South African Revisionism

South African labor history has been remarkably rich over the last three decades, and there is an interesting, and familiar, history to the writing of this history. The resurgent Marxism in Britain, France, and the US in the late 1960s encouraged a group of expatriate scholars (inside and outside the South African Communist Party) to issue a devastating “revision” of the liberal thesis that capitalist development would eventually lead to the dismantling of racist institutions in South Africa.¹

Inside the country several prominent scholars crossed the boundary between the unions and the academy in the 1970s, and a set of journals, notably the South African Labour Bulletin, fostered direct connections between shop-floor organization and academic scholarship.² By the start of the next decade a second generation of penetrating and evocative social histories of the urban African working class had been published.³ At this time the relationship between the, mostly white, revisionist social historians of labor and the vibrant and powerful and, mostly black, union movement was always complicated, with union organizers and academics finding abundant opportunity for mutual admiration and disappointment.

By the middle of the 1980s the historians of the factories and mines had begun to creep up the battered roads of the rural labor reserves to explore the cultural roots of the African proletariat. South African researchers, as my colleague Jeff Guy has often noted, went in search of the political consciousness of the proletariat and instead found ethnicity and witchcraft. In the process the idea of the “pure worker” began to dissolve in the face of “workers who are not only members of a class, but also of communities, located in a specific national past, and culture.”⁴

At the same time, driven by the terrible consequences of forced removals, influx controls, and systemic poverty for the women and children trapped in the bantustans, a powerfully inflected gender analysis began to remake the male story of class formation and struggle.⁵ This attention to the rural world of the urban African working class became, very quickly, a rich and sophisticated historiography (imperfectly captured in a special edition of the Radical History Review in 1990 devoted to “History from South Africa”). By the end of the 1990s South African social history, exemplified once again in Charles van Onselen's work, had pieced together stunning connections between the intimate familial politics of the forgotten rural poor and the changing political economy of racial capitalism.⁶

What is striking about this historiography is that practically the entire corpus—and it is very large—has been driven by class analysis.⁷

There are many reasons for this. South African historiography was heavily influenced by the flowering of Marxist social histories and theory in Britain in the 1960s and 1970s. The particular explanatory hold of Marxism in South Africa had much to do with the ideology of racial and cultural difference at the heart of Verwoerdian apartheid, which makes it very difficult to make or endorse appeals for cultural or epistemological rupture. Similarly, the “eruption of race” into class in the north Atlantic, marked by the interventions of Stuart Hall, Paul Gilroy, and David Roediger, has had a much smaller impact precisely because South African historians have been preoccupied with white working-class racism since the early 1960s. A theory of racial capitalism (combining Franz Fanon and Karl Marx in almost equal measure) lies at the foundation of the revisionist understanding of South Africa. Nor is this concern with class analysis the exclusive preoccupation of white scholars. African historians, often more closely associated with a liberation movement that was itself heavily influenced by the Communist Party, have often adopted an even more determinedly universalizing and materialist Leninist analysis.⁸

These are important exceptional features of our national historiography, but there are a few more. The most compelling explanation for the enduring power of class analysis in South Africa may simply be that capital, in the form of the gigantic mining houses, has been such an intense political and economic force here. This fact, in combination with the ubiquity of race and racism, has tended to mean that class analyses have almost always had revelatory power. And the political vitality of the organized working class for much of the twentieth century clearly nurtured the field of labor history and, later, of working-class social history.

But what will happen now? In recent years the political nature of class in South Africa has undergone some astonishing changes. In addition to the basic demographic decline in the size and power of the industrial working class, there has been a bewildering reorganization of capital. The renaming and movement of the major share-holdings of Anglo-American, De Beers, Old Mutual, Gold Fields, and Gencor to London has combined with a significant intrusion of foreign capital into fields, like newspaper publishing, that have traditionally been the public face of capital. The state, and the large mining interests themselves, have been vigorously pursuing an empowerment strategy that has already had the effect of creating very powerful and profitable gold mining houses that are black-owned and controlled. Most disconcerting, the trade union movement has itself been a major investor in the creation of these black capitalist interests, and many of the most important union activists have become influential capitalists. Small wonder that the union movement finds itself fending off charges that it now protects the domain of a privileged (largely male and urban) minority against the infiltration of the (mostly female and rural) majority. Class is certainly not what it used to be.

Today these problems have been played out across the gap between town and country. Many of the rural poor are still trapped by the geography of Apartheid and the demise of industrial migration, in densely settled rural slums with little in the way of infrastructure and very few services. The HIV crisis is both effect and cause of the ongoing weakness of basic social services in these regions. But the rural residents of the former Bantustans can (and do) vote, and contemporary South African politics does require every major national politician to address their concerns. A more powerful and expanding social boundary (fostered by a state that is increasingly effective and hegemonic in the cities) lies in the division between citizens and illegal immigrants. The great majority of those who fall into this last category (universally described by the epithet Makwerekwere, an onomatopoeic noun derived from the accented speech of foreign Africans) come from rural communities beyond our borders that have been providing labor to the South African cities for over a century. A key feature of post-Apartheid South Africa, like many other industrial economies, is that the urban-rural divide increasingly straddles our national borders.

