If bungoma is not a deficient or ‘traditional’ medicine, and not a science, is it not then a form of religion? If traditional healing is not fully traditional, can it be reduced to a single task that we call ‘healing’? Does its use of ritual, formulaic language, music, song and intangible person-like agents make it a kind of African traditional religion? Some scholars have argued that it is indeed a form of traditional religion (Chidester 1996; Chidester et al. 1997; Dovey and Mjingwana 1985; Kruger 1995; Mbiti 1975, 1990; Setiloane 1973, 1976, 1986; Smith 1929) and that it must be respected as such. But the argument that traditional healing is a limited, early, or ‘traditional’ form of religion – or, in the words of the missionary Edwin Smith (1929), a ‘twilight religion’ – has the same drawback as representing it as a simpler and alternative medicine. Both positions rest on the assumption that bungoma is a deficient form of something else, either a not-quite religion, or a ‘barefoot’ medical practice rooted in superstition and misunderstanding.

By presenting the philosophy of traditional medicine in its own terms we can ask in what sense it might also be a form of medicine or a form of religion. I would claim that bungoma is certainly a form of medical practice, even though it is not only a medical practice. It is not, however a form of religion, that is, it is not a deficient theology. Surprisingly, the earliest Christian missionaries agreed that it was not religion.

In order to understand how and in what ways southern Africans got along perfectly well without ‘religion’ we can explore why early Christian missionaries agreed that there was no religion in southern Africa.