Behind Decision and Games Theory: Acting with a Co-Agent versus Acting Along with Nature

Abstract

The work to be outlined here has arisen out of my attempts, over the last decade, to say something definite, even if only schematic, about the connexion between belief and action. Most philosophers agree that the relation between having a belief and, in appropriate circumstances, acting on the belief cannot be an accidental or even merely a causal connexion. The belief in its context supplies a good and sometimes an adequate reason for the action; and this is a logical and not an empirical fact. But the relation of belief to action is complicated by the fact that it cannot be considered independently of the believer’s system of preferences: a belief that a river is deep is a good reason for a man who cannot swim and who is intent on suicide to jump into it, and a good reason for him not to jump into it if he has no such intent. What is needed for the analysis of the relations between a man’s beliefs, his preferences and his actions is to find some way of distinguishing the part played by his beliefs from the part played by his preferences in guiding his choices of action. If it were possible to make this discrimination, it would then be profitable to consider what about a man’s beliefs and preferences can be deduced from his actions; and this, for me, would be to complete the pragmatists’ programme of determining what a man is (in the way of believing and preferring) by finding out what he does.