No global processing deficit in the Navon task in 14 developmental prosopagnosics

Abstract

Faces are represented in a more configural or holistic manner than other objects. Substantial evidence indicates that this representation results from face-specific mechanisms, but some have argued that it is produced by configural mechanisms that can be applied to many objects including words. The face-specific hypothesis predicts that non-face configural processes will often be normal in prosopagnosic subjects, whereas the domain-general configural hypothesis predicts they will be deficient on all configural tasks. Although the weight of the evidence favors the face-specific hypothesis, a recent study reopened this issue when it was found that three out of five developmental prosopagnosics showed a larger local processing bias than controls in a global-local task (i.e. a Navon task). To examine this issue more thoroughly we tested a significantly larger sample of prosopagnosics (14 participants) who had severe face memory and face perception deficits. In contrast to the previous report, the developmental prosopagnosics performed normally in the global–local task. Like controls, they showed a typical global advantage and typical global-to-local consistency effects. The results demonstrate that the configural processing required by the Navon task is dissociable from face configural processing.