Abstract
Studies of individuals who were treated for congenital cataracts early in life have yielded an intriguing finding: when tested on face individuation tasks several years after treatment, they exhibit impairments in configural processing relative to those who had normal visual development. The genesis of this impairment is unclear, but explanations have tended to focus on the idea of a 'critical period' for normal development of face processing. Deprived of patterned visual inputs during this period, the brain is believed to be permanently compromised in its ability to analyze facial information. While plausible, this account suffers from two weaknesses. First, it ascribes a special status to faces as a visual class in order to explain why only face processing, but not general object processing are compromised due to early visual deprivation. Second, the account leaves unspecified why within the domain of face processing the impairment is so narrowly circumscribed, with only configural processing being adversely affected. Here we propose a parsimonious alternative hypothesis that sidesteps these weaknesses. Building upon the observation that the post-operative acuity that this group of individuals start out with is higher than that available to the typically developing visual system, we have conducted computational simulations to probe how high-resolution information may impact the learning of facial representations. These experiments reveal a detrimental effect of high initial acuity on configural encoding. This leads us to suggest that the face processing deficits reported in individuals who had suffered early visual deprivation may be due, at least in part, to having access to fine image structure beyond what normal development allows. The hypothesis has potential implications for understanding prosopagnosia, the design of therapeutic interventions, and cautions to be followed in the post-operative care of children treated for congenital blindness.
Meeting abstract presented at VSS 2016