Friday Late Afternoon Session ( 15 : 30 -

The neural basis of social perception and understanding. Tjeerd Jellema, Utrecht University. Social scenes usually contain complex stimulus configurations, derived from the postures and actions of one or more agents. A description of the social scene can be made at two successive levels. At the first level, a literal, mechanistic, description is made in terms of causes and consequences of goal-directed actions. At the second level, a mentalistic description is made, in terms of the mental states and intentions of the agents. The latter involves associations with emotional values and memories, and possibly the recruitment of mirror-neuron systems, resulting in a meaningful interpretation. A number of distinct visual cues have to be taken into account to accomplish the mechanistic, or perceptual, description. These include: (1) Form: e.g. the perspective view at which the agents are seen. (2) Motion: the articulated, or whole body, actions performed by the agents. (3) Spatial location: the relative positions of the agents with respect to each other and objects. (4) Immediate perceptual history: actions usually consist of sequences of motions/postures. Cells in the anterior part of the superior temporal sulcus (STSa) of the macaque monkey are selectively responsive to one or more of these four visual cues, or to intricate interactions between them. Therefore, the STSa seems well equipped to represent a social scene at the perceptual level. Next, concerted activity of the STS with e.g. the amygdala, cingulate cortex and orbitofrontal cortex may form the basis of the second, interpretation, level. The findings from the single cell studies have directed our thinking about what the components and neural underpinnings of social cognition are, and have guided the design of behavioural experiments. In the latter experiments a perceptual distancejudgment task was used to measure quantitatively the extent to which social stimuli are processed automatically at the interpretation level by autistic and typical participants. The judgments of typical participants were biased by the implications of the social stimuli, while the autistic participants remained unaffected. This suggests that the semantic processing of social stimuli in autistic people is not automated.