So this is an appropriate time for South African historians to consider the question that Geoff Eley and Keith Neald have recently asked: Is it time to bid farewell to the working class? Has the Marxist interest in the political struggles of production-defined classes been replaced by research into the broader effort: To use their slightly cumbersome phrase—to “hegemonize identity?” Eley and Neald, it must be said, lay out an admirably ecumenical effort to end what Cooper calls the “left's culture wars.” Theirs is a call—to choose an example from the populist struggle against the Apartheid state—for a United Democratic Front of Marxist, Foucauldian, and Derridean historians, and an argument for a kind of social history that considers the state, gender, race, sexuality and class as conceptual peers. The terrain of this scholarship would cover discourse, interiority, epistemology, and political and economic structure. Above all, class becomes one variably defined element, amongst many, and modern global capitalism would be better understood as a complexly oppressing, exploiting and enriching system that bares little correspondence to the neat class dialectic of the (old) Labor Party politics.⁹

There are moments when it seems almost as if Eley and Neald are describing something that has already happened. But it is significant that they have been attacked by representatives of each of the contending “sides” for proposing a resolution. There is something particularly compelling about Nancy Fraser's warning that the effort to combine the materialist and culturalist critiques of modern societies “courts the risk of philosophical schizophrenia.”¹⁰ Before we risk our mental health any further, I want to subject South African class-based histories to a set of tests, that, I hope, will show that our current use of class already adopts the kind of theoretical and empirical promiscuity that Eley and Neald recommend. And I hope to show that this promiscuity follows directly from the effort to find the urban industrial working class in the countryside.

The questions are straightforward. Does the study of class help us to make sense of the very real and very modern problems of social and economic exclusion? Can it also address the origins and politics of gender, ethnic and sexual identities in the chaotic mix of modern industrial capitalism? Will it account (and suggest remedies) for the misrecognitions, and mal-distributions, that distort our societies? I want to consider these issues by looking at the social history of migration to my own home town.

Rural in the Urban: The Durban City-State

I live in the city of Durban. This is a beautiful place, with long, white sandy beaches along the warm Indian Ocean and a climate that is balmy all year round. The city is a tourist mecca, particularly for lower-middle class South Africans and their European counterparts. It is also a major industrial center in its own right. Durban has one of the largest harbors in the world, and serves as the primary point of entry for high-value South African trade. Yet eThekwini—to give the municipality its isiZulu and official name—must also be one of the most rural cities in the world. Large and densely settled tribal reserves (established by the British colonial authorities in the nineteenth century as domains of African homestead production and taxation) abut the entire perimeter of the city. Huge tracts of these territories—indeed almost all of them—have recently been incorporated within municipal boundaries that now have a radius of some eighty-five kilometers. This means that the city of Durban, much more than the province of KwaZulu-Natal, has become the primary focus of economic development and political mobilization for the three million people who live in its shadow.¹¹

The city has a long history, dating back to the 1840s, as a source of work and identity for people moving from rural homes. Some of these rural homesteads were in the adjacent reserves, others much further away in northern Zululand, southern Mozambique, or the Mpondo territories of the northern Transkei, and yet others from the distant villages of Madras in south India. Surrounding the city, scattered between the colonial labor reserves, are the rolling hills and stately mansions of massive sugar plantations. Between the sugar estates, the colonial reserves and the city, Durban has attracted some three-million people with a racial mixture that is unique; more than half of these residents are of African descent, about a million trace their origins to the immigration of Indian indentured workers in the nineteenth century, about half a million are the descendants of English-speaking settlers, and the remainder fall into smaller Colored, Mauritian, and even Zanzibari communities. Urban Apartheid has left the city with a landscape fundamentally divided by race, but the three largest groupings fragment easily, and frequently, along lines of linguistic, religious, or regional affiliation. Racial identities are both overwhelming and unstable. Culture and identity are political dynamite in this city. The tensions between rural poverty and urban wealth form part of the daily experience of all its citizens.

It is for these reasons that I think that contemporary Durban makes a good test of the contemporary powers of class analysis. But Durban has another advantage, at least for my purposes here; the published historiography of the city is dramatically smaller than it is for the other two city-states: Johannesburg and Cape Town.¹²

In a pattern that reflects the way in which the city forms part of the rural world that surrounds it, the historiography of this region has concerned itself with the movement of people from countryside to town. The most important of these are three studies—by Keletso Atkins, Bill Freund, and Patrick Harries—which share a common focus on the history of the migrant workers that have labored and lived in this region over the last two centuries.

These three scholars are all forward-looking researchers; they are emphatically not content to reproduce the method of social history devised in the 1970s and 1980s and have each been energetically fostering new kinds of historical analysis. Often this has been framed in opposition to an older, deeply empiricist social history—and yet what is striking about each of these studies is how much they rely on, and reproduce, Thompsonian understandings of class and class formation.

Each author issues an unambiguous statement of analytical purpose at the outset, and, at first glance, they could scarcely be deliberately more antagonistic. Echoing Karl Marx's The German Ideology and work by Eric Hobsbawm, Bill Freund's work is driven by the effects of large-scale economic changes. As Freund notes in the preface of his study of the Indian working class in Natal, “It is the intersection of structure and agency, the interplay between economic and social history and how they impact on one another, rather than the quest for identity, that engages me.”

Compare this endorsement of political economy with Harries' call for an explicitly cultural focus on the “values, signs, and rituals of authority” that migrants from southern Mozambique brought from their homes to the sugar fields of Natal and Zululand, and out of which they fashioned a new ethnicity. “This is the history,” Harries notes in direct opposition to the political-economy that has dominated much of the revisionist historiography, “of the making of that culture.”

Keletso Atkins' analysis of nineteenth century ideas about work is yet more explicitly framed by a concern for identity, generally antagonistic to the wider body of the “self-styled radical” historiography. Writing for a black undergraduate audience in search of historical and academic role models, Atkins adopts what she calls an “African-centered approach.” “I brought to the primary sources a list of inquiries,” she notes with equanimity, “the answers to which I hoped to elicit directly or indirectly from the workers themselves.”¹³

Yet there are important and strikingly similar contradictions at work in each of these books. Freund's work—while arguing for a structuralist understanding of the making of class—produces an excellent history of a local culture and community—in which class is first submerged and then forsaken. Harries, arguing for a history forged in the countryside by symbolic understandings central to culture, produces an orthodox political-economy of the urban fulcrum of class, in which rural culture figures through a glass, darkly. (Although he does present symbolic analysis, it is comparatively thin, based upon weak ethnographic and linguistic foundations and often the grammatical sleight of hand of the passive voice. Conversely, the evidence Harries presents on the strategies of the state and capital is rich and sustained.)

