Zande magic and the Dawkins delusion

This paper considers some logical errors in the analysis of belief systems, adopting the analytic methods of two Wittgensteinian thinkers, Peter Winch and Wes Sharrock. Examining the different analyses of Zande magic provided by E. E. Evans-Pritchard and Alasdair MacIntyre, we suggest that these methods can be characterised by their identification of ‘moments’, places where such analyses go catastrophically wrong. A Winch moment is the point in an account where something not required in the analysis is smuggled in to facilitate the making of unnecessary and unwarranted claims. A Sharrock moment is an incoherent or nonsensical premise or assumption made to get an account off the ground in the first place, without which little of the account remains. Some of Richard Dawkins’ accounts of religious belief are examined to show where both Winch and Sharrock moments can be found in his arguments.