A Book About Love 1 ATTACHMENT
If I can stop one heart from breaking,

I shall not live in vain.

—EMILY DICKINSON1

JOHN BROADUS WATSON didn’t believe in love. He was one of the most influential psychologists of the twentieth century,2 yet Watson insisted that the feeling was just a fantasy, a fairy tale, a four-letter word used to sell lipstick and sonnets and movie tickets. If love were real, he said, then it must be measurable, a cause with tangible effects. But Watson concluded that love was not like that—it could not be held in the hand or weighed on a scale—and so he declared it an empty cliché, as useless as poetry.

His doubt led to a discovery. When Watson said that love was not real, other scientists tried to prove that it was, and that no subject mattered more.
The Scientist
Watson was born, in 1878, into a destitute family. His mother, Emma, was a devout Baptist. His vagabond father loved whiskey and disappeared for weeks at a time to drink in the backwoods of South Carolina.3 They scratched out a living as tenant cotton farmers; Watson remembered laboring as a young child, “handling tools, half-soling shoes and milking cows.”4 He was bullied as a boy and then became a bully.5 Fighting, specifically with African-Americans, was one of his “favorite going-home activities.” He never went to high school because the local county didn’t have a public one.

But Watson wasn’t held back by his childhood. A self-styled truth-teller, he described his life as a testament to the possibilities of the modern age, a time in which people started to shed their old superstitions. After enduring five “bitter” years at the local college,6 Watson sent a letter to the president of the University of Chicago, promising to be an “earnest student.”7 The letter worked. In 1900, Watson headed north with $50 to his name, determined to prove his worth and change the world.8

Although Watson had intended to study philosophy at Chicago, the subject “wouldn’t take hold.” (He seems to have failed his class on Kant.)9 However, Watson soon became enchanted with experimental psychology, a young field that matched his ambitions. Human nature had always been a mystery, a subject full of myth and lore, but experimental psychology promised to finally reveal the truth. It could show us who we really are.10

Like many of these new psychologists, Watson began by stripping us down, searching for the simplest laws of the mind. In his 1913 manifesto, “Psychology as the Behaviorist Views It,” he declared that researchers had wasted too much time chasing after ideas that couldn’t be quantified, such as love and consciousness.11 They’d squandered centuries speculating about emotions and dreams and other airy nothings. Watson argued, quite rightly, that any real science had to be rooted in measurement. That meant focusing on behavior, studying the link between stimulus and response and ignoring everything in between. “The behaviorist . . . recognizes no dividing line between man and brute,” Watson wrote.12 Every living thing is just a reinforcement machine, responding to the primal incentives of food and sex.13

This strict view of psychology turned Watson into an academic star, a symbol of progress and potential. (Among young psychologists, Watson was hailed as a “second Moses,” leading his field out of the wilderness.)14 Before long, he was chair of the Johns Hopkins psychology department and, at the age of thirty-six, the youngest president of the American Psychological Association. But Watson was only getting started—his real goal was to apply his new science to the practical questions of everyday life. His most famous experiment featured “Little Albert,” a nine-month-old infant.15 First, Albert was exposed to the sight of a white rat. As expected, the baby boy reacted to the rodent with curiosity, reaching out to touch the animal. However, after the rat was paired with a loud noise—Watson clanged on a steel bar, held behind the infant’s head—Albert grew to fear every kind of furred thing, including rabbits, dogs, a sealskin coat, and even a Santa Claus mask. The lesson of the experiment was that fear, like every other emotion, was a learned reflex. Children didn’t love their mothers. They simply paired her face with the pleasure of milk, just as Albert learned to pair the fur with a feeling of fear.16 The theory was compelling, and the Little Albert paper would become one of the most frequently cited studies in American psychology textbooks.17

Watson’s collaborator on these experiments was a young graduate student named Rosalie Rayner. During their research, John and Rosalie began having a passionate affair. Unfortunately for Watson, his wife discovered a stash of their letters, which would be released during their divorce trial. The affair became a public scandal, featured on the front pages of Baltimore newspapers. Watson and Rayner’s correspondence was full of behaviorist language, an awkward attempt to describe their emotions in “objective” terms: “Every cell I have is yours individually and collectively,” he wrote. “My total reactions are positive and towards you. So likewise each and every heart reaction.”18 Forced to choose between his science and his love, Watson resigned from Hopkins. It was an ironic choice for a scientist who had spent years insisting that love was not real and certainly couldn’t influence our behavior.

But Watson was not one to stay down. He soon reinvented himself as a popularizer of behaviorism, a salesman for science.I He remained convinced that a psychology focused on observable facts—and not the invisible urges of emotion—could transform society, creating a world of maximum happiness. His first popular book was a primer on child care, since he believed that parenting was still mired in the mistakes of “emotionalism.” (Watson dedicated the book to “The first Mother who brings up a happy child.”) First published in 1928, Psychological Care of Infant and Child became a bestseller and remained the definitive guide to parenting until Dr. Benjamin Spock’s Common Sense Book of Baby and Child Care was published in 1946. (In a laudatory early review, Bertrand Russell proudly endorsed Watson’s child-care techniques, while even Watson’s critics admitted that “Watsonism has become gospel and catechism in the nurseries and drawing rooms of America.”)19 The appeal of the book was obvious: Watson pitched it as an empirical guide to parenting, a how-to manual inspired by the careful study of “more than five hundred infants” at Johns Hopkins Hospital.

So what could parents learn from this science? Watson’s fundamental message was that love wasn’t just overrated—it was unsafe. In one chapter, “The Dangers of Too Much Mother Love,” Watson insisted that all the kissing and coddling of parents reinforced the very behavior it was supposed to prevent. For instance, when a child cries, the typical mother reacts with soothing affection, which only encourages the child to cry more. (The tenderness rewards the bad behavior.) The result, Watson wrote, is “invalidism,” which will “wreck your adult son or daughter’s vocational future and their chances for marital happiness.”

