Chapter One—
The Conflict and the Conflict about the Conflict

There is a conflict in South Africa that has something to do with race. That is about as far as agreement runs among many of the participants and interpreters of the conflict. Beyond that, there is disagreement over the extent to which the conflict is really about race, as opposed to being about oppression merely in the guise of race, or about nationalism among groups demarcated by race, or about contending claims to the same land. There is disagreement over the identification and even the names of the racial categories. There is disagreement over the extent to which the conflict also involves ethnic differences within each of the racial categories. There is no consensus whatever about whether a future South Africa might also be divided along ascriptive (that is, racial and ethnic) lines and, if so, how severe such divisions might become. Even among those who acknowledge the possibility of ascriptive divisions, there is discord about what measures might be required to reduce future conflicts. There is further controversy over whether democracy is possible in such a future society and, if so, over what institutions might be required to foster a democratic outcome. Moreover, the common term democracy obscures fundamentally different visions of what a democratic regime might look like. In short, the basic description of South African society, present and future, is contested.

The more uncertain is the depiction of South Africa as a society divided along racial and ethnic lines, the more certain is the depiction of South Africa as a society divided along ideological lines—lines that de-

marcate fundamental differences in how the society ought properly to be understood and organized. Whereas in other settings knowledge about a society can proceed cumulatively, in South Africa virtually every interpretive effort involves some return to first principles, even to first facts. The plethora of contributions also means that every new contribution must struggle with nuances of difference in the understandings of its predecessors. South Africa is characterized by cognitive dissensus.

There is, then, not one conflict in South Africa, or even one type of conflict, but two. There is the conflict itself, and there is the metaconflict—the conflict over the nature of the conflict. Neither is coterminous with the other; neither can be reduced to the other.

In many severely divided societies, there are differences among participants and observers in characterizations of the conflict, but rarely are the differences as deep as they are in South Africa. The existence of a conflict over the conflict greatly complicates the problems of those who wish to plan for a democratic South Africa. In polities where all groups acknowledge the divided character of the society but disagree over the justice of their respective claims, there still remain the difficult tasks of specifying the terms of intergroup relations and the modes of handling conflict. But where the basic character of the society and the identity of its components are themselves disputed, even the need to make provision for future conflict is put in issue.

Any effort to engage in planning for democracy must begin with divergent understandings of South Africa. As I shall point out later, South Africa has many features in common with other severely divided societies. One of the most difficult tasks of putative constitutional engineers in such a setting is to take account of the pervasive dissensus without abandoning their own understanding of the conflict.

Many other societies are doubly divided, but not in the same way as South Africa is. In some societies, racial or ethnic divisions coexist, alternate, and compete for political attention with class, religious, or regional differences. In others, one set of racial or ethnic conflicts competes for attention with another set. Historically, South Africa itself has been doubly divided in this sense, for White Afrikaans and English speakers formed two contending groups, whose conflict in some ways exacerbated the racial conflict. The dual conflict I am referring to here is different, however, because the conflict of understanding does not in any sense alternate with the ascriptive conflict—it relates directly and exclusively to that conflict and thereby compounds it. Just how the conflict and the metaconflict relate to each other I shall attempt to explain

after we have had a chance to examine some competing views, their origins, and the impact of the dissensus on even the terminology of the conflict.

Twelve South Africas

Practically no one in South Africa thinks it is a healthy society, but diagnoses of what ails it and prescriptions to cure it run a wide gamut. What is more, the divergent perspectives are fervently held. What I shall present are merely tendencies, rather arbitrarily selected capsule views typical of particular strands of thought but not necessarily representative of any single proponent. Virtually all strands are heterogeneous and evolving, so it would be a distortion to regard hybridized descriptions of them as anything more than paradigmatic. No doubt these capsule views do violence to nuances within each of the conceptions depicted. Even with these qualifications, a sample catalogue of this sort makes clear the scope and depth of the dissensus.

I shall enumerate the perspectives of participants and a few selected observers alike. Ordinarily, the views of participants should have priority in an exercise of this sort, but in the South African case the views of some observers either already represent a segment of more general South African opinion or are likely to form some part of the ongoing debate.^[1] The sketches emphasize alternative descriptions of contemporary South African society, the end-states that are envisioned (including prospects for democracy), and the measures advocated to bring about the desired end-states. Some positions, of course, have changed over time. Some organizations and individuals could easily subscribe to parts of more than one of the views described here. Although twelve ideal-type perspectives are enumerated, there are many commonalities among clusters of these perspectives on one point or another. While it is important to elucidate the many ways of looking at South Africa, it is equally important not to exaggerate the differences either.

1. An Official View. South Africa is a society divided into four racial groups (Whites, Coloureds, Indians, and Africans) and, within

[1] Accordingly, the views depicted are not intended as a reflection of the state of academic theorizing about South Africa. Some major attempts to integrate South African experience into a wider framework, such as the theory of capitalist development or the theory of rational choice, are therefore completely omitted. See Stanley B. Greenberg, Race and State in Capitalist Development (New Haven: Yale University Press, 1980); Michael Banton, Racial and Ethnic Competition (Cambridge: Cambridge University Press, 1983).

these groups, into a number of distinct ethnic groups or nationalities. All of the racial groups need to participate in politics at the center, albeit on a group basis (as all groups do not participate now in the White-dominated regime). To create peace and democracy in such a divided society, however, maximum power over its "own affairs" should be devolved to each of the racial (and perhaps also the ethnic) groups, reserving to the center power over those matters that affect everyone. In addition, some or all of the nationalities may need their own territorial expression, in the form of self-governing, perhaps federated, homelands. (All of this is put in rather contingent terms. Within a few years of the 1983 Constitution, the government had become equivocal about the extent to which it continued to adhere both to the own affairs—common affairs dichotomy of that Constitution and to its policy regarding independent homelands. After the legalization of several opposition groups in 1990, these ideological commitments were weakened further.)

2. A Charterist View. According to the Freedom Charter, adopted at a public meeting in 1955 and subscribed to by the African National Congress (ANC), "South Africa belongs to all who live in it, black and white."^[2] The charter did recognize the existence of several "national groups," which, it declared, "shall have equal rights!" These were to include "equal status in the bodies of state, in the courts and in the schools for all national groups and races. . . ."^[3] In 1969, however, the ANC's Revolutionary Program rejected the idea of separate political institutions for racial minorities.^[4] If South Africa manifests any racial or ethnic divisions, that is because of apartheid. Abolish apartheid, all forms of racial discrimination, and all manifestations of "separate development," and South Africa can have what is called a "nonracial [and nonethnic] democracy," although this will not necessarily be a "classical liberal democracy."^[5]

[2] Freedom Charter , (preamble adopted at the Congress of the People, Kliptown, South Africa, June 26, 1955), in Thomas G. Karis and Gwendolen M. Carter, eds., From Protest to Challenge: A Documentary History of African Politics in South Africa , vol. 3, 1882–1964 (Stanford: Hoover Institution Press, 1977), p. 205.

[3] Ibid. For the background the Freedom Charter, see Tom Lodge, Black Politics in South Africa since 1945 (London: Longman, 1983), pp. 69–74. See also Thomas G. Karis, "Revolution in the Making: Black Politics in South Africa," Foreign Affairs 62, no. 2 (Winter 1983–84): 378–406.

[4] See "African Nationalism in South Africa," in James Leatt et al., eds., Contending Ideologies in South Africa (Cape Town: David Philip, 1986), p. 102; Lodge, Black Politics in South Africa since 1945 , p. 301.

[5] Monty Narsoo, "Responses to the Report," in Charles Simkins et al., The Prisoners of Tradition and the Politics of Nation Building (Johannesburg: South African Institute of Race Relations, 1988), p. 88. For further evidence of hostility to liberal democracy, see Tom Lodge, "The United Democratic Front: Leadership and Ideology," inJohn D. Brewer, Can South Africa Survive? Five Minutes to Midnight (New York: St. Martin's Press, 1989), p. 216.

3. An Alternative Charterist View. There is less of a plural society and more of a common society in South Africa than many people realize. Commonalities among South Africans have been obscured by apartheid. If only nature could take its course without the artificial constraints of apartheid, an unremarkable democracy would develop in a nonracial society. (In some such formulations, however, rigorous safeguards, including some based on group affiliation, are regarded as necessary to guarantee the democratic character of the nonracial society.^[6] )

4. A People's Democracy View. South Africa without apartheid can easily be a nonracial democracy, with no need for pernicious "group" protections. But mere representative democracy alone is inadequate. The future of South Africa can and will be shaped by the popular opposition to apartheid, out of which is growing a new form of people's democracy. That form is not limited to periodic elections or to a few spheres of political activity but is genuinely based on local organization by oppressed people, who are prepared to take every sphere of activity that concerns their lives directly into their own hands. (Such views are held by significant segments of the United Democratic Front [UDF], the Mass Democratic Movement, and the African National Congress.^[7] )

5. An Africanist View. South Africa is not an ordinary divided society but a colonial society, defined by colonial oppression in the form of racially based capitalism. The problem of South Africa is "settler colonialism."^[8] Africans are the indigenous people and the sole legitimate owners of the soil. South Africa needs an anti-colonial revolution before it can become democratic and socialist. The resulting state will

[6] See Heribert Adam and Kogila A. Moodley, South Africa without Apartheid: Dismantling Racial Domination (Berkeley and Los Angeles: University of California Press, 1986), discussed at much greater length in Chapter 4, below.

[7] Paulus Zulu, "Resistance in the Townships—An Overview," in Fatima Meer, ed., Resistance in the Townships (Durban: Madiba, 1989), pp. 10–23; Andrew Boraine on Behalf of the United Democratic Front, "Democracy and Government: Towards a People's Struggle," Institute for a Democratic Alternative for South Africa, Occasional Paper no. 3 (1987); African National Congress, Strategy and Tactics of the ANC (1969), quoted in Roland Stanbridge, "Contemporary African Political Organizations and Movements," in Robert M. Price and Carl G. Rosberg, eds., The Apartheid Regime: Political Power and Racial Domination (Berkeley: University of California Institute of International Studies, 1980), p. 98.

[8] For the Africanist tendency in the opposition movements, see, e.g., Lodge, Black Politics in South Africa since 1945 , pp. 84–86; Karis, "Revolution in the Making," p. 382. On "settler colonialism," see, e.g., Weekly Mail (Johannesburg), November 17–23, 1989. On skepticism of parliamentary democracy, see Simkins et al., The Prisoners of Tradition and the Politics of Nation Building , pp. 25–26.

require no constitutional checks and balances. Rather than parliamentary democracy, South Africa needs representation on a class basis in a socialist society. (In some such formulations, "nonindigenous" peoples have no part to play in a future South Africa, whereas in others even Whites can become Africans, by means unspecified.)

6. A Black Consciousness View. South Africa is a society divided by the oppression of color. Once those oppressed by color find the means to overcome their oppression, South Africa can become a socialist, revolutionary state. (In some such formulations, Black Consciousness is merely a means of struggle, so that color becomes irrelevant once the struggle is won, whereas in others something like a Black state is envisioned.^[9] )

7. A Racial Self-Assertion View. South Africa is a severely divided society, in which democracy is improbable. But just as "racial dignity" has assumed priority over democracy in post-colonial African states, so may racial self-assertion be necessary in South Africa, even if democracy is not achieved.^[10] "To some, freedom means the turning of the tables. . . ."^[11]

8. A Two-Nationalisms Partitionist View. South Africa is fundamentally divided into Black and White. The interests of the two groups are so incompatible that territorial division is necessary. (There are many versions of this view circulating among dissident, sometimes exotic, White political organizations,^[12] but it is also a view that was espoused by the African National Congress in the 1920s and the All-African Convention in 1936.^[13] )

[9] See "Black Consciousness," in Leatt et al., eds., Contending Ideologies in South Africa , pp. 105–19.

