Unimpaired perceptual causality in high-functioning children with autism

Background:

Observers from six months are sensitive to physical and social causality in launching (Michotte 1946/63) and reaction (Kanizsa & Vicario 1968), schematic events involving movements of simple geometrical shapes. Early in development perceptual causality might support learning about mechanical interactions of material bodies and about the social interactions of intentional agents (Leslie 1988; Schlottmann & Surian, 1999).
A deficit/delay in perceptual causality fits with theories focusing on either the social or perceptual peculiarities characterising autism. Ray and Schlottmann (2007) reported a link to the latter, but not the former in low-functioning young children with autism.

Objectives:

Our study investigated perceptual causality for launch, reaction and related events in 20 high-functioning children with autism (mean CA=13, VMA=9.7) and 22 typically developing controls, to consider whether deficits in launching that appear for younger children with autism are overcome with higher age/verbal ability.

Methods:

We employed a picture choice methodology like Ray and Schlottmann (2007), but with more articulate verbal instructions. Children watched 14 animations (designed to test possible explanations of any launch deficit), choosing one of three pictures depicting physical or social causality, or non-causality for each.

Results:

Children with autism performed similar to controls, with no deficit on any event.

Conclusions:

Early problems with launch perception are overcome with age/higher verbal IQ. This suggests a perceptual causality delay rather than deficit in autism. Although this might still interfere with early causal learning, that it is overcome agrees with the general sparing of physical reasoning in autism. Deficits on social animations (Bowler & Thommen, 2000; Klin 2000) may only appear for complex stimuli requiring mental state rather than goal attributions. Thus, unimpaired perception of reaction in autism coexists with deficits in complex social attributions, suggesting a discontinuity between the two.