Instead of loving our children, Watson advised that we treat them like coworkers. “Shake hands with them in the morning,” he wrote. “In a week’s time, you will find how easy it is to be perfectly objective with your child and at the same time kindly. You will be utterly ashamed of the mawkish, sentimental way you have been handling it.”20 Watson’s ultimate goal was to do away with parents altogether: he imagined an America in which infants were raised in scientific nurseries, with trained caretakers doling out rewards and punishments in response to the babies’ behavior. While some mothers might protest such a system—isn’t love a natural instinct?—Watson dismissed their concerns. “Only one thing will bring out a love response in the child—stroking and touching its skin, lips, sex organs and the like,” Watson wrote. “This is the clay out of which all love—maternal, paternal, wifely or husbandly—is made. Hard to believe? But true.”21

Watson’s science of parenting no longer seems scientific. Yet, his theories of love continue to influence our lives. They endure as a collection of parental techniques—Watson has been credited with inventing the time-out as a form of punishment22—and as a larger belief that children need boundaries, and not boundless affection. What’s more, the experimental tools that Watson helped invent—his obsession with rat mazes, reinforcement, and quick changes in behavior—are still the central tools of modern psychology. If a phenomenon can’t be quantified or dissected or reduced to a list of molecular ingredients, then we assume it doesn’t exist. The study of the mind remains a study of what can be measured in the lab.

But the real legacy of Watson’s popular science is a belief about human beings. The behaviorist argued that what the poets called love was merely a sentimental lie, used to disguise a more primal set of pleasures. It was time to admit that we are just Darwinian machines, driven by a short list of biological rules and base instincts.23 Life is not romantic. Life is sex and death and survival.

Watson’s cynicism has the tang of truth. It seems like one of those disenchanting facts that science is always discovering: the earth is a lonely speck, floating at the edge of the Milky Way; man is a brute, made of monkey parts; the universe is just dust and old starlight. Maybe love is like that, too—another marvel ruined by too much reality.

But was Watson right? Is love really make-believe? What’s at stake in this debate is nothing less than the nature of human nature. If we are just a bundle of learned habits and selfish genes, a wet computer made of dopamine and instinct, then lovers are fools. Our most intimate relationships are made of flimsy stuff. What’s worse, believing in love is a dangerous illusion, a romantic mistake that leads us to spoil our kids, ruin our marriages, and become the sort of neurotic people who need pills and therapy to cope with existence. We waste our lives chasing a fiction. No wonder we aren’t happy.

Of course, if love is real—if the feeling is more than a cultural trope or a chemical trick—then it remains our great consolation, a source of meaning in a meaningless world. The poets are right. Love is a feeling we can’t live without.
The Young Thief
John Bowlby was born on February 26, 1907, the fourth child and second son of Sir Anthony Bowlby, a baronet and surgeon to King George V. John Bowlby’s childhood was typical of the British upper class. John and his siblings were raised almost entirely by a procession of wet nurses, nannies, and governesses on the top floor of a London town house.24 Every afternoon, the children spent a single hour with their mother; John recalled having to get dressed up in silk shirts and velvet shorts for the occasion.25 Amid all this luxury, what Bowlby remembered most was the loneliness, describing his childhood as leaving him “sufficiently hurt but not sufficiently damaged.”26 During infancy, Bowlby was cared for by a sweet young nursemaid named Minnie. When Bowlby was four, Minnie left the household. He never got over the loss. “For a child to be looked after entirely by a loving nanny and then for her to leave when he is two or three, or even four or five, can be almost as tragic as the loss of a mother,” Bowlby would later write.27

At the age of eight, Bowlby and his older brother were sent to boarding school. It was a miserable experience; Bowlby was desperately homesick.28 Nevertheless, he survived his education and, while an undergraduate at Cambridge University, chose to study medicine like his father.29 But Bowlby wasn’t interested in surgery or the royal court. Instead, he decided to pursue the new field of psychoanalysis, as he’d become convinced that Freudian theory could transform the lives of troubled children. After completing his psychiatric training, Bowlby began working at the Child Guidance Clinic, a mental hospital for youth in North London.30 He cared for children with all sorts of conditions, from hysteria to violence. Bowlby, however, grew most interested in those sent to the clinic for “thievery.”31 (These children had repeatedly been caught stealing or destroying property.) In addition to giving these young thieves a battery of cognitive tests, Bowlby and the social workers asked them questions about their parents and siblings.32 Their stories were heartbreaking: Fred’s mother “shouts and terrifies the children,” while Winnie’s father “often beat her.” Cyril’s mother “openly stated that she wished he had died instead of the baby,” while Kathleen’s mother “had curious sexual ideas about the children and had been seen thrashing the dogs in a sadistic way.”33

These sad stories were not unique to the thieving children. Rather, they were a common theme in the lives of many of the disturbed children at the clinic. But Bowlby would soon identify a variable of childhood that, he believed, was more unique to those who stole. His hypothesis began with a six-year-old named Derek who had been sent to the clinic for stealing and skipping school.34 At first glance, Derek’s childhood appeared perfectly ordinary; his middle-class parents were affectionate and his older brother displayed none of his symptoms. However, Derek’s medical file contained one notable event: when he was eighteen months old, Derek was hospitalized for nine months with diphtheria. He was completely isolated from his family, cut off from everyone he loved. According to Derek’s mother, this separation changed her son. When he returned home, he called her “nurse” and refused to eat. “It seemed like [I was] looking after someone else’s baby,” she said.35

Derek’s story led Bowlby to review the histories of his other thieving patients. What he discovered next would define the rest of his career. According to the case files, approximately 85 percent of “affectionless” children prone to stealing had also suffered, like Derek, from a prolonged separation in early childhood. This became their defining trauma. These kids stole candy and toys and clothes, Bowlby argued, to fill an emotional void. “Behind the mask of indifference,” he wrote, “is bottomless misery.”36

Bowlby was haunted by this apparent connection between separation from loved ones and emotional damage. His study of these young thieves would lead him, in 1939, to oppose Operation Pied Piper, the ambitious attempt to evacuate children from British cities in anticipation of a German bombing campaign. (Over four days in September 1939 nearly 3 million people—most of them children—were put on buses and trains and shipped off to live with strangers in the countryside.)37 In a letter published in the British Medical Journal, Bowlby and his coauthors warned that the noble military exercise came with an unintended cost, as the separation of kids under the age of five from their parents would result in a “very serious and widespread psychological disorder” and a subsequent increase in “juvenile delinquency.”II 38

