CONFESSIONS SUCH AS THESE, ADMITTING ONE OR OTHER GRAVELY antisocial horror, are no rarity in the courts of Africa. While they by no means prove that the actions confessed were actually committed, they certainly point to vivid and profound beliefs in witchcraft. So do many newspaper reports. Only a few days before the above incident President Kenyatta had ‘urged a crowd of 40,000 Kenyans … to give up witchcraft’. All over Africa there appear to be people who think they are witches, and who sometimes claim guilt for the crimes of witchcraft. And there appear to be far more people, even countless people, who believe in the power of wizardry, whether for good or evil; who are encouraged or dismayed by a host of signs and omens; and who lay out cash they can ill afford on spells and counterspells.

It is all very disconcerting. Only think what Dean Farrar would have made of it: surely here, if anywhere, is the daily documentation of those African features ‘darker and deadlier’ than baseness or brutality or a fatal shortage in the frontal lobes? A great many followers of the Dean have certainly said so and not a few others, one imagines, continue to think so. Even if they are wrong, the awkward question still remains: how can magical beliefs, and confessions of doing things such as eating dead baby, remain possible within a moral order which is derived from the empirical observation of real phenomena?

A sociological answer calls first of all for seeing these beliefs in a general perspective. They have been common to all prescientific societies. Every old civilisation contained them and more or less violently wrestled with them. A fairly early but characteristic European example occurred in AD 1080 when King Harald of Denmark was told by Pope Gregory VII that ‘he must no longer tolerate among his people the gruesome superstition according to which Christian priests or wicked women are held answerable for bad weather, storms, unfruitful years, or outbreaks of plague’.