There were two purposes underlying this study: to describe the sensorimotor functioning of mute autistic children and to relate their sensorimotor performance to nonverbal communication. Twelve mute children, diagnosed autistic, ranging from 4 years 9 months to 12 years of age, were administered four scales of sensorimotor development from the Uzgiris and Hunt (1975) series: object permanence, gestural imitation, means for obtaining environmental events, and causality. Subjects performed most poorly on the imitation scale with 9 of 12 performing below Piaget's fifth sensorimotor stage. In contrast, performance was highest on the object permanence scale: No child scored below Stage V. Regarding the subjects' non-scales and Stage III on the imitation scale appeared to form minimal prerequisites for intentional communication in a variety of situations. Finally, none of the subjects, even those with relatively complete sensorimotor development, spontaneously used what Bates (1976) has called "protodeclarative" gestures to point out or show objects to adults. The absence of protodeclarative gestures may represent a qualitatively distinct pattern of prelinguistic development in certain autistic children.

Several aspects of echolalic speech produced by five autistic children were investigated. We found that the incidence of echolalia was influenced by the type of question addressed to the child and, to a lesser extent, by the child's comprehension of the specific relationships expressed in the question. Additionally, acoustic analysis showed that a substantial proportion of echoes involved a prosodic modification of the examiner's question. Further analyses indicated that some of these modified echoes represent more than just a primitive conversational strategy. Specifically, they seem to reflect a higher level of processing and serve a semantic function, that of affirming the examiner's question.

The relationship between certain features of adult speech and autistic children's response adequacy was examined within the context of unstructured, dyadic conversations. On separate sessions, four verbal, nonecholalic children were observed talking with their mothers and teachers. Analysis of conversational turns showed that as the number of facilitating features contained in adults' eliciting utterances increased, the proportion of adequate replies from the children increased. In this analysis, facilitating features included the use of Yes/No questions, questions that were conceptually simple, and questions that were semantically contingent on the child's topic. In a further analysis, it was found that adults tended to modify their use of these features in response to child feedback, although this tendency was relatively small and observed only in a minority of the sequences evaluated. The findings are discussed in terms of pragmatic deficits associated with autism and implications for intervention with this population.

In this study the relationship of cognitive and affective variables to fear of strangers was examined in fearful and nonfearful groups of 9-to 10-month-old male infants. Cognitive variables included object permanence, person permanence, rate of visual habituation to familiarity, and visual attention to novelty. Affective dimensions included approach-withdrawal, positive-negative mood, and intensity of response. Fearful infants had more mature person permanence concepts, showed more overall visual attention, and scored lower in approach to new stimuli than did nonfearful infants. Object permanence, rate of habituation, attention to novelty, predominant mood state, and intensity of positive and negative responses did not differentiate between the fear and the nonfear groups.

Previous research suggests that verbal deficits among psychotic children may be paralleled by deficits in nonverbal pantomime. However, certain questions such as the level of pantomime exhibited by psychotic children, its susceptibility to modification, and its relation to other symbolic functions have not been systematically examined. To investigate these issues, 24 psychotic children were required to represent absent objects (e.g., toothbrush) via pantomime after receiving verbal instructions or instructions accompanied by a model demonstrating the pantomime. Also, measures of receptive and expressive speech, human figure drawings, and pretend play were obtained. The findings indicated very few complete failures in pantomime; higher pantomime performance when a model was provided although even in this condition most responses consisted of low-level substitutions of a body part in place of the absent object; and significant relationships between pantomime and measures of receptive vocabulary, echolalia, drawing, and play. The relationship of the findings to symbolic functioning in normal children and their relevance to understanding symbolic deficits in psychotic children are discussed.