© PHOTO ALTO
From symmetry to smell to the dance floor groove, how evolution carves our ideas of sexy
Dorothy Hopcroft got it right. Agreeing to a date at the urging of a meddling friend, she didn't quite know what to make of Frederick Turton. He arrived on a bicycle (her former beau had a car) made at the factory where he worked a tough week with little prospect of promotion, and his financial status was such that Heinz Baked Beans were something of an extravagance. But somehow that didn't matter.
His face, his body, his voice, even his smell, might have made the difference. That's before the other, less outwardly noticeable factors came into play: his sense of humor, his generosity, his chivalry. All of these were possibly part of that complex equation. And despite some social status warning signs, Dorothy fell in love. "We found that the way we danced together, the quick step and everything, was wonderful."
What is the clincher when it comes to choosing a partner? For millennia we have been content not to probe too deeply into the rules of attraction from a scientific standpoint. But is that logical? Isn't our choice of mate, with whom we conspire to further our genetic legacy, every bit as subject to scientific scrutiny as any other biological process? Dorothy's choice resulted in 63 years of happy marriage and family that now includes three great-grandchildren (including this author's sons) and counting. To her the choice was simple. "I never looked at another fella. I knew he was the one for me," she says. But the signs that drew them together, whether they're universal or local, remain elusive. More and more, evolutionary scientists are trying to pin down the traits that tell all, those universally attractive qualities that might provide a biological answer to the question: Is this the one?
We expect traits to be the way they are for a reason. Male stalk-eyed flies, for example, have (as their name suggests) eyes set so far apart that they appear at the end of stalks. Why? Because the females like them that way. For some reason lost in time, they took preference to the males with the greatest "eyespan." Such males enjoy higher reproductive success and produce greater numbers of offspring than their less well-endowed rivals.
Female stalk-eyed flies can observe and compare males before choosing to mate with the one they find most attractive: an outward display signals its bearer's quality to mates and rivals alike. Something like that happens in humans too, though males aren't quite as wide-eyed, and indeed it's not necessarily solely the female's choice. Nevertheless, as a species we universally seem to be preoccupied with the process of seducing and being seduced by the right partner. But unlike stalkies, fruit flies, and flour beetles, which can be manipulated en masse in terms of their environmental and experimental regimes, the genetic underpinnings of human mate choice are significantly trickier to get at.
"Humans are a good model in the sense that you can interview them. You can ask very specific questions relating to their behavior," says Nina Wedell, a Royal Society Research Fellow at the University of Exeter in Cornwall. "On the other hand it's very difficult to get a good estimate of what is really going on: Even though there is some variation, we don't have lots and lots of babies."
Insects lay huge numbers of eggs. Humans are significantly stingier reproducers. Wedell, who works on sexually selected traits in a range of insect species, says that such are the obstacles in unravelling human mate-choice behavior. With an insect experimental model, it is possible to manipulate the degree of relatedness, parasite load, developmental environment, and a range of other factors, which in humans would be impossible. "You can do lots of things with insects that there's just no way you could ever do with humans," Wedell says.
Sexual signals in humans are rarely as overt as in the warriors of the Karamojong tribe in northern Uganda, who stretch their penises to such extraordinary lengths that they must literally tie knots in them to prevent injury.1 For most human cultures, however, the cues are a little more subtle. (Understanding what these cues are, and how they affect mate choice, is the goal that evolutionary psychology has set for itself.)
How do most people actually choose their mates? Men are interested in indicators of fertility, health, and youth, says David Buss, a professor at the University of Texas and currently president of the Human Behavior Society. Of these, the ratio of waist-to-hip circumferences is the best studied so far. "It's linked to age, pregnancy, and several nasty diseases." Men universally view a low waist-hip ratio (WHR) - a slender waist with wider hips - as a sexually appealing trait in women. Buss argues that the evidence for the existence of "universal indicators" has accumulated to such an extent that there is now little debate over their ecological validity.
Many outward, static cues appear to be correlated. "If you've got a low WHR, you've probably got an attractive face," says Ian Penton-Voak, a psychologist at the University of Bristol, UK. He stresses the notion that whatever information is being transmitted with these cues is likely to be about biological fitness in a general way. The high level of redundancy is perhaps a means to ensure that the message gets across unambiguously. But static features can be overpowered by even more attention-grabbing stimuli. "Dynamic cues - how people behave, their moving faces, and so on - can completely drown any individual differences in the structure of the face," Penton-Voak explains.
A recent confluence of modern methods has allowed a team led by William Brown at Brunel University, UK, to do just that.2 Brown and his colleagues at Rutgers University and the University of Washington in Seattle are interested in whether or not dance (which seemed to play so heavily into Dorothy's choice of Fred) serves as a signal of mate quality. Bodily symmetry, or lack thereof, has long been postulated to be an indicator of genetic quality, developmental history, or some combination of the two. Dance, Brown's team reasoned, offers a ritualized form of movement that could serve to highlight an individual's left-right symmetry. "What got me interested in symmetry is that it isn't just confined to humans. There are many studies showing symmetry having locomotor movement performance enhancements," says Brown.
