PROPOSALS FOR A PROGRAM OF EMPIRICAL RESEARCH INTO THE PROPERTIES OF TRIANGLES*

Physical geometry is viewed in this essay as the prototype of a successful generalizing empirical science and an acceptable framework or theory must be “empirical” in Popper's sense of being subject to refutation by observation.

Abstract

ABSTRACT Physical geometry is viewed in this essay as the prototype of a successful generalizing empirical science. Suppose that physical geometry had been studied in ancient Egypt by empirical methods similar to those commonly used today by many researchers seeking general propositions useful in explaining, predicting, or controlling phenomena of interest to practicing managers. Would these methods have produced valid and useful geometric generalizations? This intriguing question is explored in Part I by means of a parable which suggests a negative answer to the question. The reasons for this negative conclusion are developed in Part II. In the successful generalizing sciences, general propositions (or theory) have not been the result of an a posteriori “induction” from observational data, but rather the result of an a priori “construction” of a framework for the coordination of measurement operations. Such a framework, whether explicitly articulated or not, is a necessary condition for the collection of coherent and reproducible observational data and, therefore, for explanation, prediction, and control of phenomena. Yet, despite its a priori character, an acceptable framework or theory must be “empirical” in Popper's sense of being subject to refutation by observation. If the framework is logically prior to data , however, what is its source? In the successful generalizing sciences, it is argued in Part III, theories have been generalized not from data but from the skills—the coordination of operations —of practitioners. This genesis “explains” the success of the sciences, because an effective skill includes within itself tests of its range of applicability and adapts itself to changing conditions. A discussion of the implications of this argument for making management “scientific” concludes the essay.