Original ArticleSexual dimorphism of facial width-to-height ratio in human skulls and faces: A meta-analytical approach
Introduction
The idea that the human face provides social information is not a new one (Darwin, 1872). We can determine the identity (Bruce & Young, 1986), sex (Burton, Bruce, & Dench, 1993), age (Rhodes, 2009), and ethnicity (Montepare & Opeyo, 2002) of a stranger with relative ease, as well as more dynamic and changing information like emotional state (Elfenbein & Ambady, 2002). There is also evidence that trait information like personality, physical and mental health, and even sexual orientation can be perceived with some accuracy from faces alone (Jones et al., 2012, Kramer and Ward, 2010, Rule et al., 2009, Scott et al., 2013).
In 2007, researchers provided evidence of one particular facial measure, the width-to-height ratio (FWHR – see Fig. 1; Weston, Friday, & Liò, 2007), which they found to be sexually dimorphic in human skulls, and has since been the subject of intense investigation as a cue to numerous behaviours. While overall size differences play a large role in general skull dimorphism (Calcagno, 1981, Lestrel, 1974, Rightmire, 1970), Weston and colleagues suggested that this ratio difference was instead due to developmental differences in shape trajectories during puberty. Specifically, the height of the upper face (defined as the nasion-prosthion distance) in adults is similar in men and women, while the (bizygomatic) width is larger in men. In other words, while sex differences in skulls are expected simply because men grow to be larger than women, the bizygomatic width in males show additional growth at puberty beyond this predicted increase. The researchers argued that this difference in skull shape might result from intersexual selection pressures, so that a region of the face has evolved which highlights the distinction between men and women.
Why evidence of sexual dimorphism predicts an association between FWHR and behaviour is less clear. If female preferences led to increased facial width in men (although evidence actually suggests that wider faces are judged to be less attractive; Geniole, Denson, Dixson, Carré, & McCormick, 2015), it may not necessarily follow that within-sex differences are correlated with behaviours. More intuitively, intrasexual selection pressures (e.g., male-male competition) could have resulted in increased success for wider-faced men, resulting in an appearance–behaviour link, especially if these two factors have the same underlying mechanism (testosterone, for instance). Might both explanations overlap, whereby a facial cue that highlights ‘maleness’ to women has become associated with masculinity (both in appearance and behaviour) in men? Of course, there is no reason to assume that the mechanisms underlying sex differences in facial development are the same as those that may drive a within-sex association between appearance and behaviour.
While the precise account and its relationship with sexual dimorphism remain unclear, FWHR does appear to function as a social cue. Levels of masculine characteristics (e.g., aggression, dominance, deception) in men correlate with FWHR, as do perceived levels of these traits (for meta-analyses, see Geniole et al., 2015, Haselhuhn et al., 2015). The explanation for this FWHR–behaviour association is thought to involve testosterone (Carré and McCormick, 2008, Sell et al., 2009), which may influence both facial development and behavioural characteristics. Indeed, initial research found significant associations between FWHR in men and baseline levels of testosterone, as well as testosterone changes in response to potential mate exposure (Lefevre, Lewis, Perrett, & Penke, 2013).
Somewhat problematically for this account, FWHR may not actually be sexually dimorphic in faces (Kramer et al., 2012, Lefevre et al., 2012, Özener, 2012) or skulls (Gómez-Valdés et al., 2013, Stirrat et al., 2012). Of course, it may be that different mechanisms drive facial development in men and women, allowing for testosterone-produced correlates of behaviour in men without differences between the sexes (Lefevre et al., 2013). In a recent meta-analysis of this field, the authors found significant (but small) sex differences when considering studies of both skulls and faces together (Geniole et al., 2015), as well as for subsets of studies (2D photographs versus other materials). However, it is not clear whether differences remain when only skulls are analysed since this distinction was not made in their analyses. It may be that skulls do not show sex differences in FWHR but faces do, perhaps through evolved cues that utilise soft tissue deposits, which differ in men and women (Enlow, 1982). This would be an important caveat when investigating the explanatory mechanisms linking behaviour and facial measures.
