain exclusivity, Landau goes to great lengths to show how inclusive and diverse these polities actually were. This is most obvious when one examines the period of the Mfecane, or Difaqane, the southern African upheavals of the early 1800s. Frequent warfare killed off the ruling elites and created streams of refugees. Power came through people, and chiefs could not afford to be too particular about whom they brought into the fold. The first Europeans to visit southeastern Africa and the interior highlands came at this time, providing rich descriptions of African societies in states of severe flux. However, through detailed analysis of these sources, oral traditions, archaeology, and historical linguistics, Landau shows that inclusiveness was not a new policy for a desperate situation. Like other scholars such as David Beach and Norman Etherington, Landau sees a long history of cycles of political centralization and fragmentation in the region. Unlike most other scholars, however, Landau does not draw a contrast between large multiethnic states and small, stable, homogeneous chiefdoms: it was openness all the way down (Landau 2010).
Of course, this flexibility cut both ways. Landau emphasizes the ease with which people could enter other polities; recent work by Garzarelli and Keeton (2014) shows that it was likewise easy (though by no means cost‐free) for precolonial southern Africans to leave the polities to which they belonged, often to go and found new ones. It is perhaps telling that Landau focuses almost entirely on entrances but discusses exits hardly at all, even though each entrance into one polity must necessarily involve an exit from another one. Voluntary or involuntary exits from polities are just one example of acts of differentiation. Some other examples from precolonial southern Africa come from the Zulu Kingdom. While the Zulu elite were inclusive toward some of the people that they conquered or otherwise absorbed, they excluded and denigrated others among their subjects, calling them the Lala (Hamilton and Wright 1990). Even some of those who were offered inclusion by the Zulu, such as the Qwabe, rejected it for the most part (Mahoney 2012).
Sociolinguistics offers some ways to conceptualize the coexistence of inclusion and differentiation when it comes to identity, whether that identity be political, kin‐based, or ethnic. For example, Heinz Kloss (1967) developed the notion of abstand (distancing) and ausbau (expanding) languages. Linguistic abstand involves differentiating a linguistic variety from others, often consciously. Ausbau, on the other hand, is linguistic inclusion, emphasizing commonalities between different but closely related linguistic varieties. Despite the positive connotations of “inclusion” here, for Kloss ausbau usually involved the transforming of what might be considered an independent language into a mere dialect of a dominant standard language, as happened with Provençal and French, Allemannisch and German, and Sardinian and Italian, to name just a few examples. Similarly, Robert Le Page and Andrée Tabouret‐Keller (1985) argued that all linguistic utterances could be seen as “acts of identity.” Depending on how one speaks, what language or linguistic register somebody uses, one could be either emphasizing their commonality with another person or emphasizing their distinctiveness. Indeed, the same linguistic act of identity could (and usually does) at the same time assert commonality with one group of people and difference from another.
These sociolinguistic approaches have many implications for the study of precolonial southern Africa. First, they show that inclusion and exclusion can, and indeed must, coexist. Scholars must be on the lookout for both and examine their interplay. Second, the act of identification, whether assigned or asserted, is done through culture, and language above all. The almost universal multilingualism and cultural hybridity that has characterized southern Africa apparently throughout its history should not lead us to ignore the fact that at certain moments people engage in what Le Page and Tabouret‐Keller call “focusing” and Kloss calls “abstand”: choosing one particular form of cultural expression instead of others for the purpose of saying, in effect, “I am this and not that.” Without this “focusing,” cultural differences would never emerge, but history has shown that they have emerged frequently, even in the precolonial past.
Similar arguments may help us interpret the ambiguous evidence of archaeology and historical linguistics, so crucial for understanding the earliest history of the region. For example, if languages are related to one another, it follows that they descended from a common ancestor language whose speakers differentiated over time. The closer the linguistic relationship the more recent the divergence; the more distant the linguistic relationship the longer the passage of time since the divergence. By measuring degrees of linguistic relatedness, historical linguists have been able to come up with family trees of African languages and to suggest original homelands and patterns of migration and interaction (Ehret 2000). Similarly, human artifacts tend to come in certain styles that co‐occur: certain ways of designing pots may be common in a certain area for a certain period of time and may coincide with certain ways of designing houses, and so on. Archaeologists refer to these co‐occurring styles of artifacts as assemblages and, more broadly, as cultures or traditions. Do these “cultures” or “traditions” represent ethnic groups? Can a historical linguist’s “proto‐languages” be linked with an archaeologist’s “cultures”? Can they both be tied to specific ethnic groups existing today, or at least to the ancestors of those ethnic groups?
