IntroductionWe are all conservatives, at least when it comes to belief retention. We are forgetful, of course, but we typically do not abandon our beliefs unless we have special reasons to do so. Our conservativeness might be viewed as a kind of deep self-trust, or less nobly, as simply part of the "inertial" character of belief (Harman 1986, Owens 2000. Whichever view one takes of the psychology, an epistemological question arises: if a subject S believes that p, does the retention of that belief thereby have some positive epistemic status for S, at least prima facie? An epistemic conservative is someone who answers this question in the affirmative. Conservatives may disagree about more specific matters, e.g., about which positive epistemic statuses (justification, blamelessness, rationality, etc.) satisfy a conservative principle or about how broadly to construe the range of defeating conditions associated with the prima facie rider. But the core of the doctrine is the epistemic efficacy of belief.In this paper, I will defend a form of conservatism about rational belief retention. I focus on rationality mainly for two reasons. First, this notion is regularly employed in ordinary epistemic evaluations of belief (arguably unlike justification). Second, rational belief retention does not require that one's belief be true, and so does not require knowledge, truth-tracking, safety, or other truth-entailing externalist statuses. Conservatism with respect to these statuses is a non-starter, at least under any reasonably narrow construal of the range of defeating conditions.
2My argument in favor of conservatism is not a variant of familiar ones based on considerations of cognitive efficiency (see Harman 1986 on "clutter avoidance" and also Lycan 1988) or based on the need to avoid skepticism in general (Chisholm 1989) or the need to cope with the evidential underdetermination of theories (Sklar 1975). Rather, I will argue that epistemic conservatism makes a good epistemology of memory. More particularly, I will argue, first, that the two standard accounts of the rationality of memory belief, preservationism and evidentialism, face insuperable problems, and second, that conservatism, properly honed, avoids those problems and incurs no unacceptable commitments. I make the case against the standard accounts in sections I and II, and defend conservatism against objections in section III. What will emerge is a moderate form of conservatism that is fundamentally an epistemological principle about memory.Before proceeding, certain terminological matters require brief discussion.Epistemologists often ask questions about a subject's rationality (justification, etc.) with respect to the statics of belief (e.g., is S rational to believe that p? Is S rational in believing/having the belief that p?), rather than the dynamics of belief (e.g., is S rational to retain/abandon the belief that p?) However, in ordinary life we regularly use the language of rationality when discussing belief dynamics, and it is clear that the sort of ra...