Research Group
in Analytic Philosophy

Abstract

Paradoxes can result when we mistake fallible heuristics for universal laws. The trick is to recognize in practice which is which. For many ordinary heuristics, such as the use of psychological availability to measure frequency, more painstaking but reliable alternatives can be applied to identify their errors. In other cases, we lack such an alternative, and deeper paradoxes arise. The challenge of distinguishing between laws and heuristics is obscured by talk of ‘intuition’. The difference is inaccessible to introspection, but open to theoretical investigation, by a broadly abductive methodology. For example, should we blame the Liar and other semantic paradoxes on heuristics for semantic terms or heuristics for logical constants? Again, some philosophers blame sorites paradoxes on classical logic, in order to preserve tolerance principles for vague expressions, but tolerance principles are a natural candidate for heuristics, highly reliable but fallible. The example to be discussed in most detail will be a suppositional heuristic for evaluating conditionals, both indicative and subjunctive. It fits our practice well, and predicts the role of conditional probability in assessing the probabilities of conditionals. However, it gives inconsistent results for inconsistent antecedents, and can also break down for the use of conditionals in testimony. Failure to identify our reliance on the suppositional heuristic has led to ill-motivated theories of ‘impossible worlds’ and the like.