Few debates about early history are as contentious as the one regarding the relationship between linguistics, archaeology, and ethnicity. W. D. Hammond‐Tooke, for example, makes several arguments against scholars identifying archaeological “cultures” with specific ethnic groups: (1) “ethnicity” as a concept is too vague; (2) it requires knowledge of social context, which archaeologists too often lack; (3) it can exist only in multiethnic yet politically and socially integrated environments; (4) material artifacts do not necessarily reflect ethnic self‐identification; (5) uniformity in material culture is as likely to be due to cross‐cultural borrowing as to ethnic homogeneity; and (6) distinctiveness in material culture was not necessary for drawing social boundaries when population densities were low and great spaces separated different social groups (Hammond‐Tooke 2000). While archaeologist Sian Jones’s treatment of the issue does not deal specifically with southern Africa, she makes several arguments that effectively counter Hammond‐Tooke’s. She is aware of how archaeology has been used and abused for political purposes by different ethnic and national groups today, but she cautions against seeing archaeological research into the history of ethnicity as an impossible task. Ethnic identification, she argues, is a human universal and likely emerged at the same time as cultural distinctiveness in material culture. Archaeologists have tools to distinguish between interethnic borrowing and intraethnic commonalities, and even the most extensive borrowing cannot erase ethnic difference. Perhaps most importantly, archaeologists now have a view of ethnicity as constantly evolving, subject to internal dispute, and inclusive as well as exclusive – a more flexible view that fits better with the archaeological evidence (Jones 1997). Roger Blench makes similar arguments about linking together linguistics, archaeology, and the study of ethnicity in Africa in particular. The task is difficult, it must be approached with a great appreciation for nuance, and few debates can be decisively settled, but this kind of research is nevertheless possible and even necessary (Blench 2006).
The fact that ethnicity has a deep precolonial history in southern Africa does not mean that Ranger, Vail, and, more recently, Landau are wrong to see the colonial period as the time when “the creation of tribalism” took place. Responding to critics who accused him of ignoring the precolonial era, Ranger conceded that “the invention of tradition” also took place in the precolonial period, but argued that the colonial period saw much more profound transformation over a short period of time, creating the situation that we are living with today, even in postcolonial Africa (Ranger 1993). Like Ranger, Landau argues that tradition and tribalism since European conquest have become more rigid and exclusive than they were before, exacerbating conflict in southern Africa (Landau 2010). Part of this is because Europeans came to nineteenth‐ and twentieth‐century southern Africa with their own racialized notions of ethnicity: “tribe” was something innate, essentially distinctive, and unchanging. One was either this or that, forever. However, another reason why Europeans promoted rigid and exclusive tribal identity was because they quite self‐consciously sought to divide and rule Africans and to legitimate racial segregation and political, legal, and economic subordination on the basis of Africans’ supposedly greater “traditionalism” (Welsh 1971; Mamdani 1996). Chiefs, elders (including elder women), and men in general bought into this because it reinforced their own power based on supposedly unchanging tradition. Christian Africans likewise accepted “tribalism” because it emphasized their links with their non‐Christian fellow Africans.
Given the long history of ethnic conflict in Africa and the way in which it undermined struggles against white rule and, later, contributed to political instability and bloodshed in postcolonial Africa, Ranger, Vail, Landau, and others could be forgiven for focusing on the divisive aspects of ethnicity. However, ethnicity has just as often involved coalition building across even narrower categories of identification. Already in the 1950s, anthropologist James Clyde Mitchell was showing how urban “tribalism” brought together different groups that often feuded in the rural areas from which they came. There were feuds in the urban areas too, but the feuding groups, the urban tribes, were larger (Mitchell 1956). More recently, Daniel Posner, like Mitchell a specialist in Zambia, is only one of the most prominent proponents of a coalition‐building view of ethnic mobilization in Africa and elsewhere: ethnic coalition building, Posner and others have argued, is easier than other types when potential coalition members have similar languages and cultures (Posner 2005).
