Reactions to children's faces

Abstract

Since cuckoldry risk is asymmetrical, we hypothesized that parental investment would be more affected by paternal than maternal resemblance. To test this hypothesis, we asked subjects hypothetical questions about investing in children under conditions in which their faces or those of other people had been morphed with photographs of children. Males were more likely to choose a face they had been morphed with as the most attractive, the child they were most likely to adopt, the child they would like to spend the most time with, the child they would spend US$50 on, and the child they would least resent having to pay child support for. Reactions to children's faces by females were much less affected by resemblance.

Introduction

Males invest less in children that they are not biologically related to and in many cases are more likely to abuse step or otherwise unrelated children Anderson et al., 1999, Daly & Wilson, 1988, Daly & Wilson, 1996, Marlowe, 1999. According to Daly and Wilson (1998) there are two ways in which a male can increase the probability that the children he is caring for are carrying his genetic material: he can monitor and/or sequester the female partner during the period that she is fertile to reduce the risk of being cuckolded, or he can attempt to assess paternity based on the degree to which the children resemble him.

There are two forms that a paternal resemblance mechanism might take: (1) the degree to which a male is told a child resembles him (i.e., a “social mirror”) and (2) the degree to which the child actually resembles him. Daly and Wilson (1982) recorded spontaneous remarks in maternity wards regarding the appearance of newborn children. Mothers and their friends and relatives were more likely to comment on how children resembled their fathers than they were to say the child resembled the mother or any other family member. When fathers displayed any doubt, the mothers were quick to reassure them of the child's resemblance. Regalski and Gaulin (1993) have replicated these findings using Mexican families.

In a population of males convicted of domestic violence, Burch and Gallup (2000) found the more males felt that their children looked like them, the better the children were treated. The childhood of the abusive males themselves were also associated with how much they thought they resembled their fathers. Perceptions of paternal resemblance were correlated with the incidence of physical and sexual abuse they experienced as children, as well as feelings of closeness to their fathers.

Christenfeld and Hill (1995) found that objective raters did no better than chance at matching pictures of 1-year-old infants to their mothers, but could more accurately match the pictures of infants to their fathers. Other studies have failed to replicate these findings Bredart & French, 1999, McLain et al., 2000, Nesse et al., 1990, but in these studies, actual paternity was never determined (e.g., by DNA testing). Since the incidence of cuckoldry may be appreciable (see Baker & Bellis, 1995, for a review), it is easy to see how instances of failed paternity could statistically mask/obscure paternal resemblance.

A more compelling test of the paternal resemblance hypothesis would be to experimentally manipulate resemblance to determine the effect it has on treatment and attitudes toward children. In this study, facial resemblance was manipulated using computerized facial morphing based on combinations of subjects and children and reactions to these images were measured with hypothetical investment scenarios.