Traditionally, perceptual processing of faces and words is considered highly specialized, strongly lateralized, and largely independent. This has, however, recently been challenged by studies showing that learning to read may affect the perceptual and neural processes involved in face recognition. In this light, investigating face processing in dyslexia, and reading in prosopagnosia becomes interesting: Do deficits in the two domains dissociate?
We present data from 11 people with developmental prosopagnosia, which is a disorder of face processing in people with no known brain injury, and in the context of normal intelligence and other cognitive abilities. The face processing deficits in developmental prosopagnosia appear early in life and seem to be the result of developmental problems that are currently poorly understood. In three experiments, we investigated whether reading performance in this group was abnormal. First, we examined if reading speed was affected by word length in any of the subjects. Secondly, we compared performance with single word and single letter stimuli using RT measures. Third, we measured the word superiority effect in accuracy of word and letter report with brief exposure durations. These data were also analysed using methods based on a Theory of Visual Attention1, to extract estimates of perceptual processing speed for words and letters.
We find that the group of developmental prosopagnosics perform well within the normal range on all reading tests. In the traditional RT test, they show normal RTs, and no abnormal word length effects. As a group, they also show an RT advantage for short words over single letters, as we have previously found in normal subjects.2 In the word superiority experiment, the group of prosopagnosics show the typical word superiority effect, reflected in better overall accuracy, a lower perceptual threshold, and higher processing speed for words compared to letters.
In sum, we find no evidence that reading skills are abnormal in developmental prosopagnosia, a finding that may challenge the recently proposed hypothesis that reading development and face processing abilities are intrinsically linked.