The formation of ethnic groups did not just unify Africans as much as it divided them; it also sometimes unified them in ways that Africa’s white rulers did not want. For example, in colonial Natal (1843–1910), the creation of tribalism involved colonizers promoting Africans’ identification with the dozens of chiefdoms in the colony, not with the larger Zulu identity that Natal Africans shared as a legacy of the precolonial past. The colonizers’ strategy succeeded for most of the colonial period, mainly because the Zulu Kingdom had created many enemies among the peoples it had conquered. Ironically, the Europeans’ ouster of the Zulu king in 1879 actually served to promote the spread of Zulu ethnic identification. As anthropologist Max Gluckman noted in 1940, “this opposition [between Africans and whites] has heightened allegiance to the chiefs, and especially the Zulu kingship. The sentiment about the king grows, helped by his lack of power, for he has no power to abuse” (Gluckman 1940: 44). The delayed impact of the worst consequences of colonization left Natal Africans seeking someone to save them, and the superpowered Zulu king of their imaginations was that person. Young male migrant workers, with their experience of wider ethnic categories in Johannesburg, promoted this new Zulu ethnic identity upon their return. Chiefs accepted this in the face of pressure from their subjects. Zulu ethnic identification was one of the foundations of the Zulu rebellion in 1906. The rebellion was crushed, but Zulu ethnicity was firmly established as a symbol of African unity and anticolonial resistance (Mahoney 2012).
The history of Zulu ethnicity after 1906 fits better with the narrative established by Ranger, Vail, and Landau, demonstrating that the social and political implications of ethnic identification can change over time. While leading Zulu politicians at first demonstrated a solid commitment to African unity through their involvement in the African National Congress (ANC) (from 1912) and the Industrial and Commercial Workers Union (from 1919), in the 1920s and 1930s the Zulu sections of these organizations broke away and pursued more localized and conservative agendas. These Zulu leaders, often prominent businessmen themselves, increasingly turned against African radicalism and cooperated instead with chiefs, the Zulu king, and white missionaries and government officials. This tendency culminated in the career of Mangosuthu Gatsha Buthelezi, a Zulu chief who participated in the widely condemned apartheid Bantustan system as the head of government of KwaZulu homeland. Buthelezi also founded what became the Inkatha Freedom Party, a movement opposed to the radicalism of the ANC, leading to an ANC–Inkatha civil war in the 1980s and 1990s which killed thousands on both sides, with Inkatha enjoying covert support from the apartheid government (Marks 1986; Maré and Hamilton 1987). On a more local level, ordinary Zulu men and elder women with no status except as men and/or elders increasingly embraced Zulu ethnicity and social and political conservatism as a way to preserve that status (Marks 1989; Hassim 1993; La Hausse de Lalouvière 2000).
In its broad outlines, the case of the Zulus parallels that of other African ethnic groups in South Africa and to some extent the whole region. This is not surprising considering the cross‐border ethnic and economic ties (especially participation in the migrant labor system) that are so common in southern Africa, as well as the fact that Zimbabwe, Botswana, Lesotho, and Swaziland were all British colonies, while Namibia was ruled for 75 years by South Africa and received many Afrikaner and South African British immigrants. Mozambique would be something of an exception. On the one hand, Mozambicans too were closely tied into the migrant labor system in South Africa and were very much influenced by both the top‐down and the bottom‐up promotion of ethnic identity that occurred in both Kimberley and Johannesburg (Harries 1994). On the other hand, Mozambique’s Portuguese rulers pursued a policy of assimilation that provided legal equality to a tiny minority of black Mozambicans while subjecting the vast majority to a homogeneous legal system called the indigenato, which created labor conditions bordering on slavery. Unlike the British, the Portuguese had little interest in the diversity of customary law, though they did recognize African chiefs and identified blacks primarily as the subjects of those chiefs. Also unlike the British, the Portuguese hardly relied at all on “tradition” and “tribalism” to maintain their rule, leaning quite heavily on force and coercion instead. As a result, while ethnic differences have had some salience in both colonial and postcolonial Mozambique, they have been less important than race, class, and regional differences in politics there (O’Laughlin 2000).
