Purchase this article with an account.
Rachel Robbins, Patrick Heck; Brad Pitt & Jude Law: Individual-contingent face aftereffects and norm- versus exemplar-based models of face-space. Journal of Vision 2009;9(8):516. doi: 10.1167/9.8.516.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Contingent aftereffects have recently been shown for faces differing in orientation (upright/inverted; Rhodes et al., 2004), sex (male/female; Little et al., 2005), race (Caucasian/Chinese or African; Jaquet et al., 2008; Little et al., 2008), age (child/adult; Little at al., 2008) and species (human/monkey; Little et al., 2008). That is, simultaneous adaptation has been shown to opposite distortions (e.g., eyes moved together vs. eyes moved apart) for different face categories. These contingent aftereffects are interpreted as evidence that different cell populations code different face categories, and sometimes as evidence that there are different “norms” for different face categories. Here we show contingent aftereffects for in-out eye-distance in two individuals (Brad Pitt and Jude Law). In the adaptation phase, half the participants rated the attractiveness of pictures of Brad with eyes moved together (eyes in) and Jude with eyes moved out (eyes out), and half rated the reverse (Brad eyes out/Jude eyes in). In the test phase, participants judged which of a pair of pictures looked most normal. In each pair, two versions of the same picture of Brad or Jude were shown, one eyes in and one eyes out. Those who saw Brad eyes out/Jude eyes in at adaptation were more likely to prefer Brad with eyes out at test compared to Jude. Those who saw Brad eyes in/Jude eyes out at adaptation were more likely to prefer Jude with eyes out at test compared to Brad. Based on previous results this might be interpreted as showing separate cell populations and norms for Brad and Jude. However, a norm for each individual seems essentially the same as exemplar-based coding. Thus, these results may suggest that face recognition is more exemplar-based than norm-based. Alternatively, the current distinction between norm- and exemplar-based coding may need to be rethought.
This PDF is available to Subscribers Only