As the war dragged on, Bowlby followed the reports from wartime orphanages. He spoke often with Anna Freud, Sigmund’s youngest daughter and the head of the Hampstead War Nursery, who described the suffering of the kids in her care. (Anna Freud was also against Operation Pied Piper, writing that “Love for parents is so great that it is a far greater shock for a child to be suddenly separated from its mother than to have a house collapse on top of him.”39) In many instances, the toddlers at the Hampstead War Nursery were simply not able to cope with the sudden absence of their family. Patrick, for instance, was a three-year-old whose mother had to work in a distant munitions factory. The boy was distraught, but he refused to cry because his parents said they wouldn’t visit if he cried. So Patrick constructed an elaborate routine, telling himself over and over that “his mother would come for him, she would put on his overcoat and would take him home with her again.” As the days turned into months, Patrick’s monologue became increasingly detailed and desperate: “She will put on my overcoat and leggings, she will zip up the zipper, she will put on my pixie hat.” When the nursemaids asked Patrick to stop talking, he began mouthing the words silently to himself in the corner.40

These tragic anecdotes made Bowlby determined to conduct his own study on the impact of an extended separation between children and parents. His subjects were patients in the pediatric wards of hospitals. British doctors enforced a strict visitation policy, as frequent family contact was believed to cause infection and emotional neediness. Many London hospitals limited parental visits to a single hour on Sundays, with no visits allowed to children under the age of three.41

Bowlby soon realized that these separations were traumatic, and that the trauma followed a predictable arc, much like the progression of a physical disease. (Bowlby would later compare the damage of separation to a vitamin deficiency, in which the lack of an “essential nutrient” causes permanent harm.)42 When first left alone at the hospital, the children collapsed in tears and wails; they didn’t trust these strangers in white coats. Their violent protest, however, would soon turn into an eerie detachment, especially if the separation lasted for more than a week. Instead of crying, the children appeared withdrawn, resigned, aloof. It was as if they had forgotten about their parents entirely.43 The hospital staff referred to this phase as “the settling down.”44 Bowlby called it despair.45 In an influential 1951 report for the World Health Organization, Bowlby reviewed his hospital data and concluded, contra Watson and the behaviorists, that “the infant and young child should experience a warm, intimate, and continuous relationship with his mother (or permanent mother substitute) in which both find satisfaction and enjoyment.”46

Although Bowlby was convinced by his data—the loneliness of these children left scars—his research was criticized. Many of the critics attacked the nature of his evidence, which they regarded as anecdotal and confused. They complained about Bowlby’s small sample sizes and failure to control for other variables, such as physical illness or nutritional deficits. How could he be so sure that the lack of “mother-love” was causing these behavioral problems? Perhaps these thieves needed more discipline, not affection? The skepticism of love ran deep.

So Bowlby went searching for more evidence. He found it in the work of Harry Harlow, a psychologist at the University of Wisconsin. In the early 1950s, Harlow decided to start a breeding colony of monkeys, as he needed subjects for his research on primate learning. He raised the baby monkeys according to the latest science, feeding them a formula of milk and sugar out of doll bottles. In addition, he gave them a slew of vitamins, antibiotics, and iron supplements.47 To minimize the spread of disease, Harlow kept the animals in individual cages, away from parents and siblings. (He had accidentally created the kind of “baby farm” dreamed of by Watson.) The resulting litter of primates looked bigger and healthier than their peers in the wild.

But the appearance of these young monkeys hid a devastating loneliness. Because their short lives had been defined by total isolation, they proved incapable of even the most basic social interactions. In the company of other primates, they appeared nervous and withdrawn, staring at the floor. “We had created a brooding, not a breeding colony,” Harlow said.48 For the Wisconsin scientists, these troubled primates demonstrated that the developing mind needed more than proper nutrition. But what did it need?

The first clue came from the cloth diapers that had been used to line the cages. Harlow noticed that his monkeys had become obsessed with these rags, clinging to the fabric like a young child clings to a favorite blanket. (The animals would throw “violent temper tantrums” when the cloth pads were removed.) This poignant behavior inspired Harlow to come up with a new experiment. He decided to raise the next generation of baby monkeys with two different pretend mothers. One was a “wire mother,” formed out of metal mesh. (An internal lightbulb provided glimmers of heat.) The second mother was a wooden sculpture, covered in soft rubber sponge and wrapped in terry cloth. In some of these cages, the wire mesh sculpture was fitted with a nipple and feeding tube, while the remaining monkeys were able to feed while cuddling with the soft terry-cloth mother. If the skeptics of love were right, and milk was the cause of the mother-child bond, then the infant monkeys should prefer whichever maternal substitute gave them food.

That wasn’t what happened. It didn’t matter which “mother” held the milk—the babies preferred the one made of rubber sponge and fabric.49 By the age of five months, the monkeys were spending nearly eighteen hours a day nuzzling with their cuddly parent; they would only climb onto the wire mother to eat.50 For Harlow, the lesson was clear: the developing mind desperately craved the pleasures of closeness. “Psychologists, at least psychologists who write textbooks, not only show no interest in the origin of love and affection, but they seem to be unaware of its existence,” Harlow said, in a speech about his monkey experiments to the American Psychological Association.51 But this was a tragic mistake. “If monkeys have taught us anything,” he later wrote, “it’s that you’ve got to learn how to love before you learn how to live.”52

Bowlby grasped the implications of Harlow’s experiments. The monkeys with their wire mesh mothers were like those toddlers alone in a hospital room: What they craved was closeness. Affection. A feeling that could not be measured in ounces or calories, but instead fulfilled a deeper need. So Bowlby concluded that love was not some frivolous luxury—it was part of a larger process that allowed children to cope with a difficult world. He called this process attachment.
The Strange Situation
Mary Ainsworth, a lively midwesterner with a fondness for basketball, board games, and dinner parties, arrived in London in the summer of 1950.53 She was following her husband, Leonard, who had been accepted into a graduate program. Although Mary had worked as a researcher at the University of Toronto—her most recent project focused on Rorschach inkblots—she struggled to find a suitable job in England. After a few months of unemployment, Ainsworth got an interview at a psychiatric hospital, where a doctor was looking for an assistant to help analyze his interviews with children.54 His name was John Bowlby.

