T HIS paper is intended to serve a dual function: first, to point up the need for a different theoretical approach to the specific of predicting or explaining the level of worker productivity; and second, to view this specific as an example of a general difficulty in sociological theory-building and to offer some suggestions for dealing with this difficulty. As the title indicates, this general is the need for systemic theories and organic models in sociology. The problem of worker like many behavioral science problems, involves an attempt to explain a variation in human behavior through the analysis of antecedent and/or situational factors. The names of the dependent variables in studies in this area vary considerably. Sometimes the dependent variable is called productivity, sometimes effectiveness, sometimes efficiency, and so on. In all cases, however, it refers to the ultimate behavior of the worker or the work group on the job. The word productivity is used here simply for the sake of consistency, although no one term is wholly adequate to encompass all of the variables studied.' When we turn to the independent variables, we find that they are legion. Motivation, morale, communication, control, allocation of authority, communication channels, attitudes (toward almost anything), rewards, needs, group position, forms of leadership, division of labor, and so on. No completely adequate classification of these many independent variables has been devised, and there is little hope of devising one for the present analysis. To simplify this discussion, however, a four-part classification system has been used.2 It is offered simply as a means of limiting the number of elements to be considered.