Rethinking class and culture in Africa: between E. P. Thompson and Pierre Bourdieu Translated title: Repenser les classes sociales et la culture en Afrique : entre E. P. Thompson et Pierre Bourdieu

Introduction

The ‘death of class’¹ in Africa, as elsewhere, heralded by the advent of postmodernisation and globalisation, has arguably rendered any attempt to analyse contemporary class formation in Africa as a dated, even futile, exercise. An earlier focus on mobilised labour and national economic structural divisions in developed economies failed, according to this end-of-class thesis, to capture the fragmentation of labour markets and the transformative dominance of international capitalism beyond borders. At the same time, in Britain, the study of class as identity has been revitalised by the ‘cultural turn’ in class analysis, in a scholarly movement similarly sceptical of foundational theories of class (Devine and Savage 2005), and, in particular, Marxist approaches reliant on apparently ‘objective’ understandings of the economy and the social order. Against the tendency in this new trend to privilege Pierre Bourdieu’s cultural analysis of stratification over the material iniquities of class (Bourdieu 1984), however, Crompton and Scott propose that ‘a renewed emphasis on identity and difference should not be allowed to obscure or to downgrade one of the major preoccupations of class analysis, which is the study of structured inequality’ (2005, 191). In particular, they criticise the move to focus on the ‘reflexive individual’ rather than collective actions, and to conflate cultural hierarchy with ‘real’ economic inequality (Ibid., 199–200).

The present article argues, against the death-of-class thesis, first, that when it comes to Africa, the ongoing significance of cosmopolitan law for African class struggles denies the presumed disjuncture between class and globalisation; second, that the move from structure to post-structural agency characterising the ‘cultural turn’ in class analysis is also discernible in the historiography of class in sub-Saharan Africa. During the 1960s to 1980s, early research on the African ‘working class’ was heavily influenced, I show, by economistic Marxist and especially Fanonian perspectives. This body of scholarship was later challenged, especially by social historians of Africa inspired by the work of E. P. Thompson. Most recently, the move has been towards theories of African stratification based on lifestyle and consumption, influenced by Pierre Bourdieu’s arguments on the interaction of habitus, cultural capital and ‘distinction’. In this article I trace these transitions and, using the example of Botswana, consider some of their unstated assumptions in order to reflect on the continued centrality of identity, class and globalisation to critical African studies.

African workers: the betrayal of the working class

In his masterly historical review of the literature on labour in Africa until the early 1980s, Bill Freund concludes by asking rhetorically: ‘To what extent and from when and why can we speak of an African working class?’ (Freund 1984, 41). This critical question is in many ways still relevant today. The two decades between 1960 and 1980 were remarkable for their discovery and documentation of industrialisation, labour mobilisation and trade unions throughout East, West and Central Africa. Docks, railways, mines and factories, and even rural plantations, all employed African workers, and this expansion of the waged African labour force was recorded and theorised initially in Marxist structural-economic terms.² Africanists had no difficulty in identifying proletarians, the bourgeoisie, the working class or the ruling class as objectively recognisable social entities, defined by their relations to the means of production. Trade unions were analysed primarily in terms of their organisation and relations, first with the colonial state and liberation movements, and then vis-à-vis new postcolonial regimes, the stress being on the economic self-interest and strategies of leaders, who were often found to be remote from the mass of their members and other workers, and prone to factionalist politics rather than universalist aspirations.

A key debate dominating many of these early studies surrounded Frantz Fanon’s argument, in The wretched of the earth (1970 [1961]), of an emergent dichotomy between an indigenous urbanised African proletariat, allegedly co-opted by the ruling colonial regime and its successor national states, and a potentially revolutionary rural peasantry. This was articulated for Africa by Arrighi and Saul (1963) as the ‘labour aristocracy’ thesis, derived from Marx’s concept of the ‘aristocracy of labour’. The thesis envisioned, in Freund’s words, ‘a minority of African workers … brought into relatively depoliticised unions, paid fairly decently by capital-intensive multi-nationals and thus in alliance with imperialism and separated from any commitment to socialist transformation’ (Freund 1984, 9). According to one critic, the thesis assumed an ‘over-simplistic conception of the relationship between objective economic position, “class consciousness” and radical political action’ (Jeffries 1978, 171). The obvious conclusion of the thesis was that the working class was fragmented: there was no single solidary class in African nations, opposed to an exploitative capitalist class and aspiring to progressive socialist reform. Although, as Peter Waterman points out, African ‘wage earners and their unions usually played a dramatic and visible part in the national independence movements, following independence the unions have been commonly co-opted by the ruling parties or military regimes’ (Waterman 1975, 59). Waterman concedes, however, that there were some ‘aristocrats among the workers’ able, at times, to challenge the regime.

