Rational Belief and Reasonable Belief, A Ramseyian Distinction
James Joyce, University of Michigan, Philosophy
Friday, 13 October 2006, 3:30 pm  
817R Cathedral of Learning

Abstract: It often seems as if Bayesians are forced into one of two untenable positions. They can follow Bruno de Finetti and Leonard Savage and ‘go subjective’ by embracing a thoroughgoing personalism that judges any set of probabilistically coherent opinions to be as legitimate as any other, and which portrays any appearance of objectivity in beliefs as being the result of intersubjective agreement secured by a sufficient influx of evidence.  Alternatively, they can “go objective” with Harold Jeffries and E. T. Jaynes and hope to show, by a priori arguments based on the principle of indifference or the requirement to maximize entropy, that there are uniquely correct ‘informationless priors’ that can be employed in many evidential situations.  Both these alternatives have serious and well-known shortcomings (some of which I will briefly rehearse).  There is, fortunately, a third option.  Near the end of his famous essay “Truth and Probability” F. P. Ramsey drew a distinction between rational belief and reasonable belief.  Whereas a set of beliefs is rational provided only that it does not “sin against the laws of formal logic or formal probability,” the reasonableness of a person’s beliefs is tied to their overall accuracy and to the reliability of the mechanisms that produced them.  I will sketch some components of an account of reasonable belief that neither lapses into an objectionable personalism nor invokes any implausible a priori constraints on priors.  The view’s two key elements are (i) the idea that degrees of beliefs can be assessed for accuracy in much the same way that full beliefs can be, and (ii) the idea that there can be plausible rules for aligning one’s degrees of belief with one’s evidence that do not invoke the principle of indifference, entropy maximization, or any other suspect a priori principle.  The ultimate result of this stance will be to blur the boundary between ‘classical’ and ‘Bayesian’ approaches to statistical reasoning.

Can Evolutionary Biology Resolve the Debate Between Egoism and Altruism
Stephen Stich, Rutgers University, Philosophy
Friday, 10 November 2006, 3:30 pm 
G-8 Cathedral of Learning (ground floor)
Reception following in 817 Cathedral of Learning

Abstract: During the last 30 years there has been a great deal of discussion of ways in which evolutionary theory might contribute to the venerable philosophical debate between psychological egoists and psychological altruists. In this talk, I will begin by setting out some crucial ideas used in the philosophical debate and explaining why the debate is important for moral theory. I will then consider several arguments aimed at using evolutionary theory to show that altruism is unlikely or impossible, and argue that all of them fail. In the remainder of the talk, I will consider Elliott Sober and David Sloan Wilson's recent attempt to use evolutionary theory to argue that humans probably do exhibit psychological altruism. That argument, I maintain, also fails. The conclusion I'll offer is that if we are to make progress in the egoism vs. altruism debate, the science to look to is psychology, not evolutionary biology.