Rutgers collaborator and noted evolutionary biologist Robert Trivers runs the Jamaican Symmetry Project.3 In 1996, he and his colleagues logged the differences between right and left sides of 288 children, mainly pupils at Hill Top Primary School. They measured, photographed, X-rayed, and interviewed their subjects. They even took dental casts. This information comprises the most detailed data on bodily symmetry in humans ever collected. It has since been followed by a second round of measurements in 2002, which provide information on the stability of each individual's symmetry over time.
With a neat study population in place, Brown needed to find a way of isolating a subject's dance moves from all the other extraneous information. The problem was how to test the effects of dance alone, in the absence of any physical characteristics of the dancer; height, weight, general appearance, and so on.
Motion-capture technology held the key. Brown and colleagues used multiple high-speed infrared cameras to mathematically extract information about the dancers' movements. These data were then mapped onto computer-generated avatars - "blank" figurines lacking in facial or bodily detail - resulting in a short film of a virtual dancer, gyrating away in an otherwise featureless world.
The animations were divided into two groups, those depicting symmetrical dancers and those depicting asymmetrical ones, and were presented to people from the same cultural background as the dancers themselves. Viewers rated the performance of each dancer on a simple scale and, as Brown predicted, gave higher scores to those avatars whose movements had been extracted from relatively symmetrical human counterparts. So despite the complete lack of physical cues for symmetry, something about the avatar's movement was giving the game away. Dance, it would appear, is a good way to convey information about one's genetic and phenotypic quality, at least in Jamaica.
For Dorothy, Fred's dancing ability confirmed his all-around athleticism. "He took me home on the bus, then ran home. He told me later that his feet hardly touched the ground all the way."
But how much of a leap of faith is it to say that body symmetry yields information about someone's genetic quality? Brown's findings are the latest in a long line of studies on the role of symmetry. Fluctuating asymmetry (FA), denoting random differences across the left-right divide, was a topic that caused great excitement among evolutionary biologists during the late 1980s and early 1990s. Studies of a wide range of animal species, from Australian sheep blowflies to African lions, suggested a link between differences in the right and left sides of various measurable traits and underlying genetic quality. The concept was readily adopted by evolutionary psychologists, who began to study FA in relation to facial attractiveness. Since then, however, FA has fallen from favor among scientists studying nonhuman species.
"From a vast array of studies on nonhuman species, and on humans, what is apparent is that sometimes, for some traits, fluctuating asymmetry appears to correlate with the quality of the individual, although for many other traits, and other species, it does not." Andrew Pomiankowski, a professor at University College London, says this makes it difficult to determine exactly what FA reveals.
Pomiankowski's studies of stalk-eyed flies suggest it is the distance between the eyes relative to the male's body size, rather than differences between right and left eyestalks, that females are interested in, he says. Moreover, for eyestalk asymmetry and two nonsexual traits (wing length and width), FA provides no information about an individual fly's developmental history or genetic quality. In fact, one study showed that larvae reared in the most "benign" environment, the one causing the least physiological stress, had a significantly higher overall level of FA than those raised in more damaging environments.4,5 Pomiankowski hasn't even found evidence that FA is heritable, undermining the notion that it might reveal something about genetic quality. "I've stopped bothering to measure fluctuating asymmetry in my studies," he says, "There is no simple metric that gives an indication of absolute fitness."
Nonetheless, studies in humans have consistently suggested that FA does play some role, even if that role isn't yet clear. Several studies have shown that relatively symmetrical faces, for example, are generally found to be more attractive than less symmetrical ones. Other results are contradictory. "When you present people with pictures of just one half of a vertically divided face, it doesn't appear to hinder their ability to make judgments about attractiveness that correlate with symmetry," says Penton-Voak. "The jury is still out," he says, "but I don't think facial symmetry is a cue to attractiveness at all."
And it's not just a right-left divide. Waist-to-hip ratios, the jewel in the evolutionary psychologist's crown of universal cues, are in fact a pretty abstract measure. WHR is calculated as the ratio between two circumferences, but most experiments have concentrated on frontal views. A recent cross-cultural study highlighted the impact this can have.
