Walker's empiricist interpretation of Walras’s tatonnement[Le "biais empiriste" dans l'interprétation de Walker du tâtonnement walrassien]

This paper is devoted to Walker's interpretation of Walras's tatonnement models. This interpretation is based upon a realist epistemology that Walker, in our opinion, assumed in Walras's thought. According to Walker, each statement of Walras has an empirical content, and can be compared to reality for its validation. However, this realist reading contradicts Walras's own rationalist method. This has three major consequences. First, the "empiricist bias" leads Walker to find two models of tatonnement where only one is usually considered. Second, Walker sometimes confuses what Walras called human facts with natural facts. Third, Walker puts forward an alternative view of Walras's edifice based on the articulation of institutions, procedures, rules and participants' behaviour. Usually, one considers Walras's Eléments d'économie politique pure as a rationalist edifice, where concepts are progressively introduced, step by step, from the simple to the more complex. Walker does not respect this architectonic unity.