In line with this, against the polarisation posited between so-called ‘proletarians’ and ‘peasants’, Adrian Peace proposed in his study of factory workers in Lagos that workers were never fully ‘urbanised’ (or permanently ‘proletarianised’) but remained embedded in multiple township and rural relationships, and dreamt of becoming self-employed entrepreneurs (Peace 1975, 1979). This ‘double rootedness’ of urban migrant workers has been repeatedly noted, but less recognised is the fact that elites in Africa too continue to invest both symbolically and materially in the countryside (cf. R. Werbner 2004). Moreover, despite a wealth of evidence to the contrary, the linear historical view of urbanisation posited by Arrighi and Saul – from peasants to proletarians, urbanites and, indeed, to ‘cosmopolitans’, still persists, as demonstrated in James Ferguson’s analysis of the predicament of urbanised miners in the face of Zambian economic decline, who allegedly had cut off their links to the countryside during times of plenty (Ferguson 1999).

One important objection to the view that trade unions are ‘bourgeois organisations primarily focused on consumption’ (Sandbrook and Cohen 1975, 3) has been evidence of unionised workers throughout West, Central and Southern Africa mobilising in political struggles against social injustice and autocratic regimes. Jeffries in particular has shown that in Ghana, railway and other public workers developed an independent, radical consciousness, across ethnic and tribal divisions, during the colonial era, and this radicalism persisted even after independence (Jeffries 1978). Importantly, under a charismatic unionist leader, the Ghanaian railway workers’ strike, which mobilised workers across different unions throughout the country, aimed to achieve social justice in the form of fairer redistributive policies, and to hold government accountable. Jeffries notes that unionists’ ‘working class’ identity was formulated not in opposition solely to employers but rather in terms of a broader opposition between ‘the common people versus the big men in government’ (Ibid., 74).

Critics have pointed out that the labour aristocracy thesis ignored the huge pay gap between highly placed civil servants and other workers, whether miners or factory workers (Sandbrook and Cohen 1975). Reiterating this, Frederick Cooper noted that ‘the argument was misplaced from the start, confusing urban workers with a truly affluent city-based elite of politicians, senior bureaucrats, and others with access to power’ (F. Cooper 1996, 461–462). Saul himself subsequently reconsidered his argument and conceded that proletarianised workers may be distinguished from the ‘bureaucratic bourgeoisie’, and that under some circumstances the former may develop a more ‘revolutionary’ ‘downward orientation’ (Saul 1975, 305–306). Undoubtedly, however, the top echelons of the African ‘salariat’ were, Robin Cohen concedes, both in lifestyle and ideology, a ‘true labour aristocracy’ (Cohen 1991, 85–86).³

Key to the debate throughout this period was the question of ‘working-class consciousness’, defined in Marxist scholarship as a wider consciousness, transcending occupational, ethnic or restricted trade union interests. In a subtle analysis of workers’ class consciousness on the Zambian Copperbelt, Jane Parpart (1987), for example, demonstrates, against the labour aristocracy thesis, that, despite strategic manipulations by the mine management, the colonial state and postcolonial Zambian government after independence, all of which endeavoured repeatedly to undermine worker solidarity, the sense of a broader, shared worker identity was established quite early on among Zambian mineworkers and only grew as inequalities in Zambia increased after independence, and as freedom of speech and association were curtailed. Workers’ opposition was articulated not simply in terms of their own, specific interests as miners but, much as Jeffries had found among railway workers in Ghana and I later found in Botswana, by deploying a broader political language which highlighted the gap between a narrow elite and the population at large.

