Confirmation, Decision, and Evidential Probability

Abstract

Henry Kyburg’s theory of Evidential Probability offers a neglected tool for approaching problems in confirmation theory and decision theory. I use Evidential Probability to examine some persistent problems within these areas of the philosophy of science. Formal tools in general and probability theory in particular have great promise for conceptual analysis in confirmation theory and decision theory, but they face many challenges.

In each chapter, I apply Evidential Probability to a specific issue in confirmation theory or decision theory. In Chapter 1, I challenge the notion that Bayesian probability offers the best basis for a probabilistic theory of evidence. In Chapter 2, I criticise the conventional measures of quantities of evidence that use the degree of imprecision of imprecise probabilities. In Chapter 3, I develop an alternative to orthodox utility-maximizing decision theory using Kyburg’s system. In Chapter 4, I confront the orthodox notion that Nelson Goodman’s New Riddle of Induction makes purely formal theories of induction untenable. Finally, in Chapter 5, I defend probabilistic theories of inductive reasoning against John D. Norton’s recent collection of criticisms.

My aim is the development of fresh perspectives on classic problems and contemporary debates. I both defend and exemplify a formal approach to the philosophy of science. I argue that Evidential Probability has great potential for clarifying our concepts of evidence and rationality.