One potential issue with previous investigations is that they have not considered populations separately based upon ethnicity or geographical origin. Given evidence of between-population differences in skulls (Gill and Rhine, 2004, İşcan and Steyn, 1999, Ousley et al., 2009), the inclusion of all groups into a single analysis will inherently suffer from this additional source of noise. It may be that FWHR dimorphism is present in some ethnicities/populations but not others, and this could account for the mixed results that have previously been found with faces. This would also be an important caveat for theories of dimorphism and signalling.
The other problem for the ‘FWHR–testosterone–behaviour’ account is that FWHR may not actually be associated with testosterone. In recent research investigating several samples and reporting a combined meta-analysis, no relationship was found between FWHR in adult men and baseline testosterone or competition-induced testosterone reactivity (Bird et al., 2016). Even during adolescence, when testosterone is hypothesised to impact facial growth (Weston et al., 2007), no relationship was found between male FWHR and testosterone levels or other known testosterone-derived traits (Hodges-Simeon, Hanson Sobraske, Samore, Gurven, & Gaulin, 2016). Indeed, FWHR showed no change during adolescence and no growth spurt, contrary to predictions.
In the current work, I focussed specifically on whether FWHR is sexually dimorphic in adult human skulls using a meta-analytical approach. Given that the popular topic of FWHR as an important facial cue originated from this initial finding (Weston et al., 2007), it is worth further examination using multiple large samples. I also considered geographical and ethnic origins as potential factors in order to allow for the likelihood that populations may differ. For this reason, I revisit the topic of FWHR sex differences in faces, again considering ethnicity as a potential influence. Importantly, prior large-scale analyses in this area have yet to consider the distinction between faces and skulls, and the possibility (and evolutionary implications) that there may be FWHR sex differences in one but not the other.
Section snippets
Previous research
All peer-reviewed and published manuscripts that investigated human skull FWHR separately for men and women were included. This involved searching through all articles that cited Weston et al. (2007), the first paper to propose this measure as a topic of interest. Conveniently, all articles prior to the end of 2014 had already been identified in the recent meta-analysis by Geniole et al. (2015), and no newer research (as of May 2016) or omissions were found. This resulted in the inclusion of
Meta-analysis of skulls for all populations
The Supplementary materials provide a summary of the eleven databases included in the meta-analysis, comprising 4918 men and 2924 women from 87 populations.
Discussion
The current meta-analyses provide evidence that casts doubt on what seems to be the currently accepted story regarding FWHR dimorphism. Although an overall analysis of human skulls found a very small (though statistically significant) effect, where men showed larger FWHR than women, there is an argument to be made for considering ethnicities separately (Gill and Rhine, 2004, İşcan and Steyn, 1999, Ousley et al., 2009). Subgroup analyses were suggestive of a moderating effect of ethnicity
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Acknowledgements
I am very grateful to Lyman Jellema for providing access to the Hamann-Todd human osteological collection, as well as background information and other details regarding the specimens. I also thank Rolando González-José for providing the data that he and his colleagues had previously collected. Finally, I thank Geoff Cumming for statistical advice, and Alex Jones and Renée Lefebvre for comments on the manuscript.
References (55)
On the applicability of sexing human skeletal material by discriminant function analysis
Journal of Human Evolution
(1981)Digit ratios, but not facial width-to-height ratios, are associated with the priority placed on attending to faces versus bodies
Personality and Individual Differences
(2013)- et al.
Facial width-to-height ratio and celebrity endorsements
Personality and Individual Differences
(2014) - et al.
No evidence for sexual dimorphism of facial width-to-height ratio in four large adult samples
Evolution and Human Behavior
(2012) - et al.
Telling facial metrics: Facial width is associated with testosterone levels in men
Evolution and Human Behavior
(2013) Facial width-to-height ratio in a Turkish population is not sexually dimorphic and is unrelated to aggressive behavior
Evolution and Human Behavior
(2012)- et al.
Female sexual orientation is perceived accurately, rapidly, and automatically from the face and its features
Journal of Experimental Social Psychology
(2009) - et al.
Facial cues to depressive symptoms and their associated personality attributions
Psychiatry Research
(2013) - et al.
Male facial width is associated with death by contact violence: Narrow-faced males are more likely to die from contact violence
Evolution and Human Behavior
(2012) - et al.
Face shape and behavior: Implications of similarities in infants and adults
Personality and Individual Differences
(2015)