Clearly, much of this early critical literature on class in Africa rested implicitly on African left-wing scholars’ ‘utopian dreams’ for a better, more egalitarian society (Freund 1984, 36), while being at the same time largely economistic, structuralist and positivist in its descriptive methodology. Nonetheless, the veritable explosion of labour studies from the 1960s onwards highlighted the complexity as well as the fractionising tendencies of African worker organisations and mobilisations, while also recording fleeting moments of broader worker solidarity.

Class, culture and identity: from Frantz Fanon to E. P. Thompson

Since the mid 1970s, a growing dissatisfaction with rigid Marxist structural accounts of African workers and class formation was partly linked to the fact that such theories had little to say about the momentous ‘strikes and revolts’ sweeping South Africa in particular (Hyslop 2016, 101). Thus, according to Jonathan Hyslop, E. P. Thompson’s ‘demolition of Althusser and his cohorts’ opened up new ways of thinking about class in Africa, especially in South Africa (Ibid.), in an intellectual movement mainly concentrated around the ‘History workshop’ at Witswatersrand University in Johannesburg.

In his preface to The making of the English working class, it will be recalled, Thompson argued that class is not a ‘thing’ (1963, 9) and indeed, he claimed, this was not Marx’s meaning. Hence,

The iniquities of the colour bar and the apartheid regime that followed it in South Africa made African worker trade union activism, where it occurred, heroic and often tragic.⁴ Another historical monograph inspired by Thompson’s oeuvre was Dunbar Moodie’s study of workers on the South African gold mines, Going for gold: men, mines and migration, published almost 20 years later (1994). Rather than Thompson’s The making of the English working class, Moodie frames his analysis around Thompson’s article on ‘The moral economy of the English crowd’ (1971). The moral economy of the South African gold mines was based, he argues, on mostly unspoken (and sometimes explicit) understandings and rules (known in the vernacular as imiteto) between management and rural labour migrants on contract work. In Moodie’s words,

Moodie proposes that as long as the mines relied on contract rural migrant labour, the moral economy of the mines did not create a wider worker (and hence working-class) consciousness, since only a small proportion of workers were urbanised and were willing to struggle for a living wage. Instead, management, with support from the police and the state, fomented ethnic rivalries (since miners lived in ethnically segregated housing) and these occasionally flared up into violent factional fights. In effect, Moodie is arguing for a reverse of the labour aristocracy thesis – rather than a privileged labour elite ignoring the ‘masses’, the mass of mine workers refused to be included, and to participate in, a broad-based struggle for workers’ rights. They turned their backs on a wider worker consciousness and an international world of worker rights. It was only when the price of gold went up considerably, and the flow of ‘foreign’ labour migrants from Mozambique, Zimbabwe and Malawi was historically stemmed for political reasons, that local, often more urbanised, South African workers were recruited who demanded higher wages, commensurate with the profits of the mining companies.

The issue of an ethnically fragmented working class is also stressed by Belinda Bozzoli (1987) in a definitive introduction to a history workshop collection in which she outlines the Thompsonian perspectives underlying the workshop’s new African labour scholarship. Bozzoli introduces the issue of ‘community formation’ as an emergent reality of ‘class ideology’. Singling out the ‘materiality of community’, she argues that:

There were exceptions, however. Iris Berger’s study of the struggles of women across the racial divide for decent wages (Berger 1992) shows that in some industries, like the Cape food and canning industry, union organisation inspired workers to reach a wider consciousness of a ‘vibrant worldwide progressive movement with a pro-labour, anti-imperialist character’ (Ibid., 206). Berger describes a politics and union-organised popular culture of the 1950s that involved women emotionally, socially and politically, leading women ‘to identify themselves as part of a self-conscious working class’ (Ibid., 238). Religion and politics ‘were perceived neither as conflicting loyalties nor as opposed forms of awareness’ (Ibid., 243). There was ‘congruence between women’s lives and culture at work, in the family, and in the community … central to strengthening working women’s active class identification, to reshaping “class” in terms that resonate with their own experience’ (Ibid., 9). She associates this ‘integrative approach’ with the work of E. P. Thompson and Herbert Gutman (Ibid., 10). At the heart of this integrative approach is the pervasiveness of ‘struggle’ which her book documents.

