Object Recognition in Prosopagnosia
Jason J. S. Barton
University of British Columbia

Prosopagnosia is the impaired ability to recognize faces. It is considered a selective visual agnosia, yet there has been controversy for many decades about whether the recognition of non-face objects is truly spared by this condition. This is related to the debate about whether the human perceptual system has a modular organization, with cortical processes dedicated to highly specific visual processes. Several criticisms have been directed at the literature, such as inadequate sample sizes, inadequate sampling of object types, testing stages of perception that don’t match the prosopagnosic deficit, and failures to consider the effect of prior expertise and interest. We assessed a sample of twelve subjects with acquired prosopagnosia and 14 with developmental prosopagnosia. We applied two experiments. The first was a test of short-term familiarity for several object types. The second examined car expertise effects through a two-part process. In the first, expertise was indexed by verbal semantic knowledge about cars. In the second, subjects performed a test of visual car recognition. In controls, the visual score is highly correlated with the verbal score. We found that both acquired and developmental prosopagnosic subjects were impaired on non-face object recognition. Both groups recognized fewer cars than predicted from their verbal score, and increases in expertise generated only half the normal gain in visual recognition scores. Nevertheless, an item-concordance analysis continued to show expertise effects in prosopagnosic groups. We conclude that a more detailed analysis of object recognition in prosopagnosia can reveal expertise-adjusted deficits in visual performance, though expertise continues to exert an effect on their object recognition. This argues against the modular hypothesis of cortical perceptual organization, and supports an overlap of neural resources devoted to different perceptual tasks.