The contradiction at the heart of Atkins' study is even more interesting, for it highlights the colonial state's own understandings of the cultural characteristics of work. For Atkins correctly indicts the settlers and officials who sought to mobilize African labor in colonial Natal for their racist ignorance, whilst relying entirely upon archival materials and documents that were produced by them. Far from demonstrating that the settlers en bloc “were woefully ignorant of the African cultures that their elaborate plans sought to destroy,” Atkins dependence upon texts produced by the missionary A.T. Bryant, and the magistrates James Stuart and Carl Faye, suggest the opposite. Towards the end of the nineteenth century a powerful faction of the colonial state, and some settlers, had an astonishingly sophisticated grasp of the cultural dynamics of Zulu society and sought to deploy this knowledge to mobilize and regulate the colonial workforce. This paradox suggests that the state's effort to mobilize what was called the African working class in South Africa only after 1920 was premised on an appreciation of the building blocks of rural identity: idioms of work, gender norms, the politics of marriage, and the grounding metaphors of time. It is these symbolic elements of 19th century Zulu culture—derived from colonial sources—that form the evidentiary core of Atkins' study of the working class in this region.

Work and Time in Nineteenth-Century Natal

The great strength of Keletso Atkins' study of the migrant workers who left the Kingdom of Zululand and crossed the Tukhela River to work and settle near the towns of the Colony of Natal is her effort to reconstruct the cultural origins of an African work ethic. In her search Atkins explores a domain of conflict between the settlers and their African servants that mirrors and refracts Thompson's famous essay on the conflicts between urban and rural time and work-discipline in the English countryside.¹⁴

Her story begins with a strike by a group of male letter carriers in 1858 who insisted on measuring the terms of their six months contract by the waning of the moon. The phrase “inyanga ifile,” or “the moon is dead,” came to stand for a long running conflict between African migrants, mostly isiZulu speaking, and their English employers over the hold of calendrical time on the sale of labor. Atkins shows how workers took what they could from the rigid calendar of the towns (celebrating public holidays and the sabbath with enthusiasm), while defending until the end of the century the power of the lunar month and the rhythms of an agricultural work cycle.

The migrants' defense of rural forms of time, the settlers' poor access to capital, and a general demand for intense physical labor produced two forms of labor organization in Durban that mirrored and reproduced rural work patterns: the Amawasha laundrymen and the Togt day laborer. The first of these, the Amawasha, emerged from the peer associations of youth cultures in the countryside and adopted and extended the guild ethos of precolonial Zulu artisans to regulate the conditions of laundry work and limit competition. These gangs of male workers dominated the extensive laundry economies of the towns in Natal for half a century, moving to the Witwatersrand in the 1890s to take advantage of the extravagant wages offered there. By the turn of the century the patriarchal guilds of Amawasha laundrymen found their monopoly of the domestic service industry broken by large-scale steam laundries and an increasingly draconian local government effort to weaken their control of the terms of the sale of their labor.¹⁵

In South Africa today the Amawasha are remembered only in the archives, but the other labor form that Atkins explores has, if anything, grown in power since the decline of the controls over the movement of African men and women were abolished at the end of the 1980s.

Long term contracts of continuous, very heavy labor have been the norm for most of the large South African industries, especially in the gold mining fulcrum. Yet for much of the last two centuries workers in Durban have experienced a very different pattern of work. Before the heyday of Apartheid, work in Durban was determined by the rules of Togt: “a minute form of contract in which Africans hired out by the day were paid upon completion of the task . . .”¹⁶

The original meaning of the word highlights the manner in which this “minute contract” bound urban labor to the rhythm of an agricultural world. Togt derived from the Dutch word for a trip or journey, and referred to the men and women who took the “well-trodden ‘Kafir paths,' the narrow pathways leading to the European center” from homesteads surrounding the expanding city after the 1850s. For much of the next century, it was the determination to defend the viability of these rural homes that prompted the laborers' grim resistance to the effort on the part of municipal and colonial state to subject them to more enduring industrial work discipline.

Like the other studies of migrant workers in southern Africa, Atkins' attention to the ways in which urban workers expressed rural imperatives and constraints draws her to an understanding of class that is powerfully inflected by culture. She shows how the work ethic of isiZulu-speaking men in the towns of Natal was derived from much older patterns of labor mobilization in Nguni societies and encouraged by the changing pattern of gender relations in the countryside.¹⁷

Indeed one of the most interesting products of her research is the implication that the nineteenth century witnessed a hardening of gender relations within these rural homesteads, and a narrowing of the work opportunities that were available to women. In the early years women sold commodities, craft-work and farm products in the hungry markets of the towns and served as the “predominant group making up the rural togt labor force.” Over the course of the century the colonial state, and African patriarchs, scrambled to encode “customary laws” that bound men to fulfil the colonial state's tax demands, and women to fulfil the labor demands of their fathers and husbands.

In all of this Atkins' study reveals much of the wide-ranging and integrative explanatory power of the Thompsonian project of the study of classes in the making. And yet her study also suggests, admittedly inadvertently, much about another research project, one usually associated with the theoretical work of Michel Foucault: the study of the ways in which the colonial state sought to comprehend and control its native subjects. Atkins argues, with some truth, that it was the missionaries who were best positioned to undermine the integrity of Zulu cultural life and that the settlers, with much more truth, were steeped in racist and arrogant ignorance about the African people they sought to exploit. Yet, without exception, the richest of Atkins' insights derive from the work of colonial missionaries and magistrates. Her understanding of precolonial Zulu society is taken very largely from the lifetime of work by the missionary A.T. Bryant and from magistrate James Stuart's massive archive of oral testimony collected a century ago. And her interest in the disjuncture between English and Zulu forms of time echoes Carl Faye's somewhat later obsession with the Zulu calendar. All three were pivotal actors in the triangular dance between colonial state, white settlers and the mission church. And their work reflects a massive effort at the end of the nineteenth century (following from Theophilus Shepstone's sphinx-like autocracy) to document the state's apprehension of its rural African subjects. The colonial sources that underpin Atkins' study—a good example is Faye's Zulu References for Interpreters and Students—suggest how important cultural understandings were in the state's effort to forge an African working class. Yet what is perhaps most significant from Atkins' study is the success with which Africans resisted becoming the kind of working class that their colonial overlords had in mind.

