Detaching in Peer Disagreement and Higher-Order Evidence

Abstract

In this three-paper project I discuss the detaching problem for principles about the rational response to higher-order evidence. In the first paper, I argue that the problem presents a more serious problem than previously thought to conciliatory principles. In the second paper I argue that the problem applies to a wide variety of potential principles regarding the rational response to higher-order evidence. In the third paper, I argue that the problem suggests that the rational response to higher-order evidence is no response at all, i.e., that it is rational to dismiss higher-order evidence. Paper 1. Should conciliating with disagreeing peers be considered a sufficient condition for reaching a rational belief? Thomas Kelly argues that when taken this way, Conciliationism allows for all too easy acquisition of rational beliefs. Two responses defending Conciliationism have been offered. One response views conciliation as sufficient for holding a rational belief, but only requires it of agents who enter into a disagreement with a rational belief. This response makes for a requirement that no one should follow. If the need to conciliate only applies to already rational agents, then an agent must conciliate only when her peer is the one irrational. The other response views conciliation as merely necessary for holding a rational belief. This response does not answer the central question of what is rational to believe when facing a disagreeing peer. Attempts to develop this response either collapse into the first response or give rise to frequent rational dilemmas. Paper 2. Certain attitudes come with strings attached. For instance, believing that chocolate is harmful to dogs commits us to not feeding our dogs chocolate. Focusing on doxastic attitudes, epistemologists habitually offer principles that aim to capture what beliefs we are committed to adopting or avoiding given other beliefs that we have. A known worry with some of these principles is that they seem to open the door to troublesome rational dilemmas. If believing ~P commits us to disbelieving P, and we believe ~P despite evidence to the contrary, then our commitment would have us disbelieve something that our evidence requires we believe. Attempts to avoid the problem share the result that only a rationally held belief commits its holder to adopting or avoiding other beliefs. This result has a surprising implication for possible principles regarding the rational response to higher-order evidence. If higher-order evidence requires belief revision only of agents who responded rationally to their original evidence, then higher-order evidence requires belief revision only when it is misleading. It would be perfectly rational of agents who notice this never to adhere to those principles, knowing that they do nothing but mislead. As a result, a variety of plausible-sounding principles regarding the rational response to higher-order evidence are incorrect. Paper 3. Suppose we learn that we have a poor track record in forming beliefs rationally, or that a brilliant colleague thinks that we believe P irrationally. Does gaining such information require us to revise those beliefs whose rationality is questioned? When we gain information suggesting that our beliefs are irrational, one of two general cases obtains. In the first case we made no mistake and our beliefs are rational. In that case the information to the contrary is misleading. In the second case we indeed believe irrationally, and our original evidence already requires us to revise our belief. In that case, information to that effect is superfluous. Thus, information to the effect that our beliefs are irrational is either misleading or superfluous, and cannot justify belief revision.