One of the main consequences of white rule in southern Africa was that it brought to the region two non‐African racial groups (whites and Asians) and created a third group (people of mixed race), with all three groups being themselves ethnically diverse. The largest and best‐known white ethnic group in Africa is the Afrikaners, and their history illustrates some of the principles of inclusion and exclusion I have discussed here. The ancestry of Afrikaners includes people of Dutch, French, German, Portuguese, British, African, and Asian ancestry. At the same time, Afrikaners have historically defined themselves in opposition to other ethnic groups, pretending that there were clear boundary lines when those lines were in fact quite blurred. The most obvious distinction Afrikaners have made has been between themselves and “nonwhites” of various races, but for most of the nineteenth and twentieth centuries the distinction between Afrikaners and white English‐speaking South Africans was just as salient and bitterly conflict‐ridden (Giliomee 2003), although some historians have cautioned against reading these distinctions back to the Great Trek (1834–1854) and before (Etherington 2001). Instead, Vivian Bickford‐Smith (1995) has identified late‐Victorian Cape Town as a particularly important crucible for the establishment of black, white, coloured, South African British, and Afrikaner identities in South Africa. Economic expansion and competition, growing social stratification, and imported “scientific” European ideas of race and nation all combined to fuel this process. Socially and economically dominant South African British whites established the hierarchy with themselves at the top. Afrikaner and coloured identities were shaped both by their subordination to the British and by their desire not to fall to the level of the blacks below. Even the blacks who found themselves at the bottom of the hierarchy embraced the racial identification that was the basis of their exclusion. In each case, though, Afrikaner, coloured, and black self‐assertion existed alongside assimilationist anglophilia in those groups (Bickford‐Smith 1995). Curiously, while much has been written about the histories of black, white, coloured, and Afrikaner identities, the history of English‐speaking white South African ethnic identity per se has received scant scholarly attention.
Indian history in South Africa has been particularly underserved by historians, but its lack of integration into broader histories of the region has been an even bigger problem. A steadily growing body of literature on the subject now includes a few surveys (Jain 1999; Dhupelia‐Mesthrie 2000). Like Africans, whites, and coloureds, Indians have formed a racial category in southern Africa despite great internal diversity. Class, caste, and above all religious diversity figure most prominently in the historiography, but Indians were also historically very ethnically diverse. While most of the “passenger Indian” elite were northerners, especially Hindi and Gujarati, the bulk of the indentured majority were southerners, especially Tamil. Because of this ethnic and linguistic diversity, South African Indians relied heavily on lingua francas to facilitate communication among themselves. Varieties of Hindi served this purpose at first, but are moribund today. In the late 1800s Natal Indians played a major role in the development of Fanakalo, a pidgin Zulu that was once widely used by people of all races throughout the region but is now largely confined to the mines and to interactions between white or Indian employers and their African employees (Mesthrie 1989). English has been the first language of most South African Indians for several decades now, and, despite the early dominance of Tamil and Telugu and the recent rise of schools that teach those languages, they are virtually extinct in South Africa (Mesthrie 2002). Like South African coloureds, Indians have both suffered and benefited from formal and informal racial orders that subordinated them to whites but privileged them above blacks. Few situations would be more conducive to the development of internal solidarity and a strong sense of distinctiveness in relation to their fellow South Africans.
White rule ended more recently in southern Africa than in any other part of the continent, mainly because the white population was at its largest and most entrenched there. Whereas most of Africa was independent by 1960, black majority rule in southern Africa was achieved only between 1966 (Botswana) and 1994 (South Africa). While southern Africa’s white rulers tried hard to promote African ethnic allegiances in order to undermine African nationalist movements, those nationalist struggles generally succeeded in creating greater black solidarity, and far less interethnic conflict, than in the rest of Africa. In other words, black–white conflict was a very effective vehicle for black unity. In southern Africa, only Mozambique has experienced a postcolonial civil war, and only Lesotho has had a successful military coup d’état. This is not, however, to say that there has been no ethnic conflict in southern Africa, just that it has been much more muted than elsewhere. Typically, at the national level racial or national identity have replaced ethnicity as inclusive identities, rendering ethnicity more exclusive in national politics even as it continues to be more inclusive at the local level.
