The First Globalisation and Transnational Labour Activism in Southern Africa
White Labourism, the IWW and the ICU, 1904–1934

In this article, I argue that the history of labour and the working class in southern Africa in the first half of the twentieth century cannot be adequately understood within an analytical framework that takes the nation-state as the primary unit of analysis.[1] The literature has, with very few exceptions (notably Bond, Miller and Ruiters 2001: 4–5), generally presented the history of labour in the region in this period as a set of discrete national labour histories for Namibia (South West Africa), South Africa, Zambia (Northern Rhodesia), Zimbabwe (Southern Rhodesia), Swaziland, and so forth. Each national labour history is presented as taking place within a distinctively national context where organisational and political boundaries correspond with the administrative borders of the state, with labour politics developing inside these boundaries in response to national conditions, culminating in the emergence of national working-class movements.

Such approaches project postcolonial borders onto the period of imperial rule, ignoring the way in which international labour markets, regional political economies, and networks of activists and propaganda operated both across, and beyond, the British Empire and southern Africa to create a transnational southern African working class in which activists, ideas and organisational models circulated. Transnational influences played a critical role in shaping working-class movements, which straddled borders and formed sections across the region and beyond it. Furthermore, ideological, ethnic and racial divides within the working class across southern Africa played a more important role in constituting divisions than state borders.

This article explores these issues by examining three moments of transnational labour activism in southern Africa in the first three decades of the twentieth century. Firstly, there was the tradition of ‘White Labourism’: rather than being a peculiarly South African phenomenon, it originated in Australia, spread to South Africa in the early 1900s, and subsequently developed into a significant factor in labour politics in the Rhodesias by the 1920s. Secondly, there was the tradition of revolutionary syndicalism, which stressed interracial working-class solidarity. As developed by the Industrial Workers of the World (the IWW or ‘Wobblies’) in the United States in 1905, this tradition came to South Africa via Scotland, where it spread from radical white labour circles to workers of colour in the 1910s, and then spilt over into the Australian IWW. Thirdly, there was the tradition of the Industrial and Commercial Workers’ Union (ICU), whose politics were an amalgam of two transcontinental currents: Garveyism and IWW syndicalism. The ICU operated regionally, spreading from South Africa in 1919 to South West Africa and the Rhodesias in the 1920s and 1930s.

Set against the backdrop of regional waves of labour activism, the history of these transnational labour currents provides important insights into the social character of southern African labour movements in the period of the ‘first’ modern globalisation, lasting from the 1880s into the 1920s. The analysis presented here is influenced by, and makes a contribution to, the new transnational labour history that ‘relativizes’ and ‘historicizes’ the nation-state as a unit of analysis, stressing the ‘need to go beyond national boundaries’ and avoid ‘methodological nationalism’ in understanding working-class formation (Van der Linden 1999: 1080– 1081). A transnational labour history yields important insights into labour and working-class history, provides a new synthesis that goes beyond old labour history, with its stress on formal organisation, and new labour history, with its stress on lived experience, and stresses the interconnections between labour worldwide.

The novelty of the nation-state

The lived reality of the late twentieth century – where large-scale imperial formations are unusual and illegitimate, and where nation-states control education, media, identity documents and a host of other instruments for ‘nationalising’ citizens (Torpey 2000) – contributes to a situation where the normality of the nation-state, and its utility as a unit of social analysis, seem self-evident. A historical perspective, however, shows that the nation-state form of the capitalist state is a fairly recent phenomenon for much of the world, and only became the normal, rather than the novel, form of the capitalist state in the second half of the twentieth century. Before the late 1940s, the archetypical modern state was the formal empire, which drew a vast range of peoples and races under a single state appar-atus. Different degrees of national autonomy might be allowed, but the empire was itself the focus of loyalty; there was no assumption that the state should correspond to a given ‘nation’. The modern ‘nation’ itself being a fairly recent and problematic conception (Hobsbawm and Ranger [1983] 1992).[2]

Contemporary globalisation – the ‘second’ modern globalisation – with its increasing integration of communications, investments and labour markets directs attention to the problems of understanding labour within hermetically sealed national frameworks. On the one hand, the current growth in power of transnational connections and supranational powers has challenged the view that social analysis can unproblematically take the nation-state as the site of social action. On the other hand, contemporary globalisation should be historicised, and understood against the backdrop of previous eras of globalisation, notably the globalisation of the late nineteenth and early twentieth centuries (Hopkins 2002). In this period, international trade, foreign direct investment and global flows of labour took place on a scale larger than is the case today (Hirst 1997; Hobsbawm 1977).[3] The period of closed national economies should be understood, then, as a phenomenon of the 1930s to the 1970s, a world of nation-states bracketed by two phases of globalisation.

Early globalisation and the creation of Southern Africa

An important consequence is the need to consider the international and transnational dimensions of labour history, and to locate the creation of the working class in southern Africa within the globalisation of the late nineteenth century. The industrial revolution that took place in South Africa from the 1880s following the discovery of large deep-level gold deposits on the Witwatersrand was pre-mised on the inflow of massive amounts of foreign direct investment. This was also a period of British imperial expansion, as the world’s largest and most important state extended its boundaries. The main African states were subjugated – starting in 1879 with the conquest of the Pedi kingdom and Zululand; in 1884 the kingdom of Basutoland (now Lesotho) became a High Commission territory, followed in 1885 by Bechuanaland (now Botswana); in 1902 the kingdom of Swaziland became a British protectorate; and from 1890 onwards colonial states were formed in Northern Rhodesia, Southern Rhodesia, and Nyasaland (now Malawi). The Anglo-Boer War of 1899–1902 closed the period of imperial conquest and consolidation in the region.

While mining capital was rapidly centralised, so was state power: the Union of South Africa formed in 1910 as a British dominion comprising Afrikaner republics and British colonies incorporated as provinces, and African kingdoms as reserves. The working class that emerged within South Africa itself was neither distinctively South African, nor the product of a discrete South African history. It arose from the twin processes of overseas investment, which required a large supply of labour power, and incorporation into the British Empire, which provided the administrative and coercive means to meet that demand.

The new working class was sourced from across southern Africa and the larger British Empire, and was a multinational and multiracial mass that continuously flowed through the vast human rivers of migration and regional and global labour markets. Skilled workers were whites, mainly from Britain, but with a significant number from Australia and New Zealand, continental Europe and the United States. By 1905, perhaps eighty-five per cent of the white underground miners on the gold mines were British-born (often arriving via other mining regions in the Americas and Australia) (Katz 1994: 65; Ticktin 1973: 3, 259–260; Visser 2001: 2), and a similar pattern prevailed on the collieries (Alexander 226 African Studies 66: 2–3, August – December 2007

2004: 188–119). In 1921, 59.8 per cent of all typesetters, 55.8 per cent of all fitters, 52.1 per cent of barbers, 48.3 per cent of carpenters, and forty per cent of electri-cians were foreign born (Freund 1989: 85). The mines proved an exception: Afrikaners, already a significant component of the underground white miners by the start of the twentieth century (Visser 2001: 4–5), became increasingly important; by the 1920s, around seventy-five per cent of white underground miners were Afrikaners (Boydell n.d.: 191), rising to perhaps ninety per cent by the 1940s (Andrews 1941: 20).

Unskilled workers were mainly male African migrant workers, who worked on the mines as cheap, indentured and unfree labour controlled by pass laws, returning periodically to rural homesteads where their families resided. This migrant labour system was insatiable, and drew in workers from South Africa and its neighbours, forming vast human rivers that linked South Africa, South West Africa, the Rhodesias, Nyasaland, and Portuguese East Africa (Mozambique) into a regional political economy that was linked into the capitalist world economy.

