Rationality as optimal choice versus rationality as valid inference

While economics distinguishes itself from the other social sciences by its treatment of human action as rational, its notion of rationality is actually based on a profound skepticism about people's ability to reason. This notion goes back to Hume, according to whom 'Reason is the slave of the passions'. Hume's belief that people cannot determine their preferences by reasoning led to the notion of rationality based on representation: people are rational when they are able to represent their preferences vividly, accurately gauging their intensity. Starting with Kant, philosophy moved to an understanding of rationality as valid inference; cognitive psychologists similarly speak of the superiority of the 'reason-based' model of choice over the economic 'value-based' one. Economics stayed with Hume's conception because the latter enabled the mathematical expression of a relation between scarcities. If it is to produce empirically adequate theories however, economics must follow the examples of philosophy and psychology.