It's more than just physical: The contribution of social category information to race-selective face aftereffects

Abstract

Opposite changes in perception (perceptual aftereffects) can be simultaneously induced for faces from different social categories—for example, Chinese and Caucasian faces (Jaquet, Rhodes & Hayward, in press), or male and female faces (Jaquet & Rhodes in press; Little, DeBruine & Jones, 2005). We investigated whether these aftereffects are generated in high-level visual coding that is sensitive to the perceived face representation (or the social category) of the faces, or in earlier visual coding that is sensitive only to simple physical differences between faces. We created face continua ranging from SuperCaucasian faces (caricatured Caucasian faces) to SuperChinese faces (caricatured Chinese faces). Seventy-six Australian Caucasian participants and 72 Hong Kong Chinese participants were adapted to oppositely distorted (contracted or expanded) face sets that were a fixed physical distance apart on the morph continua. The adapted sets either crossed the race category boundary (Chinese and Caucasian faces) or did not (SuperCaucasian and Caucasian faces or SuperChinese and Chinese faces). Larger opposite aftereffects were found when the adapted sets crossed the race category boundary than when they did not. These results suggest that opposite aftereffects for Chinese and Caucasian faces reflect the recalibration of neurons tuned to high-level face information rather than simple physical face differences. An effect of expertise was also found. Opposite aftereffects could be induced for visually distinct sets within the participants own race (e.g., SuperCaucasian and Caucasian faces for Caucasian participants), but not for other-race face sets (e.g., SuperCaucasian and Caucasian faces for Chinese participants). We discuss the implications of these results for the representation of faces in face space. Acknowledgement: This research was supported by an Australian Postgraduate Award and UWA Completion Scholarship to Emma Jaquet, an Australian Research Council grant to Gillian Rhodes, and a grant (Project No. HKU 4653/05H) from the Hong Kong Research Grants Council to William Hayward