Ainsworth got the job. She would spend the next three and a half years working closely with Bowlby, studying the long-term effects of separations at an early age from loved ones.55 They had no shortage of subjects: postwar Britain was overrun with orphans and refugees, the legacy of a violent decade. The interview data was usually heartbreaking—the children’s lives were already marked by absence—but it convinced Ainsworth that Bowlby was right. Attachment is a basic need, writ into our nature.

In 1954, Ainsworth left London behind, following her husband to his new job, this time in Uganda. Although she had no official academic position and little funding, Ainsworth began an ambitious study of infant behavior.56 She recruited twenty-six families with young babies and visited them for two hours every two weeks in their own homes, many of which were made of mud and wattle.57 (To gain access, Ainsworth offered the mothers free rides to a nearby medical clinic and dried skim milk at the wholesale price.)58 Because of her work with Bowlby, Ainsworth was primarily interested in observing the development of the mother-child relationship, but she approached the subject from a different angle. If Bowlby had measured the devastating effects of love’s absence, Ainsworth wanted to watch its ordinary beginning.

Ainsworth’s research method proved essential. An outsider to everything—Ainsworth was a midwesterner in Africa, a childless woman observing children—she watched these mothers without any preconceptions.59 In Ainsworth’s book about her field research, Infancy in Uganda: Infant Care and the Growth of Love, she makes her case not with theory or speculation, but with detailed vignettes of home life. She describes, for instance, an infant named Sulaimani, and the struggles of his mother to care for him:

Sulaimani’s mother was a slip of a girl, still in her teens. This was her first baby, and both she and he were unhappy. She had to do most of the garden work, but had no satisfactory arrangement for Sulaimani’s care while she was gone. He cried so much that his mother was at her wit’s end, and could not behave consistently. Sometimes she was tender and indulgent, and sometimes she was rough and angry in the way she picked him up, slung him over her back, and rocked him. Sometimes, she just let him cry and cry.60

Even as Ainsworth describes clear failures of parenting, she withholds judgment. She is careful to remind herself that these mothers are living arduous lives. If they are inattentive, it’s often because they need to work the fields; if they are anxious, it’s usually because they don’t have enough food to eat. The tragedy is that their child-care issues became a downward spiral. The women with the hardest lives often had the least time to attend to their children, which led to more crying and even more stress.

The Uganda observations were an important extension of Bowlby’s research, but Ainsworth knew that mere observation was insufficient. If she was going to really understand the mother-child bond, then she needed to find a way to calculate its strength. After two years in Africa, Ainsworth followed her husband for the final time, settling in Baltimore. (She would divorce Leonard in 1960.) Ainsworth accepted a position as a lecturer at Johns Hopkins University, working in the same department that Watson had once chaired.

In 1965, Ainsworth and her assistant, Barbara Wittig, pioneered an experiment known as the Strange Situation task.61 The task was a twenty-minute melodrama of science, a scripted play unfolding in eight distinct scenes. In the first scene, a mother and her one-year-old baby are introduced to a new room, filled with toys. Most children soon begin to explore the space, bouncing on the red ball and playing with the Raggedy Andy doll. A few minutes later, a strange woman enters the room and begins talking with the mother. In scene three, the first separation occurs: the mother abruptly exits the room, leaving the child alone with the stranger. A few long minutes unfold; the child is observed through a two-way mirror. Scene five is the first reunion, as the mother returns to the room and the stranger departs. Once the baby is “settled in play”—this usually takes a little while—the mother leaves again. In scene six, the baby is left alone for three long minutes (unless he or she was so distraught the experiment had to be terminated early) before the stranger returns and tries to play with the infant. Then, in the final scene, the mother enters and comforts her child.

Not surprisingly, the Strange Situation experiment was controversial when it was first introduced. Many of Ainsworth’s colleagues thought it needlessly callous, revealing nothing new about the parent-child bond. But Ainsworth was unfazed. She believed that her scripted experiment revealed a basic truth of life. Only in the midst of high drama—when a baby was repeatedly lost and found, abandoned and reunited—was Ainsworth able to learn about our attachment habits.

At first, the results seemed to confirm Bowlby’s basic theory: separation from the mother is stressful.III That’s why a clear majority of one-year-olds wept when left alone and expressed affection after their mother returned. But Ainsworth wasn’t interested in generalities—she wanted to understand every child, even those who violated her theoretical assumptions. This led her to study those outlier infants, the kids who didn’t cry when abandoned or couldn’t stop crying when the experiment was over.62

The problem with outliers is that they’re unusual. As a result, Ainsworth was forced to repeat the experiment again and again; the sound of wailing became the sound track of her lab. It took a few years, but some patterns eventually emerged from the muddle of tears and hugs. While approximately 66 percent of kids engaged in affectionate touches when their mother came back—this was a sign of secure attachment—the remaining 34 percent consistently exhibited a form of behavior Ainsworth labeled “insecure” attachment.63 In most instances, this insecurity manifested itself as indifference, as the children didn’t seem to mind the departure of the mothers; the separation wasn’t that upsetting. (Later studies revealed that these stoic infants were in fact quite stressed, as their heart rate and cortisol levels spiked whenever they were left alone.)64 When the mother returned, these children would turn away, seemingly uninterested in reunion. Ainsworth referred to these babies as having “avoidant” attachments. However, a third of the insecure infants, or 12 percent of the total sample, reacted in the opposite manner. Some of these babies never settled into the new room and refused to play with the unknown toys. Others became distraught after the first separation, clinging to the mother or pushing her away when she returned. (If the mother picked them up, these infants did not “sink in” to her embrace.)65 Ainsworth labeled this category of response “resistant” attachment, since the children seemed to resist the comfort of their caregivers. Instead of being affectionate, the interactions of mother and child had an “unmistakably angry quality.”66