[10] David D. Laitin, "South Africa: Violence, Myths, and Democratic Reform," World Politics 39, no. 2 (January 1987): 258–79, at 259 n. 3, 279. Since it is extremely unfashionable to eschew democracy in South Africa, this view is rarely, if ever, articulated there; but it probably enjoys a fair amount of support. It has, of course, some affinity to the Africanist view identified previously.

[11] Nomavenda Mathiane, South Africa: Diary of Troubled Times (New York: Freedom House, 1989), p. 117. For an interesting range of views, see ibid., pp. 115–20.

[12] Such as the Afrikaner Weerstandsbeweging, the Vereniging van Oranjewerkers, and the Afrikaner Volkswag. See Carole Cooper et al., Race Relations Survey, 1985 (Johannesburg: South African Institute of Race Relations, 1986), pp. 11–12. For an inventory, see Helen Zille, "The Right Wing in South African Politics," in Peter Berger and Bobby Godsell, eds., A Future South Africa: Visions, Strategies and Realities (Cape Town: Human & Rousseau Talfelberg, 1988), pp. 55–94.

[13] Gwendolen M. Carter, "African Concepts of Nationalism in South Africa," in Heribert Adam, ed., South Africa: Sociological Perspectives (London: Oxford University Press, 1971), p. 118.

9. A Two-Nationalisms Accommodationist View. South Africa consists of several racial and ethnic groups, but the fundamental struggle is between White (or Afrikaner) and Black (or African) nationalism. Given time, these two nationalisms could accommodate each other and perhaps grow together into a nonracial society. For the time being, however, a binational state is the best that can be hoped for. With a dual set of largely autonomous institutions, such a state could be democratic.^[14]

10. A Consociational View. South Africa is a severely divided society, with both racial and ethnic divisions. Democracy is rendered difficult to achieve by virtue of that fact—difficult but not impossible or even improbable if elaborate precautions are taken. To achieve democracy in a divided society, elites of the various groups must agree to share executive power and abide by a system of mutual vetoes and spheres of communal autonomy. (Over a period of nearly two decades, enormous energy has been invested in devising such model arrangements for South Africa.^[15] )

11. A Modified Consociational View. South Africa is a severely divided society, requiring some complex institutions if it is to function democratically. "Checks and balances" will be needed to protect minorities, including "black political minorities."^[16] But a full-blown apparatus of mutual vetoes and delegations of communal authority would render it impossible for the new polity to solve some of its most pressing problems of inequality and perhaps also, therefore, to remain democratic.^[17]

12. A Simple Majoritarian View. Whether South Africa is or is not

[14] See Hermann Giliomee, "The Communal Nature of the South African Conflict," in Hermann Giliomee and Lawrence Schlemmer, eds., Negotiating South Africa's Future (New York: St. Martin's Press, 1989), pp. 114–29.

[15] Most of these efforts are discussed in Arend Lijphart, Power-Sharing in South Africa (Berkeley: University of California Institute of International Studies, 1985), pp. 47–82. Among the major contributions are F. van Zyl Slabbert and David Welsh, South Africa's Options: Strategies for Sharing Power (New York: St. Martin's Press, 1979); Political Commission of the Study Project on Christianity in Apartheid Society, South Africa's Political Alternatives , SPRO-CAS Publication no. 10 (Johannesburg: Ravan Press, 1973); The Buthelezi Commission, The Requirements for Stability and Development in KwaZulu and Natal , vol. 1 (Durban: H & H Publications, 1982); KwaZulu Natal Indaba, Constitutional Proposals Agreed To on 28 November 1986 (Durban: n.p., n.d.; pamphlet).

[16] Charles Simkins, "Democracy and Government: A Post-Leninist Perspective," Institute for a Democratic Alternative for South Africa, Occasional Paper no. 1 (1987), p. 6.

[17] Charles Simkins, Reconstructing South African Liberalism (Johannesburg: South African Institute of Race Relations, 1986; mimeo.), pp. 74–77.

a severely divided society is irrelevant. The problem is that it is undemocratic. To make it democratic, the majority must rule, and rule unimpeded. In the ordinary course of events, majority rule will produce a Black government.^[18
]

It might be possible to reduce these twelve views to four main tendencies—official, Charterist, Africanist, and consociational—but this would do violence to the range of perspectives and would omit some important strains, both articulate and inchoate. Even a fourfold division, however, would show just how polarized ideologies are, for it would not be possible to align them on a spectrum if the test were self-perceived proximity of one ideology to the next. Or, to make the point more graphically, organizations in the Charterist stream have been engaged for years in violence not only with the official instrumentalities of force but also with organizations aligned with the Africanist stream and with Inkatha, the organization led by Chief Mangosuthu Buthelezi, an avowed consociationalist.

If all or nearly all of these perspectives are to be taken seriously, as they should be, South Africa could easily be seen as a kind of riddle society, impossible to fathom. Even within the realm of similar general diagnoses of the society's condition, there are many different prescriptions for its treatment. But despite the plurality of conceptions about South Africa, it is no riddle. South Africa is a complex place, but the cleavages reflected in these divergent views are patterned and explicable.

It is conventional to regard ideological cleavages as the fallout of historical events, conflicts, or movements.^[19] South African history contains the seeds of the ideological cleavages already identified, but with a special qualification. European South Africa was what Louis Hartz calls a "fragment society," an incipient whole society constructed out of merely a part of the European society of origin—or, in South Africa's case, out of two societies of origin, the Netherlands and Great Britain.^[20]

[18] This is sometimes the formulation of the ANC when it speaks of the "transfer of power from a white minority to a black majority." Quoted in Heribert Adam, "Exile and Resistance: The African National Congress, the South African Communist Party and the Pan Africanist Congress," in Berger and Godsell, eds., A Future South Africa , p. 104.

[19] Thus, Seymour Martin Lipset and Stein Rokkan speak of "inherited cleavages." "Cleavage Structures, Party Systems, and Voter Alignments: An Introduction," in Seymour Martin Lipset and Stein Rokkan, eds., Party Systems and Voter Alignments: Crossnational Perspectives (New York: Free Press, 1967), p. 8.

[20] Louis Hartz et al., The Founding of New Societies (New York: Harcourt, Brace & World, 1964).

As Hartz explains, the history of such societies entails working out the logic of the fragment ethos or culture as it encounters new conditions, new problems, and new people in the transplanted setting. The process of fragmentation, whereby a part becomes a whole, accounts, in Hartz's view, for the distinctive character of fragments such as French Canada or Dutch South Africa. The full social complexity of the home country was not transplanted, and so the fragments escaped social conflicts that occurred in the sending country after the migration. Eventually, however, the isolation of the fragments from world history broke down, and the distinctive fragment societies returned to worlds they thought they had eluded. "As the globe contracted, the Western revolution that the fragments escaped was spreading throughout it with increasing rapidity, so that ultimately it was bound to overtake them, if only from a distant place suddenly made near. If Holland was left behind, there was still Russia, China, the awakening African states."^[21]

The concatenation of distinctively South African historical events and reverberations of events in world history is responsible, above all, for the proliferation of ideological currents. To make sense of what otherwise appears to be a kaleidoscope, to put back together the views I have pulled apart, and to capture the extent to which ideas have migrated across cleavages and over time, at least an abbreviated historical excursion is required.

On Understanding South African Cleavages

The conflict and the metaconflict have deep roots in South African history, particularly the history of the ruling Afrikaner group and its encounters with the British and the Africans who occupied the same space. Both nationalist aspiration and racial ideology formed prominent threads of Afrikaner history.^[22]

The Cape Colony in South Africa was founded by the Dutch East India Company in the seventeenth century and finally taken over by the British in the early nineteenth century. In the interval, descendants of the first Dutch and other European settlers were well on their way to forming a distinctive group, the Afrikaners. Throughout their long conflicts with the Dutch East India Company and with the British, the Af-

[21] Ibid., pp. 20–21.

[22] Among useful histories of South Africa, see T. R. H. Davenport, South Africa: A Modern History , 3d ed. (Toronto: University of Toronto Press, 1987); C. E. de Kiewiet, A History of South Africa (London: Oxford University Press, 1957).

rikaners were also engaged in conflicts with the Africans around them. Practically every step the Afrikaners took to secure their precarious collective existence had elements of both struggles embedded in it. The Great Trek north from the Cape in the 1830s, for instance, was precipitated by the desire to escape some unpleasant features of British rule and threats to cattle and security from Africans on the eastern frontier.^[23] The Afrikaners thought they were creating or preserving a nation out of a people threatened with "extinction,"^[24] particularly by an increasingly powerful British colonialism, but they were simultaneously acting as conquerors of Africans.

For more than two centuries, Afrikaners (and later the British as well) subjugated Africans in the territories they controlled, and they developed racial attitudes typical of such relationships. Along their shifting frontiers, Afrikaners also fought the various Black ethnic groups, forming alliances with some against others, in struggles lasting until the late nineteenth century. While this process was winding down, events such as the British annexation of the Transvaal in 1877 gradually made Afrikaner relations with the British authorities and with English-speaking South Africans the overriding issue.^[25]

For half a century after the Second Anglo-Boer War (1899–1902), there was an uneasy power-sharing arrangement within the British Empire and Commonwealth. Slowly, however, the fears of the socially inferior Afrikaners prevailed over appeals to unity with the far wealthier and culturally self-assured English-speaking South Africans. In 1948, the Afrikaner-based National Party came to power and began a process of Afrikaner upliftment. Within a few decades, Afrikaners were brought to virtual economic parity with their English-speaking compatriots.^[26] Afrikaner nationalism had triumphed.

The National Party, in "its endeavor to make the country safe for Afrikanerdom . . . set up a bulwark of restrictive racial legislation."^[27] Threatened by the Africans around them, as Afrikaners had been for 300 years, the Nationalists greatly extended previous policies of racial

[23] See Hermann Giliomee, "The Growth of Afrikaner Identity," in Heribert Adam and Hermann Giliomee, eds., Ethnic Power Mobilized: Can South Africa Change? (New Haven: Yale University Press, 1979), p. 95.

[24] Ibid., p. 103.

[25] See Gwendolen M. Carter, The Politics of Inequality: South Africa since 1948 , 2d ed. (New York: Praeger, 1959), pp. 27–38.

[26] See Milton J. Esman, "Ethnic Politics and Economic Power," Comparative Politics 19, no. 4 (July 1987): 395–418, at 406–10.

[27] Giliomee, "The Growth of Afrikaner Identity," p. 115.

segregation. In the 1950s, the government put in place a network of statutes to separate the races spatially and socially. The Population Registration Act (1950) assigned every person to a racial category. The misnamed Abolition of Passes and Coordination of Documents Act (1952) required all Africans to carry a wad of identifying papers. Over the next decades, hundreds of thousands were arrested annually for failing to produce the documents upon demand. The Reservation of Separate Amenities Act (1953) provided for segregation in public facilities. The Native Laws Amendment Act (1952) and the Natives Resettlement Act (1954) limited the rights of Africans to live in urban areas and permitted authorities to relocate those not authorized to be where they were living. The Group Areas Act (1950) was designed to segregate every locality by race.