Zulu men and women in Natal consistently refused to take up the long-term contract work that was offered on the colony's sugar estates, mills, and mines. Towards the end of the 1850s, as the price of sugar began to increase dramatically, the local sugar farmers impressed upon the colonial state their need for a reliable source of long-term labor. By the 1860s officials and settlers had abandoned the idea of forcing isiZulu speaking men and women to take up plantation labor and focused instead on two groups of more distant migrants: indentured Indian workers, mostly from Madras, and people from southern Mozambique, called AmaTongas by white and black Natalians alike. The official effort to recruit and transport workers from southern Mozambique was haphazard before the 1870s, but this did not stop large numbers of workers from making the journey south from the Delagoa Bay hinterland to the sugar farms along the Natal coast, and by the late 1860s these tramping workers “had come to dominate the labor force.”¹⁸

The same workers would, in later years, come to serve as the backbone of the immense South African diamond and gold mining industries. Today, known by another name, the Shangaans, they are subjected to intense vilification, policing, and repatriation. Keletso Atkins has given us an account of the ways in which isiZulu speaking men and women successfully resisted being transformed into an urban industrial working class. Ranks of the South African working class on the mines, factories, and plantations were filled, until very recently, by “foreigners;” chief amongst these were men from southern Mozambique.

The Mozambican origins of the South African working class

The story of the migrants who made their way, by land and sea, from the villages of southern Mozambique to the towns of Natal (and then later on to the mines of Kimberley and the Witwatersrand) has for decades been the focus of Patrick Harries' research. His work is interesting in part because he argues for an explicitly cultural focus, and a turn away from a focus on class forged through struggle. “It is an assessment of the symbolic content of the workers' culture that alerts us,” he notes towards the end of his study, “to the very weakness of class as a source of group consciousness linking black miners on the Witwa-tersrand.”¹⁹

So, in the place of the metanarrative of class formation, Harries' work presents an account of cultural evolution, of the making of “Shangaan” identity.

This account is not an ethnic history. The peoples who have been described by the terms Amatonga, Tsonga, or Shangaan in South Africa were defined more by the fact of not being Zulu, Swazi, or Pedi than by a set of cultural or political institutions of their own. In the nineteenth century the term Tsonga came to describe clans that fell outside the orbit of the Zulu, the Swazi, and Gazan states, and were linked to the trading settlements at Lourenco Marques and Inhambane. There is a niggling circularity here. While Harries is arguing that rural cultural institutions determined the pattern of proletarianization, he also demonstrates that migrancy constructed a widely held Shangaan identity. In the light of his penetrating research into the makings of the Tsonga orthography, Harries cannot simply endorse the missionary-anthropologist Henri Junod's reliance on language as “the great bond which bound the Thonga clans together in past centuries.”²⁰

In the absence of large-scale cultural or political institutions, he turns to the local ties of “clan and chief”: he maps the foundations of rural social identity in the naming of kin and the libation of clan ancestors.

Given this focus on culture, one of the interesting features of the account that Harries presents of the rural world of the migrants of southern Mozambique is that it offers geography as a substitute for ethnography. Harries paints a history of the Delagoa Bay hinterland in the nineteenth century in the broad strokes of climatological, epidemiological, and military crisis. In the early 1860s, a civil war in the Gazan kingdom, widespread drought and cattle sickness throughout the subcontinent, were followed by Swazi raids into southern Mozambique to restore their own depleted cattle herds. In the three decades that followed, smallpox epidemics which had brought death to the eastern Cape a century earlier, swept through the Delagoa Bay area. Like their neighbours in the Pedi kingdom, the people of southern Mozambique were well prepared for long distance migration by the upheavals of the mid-nineteenth century.²¹

After the 1850s the combination of long distance mercantile trade from the two ports, and specialized guilds of elephant hunters called Amapisi, focused much of the regional economy around ivory export. And the great bulk of Harries insight into the workings of these societies follows the same lines; his study is richest where it reveals interaction between the network of “Banian” traders—Portuguese-speaking merchants from Goa, Diu, and Damão—and their local clients, customers, and slaves. The trade in ivory placed a premium on long-distance travel as a means of subsistence, and encouraged other connections with the distant sources of salt, iron, and copper in the interior. As Harries puts it, “a tradition of migration was ingrained in the pattern of everyday life” decades before the Natal colonists began to look north for sources of cheap labor.

If there is a single cultural practice that serves as the lynchpin of Harries' explanation of the emergence of a common Shangaan culture, it was the predominantly male experience of undertaking very long journeys across the political, linguistic, and disease frontiers of southern Africa. Especially after the 1876 wage reductions on the Kimberly diamond mines sent thousands of Sotho and Pedi workers back to their homes, men from southern Mozambique made their way across the subcontinent to take their places. Some followed the road through Zululand and Natal that had been opened by the sugar planters in the early 1870s, others travelled via the Zoutpansberg or the Oliphants river to Pretoria and then passed through the western Transvaal to Kimberley. The term that he uses to describe these extraordinary journeys—“tramping”—is not incidentally derived from Hobsbawm and Thompson's studies of the “artisan's equivalent of the Grand Tour” in Britain in the 18th century.²²

And Harries intends us to understand them as a similar nexus of culture and class. “By the middle of the 1890s the imprint of a new culture had emerged out of the tramping network,” he notes. “This culture combined and transformed elements of the old and new and bound workers together as a group.”

