Choosing a partner

The peacock’s tail

On 3 April 1860 Charles Darwin wrote to the American botanist Asa Gray with a confession. The eye, he said, had once made him cold all over, because it seemed too perfect to have arisen by natural selection. But now small, trifling details of structure disturbed him more. “The sight of a feather in a peacock’s tail, whenever I gaze at it, makes me sick!” Natural selection favors what helps an animal survive. A peacock’s tail is heavy, bright, and costly to grow, and it makes the bird easier for a fox to catch. How could it have evolved?

Darwin’s answer, in The Descent of Man, and Selection in Relation to Sex (1871), was a second kind of selection. Survival is not enough: an animal must also find a mate. If some males are more likely than others to win mates, then whatever helps them win will spread, even if it makes them less likely to survive. Sexual selection works in two ways: members of one sex compete with one another, and members of the other sex choose. In 1975 the Israeli biologist Amotz Zahavi suggested why a costly signal might be trusted: only a bird in good condition could afford to waste so much on a tail. That is the “handicap principle.”

Does it work for peacocks? The experiments are famous, and they disagree. Marion Petrie and her colleagues watched a group of ten males at Whipsnade Park in England in 1988. Eleven peahens each visited several males, and in ten of the eleven cases the male they chose had the most eyespots in his tail of those visited. When Petrie later trimmed eyespots from some males’ tails, their mating success fell, and the chicks of males with more elaborate trains grew and survived better. But in a seven-year study of free-living birds at Izu Cactus Park in Japan, Mariko Takahashi and her colleagues found no sign that peahens preferred males with more eyespots, longer trains, or more symmetrical ones. A later analysis of several populations suggested that eyespot number matters where males differ a good deal, and stops mattering where nearly all males have close to the maximum. The honest summary is that females in some places favor more eyespots, in others they do not, and “peahens choose the biggest tail” is too strong.

Who chooses?

Which sex does the choosing? In 1972 the biologist Robert Trivers put forward a principle that has organized the field ever since. Parental investment, he said, is anything a parent does for one offspring that raises its chance of surviving and costs the parent the chance to invest in others. The sex that invests more is the scarcer resource for the other sex, so the other sex competes for it, and the sex that invests more can afford to be choosy. Where females produce the large, costly sex cells, as Lesson 2 described, and carry the young, this predicts that males will compete and females will choose.

A famous earlier study seemed to prove it. In 1948 the geneticist Angus Bateman bred fruit flies, in groups of marked males and females, and concluded that males’ number of offspring varied far more than females’, and increased with their number of mates. Textbooks have repeated this ever since. But in 2007 Christine Snyder and Patricia Gowaty went back to Bateman’s data and found flaws: the way mates were counted, the sampling, the statistics. In 2012 Gowaty and colleagues repeated his experiment and found that mutations in the marked offspring biased the counts, so that Bateman’s result was largely an artifact. The data cannot stand as proof. The principle does better: in 2016 a large analysis of 66 species, by Michael Janicke and colleagues, found that sexual selection is stronger on males on average, and that the difference tracks how much more females invest in the young.

The roles can be reversed, and that is evidence for the principle. In the pipefish Syngnathus typhle females compete for males, who carry the eggs. In the pheasant-tailed jacana, a tropical water bird, most females have several mates, and defend territories, and it is the males who tend the chicks. In such species it is the sex that does the caring that does the choosing. Try changing who is busy in the model below.

The price of an egg

For people, the costs differ in several ways, and the numbers are worth knowing. A woman has hundreds of thousands of egg cells at puberty, estimates vary, and only a few hundred are ever released. A healthy man makes roughly 100 to 200 million sperm a day. A pregnancy costs the mother about 77,000 kilocalories over nine months, which is roughly 25 to 30 marathons’ worth of energy, though it is not burned all at once, and some of it is stored in the baby and in her body. Childbirth has always been dangerous. The World Health Organization estimated about 260,000 maternal deaths worldwide in 2023, 197 for every 100,000 live births; the United States had 18.6 per 100,000 that year. In historical England it was probably between a half and one percent per birth, so a woman who bore five or six children faced a risk of a few percent. The danger is far lower in rich countries now, but the biology that selection acted on was the older one.

