We sometimes speak of a person's beliefs as being warranted, justified or reasonable. We say that one ought or ought not believe certain things, and we occasionally blame a person for believing what we imagine he ought not believe. These ways of talking make it appear that the forming, if not the holding, of beliefs resembles in at least some important respects the performing of actions. The latter view -which I shall call doxastic voluntarism, or plain voluntarism for short -has it that believing something can, at least sometimes, be under the voluntary control of the believer, as it were, in his hands. A person might be free with respect to what he believes in roughly the ways in which he can be free with respect to what he does. Indeed if believing (or rather coming to believe) is a species of acting, then believing is simply one of the things an agent, suitably equipped, can do voluntarily or non-voluntarily.
One may, of course, deny that anything one does is done voluntarily, or hold that all one does is caused but that this is altogether compatible with one's doing some things voluntarily. One may argue, as well, that some of the things one does are uncaused or that they are subject to some special form of causation. These are topics I prefer not to broach here. I wish rather to discuss whether believing or the forming of beliefs can plausibly be regarded as voluntary in whatever sense (if any) ordinary actions may be regarded as voluntary.
I have suggested already that our usual ways of talking about belief appear to incorporate an element of voluntarism. Should one wish philosophical support for this way of treating beliefs, one need look no further than Descartes' Fourth Meditation. There Descartes appears to assimilate believing to one's affirming (via the will) conceptions or ideas produced in the 'understanding'. The latter, in themselves, are not beliefs. They resemble propositions towards which an agent may take a range of epistemic stances. One of Philosophical Studies 43 (