In 1920 only fifty-one per cent of African mine labour in South Africa was drawn from within the country, largely from the eastern Cape African reserves, with a further thirty-six per cent from Mozambique, and thirteen per cent from the High Commission territories: Basutoland, Bechuanaland, and Swaziland (Yudelman and Jeeves 1986: 123–124). In later years, a growing proportion would be drawn from the ‘tropical territories’, situated north of twenty-two degrees S.L., reaching ten per cent in 1945, as well as a small number from South West Africa. African workers not employed on the mines tended to be more heavily drawn from South African sources than the miners. In 1931, for instance, 47.4 per cent of newly arrived African labour on the Witwatersrand, not working in mining, came from Natal, with a further 44.2 per cent from the Transvaal, while only 3.6 per cent and 10.6 per cent of African mineworkers were from Natal and the Transvaal, respectively (Freund 1989: 83).

Given an ongoing shortage of unskilled labour, employers throughout the region found themselves in continual competition for African workers (see Beinart 1987; Berger 1974: chapter 2; Cooper 1999; McCracken 2000: chapter 5; Meebelo 1986: chapter 1; Moorson 1978; Phimister 1988; Sanderson 1961; Van Onselen 1976; Vellut 1983). Dominated by American and South African companies, the Northern Rhodesian copper mines expanded rapidly in the 1920s as the central African copper belt was developed. The Northern Rhodesian mines found themselves competing for labour with the copper mines in the Belgian Congo, which began phasing out migrant labour in favour of an urbanised African labour force from 1927 in a bid to attract a stable labour force. Moreover, Northern Rhodesia and Nyasaland competed for migrant labour with both South Africa and Southern Rhodesia. The latter, with its relatively low wages, in turn fought an ongoing battle for labour with South Africa, while employers in the Cape did their best to stop the ongoing outflow of local workers to German South West Africa in the first two decades of the twentieth century.

Within the borders of different states, similar patterns of competition emerged: the less developed northern areas of Nyasaland competed with the emerging planta-tions in the colony’s south. Within South Africa, Southern Rhodesia and South West Africa, farms competed with mines for labour, while the mines strove to prevent the outflow of labour to the growing manufacturing sector. At the same time, the mines in southern Africa were competing with mines abroad for skilled labour. In the 1890s and early 1900s, not only did skilled white miners typically earn about five times the wages of African miners, but wages for skilled miners and some categories of artisans were generally at least double – and sometimes up to five times – higher than wages for comparable categories in mining areas elsewhere (Katz 1994: 67, 75–7).

Faced with the mismatch between labour markets and state boundaries, big capital began to form its own regional organisations for managing labour flows. In South Africa, the Chamber of Mines set up the Witwatersrand Native Labour Association (WNLA) in 1901 to organise the recruitment of African workers in the Transvaal, Bechuanaland, Mozambique, Namibia, Northern Rhodesia, and Nyasaland, while the Native Recruiting Corporation, established in 1912, focused on Natal, the Cape, Basutoland and Swaziland (Jeeves 1985; Katzenellenbogen 1982). Employers in Southern Rhodesia and South West Africa established similar bodies, with government help, in order to channel labour, reduce workers’ choices, and prevent the loss of labour to other areas. Even so, competition continued: in South Africa, for instance, the collieries – facing low prices for their product and unable to compete with the relatively high wages of the gold mines – sought to attract experienced African miners by allowing a substan-tial number to settle permanently with their families near the mines from 1907, providing land for housing and small-scale farming (Alexander 2001: 510–515).

White labourism and the British labour diaspora

Given these regional and global linkages, and the large-scale migration that they entailed, the borders of South Africa did not neatly determine the experiences and lives of working class people in South Africa. Not only were wages shaped by labour markets that crossed borders, but the politics of labour throughout southern Africa was shaped by ongoing interconnections across the colonial borders. The politics of white workers were, unsurprisingly, deeply shaped by foreign models, with trade unionism from Cornwall, White Labourism from Australia, and socialism from Scotland (Hyslop 2002), playing an important role. ‘In the eighties and nineties a number of British unions opened branches in Durban, Kimberley ... Johannesburg ... [and] ... Cape Town’, including the Amalgamated Society of Carpenters and Joiners in Cape Town in 1881, the Amalgamated Society of Engineers in 1894, which organised skilled metalworkers such as blacksmiths and fitters and turners, and a South African Typographical Union for skilled printers in Durban and Pietermaritzburg in Natal (Andrews 1941: 12–13).

White workers in South Africa developed a local tradition of White Labourism by the early twentieth century, which combined social democratic demands with a platform of job colour bars, segregation and the repatriation of Asians. While this drew on the existing traditions of the local unions, it was also demonstrably influenced by the policies of the Australian Labor Party, which combined labourite parliamentary socialism with a commitment to white supremacy. While forming the world’s first labour government in 1904, creating a ‘workingman’s paradise’ (Mackenzie 1966: 132), the Australian Labor Party defended the White Australia Policy designed to maintain the ‘thinly peopled continent as an everlasting home for an untainted European, and dominantly English community’ (Kennedy 1984: 3–4).

Australian immigrants like Peter Whiteside played a key role in promoting this project in South Africa (Hyslop 1999; Katz 1976; Kennedy 1984). Born in Australia, Whiteside was an engine driver and keen trade unionist who came to the Transvaal in the 1890s where he became active in the South African Engine Drivers’ and Firemen’s Association – notable for securing the first statutory job colour bar in 1896 (Ticktin 1973: 109) – and helped form the first trades and labour councils on the Witwatersrand. Whiteside was also involved in efforts to form a white workers’ party on a segregationist platform, starting with the Political Labour League in 1905 (Grobler 1968: 57–64; Ticktin 1973: 186–209). The League was one of a number of similar initiatives, which culminated in the formation of the South African Labour Party (hereafter the SA Labour Party) in October 1909. This was backed by most of the Natal and Witwatersrand unions, as well as a bloc in the Cape unions. The party platform advocated residential segregation, the repatriation of Indians, and job reservation, alongside welfare and municipal reforms framed within a vaguely socialist aim modelled on that of the British Labour Party (South African Labour Party [1910] 1960: 73).[4]

However, if White Labourism to South Africa came from Australia, it did not stop there, soon flowing northwards from South Africa as new mines were established and new markets for skilled white labour were opened. With the opening of collieries and gold mines in Southern Rhodesia in the 1890s, the rapid growth of Lourenço Marques in Mozambique as a direct result of South African trade and investment,[5] the opening of diamond and other mines in South West Africa after the turn of the century, and the development of the copper belt spanning the Belgian Congo and Northern Rhodesia in the 1920s, many white workers moved northwards, bringing along their political traditions.

In 1924, there were 702 whites and 8,740 Africans employed in mining and prospecting in South West Africa, with three quarters on the diamond fields (Cooper 1999: 124). In 1931 the mining industry accounted for 35.1 per cent of white employment in Northern Rhodesia, with much of the remainder concentrated in the civil service or in industries linked to the mines (Berger 1974: 16–17). By 1939, the Belgian Congo had 157,250 mine employees (including 2,250 whites), Northern Rhodesia had 24,900 (including 2,700 whites), South Africa had 464,359 (including 52,693 whites), and Southern Rhodesia had 90,886 (including 3,116 whites) (Vellut 1983: 131, Table 4.2).[6] Meanwhile, the white population of Lourenço Marques grew from around seventy-six in 1862 to 6,356 people by 1900, rising to four times this figure by the 1920s (Capela 1981: 11) as the port town developed. A racial division of labour developed, with the skilled and supervisory jobs increasingly allocated to whites: between 1910 and 1925 the number of whites employed at the port complex tripled (Penvenne 1995: 79–80, 82).