Such variation isn’t surprising. Human nature is a bell curve, every behavior exhibiting a spectrum of differences. But Ainsworth discovered that these differences weren’t random, but instead were closely correlated with parenting style as observed during a series of lengthy home visits conducted every three weeks.67 (“We got to know our families very well,” Ainsworth wrote, noting that this familiarity made it easier for the mothers to forget they were being studied.68) Those infants with secure attachments were far more likely to have mothers who scored high on a sliding scale of “parental sensitivity.”69 These moms engaged in constant baby-talk and expressed more interest in the infant mind. When asked about their child, their answers were far more detailed, entertaining, and emotive. The highly sensitive mother is “able to see things from B’s [the baby’s] point of view,” Ainsworth wrote. “Her perceptions of his signals and communications are not distorted by her own needs and defenses.”70

This doesn’t mean that these sensitive mothers were pushovers, always obeying the whims of their kids. Ainsworth was careful to point out that true sensitivity also included plenty of boundaries; the most effective parents knew when to push back. “When she feels that it is best not to comply with his demands—for example, when he is too excited, over-imperious, or wants something he should not have—she is tactful in acknowledging his communication and in offering an acceptable alternative,” Ainsworth wrote.”71 While Watson and his followers dismissed the “mawkish” love of parents as a dangerous influence, weakening the will of their kids, Ainsworth pointed out that even effective discipline requires warmth. It wasn’t a coincidence that the babies with the most loving mothers spent the least time crying.

Ainsworth’s research is often taken as a guide to parenting, but its real legacy concerns the nature of measurement. John Watson and his followers disregarded love because it couldn’t be quantified. The feeling was just another vague mystery, unfit for a rigorous science. But Ainsworth’s Strange Situation experiment—her short drama of separation and reunion—found a way to test the strength of the loving bond. Although modern psychology had always been obsessed with variables of the individual—intelligence tests, personality quizzes, and so on—Ainsworth’s research documented the stunning influence of relationships. What mattered most, at least for these babies, happened between people.

In her later years, Ainsworth and her colleagues would show that attachment security closely predicted a wide range of seemingly unrelated childhood behaviors. One of her most important findings was the connection between sensitive parents and the willingness of young children to explore novel situations, whether it was a new room or a new toy. Bowlby explained this finding with an elaborate military metaphor, comparing the attachment figure to a “secure base,” which allows an expeditionary force to “press forward and take risks.”72 The same logic applies to parents. Unless children have a secure base at home—a loved one to retreat to in times of stress—they won’t be able to enjoy the world on their own.”73

This finding would later become known as the dependency paradox.74 It’s a paradox because it suggests that true independence requires that we become dependent on someone else. Children don’t explore because they are lacking something essential. They explore because they already have everything they need.IV As John wrote in the Gospels (4:18), “Perfect love casts out fear.”

Like Bowlby before her, Ainsworth faced harsh criticism from her peers. They pilloried the premise of her work—wasn’t it obvious that kids would cry when left alone in a strange place?—and mocked her use of “nonscientific” language, such as “tender” and “sensitive.” (Many of these attacks now seem tainted with misogyny, as if Ainsworth were incapable of looking at love objectively simply because she was a woman with maternal instincts.) The criticism was so fierce that Ainsworth’s studies were routinely rejected during peer review, forcing her to write books about her research instead of publishing articles.75

The scorn fueled Ainsworth’s self-doubts. On good days, she imagined herself, like Bowlby, as one of those stubborn scientists working to shift the paradigm. On bad days, she worried that her critics were right, that love wasn’t a subject fit for science. However, the force of her ideas—her conviction that our attachments mattered, and that how they mattered could be measured—was irrefutable. But it would take another study of infants left alone in a strange room, and another few decades, for that to become clear.
Licking and Grooming
In 1975, Byron Egeland, a psychologist at the University of Minnesota, began studying a group of subjects that had been largely neglected by modern science: pregnant women living in poverty. He signed up 267 expectant mothers at Hennepin County General Hospital in Minneapolis. Their demographic data reads like a laundry list of risk factors for a difficult future. In addition to poverty, many of these mothers were teenagers, lacked education (41 percent had dropped out of high school), had dietary deficiencies (37 percent weren’t getting adequate nutrition), and suffered from a shortage of social support.76 Veronica, for instance, ran away from an abusive home at the age of twelve. A few years later, she became pregnant, probably with her drug dealer’s child. When he was sent to prison, Veronica and her son, Thomas, ended up living in a shelter.77 She suffered from depression and addiction.

Egeland’s study began with limited aims: he wanted to identify the variables that predicted the “mistreatment” of children, so that social workers could offer counseling before any abuse occurred. However, the scientists were soon impressed, like Ainsworth in Uganda, by the sheer variation of parenting behavior. Some of these new mothers were attentive and supportive, even amid adversity. Others struggled to hold it together during a tantrum; Veronica, for instance, was often “harshly rejecting” toward her son.78

To better understand the impact of all this variation, Egeland teamed up with Alan Sroufe, a Minnesota psychologist and early supporter of attachment theory. After testing every infant with Ainsworth’s Strange Situation task at the age of twelve months—the initial results confirmed the correlation between sensitive parenting and attachment security—Egeland and Sroufe decided to turn their short-term risk-assessment study into a longitudinal project that would eventually become epic in scale, spanning generations and decades. “We didn’t really know what we were getting into,” Sroufe remembers. “But once you start seeing these patterns emerge, and you see the continuities in behavior, you don’t want to stop.”79

The first ambitious follow-up occurred when the children were four and five years old. To capture the complexities of early childhood, Egeland and Sroufe decided to start their own nursery school, which was offered free to forty of the children in the study. In the classrooms, a team of twenty trained observers kept detailed notes on every child’s behavior. In addition, the scientists recorded hundreds of hours of videotape, which were later coded and analyzed.80

