Abstract
Leading models of face perception posit that, after initial structural encoding, separate streams are responsible for the analysis of identity and expression. Interestingly, many individuals with developmental prosopagnosia (DP), a condition characterized by difficulties recognizing faces, are apparently unimpaired on expression recognition tasks. This pattern of performance is suggestive of a specific deficit in the identity processing stream, rather than a deficit of structural encoding. However, some authors have suggested that this apparent dissociation may reflect the use of expression recognition tasks that lack the necessary sensitivity to detect subtle impairments. The present study sought to investigate emotion recognition in DP using a task that systematically varies judgment difficulty. Sixteen adults with DP and 16 typically developing (TD) controls completed a computer-based emotion recognition task. Stimuli consisted of cropped eye-regions taken from happy, fearful, disgusted, angry, surprised and sad faces. Each source face was morphed with an image of the same actor exhibiting no emotion to obtain three sets of stimuli with varying levels of emotion intensity (100%, 66% and 33%). Results showed that DP and TD observers did not differ in emotion recognition ability at the 100% intensity level. However, as the emotion signal became weaker (in the 66% and 33% conditions), members of the DP group were increasingly likely to exhibit evidence of impairment, relative to members of the control group. Residual perceptual sensitivity, augmented by compensatory strategies, may be sufficient to judge unambiguous expressions to a broadly typical level of performance. However, as tasks become increasingly difficult, perceptual deficits may often be revealed. Where observed, co-occurring deficits of identity and expression recognition are consistent with impaired structural encoding of faces.
Meeting abstract presented at VSS 2015