Symposium on 'Cognition and Rationality: Part Ⅱ'

摘要

In the introduction to part I of the symposium (Mind and Society, Vol. 5, 2006) we stated that a rational agent could be thought of as an agent who has good reasons for its actions. In formal analyses of economic, medical, political, military and forensic decisions (and ensuing behaviours) rationality, that is the "goodness" of those reasons, is inextricably intertwined with probability. Typically, those analyses concern decisions in a particular class of uncertain situations, namely "risky" situations, where all the relevant available alternative actions are known, each or most actions are non-deterministic (that is, they can have different outcomes), and the probability of each outcome is known to some degree of approximation. As yet, formal analyses have only begun to scratch the surface of the subtleties implicated in uncertain, but not risky, situations, where the available courses of actions, and/or their possible outcomes, and/or the probabilities of those outcomes, are not all known. As a consequence, many uncertain situations are modelled as risky situations, incurring an inescapable degree of simplification that has caused endless disputes between theorists committed to different views of probability. Those disputes are interesting from a philosophical, mathematical, and practical perspective, but their settlement is not in sight as yet.