A Long Way Home. Deborah James

help, but the sons must pay some.’ He had bought cattle from other people and paid seven for his bride-wealth, to which his father added four. Thus earnings from migrant labour were gradually incorporated into the very reproduction of the homesteads. The amount paid in bride-wealth also increased: ‘Before it was two or three, but when people started going to work, it was 10 or 12 – and this was for ordinary people.’⁴

      Even by the 1940s, when men were taking more regular contracts on the mines, Native Recruiting Corporation records show that 45 per cent of recruits from Mpondoland deferred their pay in full and 60–70 per cent of the total earned by recruits from these districts was repatriated. Earnings from migrant labour were important for family subsistence, but also helped to sustain the rural economy through investment in livestock and arable production. Migrancy became necessary for most families, in order to pay taxes and purchase what became seen as necessities. However, contrary to some early analyses that saw migrancy as symptomatic of the collapse of rural production, evidence suggests that there was initially a positive, rather than inverse relationship between wage income and smallholder output.

      We can explore this in two ways – as an aggregate and with respect to individual homesteads.

      In Table 4.1, the 1914 figure shows cattle herds devastated by East Coast fever, but by 1932, numbers had reached 521 000 – almost certainly at an all-time high. Booming livestock numbers, facilitated by dipping against tick-borne diseases, coincided closely with the rise in migration. Cattle prices were low at this time, so wages went further if invested in livestock. Production of maize (Table 4.2) also doubled in the early decades of the twentieth century, so that 1939 (at 810 000 bags) was probably the all-time peak year.

      In the aggregate, the standard of living for Transkeian families, especially in the Mpondoland districts, probably improved until the 1930s. The positive relationship between migrant earnings and agricultural production continued until the 1960s – and possibly longer. Although aggregate production probably then declined, surveys emphasise the continuing link between wage income and investment in agriculture for individual homesteads. In Tsolo in 1976, families earning above R40 per month cultivated three times more land than those earning R10–20 and they also got slightly higher yields; similar correlations were found in Matatiele and in Libode in the 1980s.⁵ Since the 1990s, however, this relationship seems to have broken down for arable production, even in wealthier homesteads.

       Migrant identities, rural families and associational life

      In Going for Gold, Dunbar Moodie attempted to identify different ‘migrant cultures’ on the mines, based on interviews conducted in the 1970s. He records that ‘miners from other groups talk of the proverbial stinginess of the Mpondo’, who were there ‘on business’.⁶ Following the analysis of Philip and Iona Mayer of the relative ‘encapsulation’ of traditionalist migrant workers in East London, Moodie argued that his Mpondo interviewees shared a ‘commitment to the independence and satisfactions of patriarchal proprietorship over a rural homestead’ and that this ‘implied resistance to proletarianisation’. Their social and economic base in the rural areas remained a priority and manhood ‘was achieved essentially in presiding justly, wisely, and generously over an umzi, or rural homestead’.⁷

Table 4.1 Average cattle holdings in seven Mpondoland districts

Year	Number of cattle
1904 census	135 000 (still recovering from Rinderpest)
1911 census	280 000
1914	76 000 (after East Coast fever)
1910s	127 000 (number of figures =5)
1920s	312 000 (n=10)
1930s	500 000 (n=10)
1940s	453 000 (n=10)
1950s	401 000 (n=10)
1960s	400 000 (n=3 – published series ceases after 1963)

Table 4.2 Average maize production in seven Mpondoland districts

Year	Number of bags of maize
1920s	545 000 bags (n=7)
1930s	544 000 bags (n=7)
1940s	542 000 bags (n=4)
1950s	499 000 bags (n=10)
1960s	627 000 bags (n=5)

      Source: Figures taken from the Census, Agricultural Census and Transkeian Territories General Council reports

      Migrants from Mpondoland, however, also absorbed cultural practices and patterns of masculinity forged in underground work and the compounds that, in some respects, mitigated the harshness of conditions on the mines. They were known for their clannish behaviour and assertiveness. When I conducted interviews in the 1970s, I was not especially looking for such material, but it was sometimes forthcoming. Xatsha Cingo remembered: ‘The Mpondo were in their own rooms; the Xhosa lived separately. It was all right as long as a man was Mpondo, it did not matter if he was from Nyandeni [Western Mpondoland] but no Xhosa were allowed. We were not on good terms.’⁸ The language of difference between Mpondo and Xhosa on the mines was often remembered as focusing on circumcision. In the nineteenth century, the Mpondo, as in the case of the Zulu kings, abolished male circumcision, which remained central to most African societies in the region. Lionel Mathandabuzo from Lusikisiki recollected that ‘when they go to the mines, they meet with the Xhosas and they are insulted as boys.’⁹ Leonard Mdingi, from Bizana, worked for four contracts on the mines in the 1940s and blamed employers for the strength of ethnic identity. But he recognised the sense of difference: ‘When the Xhosas speak of Pondos as boys, the Pondos don’t take it seriously. Because the Xhosas want the Pondos to adopt their customs and the Pondos don’t want that.’¹⁰

      Moodie analysed the way that collective action by miners in the compounds enabled them to forge some space and influence over their lives. These solidarities could also result in conflict and Mpondo workers were perceived to have a predisposition to violence. Mdingi recalled: