Conservatism and Tacit Confirmation

Abstract

i. Methodological conservativism holds that believing a statement is a reason to continue to believe it. It also favours hypotheses that are introduced earlier than competitors. Other conservative theses have been defended, but these two already contain enough to motivate the main objection to conservatism. Since neither being believed nor the time of introduction are relevant to showing greater approximation to the truth, conservatism amounts to a prejudice, it is alleged, creating fundamental epistemic differences on non-epistemic grounds. Nevertheless, I want to defend conservatism. In part I, I argue that a statement's being believed or introduced earlier provides good reasons to expect it to have traditional epistemic advantages over some competitors, equally compatible with the data. Part II takes up the objection that because of underdetermination many hypotheses will elude the above defence. Against this objection I argue that if one accepts the underdetermination thesis as raising a serious methodological problem, then we should distinguish knowing that there are equally good competing hypotheses from actually producing them. Underdetermination does not itself guarantee the latter, and that is what counts. Standard objections to conservatism appeal to intuitions, and conceptions of justification, which are responsive solely to an interest in truth or correspondence with reality. The difficulty in defending conservatism is that it must concede its failure on these counts. In Part III it is argued that this failure does not fall on it differentially. Conservatism's claim to our allegiance shifts then to showing that, for all its defects, it remains the best policy we know.