Fathers face a different problem, uncertainty. A mother knows a child is hers; a father, before DNA tests, could not. “Mama’s baby, papa’s maybe,” runs the English saying. How often is the saying true? Not as often as people think. Reviews by Maarten Larmuseau and colleagues put the rate of children who are not their legal father’s at about one to two percent, and about one percent per generation in European family trees going back centuries. The popular figures of ten percent or more come from laboratories whose clients already doubted their paternity.

A famous experiment

If the costs differ, the sexes should differ in how they respond to a stranger’s offer of casual sex. In 1978 and 1982, on the Florida State University campus, the psychologists Russell Clark and Elaine Hatfield sent attractive young men and women, working as confederates and told to approach only people they found attractive, to approach students of the other sex. They asked one of three questions: would you go out with me tonight, would you come to my apartment, and would you go to bed with me. The answers, from their combined results and a later repeat, are shown in the table.

Two points are often lost. Men and women were equally willing to go on a date, so it is not true that men turn down every ordinary invitation. And the difference is real: it has been found again in Denmark, Germany, and France, in street studies in which, for example, about 38 percent of men and 2 percent of women in Denmark accepted the blunt invitation, and 32 percent and 2 percent in Germany. The question is why. Terri Conley found in 2011 that both women and men judged women who made such an offer as more intelligent, successful, and sexually skilled than men who made it, that expected skill and pleasure predicted who accepted, and that the gap in acceptance disappeared when the person asking was a famous person, a friend, or of the same sex. A fear of being judged badly explains part of the gap as well, and in one small German study it vanished in a laboratory set-up in which women felt safer. The cost of reproduction may play a part, but so do safety, expectations, and a double standard that praises men and criticizes women for the same behavior.

What we say, and what we choose

Maybe the choice of partner depends on who does the approaching. Eli Finkel and Paul Eastwick tested this at speed-dating events, with 350 people. At some events the men rotated from table to table while the women stayed seated, and at others the reverse. Whoever rotated was less selective, apparently because they felt more confident, and the usual sex difference in selectivity disappeared when the women were the ones who got up and moved. A custom that looks natural turned out to shape behavior.

The same researchers tested what people say they want. Asked in advance, men said they wanted attractive partners more than women did, and women said they wanted partners with good earning prospects more than men did. But when they met real people, the way attractiveness and earning prospects affected who they actually liked was the same for women and men. A 2014 analysis of 97 studies found that a partner’s attractiveness predicted romantic interest at a correlation of about 0.40 and earning prospects at about 0.10, for both sexes. What people say is one thing; what they do at a table is another.

The broadest survey is David Buss’s of 1989: 10,047 people in 37 samples from 33 countries. Women rated a partner’s financial prospects higher than men did in 36 of the 37 samples; men rated good looks higher in all 37. Men preferred partners about 2.7 years younger and women preferred partners about 3.4 years older. Both sexes put kindness and intelligence at or near the top. In 2020 Kelly Walter and a large team repeated the study in 45 countries, with 14,399 people. The sex differences replicated and were moderate: about 0.3 of a standard deviation for earning prospects, and about 0.27 for looks. The country with the largest gap in valuing financial prospects was China, where the gap in valuing looks was among the smallest. Do the gaps shrink as women gain equality? A 2012 study by Marcel Zentner and Klaudia Mitura said yes. The 2020 team did not find that robustly. The question is open.