The results were convincing. In the nursery school, children with secure attachment histories were rated as more independent, sociable, and popular than their less secure peers. They were far less likely to bully or get bullied. They exhibited more self-control, scored higher on measurements of intelligence, self-esteem, and resilience,81 and displayed more empathy for other children.82 (All ratings were done blindly, as the observers had no knowledge of how the subjects had been categorized as infants.) In the laboratory Barrier Box task, for instance, a child was introduced to a collection of toys, such as LEGOs, superhero figurines, and dolls. After a few minutes of free play, a scientist came along and told the children that the toys they’d been enjoying belonged in the other room. If the children wanted to keep playing, they had to get a new set of toys out of a clear plastic box that was virtually impossible to open. It was a deliberately frustrating task. However, the way these children coped with their frustration was revealing. While most preschoolers struggled with the task, either quickly giving up or resorting to brute force and displays of anger, some kids excelled at the Barrier Box. They never managed to open it, but they remained focused and persistent, patiently trying out a variety of strategies. After ten minutes, the scientists ended the experiment, opened the box, and let every child play with the toys.

What explained these differences in performance? The best answer was early attachment. Approximately 40 percent of securely attached children were given the highest ratings on the Barrier Box task, while not a single preschooler with an insecure attachment history was so rated. What’s more, insecurely attached kids accounted for 75 percent of the subjects given the lowest ratings. Although the children had been given a challenge that had nothing to do with their closest family relationships, the legacy of those first attachments shadowed their behavior. The variable couldn’t be escaped.83

Five years later. The children were invited to spend four weeks at a summer camp on the University of Minnesota campus. Like the preschool, this immersive setup allowed the scientists to gather an unprecedented amount of data as they watched the children play soccer and softball, swim in the pool, and work together on art projects. Once again, the results were a stark reminder that attachment has lasting consequences. Children with secure attachments to their parents generally showed much higher levels of “social competence,” better able to develop and maintain relationships with other kids at the camp. As a result, they spent 40 percent more time with their friends. When the ten-year-olds were given challenging tasks by the scientists, such as having to navigate an obstacle course, those with “secure histories organized themselves in more effective ways, avoided scapegoating, and performed dramatically better.”84

As Egeland and Sroufe followed the children into adolescence, they uncovered a most unexpected finding: the connection between infant attachment and the behavior of the children was getting stronger. The teenagers were even more influenced by their earliest relationships than when they were five or ten years old. “Isn’t that remarkable?” Sroufe says. “These kids often have the physical characteristics of adults, and yet what we found is that so much of what they are doing was correlated with these attachment measures from when they were twelve months old.” The Minnesota scientists found that adolescents with secure infant attachments also performed better in high school, with stronger attachments leading to higher standardized-test scores. (Attachment security was also inversely correlated with disciplinary problems.) The presence of a supportive and sensitive parent before the age of three and a half was better than IQ scores at predicting whether the children would graduate from high school. 85

Why are teenagers so shaped by their attachment history? Like infants, adolescents are also struggling to form close relationships as they spend increasing amounts of time with their friends. “They’re [teenagers] looking for a kind of intimacy,” Sroufe says. “But if you’re going to get that intimacy, then you need to trust people. You need to tell them how you feel. You need to open up. And that builds on emotional skills that are very much connected to early attachment experience.”86 Just as babies must learn how to express their vulnerability—that usually means crying for an attachment figure—so do teenagers have to find ways to reach out and make connections with their peers. Being vulnerable is not a sign of weakness. It’s how we let people in.

The Minnesota subjects are now turning forty, with families and children of their own. Nevertheless, the correlations continue; love echoes across the generations.87 The scientists have found that attachment quality in the first year of life helps predict adult health, as “resistant” infants—those babies who refused to be comforted by their mothers upon their return, even though they cried in her absence—were 2.85 times more likely to report a chronic illness at the age of thirty-two when compared to those who were securely attached.V 88 Or consider a recent paper led by the psychologist Lee Raby, which found that the quality of parenting during childhood was correlated with the romantic behavior of the Minnesota subjects more than thirty years later. In particular, those with less sensitive mothers exhibited a spike in skin reactivity when having discussions about a source of conflict with their partners and spouses. One explanation for these changes in skin reactivity is that subjects are holding back their feelings, as increased reactivity is associated with fear and emotional inhibition. (Lie detectors work on this principle.) Because these adults had parents who struggled to respond to their emotional needs, they learned to hide their worries.89 Intimacy requires candor, a willingness to show your cracks, but that can be hard for people who had insecure childhood attachments.

As the years pass, this inability to discuss relationship issues can undo the relationship, which is why those with less secure attachments as children also experienced shorter and less satisfying adult attachments.90 In a separate analysis, the Minnesota researchers looked at old videotapes of the subjects at the age of two, as the toddlers attempted to learn a new skill from their mothers. If this interaction was judged unsuccessful—the mother wasn’t supportive enough, or the child refused help—the subjects were also more likely to get described as a “weak link” by their adult partners. Thirty years had passed, but they still struggled to get close to other people.91

In their descriptions of the study, the Minnesota scientists repeatedly return to the story of a boy named Tony, whose complicated life reflects the complicated themes of the research. In 1977, when the Minnesota scientists first tested him on the Strange Situation task, Tony appeared securely attached, the happy infant of loving parents. He excelled in preschool and scored well on nearly every measurement of early childhood, from the Barrier Box to literacy skills. But then, when Tony was six, his parents went through a bitter divorce. He rarely saw his father. When he was thirteen, Tony’s mother died in a car accident. His father then decided to move to another state, taking Tony’s two siblings with him. Tony was forced to live with his elderly aunt and uncle.92

Given this tragic series of events, it’s not surprising that Tony spiraled downward in his teenage years. He failed several classes at school, had few stable friendships, and oversaw a burglary ring in an “administrative capacity.” (Like Bowlby’s young thieves, Tony tried to appease his sadness with stolen things.) After interviewing Tony at the age of fifteen, the psychologists described his fall: “The light inside of him seemed to have gone out. He was visibly depressed and isolated.”

