Doxastic voluntarism and self-deception

Direct Doxastic Voluntarism � the notion that we have direct (un-mediated) voluntary control over our beliefs � has widely been held to be false. There are, however, two ways to interpret the impossibility of our having doxastic control: as either a conceptual/ logical/metaphysical impossibility or as a psychological impossibility. In this paper I analyse the arguments for (Williams 1973; Scott- Kakures 1993; Adler 2002) and against (Bennett 1990; Radcliffe 1997) both types of claim and, in particular, evaluate the bearing that putative cases of self-deception have on the arguments in defence of voluntarism about belief. For it would seem that if it is the case that self-induced cases of self-deception are indeed possible, then voluntarism about belief could be true after all. Bennett claims that Williams� argument for the impossibility case proves too much in that if it is successful in ruling out direct doxastic voluntarism, it is also successful in ruling out cases of indirect doxastic voluntarism. If cases of self-deception can also be cases of indirect doxastic voluntarism, then such cases support the argument against the impossibility case. I argue that Bennett is right in claiming that Williams� argument proves too much, that cases of self-deception are indeed also sometimes cases of indirect self-deception and so that they cause genuine trouble for the conceptual impossibility case. However, I also argue that this is the only genuine worry for Williams� argument. I end, while considering whether cases of self-deception can tell us anything about the psychological possibility of direct doxastic control, by suggesting a way of establishing the conceptual impossibility of direct doxastic control that circumvents Bennett�s counter-argument