Recognition of Context-Dependent Emotion In Autism

Background: Social cognition is a key feature of primate, and especially human, behavior which depends on the ability to evaluate the emotional and social meaning of complex stimuli as they occur in a given context. The process of emotion recognition can thus be thought of as consisting of several components, including initial perception of a stimulus and its integration with context. While complex emotion recognition is impaired in people with autism, it remains unknown which of these particular components are most affected.

Objectives: To address this issue, we compared the influence of context on the emotion recognition performance of adults with high-functioning autism. The core task consisted of photographs of people displaying strong, but ambiguous emotions, preceded by a context stimulus that served to disambiguate the target scene.

Methods: We have tested 15 autism participants and 16 controls. Participants were shown a series of photographs of people displaying strong emotions, in natural environments, ranging from laughter to anger and crying. These target stimuli ranged from very clear meanings to ones that were quite ambiguous in that they could be interpreted both positively and negatively (for instance a picture of people with wide eyes and open mouths looking upwards could be positively interpreted as pleasantly surprised people, who have just seen a mesmerizing fireworks show, or negatively interpreted as horrified people watching a drama unfold at an upper level of a building). We tested the influence of prior context information on the interpretation of these stimuli by presenting a short text (1 to 4 words) (for the text condition) or an emoticon (for the cartoon condition) immediately preceding the target stimulus.

Results: The pattern of responses shows a bias towards more negative interpretations of emotions for the autism participants compared to controls. Preliminary data indicates that the control participants were more likely to rate the photographs as positive emotions on the first presentation of each image, and more likely to be influenced in their rating by a positive context, whereas participants with autism were less likely to let themselves influenced by the suggestion of a positive interpretation for a given ambiguous photograph. This effect was particularly strong for the text condition.

Conclusions: Our results indicate a possible impairment in context integration and a tendency to ignore a positive interpretation in an ambiguous emotional stimulus.