The Evolution of Leadership Structure

The purpose of this study was to examine the applicability of reinforcement theory to the development of a leadership structure in a newly formed group. The results indicate that the ability to provide rewards is related to leadership as measured by sociometric, interaction, and influence scales. The more valuable rewards an individual could provide, the more closely was this ability related to measures of leadership. A further confirmation of the utility of a reinforcement model for explaining the emergence of a leadership structure was given by the suggestion that actual influence, attempted leadership, and rewards for initiating leadership increased over the course of a meeting. Additional application of the reinforcement model to an understanding of the evolution of leadership structure is discussed. The most common definition of leadership in the experimental literature and the one closest to ordinary usage describes leadership in terms of interpersonal influence-compliance relationships. In an operational sense, when one speaks of leadership, he means the number of people influenced by the leader, the frequency with which they are influenced, and the range of activities over which they are influenced.' In light of this definition, understanding the emergence of a leadership structure in a newly formed group involves at least an understanding of the factors which affect the frequencies of-and inter-relations between-the ascendant and submissive behaviors that occur during group interaction. The reinforcement theory of learning appears to be an economical way of describing changes in such frequencies. Reinforcement theory predicts that those behaviors which lead to rewards will increase: thus we should expect leadership and compliance behavior to increase as the interaction is mutually rewarding. Supporting this proposition * This research was sponsored in part by the Harvard Laboratory of Social Relations and National Science Foundation Grant G11309, Part of the work reported in this paper was done under the direction of Professor George C. Homans for an undergraduate honors thesis at Harvard College.