27 AUGUST 1998 In a jaw-mashing blow to macho movie men, a Scottish psychologist says that men's faces, on average, are more attractive when they're made slightly more feminine. David Perrett of St. Andrews University used computer graphics to create average faces. Then he allowed research subjects to vary the facial structure to make them slightly more masculine or more feminine. Masculinizing a photo, for example, would enlarge the jawbone, while feminizing it would enlarge the lips or alter the forehead.
As expected, both men and women preferred women with a slightly "feminized" face. But in a finding that could flatten the average thin-lipped, jut-jawed, heavy-browed action hero, both men and women also preferred a slight "feminizing" of the men's faces.
And while it's clear that people don't base mating decisions on faces alone (the body, personality, and for all we know even the bank account could also play their parts) the finding does seem a slap at the notoriously fragile male ego.
More important -- at least to scientists -- it could also undermine current theories of mate selection. Psychologists who take evolution seriously expect men to choose women with faces having a bone structure that reflects the presence of high levels of estrogen, the female sex hormone, which, in turn, is associated with high fertility and longevity. No problem so far -- that's exactly what Perrett and associates found.
Time to shave that beard?
It's the flip side that's causing the turmoil. Presumably, a male face whose heavy bone structure reflects high levels of the male hormone testosterone would indicate a strong, dominant male. And evolutionary psychologists expect that evolution would have "trained" women to seek qualities like these in a mate. Why? Simply because power and strength should help the couple's children to survive.
But Perrett actually found the opposite. While the subjects did consider the more masculine looking males "dominant," they also found them cold and intimidating. And when asked to select the most "attractive" faces, they chose faces that were 15 to 20 percent more feminine than average.
The Method
During a decade of research into facial attractiveness, Perrett and his colleagues began wondering about the negative traits that volunteers were associating with the hyper-male faces that they chose as most "dominant." (Researchers used that term, Perrett says, to indicate the man's reproductive fitness, but the exact definition was left to the subjects.)
Using technology borrowed from the special-effects industry, Perrett's group digitized photos of people in Japan and Scotland and produced separate average faces for men and women from each location. Then they taught a computer to distort the average images in masculine or feminine directions.
During the test, each subject viewed a computer image of a male or female face, then used the mouse to maximize attractiveness (again defining the term as they wished). The use of computer morphing gave an exact measure of how far the optimum face departed from average.
The madness
To Perrett, the fact that feminized male faces were more attractive indicates a problem for psychologists who look at evolution and mate selection. He grants that strength has its evolutionary role: "Humans have had to cope with defensive situations, with predators, with hunting, and all that requires considerable musculature. But on balance, when a woman is looking for a mate, she may be looking for a multitude of attributes." Dominance "is not the only criteria that people use to assess attractiveness."
Alternatively, the unusual finding may be revealing something about the reproductive state of the women subjects. Randy Thornhill, professor of biology at the University of New Mexico, says that his work on sexual attractiveness indicates that women have varying preferences during different stages of the menstrual cycle: "I'd predict that if they controlled for female reproductive state, the results would show two kinds of preferences." During the fertile, mid-cycle time, women may be more attracted by high-testosterone males. At other times, they may be more interested in men who seem likely to make a greater investment in the children. (Intriguingly, he says, women who take birth-control pills would be likely to make only "sensible" choices for good providers, since they are never fertile).
Yet simply by contradicting previous studies, Thornhill says, Perrett's work is useful. "We have varying results -- that's always important in scientific progress. All these studies are solid, so the question becomes, what is causing this variation?"
Of Arnold Schwarzenegger and gorillas
Yet if women were indeed selecting men with highly "masculine" faces, while men were selecting highly "feminine" women, men's faces would differ dramatically from women's. By questioning the mechanism of selection, Perrett's work offers a reason why they don't differ so much.
So stand back, Chuck Norris. If Perrett is right, the most attractive man may resemble Will Smith, the ace rocket jockey in the movie Independence Day. Is the man tough? You betcha -- no gang of Evil McNasty aliens will rattle him. But is he a good mate? No question. Smith's going to save this Earth if only because it's home to his mate and his child.
The new study could also explain why men and women look as similar as they do (in contrast, male baboons and male mountain gorillas are almost twice as large as their female counterparts). More to the point, human faces can be so similar that it may be hard to distinguish one gender from another.
-- David Tenenbaum
Credits | Feedback | Search