So working through the logic rigorously is important. Once you have checked out the logic of the the prisoners dilemma you can infer the implications of the theory and say that rational self interest implies mutual shafting in one-shot deals. Then you can observe that that is not what people do, so "rational self interest" is a false theory of human behaviour. You need "human psychology".

But you still want to distinguish between necessity and sufficiency. "A sense of fairness" is sufficient to explain the facts, but is it necessary? Rational self-interest leads to co-operation in iterated prisoners' dilemmas. Perhaps people aren't so much fair as limited: they always work in iterated mode, failing to take advantage of one-shot situations. What kind of evidence would favour one theory over the other?

[ Parent ]