Abstract
The purpose of this study was to shape manding behavior in a severely retarded child who showed perseverative behavior, and to obtain generalization of the manding. Perseverative behavior has been regarded as an inappropriate behavior, and is generally targeted as a behavior to be suppressed. However, in this study, the perseverative behavior was used to promote appropriate verbal behavior. In a study by Hung (1978), self-stimulation was used as a reinforcer, contingent on verbal behavior. However, Hung (1978) reported some difficulty in obtaining generalization of the newly learned responses to daily living settings. One possible reason for this was that the subject discriminated the difference between the training settings and daily living settings. Therefore, in the present study, it was attempted to chain the verbal behavior to discriminative stimuli of daily life, in order to promote generalization. It was also hypotheized that if low-frequency responses were chained to high-frequency responses, the former would increase in frequency. Specifically, high-rate perseverative behavior and modified perseverative behavior was made contingent on mand behavior, which was initially low in frequency. Modified perseverative behavior was defined as perseverative behavior in which the trainer became able to intervene. The procedure for shaping the mand behavior was as follows: (a) Pretraining: to increase spontaneous pointing and vocalization by positive reinforcement, including both feedback of the same vocalization by the trainer contingent on the subject's spontaneous vocalization, and the procedure of Differential Outcome Effect (DOE). (b) First step: to try to chain the vocalization to the trainer's intervention, and then to introduce a temporary deprivation of reinforcement by changing the reinforcement from immediate to delayed. The reinforcers were handing pegs to the subject, and the trainer's approval. (c) Second step: delayed reinforcement and presentation of the appropriate verbal model, if the child did not emit appropriate mand behavior. The results showed that mand behavior could be developed by chaining it to perseverative behavior and modified perseverative behavior. In the present study, the attempt to associate manding with the discriminative stimuli of daily life is considered the main cause of the generalization observed. The mand behavior acquired two functions, that is, obtaining pegs and playing with them; this was considered to be another cause.
