Epistemological problems and the personality and social system in social psychology

If we understand social psychology to be an area where sociology and psychology overlap, or more precisely where we try to explain interaction on the basis of psychological and sociological propositions and concepts, we have singled out a field that should be quite challenging not only in theory and method but in the fundamental questions it raises for both sociology and psychology. Actually, the discipline is not that well integrated and is constituted by such disparate approaches as reinforcement theory, field theory, role theory, small group theory, game theory and psychoanalysis. Many sociologists have abandoned the field altogether. Nor have the proponents of these sub-fields made much effort to consolidate, integrate or reconcile their methodologies. Epistemological questions have been notably absent and only now have arguments from the philosophy of science point of view reemerged to revive the critical and potentially fruitful methodological discussions of earlier theorists (F. Allport; Lewin; Mead; Simmel) and their more recent followers (Homans; Malewski). After considering epistemological problems dealing with the generality of theory and explanation, behavioristic vs. action approaches, operational and model structural implications, we want to argue for a better understanding of social system variables besides those of the personality system and of system theory in general. In our discussion we use examples from the area of sport because it composes a complex system, that is not too difficult to observe at the same time that it shows in relative clarity all of the different levels of an action system. It has, furthermore, many features of an almost experimental design in a natural field. In so far it is a model area to allow due consideration for our demand that social psychology rediscover the method of field studies. This will help to reverse the trend characterized by a general neglect of theory that has resulted from behavioral dogmatism and the expedience of research pragmatism based mostly on two-variable linear models. This is not to say that we disfavor rigorous research design and data analysis - to the contrary. We just want it to be done in the context of broader theoretical concerns and in clear recognition of the pitfalls of operationalism and the merits of action theory.