Normal social evaluations of faces in acquired prosopagnosia.

People frequently evaluate facial attractiveness and make We tested four acquired prosopagnosics. Florence is a 29inferences about character traits from faces, and these evaluations are consistent across observers (Todorov, Said, & Verosky, 2011). Do these face evaluations rely on the same perceptual mechanisms as facial identity recognition? Recent studies claim they do not. Some individuals with developmental prosopagnosia make typical trustworthiness (Todorov & Duchaine, 2008) and attractiveness (Carbon, Grüter, Grüter, Weber, & Lueschow, 2010) judgments, and one acquired prosopagnosic judged attractiveness and several character traits normally (Quadflieg, Todorov, Laguesse, & Rossion, 2012). These studies however suffer from three methodological limitations that raise concerns about the validity of their findings. First, face evaluations were measured only with ratings, a format that may not be sensitive enough to detect subtle impairments. Second, while face evaluations were measured with ratings, identity perception was assessed with different formats. Third, the faces evaluated by prosopagnosics in two studies (Carbon et al., 2010; Quadflieg et al., 2012) were in color and included hair. Both hair and color can influence evaluations (Stephen, Coetzee, & Perrett, 2011; Todorov & Duchaine, 2008) and some prosopagnosics can perceive these cues normally (Fox, Hanif, Iaria, Duchaine, & Barton, , 2011; Garrido et al., 2009). To overcome these limitations here we investigated perceptions of facial attractiveness, trustworthiness and aggressiveness for greyscale faces without external facial cues. We assessed each evaluation using three formats: rating, sorting, and categorisation. We chose attractiveness because of its evolutionary importance (Rhodes, 2006), and trustworthiness and aggressiveness because of their central roles in a leading model of face evaluation (Oosterhof & Todorov, 2008).