Date of Award
Doctor of Philosophy (PhD)
Professor Margo Wilson
Professor Martin Daly
Two lines of reasoning predict that highly social species will have mechanisms to influence behaviour towards individuals depending on their degree of relatedness. First, inclusive fitness theory (Hamilton, 1964) leads to the prediction that organisms will preferentially help closely-related kin over more distantly-related individuals. Second, evaluation of the relative costs and potential benefits of inbreeding suggests that the degree of kinship should also be considered when choosing a mate. In order to behaviourally discriminate between individuals with different levels of relatedness, organisms must be able to discriminate cues of kinship. Facial resemblance is one such potential cue in humans. Computer-graphic manipulation of face images has made it possible to experimentally test hypotheses about human kin recognition by facial phenotype matching. I show that humans respond to facial resemblance in ways consistent with inclusive fitness theory and considerations of the costs of inbreeding, namely by increasing prosocial behaviour and positive attributions toward self-resembling images and selectively tempering attributions of attractiveness to other-sex faces in the context of a sexual relationship. The four studies presented in this thesis use face images made to resemble either the experimental participants or unfamiliar people in order to test the impact of facial resemblance on behaviour and attributions. In Chapter 2, I demonstrated that people favour individuals represented by resembling faces in an economic game. In chapter 3, I tested the hypothesis that men, because of paternity uncertainty, should be more sensitive to phenotypic cues of relatedness to children than women should be. Although facial resemblance increased the attractiveness of and self-reported willingness to invest in pictured children, I found no evidence for a greater effect in men than women. To test whether responses to a kinship cue are sensitive to contexts in which the adaptive response should be different, I tested the perceived attractiveness of same-sex and other-sex self-similar faces in Chapter 4. Although self-resemblance increased the perceived "averageness" of male and female faces equally, self-resemblance increased the attractiveness of same-sex faces only. Additionally, in Chapter 5 I tested other-sex faces for trustworthiness, attractiveness in the context of a long-term relationship, and attractiveness in the context of a short-term relationship. Self-resemblance of other-sex faces increased attributions of trustworthiness, had no effect on attractiveness for a long-term relationship, and decreased attractiveness for a short-term relationship. The results of these four studies indicate that humans respond to self-resembling faces in ways that are consistent with facial resemblance being a cue of kinship. Nonadaptive hypothesis explaining these results as byproducts of general perceptual mechanisms were tested and not supported. Specifically, the finding that same-sex faces were perceived as no more average than other-sex faces was evidence against the hypothesis that separate mental representations of male and female faces could explain the increased attractiveness of same-sex self-resembling faces relative to other-sex selfresembling faces in Chapter 4. The opposite responses to an other-sex self-resembling face for judgments of trustworthiness versus judgments of attractiveness in the context of a short-term relationship in Chapter 5 provided evidence against the "mere familiarity" hypothesis whereby exposure globally increases liking. Consequently, responses to facial resemblance seem best interpreted as evidence of specialized adaptations to the problems of kin recognition in the domains of mate choice and prosocial behaviour.
DeBruine, Lisa M., "Facial Resemblance Affects Behaviour and Attributions" (2004). Open Access Dissertations and Theses. Paper 1600.