In the 1960s, under the prime ministership of H. F. Verwoerd (1958–66), the apartheid system was refined, perfected, and—above all—enforced. Under the euphemism of "influx control," tough restrictions were placed on Black migration to cities. For those permitted to migrate, urban life was made unattractive. Large numbers of Black workers were forced to live in single-sex hostels far from their families, whom they got to visit only occasionally. There were restrictions on the movement of Indians and a complete ban on their residence in one province, the Orange Free State. Whole business and residential areas, declared "White" under the Group Areas Act, were eventually cleared of people who had worked or lived there for decades. Blacks were relegated to a standardized, inferior "Bantu education." Higher education was also segregated, and several separate universities were established. Verwoerd propounded an ideology of "separate development."

With the Bantu Self-Government Act (1959), the regime initiated a policy of consolidating eight "homelands" that might eventually become independent African states. Land ownership outside the homelands became virtually impossible for Blacks, all of whom were ultimately intended to become citizens of their respective homelands and hence foreign workers in South Africa proper. In the meantime, the citizenship of all except Whites had been devalued. The regime meticulously removed all but White voters from the electorate, in part to insure that Afrikaners were not outnumbered at the polls by English-speaking Whites and their "non-White" allies.

At its height, apartheid thus had two complementary thrusts: racial separation within South Africa proper and ethnic "self-determination" in the Bantustans. Both were out of step with international standards.

As segregation was being phased out in the United States, it was being reinforced in South Africa. As Britain, France, and Belgium were granting African independence within existing colonial boundaries, South Africa was contemplating balkanized independence for impoverished, disconnected dots on a map. As Hartz suggests, developments outside South Africa could no longer be ignored.

As the institutions of apartheid were being strengthened, the ideology supporting it was being undermined. The standards of the outside world had an insidious influence on apartheid even as they were being defied. Contact with those standards was intensified, ironically enough, by the very success of Afrikaner nationalism, which produced a class of Afrikaner academics, professionals, writers, and business people who understood emerging international norms and were, at the very least, embarrassed by the failure of South Africa to conform to those norms. At the same time, the economy engendered major changes in Black educational, occupational, and income status. Between 1960 and 1985, the number of African students in secondary school increased more than 20-fold, those in the highest primary-school standard more than 40-fold, and those in university more than 25-fold.^[28] Between 1970 and 1980, Africans increased their share of higher-level professional and technical employees by more than a third and of administrative, managerial, and clerical workers by more than 50 percent.^[29] Although average White income was still more than three times average Black income, Black incomes rose six times as fast as White incomes from 1970 onward.^[30] Since surveys show that tolerance among Africans for the regime of apartheid declined sharply with increasing education,^[31] the expansion of Black education—and probably also the increases in income—paved the way for widespread Black opposition. Just as apartheid was fully constructed, it had powerful forces arrayed against it.

Black opposition to White rule has a rich history, long antedating apartheid. Facing a regime that had largely excluded them from political participation, opposition organizations had to answer at least three

[28] Hermann Giliomee and Lawrence Schlemmer, From Apartheid to Nation-Building (Cape Town: Oxford University Press, 1989), p. 118.

[29] Ibid.

[30] Ibid., pp. 125–26; Lawrence Schlemmer, "Change in South Africa: Opportunities and Constraints," in Price and Rosberg, eds., The Apartheid Regime , p. 238; Heribert Adam, "Three Perspectives on the Future of South Africa," International Journal of Comparative Sociology 20, nos. 1–2 (March–June 1979): 122–36, at 128–29.

[31] Giliomee and Schlemmer, From Apartheid to Nation-Building , p. 119; John D. Brewer, "Black Protest in South Africa's Crisis: A Comment on Legassick," African Affairs 85, no. 2 (April 1986): 283–94, at 287–90.

interrelated questions: (1) Who properly belongs to South Africa? (2) Who should be included in the struggle against the regime? (3) What institutions are appropriate for a future South Africa? Answers to these questions, which constitute components of the metaconflict, have varied widely over time and across organizations. The answers, however, were not formulated in isolation but as part of a response to White, particularly Afrikaner, pretensions; and the answers affected White responses in turn.

As I have suggested, Afrikaner political behavior over the centuries was underpinned by two motivating and justifying forces: nationalist aspiration and racial ideology. Once contending African groups had been defeated, racial ideology might well have occupied the field alone, had it not been for the enduring struggle against the British, which had the effect of sharpening the keen Afrikaner sense of a precarious national destiny. Black opposition to White domination also had, in the first instance, two sides: African nationalism, on the one hand, and non-racial ideology, on the other.

As early as the nineteenth century, a movement had begun to keep "Africa for the Africans."^[32] This was an enduring theme in the decades that followed, but it alternated with more inclusive conceptions of belonging. The African National Congress had operated on racially inclusive assumptions in choosing its allies and in advancing plans for a future South Africa. The Bill of Rights it published in 1923 went no further than to demand "equality of treatment and equality of citizenship in the land, irrespective of race, class, creed or origin."^[33] By 1944, however, an "Africanist" tendency was discernible in the ANC Youth League, led by Anton M. Lembede. Lembede's Youth League Policy Statement declared that "Africa is a Black man's country."^[34] The statement had implications at two levels. First, it implied that Blacks might have an exclusive or at least prior right to South Africa; perhaps others could stay only on sufferance. Second, it was critical of ANC cooperation with Whites, particularly with White communists, with whom the ANC was working. ANC leaders, in turn, saw the Africanists' thinking as a mirror image of Afrikaner nationalist ideology.

To the ANC's declaration in the Freedom Charter of 1955 that "South

[32] "African Nationalism in South Africa," p. 89.

[33] Quoted in ibid., p. 91. For the modest goals of the ANC's predecessor at its inception in 1912, see Brian Willan, Sol Plaatje, South African Nationalist, 1876–1932 (Berkeley and Los Angeles: University of California Press, 1984), pp. 150–54.

[34] Quoted in Lodge, Black Politics in South Africa since 1945 , p. 21.

Africa belongs to all who live in it, black and white . . .,"^[35] the Africanists rejoined in their journal that "[t]he African people have an inalienable claim on every inch of the African soil. . . . The non-Africans are guests of the Africans . . . [and] have to adjust themselves to the interests of Africa, their new home."^[36] In 1959, as one African country after another was receiving its independence, the ANC Africanists withdrew from the organization and founded the Pan Africanist Congress (PAC). The early Africanists did not quite resolve the oscillation in their pronouncements that implied at times that Whites did not belong in South Africa at all (or belonged as mere guests) and at other times that Whites were only disqualified from playing a part in the struggle for a South Africa that might nevertheless ultimately include them on equal terms. In general, however, the PAC's view has been to refer to Whites as colonial "settlers" in "occupied Azania."^[37] At the end of 1989, a new Pan Africanist Movement, an internal counterpart to the PAC, was created. Its founding document declared that "colonialism in occupied Azania has over the years developed to be internal colonization of the African people by a white settler minority," and participants at its inaugural meeting chanted "One settler—one bullet!"^[38]

In its choice of allies, the ANC also oscillated. Under Africanist influence, the organization adopted a militant Programme of Action in 1949 that espoused Black self-determination and eschewed cooperation with other racial groups, a position that was reversed at the time of the ANC's Defiance Campaign of 1952, following which the ANC returned to a policy of cooperating with the multiracial, but disproportionately White and Indian, South African Communist Party. Even the ANC, however, referred from time to time to the leading role Africans had to play in the liberation struggle, and some in the Charterist stream have spoken of "settlers," thieves who have "come and stolen my land."^[39]

These debates continued in the 1960s and 1970s as new organizations proliferated.^[40] Many of these wore the emblem of Black Con-

[35] Freedom Charter , preamble, p. 205.

[36] Quoted in "African Nationalism in South Africa," p. 98.

[37] Weekly Mail , October 7–13, 1988, quoting the Pan Africanist Congress; Colin Legum, ed., Africa Contemporary Record, 1986–87 (New York: Africana, 1988), p. B747, quoting PAC leader Johnson Mlambo. Azania is the Africanist name for South Africa.

[38] Quoted in New York Times , December 27, 1989.

[39] Winnie Mandela, Part of My Soul Went with Him (New York: Norton, 1984), pp. 42, 127.

[40] See Stanbridge, "Contemporary African Political Organizations and Movements," pp. 77–92.

sciousness, which drew some inspiration from anti-colonial struggles and perhaps even more from Black struggles in the West. Like the Africanists, the Black Consciousness groups were ambiguous on the ultimate place of those who were not Black in South Africa. For those, such as Steve Biko, seeking to consolidate Black solidarity before the struggle began, Black Consciousness could be reconciled with the inclusion of Whites on equal terms once the struggle was over. The Azanian People's Organization (AZAPO), founded in 1978 and working within the Black Consciousness tradition, spoke, on the other hand, of "repossessing the land," and Black Consciousness leaders repeatedly described Azania as belonging exclusively to Blacks.^[41] While AZAPO and its affiliate, the National Forum, incorporated strong elements of class analysis in their thinking, it was not the sort of orthodox class analysis that rendered thinking along racial lines illegitimate. Virtually all of the Africanist and Black Consciousness groups took the position that the struggle had to proceed without significant White participation.

In planning for a future South Africa, a full array of proposals was presented by the opposition over time. As I mentioned earlier, the ANC in the 1920s and the All-African Convention in 1936 advanced overtly partitionist positions, contemplating a division of South Africa into one Black and one White state. In the 1940s, the ANC recognized the existence of racial pluralism in a future, undivided South Africa and proposed parliamentary representation based on separate voters' rolls and reserved seats for various races. Later, the organization changed its position and demanded a single voters' roll.^[42] Nevertheless, recognition of pluralism was present even in the Freedom Charter, for it promised, as we have seen, "equal status" for "all national groups and races." The possibility of a racially federated South Africa was thus opened, but since 1969 the ANC has set its face against separate political institutions for racial minorities. The government's emphasis on racial categories and groups had made group-based arrangements anathema to the opposition.

The first president of the PAC, Robert Sobukwe, declared as early as

[41] "African Nationalism in South Africa," p. 113; Theodor Hanf et al., South Africa: The Prospects of Peaceful Change (London: Rex Collings, 1981), p. 285; "Tell No Lies . . . Claim No Easy Victories, a Position Paper of the Black Consciousness Movement on the Events Culminating in the United Democratic Front (UDF) and the Congress of South African Trade Unions (COSATU) Distancing Themselves from Mrs. Winnie Mandela" (1989; mimeo.), p. 3.

[42] "African Nationalism in South Africa," p. 93.

