Carefully hidden away: excavating the archive of the Mapungubwe dead and their possessions

Ever since Jerry Van Graan first stumbled upon golden artefacts in 1933, Mapungubwe - an Iron Age civilisation that existed in the confluence of the Shashe and Limpopo rivers between 900 and 1300 AD - has been the subject of contestation. Initially knowledge production about Mapungubwe was informed by the need to make a case for the late arrival of Bantu-speaking people in Southern Africa ? a now discredited theory used to justify the subjugation of Africans. In the post-apartheid era, Mapungubwe became a focal point for a new form of myth-building: the myth of liberation and a romantic past but, in my view, with a neo-liberal bias. In this dissertation I interrogate the role played by the disciplines of archaeology and physical anthropology in the political contestation that has surrounded Mapungubwe, focusing on the production of knowledge. I do this by investigating the claim that Mapungubwe was shrouded or hidden away. In particular, I ask: What happens when disciplinary workings, in the course of knowledge production, construe an archive? What do museums, archives and other memory institutions hide and what do they reveal? What gets acknowledged as archive and what is disregarded? How is this knowledge presented in the public domain over time? Lastly, what happens when the archive is construed differently? My interrogation lays bare the continued discomfort and improvisation that prevails among those disciplines or institutions that engage with Mapungubwe. I have chosen to organise the core chapters of the thesis according to specific timeframes: before apartheid, during apartheid and after apartheid. This is done to demonstrate how archaeology, claimed as a science, was a powerful strategy deployed to exchange the messiness for the "true" knowledge of the past. The research on Mapungubwe, by way of the Greefswald Archaeological Project, was the most prolonged research project in the history of South Africa. Its four research phases, which began in 1933 and ended in 2000, mutated as the broader political landscape shifted. As a result, everything that can possibly play itself out in broader post-apartheid South Africa is present in Mapungubwe: contested claims, racial history, land dispossession, apartheid and the military, repatriation, post-apartheid claims, nationalism, pan-Africanism, ethnicity and more. This thesis demonstrates how the disciplinary practices of archaeology were instrumental in keeping Mapungubwe shrouded. An example of this "shrouding" is the deployment of highly technical language in writing about Mapungubwe. Before the end of apartheid, this epistemic hiding offered a convenient retreat for the discipline, to avoid engaging with issues facing South African society at large. This placed the discipline in a position of power, a position of "truth" and "objectivity". All inconvenient forms of knowledge were simply disregarded or silenced through choices, made by powerful institutions and individuals, about what was worthy of being archived. However, when the archive is differently construed, a different picture emerges