This is an evocative and suggestive idea, but what kind of cultural history can be inferred here? Unlike Thompson's itinerant artisanry what is missing here is precisely the phenomenology of culture. It certainly makes sense that people who overcame these extraordinary journeys through bitterly cold Highveld winter nights, past dangerous predators, deadly diseases, and the efforts of impoverished white farmers to hold them for labor, might construct a common identity. But what it actually might have meant to the migrants remains completely obscure. What Harries has described as “tramping” might easily have been understood as a variation of forced labor, or, at best, of military servitude. After all the careful theoretical distinction that emerged between waged and forced labor in the wake of the abolitionist movement was easily lost in the context of armed recruiters, debt servitude, and generational despotism.²³

The common identity that emerged amongst the East Coasters might just as easily have derived from their mutual subjection to the political institutions which delivered and controlled the Mozambican workforce, and against which they have struggled for most of the last century. The irony here is that Harries provides us with a materialist account of the parameters of migrant social life; we are left to speculate about the content of culture itself.²⁴

The problem of scarce phenomenological evidence is acute for the rural areas of Delagoa Bay but it is not confined to them. Consider Harries' groundbreaking examination of literacy amongst the miners at Kimberley and on the Witwatersrand. “At Kimberley,” he points, “the power of the written word filled everything from travel and work passes to the chit of paper allowing the purchase of liquor.”²⁵

The lessons in basic literacy offered by the mission and compound churches equipped workers with tools that they could use to manipulate the web of documents, and, sometimes, use to access better paid and less arduous clerical work for the mines or their fellow workers. It was for these reasons, Harries argues, that workers and their families supported the missions located in the Kimberley locations.

Harries carries his observations about the potentials of literacy further, arguing that the “the power that accompanies literacy was to become an important factor drawing men to the diamond fields.” This is not an unlikely proposition, but I want to draw attention to the way in which Harries constructs the case about the cultural significance of literacy for Mozambican migrants. Precisely where it is necessary for him to present evidence about the ways in which workers apprehended literacy he shifts into the passive voice. “Writing,” he argues later in the book,

This is very important, fascinating evidence about the politics of literacy, indeed it suggests that migrant workers adopted literacy because it helped them bridge the cultural and emotional gaps between town and countryside. But who believed these things? When? Where? The reference for this crucial sentence directs us to the correspondence of the Swiss Mission, a footnote in Junod's ethnography, and the journal of one of the missionaries. There are then good reasons for the use of the passive voice here, for Harries' sources are not the workers, but the missionaries in the countryside. I hasten to add that the missionary archives are an invaluable source and, like Harries' account, reveal much about the cultural struggles that took place in the countryside in the late nineteenth and early twentieth century.²⁷

But they are certainly not unproblematic witnesses of the migrants' apprehension of literacy. It is not Harries' materialist account of the use of documents, or the establishment of schools that fails here; it is his description of culture. The evidence he has will simply not support the epistemological claims he makes about migrant culture. Yet this failure is richly compensated by the explanation Harries offers of the political history of long-distance labor migration.

The real power of Harries' history of the Shangaan migrants is in the political economy, concerned with the changing character of the state, the economy and the overlapping agents of each. Setting aside, for the moment, Harries' brilliant insights into the workings of the network of traders in southern Mozambique, his study is especially strong on the history of southern African labor law. He documents the Natal colonial state's unsteady participation in the recruitment and control of migrant labor from Mozambique after the 1857 sugar boom. By the early 1870s two characteristic features of the state in South Africa—direct official involvement in the control of migrant workers without any fiscal commitment to recruitment—had been passed into law in Natal. And, in a local variation of plantation politics, the agents primarily responsible for maintaining the discipline of indentured and migrant workers were the colonial magistrates wielding the blunt instrument of the 1856 Masters and Servants Act. This law was distinguished from the wider body of Masters and Servants law by the application of criminal penalties exclusively to African and Indian workers. Offenses that were subject to a mix of criminal and economic penalties in the Anglo-American common law—disobedience, misconduct and insolence—were transformed into peculiarly racial crimes in South Africa.²⁸

Harries follows the politics of South African labour law from the coastal plantations of Natal to the Witwatersrand in the 1890s. The formation of the Transvaal Chamber of Mines brought most of the significant mines together to consolidate control of the migrant labor force. And they immediately turned their attention to the law. William Grant, “the former chief recruiter of the Natal Planters' Labour League, drew up and presented to the Volksraad a special pass law for the Witwatersrand.” This attempt to involve the state in the mining industry's civil war was initially resisted by the Republican government. But five years later, using what would become one of the political hallmarks of the Chamber, the industry turned speculative panic to its advantage, persuading the Volksraad to accept a bill, which had been prepared “in every detail” by mine officials and was modelled on the systems of labour control in Natal.²⁹

The movement of law from the plantation complex on the Natal coast to the Witwatersrand industrial metropolis was sustained after 1900, when Henry Taberer moved from his position as the magistrate of Eshowe to take up the job of Witwatersrand Native Commissioner, and, after 1910, the management of the Native Recruiting Corporation monopsony.

The political history that Harries builds around the Mozambican migrants moves from the peaks of state power to the preeminent urban site of rural migration: the single-sex compound. The nineteenth-century mine compounds were modeled on Brazilian slave barracks in the late 1870s and they have come to dominate almost every South African city and township.³⁰

Harries shows that in the short space of four years after 1885 the closed compound put down powerful social roots: both managers and workers invested heavily in the societies nurtured within them. In a deliberate turn away from the austere institutions of social control described in van Onselen's Chibaro, Harries argues that workers within the compounds controlled their informal economy, determined the pace of their work, and ate “when, what, with whom, and how much they wished.”

