Why rationality is not a consequence of Hume's theory of choice

Abstract

This paper argues that the theory of action proposed by Hume in the Treatise does not imply that individuals are rational in the sense of modern choice theory. An individual's behaviour is non-rational if his/her choices systematically contravene the consistency axioms of the theory, and if the causal explanation of those choices cannot credibly be offered as a reason for making them. Hume proposes a theory of causal relationships between mental states, based on associations of ideas. The relationships he postulates are liable to induce various forms of non-rational behaviour, some of which have since been observed in controlled experiments.

Keywords:

• Hume
• choice theory
• rationality

Notes

For more on this formulation of choice theory, see Sen (1971).

This effect was first reported by Huber et al. (1982). Shafir et al. (2000) reviewed some of the accumulated evidence. Waite (2001) reported an asymmetric dominance effect for grey jays. The effect has also been found for bees and humming-birds.

For a review of current knowledge about temporal inconsistency, see Frederick et al. (2002).