by Vanessa Wilson and Drew Altschul
Over the last 16 years, the study of human facial morphometry, and its potential links to social behaviour, has received a lot of attention. One metric in particular, the facial Width-to-Height ratio, has been linked to an array of dominance-like traits in males, such as aggression, achievement drive, psychopathy and (low) cooperation. In other words, men with higher ratios (i.e. relatively wider faces) tend to exhibit higher levels of dominance-like traits. Moreover, studies have found that people also tend to associate wider faces with being more dominant. The idea that we can ‘read’ certain traits in a person’s face is certainly an attractive one. But in reality, is such a relationship really so straight forward?
A number of studies have now cast doubt on the idea that these relationships are anything more than spurious. Indeed, early proposals that testosterone is the underlying mechanism that drives the relationship between aggressive, dominant male behaviours and wider faces are not strongly supported in the literature. There are several alternative explanations for what seems to be a tentative link between facial morphology and social behaviour. Amongst them is the evolutionary mismatch hypothesis, which proposes that facial cues to behavioural tendencies might have been salient in ancestral environments, yet no longer predict those tendencies in modern environments. Nevertheless, ancestral biases remain, resulting in a mismatch between existing behavioural traits, and how people predict those traits from facial features.
Where then, do nonhuman primates come into this? The problem with human-focuses studies is that, aside to the fact that most studies draw their samples from WEIRD populations (western, educated, industrialised, rich and democratic), by concentrating research on only humans, we learn little about the origins of facial morphology as a cue to social behaviour. If we really want to get into the evolution of the facial Width-to-Height ratio, or any other metric, as a behavioural cue, then we need to consider whether parallel relationships exist in other species. One theory that addresses the mechanisms underlying the link between face width and aggression, proposes that a wider zygomatic arch (producing a wider face) could improve skull strength, and therefore offers advantage in physical combat. Since humans live in relatively egalitarian societies, with low levels of in-group aggression, this could explain the weak findings for facial metrics as cues to traits pertaining to combat advantage. By examining whether similar relationships exist in species with lower levels of social tolerance and higher physical aggression, one can test this hypothesis.
“By concentrating research on only humans, we learn little about the origins of facial morphology as a cue to social behaviour.”
In our recent study, we examined links between facial morphometry and personality ratings in chimpanzees, a species that live in male-dominated, fission-fusion societies, where male displays of strength and physical aggression play an important role in social interactions. Notably, examination of chimpanzee personality dimensions on a 54-item scale reveals six components, five of which resemble the human ‘Big five’ (Openness, Conscientiousness, Extraversion, Agreeableness, Neuroticism), and a sixth that is labelled Dominance. Several of these components have been linked to observations of aggression in chimpanzees, in particular Dominance. We examined keeper-ratings on these six dimensions in relation to photo-derived measures of the facial Width-to-Height ratio, in a sample of 131 captive chimpanzees from zoos and research facilities in Japan, the UK and the USA. We found no evidence of sexual dimorphism in the facial Width-to-Height ratio, consistent with findings in humans. We did find a relationship between facial Width-to-Height ratio and the dimension Dominance, but in females, rather than males (as has been reported in the human literature), which was specific only to the western subspecies (Pan troglodytes verus).
At first glance these results may seem surprising. The differences in findings for males and females could, however, be explained by differences in rank stability with age. That is, females tend to maintain somewhat stable ranks across the lifespan, whereas males compete over social status, which tends to vary across the lifespan. Indeed, since male chimpanzees tend to be more aggressive than females, these findings do not support the hypothesis that the facial Width-to-Height ratio provides a combat advantage. The picture is perhaps more complicated than that, since our findings suggest that how social status is both achieved, and maintained, could play a role in whether the facial Width-to-Height ratio predicts dominant tendencies.
Moreover, we compared our findings with those from previous studies on humans, and bonobos, by examining differences in the strength of relationship between facial Width-to-Height ratio and social traits. We found that, compared to results for chimpanzee females and bonobos, the effect sizes for both humans and chimpanzee males were negligible, with the strongest effect emerging for bonobo Affiliative Dominance. Given that both bonobos and humans are considered to be more socially tolerant than chimpanzees, these differences suggest that social style (i.e. the level of tolerance a species exhibits) is not a strong predictor of trait-related variance in the facial Width-to-Height ratio, at least amongst apes.
So where do these findings stand with regard to current interpretations of the human literature? Alongside the accumulating results from brown capuchins, the Macaca genus, and bonobos, these findings suggest that the facial Width-to-Height ratio existed in a common primate ancestor, before the divergence of New World monkeys. Whilst questions remain about both the existence and strength of the facial Width-to-Height ratio as a cue to behavioural tendencies in humans, answers to these questions may be found by taking a comparative approach to understanding what selection pressures have driven this relationship. Specifically, examining broader differences in social style and rank stability could hold the key to understanding the variable nature of this relationship amongst primates.
“Our own recent evolutionary history, with the development of culture and language, could have provided alternative means to communicate dominance, reducing the need for facial cues to such traits”
For now, our own interpretations support the proposal of Wang and colleagues for an evolutionary mismatch in humans: our own recent evolutionary history, with the development of culture and language, could have provided alternative means to communicate dominance, reducing the need for facial cues to such traits. Yet the floor remains open for more thorough investigation, and we strongly encourage other researchers pondering this field to embrace comparative research as well as non-WEIRD samples, in order to understand the existence, origins and mechanisms of the relationship between facial morphology and dominance behaviour.