The keystone of Harries' argument about early Kimberley is his description of the emergence, and decline, of a “racial paternalism” that white supervisors modeled on a reified idea of rural politics. Before the 1890s workers had substantial freedom within the brute incarceration of the closed compound to order their lives as they pleased; more importantly, they were highly skilled and demanded wages that were very high by international and regional comparison. It was in these years that mine managers developed the portentous idea of the compound as a “fictive chiefdom.” “This notion,” Harries argues, “was less a ‘mine owners' fantasy' than a managerial strategy through which De Beers battened onto, reinforced, and fed off rural structures of authority.”³¹

Rural culture, in this sense at least, was integral to the mobilization and control of the urban industrial working class.

This study of the making of the Kimberley compound system foregrounds the most impressive feature of Harries' work: his relentless effort to present the political dialectic between the brightly-lit urban centers and the shadowy rural social world of southern Mozambique. One of the primary findings to emerge from this double-sided research is a careful periodization of the economic effects of structural migrant labor in the countryside. Like most of southern Africa, the nineteenth century was difficult in Mozambique, but Harries suggests that the steady tempo of military, epidemiological, and climatological crises was muted by a commodity boom that coincided with the period of racial paternalism on the diamond fields. This boom, which was marked by the dramatic increase in the monetary value of bridewealth, began to collapse in the 1890s as mine managers sought ways to drive down wages on the mines, and the Portuguese set in place the military and fiscal arms of the colonial state in the countryside.

Starting with a massive increase in the amount and enforcement of hut tax in 1894, the Portuguese used the export of migrant labor from the Delagoa Bay area as the mainstay of the fledgling colonial state. Over the next two decades hut tax rose from 4s to 22s. In the wake of the defeated 1894 uprising the area of tax collection, and colonial administration, expanded from the immediate vicinity of the ports to include the previously independent chiefdoms in the interior and, most notably, Gazan kingdom. The emphasis that the colonial officials placed upon raising revenue served both to weaken the financial independence of individual homesteads, and dramatically to increase the movement of workers to the industrial centers of Kimberley and the Witwatersrand.

It should be obvious from all of this that Harries' research offers substantial new ways of conceptualizing the relationship between the industrial centres of South Africa and southern Mozambique. His work establishes a new periodization, and highlights the dialectic between political changes in the countryside and the shifts taking place in South African urban centers. It is doubly ironic that, in the light of this strength and Harries' earlier statement of purpose, the cultural history offered here is sometimes perfunctory, and always incomplete.

Given the tremendous power of the political economy in Harries' study it is useful to consider the meaning of class here. Harries' core argument concerning class is developed only towards the end of his study (this late introduction is unfortunate as he makes an important, and controversial, analytical point). Here he suggests that the preoccupation with class itself has distorted our historical understanding. Working-class consciousness was, at best, shaped by a “fitful and sporadic class conflict;” on the contrary, workers most commonly defined themselves in ways that were antithetical to class and more powerful. But the analytical point of this reassessment of class is a major reproach to the revisionist historians who have “tended to repress a complex and contradictory working-class discourse that included apathy, collaboration, and a bundle of identities at best only obliquely related to resistance.”³²

Yet, in the body of his research, Harries uses class in ways that are partly definitional, determined by economics and production, and partly commonsensical, especially as it relates to power. The first sense is most clearly suggested by the description of Kimberley as the “cradle of a new working class,” and the second by Harries' concluding description of the “long defeat of black labour” after 1896.³³

Similarly, he can not do without the notion of the “capitalist class” to account for the actions and strategy of those who controlled the sugar estates and mines. Even at the heart of his thesis that political conflict on the mines was determined by the evolution of customary limits on management's power and workers' responsibilities, Harries is forced to return to the old dialectic of those who purchase labor and those who must sell it. The moral economy—yet another idea from Thompson—that established the cultural rules of the game binding most participants in the mining industry was formed, as he notes, by the “informal struggles between labour and capital.” But if Harries uses class conflict in the commonsensical way that is common in labor history, he also intends us to understand it as a struggle that was shaped, in the first instance, by rural cultural imperatives. Workers, he notes, “deserted, stole, and loafed more as a means of protecting the rhythms of their noncapitalist work culture than as a means of expressing their opposition to a class of capitalists.”³⁴

This understanding of the political centrality of rural culture poses some very difficult problems for research, problems that seem to be unanswerable, at least for the moment. For the rural social world, especially in Mozambique, was well hidden from the eyes of the colonial state, and it is remembered only by shadows in the massive archival collections. In its place Harries offers us an account of the making and mobilization of workers that reflects the political and economic concerns of his evidence. It is an account of the making of class very like the political and economic explanations that Marx offered in the Eighteenth Brumaire. That makes it different from his own stated objectives, but certainly not less useful or insightful in accounting for the predicament of the modern Shangaan migrants—who find themselves, once again, the subjects of plans hatched at the dizzy heights of the South African state.

Was there an “Indian” working class in Natal?

When the Mozambican migrants abandoned the sugar fields of Natal in the 1870s, their places were filled by men and women from India who were brought to Natal under a rigid program of indenture between 1860 and 1911. From the outset Bill Freund's study of the history of these workers and their descendants is vexed by the relationship between the indentured working class—a legally constituted class whose existence was transient despite the planters' best efforts institutionalize indenture—and a set of identities: Hindu, Muslim, urban, rural, poor, small-scale peasant, merchant, and, eventually, the all-embracing category, Indian.³⁵

Again the problem here is not a product of the peculiar circumstances of working-class life in Durban. The dilemma of attributing political movements to class struggle or the defense of community, particularly as it has found expression in Thompson's study of the English working class, presents a thicket of sociological problems. Twenty years ago Craig Calhoun convincingly argued that the struggles that form the heart of Thompson's study of the making of the working class ought to be understood as radical populism motivated by older bonds of community, rather than a working class institutionally or culturally bound to trade unions.³⁶

But the relationship between class and community is in some respects more interesting in Durban precisely because the community, and many of its traditions, have no simple antecedents. The 150,000 indentured workers brought to Natal were recruited across large areas of northeastern and southeastern India over a fifty-year period. The “Indian Community” in Natal, then, is as much a product of the modern history of this region, and of the process of industrial development (at least in agriculture), as the working class might be.