But Tony’s tale doesn’t end there. In his early twenties, he met a woman at the local community college. She was attracted “to his quietness and tender heart.” They got married a few years later and had a daughter. When the scientists watched Tony with his toddler, they were struck by his devotion, how he hadn’t forgotten the lessons of early attachment. “An extraordinarily supportive father,” wrote the scientists. “He was patient, involved, warm and available, providing the structure, limits, and encouragement the child needed.” Although Tony had struggled as a teenager to talk about his mother’s death—he insisted that the loss “didn’t mean that much”—he was now able to share his feelings and speculate on how her passing had shaped his life.93 In The Development of the Person, the Minnesota scientists compare our early attachment experiences to the foundation of a house. While the foundation is not sufficient for shelter—you also need solid beams and a sturdy roof—they note that “a house cannot be stronger than its foundation.” That’s what we get as young children: a foundation for attachment. The beginnings of a structure on which everything else is built.94

Such stories are just stories; nothing is proved when the n is 1. We will never know if Tony’s securely attached childhood enabled his adult recovery or if his experience of early love made his later resilience possible. It’s important to remember that these correlations of personality are not certainties: many insecure babies become supportive spouses, just as many secure children drop out of high school and struggle later in life. As Bowlby first pointed out, the human attachment system is responsive to changing conditions. That means we can still learn how to love, even if our childhood was marked by loss and insecurity. Attachment is not a fixed state, or a permanent diagnosis. It is a continual process, a working model of relationships that can always be revised.

The Minnesota study looks at life with a telescope. By tracking its subjects for years, it illuminates our human arcs, the subtle patterns that underpin our lives. But some questions require microscopes. If we want to understand how love changes us—why it makes us tougher, more resilient, less likely to break—then we need to zoom in. Because the feeling is not an airy nothing. Love literally marks the mind.

Michael Meaney, a neuroscientist at McGill University, began studying attachment because of an accidental observation. After he returned rat pups to their Plexiglas cages, Meaney noticed that some rodent mothers were much quicker to comfort their offspring than others, licking and grooming the young animal until its pulse returned to baseline. Meaney became interested in this maternal variation, which led him and his graduate students to spend eight hours a day closely observing the interactions of rat families. Sure enough, some mothers spent far more time (about 50 percent more) licking and grooming their offspring. When the pups were a hundred days old—that’s late adolescence in rat years—they were run through a battery of stress and intelligence tests.95

The results were stunning. One of Meaney’s main behavioral measures is known as the open-field test. It’s a simple enough design: A rat is placed in a round box for five minutes. Nervous animals tend to be wallflowers, clinging to the edge of the box like teenagers at a school dance. Less anxious rats tend to explore their surroundings, wandering into the center of the enclosure in search of food. According to Meaney’s data, those pups born to the most comforting mothers—the high-LG [licking/grooming] animals—spent thirty-five seconds, on average, in the center of the box. The low-LG pups, on the other hand, spent less than five seconds there.96 Similar patterns appeared across a wide range of assays and measurements. High-LG were less aggressive with their peers.97 They released fewer stress hormones when restrained.98 They solved mazes more quickly.99 They were better at learning from their littermates.100 They were more nurturing with their own pups.

In recent years, Meaney has shown how these feelings of affection alter the brain. High-LG rat pups have fewer receptors for stress hormone and more receptors for the chemicals that attenuate the stress response; they show less activity in parts of the cortex, such as the amygdala, closely associated with fear and anxiety; they exhibit more synaptic growth in the hippocampus, a part of the cortex associated with learning and memory;101 even their DNA is read differently as all that maternal care activates a genetic switch that protects rats against chronic stress.102 These neural and genetic differences suggest that pups raised by the most comforting mothers are better able to handle the upheavals of life, whether it’s a scary new cage or a strange new scientist.

The same lesson applies to people. A natural experiment took place during World War II, when more than seventy thousand young Finnish children were evacuated to temporary foster homes in Sweden and Denmark.103 For the kids who stayed behind in Finland, life was filled with moments of acute stress—regular air bombardments and invasions by the Russians and the Germans. But for those sent away, the stress of being separated from their parents was unceasing. They lacked what they needed most.

This early shock had lifelong consequences. A 2009 study found that Finnish adults who had been sent away from their parents between 1939 and 1944 showed an 86 percent increase in deaths due to cardiovascular illness compared to those who had stayed at home.104 Although more than sixty years had passed since the war, these temporary orphans were also significantly more likely to have high blood pressure and type 2 diabetes. Other studies have documented elevated levels of stress hormone105 and increased risk of severe depressive symptoms among the wartime evacuees.106

The feeling of love is not just a source of pleasure. It’s also a kind of protection.107
Adaptation to Life
In the late 1930s, Arlie Bock, a professor of hygiene and the head of Health Services at Harvard University, received funding from the department-store magnate W. T. Grant for a long-term study of healthy undergraduates. Science was too focused on sickness, Bock said. It was time to “make a systematic inquiry into the kinds of people who are well and do well.”108 Grant agreed and hoped that the research would help him better identify those who would be successful store managers.

The project began with the Harvard College class of 1939 and continued to recruit normal undergraduates—men “able to paddle their own canoe,” Bock wrote109—for the next five years. Bock ended up with a cohort of 268 Harvard students.110 The subjects were run through a gauntlet of medical and psychiatric tests, from handwriting analysis to inkblot cards. They were asked about their personal and medical histories—When did you stop wetting the bed? How much sugar do you put in your coffee? Do you masturbate?111—and given a litany of intelligence tests, most borrowed from the army. The staff doctors measured the brow ridges on their foreheads, the circumference of their chests, and the hanging length of their scrotums.112 They were asked to make sense of inkblots and sprint for five minutes on a treadmill.