Divided workers and the regional strike wave of 1917–1925

From 1917 a large strike wave swept through parts of southern Africa. Out of 199 strikes recorded in government yearbooks for 1906 to 1920, a full 168 took place between 1916 and 1920 (Pike 1988: 103–5). There were 205 strikes from 1916 to 1922, involving 175,664 workers (Simons and Simons [1969] 1983: 333). Unionism also grew rapidly, with union membership rising from perhaps 9,178 in 1914 to 40,000 in 1917 to 135,140 in 1920 (also see Cope c1943: 200; Van Duin 1990: 640: 39). An important development in this period was the emergence of unions amongst Africans, coloureds and Indians in Bloemfontein, Durban, Cape Town, Johannesburg, Kimberley, and Port Elizabeth. Mozambique, meanwhile, was rocked by a strike wave from 1917 to 1921, headed by the powerful Port and Railway Employees Association (see Capela 1981; Penvenne 1984).[7] In South West Africa, popular unrest also grew in this period, and rumours of a general uprising by the Africans swept the country in 1922 and 1923 (also see Emmett 1986: 8–15; Gottschalk 1978; La Guma [1964] 1997: 17–21; Katjavivi 1988: 17–19). While rising inflation played an important role in the unrest of the times, the international wave of proletarian and colonial revolt that started in 1916 in Ireland and Mexico, and the hopes of change that followed, was also important. Local developments were often important catalysts: the unrest in South West Africa, for example, followed directly from the rising (and quickly disappointed) expectations following the South African takeover in 1915.

Across the region white workers fought bitter battles against employers and the state in the 1910s and 1920s around issues ranging from wages to replacement by African labour, and the ideology of White Labourism played a central role. The best-known case is, of course, the struggles in South Africa that culminated in the Rand Revolt of 1922, a general strike centred on the resistance to the replacement of white by African miners, which spiralled into a violent insurrection on the Witwatersrand that was suppressed with martial law.

However, the Rand Revolt was only one episode in the regional wave of class struggles, and only one of a series of dramatic confrontations between white labour and white capital. As early as 1916, a European Railway Workers’ Union had managed to establish the job colour bar on the Rhodesian Railways, a private company that operated in both the Rhodesias (Berger 1974: 45). In 1920, however, the (Southern) Rhodesian Mine Owners’ Association was formed to wage an onslaught on white miners, who had organised a general union from 1919, and the Association succeeded in inflicting a decisive defeat in early 1921 with a three-week lockout. Inspired by this victory, employers took the offensive on the railways, inflicting a crushing defeat on the railway union in 1929.

An explicit ideology of White Labourism was not, it seems, developed in the labour movement of Lourenço Marques, perhaps because many unions retained close links to anarcho-syndicalist and socialist movements in Portugal. However, the unions that operated from 1905 to 1925 in Lourenço Marques were predominantly based amongst the Portuguese (Freire 2001: 41). Few attempts were made to organise workers of colour, and the African workers who did join in the strikes of the time alongside the whites were routinely marginalised in strike settlements (Penvenne 1984: 270–277; Penvenne 1995: 84–5). The strike wave in Mozambique faced increasing repression and was eventually suppressed by the military. This was echoed in Portuguese Angola, where troops were used to break a massive strike by white railwaymen in 1923.

By 1925 the Port and Railway Employees Association was once again ascendant, and was strong enough to organise a major strike on the railways: this was crushed early the next year, when the railway service was militarised, with strikers evicted from their homes, hundreds fired, and many key figures deported.[8] The unprecedented scale of the repression was a ‘shocking example for all workers in the city’, white and African alike, and effectively ended all labour action until the 1930s (Penvenne 1984: 273–4, 278). In May 1926, a fascist dictatorship was established in Portugal, and extended to the colonies, and a second strike by electricity workers and railwaymen in 1932 was given short shrift (Penvenne 1996: 460–461).

Following the crushing defeat of 1921, and the subsequent demise of their union, white miners in Southern Rhodesia were unorganised until the late 1930s (Phimister 1977: 196–7). From 1921, however, there were attempts to enrol these workers in the South African Mine Workers’ Union (the largest South African union, and a bastion of colour bar politics), which formed branches in both Rhodesias in 1936 (Berger 1974: 49; Phimister 1977: 196–7). The Southern Rhodesian section initially competed with a local union, later merging to form the Associated Mineworkers of Rhodesia, which was recognised by employers in 1938 (Phimister 1977: 197).

In the meantime, several labour parties were formed in Southern Rhodesia with trade union backing, championing job colour bars. In 1933 the Reform Party won the elections, with the support of white labour, and implemented a programme of industrial relations reform, job reservation, and import-substitution-industrialisation (Phimister 1988: 173, 179–181, 189–192, 195, 248; also see Bond, et al. 2001). This was very similar to the platform of the Pact government elected in South Africa in 1924, a coalition of the National Party, an Afrikaner nationalist group that largely represented agrarian interests,[9] and the SA Labour Party.

Charlie Harris, secretary of the South African Mine Workers’ Union, travelled north to ‘help make Northern Rhodesia a white country’ and ensure that ‘union men only would be employed on the copper belt’ (quoted in Meebelo 1986: 107–8). The colony’s branch of the South African union eventually evolved into a separate body (Berger 1974: 49; Meebelo 1986: 107–8), the Northern Rhodesian (or European) Mineworkers’ Union. This was recognised by employers in 1937, and obtained a ‘gentlemen’s agreement’ securing the job colour bar in 1938. It faced competition from a Mine Workers’ Federation, but in 1939 organised 1,000 of the 2,500 daily paid white workers (Berger 1974: 49), and in 1940, militants from the union led a week-long strike.

Harris found a particularly sympathetic audience amongst white South African miners. Most white mine officials and miners came from South Africa (around 417 out of every 1,000 whites, as compared to 305 from the United Kingdom and Ireland (Berger 1974: 49–51), many of whom brought with them the segregationist politics of mainstream white South African labour (Meebelo 1986: 64). A labour party was formed in Northern Rhodesia in 1941, which won several seats on the legislative council in 1941 and 1944; it collapsed in 1944, but several unionists were elected in the late 1940s (Berger 1974: 65–6, 98–9), as I will show below.

Radical White Labour, Syndicalism and Communism

It is, however, important to understand that the politics of the white working class in southern Africa were not homogenous. The rise of White Labourism in the larger British working class diaspora was not to be seen simply as a process whereby the ‘The Imperial Working Class makes itself “White”’, stressing its ‘whiteness’ to claim racial privileges (contra. Hyslop 1999). If the structure of capitalism in southern Africa in the period of the first modern globalisation – characterised by ongoing attempts to substitute cheap unfree African labour for expensive free white labour – prompted many white workers to champion a ‘job colour bar’ as a defensive response (Johnstone 1976: 13–75), it simply does not follow that the ‘class interests’ of white workers were ‘dependent’ on ‘some degree of a job colour bar’ (Phimister 1977: 203), that White Labourism was inevitable, or that white identity was simply (or even primarily) a vehicle to advance the politics of the colour bar.

The choice between exclusive forms of labour organisation and more inclusive approaches that played out elsewhere in the world also played out in southern Africa. If many white workers ‘imagined themselves as citizens of what might be called the Empire of Labour’ that combined a ‘project of racial domination and class struggle’ (Hyslop 2002: 3–4), the ‘Empire of Labour’ was only one of many projects amongst white workers. The British labour diaspora in southern Africa also had more radical and internationalist wings, including more left wing and interracial forms of labourism, as well as syndicalism and, later, Communism that could potentially foster interracial movments.

The Northern Rhodesian Mine Workers’ Union, for example, made an unsuccessful move in 1943 to form an African wing to forestall the emergence of independent African unions (Berger 1974: 89). In Southern Rhodesia, the miners’ union and the Labour Party were deeply divided over the racial question, and the latter split in 1941 into an ‘exclusive white socialist’ Labour Party and a ‘non-racial’ Southern Rhodesian Labour Party that set out to organise an African Branch (see Lessing 1995: chapter 14; Phimister 1977: 199–201; Raftopolous 1997: 72; Ranger 1970: 168–169).