Elsewhere in Africa, studies have repeatedly noted the extent to which class encompasses ethnicity in trade union worker activism: on the Zambian Copperbelt (Epstein 1958; Parpart 1987), among Ghanaian public workers (Jeffries 1978) and even, as we saw, on the Rhodesian mines. Frederick Cooper, however, raises the question whether ‘the content of Thompson’s text has been as influential as his title’ in Africanist labour studies (F. Cooper 1995, 235). He finds ‘disappointing’ the fact that ‘[l]abor historians emphasize what the workplace brought to African workers rather than the other way around’ (Ibid., 236). With few exceptions, however, given the inspiration derived from Thompson, equally disappointing, in my view, has been that few of the Thompsonian-inspired studies produced comprehensive, ‘integrative’ ethnographies of working-class formation.

Botswana: the making of an African working class

In Botswana the historical emergence of class identity and class consciousness in an African postcolonial nation was an active process, forged in the struggle of low-paid workers for public dignity and a living wage (Thompson 1963, 8). Elsewhere I trace this struggle in detail through a series of historical events that have shaped Botswana’s low-paid public service workers’ individual and collective identity in what Thompson called an ‘historical relationship’ – above all, between workers and the Botswana government, their employer, but also with churches, the opposition parties, fellow unionists and other key actors in civil society. As in industrialising England, in Botswana, too, the working class, newly formed after independence in 1966, cannot be regarded as a fixed, reified social entity. As Muhammad Talib has argued in relation to Indian workers, class cannot be measured simply by ‘concrete indices’ (Talib 2010, 230). It is, instead, in the words of Sharryn Kasmir, ‘multiple and contradictory’ (Kasmir 2005, 81). Above all, class is a constantly evolving imaginary, created through concrete interactions among workers, in practice and performance, vis-à-vis a dominant, hegemonic employer. The experience of class, I found in my research, is mediated by cultural images, labour songs, vernacular oratory, numerous public meetings, rallies, strikes, legal mobilisation at courts and popular cultural performances. Together these contribute to a sense of continuity, linking past and present, even as these images and discourses are refigured and reincorporated into new discourses whenever workers encounter new cosmopolitan ideas about labour rights or face new local crucibles. We may say that this changing yet continuous repertoire of signs orders experience and infuses it with meaning (Harries 1994, xvi). As Iliffe, following Thompson, early on argued, dock workers in colonial Tanzania developed class consciousness through the very act of working together and acting together to advance their interests (Iliffe 1975, 50).

Such cultural dimensions of trade union and worker activism are often neglected in other African labour studies. In particular, I found – against a simplistic unidirectional African ‘proletarianisation’ thesis – that the fusing of cosmopolitan and local popular culture has created a distinctive, vernacular way of being a worker in Botswana that does not deny workers’ roots at ‘home’, in the countryside. Although Thompson’s humanist interpretation of class consciousness and of the ‘acting subject’ has met with some scepticism, others have rejected the widespread ‘end of ideology’ view that charts a decline in class consciousness, allegedly replaced by non-class, identity-based social movements (Kasmir 2005, 81). Citing E. P. Thompson, Kasmir proposes that anthropologists are particularly well placed to understand the ‘ideas of working people in specific social and historical contexts’ (Ibid.). In portraying the life histories of union activists, women and men, I was able to uncover manual workers’ evolving agency as they strive to shape their futures by strategising and struggling against the handicaps of low pay, working poverty and a lack of formal education.

There was a redemptive quality to the self-conception of struggle in the union, expressed in their labour songs, a fusion of international labour and local traditional melodies and lyrics, and the pervasive sense that their ultimate aim was to liberate the poor and oppressed – this, despite the fact that union daily activities are marked by nitty-gritty, pragmatic negotiations over wages and rights. Members of the union never tired of telling me that, despite their poverty and lack of education, they could ‘teach’ the other unions and occupational associations in Botswana how to be unionists, how to protest, how to mobilise effectively. This was a source of pride and distinction for them.

