Robert Kurzban : University of Pennsylvania 

Evolutionary psychologists are committed to the view that form follows function. This commitment carries an epistemic corollary: if a mechanism with a proposed function does not have the form that is required to perform that function, confidence in the proposed function should be reduced. The view that moralistic punishment – imposing costs on those who violate a moral norm – functions to deter harm requires that moralistic punishment have a number of features to implement that function. First, moralistic punishment should be desired and deployed (only) in cases in which agents intend that harm come about; without intent, harmful acts can’t be deterred. Second, the magnitude of punishment desired should relate systematically to the benefits perpetrators receive, in line with standard decision theory. Third, and related, no punishment should be desired when no harm is intended and no harm comes about; there should be no victimless norm violations. Fourth, whether and how much punishment is desired should depend only on the costs and benefits of perpetrators’ actions/inactions; how violations occur is irrelevant to deterrence. All four of these predicted features of moralistic punishment have been empirically falsified. In this talk, I will briefly review this evidence and discuss the implications.