1959 that the PAC would "guarantee no minority rights, because we think in terms of individuals, not groups. . . ."^[43] Sobukwe equated multiracialism with "racialism multiplied."^[44] On this score, the Black Consciousness organizations were largely in agreement: they favored the term nonracialism over multiracialism , because they, too, wished to avoid thinking, at least ultimately, in terms of ascriptive groups. Even now, AZAPO argues that race does not really exist except as an artifact of economic exploitation and political oppression.^[45] With the infusion of class analysis, the prospect of liberal democracy has receded. AZAPO and the National Forum speak of the present system in terms of "racial capitalism"; AZAPO is committed to a "dictatorship of the proletariat," the PAC to African socialism. The PAC and AZAPO are both highly skeptical of multiparty democracy.^[46]

Even this cursory survey demonstrates the extent to which the various movements have either oscillated or proved ambiguous on the three questions of membership in the South African community, inclusion in the struggle against the regime, and the nature of future South African institutions. The customary demarcation between Africanist and Charterist organizations and ideas identifies the latter as being nonracial and inclusive. In answering the three questions posed, however, so clear a demarcation is not possible. Ideas have traveled across organizational boundaries. The ANC, committed to nonracialism, nevertheless included non-Africans (Whites, Coloureds, and Indians) in its National Executive Committee only as late as 1985. The ANC has referred to the South African government as a "colonial regime," and individual United Democratic Front leaders have referred to White "settlers."^[47] Sobukwe's dictum that the PAC thinks in terms only of individual rather than group rights has become and article of faith in virtually all extra-

[43] Quoted in Carter, "African Concepts of Nationalism in South Africa," p. 116.

[44] Quoted in Gail Gerhart, Black Power in South Africa: The Evolution of an Ideology (Berkeley and Los Angeles: University of California Press, 1978), p. 195.

[45] Narsoo, "Responses to the Report," p. 84.

[46] Simkins et al., The Prisoners of Tradition and the Politics of Nation Building , pp. 24–26.

[47] Proceedings of the Second National Consultative Congress of the ANC , Lusaka, Zambia, June 16–23, 1985, quoted in F. van Zyl Slabbert, "The Dynamics of Reform and Revolt in Current South Africa," lecture 3, "The Dynamics of Reform: Patterns of Resistance and Revolt," Institute for a Democratic Alternative for South Africa, Occasional Paper no. 9 (October 1987), p. 4; Hanf et al., South Africa: The Prospects of Peaceful Change , p. 285, quoting an ANC leader; Donald L. Horowitz, "After Apartheid: How Majority Rule Can Work," New Republic , November 4, 1985, p. 23, quoting a UDF leader.

parliamentary opposition thinking, despite the willingness of the ANC, in particular, to think quite differently throughout a good part of its history. Nonracial has eclipsed multiracial as the norm for the entire extraparliamentary opposition. The early Africanist notion that Whites, although not automatically part of the political community, could be accepted as Africans if they made a proper adjustment has also had influence far beyond its origins. Even Afrikaners in the parliamentary opposition have been heard to say that Whites must learn to be, not White, but merely South African. The terminology of some Black Consciousness organizations, which used the word Black to include all those oppressed by color, has also spread very widely among the opposition, as I shall note later. In some ways at least, one can say that the ANC has been winning the organizational battle among the opposition but that the Africanists have been winning the ideological battle.

The completeness of racial exclusion under the Nationalists, coupled with the emerging doctrines of the Black opposition, had an impact on Whites who were opposed to the regime. As early as 1959, the then—Progressive Party, the largely English-speaking parliamentary opposition, began to think seriously about ways for excluded groups to participate in South African politics. The report of the party's Molteno Commission was the first of several efforts to provide a political framework in reaction to apartheid and in recognition of the plural character of South African society.^[48] Over time, these efforts drew increasingly on theories of consociational democracy and emphasized formal devices to provide minority security. Their proponents were eventually joined by Chief Buthelezi, who, as a result of the government's homeland policy, had become chief minister of the Zulu homeland, although he refused the offer of independence for it. The effect was to provide a counterpoint to the undifferentiated majority rule propounded by the ANC and to the restrictive formulations of the Africanists. White politics and Black politics, often treated separately, do not exist in isolation.

The Soweto rising of 1976 and the introduction in 1984 of a new constitution that continued to exclude Africans from even limited parliamentary participation both produced surges of Black opposition. New organizational structures were created, and the opposition became a mass movement or, rather, several mass movements. The UDF, formed

[48] See Donald B. Molteno et al., A Report Prepared for the Progressive Party of South Africa by a Commission of Experts (N.p. 1960 and 1962).

to oppose the constitution, consisted of some 600 organizations. The increasing breadth and depth of opposition helped elicit additional perspectives on South Africa.

The new activity coincided with a less self-confident apartheid. In the late 1970s, some features of "petty apartheid" (segregated facilities) began to disappear. White universities were again permitted to accept some Black students. In 1975, the South African Defence Force "announced that black soldiers would enjoy the same status as whites of equal rank, and that whites would have to take orders from black officers."^[49] In the late 1970s, the regime set itself on a path leading to the recognition of Black trade unions. Some mixed residential areas were tolerated, although not fully legitimated until 1988. The anti-miscegenation acts were repealed in 1985. Influx control and the pass laws were abolished in 1986. Freehold title was made possible for Africans in the same year. Segregation in hotels, beaches, restaurants, and business districts declined rapidly in the late 1980s. The regime had embarked on an effort at "reform."

Complete racial separation was seen to be less and less feasible, and in the 1980s the regime began to speak, not of separate development, but of "cooperative coexistence," within "a system in which there is no domination of one population group over another, which in turn requires self-determination for each group over its own affairs and joint responsibility for and co-operation on common interests."^[50] Not that the regime was consistent about any of this. Just as the regime was finally abandoning measures that would have relegated all Africans to homeland citizenship, further territories were being added to some of the homelands. Untouched by "reform apartheid" were the central pillars of the system: segregated primary and secondary education, the Group Areas Act (the cornerstone of segregated living), the Population Registration Act (which assigned everyone to a race), and—above all—the segregated political system, in which Africans had no significant voice.

[49] Giliomee and Schlemmer, From Apartheid to Nation-Building , p. 124. The other changes recited in this paragraph are drawn from ibid., pp. 123–26, and from Hendrik W. van der Merwe, Pursuing Justice and Peace in South Africa (London: Routledge, 1989), p. 106. On the nonracial definition of "superior officer," see also Kenneth W. Grundy, Soldiers without Politics: Blacks in the South African Armed Forces (Berkeley and Los Angeles: University of California Press, 1983), p. 191.

[50] P. W. Botha, "Speech by the State President, Mr. P. W. Botha, on the Occasion of the Opening of the Second Session of the Eighth Parliament of the Republic of South Africa, 25 January 1985" (mimeo.), p. 1.

At the height of the UDF protests of 1985, the government was as ambiguous about the future as the opposition was. A National Party pamphlet, entitled . . . And What about the Black People?^[51] confessed to the bankruptcy of earlier plans to consign all Blacks to their homelands to find their political rights, but it rejected one person, one vote and a parliamentary chamber for Blacks. Having stated flatly that "[i]t has long been clear that political rights cannot be withheld from Black people forever," the pamphlet merely said that any future solution would have to take the "group nature of South African society . . . as the point of departure" and that no group could be placed in a position where it might be dominated by another.^[52] Vague demands for a nonracial state had been met by a vague plan for a multiracial state.

Many Whites took up more conservative positions, some advocating a return to pure apartheid and others, convinced that White control over an undivided South Africa was untenable, devising a variety of schemes for partition along racial lines. Among White academics, the old regime of race relations had lost its legitimacy. Some became committed to Charterist positions; others, along with members of the business community, developed an interest in consociational thinking; and a few, unconvinced by the feasibility either of power-sharing plans in an undivided South Africa or of territorial partition, pursued the theme of two nationalisms and argued for a binational state. This last theme produced a strong reaction because of its contradiction of a fundamental tenet of opposition ideology—namely, that race does not demarcate a legitimate pluralism worth preserving after apartheid. On this view, race is purely a device for oppression, not an affiliation that could be transformed into a benign nationalism.

Currents of thought in the Black townships are rather more difficult to ascertain than are the elaborate programs of established organizations to which many, but not necessarily all, township residents subscribe. At least two strands, however, can be identified.

The first might be labeled majoritarianism coupled with deep suspicion of minority protection. "God made us the majority in South Africa. As soon as Blacks do what everyone else does [practice majority rule]," there will be objections.^[53] "For centuries we have been oppressed, exploited and downtrodden both as individuals and as a peo-

[51] Stoffel van der Merwe, . . . And What about the Black People? (Pretoria: Information Service of the National Party, 1985; pamphlet).

[52] Ibid., pp. 12, 13.

[53] Conversations with a youth group from Soweto, August 5, 1989.

ple. Now that whites fear the day of reckoning we hear all this moralistic nonsense. It's another attempt to maintain power."^[54] On this view, any form of minority protection is out of the question.^[55] In some such formulations, there is more than a hint of revenge.^[56]

A second view might aptly be labeled Leninist. It is focused on revolutionary transformation, and it is quite prepared for things to get worse before they get better. One example: a comrade from Soweto argued, in terms reminiscent of Weimar Germany, that a Conservative Party government, prepared to restore apartheid, would be preferable to the status quo, because then "the people would know they are oppressed."^[57] Leninism may be dead in Eastern Europe, but it is not dead in the townships.^[58]

I have not listed such revolutionary formulations separately, but there is no doubt that many in several streams of the extraparliamentary opposition entertain transformative objectives. The ANC has said: "Victory must embrace more than formal political democracy. There must be a return of the wealth of the land to the people. . . . The perspective of a speedy progression from formal liberation to genuine and lasting emancipation is made more real by the existence in our country of a large and growing working class whose class consciousness complements national consciousness."^[59] Leaving aside the Marxist teleology implicit in the statement—that is, bourgeois nationalism can be converted more quickly to socialism because of the large working class—there is the same dichotomy of "formal" democracy, presumably liberal democracy, on the one hand, and "genuine liberation," "people's de-

[54] Quoted from a group discussion reported in Paulus Zulu, "The Politics of Internal Resistance Groupings," in Berger and Godsell, eds., A Future South Africa , p. 147.

[55] See Mandela, Part of My Soul Went with Him , p. 125:

It is not us, but the white man, who should be thinking about how he will fit into our future society. It's his problem. He has the audacity to talk about the protection of minority groups when he is oppressing the majority. The arrogance! That he sits in power for over four hundred years, legislating against millions and millions of people and oppressing us for all these generations, and now we must worry about the protection of minority rights, and of his property and his culture.

[56] "There are many stories of what will happen when blacks take over. One woman maintains she will hunt down all the whites who were responsible for black suffering, lock them up at Voortrekkerhoogte [a military township] and be their warder. On Sunday, her prisoners will go without shoes, winter or summer. Similar stories are common, half in jest and half serious." Mathiane, South Africa , p. 117.

[57] Conversations with a youth group from Soweto, August 5, 1989.

[58] See Robert K. Massey, Jr., "Great Expectations," New Republic , February 26, 1990, pp. 17–20, reporting township criticism, directed at two Soviet visitors, of the Russian rejection of socialism under Gorbachev.

[59] African National Congress, Strategy and Tactics of the ANC , p. 98.

mocracy," on the other. As Charles Simkins has noted,^[60] in a racially stratified society, where capital is mainly White and labor is mainly Black, such formulations translate into racial revolution at least as readily as they do into class revolution.

