Item

Abstract

We investigated whether male and female faces are discrete categories at the perceptual level. We
created artificial gender continua between male and female faces using a 3D-morphing algorithm and used classical
categorization and discrimination tasks to investigate categorical perception of gender. In Experiments 1 and 3,
3D morphs were computed between male and female faces. The results of the discrimination task suggest that the
gender of unfamiliar faces is not categorically perceived. When participants were familiarized with the male and
female endpoint faces before testing (Experiment 3), a categorical effect was found. In Experiment 2, only shape
or texture of unfamiliar 3D morphs was indicative of gender, while other information (e.g. texture or shape) was
kept constant. Again there was no evidence of a categorical effect in the discrimination task. In Experiments 1, 2
and 3, changes in the gender of a face were also coupled with changes in identity which may have confounded the
findings. In Experiments 4 and 5, we used face continua in which only the gender of the facial features changed,
while the characteristic of the facial features remained constant. When the faces were unfamiliar (Experiment 4),
there was no evidence of categorical perception of gender. In Experiment 5, participants learned to classify the
face images in two gender categories using a feedback procedure. A clear categorical effect for gender was present
after training. Our findings suggest that despite the importance of faces, gender information present in faces is not
naturally perceived categorically. Consequently participants showed categorical perception of gender only after
training with the face stimulus set.