As Thompson and other early historians too found (e.g., Hobsbawm 1959), the redemptive quality of their vision may be linked in part to members of the Manual Workers Union’s almost universal affiliation to churches, whether established churches like the Anglican, Congregationalists and Catholic, or local ‘Spiritual’ churches, ‘Zionists’ or Pentecostals. Many union leaders stressed that they had cut their teeth in church preaching and organisation. ‘Political prayers’ were a common feature of union rallies. The struggle for workers’ rights in Botswana drew ‘spiritual’ legitimacy from biblical truths learned in the myriad of churches workers regularly attend (Gutman 1966, 82; see also P. Werbner 2014, 2016). As in American and British trade unions, in Botswana too the Christian God justifies and sanctions workers’ claims and ‘ennobles’ their labour (Gutman 1966). Christian faith and current politics are interwoven in Botswana’s union activism into a seamless whole, as God is beseeched in ‘political prayers’ to protect workers from greedy employers who rob the workers of their fair wages.

Unionists’ horizons expanded as they came to be aware of worker struggles elsewhere, joined and drew support from international labour organisations and gained consciousness of their rights, enshrined in international law. Cooper’s monumental comparative study of the history of African trade unions in anglophone and francophone colonies from the 1930s to 1950s highlights both the internationalisation of the trade union movement in Africa within the context of colonial labour relations in the period leading up to decolonisation, and the awkward relations between unions and nationalist leaders once independence became imminent (F. Cooper 1996). As Cooper has argued, in the period leading to independence ‘the question of free labour was becoming caught up with questions about the entire political and ideological fabric of colonial rule itself’ (F. Cooper 2000, 113). The slogan in the 1945 strike in the port of Dakar was ‘equal pay for equal work’, underlining the fact that Africans were acting ‘in accordance with the idea of a “universal” worker’, encompassing citizens and recent migrants ‘striking side by side’ (Ibid., 144).

The complexity inherent in African workers’ identities and subjectivities has been revealed, as we have seen, in studies that show trade unions to be fractionised by internal racial, class and educational divisions, with clerical and skilled workers often suspected of being stooges or sell-outs. Such divisions have been apparent most recently, in 2012, in the platinum mine strikes in South Africa, with tragic consequences. The absence of a racially based labour politics in Botswana, however, has meant that labour relations in Botswana are wholly African and thus mainly class based, so that radical politics too are framed ideologically in the language of class warfare.

Unions are institutions capable of creating solidarity across ethnic divisions and class fractions. This has been true in Botswana where, despite their multiple ethnic and linguistic affiliations, members of the Manual Workers Union tended to minimise the divisions among themselves. In transcending ethnic and regional divisions, their project can be constituted paradoxically as one of nation-building, even as they struggle against the government, their employer. This capacity is also arguably why, historically, African trade unions were at the forefront of the liberation struggles, including the fight against apartheid in South Africa.

Seen worldwide, unions in the public sector are some of the largest and most influential, especially when it comes to enabling new, more progressive labour laws (on this in the USA see McCann 1994; Snarr 2011). More than other trade unions, public service unions are likely to engage in public policy issues relating to the welfare state and universal social justice (Carpenter 2000). They are also adept at using judicial review in the High Court and Court of Appeal to hold government accountable. In Botswana unions in the public sector form a dominant numerical presence, with low-paid public sector workers being at the forefront of the struggle for a ‘living wage’.

Their neglect is significant: public sector unions form a special class of unions. They have certain inherent advantages (for example, job security), are subject to specific legal regimes and mirror in the span of their spatial organisation the outer reaches of the state. Unlike mine, factory or dock workers, usually concentrated at specific sites, public sector workers are dispersed nationwide, in large cities and small towns, in the capital’s government enclave and in remote border outposts, and in urban and rural areas. This means that even when public sector unions mobilise for a national strike against the government, as they did in 1991 and 2011, they are nevertheless ‘building’ the nation.

Class unities are constantly threatened: by 2016 the solidarity of the umbrella association of public sector unions in Botswana was shattered by divisions across internal class and factional lines, with non-manual public servants abandoning the class solidarity they had displayed previously in mobilising for the national strike. By April 2016, their dispute was being fought out in the High Court and Court of Appeal. It highlighted the continuous impact of the nation-state, its politics and law, on class formation in Africa. Such divisions seem also to show the fragility of working-class solidarities in the face of huge disparities of income and lifestyle.

