Abstract
The article discusses morality from the perspective of compliance. The expectation that rules easily will be internalized and then followed makes very unrealistic assumptions about human behavior. There are good reasons to put more trust in prudence - amorality supported by incentives - than pure morality. There are reasons to be critical to morals that have no prospect to be developed into prudence. Such a preference for a prudent morality is also a criterion for selection among different suggestions of moral rules. A normative suggestion in the Prisoner' s Dilemma exemplifies the case of prudent morality. The player should play contribute if the expected value of that choice exceeds his payoff of mutua! defection. This suggestion is in line with Gauthier' s rules for constrained maximizers. Empirical evidence indicates strong preferences for symmetric outcomes: People tend to contribute rather than defect if the other player also is expected to contribute; otherwise mutual defection is the preferred outcome. Fruitful revisions of the game are more likely to be found in modifying the payoffs of playing defection than in seeing the players' priority of their personal payoff as the main obstacle for cooperation. The article questions the value of an absolute morality and suggests a morality that is more of a prudent policy than a categorical imperative. A conviction of a good average result in the long run is the most relevant argument for complying with a rule.