THE EMERGENCE OF AN ACCURATE VERBAL DESCRIPTION AND EFFECTIVE NONVERBAL PERFORMANCE (RULE-GOVERNED BEHAVIOR, MICROCOMPUTERS, SHAPINE)

Human performance of a simple operant, such as pressing a button for token reinforcement, is often not reliably consistent with the operant behavior of other species, raising questions about the generality of reinforcement principles. Verbal behavior, in the form of experimenter instructions or self-generated rules, has been suggested as a critical variable affecting human operant performance. The interaction of verbal and nonverbal repertoires was examined in the present work.; College students responded on a microcomputer keyboard in a game scenario. They were told that their job was to operate a packaging assembly line in a paint factory and that periodically they would be asked to provide operation descriptions to new employees. Subjects were assigned to one of three conditions for establishing verbal descriptions: shaping, instructions, or nondifferential reinforcement.; In Phase I, the assembly line allowed subjects to package paint according to a mult DRL DRH schedule of reinforcement. Subjects' descriptions received different amounts of "bonus pay" according to their treatment condition. In Phase II, two cycles of the mult DRL DRH were followed by a mult RI RI schedule for the remaining eight cycles of the phase. In Phase III, half of the subjects in each condition were exposed to a reversal of their original mult DRL DRH contingencies from Phase I and all descriptions received differential reinforcement. The remaining subjects in each condition were exposed to the mult RI RI and provided with inaccurate instructions to describe their performance as if the mult DRL DRH were in effect.; Two significant findings emerged. First, differential responding on the mult DRL DRH often developed in the absence of accurate descriptions for the two uninstructed groups, suggesting that accurate descriptions were not a prerequisite for effective performance. Second, when presented with misleading instructions, subjects who had been instructed initially showed greater sensitivity to the actual contingencies while subjects who had been differentially reinforced showed greater correspondence between descriptions and performance. The results were discussed in terms of the role of verbal behavior in determining the acquisition and persistence of related nonverbal behavior with a focus on the issue of establishing correspondence between the two repertoires.