Subsequent studies of magic in non-Western, tribal cultures, while sharing EvansPritchard’s presumption of the empirical non-efficacy of magic and related practices such as witchcraft and sorcery, have tended, probably from a laudable concern to protect ‘primitive’ peoples from accusations of irrationality, to emphasize the ‘symbolic’ and ‘metaphorical’ elements in these practices. In the words of †John Beattie (1964), in magic ‘we are dealing with the imagined potency of symbols and symbolism’. The whole basis of this approach, which rested on a †worldview inherited from the Enlightenment, has been brought into question since the so-called *postmodern critique of anthropology of the late 1970s and 1980s. Inspired in some case by the controversial works of Carlos Castaneda, which introduced readers to the ‘separate reality’ supposedly known to the Yaqui Native Americans of the Mexican Highlands, certain anthropologists have written of magical phenomena as objectively real, even if inexplicable in terms of Western scientific knowledge. An early example is Michael Harner’s account of the magical world of Conibo and Jivaro *shamans of South America. Other recent examples of this school are the works of Stoller and Olkes, de Surgy and Edith Turner, all based in African field experience. Michael Jackson, another anthropologist of Africa who also adheres to this approach, has dubbed it ‘radical empiricism’.