ABSTRACT OF PAPER

Title: From the classical to the representational notion of measurement in utility theory: the debate of the 1930s
Author: MOSCATI Ivan


This paper belongs to a larger project that investigates how the notion of measurement evolved in utility theory during the 20th century, and more precisely how it shifted from the classical to the representational view of measurement. I trace this shift from a historical and methodological perspective, relating it to modifications in the mathematical notion of number, the popularity of Percy Bridgman’s operational philosophy of science, and attempts made by experimental psychologists to measure sensations. The canonical narrative regarding the evolution of utility measurement is all about the contrast between ordinal and cardinal views of utility. I argue that, in the debates about cardinal vs. ordinal utility, economists found their own way to move from the classical to the representational understanding of measurement, and that we can gain important insights into the history and methodology of utility theory by keeping in mind this measurement-theoretic dimension. According to the canonical narrative, in the period 1870-1900 economists were cardinalists. The ordinal view emerged around 1900 in the works of I. Fisher and V. Pareto, became dominant in the 1930s, and found its classical formulation in J.R. Hicks’ "Value and Capital" (1939). The publication of "Theory of Games and Economic Behaviour" by J. von Neumann and O. Morgenstern in 1944 offered new arguments to the cardinal view because their Expected Utility Theory features a utility function which is cardinally measurable. The present paper is devoted to the important debate about the measurability of utility beginning in 1934 with an article by O. Lange and ending in 1938 with a contribution by P. Samuelson. The paper shows that the canonical narrative is misleading because it erroneously projects the current notion of cardinal measurement into the past. In fact, until the mid-1930s, economists maintained a classical understanding of utility measurement as ratio assessment, and hence were not at all cardinalist in the current, ranking-of-utility-difference sense of the term. It was only in the course of that debate that the words “cardinal” and “ordinal” utility settled in economics, and “cardinally measurable” acquired the meaning it currently holds in economics. Moreover, in the contributions to the debate the first clear signs of a representational understanding of measurement appeared in economics. As far as I know, there is no detailed analysis of this important discussion of the 1930s, and my paper fills this lacuna.

Registred web users only can download this paper - Go back


Please note that files available for download have not been checked for viruses. These files have been submitted by authors of the conference to this web site. Conference organisers can't accept any responsibility for damages caused to users by downloading such files.