What draws us together

Before any preference has a chance to work, there is geography. In 1932 the sociologist James Bossard looked at 5,000 consecutive marriage licenses in Philadelphia and found that a third of the couples had lived within five blocks of each other before they married. At MIT in the 1940s, Leon Festinger and colleagues found that residents of married-student housing were far likelier to be friends with their next-door neighbors than with people at the far end of a court. Proximity matters because it makes meeting possible, because we tend to like those we expect to see again, and because familiar things become pleasant: Robert Zajonc’s “mere exposure” effect, in which repeated sight of a face or a shape raises liking. A 1977 study found that people preferred a photograph of their own face as they see it in the mirror, while their friends preferred it the way others see it.

Two more effects come from single, small studies and should be held loosely. In 1994 Daniel Wegner and colleagues had two mixed-sex pairs of strangers play cards, with one man and one woman playing footsie under the table, and found that the contact raised attraction more when it was secret than when it was open. In 2011 Erin Whitchurch and colleagues told women that men had looked at their Facebook profiles and liked them a lot, a little, or either, and found that the uncertain condition made the men most attractive. Neither has been widely replicated. A 2017 study by Samantha Joel and colleagues used machine learning on speed-dating data and found that it could predict a little about who is generally liked and who generally likes, but not who would click with whom.

Faces, bodies, voices, and smells

What do people find attractive? Faces first. In 1990 Judith Langlois and Lori Roggman averaged photographs of faces by computer and found that the composites were judged more attractive than almost all of the faces that made them, and the more faces were blended, the more attractive the result. But the single most attractive real faces beat the averages, and the claim that average or symmetrical faces signal good health is weak: reviews in 2006 and 2017 found only small links. Raters in different cultures agree about which faces are attractive to a surprising degree, with correlations around 0.9.

Bodies and the waist-to-hip ratio have a more complicated story. In 1993 Devendra Singh reported that men judged line drawings of women with a ratio of about 0.7 as the most attractive and healthiest, and that Miss America winners and Playboy centerfolds over decades kept to ratios near 0.7, though their weight fell. Later work cut this down. In studies by Martin Tovée and colleagues, body mass index mattered more than waist-to-hip ratio. A 1998 study by Douglas Yu and Glenn Shepard found that this preference was not shared by men of an Amazonian group, the Matsigenka of Peru, and a 2010 survey of 7,434 people in 26 countries found heavier bodies preferred in low-income places. Hungry men preferred heavier figures than well-fed ones did. Fashions change as well: foot binding, which began in China around the tenth century and spread down the social ladder, was a standard of beauty that nobody now defends. The fair summary is that preferences for relatively low waist-to-hip ratios are common, but 0.7 is not a universal constant.

Other claims have fared worse. In 2007 Coren Apicella and colleagues found that Hadza hunter-gatherer men with lower-pitched voices had more children, among 49 men; a later reanalysis found that the link vanished once a man’s reputation as a hunter was counted. In 1995 Claus Wedekind found that women rated the smell of T-shirts worn by men whose immune-system genes, known as the MHC, differed from their own as more pleasant. A 2017 analysis of 58 results from 30 studies found no general tendency of this kind. And although it is often said that a kiss exchanges DNA, there is no evidence for it: a ten-second kiss transfers about 80 million bacteria. Romantic kissing is not universal either. An anthropological survey found it in only 46 percent of 168 cultures.

Like with like

Do opposites attract? A 2008 analysis of 313 studies by Matthew Montoya and colleagues found that actual similarity and attraction go together, at a correlation of about 0.47. In existing relationships, though, it is perceived similarity that matters. Among married couples, the largest similarities are in political values, around 0.58, religion, 0.56, and education, 0.55, according to a 2023 analysis of 199 studies. Spouses are also alike in intelligence, 0.44, and, to a smaller degree, in height, 0.24. Personality is the exception, at only 0.08 to 0.21. Sex is the one respect in which most people look for a difference. We also tend to like people who seem to like us. One suggested reason for liking similar people is that they are easier to live with and confirm our choices.