Evidentialism and the Numbers Game1


  • 1

    This paper has benefited significantly from a number of comments. I owe particular thanks to Bill Brewer, John Broome, Roger Crisp, Sven Nyholm. Jeff Speaks, Wlodek Rabinowicz, and two anonymous referees for Theoriu. I have also benefited from comments given at presentations of this work in a slightly different form at the Oxford Moral Philosophy Seminar and in much the present form at the Canadian Society for Epistemology's International Symposium on Rationality in Contemporary Epistemology.


Abstract: This paper poses a puzzle concerning a broadly held view about normative reasons for belief: evidentialism. Evidentialism is the highly intuitive view that the only normative reasons for belief are evidential reasons. I shall argue that in certain circumstances, evidentialism is unable to generate the correct reasons for belief; these reasons can only be provided by other kinds of epistemic reasons apart from evidential ones. I am not arguing that reasons in ordinary cases for belief are non-evidential, but that evidentialism is too narrow an account of normative reasons for belief to serve as a complete theory of epistemic reasons.