Many academics and nonacademics alike have held up Botswana as an African success story, and have often attributed that success to Botswana’s great ethnic homogeneity. There are, however, two main problems with this argument. First, Botswana’s ethnic homogeneity is in many ways very similar to Somalia’s, but the two countries could hardly be more different when it comes to civil strife. In both cases, subethnic categories existed that could, and in Somalia’s case did, form the basis of intense domestic conflict. Somalia’s deep traditions of feuding, its more predatory colonial regime, which prevented the emergence of inclusive indigenous elites, its colonial division into two separate territories ruled by two different colonial powers, and its role in Cold War power politics all help to explain the difference (Samatar 1997). Second, while Botswana is more ethnically homogeneous than most countries in Africa or, indeed, the world, it is still ethnically diverse to some degree. Botswana’s Bushmen have long campaigned, with international support, against exclusive and paternalistic government policies, not to mention generalized informal prejudice. Similarly, even Bantu‐speaking agriculturalists like the Kgalagadi and the Kalanga have chafed against Tswana supremacy. In all these cases, success has been limited but at the same time sufficient to prevent the escalation of conflict. Botswana’s ethnic minorities have for the most part not relied on a simple oppositional ethnicity to further their political goals. Instead, they have pragmatically deployed a host of less confrontational strategies: international alliances with nongovernmental organizations, the embrace of minority rights discourse as opposed to any kind of separatism, and cultivation of political and economic linkages with sympathetic members of the Tswana establishment (Motzafi‐Haller 1994; Solway 1994; Van Binsbergen 1994; Werbner 2002; Wilmsen 2002; Pelican and Maruyama 2015).
Lesotho and Swaziland have much in common with Botswana. Though all three countries are dominated by ethnic groups that actually have more people living in South Africa, segments of those ethnic groups managed to avoid incorporation into South Africa through judicious alliances with missionaries and elements in the British government. Nevertheless, labor migration and other economic ties have meant that these countries cannot be examined in isolation from South Africa. Thus, in regards to Lesotho, one can speak of ethnic exclusion based on various Sotho subgroups within the country (Khalanyane 2012) while, at the same time, seeing even the Lesotho Sotho as an ethnic minority within the South African political and socioeconomic system (Harris 1988). But there are differences between the three countries. The discovery of diamonds in Botswana has greatly lessened Botswana’s dependence on labor migration to South Africa, in stark contrast to Lesotho and Swaziland. Also, the Tswana chiefs and the Sotho king have been politically marginalized for the most part, while the Swazi king is still essentially an absolute monarch. Already, 30 years ago, in an article that seems amazingly prescient in the light of later developments, historian Hugh MacMillan argued that this situation developed through the process of decolonization. The Swazi kings, royal family, and aristocracy deftly manipulated Swazi tradition and Swazi ethnic mobilization against both Boer settlers and a Zulu minority that promoted African nationalism and labor activism, all factors that had no analogues in Botswana or Lesotho (MacMillan 1985).
Of all the countries in southern Africa, Mozambique and Zimbabwe have probably experienced the greatest levels of violence since the end of white rule, in 1975 and 1980, respectively. As we have seen, the nature of Portuguese colonial rule in Mozambique meant that race, class, and region were more important political categories than ethnicity was. Still, Mozambique’s main liberation movement, FRELIMO, often identified “tribalism” as a major problem and went so far as to abolish chiefship in the country. After independence, the 1977–1992 Mozambican civil war certainly had an ethnic component, with FRELIMO accusing their opponents, RENAMO, of being narrowly tribally based (in the central and northern parts of the country), while RENAMO accused FRELIMO of hiding its own tribal favoritism (toward the southern groups) behind a rhetoric of national unity. Though both arguments had some truth to them, in practice the situation was far more complicated: the political orientations of the members of, say, the Maconde of the north have always been diverse, and Mozambicans’ attitudes toward FRELIMO cannot be easily reduced to either support or opposition (Cahen 1999, 2000; Virtanen 2005). In Zimbabwe, the most notable case of ethnic conflict was the Gukurahundi, a government action during the mid‐1980s, in principle against regime opponents, although in practice it seems that many people who were killed were Ndebele, who were assumed to be opponents simply by virtue of their ethnicity. Shona–Ndebele conflict had a long history in Zimbabwe even before the Gukurahundi, from the Ndebele king Mzikilazi’s crossing of the Limpopo in the late 1830s to the divisions that tore apart the liberation movement. However, political divisions among the majority Shona have an even deeper history and are painfully obvious today, and have often involved interethnic coalitions on both sides of any divide. Under such circumstances, it is difficult to determine the specific salience of ethnicity as opposed to a whole host of other factors (Muzondidya and Ndlovu‐Gatsheni 2007; Ndlovu‐Gatsheni 2012).
