This paper explores the influence of operationalism and its corollary, descriptivism, on Paul Samuelson's revealed preference theory as it developed between 1937 and 1948. Samuelson urged the disencumbering of metaphysics from economic theory. As an illustration, he showed how utility could be operationally redefined as revealed preference, and, furthermore, how from hypotheses such as maximizing behavior, operationally meaningful theorems could be deduced, thereby satisfying his demand for a scientific, empirical approach toward consumer behavior theory. In this paper I discuss the ensuing debate during the 1950s and 1960s on Samuelson's operationalism that raised doubts about its efficacy. In addition, I argue that certain concepts (revealed preference, equilibrium) and theorems (e.g., weak and strong axioms) that are supposedly operational in revealed preference theory, lack operational meaning, not withstanding their mathematical implications. Finally, I suggest that, although Samuelson's methodological rhetoric did not correspond with his implicit aprioristic theorizing, he possibly thought that his methodology and theorizing would converge in the long run.
