Helen of Troy might have had "the face that launched a thousand ships," but that visage might not float anyone's vessel in Fiji.

It turns out that a person's bias toward highly feminine or masculine facial features in the opposite sex might have more to do with how many people live close by than with anything else, according to a study published this week in Proceedings of the National Academy of Sciences.

Researchers went to rain forests, steppes, islands, cities and campuses to get more than 900 people from 12 cultural groups to choose the most appealing face of the opposite sex. They selected among several similar composite mug shots made up of features typical to their region. Two of the composites, however, were slightly caricatured โ€” eyes, lips, cheeks, brows and other features slightly altered to appear more "masculine" or "feminine."

A growing number of studies report a preference toward facial features that are more sharply masculine or feminine. The popular hypothesis to explain the apparent bias holds that these features accurately signal fitness and fertility. Features deemed "feminine" correlate with estrogen and fertility, while masculinized faces correlate with testosterone, which increases aggression and tunes down the immune system's inflammatory reactions to pathogens.

If these hypotheses are true, researchers surmised, then preference for caricatured features ought to be greater in less economically developed societies, which are under greater stress and privation. And everyone across cultures ought to associate masculine faces with aggression, the researchers reasoned.

"Instead, what we found was they didn't really share these perceptions at all," said the study's lead author, Isabel Scott of Brunel University, Uxbridge. She called the results "a great surprise" that changed her view of the going theories espoused by fellow evolutionary anthropologists.

Preference for exaggerated features was strongest in large urban societies and groups with low rates of disease, homicide and fertility, the study found. In some small-scale societies, in fact, women's preferences skewed away from the square-jawed and strong-browed man, according to the study.

And while the association between aggression and masculinity of features was detected across groups, it was strongest in the wealthier urban samples, the study found. Urbanization, in fact, was the variable that best predicted the strength of that association, according to the study.

"It really kind of jumped out as explaining a lot of the variation in the data," Scott said.

Preference for strongly male or female features not only appears to be less "universal," according to the study, it may be new to human evolution โ€” a product of densely populated settlement and the social demands it entails.

At the very least, the study suggests that some scientists trying to read the evolutionary meaning behind facial variety might just be staring at a mirror of their own culture.

Still, what happens on paper may not be what happens in bed.

"The next step in general is trying to find out how these preferences actually play out in terms of mating," said Michael Sheehan, an evolutionary biologist at the University of California, Berkeley who was not involved in the study.

Sheehan is interested in a different set of questions over a longer period of time: how human faces came to be so varied in the first place. A genetic study he published in Nature Communications last week suggests that difference is long-lived in our genome.

Sheehan and his co-author, Michael Nachman, went to the source that cares a lot about measuring human features: the U.S. Army, which has to design equipment to fit its soldiers. Its database of troops of European and African-American ancestry confirms that there is more variance among facial features than among any other body traits.

That's exactly what would be expected if faces evolved in favor of helping us distinguish among individuals, said Sheehan.

"The traits on the face should be less correlated with each other than they are for other parts of the body," Sheehan said. "If you tell me how widely spaced your eyes are, I cannot predict at all how long your nose is going to be. They are not related to each other." But width of a hand can reliably predict its length, he noted.

The researchers searched the human genome for clues about whether this high variability was a byproduct of other changes over evolutionary time, or selected by evolution. They found evidence of selective pressure.

Stretches of DNA associated with facial features were more diverse than the overall genome or than stretches associated with other physical features. Mutations that popped up around these areas also showed unusual diversity.

Lastly, comparisons with equivalent stretches of DNA from extinct ancestors such as the Neanderthals also suggested that selection began well before their split with modern humans, Sheehan said.

Many studies have shown that humans developed a brain with unrivaled social cognition โ€” an ability to distinguish identity and glean information from faces. The reciprocal effect may be just as profound, according to the study. Social evolution may have changed the very shape of our faces.

Social factors that drove facial diversity don't leave fossils. But it's highly unlikely that sexual selection was the sole or even paramount factor, Sheehan said.

Scott's work, in fact, suggests sexual selection for facial features is unlikely to have been the same across cultures and over time.

Read or Share this story: