On the Measurement of Action

Abstract

Most social theories have actions among their objects. This may not seem obvious for pure macro-theories ¹ and for purely numerical mathematical models ² where actions can not be made out. Even in these cases, however, actions do play a central role. They form an underlying level from which a theory’s terms either are obtained by statistical procedures of aggregation or are obtained as numerical representations. In the case of pure macro-theories the determination of action does not appear as a major methodological difficulty because variation of action apparently is cancelled out by aggregation. In this paper, I have nothing to say about theories of the latter kind which does not mean that they are without problems. Also, I will not discuss the problems of ‘representationalist’ theories in which actions are represented by numerical terms. What I have to say here will however become pertinent to such cases once the representation or aggregation is no longer taken for granted, and itself becomes the object of theoretical investigation so that representation or aggregation itself explicitly occurs in the theoretical model. In the present paper, I concentrate on the main case which is most important methodologically and perhaps also by the number of its occurrences in science. In this case a social theory has actions among its objects. In the following I will refer to theories of this kind as primary social theories.