When Kier Hardie of the Independent Labour Party (formed in Britain in 1893), and then part of the British Labour Party, visited South Africa in 1908, he was met with violent opposition by sections of white workers and the white middle class for his outspoken criticisms of colonial policies. In Cape Town, however, the colour bar tradition was weaker: the racial division of labour was less stark, coloureds formed an important part of the artisan layer, unfree African workers were a minority, a number of craft unions had a multiracial membership,[10] and a radical left was fairly well-established. There were no incidents surrounding Hardie’s arrival, although a planned reception by the local trades and labour council was withdrawn following the events elsewhere in the country (Harrison n.d.: – 22).

Hardie was, instead, hosted by the local Social Democratic Federation (SDF), and held a meeting that was ‘far and away the most enthusiastic I had in South Africa’.[11] The local SDF, led by the self-described ‘philosophical anarchist’ Wilfred Harrison, was a far more open body than its British namesake, with a membership including Marxists, ‘anarchists, reform socialists, [and] guild socialists’ (Harrison n.d.: 16, 118–119). An active body, the SDF was an important influence on local popular political culture, formed an interracial General Workers’ Union, operated in District Six, the heartland of the coloured working class, and organised multiracial demonstrations of the unemployed in 1906. In 1908, the SDF ran a candidate in District Six, where a number of coloureds had the vote, benefiting from its image as a champion of coloured working-class interests (Ticktin 1973: 337–8), losing the poll but still receiving hundreds of votes.

The SDF, with its large ‘anarchist section’, was part of a larger, if minority, liber-tarian socialist current in the white working class. Outside of Cape Town, this current was often influenced by syndicalism and the IWW – currents rooted in anarchism that wished to form a revolutionary ‘one big union’ of all workers to expropriate the means of production. The 1890s to the 1920s were the ‘glorious period’ of syndicalism worldwide (Beyer-Arnesen 1997–1998: 20), and in many countries ‘the bulk of the revolutionary left was anarcho-syndicalist, or at least much closer to the ideas and the mood of anarcho-syndicalism than to that of clas-sical Marxism’ (Hobsbawm 1993: 72–3). The IWW in the United States was an important part of this upsurge, and it inspired IWW unions and groups across the settler colonies of the British Empire as well as IWW unions in Latin America and elsewhere.

Given the transnational character of labour in South Africa, it is not surprising that ‘In common with the labour movements elsewhere in the world, South Africa passed through a period of ... disillusion ... in the value of parliamentary reform’ that spread from ‘Britain, America, Australia and New Zealand’ and took up the ‘ringing call to action’ of the IWW and ‘the doctrines of the revolutionary syndicalists’ (Cope c1943: 108–110, also see 99–100). On the Witwatersrand, Scottish activists like Andrew Dunbar and JM Gibson promoted the IWW ideas that they had imbibed in Edinburgh and Glasgow, the Irishman Tom Glynn championed the syndicalism he had learnt in New Zealand, and Jewish immigrants from the Pale in East Europe snapped up copies of the anarcho-syndicalist Kleb i Volya (Bread and Liberty), produced by Peter Kropotkin’s Freedom group in London.[12] The Socialist, the syndicalist paper from Edinburgh, was available locally, as was the International Socialist Review of Chicago, which carried a wide range of syndicalist material.

The Voice of Labour, published in Johannesburg by the immigrant radicals Archie Crawford and Mary Fitzgerald from 1908, also provided an important conduit for syndicalism and anarchism more generally, and provided the space where activists like the anarchist Henry Glasse, who was closely associated with Freedom, could promote the ‘one big union’: ‘For a white worker in this South Africa to pretend he can successfully fight his battle independent of the coloured wage slaves – the vast majority – is, to my mind, simply idiocy.’[13] A visit to South Africa by Tom Mann in early 1910 gave local syndicalism a further boost. Mann, a founder member of the Independent Labour Party in Britain, had broken with Labourism while living in Australia, and was en route to Britain and France to promote and study his new-found syndicalist doctrines.

Mann’s visit was the immediate impetus to the formation of two syndicalist groups on the Witwatersrand in 1910: the IWW, associated with Dunbar and Glynn, and mainly based in Johannesburg, with groups in Durban and Pretoria; and the Socialist Labour Party, associated with Gibson, PR Roux and others, which was primarily a Johannesburg group. Following a dramatic IWW strike on the tramways of Johannesburg, Glynn returned to Ireland in 1911, then moved to the United States, and ended up in Australia, where he edited the IWW’s Direct Action, campaigned against the White Australia Policy, and was arrested for treason in 1916 (Burgmann 1995: 36, 77, 88, 207).

While the Witwatersrand left was in a state of collapse at the start of 1914, organised syndicalism revived after the formation of a War on War League after the outbreak of the First World War. Initially a group that struggled, ultimately without success, to keep the SA Labour Party to an anti-war position, the group provided a pole of attraction for SA Labour Party radicals, and the IWW and Socialist Labour Party veterans. It was re-organised as the syndicalist International Socialist League (ISL) in 1915, which left the SA Labour Party and resolved ‘that we encourage the organisation of the workers on industrial or class lines, irrespec-tive of race, colour or creed, as the most effective means of providing the necess-ary force for the emancipation of the workers’.[14] The ISL established sections across the Witwatersrand, had sections in Natal and the Cape, and worked closely with the SDF, which contributed to, and distributed, the ISL’s weekly paper, The International.

Drawing initially on radical British workers, the ISL recruited increasingly amongst Jewish immigrants. If the IWW and Socialist Labour Party had opposed racial discrimination and prejudice in principle, but done little in the way of organising across racial lines, the ISL took a more activist line, calling on Africans to ‘organise industrially’ against ‘any tyrannical law’.[15] In addition to ongoing attempts to reform the white unions, its great achievement was to form syndicalist unions amongst workers of colour from 1917: the Indian Workers’ Industrial Union in Durban, the Clothing Workers’ Industrial Union in Durban, Johannesburg and Kimberley and the Industrial Workers’ of Africa in Cape Town and Johannesburg: modelled on the IWW,[16] probably the first union for Africans in Britain’s African empire.

Through these initiatives, syndicalism became increasingly multiracial, as the ISL recruiting activists from the syndicalist unions like Reuben (Alfred) Cetiwe, Johnny Gomas, Hamilton Kraai, Bernard Sigamoney, and TW Thibedi, who argued that ‘as soon as all of your fellow workers are organised, then we can see what we can do to abolish the capitalist system’.[17] In Cape Town, meanwhile, the SDF split, giving rise to an Industrial Socialist League with an IWW platform.[18] Influenced by ongoing visits by IWW sailors,[19] it also managed to involve ‘coloured and Malay comrades in our propaganda ... amongst the coloured and native workers’,[20] and formed a syndicalist Sweet and Jam Workers’ Industrial Union. Initiatives like the Industrial Workers of Africa profoundly impressed a layer in the coloured rights group the African Peoples’ Organisation, and formed an early bridge between local socialists and nationalists. Cetiwe, Kraai and Thibedi were, moreover, active in the left-wing of the South African Native National Congress (renamed the African National Congress, or ANC, in 1923) (for more information, see Van der Walt 1999, 2004).

These syndicalist unions were formed, in short, in the context of the rising tide of labour militancy in the region, in which they played a pioneering role, and should also be seen as part of the global wave of syndicalism then sweeping the world. They made an important contribution to the growth of socialism amongst workers of colour in South Africa – and, as I will show, this influence later echoed across the region.

As popular militancy rose, and radical ideas gained currency, there was even a move to launch a general strike amongst the African workers of the Witwatersrand. Following the arrest of African municipal workers for strike action, a series of mass meetings were held in Johannesburg in June 1918, presided over by the Industrial Workers of Africa, the ISL, and the Transvaal Native Congress, a section of the African nationalist South African Native National Congress. A general strike was proposed by members of the Industrial Workers of Africa, supported by the ISL and the left wing of the Transvaal Native Congress, and set for 1 July. It was called off at the last moment, but several thousand African miners did not hear the news on time, and clashed violently with armed police.