Like African slaves in the Americas, the indentured workers shared a definitive social and economic institution in the form of unfree labor on the sugar plantations. Throughout Freund works hard to inject an analysis of social structure and economics into his work, but he chooses not to argue that the experience of indenture was the critical formative moment in the history of a working class. His reasons for not making this argument seem to stem from the fact that indentured workers were not marked apart from the wider community of south Asian immigrants. Ex-indentured workers usually sought to return to the land as small-scale Indian peasants, to eke out a living on the margins of the sugar industry and on the peripheries of the towns.

There are some basic economic facts about the history of Indians in Natal that warrant repeating today. Before the 1950s the single most important economic characteristic of Indian people was their poverty. In the countryside the racial segregation of land and financial facilities enforced the steady decline of market-oriented Indian peasantry which, in turn, bolstered a large population of urban poor in the city. Almost all urban laborers were defined as unskilled, and nine out of ten lived in wattle-and-daub or wood and iron houses before World War Two. As Freund stresses, prior to the 1960s, very few Indians occupied the intermediate economic strata they hold today; most were desperately poor, and in many cases poorer than African migrant laborers who were also struggling to retain access to rural agricultural resources.

The boundaries between peasants, urban workers, unemployed, and traders were particularly obscure in the jumbled landscape of the Durban hinterland before World War Two. Freund deals with this confusion with a simple but evocative geography of prewar Durban. He describes both the misery and the license of Outcast Durban, and the deliberate effort of the municipal planners in 1931, following an earlier metropolitan example, to extend the city limits to absorb its undisciplined perimeter.³⁷

(Both the Victorian effort to regulate the slums of London and the twentieth century Durban planners' concerns with the slums of the city periphery pale in comparison with the staggering recent extension of the city's jurisdiction and responsibilities.) This description of the city explains much about the interconnectedness of class and community. If there was structure to the demography of the Indian city before World War Two, it was essentially cultural: the pivotal institution of working life was the Grey Street mosque and the fabric of leisure time was held together by the twelve Indian-owned cinemas in the city.³⁸

Many of the forces shaping the preeminence of a racially defined community and the defeat of class were imposed from above by white employers and the state. This was as true for their presence in the city of Durban as it was on the sugar fields: ethnicity defined the work experience of Indian workers. “The policy of replacing Indians by Europeans has been carried out as far as the Council can, with advantage,” wrote the Town Clerk in 1935, “There are certain jobs, however, of a ‘dead end' nature which an Indian values, gives effective service therein, and is prepared to spend his working lifetime thereat, which would not satisfy nor be carried out efficiently by a European lad.”³⁹

There is also evidence that the workplace was a source of class identity for Indian workers, but, on the whole, it is the weakness of institutional trade unionism that emerges most starkly. Following a pattern that was true for the entire country, trade union activism in Durban was concentrated around the two world wars. By the end of the period two powerful racial imperatives effectively flattened these tentative moves towards a broad class-based identity. The first was the use of African scab labor to replace Indian and African strikers in the prolonged disputes in the laundry and textile industries in the 1940s and 1950s. The second was the much more traumatic and devastating “race riot” in Cato Manor in January 1949.

The conflict between African tenants and Indian shopkeepers, landlords, and residents in the Mkhumbane slums of Cato Manor, within walking distance of the Durban City Hall, produced an extended and terrifying period of murderous violence. This conflict was one of the definitive moments of exclusion in the evolution of a politically and geographically distinct “Indian” community in Natal. The sense of being caught between the racist white elite and the rapidly increasing African urban population is well captured by folk accounts of the violence which describe white instigators dressed as Africans, and the mayor of Durban egging on the rioters. The violence in Cato Manor was a fork in the road for Indian politics. On the left, the path lay with the leaders of the Natal Indian Congress and the Transvaal Indian Congress who allied themselves with the ANC. On the right, an increasingly defensive, inward-looking and accommodationist separatism. As Freund explains, in the wake of the unsuccessful non-racial strikes of the 1940s and the Mkhumbane disaster, “most Indian workers became members of the moderate Southern African Trade Union Council rather than the [African National] Congress-aligned SACTU.”⁴⁰

By the time his story enters the 1950s Freund's study of class has been completely overwhelmed by a larger ethnic identity. This shift reflects an internal retreat from the non-racial working class of the 1940s, but it was also a result of the enforced communal identity imposed by the combination of the Population Registration and Group Areas Acts. After 1950 the legal terms of address used by the state adandoned the economic definitions of indenture in favor of the biological and cultural definitions of Verwoerdian social theory. If the working class was submerged by ethinicity in the pre-1948 period, by the time of high-apartheid it was dissolved by the residential segregation that followed forced removals, and the establishment of separate “Indian” towns like Chatsworth and Phoenix.