Early experiments suggested a difference in men's preferences in the United States and those of the Hadza tribe in Tanzania, with the average preferred WHR of the former being lower (i.e., curvier) than the latter. Frank Marlowe and colleagues asked subjects to rate profile views. The overall difference between the two populations disappeared: Hadza men have a greater penchant for protruding buttocks than their US rivals, effectively leveling the score. The WHR emerges relatively unscathed as a universal indicator of female attractiveness, but there are subtle geographic variations.6
Nevertheless, studies like this underscore the caution with which static measures should be treated. Just measuring them is difficult. Finding out how people actually use the cues in real life is another matter altogether. Marlene Zuk, a professor of biology at the University of California, Riverside, states what should be perfectly obvious. "We're not peacocks." Having worked on mate choice in several species and coauthored a seminal paper on the relationship between bright plumage and parasite burdens, she is skeptical of the way that some of the concepts developed in studies of animal behavior have been caricatured by some proponents of evolutionary psychology. It simply isn't possible to think about humans in the same way, she says, because there are so many complicating factors, culture being just one.
Brown's findings mark a shift at least from the study of static cues such as facial symmetry and WHR to more dynamic, expressive indicators. He wants to explore the idea that dancing ability signals not just developmental stability but also mental agility. Because human dance is often complex, he says, it could be signaling something about the neuromuscular coordination of the dancer in addition to his or her creativity. These more complex cues, according to Geoffrey Miller, an evolutionary psychologist at the University of New Mexico, allow some insight into the mental workings of the actor, which is perhaps what people are really interested in.
"Looks are important, but only for the first five minutes," he says. People use those cues to filter out the no-hopers, but other factors soon come into play. "In a singles bar, women might initially pay attention to cues of status and wealth, and men might initially pay attention to cues of physical attractiveness, but then once they start talking they'll start to shift into focusing on the mental and personality traits." Miller's studies suggest that a man's sense of humor, and his musical or artistic ability - traits that typically do not bring material advantages - seem to be valued by women as indicators of high intelligence and creativity. He is currently conducting experiments to test these ideas.
Perhaps there are aspects to human mate choice that evolutionary psychology cannot discern, but which might bring the field more into the mainstream of evolutionary biology. "What we've learned in 40 years [of behavioral ecology] is that a lot of mate choice happens after copulation," says Wedell. "None of that has been applied to humans, but it's possible that it might have some potential explanation for variation in human fertility." In particular, she says, the link between mate choice and the human major histocompatibility complex (MHC), a genomic region responsible for immune function, could be important.
Several experimental studies have linked the MHC to measures of attractiveness (see "Match 'n' Sniff"). In many other species, particularly insects such as crickets, females mate with more than one partner but have internal mechanisms that allow them to give preference to the sperm of one (usually less related) male over another. Wedell suggests a human parallel mediated by MHC compatibility: If precopulatory mate choice is influenced by biases in perceived attractiveness that are correlated to MHC differences, it could serve to increase fertility just as surely as the female cricket's sperm-sorting apparatus. "I think these kinds of ideas might help us to understand human fertility issues."
By reducing mate choice to a handful of static traits, research by people such as Brown is now embracing more complex questions. As the technology advances, methods can become more sophisticated, adding extra layers to the simple scenarios that have so far been studied. "Perhaps we'll be able to understand some of the most costly aspects of human behavior," he says, "those involved in sexual selection."
Brown says he intends to delve further into the factors that determine perceptions about attractiveness and dynamic traits. Great strides have been made in imaging movement, he says, which can also be brought to the study of facial movement as well as dance. Understanding how facial movement interacts with facial symmetry could be revealing. "No one's actually looked at that, and I really think it's the future for the field, at least for now."
Although the nature of love will remain clay for the artisans of the world, scientists are steadily gaining a foothold on the nature of the choices that underlie it. Dorothy and Fred probably made hundreds of calculations on the static and dynamic attributes they saw in each other so many decades ago, and their decision paid off. Whether science can bring about the reduction of seduction isn't the question, but to the evolutionary psychologists who work to dissect it, it's a quest with an attractive challenge. The literature on human mate choice was "extraordinarily simplistic" before the advent of evolutionary psychology, says David Buss. As new studies tackle increasingly sophisticated and subtle questions, that picture is now changing. But, he says, "there's still a very long way to go."
2. W. M. Brown et al., "Dance reveals symmetry especially in young men," Nature, 438:1148-50, 2005.
3. R. Trivers et al., "Jamaican symmetry project: long-term study of fluctuating asymmetry in rural Jamaican children," Hum Biol, 71:417-30, 1999.
4. P. David et al., "Male sexual ornament size but not asymmetry reflects condition in stalk-eyed flies," Proc Royal Soc Lond B, 265:2211-6, 1998.
5. T. Bjorksten et al., "Fluctuating asymmetry of sexual and nonsexual traits in stalk-eyed flies: a poor indicator of developmental stress and genetic quality," J Evol Biol, 13:89-97, 2000.
6. F. Marlowe et al., "Men's preferences for women's profile waist-to-hip ratio in two societies," Evol Hum Behav, 26:458-68, 2005.