Soon afterwards, the authorities arrested and charged five Africans and three whites for incitement to public violence for their role in the strike movement. Six of the arrestees were members of syndicalist groups, and included Cetiwe and Kraai in their number, as well as other ISL activists like SP Bunting; the remaining two people arrested were Thomas Levi Mvabaza and Daniel Simon Letanka, who produced the South African Native Congress paper, Abantu-Batho. ‘A matter of exceptional interest in this case is the fact that for the first time in South Africa, members of the European and Native races, in common cause united, were arrested and charged together for their political activities’ (Skota n.d.: 171).

There was also a radical wing in the labour movement in Lourenço Marques. Anarchism was an important factor in Portugal since the nineteenth century, and syndicalists came to dominate the unions in the early decades of the twentieth century (for an overview, see Freire 2001; Van der Linden 1999). Anarchism and syndicalism spread into Mozambique by the immigration of Portuguese activists, as well as due to the official practice of deporting radicals for imprisonment in the colonies (Capela 1981: 20; Freire 2001: 16). One notable deportee was the anarchist José Estevam: released in Lourenço Marques in 1910, he was returned to jail after forming a revolutionary league, only to reappear as a leading figure in the railway unrest of 1915 (Capela 1981: 65; Freire 2001: 16).

A Batalha, the Portuguese syndicalist daily, was widely distributed in Lourenço Marques, and the anarchists and syndicalists were a vocal minority in the Port and Railway Employees Association and the local labour press (Capela 1981: 19–36). They also set up radical circles, like the Grupo Libertaŕio Francisco Ferrer (named after a Spanish anarchist martyr), and also had a presence at the May Day demonstrations (Capela 1981: 277–228). Although the anarchist and syndicalist movement in Mozambique developed very few links to its South African counterpart, the ISL press was distributed in Lourenço Marques and some contacts were made.[21]

Anarchists and syndicalists also had some influence in the town’s lively associational life and intellectual culture, which drew in educated Africans, coloureds (mulattos) and Goans (Penvenne 1996: 428, 443, 458). Joao Dos Santos Albasini – the town’s most prominent intellectual, its leading African personality, and editor of O Africano, later relaunched as O Brado Africano – was, for example, familiar with a wide range of ideas circulating within the city’s ‘intense cafećulture’, including ‘socialist, anarco-syndicalist [sic], Masonic and other’ currents. Basically a moderate republican, Albasini was nonetheless sympathetic to labour of all races, and his articles periodically echoed the socialist rhetoric of the left wing of white labour (Penvenne 1996: 443–4, 449, 450–451).

The 1920s saw the rise of the communist parties worldwide, and this had some impact in southern Africa. The Communist Party of South Africa (CPSA), founded in 1921, drew heavily on the older anarchist and syndicalist milieu, and had a syndicalist current in its early years. The Communist Party of Portugal (PCP) was formed the same year, but does not seem to have had a section in Mozambique. It is, however, noteworthy that several CPSA activists were in Lourenço Marques during the 1925–1926 strike, and were deported in the repression;[22] Faustino da Silva, one of the few Mozambican unionists to identify openly with Communism, was exiled to South Africa at the same time (Capela 1981: 34–36).

Parties do not seem to have been formed elsewhere in the region at the time. There was an overwhelmingly white and short-lived Communist Group (later styled the Communist Party) in Salisbury in the 1930s and 1940s, which had links to the left wing of the in Southern Rhodesian Labour Party (Lessing 1995: chapters 11 to 17). It had ongoing contact with the CPSA, which in turn had some impact in the region, as will be noted below.

Jack Hodgson, a white miner from South Africa active in unions in Northern Rhodesia in the late 1930s, was barred from returning in 1940 after a visit to South Africa. He joined the CPSA and Springbok Legion as a serviceman during the Second World War, was active in the anti-apartheid struggle in the 1950s and 1960s, and fled abroad in 1963. In the 1930s, Hodgson worked closely with Roy Welensky, who was the Southern Rhodesia-born chairman of the Northern Rhodesian branch of the European Railway Workers’ Union, and an opponent of the small (but vociferous) local section of the British Union of Fascists (Macmillan and Shapiro 1999: 233–239). A founder member of the Northern Rhodesia Labour Party in 1941, Welensky was elected to the Northern Rhodesian Legislative Council in 1938 for the mining constituency of Broken Hill (now Kabwe), retained a seat after the party collapsed, and became the president of the short-lived Federation of Rhodesia and Nyasaland in the 1950s.

The ICU: Garveyism meets the IWW

Transnational connections also, of course, played an important role in nationalist politics in southern Africa. The African intelligentsia of South Africa was deeply influenced by African-American intellectuals, notably Alexander Crummel, Booker T Washington and WEB Du Bois (see, for example Masilela 2003). Not only were pioneer African nationalists like John L Dube, Sol Plaatje and DDT

Jabavu educated in black colleges in the United States, but the early debates between Washington and Du Bois were played out in South African circles in the late nineteenth and early twentieth centuries. These African nationalists also looked to Asia for inspiration, and it was the Indian National Congress that provided the template for the South African Native National Congress, as well as the Natal Indian Congress. The Native Congress provided, in turn, the template for the early African nationalist groups in Southern Rhodesia, and it is not a coinci-dence that the main African nationalist groups in both of the Rhodesias were named the ANC for many years.

Labour migration and political interchange were major elements of these African connections. The South African ANC provided advice, aid and regular visitors to the nascent African nationalist movement in Southern Rhodesia in the 1910s and 1920s, also maintaining close connections with ‘black settlers’ – the Africans who had accompanied Cecil John Rhodes north (Phimister 1988: 148–153; Ranger 1970: 56–63). If these early connections were largely an affair of the small African elite, African working class migration also played an important role in spreading ideas and movements. Migrant workers returning from the mines of the Witwatersrand spread Protestant Christianity across the hinterland of nominally Catholic Mozambique (Harries 1994:chapter 4). Labour migration was also crucial in the spread of the Watch Tower Christian sect in the Rhodesias and Nyasaland, which was a millenarian movement with anti-colonial elements that drew on the doctrines of Charles Taze Russell, founder of the Jehovah’s Witnesses.

Marcus Garvey – a Jamaican opponent of Du Bois, and, surprisingly, an admirer of the more moderate Washington – formed the Universal Negro Improvement Association (UNIA) in 1914, an organisation that would have an important impact in southern Africa. By the time of the UNIA’s 1920 world congress in New York, the organisation claimed branches worldwide, claiming four million members committed to Garvey’s Pan-Africanist vision of racial unity and ‘Africa for the Africans’ (see Cronon [1955] 1969). This is an improbable figure, and Du Bois, for one, believed the UNIA membership to be rather smaller;[23] a staunch critic of Garvey, he organised separate Pan-African congresses in 1919, 1920, 1921, 1923, 1927 and 1945. Even so, there can be no doubt that Garvey’s doctrines were immensely popular worldwide, that his movement grew at a staggering pace before his conviction for mail fraud in 1923, and that it was far more important than Du Bois’ Pan-African movement for many years.

While the UNIA’s Negro World and Black Star Line of ships played an important propaganda role, it was migration, above all, that played the crucial part in the spread of Garveyism into southern Africa. Black sailors – and, more particularly Afro-Caribbean sailors – were central actors, bringing Garveyism to ports in South Africa and South West Africa (Hill and Pirio 1987). Many of these sailors were put ashore after the First World War as white sailors reclaimed their jobs: often assimilated into local coloured communities, many promoted ‘Ethiopian’ churches and the UNIA (see Cobley 1992: 357–368).

