While writing this book I became increasingly sceptical about the received categories that framed this discourse in anthropology and, more generally, in African Studies and the popular literature on ‘traditional healing’. Throughout this text, then, I challenge the many misconceptions held in the worlds of government, public health and medicine, in popular literature and in much of the scholarly literature, including anthropology, my own discipline. Most importantly, I argue that bungoma is not, as is often supposed, about witchcraft or ‘muti killings’. It is also not primarily concerned with curing physical disease. It is not a makeshift psychology and does not provide either ‘barefoot doctor’ services on the cheap, or bush psychotherapy. It stands on its own, as I have tried to show.

Critique would be sterile if I could not offer an alternative way of seeing – that is theorising – the practices, knowledge and beliefs in the context and time that concern me here. I have tried, then, to present a more-or- less coherent set of terms and concepts that might allow us to re-imagine this rich fund of African knowledge and give it its due.

Much of what I say about the little corner of South Africa in which I worked can be generalised. Ngoma, as practiced in Barberton and Umjindi, is but one facet of a vast set of flexible yet sturdy concepts and practices – all designated by the root word – ‘ngoma’ – that are pervasive throughout the southern half of Africa. Ngoma in its many forms has been understood as ‘cults of affliction’, as ‘deep knowledge’, as ‘tradition’ itself. Its simplest, referential meaning is ‘drum’ and ‘dance’. But it is all of these and more. I have tried to show ways in which it is, or can be understood as political, economic, ritual, ecological, philosophical, medical and even empirical. Whatever else it is, however, it is not religion and it is not tribal. Thus, I argue that it cannot be usefully comprehended as, for instance, ‘Zulu religion’.