Suffer A Witch

In 1937, anthropologist Edward Evan Evans-Pritchard (father of the less awkwardly named contemporary journalist Ambrose Evans-Pritchard) published his famous monograph Witchcraft, Oracles, and Magic Among the Azande, in which he claimed that the beliefs of the Azande (a tribe of north central Africa) concerning witchcraft were logically contradictory. Given the Azande’s beliefs about how witchcraft is inherited, together with their beliefs about which members of the tribe actually are witches, it logically follows, Evans-Pritchard assures us, that every member of the tribe is a witch. Yet the Azande do not draw this conclusion, even when it is pointed out to them: “Azande see the sense of the argument but they do not accept its conclusions, and it would involve the whole notion of witchcraft in contradiction were they to do so. … They saw the objection when I raised it but they were not incommoded by it.”

In the years since, philosophers and social scientists have debated ad nauseam how to interpret these facts (assuming they are facts – one would hope that further anthropological studies have been done to confirm or disconfirm Evans-Pritchard’s claims, but if so I haven’t heard about them). Are the Azande incompetent practitioners of logical inference? Or are they, perhaps, competent practitioners of some alternative logic, perhaps a three-valued logic? Or are their pronouncements about witchcraft best understood as something other than straightforward declarative statements about a language-independent reality, making the application of logic somehow irrelevant? (See Mark Risjord’s Woodcutters and Witchcraft: Rationality and Interpretive Change in the Social Sciences for discussion of the options.)

What I find puzzling about this debate is that it proceeds on the assumption that in (purportedly) manifesting this inconsistency in belief, the Azande are showing themselves to be importantly different from us in some way that requires special explanation. But in fact nothing is more common than for people to see the force of an argument and yet reject the conclusion, on the grounds that the conclusion is so contrary to their basic worldview that they assume there must be something wrong with the argument even if they can’t see what.

Consider how people – especially non-philosophers – react to philosophical paradoxes like the Liar, or the Sorites, or one of Zeno’s paradoxes of motion. Or consider how atheists react when confronted with the ontological argument, or how theists react when confronted with the argument from evil. Or, again, how statists react when confronted with the contradictions in statist morality (e.g., taking property without the owner’s permission is wrong, taxation involves taking property without the owner’s permission, yet taxation is not wrong), or as slaveholders formerly reacted when confronted with the analogous contradictions in slaveholding morality, or as Socrates’ interlocutors reacted when he exposed their inconsistent triads. In all these cases, there’s a tendency to assume that the argument is a sophism, that given its unacceptable conclusion it must have some flaw justifying its dismissal, even if this flaw can’t easily be identified or articulated. In Pericles’ words: “At your age we were clever hands at such quibbles ourselves.”

As these examples suggest, this reaction is sometimes justified and sometimes not. In the case of philosophical paradoxes, I agree with Moore’s argument that we are perfectly justified in rejecting the case for a crazy conclusion even when we can’t pinpoint where it goes wrong. We don’t, e.g., have to solve the Liar Paradox before we’re entitled to keep on using the concepts of truth and falsity. (Though I do think lack of curiosity about what’s wrong with the argument is an intellectual vice.) But there are other cases, like the slavery and statism ones, where the reaction was not justified. And that raises the question of how to distinguish propositions that really are fundamental data of common sense from those that merely strike us as fundamental data of common sense.

That’s a thorny philosophical question which I don’t intend to tackle in this post. My present target is smaller game: I merely wish to suggest that if the Azande did in fact a) hold the beliefs Evans-Pritchard describes, b) understand his argument, and c) reject his conclusion, we needn’t ascribe to them anything bizarre or unusual to explain this. Why not instead assume that they, just like us, tend (whether justifiably or unjustifiably) to dismiss, as probably flawed in some yet-to-be-identified way, arguments for conclusions that run against their basic worldview.

As Mises wrote:

Explorers and missionaries report that in Africa and Polynesia primitive man stops short at his earliest perception of things and never reasons if he can in any way avoid it. European and American educators sometimes report the same of their students. With regard to the Mossi on the Niger Levy-Bruhl quotes a missionary’s observation: “Conversation with them turns only upon women, food, and (in the rainy season) the crops.” What other subjects did many contemporaries and neighbors of Newton, Kant, and Levy-Bruhl prefer? … No facts provided by ethnology or history contradict the assertion that the logical structure of mind is uniform with all men of all races, ages, and countries.