Faces on Her and His Mind: Female and Likable

Abstract

Faces are a valuable source of non-verbal information for daily life social interaction. Mounting evidence points to gender specificity in face perception. Here we search for the factors that can potentially trigger gender differences in tuning to faces. By using a set of Face-n-Food images slightly bordering on the Giuseppe Arcimboldo style, we examine: (i) whether face resemblance is linked to gender specific face impression, and, if so, whether this association is perceiver gender specific; and (ii) whether images most resembling a face are also most likable for female and male perceivers. First, in a spontaneous recognition task, participants were shown a set of Face-n-Food images in a predetermined order from the least to most resembling a face. Then in a two-alternative forced-choice (2AFC) task, participants judged whether each face appeared for them (i) either female or male (Exp. 1); or (ii) either likable or unlikable (Exp. 2). Remarkably, face resemblance is closely connected to gender specific impressions: images more resembling a face elicit also more female-face responses. This link is not perceiver gender specific as it occurs for both females and males. Moreover, face resemblance is positively linked to face likability, but this holds true only for female perceivers. The findings shed light on gender specificity in tuning to faces, and help to clarify abnormalities of the social brain in neurodevelopmental, psychiatric and psychosomatic disorders.

Fig 1. An example of portraits by Giuseppe Arcimboldo.

“The Gardner” by Guiseppe Arcimboldo (1526–1593), an Italian painter best known for creating fascinating (often grotesque and allegoric) imaginative portraits composed entirely of fruits, vegetables, plants, tree roots, flowers, and even books and human bodies (http://www.wikiart.org/en/giuseppe-arcimboldo/the-gardner; public domain).

Fig 2. An example of Face-n-Food images.

One of the most resembling a face, eliciting most female-face responses, and at the same time, one of the most likable Face-n-Food images.

Fig 3. Proportion of female-face responses.

Left: Proportion of female-face responses to each Face-n-Food image pooled across female and male perceivers. Vertical bars represent 95% CI (confidence interval). Right: Relationship between face resemblance and female-face impression shown separately for female and male perceivers. The image number reflects its face resemblance (1 –the least recognizable as a face, 10 –the most recognizable); trendlines are shown for illustrative purposes only.

Fig 4. Face resemblance and face gender.

Close relationship occurs between gender specific impression (as proportion of female-face responses) and face resemblance (as proportion of face responses) for female, left panel, and male, right panel, participants. In both female and male perceivers, face resemblance positively correlates with female-face impression; trendlines are shown for illustrative purposes only.

Fig 5. Face resemblance and face likability.

Relationship between likability (as proportion of likable-face responses) and face resemblance (as proportion of face responses) for female, left panel, and male, right panel, participants; trendlines are shown for illustrative purposes only.