THE STUDY OF RELIGION, WE ARE TOLD, REMAINS EVEN TODAY ‘AN enormous and unfilled field for research’. And it is understandably one that scientists have seldom cared to enter, at any rate since Bruno was burned at the stake and Galileo bullied into saying that the Earth was flat. Even when they have entered this field, they have generally done so with the suspicion of being sadly, even rather shamefully, out of place. To many anthropologists of the nineteenth century, Professor Evans-Pritchard has lately written, ‘religious belief was … absurd, and it is so to most anthropologists of yesterday and today’.

Yet it is now perfectly evident that scientific elucidation of all those systems of thought and belief which have passed in Africa under the catch-all label of ‘religion’—and I use this often misleading word only for want of a better—is vital to any genuine social analysis. We cannot begin to understand the drift and logic of African apprehension of reality unless, for example, we have grasped the reasons why appointed ancestors should have become ‘the jealous guardians of the highest moral values … the axiomatic values from which all ideal conduct (has been) deemed to flow’. In this large sense the study of religion is the study of social fact.

When the lords of the Karanga carved their empire from the lands between Zambesi and Limpopo long ago, and built their stone dwellings at Zimbabwe, they set up a shrine to hungwe, the fish-eagle, and erected soapstone effigies to the power they also called Shirichena, the Bird of Bright Plumage, or Shiri ya Mwari, the Bird of God. Attending this shrine the priest or medium of the most powerful of the appointed ancestors, Chaminuka the great mhondoro, was required to interpret the meaning of the cries of hungwe, and crucial decisions of state were influenced by what he said. For more than three centuries before 1830 it was to the spirit of Chaminuka, and its oracle the Bird of Bright Plumage, that the kings of the Karanga appear to have turned for guidance in their testing problems of state.