An experimental demonstration of discriminated (nongeneralized) imitation

Abstract

In a number of experiments, nonreinforced imitation has been found to persist at high rates despite the variety of procedures which have been employed to eliminate such responses. In the present experiment, powerful stimulus control was established over imitation by providing an alternative response which was reinforcible. Four children participated in a multiple-schedule experiment in which imitation was reinforced in the presence of one light and bar pressing in the presence of a second light; throughout the experiment, responses were modeled on each trial: hand-arm responses in the presence of the red light and leg responses in the presence of the yellow light. Initially, imitation was reinforced in the presence of the red light and button pressing in the presence of the yellow light. In a within-subjects design, stimulus control was demonstrated by reversing the association of lights and contingencies, then reinstating the original contingencies. The children imitated when reinforced for imitation and pressed the button when button pressing was reinforced. The results demonstrate stimulus control over imitation which is more powerful than in previous investigations and indicate that the prevailing reinforcement contingencies determine whether or not a child will imitate on a particular occasion.