Passion, companionship, and what to keep

A corner of People’s Park

On weekend afternoons, in a corner of People’s Park in the middle of Shanghai, parents stand beside umbrellas and rows of paper sheets. Each sheet advertises a grown-up child: age, height, education, job, income, the zodiac animal of the birth year, and often whether there is a flat. The parents do the browsing, the haggling, and sometimes the hand-wringing, while the children themselves are often nowhere to be seen. The “marriage market,” 相亲角 (xiāngqīn jiǎo, “blind-date corner”), has been held there since about 2004, according to press accounts, and similar corners have appeared in parks across Chinese cities. A 2013 news report put women at about four in five of the people advertised. A study of the Beijing corners found that for women, parents advertised youth and good looks, and for men, education, income, and wealth.

Take a flight to the United States and the market looks different, though the idea is the same. In a 2022 survey, 30 percent of American adults told the Pew Research Center that they had used a dating site or app, and about one in ten of those who were married or partnered had met their partner that way; among partnered people under 30, one in five. A profile is a sheet of paper that fits in a hand, and a swipe is a decision taken in a second. In both places a person is a list of attributes. This lesson asks what evolutionary psychology and the other sciences of the mind have to say about why people marry, how they love, and how they should treat the answers.

Why people marry

A common explanation of marriage is a neat one. Human babies are born helpless, because our big brains make birth hard, and a baby takes years to become independent. So the two parents bargain: they agree to share the work of raising a child, and marriage is the name of the bargain. The first part is true. A newborn chimpanzee’s brain is about 40 percent of its adult size; a newborn human’s is about 28 percent. Birth is difficult, and human babies are very dependent.

The popular explanation for why is less certain. For half a century it was said that human babies are born early because upright walking narrowed the pelvis while big heads kept growing, the “obstetrical dilemma” of the anthropologist Sherwood Washburn. It is sometimes added that babies would need a year and a half, or more, in the womb. But in 2012 Holly Dunsworth and colleagues pointed out that human pregnancy is actually longer than expected for a primate of our size, and proposed that it ends when a mother can no longer supply the energy, at about twice her resting metabolic rate. Studies of the pelvis have found that its width does not predict how much energy walking costs. A different model suggests that the narrow birth canal arises from weak selection on both the baby’s size and the pelvis’s width. The debate is live: the safe statement is that birth is hard and newborns are immature, and that pelvic width as the explanation is not established.

Nor are parents the whole of the story. Human children have always been raised by many people. The anthropologist Kristen Hawkes and colleagues argued in 1998 that grandmothers, who forage and share food, helped women have more children, and Sarah Blaffer Hrdy has shown in Mothers and Others (2009) that human infants are unusual among apes in being cared for by siblings, aunts, grandparents, and fathers as well as mothers. A pair-bond sits inside that web. Only a small minority of mammals, roughly 3 to 9 percent depending on how it is counted, live in lasting pairs. Comparative studies disagree about why. One finds that pairing evolved where males could reduce the killing of infants by rival males; another finds that pairing arises when a male cannot defend several females, and that fathers’ care follows from pairing, rather than causing it. There is no single accepted answer for people.

Human marriage also takes many forms. In the Ethnographic Atlas, a catalog of 1,231 societies compiled by the anthropologist George Murdock, 186 societies were classed as monogamous, 453 allowed occasional polygyny, 588 frequent polygyny, and 4 polyandry, in which one woman has several husbands. That means about 85 percent of the societies permitted a man to have more than one wife. It does not mean that most men had several: since there are about as many men as women, in most such societies most marriages must still have been between one man and one woman. Monogamy as the rule, in other words, is a historical rarity that became common with the rise of certain large societies, a puzzle that Joseph Henrich and colleagues discuss in a 2012 paper.

Love and marriage

A claim that often accompanies this explanation is that love is not the reason people marry. As a matter of history there is truth in it, and as a matter of human nature there is not. In a survey of anthropological records, William Jankowiak and Edward Fischer found evidence of romantic love in 147 of 166 societies, and in the other 19 no one had reported on it either way. People have fallen in love everywhere. What is new is the idea that love should be the reason to marry. The historian Stephanie Coontz has argued that for most of history marriage was an economic and political arrangement between families, with love a hoped-for bonus, and that when love became the stated basis of marriage in the last few centuries, the old reasons weakened and expectations rose.

