In this post I will claim that Williamson’s analysis of epistemic conservatism is based on a mistake. Williamson’s mistake in chapter 7, section 6 of Philosophy of Philosophy (POP) involves including the belief that p among one’s reasons for believing that p. To flesh this out I first need to put a few things in place. Consider the principle of epistemic conservatism as formulated by Kevin McCain (2008: 189):
(PEC): If S believes that p and p is not incoherent, then S is justified in retaining the belief that p and S remains justified in believing that p so long as p is not defeated for S.
PEC captures the notion that one has a defeasible right to one’s beliefs. One loses one’s right to one’s beliefs given two conditions of defeat:
(DC1): If S has better reasons for believing that ∼p than S’s reasons for believing that p, then S is no longer justified in believing that p.
(DC2): If S has reasons for believing that ∼p which are as good as S’s reasons for believing that p and the belief that ∼p coheres equally as well or better than the belief that p does with S’s other beliefs, then S is no longer justified in believing that p.
Given PEC, the justification for believing p is analogous to the justification that S’s lacking a defeater provides. Lacking a defeater provides some justification, but it does not count as part of S’s reasons for believing. As McCain mentions, “S’s justification for believing that p is bolstered by her believing that p, but her belief that p does not count among her reasons for believing that p” (2008: 187). In short, belief that p cannot be used as a reason for believing that p. In a situation where S has another belief (or inclination to believe) that is inconsistent with p, S cannot use her believing that p as a reason to continue believing p. DC1 indicates that reasons for believing that ~p can act as defeaters and eliminate S’s justification for believing that p. DC2 indicates that if reasons for believing that ~p rival reasons for believing that p, and the belief that ~p coheres better with S’s other beliefs, then S has lost her justification for believing that p.
Williamson mentions that if intuitions are beliefs then they fall under epistemic conservatism. Do inclinations to believe also give one a defeasible right to one’s beliefs? What does epistemic conservatism council one to do when one has an inclination to believe something that is inconsistent with a belief one is currently committed to? Williamson uses a Gettier scenario to show that one cannot use an inclination to believe to arrive at a new belief. One can be inclined to believe something without believing it, and inclinations can conflict. When an inclination to believe something conflicts with a currently held belief, then, given epistemic conservatism, the currently held belief can be retained. Williamson (2007) arrives at this conclusion by claiming:
If I currently believe p, I am currently committed to the belief that any inclination to believe something inconsistent with p is an inclination to believe something false. I am not committed to the beliefs I am merely inclined to have in the way I am committed to my current beliefs (p. 243).
Given the PEC/DC1/DC2 package, an inclination to believe something inconsistent with a currently held belief (p), namely ~p, can serve as a reason to believe that ~p. That one believes that p cannot be used as a positive reason for retaining the belief that p in the face of reasons against that belief. Conservatism does not commit one to dogmatism. Simply because p is a belief (or because it is believed), and the reason to believe that ~p is arrived at via an inclination to believe, does not warrant retaining the belief that p. The inclination counts as a reason to favor ~p, so it is a potential defeater that must be overcome by reasons in favor of retaining p. Williamson does not offer any. What Williamson argues is that because an inclination is not fully believed it is not enough to overcome a belief that is actually believed or firmly believed. Williamson uses that fact that p is believed as a reason to retain p, which is a violation of PEC.
By contrast, what Williamson needs to argue is that an inclination to believe that ~p is not a reason that trumps the reasons in favor of believing that p. However, as the case is currently constructed, Williamson is not able to do this. In the Gettier case Williamson describes the reason Justin has for believing that knowledge is equal to justified true belief is that “Justin has been brought up to believe” that JTB theory is true. Is familial inculcation a reason for believing p that trumps the intuition that when presented with a Gettier case Justin judges that the Gettier subject has a JTB without knowledge? If anything, familial inculcation is often cited as a source of bias, blind belief and wishful thinking in the face of contrary evidence. This suggests that the inclination to believe that ~p, which is formed when presented with the Gettier case, is stronger than the reasons Williamson presents for favoring the belief that p. Thus, the intuition can serve as a defeater in this case, and epistemic conservatism councils Justin to abandon his inculcated belief and move to the new belief that the subject in the Gettier case has a JTB without knowledge or that JTB theory is false.