For line drawings of colleagues' faces more extreme caricatures look g
ood (Rhodes, Brennan & Carey, 1987) than for photographic caricatures
of famous faces (Benson & Perrett, 1991). The experiment reported here
used line drawings of famous faces to see whether this discrepancy is
due to the different kinds of faces used or to the different picture
media used. The mean caricature level of the preferred picture was +5.
5%, which is close to the +4.4% reported by Benson and Perrett, but mu
ch lower than the +16% reported by Rhodes, et al. Apparently more extr
eme caricatures look good for colleagues' faces than for famous faces,
which is consistent with the idea that highly familiar faces (e.g., c
olleagues' faces) are ''caricatured'' in memory. Recognizability judgm
ents were also obtained for pictures of birds-the undistorted drawings
were preferred.