By 1921 there were UNIA branches in Cape Town, Durban and Windhoek, forming a regional base for the expansion of Garveyism throughout the subconti-nent (see West 2002). In one sense, the wheel had turned full circle: illustrating that political influences circulate, rather than simply diffuse outwards from greater Europe (Weinstein 2005). It seems that Garvey was inspired to form the UNIA following a shipboard conversation about conditions in colonial Africa with a West African friend and his Basutoland wife (Garvey 1923).

The ICU is best known for its explosive growth in South Africa, where it started with a few hundred members in Cape Town in 1919, and reached at least 100,000 members by 1927, but perhaps many more (the ICU’s record-keeping was poor, at best). However, the South African ICU was only part of the history of the union: it was an international body, with sections formed in South West Africa in 1920, in Southern Rhodesia in 1927, and in Northern Rhodesia in 1931. If the ICU was, in the most literal sense, a transnational movement, it was also a movement profoundly influenced by ideas circulating in the international proletarian public sphere of the times.

ICU doctrines were an unstable mixture of elements: in addition to moderate nationalism, liberalism and millenarian African Christianity, the union was profoundly and primarily influenced by the ideas of the UNIA and the IWW. Clements Kadalie, the ICU’s charismatic South African-based leader, was a great admirer of Garvey, while Afro-Caribbeans played an important role in the South African ICU (also see Bradford 1987: chapters 3 and 4; Cobley 1992: 368–370). South Africans and West Indians headed the ICU in South West Africa and there was a clear overlap, and close working relationship, between the ICU and the local UNIA (see Emmett 1986: 20–29); the same was later true of the ICU in Southern Rhodesia (West 2002: 338–340, 347–353).

This ‘transatlantic connection’ (Couzens 1982) between black America, the UNIA and Africans in southern Africa is well known, but there was another transatlantic connection – to working-class America and the interracial syndicalism of the IWW – that needs to be stressed. Like the historical importance of syndicalism more generally, the impact of syndicalism on the ICU has been greatly underesti-mated.[24] The ICU was not simply a nationalist movement, but consistently invoked the ideas of interracial workers’ solidarity, ‘one big union’ and the general strike. The 1925 ICU constitution, for instance, drew directly on that of the IWW (ICU [1925] 1972: 325–326):

Whereas the interest of the workers and those of the employers are opposed to each other ... a struggle must always obtain about the division of the products of human labour, until the workers through their industrial organisations take from the capitalist class the means of production, to be owned and controlled by the workers for the benefit of all, instead of for the profit of a few ... This is the goal for which the ICU strives along with all other organised workers throughout the world.

At the ICU conference in 1926, Kadalie called for ‘one big union’ of the workers of all races to ‘assist in abolishing the capitalist class, who were in reality only a small body but owned practically everything’; other speakers called for a general strike across South Africa.[25] Allison WG Champion, the most prominent ICU leader in South Africa after Kadalie, described the ICU a year later as ‘an industrial organisation’ for ‘industrial and political democracy by and through the emancipation ... of the African worker’ and the ‘co-operative commonwealth’, ‘nothing more and nothing less’.[26] That year the ICU congress returned to the theme of the general strike that ‘would paralyse South Africa as not a native would work’ (quoted in Bonner n.d.: 9–10).

The influence of IWW syndicalism on the ICU can be partly explained as a consequence of the rising popularity of syndicalism in its ‘glorious period’, with many syndicalist unions peaking after 1917 as part of global climate of unrest. The respect that the IWW won from many black activists in the United States doubtless also played a role: Du Bois, for example, had famously stated that ‘We respect the Industrial Workers of the World as one of the social and political movements in modern times that draws no colour line’ (quoted in Foner 1974: 159), while the black socialists linked to the Messenger in New York, like Chandler Owen and A Phillip Randolph, held joint meetings with IWW speakers, and championed the IWW and its politics amongst black workers (Spero and Harris 1931: 391–2).[27]

Official fears about the prospect of a ‘potentially explosive’ alliance between the IWW and the UNIA (Pawa 1975: 272–273) were, however, misplaced. Garvey was quite hostile to the American left, and while Hubert Harrison – the main editor of the Negro World in its early years – was a former Wobbly, he had moved to a ‘race first’ position, and had partially broken with the left (see Perry 2001).

Perhaps a more important vector for syndicalist influence on the ICU operated within South Africa: the ISL and the Industrial Workers of Africa. In the middle of 1919, Cetiwe and Kraai left for Cape Town where they formed an Industrial Workers of Africa branch in the Ndabeni township, which housed African dockworkers. While the syndicalist union initially competed with the ICU, the two bodies cooperated in organising a dramatic strike of 2,000 African and coloured dockworkers in December 1919, which was supported by the left wing of white labour, the ISL and the Industrial Socialist League (Kadalie 1970: 42; Wickens 1973: 67–80). The strike was not a success, for which the ICU blamed the Industrial Workers of Africa and its allies in the Cape Native Congress (Wickens 1973: 79–80).

It is noteworthy that Kadalie never mentioned the Industrial Workers of Africa in his later accounts of the 1919 strike (Kadalie 1970: 717), the event that propelled the ICU into the public eye. Perhaps this was partly a result of the controversy over the strike – Kadalie’s own tendency to identify the history of the ICU with his own person was also doubtless a factor, and it is a pity that later scholars have not viewed Kadalie’s claims about the union’s history with more scepticism.

Nonetheless, the Cape section of the Industrial Workers of Africa seems to have merged into the ICU sometime in 1920, and the ICU would long bear the imprint of the politics of the IWW. When a national conference of African and coloured unions was held in Bloemfontein in 1920, attended by the ICU, the Industrial Workers of Africa and other bodies, it aimed to ‘form one great union of skilled and unskilled workers of South Africa, south of the Zambesi [sic]’, although it also wanted to ‘settle differences’ with employers by ‘amicable and conciliatory means’ (quoted in Wickens 1973: 145–146). ‘We had,’ Kadalie later recalled, ‘the “one big union movement” in view’ (Kadalie in 1923, as quoted in Wickens 1973: 97).

The ICU cannot be understood, then, unless it is properly grasped as a movement that was influenced by two major currents, Garveyism and syndicalism, that came from across the Atlantic and which were spread by Afro-Caribbean and white immigrants, respectively. At the same time, the rivers of labour flowing within the sub-continent profoundly shaped the spread of the ICU northwards from South Africa, which gave it an additional element of transnationalism.

The first move to establish the ICU outside South Africa came from a South African migrant, James Arnold ‘Jimmy’ La Guma, a young coloured worker from Cape Town. Born in Bloemfontein in 1894, the son of an itinerant cobbler, La Guma was raised in District Six where he encountered ‘working class literature’ and was drawn into an interracial movement of the unemployed organised by the SDF in 1906 (La Guma [1964] 1997: 18). To Harrison’s annoy-ance (he was not in favour of violent action, although he was regularly arrested for inflammatory utterances), some marches turned into what were called the ‘hooli-gan riots’ (Harrison n.d.: 8–9). La Guma was actively involved in the looting that took place, which he found ‘a mixture of fun, adventure and participation in the class struggle’ (La Guma [1964] 1997: 18).

Still unable to find work, and ‘seized’ with ‘wanderlust’, La Guma left for South West Africa around 1907 or 1908 (La Guma [1964] 1997: 17–21), one of thousands of local Africans and coloureds recruited by the German colonial authorities at Cape Town (Beinart 1981; Beinart 1987:especially 167–170, 178–9; Moorson 1978). Working on farms, diamond fields and the docks, La Guma organised a strike in 1918, part of the strike wave that followed the South African takeover (also see Emmett 1986: 8–15; Gottschalk 1978; La Guma [1964] 1997: 17–21; Katjavivi 1988: 17–19). Upon hearing of the 1919 Cape Town strike, La Guma got in touch with Kadalie, and set up an ICU branch at Luderitz in December 1920 (La Guma [1964] 1997: 714). He apparently considered himself a socialist by this stage (La Guma [1964] 1997: 18, 24–9), although of what brand, it is dif-ficult to say; the president of the South West African ICU was John de Clue, an Afro-Caribbean influenced by Garveyism (Emmett 1986: 21).