I have also not listed separately a set of township orientations that is easier to infer from behavior than from documents. It is simultaneously militant and pragmatic, but otherwise difficult to categorize. It derives from the UDF protest activity of the mid-1980s and after. At the local level, UDF affiliates often undertook consumer boycotts of White businesses. These boycotts sometimes led to negotiations and agreements, often wide-ranging, with White merchants and local authorities. Since they did not produce anything like an articulated perspective, I shall reserve further treatment of them for a later discussion of models of accommodation.^[61]

In the United Democratic Front, there is, in addition to majoritarian and Leninist strains, a strong current of support for what is called people's democracy, which is distinct from liberal democracy:

. . . not only are we opposed to the present parliament because we are excluded, but because parliamentary-type representation in itself represents a limited and narrow idea of democracy.

Millions of South Africans have for decades not only been denied political representation, but have also been oppressed and exploited. Our democratic aim therefore is control over every aspect of our lives, and not just the right (important as it is) to vote for a central government every four to five years.

When we speak of majority rule, we do not mean that black faces must simply replace white faces in parliament. A democratic solution in South Africa involves all South Africans, and in particular the working class, having control over all areas of daily existence—from national policy to housing, from schooling to working conditions, from transport to consumption of food. This for us is the essence of democracy. When we say that the people shall govern, we mean at all levels and in all spheres, and we demand that there be real, effective control on a daily basis.

This understanding of democracy tends to be fundamentally different from the various abstract constitutional models which tend to be put forward as solutions. Most of these are concerned with the question of how central political representation can be arranged so that "groups" cannot dominate each other, or how what is referred to as the "tyranny of the majority" can be avoided. . . .

The rudimentary organs of people's power that have begun to emerge in South Africa (street committees, defence committees, shop-steward struc-

[60] The Prisoners of Tradition and the Politics of Nation Building , p. 22.

[61] See Chapter 4, below.

tures, student representative councils, parent / teacher / student associations) represent in many ways the beginnings of the kind of democracy that we are striving for.^[62]

Formulations of this sort are very common.^[63] Among the Black opposition, there is a large measure of distrust of representative democracy,^[64] deriving from fear of cooptation and a long history of being far from the center of power. Needless to say, however, the transformative objectives of the opposition would require a very strong central government; there is tension between direct, bottom-up democracy and a strong center. There is also tension between direct democracy and multipartysim. As Simkins notes, when township street committees sprung up in 1984–86, they took one of two forms: either partisan differences were ignored or one organization took charge of an area and forcibly excluded all rivals.^[65] Neither pattern resembles multiparty competition. The conflicts among direct popular democracy, Leninism, and liberal democracy have only begun to surface, but they are profound.

Many of these perspectives, including some only barely touched on here, will be analyzed more thoroughly in later chapters. So, too, will the quite different dimensions of consensus and dissensus that derive from sample surveys of popular attitudes. All that needs to be said for now is that nearly every historical source of ideology, internal and external, was structured to produce dissensus. In opposing the regime, it became routine to oppose its definitions, its categories, its assumptions, its visions. Theodor Hanf has pointed out that the ANC's rejection of group-based political thinking occurred as a reaction to the rigid formulations of the White regime. Previously, the ANC itself was organized "plurally: based on collaboration between Africans, Indians and Whites as different but like-minded groups. It was government policy—the unilateral definition of groups, the manipulation and ascription of ethnicity—that induced the ANC to reject all group formulas, regardless of their nature and reason, as expression and symbol of a system of oppression."^[66] The very organizational structure of the opposition re-

[62] Boraine on Behalf of the United Democratic Front, "Democracy and Government: Towards a People's Struggle," pp. 4–5 (emphases in the original).

[63] See, e.g., Zulu, "Resistance in the Townships—An Overview," p. 17; "The Western Cape," in Meer, ed., Resistance in the Townships , p. 61.

[64] See Simkins et al., The Prisoners of Tradition and the Politics of Nation Building , p. 24.

[65] Ibid., p. 30.

[66] Theodor Hanf, "The Prospects of Accommodation in Communal Conflicts: A Comparative Study," in Giliomee and Schlemmer, eds., Negotiating South Africa's Future , p. 108.

sponded, by 180-degree reversal, to official formulations, so inimical were the regime's conceptions to the interests of all but its White subjects.

To be sure, the independence of Africa had an effect that cut across several conflicting positions. After decolonization, a White regime looked anomalous, even to itself. After decolonization, equal White participation became problematic for the regime's antagonists. By raising the question of a settler-dominated polity, African developments simultaneously undermined regime exclusiveness and opposition inclusiveness. And Marxist revolutionary formulations, superimposed on racial classes, had an inescapable element of racial reversal about them. Yet, in the formulation of both cognitive and prescriptive positions, the shaping influence of the South African state and the experience of apartheid stand out most prominently, if only by virtue of negation. In such a setting, the very term opposition takes on a totalistic meaning.

The Terminology of Dissensus

Terminological disputes reflect some of the basic differences of cognition. The government's division of the society based on racial groups generally yields four main categories: Whites, Coloureds, Indians (or Asians, thus including the very small Chinese minority), and Blacks (meaning Black Africans).^[67] A division of the society based on the oppression of color, however, often yields only two categories: the privileged, who are White, and the oppressed, who are Black. The Black category can then embrace all who are oppressed on grounds of color—Coloureds, Indians, and Africans alike. The origin of the inclusive term Black lies in Black Consciousness thought of the 1970s. The inclusive usage is entirely congruent with an older, Eurocentric division of White-dominated societies into Whites and non-Whites—only now, of course, the conversion of non-White to Black signifies an opposite valence.^[68]

The term Coloured is also a matter of sharp contention, reflecting

[67] We shall review population statistics more fully in Chapter 2. The racial breakdown (including all the so-called national homelands) is approximately as follows: Black Africans, 74.0 percent; Whites, 15.0 percent; Coloureds, 8.4 percent; and Indians, 2.6 percent.

[68] " . . . Black means anyone who is not White, anyone who is discriminated against because of colour." South African Students' Organization manifesto, quoted by Patrick Lawrence in Star (Johannesburg), February 16, 1972, and requoted in Brian M. du Toit, "Consciousness, Identification, and Resistance in South Africa," Journal of Modern African Studies 21, no. 3 (September 1983): 365–95, at 378.

unease with the externally determined name of the category but, for some, denial that the category should exist at all. The "Coloured" category is a complex product of contact among Europeans, Khoi, San, other Africans, and "Malays" brought by the Dutch from Indonesia. As in many other colonial societies with a significant number of Eurafricans, Coloureds were placed in a middle position between Europeans and Africans, their group boundaries, status, and prerogatives defined by the Europeans.^[69] The minimalist objection to the name Coloured is expressed in alternative terminology: the "so-called Coloured community." Maximalists are more likely to refer to so-called Coloureds as merely Black.

Were there only an objection to the name, a new name would surely be adopted. New names commonly are adopted—and even more frequently proposed—for groups in conflict, always with political purposes and often reflecting an aspiration for an improved collective status or a different conflict alignment.^[70] In South Africa, the term Afrikaner was originally employed as the rough equivalent of Creole —that is, locally born people—in contrast to the Europeans who administered the Dutch East India Company. The term Boer (farmer) was later used, but its backwoods, pejorative quality, when uttered by the British, rendered it obsolete for those who bore the designation, particularly since a movement to promote an "Afrikaans" language was then under way.^[71] In the same way, the offensive term Bantu was abandoned in favor of Black or African;^[72] and European , which might suggest a temporary status in South Africa, has given way to White . In the case of the so-

[69] Borrowing terminology from J. A. Laponce, The Protection of Minorities (Berkeley and Los Angeles: University of California Press, 1960), Pierre du Toit and François Theron suggest that the Coloured population is a "minority by force" rather than a "minority by will." "Ethnic and Minority Groups and Constitutional Change in South Africa," Journal of Contemporary African Studies 7, nos. 1–2 (April–October 1988): 133–47, at 142–43. For the emergence and nonemergence of comparable categories in the Western Hemisphere, see Donald L. Horowitz, "Color Differentiation in the American Systems of Slavery," Journal of Interdisciplinary History 3, no. 3 (Winter 1973): 509–41.

[70] See, e.g., Harold R. Isaacs, Idols of the Tribe: Group Identity Political Change (New York: Harper & Row, 1975), pp. 77–92; Donald L. Horowitz, Ethnic Groups in Conflict (Berkeley and Los Angeles: University of California Press, 1985), pp. 33–34.

[71] For the language movement, see Isabel Hofmeyr, "Building a Nation from Words: Afrikaans Language, Literature and Ethnic Identity," in Shula Marks and Stanley Trapido, eds., The Politics of Race, Class and Nationalism in Twentieth-Century South Africa (London: Longman, 1987), pp. 95–123.

[72] For African resentment of the Bantu label, see Melville Leonard Edelstein, "An Attitude Survey of Urban Bantu Matric Pupils in Soweto, with Special Reference to Stereotyping and Social Distance: A Sociological Study" (M.A. thesis, University of Pretoria, 1971), pp. 47, 137.

called Coloureds, name change appears precluded, at least temporarily, because of the larger South African dissensus: as indicated, some putative group members deny that the group should exist separately from the Black category. This difference reflects the fundamental conflict over the identification of the racial categories and the conflict over the character of the conflict.

Now, for social scientists, such issues pose some problems of epistemology, but they are not insoluble. Take the question of political party affiliations or group attitudes. Using sample survey instruments, it is possible to determine whether, for example, so-called Coloureds or Indians hold distinctive patterns of affiliations or attitudes or are regarded distinctively by others. Racial or ethnic identity is a matter of the interplay between self-definition and other-definition. The epistemological problem is to determine, in the first instance, who among the respondents is Coloured or Indian. If identification becomes too difficult, because social convention makes it so, it is then unlikely in any event that distinctive patterns would emerge in the data. If, however, respondent identification is possible and if distinctive patterns do emerge, social scientists might conclude that these categories of respondent were acting as if they constituted distinctive groups.^[73] As we shall see, research—including studies conducted by investigators located at widely distant points along the political spectrum—discloses some important, patterned differences among the four racial categories, which is why I need to retain separate terms for Coloureds and Indians.^[74]

Nevertheless, this does not resolve the matter for participants, for although I have labeled the issue a cognitive one, it is by no means exclusively that. Group names, as I have suggested, have an aspirational

[73] For the distinction between ethnic category and ethnic group, see Banton, Racial and Ethnic Competition, p. 105; R. D. Grillo, "Ethnic Identity and Social Stratification on a Kampala Housing Estate," in Abner Cohen, ed., Urban Ethnicity (London: Tavistock, 1974), p. 160. For the intricate relations between ethnic group labels and ethnic group behavior, see Frederik Barth, "Introduction," in Frederik Barth, ed., Ethnic Groups and Boundaries (Boston: Little, Brown, 1969), pp. 29–30.

[74] For clarity, I shall also reserve the term Black for Africans, and I shall use these two designations interchangeably. In this work, I shall also use the words race and racial for the differences demarcated by color. I shall use the terms ethnicity and ethnic when referring specifically to Afrikaners, English-speakers, Xhosa, Tswana, etc. This is not because race and ethnicity are, by themselves, qualitatively separate phenomena but because this is the usage in South Africa, where differences along color lines have been the more important ones. In referring inclusively to groups based on race and ethnicity, I shall generally use the term ascriptive, although in parts of the theoretical discussion not confined to South Africa the term ethnic will also be used. Finally, I shall refer to North Sotho as Pedi, to South Sotho simply as Sotho, and to Shangaan or Tsonga by either designation interchangeably, as is conventional in South Africa.

aspect. The cognition that Coloureds and Africans share a common fate is bound up with the hope for solidarity among the oppressed.