Class, identity and culture

Which brings me back to the cultural turn in class analysis raised at the outset, and with it, the problem of class ‘identity’. There are many ways of approaching identity. It may be seen as oppositional or experiential, cultural or social. Class identities have been construed, subtly, as inflected by lifestyle⁵ and theorised as intersecting with other, equally determining social identities – of gender, age, ethnicity or race. My study followed E. P. Thompson and other historians of class in attempting to describe the growing salience, depth and uniqueness of working-class identity in Botswana. My objective in using ethnographic methods of thick description was to portray a nuanced picture of the various strands that constitute Botswana worker identity: steeped in Christian faith and the law, mobilised in popular culture, threatened by the precarity of job loss, rooted in ‘home’ villages, yet aspiring to middle-class lifestyles.

Although the very existence of a working class in Botswana is relatively new – given that class identity has been shaped among workers who grew up in a predominantly rural society – as an identity it was evident that class is, nevertheless, culturally compelling, embodied and entangled in powerfully felt subjective notions of honour and dignity. Working-class identity, both individually and collectively, constitutes for workers a narrative of virtue, moral commitment and self-respect, seriti, to be defended against those in power.⁶ To lose this identity is deeply traumatising. As one of the workers who had lost her job, attending the Court of Appeal hearing, told me: ‘We are just sitting at home. We have lost everything, lost our dignity. This is not our country.’

The making of working-class identity in Botswana should be conceived of, I argue, as process, responsive to current predicaments and crucibles that have made it historically significant. This has been so particularly with regard to the development of a more enlightened, ethically grounded body of case law in Botswana. The question for workers is: how to achieve social change and render the struggle for justice effective from their position of uneducated manual workers? There have been notable constitutional victories in Botswana’s courts, but the challenge of translating these victories into lasting rights for workers still remains, in the face of an intransigent and powerful employer, willing to invoke nightmarish visions of a future Armageddon to justify low-paid government salaries being kept well below a living wage, even as the salary gap with top civil servants widens. This challenge, of transforming worker rights achieved through legislation and the courts into a better, more humane and just working environment, is familiar to workers throughout the industrial world.

Class identity in the Thompsonian ‘integrative’ tradition would seem to imply a total way of being – material, embodied, performative, dialogical, culturally rich, nuanced and complex – rather than a segmented, situationally shifting, identification or individual status descriptor. Even if workers’ sense of identity is fiercely oppositional, it is not conceived of as part-time. The same can be said ideally of Bourdieu’s notion of identity as distinction (Bourdieu 1984): it is inscribed in the very disposition or ‘habitus’ of a person, in his or her manner of bodily conduct (‘hexas’), choice of lifestyle, cultural knowledge (‘capital’), aesthetics of ‘taste’, as well as symbolic and economic capital (social networks, influence, property). Bourdieu’s elites do not make a habit of slumming and, if they did, this would itself become a mark of distinction. The semiotics of distinction, Bourdieu shows, is a form of cultural consumption that is constantly transformative; but throughout, upper classes maintain their dominance.

For both Thompson and Bourdieu, then, class is ‘cultural’ and for both, it is overarching, subsuming and encompassing of all other identities (such as gender, ethnicity etc.).

From working-class to middle-class culture and identity in Africa

Not surprisingly, perhaps, the growing trend in African class analysis concerns the emergence of an African middle class or of African ‘elites’ engaged in consumption practices.⁷ The aesthetics of African consumption and distinction have in fact, however, been a familiar topic with a long and noteworthy genealogy from quite early on in the study of African towns – starting with pioneering studies of dress on the Zambian Copperbelt in the 1940s by Godfrey Wilson (Hansen 2000, 27, 34, 40) and Clyde Mitchell, in his analysis of the Kalela dance (1956, 12–14), which both highlighted the prestige of Western attire for urban mine workers. In the Congo, Friedman’s study of the society of elegant people (Société des ambianceurs et personnes élégantes – SAPE) revealed the sapeurs to be the eponymous example of sartorial elegance and ambiance both at home and in the diaspora, spending vast sums on trips to Paris where they purchased authentic designer clothes (Friedman 1994, chapter 7). John and Jean Comaroff speak of the early European missionaries cooperating with other imperialists in seeking to ‘conquer by implanting new cultures of consumption’ (Comaroff and Comaroff 1997, 219 and passim, emphasis in the original). The Christian evangelists focused on the body – ‘In cleaning it, housing it, curing it, and clothing it lay the very essence of civility’ (Ibid., 220). Western clothes were an ‘enchantment’, a sign of ‘exotic forces’ (Ibid., 234), but so too was the use of soap, hygiene and cleanliness (Burke 1996) and, very importantly, education (West 2002).

