action theory, action frame of reference These terms are not interchangeable but are closely related and carry a number of implications about the way we regard sociology as a science. It is usual, for example, to juxtapose action to structure as alternative starting-points for sociological investigation. Action theories are those which see the major or only object for sociology as human action. This group includes Weberian sociology, phenomenological and hermeneutic sociology, symbolic interactionism, ethnomethodology, and structuration theory (all of which are dealt with under separate headings in this dictionary). These approaches concern themselves, not only with the nature of action, but also with meaning and interpretation. A defining quality of action is that, unlike behaviour, it carries a subjective meaning for the actor. Max Weber distinguished four types of action—traditional (customary), affective (emotional), value-oriented, and instrumental (rational means-end)—although, historically, much of sociological analysis has concentrated on the last two of these. Action theories, therefore, do not see sociology as a science like the natural sciences, dealing with external, independent objects; rather, sociology is scientific because it gives a rational, coherent account of peoples' actions, thoughts, and relationships.

The action frame of reference is associated with the name of Talcott Parsons, whose theory starts with a systematic analysis of action which sees the social actor as choosing between different means and ends, in an environment which limits choice both physically and socially. The most important social limitations on choice are norms and values. From this, Parsons builds up an elaborate model of the social system, such that his theory arguably loses its voluntaristic character: that is, the notion of the choosing actor disappears, in favour of a theory of structural determination in which norms and values play the determining role.

Apart from Parsons's theory, modern action theories in sociology have three different concerns. The first is the nature of rationality and rational action itself. This focus arises out of Weber's typology and poses questions about the possibility of causal explanations of action. (Are the reasons for doing something a cause in the same way that heating a piece of metal causes it to expand?) It also addresses the issue of whether there are any absolute criteria of rationality, or wether sociological explanations are always in some sense relative. Jon Elster's rational choice theory takes up some of these problems in a more substantive way. The second concern is the taken-for-granted rules and stock of knowledge that underlie action—a theme pursued notably by ethnomethodology and phenomenology. The third, addressed by symbolic interactionism, is the learning and negotiation of meaning that goes on between actors.

All action theories have something to say, implicitly or explicitly, about the rationality of the actor—if only that he or she behaves rationally. See also EXCHANGE THEORY.