Covert face recognition relies on affective valence in congenital prosopagnosia

Abstract

Dominant accounts of covert recognition in prosopagnosia assume subthreshold activation of face representations created prior to onset of the disorder. Yet, such accounts cannot explain covert recognition in congenital prosopagnosia, where the impairment is present from birth. Alternatively, covert recognition may rely on affective valence, yet no study has explored this possibility. The current study addressed this issue in 3 individuals with congenital prosopagnosia, using measures of the scanpath to indicate recognition. Participants were asked to memorize 30 faces paired with descriptions of aggressive, nice, or neutral behaviours. In a later recognition test, eye movements were monitored while participants discriminated studied from novel faces. Sampling was reduced for studied–nice compared to studied–aggressive faces, and performance for studied–neutral and novel faces fell between these two conditions. This pattern of findings suggests that (a) positive emotion can facilitate processing in prosopagnosia, and (b) covert recognition may rely on emotional valence rather than familiarity.

Keywords:

• Covert recognition
• Prosopagnosia
• Affective valence
• Face recognition
• Visual scanpath

Acknowledgments

This research was supported by an Economic and Social Research Council (ESRC)/ Medical Research Council (MRC) interdisciplinary studentship (PTA–037–2006–00006) and an ESRC project grant (RES–000–22–1910). The authors would like to thank Alexander Todorov and James Uleman for sharing their trait inference behaviours and norming data with us. The research in this paper use the FERET (Facial Recognition Technology) database of facial images collected under the FERET programme, sponsored by the Department of Defense (DOD) Counterdrug Technology Development Program Office. The authors would also like to thank Brad Duchaine for access to the Cambridge Face Memory Test and Cambridge Face Perception Test.