I use these terminological disputes to illustrate the fundamental dissensus with which I began. Nor have I exhausted them: there are Afrikaners who object to the use of the term African for Black Africans alone, on the ground that Afrikaners are White Africans.^[75] And, again, historical understanding is linked to group legitimacy, for to be called African is to belong, fully and properly, to Africa.

The possibility, raised by the Africanist stream, that Whites could ultimately become full members of the African community in South Africa has a close historical parallel in debates between Afrikaner leaders over who, among Whites, was a genuine Afrikaner. Many alternatives were advanced in the late nineteenth and early twentieth centuries.^[76] Among the possibilities, one, proposed by J. H. Hofmeyr, was a conception that would embrace English-speakers who had settled permanently in South Africa. For the next century, a much narrower definition, based on ancestry and the Afrikaans language, prevailed. The newer White category may have roughly the same inclusive effect as Hofmeyr's version of Afrikanerdom. But in a severely divided society, as White society alone was for much of the twentieth century, identifying terminology that includes the contending ethnic groups is foreclosed.

Every society with ascriptive divisions, as I have suggested, has such terminological issues, but South Africa's are more severe than those of most other deeply divided societies. A glance at one terminological issue in one such society is sufficient to make the point.

After Malaysia was created out of Malaya and two former British colonies in Borneo, the term Bumiputera was adopted to embrace Malays and the various indigenous (but not Malay) peoples of the Borneo states, in contradistinction to the Malaysian Chinese and Indians. Bumiputera means "son of the soil," and the clear implication is that non-Bumiputera (Chinese and Indians) are immigrants and, on that account, entitled to fewer of the rewards and opportunities the state might dispense. What began as an effort to find an inclusive term for indigenes, only some of whom were Malays, also proceeded on the assumption

[75] "After more than ten generations I regard myself as 'African' as any other African." Stoffel van der Merwe, "Not Just Two Groups," Sunday Times (Cape Town), August 9, 1987, reprinted in Giliomee and Schlemmer, eds., Negotiating South Africa's Future, p. 15.

[76] Giliomee, "The Growth of Afrikaner Identity," pp. 101–03, 105, 111.

that the term would be exclusive of others, and it ended with an implicit declaration of differential legitimacy underpinning differential access to certain "Bumiputera privileges."^[77]

How, then, is South Africa different? Not because of official categorizations: contrary to some popular South African myths, those are present in many countries, though rarely, if ever, with comparable disabilities attached to them. Rather, because official categorization in South Africa had much harsher results for life chances, the categories themselves came to be contested. In Malaysia, although Chinese and Indians always regarded the consequences of division into Bumiputera and non-Bumiputera categories as unfair, they acknowledged that the categories identified the main actors in a conflict-prone polity. While the argument was made, as it always is, that group leaders were merely using ethnicity to benefit from it, hardly anyone thought that this fact, if true, vitiated the operational significance of the categories for political and social life. In South Africa, by contrast, a strong body of opinion holds, as I have said, that the categories are incorrectly and illegitimately demarcated. The intensity of terminological dispute shows, above all, the unusual depth of cognitive and normative dissensus in South Africa.

The Burden of the Metaconflict

The conflict about the conflict makes at least three interrelated problems of severely divided societies more difficult in South Africa. First, the metaconflict places severe constraints on political discourse. Second, it contracts the range of acceptable innovations and future arrangements. Third, it increases the difficulty of finding ways for the parties to the conflict to seek accommodation and inclusion rather than hegemoney and exclusion. In Hanf's apt words, "The real load of conflict, already burdened by the huge disparities in power and prosperity, is enormously increased by the fundamental ideological cleavage[s]."^[78]

The most useful way to illustrate the constraints on discourse is to

[77] The Malaysian Constitution has, since independence, provided that indigenous people enjoy a "special position" and are entitled to superior access to scholarships, licenses, and employment in the civil service. Even before the merger with the Borneo states, who was a Malay was specified by the Constitution. Constitution of Malaysia, article 153, in Albert P. Blaustein and Gisbert H. Flanz, eds., Constitutions of the Countries of the World, vol. 9 (Dobbs Ferry, N.Y.: Oceana, 1988), pt. 12, pp. 153–56. For the official definition of a Malay, see article 160, in ibid., pp. 161–62. For its rationale, see K. J. Ratnam, Communalism and the Political Process in Malaya (Kuala Lumpur: University of Malaya Press, 1965), pp. 78–79.

[78] "The Prospects of Accommodation in Communal Conflicts," p. 111.

probe the significance of the term nonracial . The term arose in response to the emphasis of the government on ethnic divisions among Africans and racial divisions among everyone in South Africa. In assigning Africans to their national "homelands," the regime acted on the assertion that "there exist great differences between the Zulu and the Tswana, for example, or between the Xhosa and the Venda, differences more important than those between the Swedes and the Spaniards."^[79] The government may also have created some further divisions, such as the division between the two Xhosa "homelands" of Ciskei and Transkei.^[80] The whole policy of separate development assumes there is something separate to develop. The vision of a nonracial future is therefore diametrically opposed to the policy of divide and rule.

If the future is to be nonracial—not, it should be emphasized, multiracial, for multiracial is a term the regime has used—it must also be nonethnic. If nonracial merely meant nondiscriminatory, practically no one but avowed proponents of a return to pure apartheid would contest its usage. But it means more. At a minimum, it means that race and ethnicity can have no officially recognized part in public life and, for some, also that those affiliations should play no part, even informally, in political alignments or collective action. The nonracial society is the plural society's analogue to the utopian aspiration for a classless society. Both suggest so dramatic a transformation as to make the former society unrecognizable. The regime's contention that ascriptive groups are building blocks of the society is met with the contention that they cannot and should not be regarded as components of a good social order at all. So strong is the aversion to official group categories that the very existence of politically significant ethnic groups is denied. And denial is not too strong a term for describing the studied neglect of ethnicity that characterizes current discourse in South Africa.

Of course, South Africans are not alone in rejecting the significance of ethnicity. Western scholarship often sees ethnic affiliations as a mask for what are really class affiliations, as diversions from more important forms of conflict, as a cover for elites to secure support, as just about anything but ethnic affiliations—that is, anything but sources of com-

[79] P. W. Botha, quoted in Le Figaro (Paris), December 5, 1986, as requoted in Marianne Cornevin, "Populations noires d'Afrique du Sud," Afrique Contemporaine, no. 141 (January 1987): 34–49, at 34.

[80] See "Ethnicity and Pseudo-Ethnicity in the Ciskei," in Leroy Vail, ed., The Creation of Tribalism in Southern Africa (Berkeley and Los Angeles: University of California Press, 1989), pp. 395-413.

munity (as well as conflict), familiarity (as well as distance), and assistance (as well as discrimination). The reasons for this bias against ethnicity run deep. They relate to the general disfavor in which ascriptive affiliations of all kinds are held in the West and to the growth of materialist explanations for social phenomena. But the confluence of South African opposition with this more general stream of scholarship has produced something close to silence on ethnicity in South African scholarly and political discourse.

When scholars tread close to ethnicity, they deal in euphemisms or walk on eggs. The "Eastern Cape" is "an ANC stronghold"; the Xhosa who live there go unmentioned.^[81] In 1976, "a section of the migrant worker population of Mzimhlope Hostel in Soweto"^[82] killed young people in its path; Zulu go unmentioned.^[83] The odd reference to ethnicity typically explains how it will be overcome: "the uniform exploitation of African workers at the bottom of the pay scale and labour hierarchy promises to become the binding grievance that bridges black heterogeneity."^[84] Careful researchers who find ethnic phenomena in their data weave them skillfully, if obliquely, into their explanations, so that they will be available to sharp eyes and alert minds.^[85] Rarely is there explicit discussion of even the possibility of protecting Black ethnic minorities.

As is common in divided societies, there is a belief that talking about ethnicity creates or reinforces ethnic divisions—in South Africa, that such talk does the government's dirty work for it—even when the talk is directed at how to prevent such divisions from overwhelming a future democratic state. This is a notion that, in some measure, spans the otherwise significant chasm between the extraparliamentary and the parliamentary opposition. When, in a speech at the Institute for a Democratic Alternative for South Africa (IDASA), an Afrikaner academic suggested

[81] Karis, "Revolution in the Making," p. 391.

[82] Lodge, Black Politics in South Africa since 1945, p. 329 (emphasis supplied).

[83] To the same effect, see Martin E. West, "The 'Apex of Subordination': The Urban African Population of South Africa," in Price and Rosberg, eds., The Apartheid Regime, p. 140.

[84] Heribert Adam, "Variations of Ethnicity: Afrikaner and Black Nationalism in South Africa," in Anand C. Paranjpe, ed., Ethnic Identities and Prejudices: Perspectives from the Third World (Leiden: E. J. Brill, 1986), p. 43.

[85] Cf. William Beinart, "Worker Consciousness, Ethnic Particularism and Nationalism: The Experiences of a South African Migrant, 1930–1960," in Marks and Trapido, eds., The Politics of Race, Class and Nationalism in Twentieth-Century South Africa, p. 306.

that South Africa might actually house more than one nationalism that would have to be accommodated, he was advised that this kind of thinking was "exactly what IDASA was trying to get away from."^[86]

In Chapter 2, I shall provide evidence that ethnic heterogeneity is likely to have some political significance in a fully democratic South Africa. There is now a fairly strong taboo on public discussion of whether such a forecast is accurate and, if it is, what the consequences might be.

The result of this taboo has been a paucity, albeit fortunately not a complete absence, of research on demographic, attitudinal, and behavioral differences among Ndebele, Pedi, Shangaan (Tsonga), Sotho, Swazi, Tswana, Venda, Xhosa, and Zulu. Research on this subject, elsewhere recognized as increasingly important, is in South Africa vulnerable to the suspicion that it could foment ethnic divisions where no significant divisions exist. From research on divided societies, it is well established that the context of interactions, including the political context, affects the operative level and felt intensity of group identities.^[87] Under apartheid, race has been preeminent. After apartheid, this could change, and African ethnic identities could become more important. But the ideological reaction to apartheid has been such that the matter has not been considered in these terms. Ideology about the conflict presses in on understanding of the conflict at every level.

The metaconflict, then, can be a constraint on imagination and a constraint on innovation. The response of analysts to the two constraints should be different. As I shall suggest, to defer to ideas about the conflict, even mistaken ideas, in institutional innovation is to acknowledge the power of ideas to defeat the work of otherwise well-designed institutions. To defer to mistaken ideas about the conflict in analyzing the conflict—except to the limited extent that ideas about the conflict actually remake the conflict—is to perpetuate illusions. In the ethnicity example I have provided, these could turn out to be very dangerous illusions for the polity. And so the analyst who aims to ameliorate the conflict is confronted with the difficult task of combining relentless pursuit of the truth about the conflict—and persuasion of the participants that it is the truth—with deference to their sensibilities when the time comes for design.

As I have indicated, in the field of ethnic and racial conflict, the ideas

[86] Alex Boraine, quoted in Pierre de Vos, "NP Shift Poses New Challenge," Democracy in Action (Cape Town), May 1989, pp. 1, 5, at p. 1.

[87] See, e.g., Crawford Young, The Politics of Cultural Pluralism (Madison: University of Wisconsin Press, 1976), pp. 114–21.

of the participants have only limited ability to reshape the conflict. To demonstrate that point would require a volume in itself. An illustration of the repeated failure of a remarkably resilient idea will have to suffice.

