Event Detail

Many people – including Issac Levi, Peter Walley, Teddy Seidenfeld, Richard Jeffrey and Mark Kaplan – have suggested that uncertain beliefs in light of evidence are best represented by sets of probability functions rather than individual probability functions. I will defend the use of such “imprecise credal states” in modeling beliefs against some recent objections, raised by Roger White and others, which concern the phenomenon of dilation. Dilation occurs when learning some definite fact forces a person’s beliefs about an event to shift from a fully determinate subjective probability to an imprecise spread of probabilities. A number of commentators have found aspects of dilation disturbing, both from an epistemic and decision-theoretic perspective, and have placed the blame on the idea that beliefs can be imprecise. I shall argue that these worries are based on an overly narrow conception of imprecise belief states which assumes that we know everything there is to know a person’s doxastic at titude toward an event once we know all possible values of her subjective probability for that event. A well-developed theory of imprecise beliefs has the resources to characterize a rich family of relationships among doxastic attitudes that are essential to a complete understanding of rational belief. It can only do so, however, if it is purged of an overly narrow conception of belief states. Once this change is made dilation does not seem so disturbing, and certain decision problems that often seem perplexing can be resolved in a satisfactory manner.

Prof. Joyce recommends the following papers as (optional) background reading:

(1) http://www.fitelson.org/few/few_08/white.pdf

(2) http://www.fitelson.org/joyce_hpre.pdf

Note: Jim will also be leading Branden Fitelson’s seminar on Thursday January 28 from 4-6pm in the Dennes Room. This talk will be entitled: “Inference and Decision Making With Imprecise Probabilities”. Please feel welcome to attend.