More recent studies have also stressed the way that African workers and the middle classes alike continue to follow European fashions, whether by buying second-hand salaula clothing in Zambia (Hansen 2000) or hip-hop outfits adopted by young men in Mombasa’s slums (Behrend 2002). Those with higher incomes who work in the Kenyan tourist industry are said to ‘shine’ – they display their sophisticated cosmopolitan sartorial competence ‘to effect a Europeanised look that even Europeans can’t quite match’ (Schoss 1996, 170). Many purchase much of their clothing directly from Europe (Ibid., 171) and follow current fashions (Ibid., 174). By contrast, beach boys cultivate ‘indigenous chic’ (Ibid., 172). Equally important are accessories: Ray-Ban sunglasses, gold jewellery, and Italian leather shoes and belts (Ibid., 174).

In Tanzania, a housing boom has generated a wide variation in housing styles among different fractions of the growing middle class (Mercer 2014). To be fully middle class, however, requires a high income, extensive property, and a full range of consumer goods and cultural competences, as Hansen describes for affluent Zambians, the apamwamba:

In Nairobi too leisure is a mark of exclusivity and cultural competence: young, upwardly mobile professionals in Nairobi ‘frequent the dozens of upmarket bars, clubs, and restaurants that dot the urban nightscape’ (Spronk 2014, 106), dancing, drinking, surfing the Internet, reading magazines, listening to music and watching films, swimming in hotel pools, exercising in gyms and sports centres, in a subculture ‘where they meet their peers’ (Ibid., 107).

Although African manual workers cannot possibly participate fully in this globalised consumer culture, in Botswana and elsewhere, they can, selectively, buy into it: purchasing some consumer goods, especially if they live in town – smartphones, flat-screen televisions, cars. The wide availability of credit today makes this possible (Hansen 2000, 192; James 2012).⁸ In Botswana they often own multiple homes, however modest, invest in cattle and acquire plots through government schemes. Seen purely in terms of consumption, then, lines between the working classes and the middle classes are often blurred. As in Ghana, everywhere on the continent weddings and funerals involve huge feasts and competitive displays of sartorial distinction (Fumanti 2013; Lentz 2009, 2015). Indigenous African women’s dresses, known as chitenge in Zambia (Hansen 2000), often with West African designs, add to a growing creative African market in popular music, dance, films, TV soaps, advertising, design and other media, as well as live performances, with the West no longer exclusively the enchanted other. African culture industries, cultural entrepreneurs and cultural capital are increasingly dominating everyday leisure throughout the continent.

Concluding remarks: middle-class morality and the political field

To conclude this historical overview of the increasingly cultural ‘turn’ in the study of class in Africa, we need, however, to reflect on the meaning of middle-class consciousness. We saw that the link between culture, consumption, morality and the Puritan ethic that Max Weber disclosed was later imported into Africa by the missionaries. Beyond cultural consumption, however, to the extent that the growing African middle class today is not a singular, homogeneous entity but is divided and fractionised, what kind of collective ethos and political consciousness is discernible in its different class fractions?

In a little-discussed chapter of Distinction (1984, chapter 8), Bourdieu makes the important point that personal experience in itself cannot generate broader political discourses. The ‘essential problem of politics’, he says, is ‘the question of the transmutation of experience into discourse, of the unformulated ethos into a constituted, constituting logos, of a class sense’ (Ibid., 460). Within the political field, he argues, only a select few actors acquire the legitimacy authorising them to make ‘universal’ judgements that are a precondition for collective mobilisation; paralleling other cultural producers in the cultural field, these experts specialise in the ‘production of political opinions’ in the field of politics (Ibid., 399 and passim).