The history of postcolonial ethnicity in Namibia has paralleled that in other southern African countries in many ways. Like Botswana, Lesotho, Swaziland, and Zimbabwe, Namibia has one ethnic group, the Ovambo, which makes up half or more of the country’s total population. Moreover, as in the other cases, this apparent ethnic unity covers up substantial subethnic diversity and even conflict. Since independence Namibia has tried to forge national unity through the classic strategy of multiethnic countries worldwide: nation building through government propaganda in the media, schools, and museums. The results have been mixed (Schildkrout 1995). Like the liberation movements in South Africa, Zimbabwe, and Mozambique, Namibia’s SWAPO strongly criticized chiefship during the liberation struggle but, like the ANC in South Africa and ZANU‐PF in Zimbabwe and unlike Mozambique’s FRELIMO, ended up forging strong alliances with chiefs after independence (Friedman 2005). As in Botswana, Namibia’s most notable postcolonial ethnic conflicts have been between the Bantu‐speaking majority and the Khoisan/colored minority, including, inter alia, the Damara, Nama, coloreds, San, and the Rehoboth Basters, who together constitute more than a fifth of Namibia’s population and whose memberships shade into one another. For example, many San fought on the South African side during the Namibian War of Independence, while the Rehoboth Basters tried to declare independence rather than accept Namibian rule after South Africa withdrew. Tensions linger, and Namibia’s Khoisan/coloreds suffer from substantial political exclusion and economic disadvantage. They have turned to indigenous rights rhetoric, while SWAPO has often questioned their loyalty to the nation (Forrest 1994; Widlok 1996; Kjæret and Stokke 2003; Lee 2003; Taylor 2008).
Finally, the end of apartheid in South Africa has in some ways fractured and in some ways focused ethnic identities there. One of the most important developments in postapartheid South Africa has been the emergence of a cosmopolitan, multiracial, English‐speaking middle class. The growing dominance of English, the decline of state support for minority languages, and the emigration of many Afrikaans speakers have all seriously eroded Afrikaner ethnic identity (Giliomee 2014), although during the apartheid era many white Afrikaans speakers were already rejecting Afrikaner ethnic identity as it was conventionally defined (Louw‐Potgieter 1988). Though similar processes have affected black ethnicities such as the Zulu, class limits black access to English‐speaking cosmopolitan identity, and township residents who try to do so risk certain local exclusion for not necessarily guaranteed privilege and acceptance by other English speakers (Rudwick 2008). The rise of the “Zulu clique” under Jacob Zuma’s presidency from 2009 to 2018 has raised the specter of ethnic conflict in a country that has enjoyed relative ethnic peace for some 20 years. However, anthropologist Hylton White sees a new and different kind of ethnic politics emerging here. Views of ethnic groups as discrete populations with shared culture and language are becoming harder to maintain in mixed urban environments, and traditional authorities are becoming increasingly irrelevant in a black South African population that became majority urban in the early 2000s. In this urban environment, crime and unemployment and underemployment are the biggest concerns. Zuma’s Zuluness was more about making him relatable to the average black South African, and about raising the possibility that one could achieve personal prosperity and security through some kind of ethnically grounded relationship with Zuma’s state (White 2012). White sees a troubling potential here for growing authoritarianism, but he does not mention another troubling prospect: ethnic conflict. By focusing on the inclusiveness of this new Zulu identity, White is perhaps making the opposite mistake to that which Ranger, Vail, and Landau made in focusing so much on the exclusiveness of ethnicity. One senses in South Africa in particular, and in southern Africa in general, that the region’s success in containing growing social discontent cannot last forever. When major explosions start happening, one wonders what role ethnicity might play in them.
I would like to acknowledge the help provided by my research assistant, Cecelia Ward of Ripon College, in putting together the bibliography upon which this review essay is based.