Choice, belief and the role of evidence

The papers are considered Draft Only and are not to be cited without the permission of the author. Centre for Applied Philosophy and Public Ethics Working Paper number 2001/5.Choices can be rational. So can beliefs. But what is the relationship between rational choice and rational belief? I will argue that rational choice imposes an evidentiary constraint on the beliefs that inform decision-making, and that Bayesian decision theory violates this constraint. We need a modified decision theory.Our actions are more likely to be successful if the beliefs they are based on are true. A rational agent should want to reason in a way that leads to true belief. Talk of 'true belief', however, should not be taken literally here. Although we should not limit our talk of belief to degrees of belief, the processes of rational choice only require degrees of belief. But what constraints of choosing rationally impose on how we form our degrees of belief? I will argue against orthodox Bayesianism that evidence about probabilities bears directly on the rationality of degrees of belief, and that rational decision-makers should only form degrees of belief where there is positive evidence about probabilities. Bayesianism is flawed because it entails degrees of belief in situations devoid of evidence. Decision theory can be justified as a regulative ideal, but Bayesianism fails on this interpretation as wel