Class politics is an adversarial, conflictual politics.⁹ I have shown that African trade unions, their leaders and activists formulate and articulate the political and economic interests of the working classes, based on broader, universal principles beyond ‘inarticulate’ individual experience; indeed, they draw on cosmopolitan labour principles. Within the African middle classes certain groups, generational cohorts or social circles acquire the sense of legitimacy and competence to assert morally informed political principles. Thus in Kenya, the yuppies studied by Spronk saw themselves as ‘inter-ethnic’, and were committed to ‘overcome “tribalism” perceived as one of the worst socio-political maladies in postcolonial Kenya’ (2014, 103). One of them set up a website in response to the post-election ethnic violence in 2008, which became an online platform enabling individuals to report via SMS or email on acts of violence or trouble spots. As Spronk comments, ‘[t]his came from a group of typical urban young professionals who had the cultural capital needed to do so. And unlike their fellow Kenyans they had the means, the knowledge, and the contacts to enact their social and political aspirations’ (Ibid., 108).

This kind of morally grounded, rooted, vernacular cosmopolitanism is evident elsewhere as well. In Ghanaian ‘civil society’, Thomas Yarrow shows, leaders of non-governmental organisations (NGOs) promote democratic values based on an ideology of transparency, social equity, social justice, freedom of speech, sacrifice and patriotism. He thus argues against the ‘widespread assumption that for African NGOs moral discourses are nothing more than a charade behind which selfish acts of accumulation and aggrandizement are concealed’ (2008, 335). He cites approvingly Richard Werbner’s critique of the Afro-pessimistic literature on African elites that ‘focuses attention on the apparently irrational acts of kleptomaniacs and self-serving bureaucrats, foreclosing understanding of African concerns for the public good’ (Ibid.). Many of the NGO activists in Ghana, Yarrow tells us, have ‘longstanding friendships and shared ideological positions’ (Ibid., 338), and conceive their own selves and subjectivities, as told through personal narratives, as deeply intertwined with the development of the nation (Ibid., 344). Activists’ ‘agency’ has thus unfolded in tandem with the nation. Many had suffered and even spent time in jail for their beliefs. One activist told Yarrow that in the 1980s, the ‘destruction of those [socialist] movements meant that NGOs were the only places that you could raise those issues of policy. So charity work was directly turned into political work.’ Yarrow concludes, against the Afro-pessimist narrative, that NGO workers and activists in Ghana today are committed to the transformation and improvement of society (Ibid., 353).

In Namibia, Mattia Fumanti shows that elite generational cohorts, middle class in education, income and lifestyle, understand and interpret national and local politics quite differently, with the younger generation imitating and satirising their elders in their own private hangouts, bars and pubs (Fumanti 2016). Nonetheless, Fumanti shows, both generations are committed to the development of the Namibian nation. Finally, in Botswana, Richard Werbner describes how Kalanga elites form a post-civil service cohort committed and working for the public good of the nation (R. Werbner 2004). Werbner speaks of ‘cosmopolitan Ethnicity’ – as members of a minority group, the Kalanga middle classes struggle for an open, multicultural society for all.

In all these cases, and no doubt in many others, African middle-class consciousness clearly goes beyond narrow circles of leisure and hedonistic consumption, is often adversarial and is deeply cognisant of the responsibilities that come with privilege. Like African working-class consciousness, middle-class consciousness too is here an achievement, and, like working-class consciousness, it is integrative, encompassing and deeply imbricated in a vernacular culture and in local cultural and political understandings as well as a wider, globalised consciousness.

The present article has attempted to develop a comparative critical analysis of the historiography and present state of the African political economy of labour, as expressed historically and contemporaneously in class formation(s) and class struggle(s). My aim has been to tease out the cultural dimension of class, seen as a locus of power, conflict and solidarity, and interrogate the way that an ethos of class is imbricated in ethical notions of fairness and justice, and in popular cultural as well as pragmatic considerations. In all these respects, recent cultural analyses of class as just another intersectional ‘identity’ label tend, I have argued, to miss the encompassing moral, experiential and communal dimensions of class. The individualist, fragmentary class constructions they advocate also tend to miss the ‘hard’ political-economic bases in which class is embedded. A consciousness of these is perhaps most salient among trade unionists representing low-paid workers, on the one hand, and an African super-elite, bent on protecting its privileges, on the other. Current studies of African middle-class consumption need thus to be ever-cognisant of the ‘integrative’ dimensions of class as spelled out perhaps most profoundly by E. P. Thompson.

No potential conflict of interest was reported by the author.