The picture of Greek husbands who thought love for a wife an illness is too simple. A famous line in a court speech from fourth-century BCE Athens, spoken by a litigant named Apollodoros, says that men keep courtesans for pleasure, concubines for daily care, and wives to bear legitimate children and guard the household. But it is a lawyer’s speech aimed at a jury, not a statement of what every husband felt. In the Odyssey, Odysseus wishes the young Nausicaa a husband, and says nothing is better than when “man and wife should be of one mind in a house,” in Samuel Butler’s translation. Xenophon described a wife as a partner in the household, and Aristotle thought the friendship of husband and wife could be of the highest kind.

In imperial China, marriages were arranged by parents and go-betweens. The philosopher Mencius, in the fourth century BCE, said that young people who dig holes to peek at each other without waiting for their parents’ orders and the go-between’s words will be despised by everyone. By the Ming code, a first marriage was arranged by grandparents or parents. A man who took a second wife while the first was living was liable to ninety blows and the marriage was annulled; commoners could take a concubine only at forty and only if they had no son. The popular phrase 三妻四妾 (sān qī sì qiè, “three wives and four concubines”) was a rich man’s privilege, never the norm. And romance has been a theme of Chinese writing for a very long time: the Book of Songs poem “Guanju” tells of a man tossing and turning with longing, and the playwright Tang Xianzu wrote in the preface to The Peony Pavilion (1598), 情不知所起，一往而深, “Love arises, no one knows how, and grows ever deeper.” Shen Fu’s memoir Six Records of a Floating Life, written around 1807, is a love story about his marriage to Chen Yun.

Law changed this in the twentieth century. China’s Marriage Law, passed on 13 April 1950 and in force from 1 May, banned arranged and forced marriage, bigamy, concubinage, and child betrothal, and set the minimum marrying ages at 20 for men and 18 for women. Today, how much love matters for marriage still differs among cultures. A 1995 study of university students in eleven countries by Robert Levine and colleagues, who asked whether they would marry someone with every quality they wanted who was not someone they loved, found that love mattered most to students in Western countries and least to those in India, Pakistan, and Thailand.

Passion and companionship

Psychologists divide romantic love into two kinds. Passionate love is the state of intense longing and absorption in another person that most of us mean by “falling in love.” Companionate love is the affection and deep attachment we feel for someone whose life is closely tied to ours. Ellen Berscheid and Elaine Hatfield drew the distinction in the 1970s, and Robert Sternberg in 1986 proposed a “triangular theory” in which love has three parts, intimacy, passion, and commitment, and different mixes make different kinds of love. A relationship with all three, he called consummate.

A familiar sketch shows two curves: passion shoots up as strangers meet and then falls, and companionate love begins at nothing and climbs slowly. As a picture of the average, there is something in it. A large review of 115 studies of marriages by Benjamin Karney and Thomas Bradbury found that, on average, marital satisfaction declines over time. In an analysis of life satisfaction in more than 24,000 people over fifteen years by Richard Lucas and colleagues, marriage gave people a boost that, on average, faded back to where they started within a couple of years. But the average hides a great deal. Try drawing a few couples below.

In 2012 K. Daniel O’Leary, Arthur Aron, and colleagues asked a random sample of 274 Americans who had been married for more than ten years how in love they were with their partner. About 40 percent said “very intensely in love”; the figure was 29 percent in a second sample, of 322 New Yorkers. A team led by Bianca Acevedo also scanned the brains of 17 people who had been married for an average of 21 years and who said they were still madly in love, and found activity in reward regions like that of people newly in love. That study is small, and the volunteers answered an advertisement asking who was “still madly in love,” so it shows what is possible, not what is typical. A study that followed 464 newlywed spouses for four years found that about four in five stayed happy or declined only slightly. A 2021 analysis of 165 samples found that, on average, satisfaction falls between ages 20 and 40, then rises into the 60s. The honest summary is that the first obsessive months do cool, that on average satisfaction slips early, and that many couples stay in love for decades. “Passion fades, friendship grows” is an average, not a law.

