A thesis submitted to the Faculty of Humanities, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the Degree of Doctor of Philosophy. Johannesburg, August 2015. / This thesis explores the hypothesis that South Africa’s land reform programme is based on a set of assumptions about the country’s past which are inadequate and have contributed to the failure of policy. The impact of these assumptions is that they support particular models for restitution and rural economic development which became ‘accepted wisdom’ within international development agencies, government, and amongst land activists in South Africa, but which were and still are inappropriate in the South African context. To test this hypothesis I look at the history of land settlement and agriculture in Mopani district of Limpopo province. In particular, I look at how ordinary people accessed and lost rights to land over the nineteenth and twentieth century, and how land became tied up in struggles for political authority and access to resources. I show how the importance of ethnic identities and a sense that land belongs to ‘indigenous’ people of a particular ethnicity, deepened during the Bantustan era. I argue that policymakers could have learned from past government policies. This includes the 1936 Native Trust and Land Act which called for the state to purchase farms from private landowners for African settlement, and smallholder irrigation schemes and co-operatives, which were intended to improve agricultural production in the reserves and homelands. What this history reveals is that land settlement patterns and experiences of land loss were far more complex than the simplified narratives on which land restitution was based. The poor performance of agriculture in reserves and bantustans cannot be blamed on past government policies intended to destroy a peasantry, or on land loss alone, rather there were many challenges and constraints. Women maintained an interest in agriculture throughout the twentieth century, but were not taken seriously by those attempting to improve African agriculture. Africans interested in commercial farming were constrained in how much land they could access. The idea that Africans are naturally communal, and that restitution and development should target ‘communities’ is deeply problematic. Policy failed to take into account these realities. The consequences have been that land restitution has failed to bring redress, restituted farms have failed as commercial operations, those with a real interest in agriculture continue to face constraints, and smallholder irrigation schemes continue to perform poorly.