The history of efforts to organize noncommunal socialist parties in divided societies—efforts based on the premise that the "real" divisions are class rather than ethnic divisions—testifies to the ability of ascriptive affiliations to overcome such ideas. The reason is, in part, that conceptualization exclusively in terms of class is disproportionately the propensity of educated elites. The participants, especially those at lower levels, also entertain other ideas, some of which are at odds with this conceptualization of the conflict. In the daily interactions that shape political cleavages, they are more likely to act on the basis of those other (smaller) ideas and on the basis of interests that hardly rise to the ideational level at all. And, cumulatively, a dynamic is set in motion, so that, before too long, socialist parties are either monoethnic, de facto ethnically based parties or they are out of electoral business altogether.^[88] No taboo on discussion or research can change this dynamic where the ingredients for it are otherwise in place. The participants' views of the conflict feed into it, but the conflict cannot be reduced to what some of its more organized and articulate participants think it is about. The dynamics of the conflict, even apart from the discourse, place people in positions and relationships that exercise an independent influence on the course of politics.

If the metaconflict should not inhibit imagination, it nevertheless must, as I have said, constrain innovation. The feasibility of what might be desirable political arrangements is affected by the way interested parties understand and characterize the conflict. Some institutions that have operated well in environments more frankly hospitable to the legitimacy of group identities will be regarded as illegitimate ab initio in a South Africa where some of the major actors identify groups only with discrimination and are willing to accommodate difference only if it works no fundamental contradiction to their way of seeing the world.

No doubt some actors can and will modify their ideological predispositions. Learning takes place as events proceed. But it is fanciful to assume that ideologies developed over decades will be transformed during the shorter period in which the adoption and implementation of fundamental constitutional change takes place. That does not obviate

[88] The evidence is assembled in Horowitz, Ethnic Groups in Conflict , pp. 9–10, 334–40.

the need to argue the merits of conflict-reducing institutions. Without a good deal of candid argument and a generous measure of persuasion, the South African future will almost surely be very bleak. But it would be a grave mistake, even if it were possible, to force those institutions on unwilling participants as the price of change or the price of peace, for that is virtually to invite them to delegitimize the institutions, to accept them insincerely, to play against their rules, and to overthrow them when they can. I shall suggest later that the first set of inclusively democratic institutions is likely to assume inordinate importance for the future of democracy. If that is so, grudging acceptance as the prelude to something else is not good for democracy. To insure durability, institutions to effectuate democratic consent need to aim at, among other things, genuine consent in the adoption.

It follows that those who propose institutional innovations for South Africa bear an unusual burden. It is not adequate to be correct about the conflict. That is difficult enough in divided societies. In addition, it is necessary to be correct in relation to the metaconflict—that is, to propose nothing that is at the center of the conflict of visions, for that, like the conflict of races, needs to be ameliorated. This emphatically does not mean that innovations should somehow split the difference, for such an approach could produce a grotesque combination of institutions, ineffective for accommodating the underlying conflict. What is required is the design of institutions drawn from the realm of what is acceptable.

To take a crude example of the extra constraints imposed by the metaconflict, I shall note in Chapter 5 that the former Lebanese electoral system, based on seats reserved for candidates of particular ascriptive identities, performed accommodative functions in Lebanon. The usual argument against ethnically reserved seats in divided societies is the a priori assertion that such arrangements entrench or perpetuate the divisions they seek to heal. In the case of Lebanon and in many other such societies, the argument is without substance. Lebanon disintegrated despite, rather than because of, these arrangements. But no Lebanese-style electoral system could perform similarly accommodative functions in a South Africa that has had more than enough of officially designated, group-based institutions.

Innovations always must meet the dual standards of aptness and acceptability. In no society are the two coterminous, but in South Africa, the gap between what would be apt and what would be acceptable is often very wide. Constitutional engineers should tell South Africans which

bridges will and will not stand, but in designing bridges for South Africa they will need to take the idiosyncratic South African institutional aesthetics into account. This, of course, is not easily done, since the aesthetics are themselves in conflict: for example, one is averse to official formulations based on ascriptive groups, while another is averse to the abandonment of group identities. This is a central burden of the conflict over the nature of the conflict.

All of this leads to the final and most serious potential burden of the metaconflict, which is perhaps best stated as a question. Does the formidable degree of ideological dissensus incline the participants in the ideological contest to hegemonic rather than accommodative aspirations, rendering democracy even more difficult to attain?

A willingness to share power or alternate in power is necessary for democracy to succeed. In a very important paper, Pierre du Toit argues that such a willingness is not likely to emerge in South Africa.^[89] In severely divided societies, he suggests, contenders for power are engaged in a contest for hegemony based on competing models of the regime they prefer. If the models are incompatible, as they are in South Africa, no constitutional scheme alone will be sufficient to insure democratic outcomes. The aim is simply to capture the state for one's own side. Examples might be the government's recurrent efforts to share power without losing it or the ANC's oft-stated position that it was prepared to negotiate only the transfer of power to itself, as agent for "the people."

The appropriate analogy for du Toit is the international system. There regulatory regimes can gain the consent of contending actors once the application of strategic rationality leads them to "a common perceptual frame about the nature of the conflict, which complies with the requirements of cooperation theory. . . ."^[90] In severely divided societies, there is no "community of consent about the basic structure of society," and "the opposing parties do not share a common perceptual frame through which to assess societal conflict"^[91] —which is to say that they are engaged in a conflict about the conflict. This is because no learning process, comparable to that which produces international cooperation, has taken place. What is needed, therefore, is a process of what du Toit

[89] Pierre du Toit, "Contending Regime Models and the Contest for Hegemony in Divided Societies" (University of Stellenbosch, 1989; unpublished paper).

[90] Ibid., p. 14.

[91] Ibid., p. 19. In du Toit's formulation, the lack of a common perceptual frame is a feature of all divided societies, but I would argue that the extent to which this is so is a variable and not a constant in such societies.

calls, significantly, "meta-bargaining, or 'bargaining about bargaining,'" to produce "a joint commitment by all contenders to the view that the conflict cannot be won by either party on its own terms but instead, that a mutually profitable, and therefore mutually acceptable, settlement should be sought."^[92] Without that commitment, the antagonists will see the choice as being merely to dominate or to be dominated and so will engage in behavior that aims at hegemony.

As the term metabargaining suggests, it is not, for du Toit, the mere existence of serious racial or ethnic divisions that creates the special problem, but the contending regime models, which in South Africa he reduces to liberal democracy or polyarchy, with its emphasis on procedure, versus "populist democracy," including "the deployment of authoritarian methods by a pro-active state, or even the establishment of an authoritarian regime, in the pursuit of substantive justice in the post-apartheid South Africa. . . ."^[93] The difficulty, in short, derives from the conflict compounded by some version of the metaconflict, the conflict of ideologies and visions. No settlement is likely, even if there is formal agreement, so long as one contender suspects another of harboring an incompatible vision.

How, then, to achieve what du Toit calls a community of consent about the threshold question of the regime model entertained by the contenders? The process he advocates entails a blend of learning, deterrence, and confidence building. In international metabargaining, the adversaries begin to understand the high costs of a contest with an uncertain outcome; they slowly create a zone of overlapping rather than zero-sum agendas; and the contest for hegemony is ultimately dissolved. On this view, the task of the constitutional engineer, even before devising arrangements to make liberal democracy work in the face of ethnic or racial conflict, is to unravel for the participants the social costs of hegemonic aspirations—to facilitate, in other words, the learning that leads to deterrence, which leads in turn to a widened zone of agreement. The constitutional engineer is perforce engaged in "demystification"—a task that should not be impossible, since "rewards for cooperation, and the inability to escape the retaliatory consequences of extremism and unilateral actions, ought to be even more compellingly obvious within divided societies than within international society."^[94]

Perhaps they ought to be, since the contenders live at closer quarters.

[92] Ibid., p. 20.

[93] Ibid., p. 17.

[94] Ibid., p. 23.

Yet, the experience of South Africa suggests that these lessons will not be easy to learn. The argument can easily be made to sound tautologous. The society is severely divided because the actors believe the myths; if the myths could be dissolved, the society would not be severely divided in the first place, so a learning model simply wishes away the problem.

Moreover, there are other models of how liberal democracy can come into being in a society as severely conflicted as South Africa. The growth of international regimes in an otherwise anarchic environment itself suggests a different account. There the contenders learn to modify their goals through long and bitter experience, not through mere counsels of demystification. If the international analogy is truly apt, it might imply the unavailability of any rapid path to a democratic outcome in South Africa. It seems doubtful that the learning process can be short-circuited. Alternatively, perhaps engineering has a design role to play, rather than a demystification role, on these ideological issues as well as on those that derive from ethnic divisions per se. Perhaps institutions can be designed to prevent hegemony or to make it very costly to attain once a contender subscribes to the institutions. If so, potential hegemons could become cooperating democrats in spite of themselves.

As these speculations imply, models of the transition to democracy are indeed various. In Chapter 7, where we return to this question, we shall examine some alternative models. The point here is to raise the crucial community of consent issue as separate and to underscore that it arises as much from the metaconflict as from the conflict itself.

The Path from Subordination and the Functions of Ethnicity

Up to this point, we have concentrated more on the metaconflict than on the conflict itself. When we focus on the conflict in comparative perspective, the gap between the two becomes apparent.

South African society is not difficult to classify. It is characterized, above all, by what is appropriately called ascriptive ranking . There are superordinates and subordinates, largely defined by birth criteria. To be sure, within the ranks of each stratum, there are also cleavages that divide, in some variable measure, Afrikaans speakers from English speakers, Zulu from Xhosa and Tswana, and so on. But the overall design of the society is predicated on racial hierarchy, and the significance of those alternative cleavages is, at least temporarily, suppressed.

We can identify other societies organized along similar lines, and we can trace their similar behavioral propensities: the enforcement of an etiquette of subordination, the denial of prestige and mobility opportunities to ranked subordinates, the social impermissibility of intermarriage and interdining, and what Max Weber calls the "acknowledgment of 'more honor' in favor of the privileged caste and status groups."^[95]

In the United States South and the West Indies, as in South Africa, the indicia of difference were color and physiognomy, but in caste-ridden India, a more complex set of indicia of birth-based membership is employed. In Japan, differences between the subordinated Burakumin and other Japanese can be detected only through clues about family origin, which led to the description of the Burakumin as Japan's "invisible race."^[96] In Rwanda and Burundi, yet other indicia of differences are employed.^[97] Color is a powerful differentiator, to be sure, but in analyzing systems of subordination one should not fall victim to the parochial fallacy that mistakes the indicator of group identity for the relationship itself. Many differentiating attributes will do just as well as color, as the victims of caste violence in India or of ethnic violence in Burundi will attest. One should not succumb to the assumptions of the system of subordination in reverse and fall prey to what I call the figment of the pigment.

Slavery on an ascriptive basis, an institution formerly well known in South Africa, has been far more widespread than we might like to think. In many places, it forms the basis for post-emancipation superior-subordinate relations.

A somewhat obscure example is enough to demonstrate the persistent regularities that characterize such relationships. Among the Ibo of Eastern Nigeria, two different kinds of slavery were practiced until fairly recently.^[98] One involved the Osu, slaves serving deities that protected villages. The other involved the Ohu, whose servitude originated in con-

[95] H. H. Gerth and C. Wright Mills, eds., From Max Weber: Essays i Sociology (New York: Free Press, 1958), p. 189.

[96] George De Vos and Hiroshi Wagatsuma, eds., Japan's Invisible Race (Berkeley and Los Angeles: University of California Press, 1966).

[97] See René Lemarchand, Rwanda and Burundi (New York: Praeger, 1972). See also Reginald Kay, Burundi since the Genocide (London: Minority Rights Group, 1987); René Lemarchand and David Martin, Selective Genocide in Burundi (London: Minority Rights Group, 1974); Leo Kuper, The Pity of It All: Polarization of Racial and Ethnic Relations (Minneapolis: University of Minnesota Press, 1977).

[98] M. M. Green, Ibo Village Affairs , 2d ed. (New York: Praeger, 1964), pp. x, 12, 23, 50–51, 58, 62, 158–59; W. R. C. Horton, "The Ohu System of Slavery in a Northern Ibo Village-Group," Africa 24, no. 4 (October 1954): 310–36.

quest (or sometimes conquest and payment). In both cases, slavery was intergenerational. Osu and Ohu alike were residentially segregated, either in peripheral areas of a village or in separate villages. Intermarriage was strictly forbidden and regarded with utter repugnance by other Ibo. Ohu and Osu were treated as property rather than persons, and Osu were occasionally used as human sacrifices. When the system was formally abolished in 1956, proponents of the abolition explicitly drew the parallel to South African apartheid. After abolition, escaping the subordinate identity still proved virtually impossible, even by migration to more anonymous, urban communities. Even where Ohu had served as priests or where Osu had amassed wealth, they were denied the respect generally accorded ordinary Ibo. Much to the discomfort of progressive Ibo, the stigma of subordination has endured.^[99]

At various times in the modern period, slavery and subordination based on ascriptive criteria have been present on every inhabited continent except possibly Australia. None of this mitigates racial domination in South Africa or elsewhere; but there is no basis at all for the view that Whites have a monopoly on this kind of behavior. No one familiar with even the outlines of, for example, Burundian ethnic relations could entertain such a view. René Lemarchand has described Burundi as "the one state in Africa that displays the most systematic and blatant violations of human rights," a state in which the Tutsi minority of 14 percent (about the same percentage as Whites in South Africa) controls well over 90 percent of all high governmental, political party, foreign service, private sector, educational, judicial, and health care positions, not to mention its 99.7 percent of all military positions, top to bottom.^[100] In this setting, the assertion that Tutsi do not "claim the right to dominate power, still less to monopolize it" but "are ready to share power with their Hutu brothers, if the act of sharing does not constitute a threat to their own safety,"^[101] has a familiar, Pretorian ring to it. There

[99] See, e.g., Igwebuike Romeo Okeke, The "Osu" Concept in Igboland: A Study of the Types of Slavery in Igbo-speaking Areas of Nigeria (Enugu, Nigeria: Access, 1986). For the debate on abolition in the former Eastern Region House of Assembly, see ibid., pp. 142–59.

[100] René Lemarchand, "Burundi: The Killing Fields Revisited" (University of Florida, 1989; unpublished paper), pp. 1, 4.

[101] Juvénal Mugiraneza [pseud.], The Origins of the Ethnic Problem in Burundi (Bujumbura, Burundi: n.p., 1988; pamphlet), p. 6. In this twelve-and-one-half-page, progovernment pamphlet, the apprehension of a Tutsi genocide by Hutu is pervasive. The word extermination appears at least eight times, synonyms like final solution, genocide , and massacre many more times. Compare the Afrikaner apprehension of extermination; see note 24, above, and accompanying text. I am indebted to René Lemarchand for a copy of this pamphlet, but the interpretation is my own.

is, then, a common behavioral tendency to subordinate others, a tendency that bitterly reaffirms the common humanity and the common failings of Whites and Blacks—and Asians as well.

Rudimentary knowledge of the common phenomenon of ascriptive subordination may be of small comfort to those who are subordinated. But it ought at least to enable us to think more coolly than has generally been done so far about how to change the status quo in ways that promise decent, peaceful relations in the future. One thing it means is that we can learn from the experience of those who have tried to move to a different system. When we do this, we shall see that learning to live without predictable relations of superordination and subordination is not the same as living without ascriptive groups at all.

The rather antiseptic description I have provided to this point serves the purpose of countering the inevitable parochialism that comes with immersion in a particular conflict. South Africa is certainly not unique in enduring a harsh system of ranked subordination, but it has had a far more elaborate legal apparatus deployed in the service of ranking than other systems have. With respect to the Burakumin, the Japanese have never had a formal apparatus of discrimination comparable to apartheid, although a cynic might say that the tight structure of Japanese society hardly requires legal sanctions. There are, for instance, published directories of Burakumin names available to employers who wish to discriminate against them. In Burundi, where the oppression of ranked subordinates has been far more lethal than almost anywhere else, South Africa's efficient legalism has nevertheless been absent.

Once we place South Africa in the category of ranked systems, we can see why it is in a process of change. Above all, the source of change lies in the realm of ideas. The spread of conceptions of human equality around the globe has made ascriptive stratification ideologically obsolete. Those who practice it must pretend that they do not. Even in the Jim Crow American South, the claim was not "separate and appropriately inferior"; rather it was "separate but equal." The importance of the international diffusion of ideas is attested by the adoption of the tactics of the American Civil Rights Movement—with a good many Japanese refinements!—by the Buraku Liberation League in Japan,^[102] and it is reaffirmed by the very name of such groups of Indian ex-Un-

[102] Frank Upham, Law and Social Change in Postwar Japan (Cambridge: Harvard University Press, 1987), pp. 78–153.

touchables as the Dalit Panthers, obviously modeled on the Black Panthers.

Several paths lead away from relations of ascriptive subordination, but the path leading to a wholly nonascriptive society is the path least traveled, even when it is the path on which many wish to journey. Far more common is an improvement in the status, wealth, and power of the formerly subordinate group, so that now it encounters other groups, not from a fixed position of submission and inferior ranking, but from a more fluid position, in which individual group members may rank low or high but in which the groups themselves are unranked in relation to each other. The relations between groups have changed from ranked to unranked, but groups still exist.

It is easy to see why the shift from ranked subordination to unranked group relations is so common. When subordination on a birth-determined basis is undermined, subordinated people challenge the system of ranking by collective action, by action on the basis of group identity, even as they deny its importance. Group networks are not abolished with a change of ethnic status, particularly as endogamy, which perpetuates group identity and recruitment by birth, is unlikely to change over the short term. Moreover, in post-subordination encounters among members of various groups, intergroup stereotypes, divisions of labor, and divergent patterns of political interest all persist.

In the United States and in India, there have been considerable changes in systems of ascriptive subordination. Although the officially declared aim for race relations in the United States and for caste relations in India was an end to public action on the basis of ascriptive identity—in American parlance, a color-blind society, and, Indian terms, a society free of casteism—the pursuit of equality has taken a rather different form. Although subordination is no longer the norm, or at least it is no longer the exclusive norm, ascriptive identities have not been abolished. Race and caste groups still encounter other race and caste groups as groups, but the behavior, the rights, the claims of the former subordinates, are no longer limited and rigidly specified.

The paths to status enhancement in India and the United States have been somewhat different.^[103] But in both, intergroup interactions are

[103] For recent American trends, see Howard Schuman et al., Racial Attitudes in America (Cambridge: Harvard University Press, 1985); Reynolds Farley, Blacks and Whites: Narrowing the Gap? (Cambridge: Harvard University Press, 1984). For two aspects of the complex trends in India, see Marc Galanter, Competing Equalities: Law and theBackward Classes in India (Berkeley and Los Angeles: University of California Press, 1984); Owen M. Lynch, The Politics of Untouchability: Social Mobility and Social Change in a City of India (New York: Columbia University Press, 1969).

more complex and less predictably unequal than they formerly were. Groups have more autonomy than before, and now they function in a system of unranked relations that has far greater legitimacy than the former system of subordination could have in the egalitarian ideological climate of the present day. Race and caste groups remain self-conscious entities capable of political mobilization and collectivities to which members can repair for mutual support.

Since not everyone believed that race and caste groups were truly equal, the initial consensus that race and caste should disappear was far from complete in the United States and India. The persistence of race and caste groups owes something to the incompleteness of the consensus and to the persistence of discriminatory behavior.^[104] But continuing inequality is not the whole explanation for the durability of ascriptive groups. For virtually everywhere—and especially in a mobile society like South Africa's—ethnicity serves a number of useful functions. These range from assistance far from home to a sense of familiarity and neighborliness, even fictive kinship, and diffuse obligations in what can otherwise be a rather unfriendly and impersonal world. More than this, there is a strong human tendency in a common environment for people to define themselves by contrast and to seek confirmation of their collective worth. It is no accident that, given an opportunity to evaluate themselves on a survey instrument, virtually all ethnic groups everywhere rank themselves first.^[105] Nor is it surprising that, in the quest for collective well-being, ethnic groups should stereotype themselves and others, should feel closer to some groups, further from others, and even hostile to others.

These tendencies have been documented in scores of environments. They are as applicable to South Africa as to other countries, and they will persist long after apartheid is gone. No doubt the creation of the homelands has induced a good many Black South Africans to react by relegating their ethnic identity to the equivalent of their mental attic, putting it in deep storage. They reject the stigma of belonging to a "tribe,"

[104] For surveys on White attitudes to racial integration in the United States over time, see Schuman et al., Racial Attitudes in America , pp. 74–75. For the early movements to abolish caste distinctions and the civil rights legislation that followed, see Galanter, Competing Equalities , pp. 23–44.

[105] For perfect South African consistency with the general rule, see J. W. Mann, "Attitudes towards Ethnic Groups," in Adam, ed., South Africa: Sociological Perspectives , pp. 53–54. Further South African evidence on this is reviewed in Chapter 2.

an affiliation regarded as premodern, even "primitive." In this respect, apartheid has ill prepared South Africans for the future. Ethnic affiliations coexist very nicely with modernity all around the world. If the Québecois, the Basques, the Walloons, the Sinhalese, the Berbers, the Yoruba, the Ilocano, and the Kikuyu can all retain their ethnicity, why cannot South Africans equally belong to an array of ethnic groups, recognizing full well that, without precautions, interethnic competition can turn to severe conflict? The only reasons I can think of are the regime's effort to use such affiliations for its own purposes and the overarching present significance of color and race rather than ethnicity per se.

Nevertheless, affiliations now in the attic can be dusted off. Politics all over Africa—and nearly all over the world—has a strong ethnic component. With some qualifications, what is true of Zimbabwe, Nigeria, Zambia, Kenya, and Mauritania is also likely to be true of South Africa.

If that is so, is the aspiration to a nonracial, nonethnic South Africa doomed? If that aspiration requires that racial and ethnic groups disappear from the social and political landscape, this they will surely not do. If, however, the aspiration means that no official categories and no state institutions based on racial and ethnic groups shall be erected, that is an aspiration quite compatible with life even in a thoroughly plural society. As we shall see, it is possible to devise a range of political innovations apt for South Africa's ascriptive conflicts and located well within the narrower range of acceptability dictated by South Africa's ideological conflicts. There are nonracial, nonethnic institutions available to serve the multiracial, multiethnic society that South Africa will be for the foreseeable future.