The Luderitz ICU was viewed by authorities as a ‘movement of well-educated coloureds’, albeit one that with a real potential to spread to local Africans (Emmett 1986: 32; Peltola 1995: 77–78), and moves were consequently made to restrict the ICU to the Luderitz townships and the ‘educated class’, as well as to isolate it from South Africa (Emmett 1986: 32). Nonetheless, communication was maintained through the ICU press and correspondence with Kadalie, and the South West African ICU’s ideas seem to have been very similar to those of the South African ICU (Emmett 1986: 21–22, 32, 37).

If the spread of the ICU into South West Africa was linked to the flow of labour from South Africa to the north, the spread of the ICU into the Rhodesias was linked to the flow of labour from the north into South Africa and Southern Rhodesia. Kadalie himself was an educated migrant from Nyasaland: unable to speak local African languages, and far better educated than the average ICU members, he was often taken to be ‘an American Negro’ (Kadalie 1970: 54–5). Born in Chifira Village near the Bandawe Mission Station on the shores of Lake Nyasa to a prominent Tonga chiefly lineage, Kadalie trained at the Overtoun Institution of the Livingstonia Mission as a schoolteacher, but left his country in 1915 ‘in quest of a higher civilized life’ (Kadalie 1970: 33).

Kadalie was part of a larger group of mission-educated men from northern Nyasaland, who left the country in search of employment elsewhere (see McCracken 2000:chapters 5 and 6). While Africans were employed in administrative and artisan positions in Nyasaland to an extent unknown elsewhere in the region, the large network of mission schools, coupled to limited job opportunities and commercial opportunities, led to a continual flow to the south (as well as the Belgian Congo in the west) of educated Nyasas (McCracken 2000: 155–156, 179–180, 187–188, 194–195). The group was heavily drawn from northern Nyasaland, from which Overtoun drew most of its students, and were dispropor-tionately represented amongst African mining clerks across southern Africa, a distinctive layer marked by education and class identity, but trapped by the colour bar (Van Onselen 1976: 118–121, 206–7). Kadalie, for example, worked first in Mozambique, then in Southern Rhodesia as a clerk, and it was here that white racial prejudice radicalised him, before he left for South Africa (Kadalie 1970: 34–7; Ranger 1970: 150).

The network of educated northern Nyasas was crucial in spreading the ICU to the two Rhodesias. The British South Africa Company’s fear that ‘one big union’ would emerge in the northern territories began to be realised when African workers from Southern Rhodesia requested aid at the 1927 ICU conference (Phimister 1988: 189; Phimister and Van Onselen 1997: 39; Van Onselen 1976: 210). Kadalie retained a deep interest in Southern Rhodesia, and the conference agreed to establish a Southern Rhodesian ICU, which was supposed to transfer twenty per cent of its income to the parent body (Ranger 1970: 150, 155, 163). The task of setting up the ICU in Rhodesia was given to Robert Sambo – an educated Nyasa friend of Kadalie’s – and, with the aid of yet another compatriot, John Mphamba, Sambo founded the ICU in Bulawayo’s African location (Ranger 1970: 151; also see West 2002: 351).

In March 1927 the union claimed 155 members and was growing rapidly when Sambo was deported, returning to Nyasaland where he later became a leading figure in the African independent church movement (McCracken 2000: 322). Kadalie described Sambo’s deportation as ‘in the best traditions of a capitalists’ democracy’, and quickly dispatched Masotsha Ndhlovu to Bulawayo (Parry 1999: 81; Phimister 1988: 158; Ranger 1970: 152; Van Onselen 1976: 210–1; West 2002: 351). Exemplifying the importance of migrancy to the spread of the ICU, Ndhlovu was a Southern Rhodesian who had worked in South Africa for many years, mainly in Cape Town (West 2002: 351). With the aid of Job Matabasi Dumbutjena Ndhlovu ‘travelled extensively, addressing meetings all along the labour route’, forming an ICU branch at Salisbury in 1929, and a further eleven branches over the next two years, including branches in the countryside (Phimister 1988: 158; Van Onselen 1976: 211–212).

In the meantime, the ICU appeared in Northern Rhodesia, and transnational migrant networks – with educated Nyasas again very prominent – played a key role. It ‘would have been surprising if some contacts between the northern labour force and more politically active southern workers did not exist’, given the human rivers linking South Africa and the Rhodesias (Berger 1974: 94). In the 1920s authorities in Northern Rhodesia were ‘painfully aware’ that returning migrants brought subversive views, worrying about the spread of ‘advanced ideas from the type of natives who recently sent a deputation to Moscow’ (quoted in Meebelo 1986: 46–7).

In 1924 police intercepted letters between Kadalie and his uncle Isaac Clements Muwamba – an Overtoun graduate who worked as a senior clerk in Lusaka (McCracken 2000: 190) – revealing that he was regularly receiving copies of the ICU’s paper, the Workers’ Herald, as well as the Messenger (Meebelo 1986: 47). Muwamba and Sam KK Mwase of Livingston, another Nyasa, ‘seem to have imbibed the proletarian ideas of Clements Kadalie and his Industrial and Commercial Workers’ Union’ although neither ‘made any attempt to form a truly workers’ organisation along the lines of Kadalie’s union’ (Meebelo 1986: 52). Then in early 1932 an educated man called Joseph Kazembe ‘made some effort ... to form a branch of the Industrial and Commercial Workers’ Union in Livingstone’ (Meebelo 1986: 52, 101, 161), with apparent success. Kazembe had just been deported back to Northern Rhodesia from South Africa (Meebelo 1986: 52, 81, 161).

The politics of the ICU – a peculiar meeting of the ideas of syndicalism and Pan-Africanism – were inevitably wracked with tensions and ambiguities between internationalist class politics and nationalist race politics. The different elements jostled with one another across the region. In Southern Rhodesia for example, ICU speakers spoke, by turns, as moderate members of the middle class, as aspirant traders, as African nationalists, and as militant socialists. The ICU in Rhodesia sometimes saw whites in general as the enemy, but sometimes focused on the ‘capitalist class ... the capitalists who kept power by means of the government and the missionaries’ (quoted in Ranger 1970: 159, 164): this was ‘why we say let us organise’ in one union regardless of ethnic background or place of origin (quoted in Van Onselen 1976: 212). ‘The ICU is for proletarian people ... We are the proletarian people’ (quoted in Ranger 1970: 164–165). Kazembe in Northern Rhodesia was more likely to conflate race and class issues: in a speech during the 1932 African strike in Ndola, he stated that ‘the comfort enjoyed by the Europeans’ was ‘procured at the expense of the exploita-tion of the natives’ (Meebelo 1986: 53).

While the ICU was generally a spent force in the region by the mid-1930s, radical ideological currents continued to flow from South Africa northwards. The major union drives in South West Africa in the 1940s at Luderitz took place at the initiative of the Food and Canning Workers’ Union of South Africa, led by radicals and CPSA members (Peltola 1995: 78–79). In Southern Rhodesia, there was some contact between ICU activists and the CPSA (West 2002: 346–347), and the CPSA’s Umsebenzi also ‘circulated amongst a tiny audience of black schoolteachers and workers’ in the early 1930s: one Malikongwa Shoko even used CPSA material in his teaching (Phimister 1988: 198). The colonial government was also alarmed by the appearance of African trade union agitators from South Africa, like John Meshack Chamalula, graduate of a CPSA course on labour organising (Berger 1974: 94).

