Evidence for Thinking in Pictures as a Cognitive Account of Autism

Background:

Many theories have been proposed that seek to explain autism at the cognitive level, such as Mindblindness, Weak Central Coherence, and Executive Dysfunction, although no consensus has yet been reached, and even the idea that a single cognitive explanation can be found has been called into question.  One view of cognition in autism that has not, however, been investigated in a comprehensive manner in the scientific literature is that of a bias towards "visual thinking" in autism, as exemplified by Temple Grandin's autobiographical book "Thinking in Pictures" and other introspective accounts.

Objectives:

To develop a well-defined hypothesis of cognition in autism centered around a reliance on pictorial representations versus verbal representations.  Also, to evaluate this hypothesis with respect to existing empirical data and to examine the relationships between this hypothesis and existing cognitive theories of autism.

Methods:

First, a set of minimal definitions was developed to distinguish between pictorial and verbal representations in terms of their computational properties and to formulate a "Thinking in Pictures" hypothesis of cognition in autism.  Then, existing empirical data from relevant published studies in the areas of behavior, cognition, and neurobiology were evaluated in light of this hypothesis, as were data typically cited in support of three existing cognitive theories--Mindblindness, Weak Central Coherence, and Executive Dysfunction.

Results:

Behavioral predictions made from the Thinking in Pictures hypothesis are consistent with many of the atypical behavioral characteristics of autism.  Empirical evidence that is relevant to the Thinking in Pictures hypothesis includes cognitive studies of individuals with autism showing a bias towards pictorial versus verbal representations while performing different activities, such as serial recall, task-switching, and semantic retrieval, although contrasting studies suggest that such a bias may not be uniform across the population.  Many individuals with autism have often demonstrated intact or even enhanced visual reasoning skills relative to the typically developing population, such as on the Embedded Figures Task and other visual search tasks as well as on visual tests of reasoning, such as Raven's Progressive Matrices.  Also, neuroimaging studies of individuals with autism performing certain tasks have shown greater brain activation in posterior sensory processing and imagery areas in comparison to the typically developing population, who show activation in verbal and other more frontal regions for the same tasks.

Conclusions:

This "Thinking in Pictures" account of cognition in autism shows significant potential for explaining many behaviors and cognitive characteristics in autism.  However, open questions remain as to whether only a specific subset of behaviors might be explained by this hypothesis and whether this hypothesis might apply to only a specific subset of individuals with autism.  With additional research, this hypothesis may lead to new and innovative methods of communication with and education of individuals on the autism spectrum.