on the other hand, holds that some conditioning occurs prior to aware ness and that awareness is both a result of past improvement and a condition of further improvement in conditioning performance. Im portant in generalizing about the role of awareness in conditioning is the fact that the present study was uniquely different from other studies of awareness in several ways. The reinforced response class was a physical one rather than a verbal one and involved the behavior of two individuals instead of one, candy instead of a verbal stimulus was used as the reinforcer, the experimental nature of the study was not apparent to the subjects, who were children rather than adults, and most importantly, the experimental design was unique among studies of awareness in providing a sensitive measure of the temporal relation ship of awareness to conditioning rate. The cooperative response was defined according to a four-dimensional scheme involving the sequential vs. non-sequential, latency, duration, and topography dimensions of the vii viii response. Three classes of verbal behavior were used to objectively define the requirements of awareness of the response-reinforcement contingency. The apparatus, the experimental procedure, and the method of recording behavioral observations is described. Ten pairs of children six to eight years of age served as subjects. On the basis of verbalizations during the experiment, five pairs were classified as aware and five as unaware. Only the aware subjects conditioned thus supporting the hypothesis that awareness is necessary for conditioning to occur. An increase in conditioning rate was observed just prior to awareness, and an even greater increase occurred follow ing awareness. These data support the Postman and Sassenrath position that awareness is both the result of past improvement and a condition of further improvement. Since all unaware subjects received some reinforcement and yet failed to become aware, the idea that the effects of reinforcement are "automatic," in the sense that reinforcement is a sufficient condition to bring about awareness, was not supported.The data suggest that reinforcement is a necessary condition for awareness to occur, and that other variables influencing awareness are partially correct hypotheses, total response rate, attention to the task, mutual compliance to test hypotheses, and the nature of the conditioning task. Each of these variables probably interacts with, or in some way is influenced by, reinforcement, instructions, and his torical variables. Manipulatable conditions which would enhance the liklihood of awareness, and hence conditioning, are: a response with a longer latency and a simpler topography; a restricted operant which would slow the total response rate and provide time between responses to observe the consequences of each response emitted; instructions which induce a problem-solving set. As a cognitive process which may be an essential mediating event in the conditioning of many human behaviors, a detailed knowledge of the proces...