What Storytelling Reveals about Affective and Causal Knowledge

Molly Losh, David Sobel, Graduate Reseachers and

Professor Lisa Capps, Department of Psychology, UC Berkeley

    Recent investigations using the wordless picture book Frog, Where are You? (Mayer, 1969) have highlighted autistic individuals' difficulties with the evaluative dimensions of narrative. This study extends existing research by analyzing the structure and use of evaluation in the storytelling of children and adolescents with autism.

    Fifteen Child Study Center children were compared with 15 non-retarded individuals with autism. Narratives of personal experience were elicited during a brief conversation in which children were asked to tell stories about various events in their lives. Story topics included, best friends, special occasions (e.g., birthdays and holidays), as well as topics of special interest to the child (e.g., computers). The children's stories were audio- and videotaped and then transcribed and coded for narrative organization and type and frequency of evaluative remarks. Evaluative remarks included comments and inferences of mental and emotional states, causal expressions, using phrases and exclamations to capture addressee attention, character voice and sound effects, etc.

    Overall, individuals with autism provided less coherent narratives, often including bizarre responses to the experimenter's questions and failing to provide adequate background information when introducing new information. In addition, the autistic group required significantly more prompts throughout their narratives than comparison children. Although no differences emerged in the total frequency of evaluative remarks, our results revealed dramatic differences between groups with respect to the type and range of evaluative remarks used. Specifically, the normally developing comparison (Child Study Center) children relied on evaluation in general and causal expressions in particular as a basis for establishing hierarchical organization in their narratives. Moreover, causal statements were frequently used to enrich narratives by reasoning about their own and others' emotions and mental states. In contrast, children with autism used causal expressions for more local functions, most often explaining actions and behaviors which were not necessarily central to the narrative and rarely if ever providing causal explanations for emotions and mental states. Furthermore, relative to comparison children, those with autism used a more restricted range of remarks and relied on less sophisticated evaluative strategies.

    These results confirm findings using the frog story narratives, indicating impairments in narrative organization and the linguistic encoding of evaluation. Although differences in the total frequency of evaluative remarks were not found, differences detected in the functions of evaluation between groups hold significant implications for linguistic and communicative development in autism. Because evaluation represents an integration of linguistic and social-cognitive aspects of narrative, these results may reflect an interaction of deficits across these domains. As such, these results illuminate the nature of linguistic and social-cognitive impairments among individuals with autism.