Inference and the Logical "Ought"

Abstract

This paper is the beginning of an explication of the normative-descriptive or ought-is distinction by way of the notion that our knowledge of other minds is the result of our imposition of constraints on the interpretation of events as actions by agents. My hope is that a general theory of rationality and the normative can be derived from an examination of the constraints it is rational to impose on agent-interpretation, i.e., of the fundamental knowledge we have of persons as persons. My attempt at an explication of the ought-is distinction takes the following form: I want to find an absolutely general way of determining when ought-sentences are true. Since the extensions of the account given below to interesting cases of ought-sentences such as moral and prudential cases depend on relatively complicated constraints on agent-interpretation,l this paper will deal only with the simplest case of ought-sentences, the ought. logic is thought of as a normative science of belief, it yields one of the simplest cases of the normative-descriptive dichotomy. By the I understand what might be called consequences of the canons of obedience to the laws of thought. An instance of such an occurs in If you believe that frogs are green, you ought to believe that anything that's not green is not a frog. The is, as it were, the minimal rational ought, the one that prescribes closure of belief under logical consequence and proscribes inconsistency of belief. It should be pointed out that the principles of the often come into conflict with other canons of rationality, just as principles of moral oughts come into conflict with each other. The example above is surely true even if