Understanding photos, models, and beliefs: A test of the modularity thesis of theory of mind

Abstract

Recent experimental findings suggest that there is a dissociation between the performance by children with autism on False Belief tasks, on which they do poorly, and False Photograph, False Maps, and False Drawing tasks, on which they do well. This may be because only the False Belief task taps the capacity for metarepresentation, at least as defined by Leslie and Thaiss (1992). In an attempt to test the modularity of this capacity further, the performance of participants with autism on DeLoache's (1987, 1991) Model and Photograph tasks (which test understanding of the symbolic function of models and photographs), and on a standard False Belief task, was compared to that of mental handicap controls. The majority of participants from both groups passed the Photograph and Model tasks, and on neither task were there group differences. However, participants with mental handicap were significantly more successful on the False Belief task than those with autism. These results provide further support for the modularity of theory of mind, and the specificity of the metarepresentational deficit in autism.