Arranged and chosen

Are arranged marriages less loving than marriages of choice? The best studies are small. A 2005 study by Jane Myers and colleagues compared 45 people in arranged marriages in India with existing American data and found no difference in marital satisfaction. A 2012 study of 58 Indian Americans, 28 in arranged and 30 in love-based marriages, found no differences in satisfaction, commitment, or love. Critics point out the small samples. Selection matters, since arranged marriages are not a random draw from the people who would choose them. So does stigma against divorce, which keeps unhappy couples together, and social pressure to report being happy.

The picture in China has changed greatly. A 1990 study by Xu Xiaohe and Martin Whyte of 586 ever-married Chinese women found that parents’ role in choosing had fallen sharply and that young people now dominated the choice. Yet parents have not left the stage: a 2000 study of urban China found that both the kind of marriage and the parents’ approval still bore on marital quality, and in People’s Park, the parents are at work. Many Chinese Americans will recognize the middle position: a family’s opinion that matters a great deal, in a society where the individual chooses.

Reading the divorce numbers

You may see a divorce rate for Beijing in 2002 of 59 percent, and the conclusion drawn that more than half of marriages fail. Be careful with numbers like this one. Almost certainly it is the number of divorces registered in a year divided by the number of marriages registered in the same year. That is not the share of marriages that end, because the divorces come from marriages made over decades, while the denominator counts only this year’s newlyweds. The ratio rises when the number of marriages falls, even if no more couples are splitting up.

China’s own figures show it. In 2019 about 9.3 million couples registered to marry and about 4.7 million registered a divorce, a ratio of 51 percent, and the divorce rate per 1,000 people was 3.36. In 2021, when a thirty-day cooling-off period before an agreed divorce came into force, the number of divorces registered fell by about a third, to 2.8 million. In 2024, there were 6.1 million marriages and 3.5 million divorces, a ratio of 57 percent, although the rate per 1,000 people, 2.49, was lower than in 2019. In Beijing in 2024 the ratio was about 59 percent. In the United States in 2023 there were 6.1 marriages and 2.4 divorces for every 1,000 people, and the number of married women who divorced in a year had fallen from 22.6 per 1,000 in 1980 to 14.4 per 1,000. As a rough guide, about 40 percent of first marriages end in divorce, depending on the period. Try the calculation below.

The stamp and the choice

Darwin’s The Descent of Man ends with a famous sentence: for all his noble qualities, man “still bears in his bodily frame the indelible stamp of his lowly origin.” Evolutionary psychology adds that the stamp is on our wants as well as our bodies. That deserves to be taken seriously, and it deserves limits. The first limit is the gap between what is and what ought to be. David Hume noticed in the eighteenth century that writers slide from “is” to “ought” without saying how, and the slide is still common: it is natural to want variety, therefore it is right to seek it; it is natural to be jealous, therefore it is harmless. Nothing follows. A second limit is that “genetic” does not mean “fixed.” Adult height is among the most heritable of human traits: a study of 30,111 pairs of twins gave estimates of about 0.9 for men and 0.7 to 0.8 for women. Yet Dutch men grew by roughly 18 to 20 centimeters in about 150 years, from around 165 to around 183, because heredity measures how much of the variation in a given place and time is associated with genes, and says nothing about what changing the environment will do.

The third limit is about explanation. You can describe a person as a swarm of neurons, or the Mona Lisa as pigment and atoms, or a Chopin nocturne as vibrations in the air. Each description is true, and none replaces the experience of the person, the painting, or the music. Reducing love to its history, or to the chemistry of reward, leaves out what it is like, and what it asks of those who feel it. Richard Dawkins, who made the “selfish gene” famous, ended his book with the thought that we alone on earth can rebel against the tyranny of the selfish replicators. Whatever the origin of our wants, we can ask what we want to want.

Darwin’s own story is a good place to end. The man who had weighed marriage on a list married his cousin Emma Wedgwood, and they stayed together for forty-three years, until his death in 1882. Emma wrote to him early in their marriage, worrying that his scientific habits of mind would keep him from a faith she held, and saying that it would make her unhappy if they “did not belong to each other forever.” Darwin kept her letter all his life. They had ten children, and seven lived to adulthood; their daughter Annie died at the age of ten, in 1851, and her death devastated him. Near the end of his life, writing his autobiography for his family, he said that he marveled at his good fortune that she, “so infinitely my superior in every single moral quality,” had agreed to be his wife. His list had been a calculation. His marriage was a life. No account of where our wants came from, however good, can tell us how to live; for that, we have to choose.