An inconvenient association between familiarity and distinctiveness ratings of familiar faces

New ratings data suggest that the more a famous face is judged familiar, the more it is judged distinctive, that is different from other faces. The same ratings obtained on unfamiliar faces do not show the same positive association. High distinctiveness ratings on celebrities might thus reflect an illusion of distinctiveness—a subjective impression that does not match the objective distinctiveness of a given face’s features. Such illusion, making objective distinctiveness ratings difficult to obtain for faces that are widely known, could be problematic for building well-controlled materials to study familiar face recognition, the impact of distinctiveness itself or individual differences in face recognition abilities. To objectify this phenomenon and test if it can be mitigated, we examined the links between familiarity and distinctiveness ratings obtained from full headshots of celebrities and between familiarity and compound distinctiveness ratings calculated from ratings on four different isolated facial parts. Isolating facial parts is known to disrupt facial recognition, and so ratings of isolated parts should be less influenced by familiarity than ratings of headshots. We found that the association between familiarity and distinctiveness was stronger than between familiarity and compound distinctiveness, both at the individual participant level and at the face item level. The eyes and external features were the most distinctive while the mouth and the nose areas were the least. We will discuss the implications of this illusion of distinctiveness for future research on familiar face recognition and how it might further our understanding of information contained in facial representations.