Self-stimulatory behavior is repetitive, stereotyped, functionally autonomous behavior seen in both normal and developmentally disabled populations, yet no satisfactory theory of its development and major characteristics has previously been offered. We present here a detailed hypothesis of the acquisition and maintenance of self-stimulatory behavior, proposing that the behaviors are operant responses whose reinforcers are automatically produced interoceptive and exteroceptive perceptual consequences. The concept of perceptual stimuli and reinforcers, the durability of self-stimulatory behaviors, the sensory extinction effect, the inverse relationship between self-stimulatory and other behaviors, the blocking effect of self-stimulatory behavior on new learning, and response substitution effects are discussed in terms of the hypothesis. Support for the hypothesis from the areas of sensory reinforcement and sensory deprivation is also reviewed. Psychologists and ethologists, who have been studying these behaviors for some time, have referred to them as "abnormal stereotyped acts," "mannerisms," "motility disturbances," "ritualistic acts," "rhythmic habit patterns," "blindisms," or "autisms" (e.g., Baumeister & Forehand, 1973;Berkson, 1967;Mitchell & Etches, 1977 (Lovaas, Varni, Koegel, & Lorsch, 1977) despite our discouragement of it. Other unexpected and elaborate repertoires also developed during treatment, such as obsession with numbers, compulsive arrangement of letters, and frequent assembly and reassembly of the same jigsaw puzzle (Epstein, Taubman, & Lovaas, 1985). The children who made the largest gains in treatment were also the ones who emerged with the largest repertoire of such "personally contributed" behaviors.To date, no comprehensive theory of self-stimulatory behavior in the developmentally disabled has been proposed. Most theorizing in this area has consisted of unelaborated accounts of specific experimental results, often focusing on body-rock-46 SELF-STIMULATORY BEHAVIOR ing as the "prototypical" self-stimulatory behavior. As a result, Baumeister (1978) found it necessary to create a five-way classification scheme simply to impose some order on the many different accounts, noting that none explained very many findings. The purpose of this paper is to present in some detail the hypothesis that self-stimulatory behavior is operant behavior that is maintained automatically by the reinforcing perceptual stimuli that it produces. This hypothesis will be used to explain the acquisition of representative behaviors and the etiology of self-stimulatory behavior in developmentally disabled persons, as well as certain prominent characteristics of self-stimulatory behavior. Empirical support for this hypothesis is offered from relevant areas of research. The Discussion considers two representative alternative theories and describes implications of our hypothesis for the treatment of developmentally disabled persons and the understanding of apparently related behaviors in other populations. PERCEPTUAL REINFO...

This study attemped to isolate some of the stimulus variables that controlled the self-destructive behavior of a psychotic child. In Experiment 1, the child was exposed to several demand and nondemand situations. In Experiment 2, the situation containing demands was modified so that demands now occurred in the context of a positive, ongoing interaction between the child and the adult therapist. The rates of self-destructive behavior underwent several orderly changes: (1) Rates were high in demand situations and low in nondemand and modifieddemand situations; (2) rates decreased sharply when a stimulus correlated with the termination of demands was introduced; and (3) rates of self-destruction typically showed gradual increases within each of those sessions which contained only demands. These results were interpreted as suggesting that (1) self-destruction, under certain circumstances, may be conceptualized as an escape response which is negatively reinforced by the termination of a demand situation and (2) certain modifications of the social environment may provide discriminative stimuli for behaviors other than self-destruction, thereby decreasing this behavior.

This study sought to identify some of the variables controlling the severely aggressive behavior of two retarded children. In Experiment 1, each child was presented with several demand and nondemand situations. Aggression was frequent in the demand situations and rare in the nondemand situations. When a stimulus correlated with the termination of demands was introduced, aggression fell to a near zero level. In Experiment 2, for one child, a variety of preferred reinforcers was introduced into the demand situation contingent on correct responding. Aggression abruptly decreased to a low level. Experiments 3 and 4 involved the second child. In Experiment 3, this child was permitted, in one condition, to leave the demand situation if he emitted a nonaggressive response. Aggression decreased to a low level. In Experiment 4, he was prevented, in one condition, from leaving the demand situation in spite of high levels of aggression. Aggression fell to a near zero level. In Experiments 3 and 4, he was permitted, in several conditions, to leave the demand situation following aggressive behavior. Aggression increased to a high level. The results suggested that: (1) aggression can sometimes function as an escape response; and (2) escape-motivated aggression can be controlled by:(a) introducing strongly preferred reinforcers to attenuate the aversiveness of the demand situation; (b) strengthening an alternative, nonaggressive escape response; or (c) using an escape-extinction procedure.DESCRIPTORS: aggressive behavior, escape, reinforcement, extinction, retarded children No human behavior arouses more social concern than that of aggression. At the level of the individual child, high-frequency aggressive behavior often has the effect of disrupting normal family functioning as well as making the child unteachable in school. Because of these negative effects, many aggressive children are

A social skills group intervention was developed and evaluated for young children with autism. Twenty-five 4- to 6-year-old (diagnosed) children were assigned to one of two kinds of social skills groups: the direct teaching group or the play activities group. The direct teaching group used a video-modeling format to teach play and social skills over the course of the intervention, while the play activities group engaged in unstructured play during the sessions. Groups met for 5 weeks, three times per week, 1 h each time. Data were derived and coded from videotapes of pre- and post-treatment unstructured play sessions. Findings indicated that while members of both groups increased prosocial behaviors, the direct teaching group made more gains in social skills.

Thirteen autistic children were compared to 13 normal children matched to them in mental age, on performance of a visual discrimination task. Form, color, and size were relevant and redundant cues. The groups did not differ significantly in mean trials to reach criterion or in breadth of learning, and both groups increased their breadth of learning after 50 trials of overtraining. Form was preferred to color and size by both autistic and normal children. Within each group, rank on mental age was highly correlated with rank in breadth of learning. Verbal and nonverbal autistic children did not differ in breadth of learning or in dimensional preference. Even nonverbal autistics equaled the performance of their normal controls. Our results suggest that overselective attention is better understood as part of a general developmental lag in cognition in autistic children than as a specific deficit underlying psychotic behavior.