The Communist Party of Southern Rhodesia distributed the CPSA’s The Guardian in the 1940s, mainly amongst coloureds and whites (Lessing 1995: 269–270, 279–288). Charles Mzingeli revived the ICU in Salisbury in the 1940s as the Reformed ICU: it was one of the most important African political groups in the country well into the 1950s. Mzingeli worked with the Southern Rhodesian Labour Party and the local left Book Club, and wrote for the CPSA’s Inkululeko and The Guardian, although he was by no means a communist (Lessing 1995: 304–311; Phimister 1977: 199–201; Raftopolous 1997: 72; Ranger 1970: 168–169).

Conclusion: Transnational working class, transnational politics

In southern Africa before the 1940s, human rivers of labour transcended countries – even empires – and it was along these flows that activists, ideas and models moved. The southern African working class was, moreover, linked into larger intercontinental flows within and beyond the British Empire. If the capital invested in southern Africa was often international, the working class it helped generate was equally sprawled across borders. Many of the experiences of this working class could not be described as ‘national’, nor could its ideological influences and organisational development be neatly cordoned off, or effectively understood, within national borders.

The history of the working class in southern Africa, then, whether African, coloured, white or Indian, simply cannot be adequately understood through the use of a framework that takes the nation-state as the basic unit of analysis in working-class formation. Labour markets, ideas, union movements, and political groups repeatedly transgressed the borders of the colonial states. The lived experience of a regional political economy, linked into a world economy, and of imperial control rather than national sovereignty, was critical to this process, and the solidarities that developed within the popular classes – solidarities of class as well as ethnicity as well as race – were only occasionally linked to a definitively nationalist politics that aimed to carve the region into a set of nation-states before the 1940s. Such a situation cannot be understood if we remain confined by methodological nationalism.

The assumption that working classes are somehow organically national cannot easily be reconciled with the objective existence of transnational working classes, such as that of southern Africa. If a national model makes some sense in the era of closed national economies, lasting from the 1930s to the 1970s, it has limited usefulness for the analysis of labour and the working class in the two eras of globalisation that bracket that era. In periods like the ‘first’ and the ‘second’ globalisation, the possibility of labour internationalism need not be merely imagined: it corresponds to a lived reality. In such a context, with its inter-linked labour markets, corporations and states, ideas of international solidarity, based on ‘ties of economic solidarity and fraternal sentiment’ between the ‘workers in all occupations in all lands’ (Bakunin [1871] 1971: 249, 252), are by no means utopian.

References

[1] I would like to thank Phil Bonner, Jens Andersson, Hugh Macmillan, and the reviewers of this journal for their valuable comments on an earlier version of this article, which was presented at ‘Rethinking Worlds of Labour: Southern African Labour History in International Context’, organised by the History Workshop and the Sociology of Work Unit at the University of the Witwatersrand from 28 to 31 July 2006.

[2] An early and unduly neglected statement of this argument may be found in Rocker ([1937] 1978).

[3] South Africa exemplifies this process, while drawing attention to the rise of large-scale African migrancy: in 1886 Johannesburg had 3,000 prospectors; ten years later, it was a city of 100,000; by 1913 it was home to around 250,000 (Krut 1988: 135–136).

[4] Key party policies and documents from 1909 and 1910 may be found in Ticktin (1973: 487–495, 526–533), while the 1912–1913 constitution and platform may be found in Grobler (1968:appendix one, 498–519).

[5] Mozambique’s foreign trade rose 300 per cent between 1877 and 1892 (more merchandise passed through that port in the first six months of 1893 than in the previous five years), a railway line linked Delagoa Bay to the Witwatersrand from 1895, the systematic development of the harbour took place from 1900 onwards, and large-scale South African-based investments in real estate and construction occurred. By 1910 South African-based interests controlled utilities, shipping and handling, insurance and banking, but local business interests responded aggressively and the colonial state rapidly displaced foreign investors in these sectors (see Harries 1994: 141; Penvenne 1995: 17, 35).

[6] There was also a small mining sector in Portuguese Angola, north of South West Africa, with 8,697 employees (including 160 whites).

[7] ‘The Congo and Delagoa Strikes: the “white labour” fallacy revived’, The International 10 September 1920; ‘Troops: what for?’, The International 17 September 1920.

[8] Divisional Criminal Investigations Officer, Witwatersrand Division, 5 July 1926. ‘Confidential Report to Deputy Commissioner, South African Police, Witwatersrand Division, Johannesburg, in Department of Justice file, JUS 915 1/18/26 part 3’. Pretoria: National Archives.

[9] Another outcome of white migration in the region was the somewhat less successful, but far from insignificant, spread of Afrikaner nationalism. It developed a base in South West Africa, where there was a growing population of Afrikaner immigrants from both Angola and South Africa, and in 1924 the National Party of South West Africa was established. It was, however, unable to establish much of a foothold elsewhere, despite the widespread existence of Afrikaner communities. The profound alienation from Afrikaner nationalism felt by most whites in Southern Rhodesia accounted, in part, for the defeat of proposals in the 1920s to join South Africa as its fifth province.

[10] While the ‘Cape tradition’ should not be unduly exaggerated, it was very real (Bickford-Smith 1995). The number of coloureds in commercial and industrial occupations in the Western Cape rose from nine per cent in 1891 to twenty per cent in 1904, with the number of clerks, storekeepers and hawkers tripling and the number of masons doubling in this period (Goldin 1987).

[11] James Kier Hardie, ‘South Africa: Conclusions’, The Labour Leader 22 May 1908.

[12] Henry Glasse, ‘International Notes: South Africa’, Freedom November-December 1905.

[13] See, for example, The Voice of Labour 26 January 1912, letter from Henry Glasse.

[14] ‘League Conference’, The International 7 January 1916; ‘The First Conference of the League’, The International 14 January 1916.

[15] ‘The Pass Laws: organise for their abolition’, The International 19 October 1917.

[16] It was actually named the IWW at first, and officials regarded it as a ‘a branch of the wider organisation which would appear to have been suppressed in Australia and New Zealand’: Secretary of Native Affairs to Commissioner of Police, 14 November 1917, 983/17/F.473, in Department of Justice, ‘The ISL and Coloured Workers’, JD 3/527/17. Pretoria: National Archives.

[17] Cetiwe, reported by police detective in an unlabelled report, May 1918 (full date illegible), in Department of Justice, ‘The ISL and Coloured Workers’, JD 3/527/17. Pretoria: National Archives.

[18] Industrial Socialist League, February 1920, ‘Where WE Stand’, The Bolshevik.

[19] Manuel Lopes, ‘Cape Notes’, The International 24 January 1919.

[20] The Workers’ Dreadnought 7 August 1920, letter from Manuel Lopes.

[21] David Ivon Jones, ‘Friends across the Border’, The International 10 September 1920.

[22] Divisional Criminal Investigations Officer, Witwatersrand Division, 5 July 1926. ‘Confidential Report to Deputy Commissioner, South African Police, Witwatersrand Division, Johannesburg, in Department of Justice file, JUS 915 1/18/26 part 3’. Pretoria: National Archives.

[23] Du Bois estimated that the UNIA had 10,000 paid-up members, at most 20,000 active members, and perhaps 100,000 nominal members (see Rudwick 1959: 428).

[24] An important exception is P Bonner who characterises the ICU as ‘millenarian syndicalism’ (Bonner n.d.).

[25] Divisional Criminal Investigations Officer, Witwatersrand Division, 1 May 1926. ‘Confidential Report to Deputy Commissioner, South African Police, Witwatersrand Division, Johannesburg, in Department of Justice file, JUS 915 1/18/26 part 2’. Pretoria: National Archives.

[26] Allison WG Champion, 1927, The Truth About the ICU, the Roberts Printing Works for the African Workers’ Club, Durban: 5–7, folder on the Industrial and Commercial Workers’ Union in the IWW Collection, Collections of Archives of Labour and Urban Affairs, Walter P Reuther Library, Wayne State University, Detroit.

[27] Also see Claude McKay, ‘Socialism and the Negro’, The Workers’ Dreadnought 31 January 1920.