Freund moves quickly over the ideological position occupied by Indians in the ideology of Afrikaner Nationalism before Verwoerd's pragmatic recognition that they “were here to stay.” Yet Dan O'Meara's research has shown clearly that Malan's NP cut its “black peril” teeth on the platteland Indian community. Drawing on the predicament of impoverished white farmers on the highveld, much of the Nationalist Party's racist vitriol in the 1930 and 1940s was directed at Indian merchants (following the pattern set by European anti-Semitism).⁴¹

This is important because the Nationalists abandonment of the compulsory repatriation of Indians did little to dampen their enthusiasm for removing them from the white cities. Freund's argument focuses on the fact that the English-speaking leaders of the Durban municipality found common cause with the Nationalists over the introduction of the “Pegging Act” which prevented Indians from purchasing property in 1943 and propelled the later policy of forced removals.⁴²

While the years after 1960 were marked by the destruction of communities and households, and the establishment of austere, and isolated “New Towns,” they also witnessed the dissolution of the pillars of the prewar Indian working class: structural unemployment and widespread poverty. The national industrial boom that followed the Sharpeville crisis (1960–1973) was particularly pronounced in the Durban region. Coincident with the development of manufacturing in the region was the expansion of compulsory education and the establishment of the university and technical institutions of tertiary learning. By 1984, just thirteen percent of those classified as Indian earned under R5,000 per annum, a figure that stands in marked contrast with the widespread indigence of the prewar era and the continuing poverty of Coloreds and Africans (whose percentages for the same income bracket were forty-eight and sixty-two percent respectively). By the middle of the 1980s the Indian working class had been subsumed by a comparatively prosperous and characteristically intermediate class of skilled workers.

But this period of prosperity was not without peril. In 1985, violence much like the earlier ethnic conflict in Mkumbane broke out on the boundaries between Inanda—the sprawling mix of formal and informal settlement that was home to a large proportion of Durban's African population—and Ghandi's Phoenix settlement, once again stressing the power of ethnic exclusion directed against Indians.

What can we make of the Indian working class in 2003? Here Freund doesn't provide any easy answers. The high proportion of subeconomic rents in Chatsworth, and the decay of Phoenix are mute testimony to the continuing experience of poverty that continues to drive protest. Recently, one of the elderly Chatsworth protestors answered an African Durban councillor who wanted to know why “Indians were resisting evictions?” with the now famous phrase: “We are not Indians, we are the poors.”⁴³

But neither poverty nor stratification make class, and it is hard to avoid the conclusion that by the early 1990s, as Freund discerns, “a new kind of ethnicity” dominated the social and political identity of Durban's Indian population:

In the process of the search for an economic and political class Freund highlights the invention of this identity and the community around it. It is precisely by following the history of the working poor from the experience of indenture on the sugar fields through the bitter poverty of peri-urban agricultural subsistence to the austere segregation of high Apartheid that Freund's account documents the startling unmaking of class and the definition of community. Yet, in contrast with both Harries' and Atkins' accounts of rural African migration migration, his preoccupation with urban working class institutions also limits Freund's history of the making of an Indian culture.

The Analytical Promiscuity of Class

I want to suggest that there is more at stake in these inversions than the always fertile possibilities of mischievous reading. The fact that each of these scholars achieves substantially the opposite of what was intended implies something important about the constraints and capacities of class-based historical analysis. Over the last few years, mostly under the influence of Subaltern Studies, we have become quite familiar with the ways in which the analytical repertoires available to historians are determined by the archive. This is not a question of bias. These evidentiary limits and accumulations, as Rolph Trouillot has shown, are themselves products of the workings of power—and they can explain much about the kind of world we now find ourselves in.⁴⁵

In South Africa the colonial governments' preoccupation with the long-distance mobilization of migrant workers from India and Mozambique necessarily organizes the labor history of these subjects around the state's work. The study of the migrant working class is a study of state power. The study of working-class culture is also necessarily study of official culture. Little has in fact changed about that.

But it is not the politics of the archive that forces class analysis—at least the Thompsonian variant that dominates southern African social history—to consider a host of other areas of research that are not conventionally included in the method of labor history. The dialectic between culture and class that has formed the basis of working class history since the early 1980s in South Africa necessarily pushes historians into fields of investigation that might more properly fall into a research project that has little concern for political economy. Consider the preoccupation in Keletso Atkins' study of the nineteenth-century Zulu migrants with the changing affective economy of marriage and gender roles before and after the arrival of colonialism. And note how Patrick Harries' study of the making of a Shangaan identity around the experience of migrancy concludes with an extended study of the changing patterns of gender socialisation and sexuality on the mines and in the villages of southern Mozambique. This concern with the sexual making of the migrant working class is widespread.⁴⁶

By the end of the 1990s, class-based history in South Africa had become remorselessly inquisitive and promiscuous, binding together the affective and the economic, state power and family politics, the urban industrial areas and isolated rural villages. Yet a concern with the compelling structural effects of capitalist development in our region was at the heart of all of these works. The fundamental analytical strength of South African labor history (an accidental effect of the scale of the country, a radically centralized state and tremendous concentrations of capital) remained the ease with which it addressed the political strategies of capital in shaping the biographies of individual workers.

The explanatory power of this historiography is prodigious. Contemporary South Africa is fraught by frightening problems: violence directed against women and children by their male relatives, HIV and tuberculosis epidemics, structural unemployment, and very high levels of criminality. In each of these fields the research results of a promiscuous historiography that relentlessly covers the ground between town and country have been amongst the very few sources of meaningful remedy.⁴⁷

Bill Freund recently observed that study of production-defined classes and their relationships may not be the best way to understand the new South Africa—or Tony Blair's Britain or George W. Bush's USA—but that is a problem that historians in one or two generations may be better equipped to confront than we are now. The acid test of historical analysis is not how well it accounts for contemporary (or future) political events, processes and affiliations. History has to pass a different examination. If we agree, as most academic historians seem to, that the fundamental task of history is to aid people in the development of an understanding of why many different things are as they are and assist many different people in shaping a politics that will help make things better (projects that most people will pursue without help from academic historians), then the histories of the making of the migrant working class have certainly achieved both in South Africa. They were able to do this because of a methodological requirement that practically forced urban labor historians to investigate rural culture.

We will be very, very lucky to do anything quite like that again.