Research into how humans choose a mate is often guided by evolutionary theory: because people’s choice of mate is assumed to have consequences for reproductive success, it must therefore be subject to Darwinian selection.
We know that facial judgements of personality are often made very quickly (in milliseconds) and automatically. And faces seem to be particularly important, so preferences for specific facial characteristics – symmetrical, masculine and feminine faces, for example – are often assumed to be the product of selective forces operating throughout our ancestral past.
But do particularly male or female faces indicate anything important about their owners that might interest potential mates?
One hypothesis is that faces indicate hormonal status. For women, a particularly female face may signal a lot of oestrogen and therefore fertility. It’s more complex for men, but a very male face may indicate a lot of testosterone. Too much testosterone can actually suppress the immune system, but the fact that it is costly may be the point – it demonstrates that the face’s owner is healthy enough to bear the burden, so he must be relatively resistant to local pathogens. High testosterone can also lead to competitiveness, with both good and bad implications.
Many studies have found that men consider particularly female faces to be more attractive than average ones. The evidence is more equivocal for men, but some studies do show that, when presented with masculinised and feminised versions of the same male face, women prefer the more masculine option – particularly in contexts where the benefits of “good genes” are more relevant, such as for short-term relationships – where longer-term parental investment is not on offer – and during the fertile phase of the menstrual cycle (although the importance of menstrual cycle effects is a contentious area).
Looking further afield
These findings have been replicated in a number of different populations: European, American and Japanese students. But most studies have been conducted with participants who live in modern, developed economies, sometimes described as WEIRD (Westernised, Educated, Industrialised, Rich, and Democratic) populations. As the acronym suggests, these participants may not be entirely representative of humanity as it is normally practised, and certainly not as it was practised for most of our evolutionary history.
These populations are over-represented because they are much easier for most researchers to access – but would we still see the same preference patterns in completely different environments?
To test this question, 21 co-authors and I ran a study looking at perceptions of faces from 12 different populations, including ones from traditional small-scale subsistence societies. We showed participants (almost 1,000 in total) photos depicting masculinised, neutral and feminised versions of an opposite-sex face and asked them to choose which version was most attractive, and also which was the most aggressive-looking.
We expected that preferences for facial sex-typicality (and perceptions that masculine faces are aggressive) would been seen everywhere, and would perhaps even be stronger in populations where conditions were closest to those experienced by our hunter-gathering ancestors. What we found, however, was the exact opposite.
In smaller-scale groups, preferences for sex-typical faces were attenuated or even reversed. All groups agreed that masculinised male faces were the nastiest, but the participants from more urbanised groups perceived this much more strongly.
What these results imply is that preferences for extreme sex-typicality are more typical of newer, modern societies and were probably not a feature of our ancestral past. In particular, the hypothesis that masculine faces are valued because they advertise disease resistance looks particularly weak, since we’d expect these preferences to be greatest where disease burden was highest – and in fact they were strongest in those countries with the lowest disease burdens.
So why do these preferences appear in modern developed societies if they aren’t part of an historical legacy? We’re not yet sure, but we speculate that the sheer number of strangers we encounter in urbanised societies may be the key.
Personal experience is a less reliable source of information when confronted with so many strangers, but the huge number of faces to which we are exposed may allow us to detect whatever weak relationships do exist between facial characteristics and desirable traits within this environment, essentially both forcing and allowing us to lean more heavily on the use of facial stereotypes.
So are these findings a problem for the idea that human behaviour has been shaped by evolutionary processes? In a word, no. Evolution via selection is a simple process, but it can unfold in countless ways, and in this case our evidence is showing that it has unfolded in a slightly different way than many previously imagined.
The idea that humans can be functionally responsive to their environment is not controversial among researchers interested in human evolution. But the findings do pose problems for the view that human beauty ideals are just random cultural constructs (so unrelated to any function other than a product of things such as history, fashion, advertising), because the patterns of preference we observed were very well predicted by the environmental people were in. It suggests that the development of our preferences is decidedly non-arbitrary.