The results were not interesting. Although Bock and his colleagues hoped that their medical measurements would predict life outcomes, that didn’t happen. The “masculinity” of body types bore no relationship to rank in the military during World War II, nor did their responses to the inkblots predict anything about their sex lives. Forehead structure wasn’t linked to intelligence, and intelligence wasn’t significantly linked to income.113

Despite these failures, the Grant Study—its official name is the Harvard Study of Adult Development—continued to track the men into middle age. The subjects were asked about their drinking habits, political beliefs, and favorite sports to play. They were interviewed about their experience in the war and given a long list of questions about smoking and cigarettes. (After W. T. Grant withdrew funding, the study was supported in part by Philip Morris.)114 Given the privileged backgrounds of the Harvard men, it’s not surprising that many of them became eminent and successful. One became a governor; four ran for the US Senate; another became the president. (While almost all of the Grant subjects remain anonymous, John F. Kennedy’s file has been sealed until 2040, thus forcing the scientists to confirm his involvement.)115 As the years went by, the Harvard study remained most notable for what it could not find. The scientists had amassed so much detailed information—each man had his own folder, as thick as a dictionary116—but it all proved useless, at least when it came to making sense of adult outcomes. The expected associations never appeared; the secrets of the good life remained a secret.

Most worrisome, perhaps, was that the fundamental assumption of the project—it was supposed to be an investigation of health, not sickness—was slowly being dismantled. By the time the Harvard men turned fifty, nearly 30 percent of them had suffered from mental illness, such as alcoholism and manic depression.117 (Three percent of them had been hospitalized for their psychiatric problems.)118 “They must have been spoiled by you psychiatrists,” Bock would later complain. “They didn’t have problems like that when I was running the study!”119

If not for a young doctor named George Eman Vaillant, the Grant Study may have remained an expensive failure. It would be remembered, to the extent anyone remembered it, as a testament to the peculiar theories of midcentury psychology. But Vaillant realized that this longitudinal project had vast potential if the right questions were asked. Unlike his predecessors, Vaillant didn’t believe that measures of chest circumference would ever predict life outcomes, or that such a thing as “the normal person” even existed. “I had gone to Harvard myself,” Vaillant says, “and I knew that even Harvard men had their issues.”120

I met George Vaillant on a hot summer day in Orange, California, where he lives with his current wife, who is also a Harvard-trained psychiatrist. Their gated clapboard house is on a quiet suburban street, situated among towering trees that were planted before there were any houses, when the neighborhood was a private park surrounded by citrus groves. After the Great Depression, the park was sold off to local doctors, who built these estates.

Vaillant suggested that we sit on the front porch and enjoy the breeze and the birds. But then the breeze disappeared and the only birds around were crows, fighting over the treetops. After a few minutes, both of us were sweating through our shirts, water glasses already empty.

Vaillant is a spry eighty-year-old. His gray hair is fine and parted, tamed by decades of being combed in the exact same manner. He talks slowly, always pausing thoughtfully in midsentence, even when answering questions he’s probably answered a hundred times before. When Vaillant slips into a soliloquy—and even straightforward questions have a way of turning into digressive lectures on Freud and fMRI research—he tends to close his eyes and run his fingers delicately over his eyelids. It almost looks as if he were praying.

The oldest son of a banker’s daughter, Vaillant grew up in the Great Depression in a wealthy family. His mother, he says ruefully, “came from the John Watson school of parenting,” leaving young George starved for affection. When he was ten years old, his father walked into the backyard of their mansion in Chester County, Pennsylvania, and shot himself in the head with a revolver. George was the last person to see him alive.

The suicide defined the rest of Vaillant’s childhood. His mother took the kids away to Arizona; they didn’t even stay for the funeral. A few years later, Vaillant was sent to boarding school in New England. He went on to Harvard College, where he studied literature. He stayed in Cambridge for medical school—“I had a vague desire to help people, which I guess is why I became a doctor,” Vaillant says—and chose to specialize in psychiatry. When I ask him why, he holds out his trembling hands. “Look at the shake. I could never hold a knife.”

As a young resident, Vaillant became interested in schizophrenic patients who no longer displayed symptoms of their illness. “Remember, schizophrenics were supposed to be incurable,” Vaillant says. “They were never supposed to stop hearing voices.” To understand their recovery, Vaillant began paying careful attention to the extended case histories of his patients, noticing all the ways in which their symptoms were shifted by circumstance, treatment, and even personal relationships. “This probably sounds like a very obvious idea, but I was impressed by the notion that if you wanted to understand mental health, then you needed to follow people for years and years. You couldn’t just take a slice of a life and say you understood. Understanding takes time.”

It was at this point in his career that Vaillant started working on the Grant Study. While he was enthralled with its longitudinal approach—“To be able to study lives in such depth, over so many decades . . . was like looking through the Mount Palomar telescope,” he said in a 2009 interview in the Atlantic with Joshua Wolf Shenk—Vaillant was disappointed with his subjects, at least at first. “I didn’t particularly want to work on healthy people,” he says. “Normal seemed boring.” However, once Vaillant read the men’s case files, he realized that their upper-class lives hid a vast amount of angst and illness, just like his own. These subjects had been selected because they seemed so healthy and fortunate—they were the most privileged young men in the most privileged country on earth—but Vaillant came to see that none of them were living happily ever after. “Their files felt to me like a Tolstoy story or a Eugene O’Neill play,” he says. “They were filled with drama. I couldn’t stop reading them.”

Vaillant’s literary allusions are not an accident. The Grant Study began as an attempt to fit the messiness of life into a neat model, to find the biological variables that predicted health, wealth, and happiness. But the initial failures of the study to predict anything led Vaillant to conclude that a new approach was needed. “You can’t fit a man on an IBM [punch] card,” Vaillant says, referring to the first form of data storage used by the Grant scientists. Instead of trying to quantify the lives of his subjects, measuring their skulls and blood pressure and scrotums, Vaillant treated them like characters in a complicated novel. He wanted to hear their stories, which was why he started his interviews with a long list of open-ended questions. He asked about their wives and mistresses, their kids and their cocktails, what they did for fun and how they coped with despair. Vaillant was a good listener. The conversations were his best data.

These interviews transformed the Grant Study. If the early Grant follow-ups had been modeled on a routine physical exam—the body was supposed to be destiny—Vaillant turned it into a therapy session. For the most part, the men responded with candor, grateful for the opportunity to unburden themselves for science. Their answers confirmed Vaillant’s suspicions, as the middle-aged subjects frequently confessed that all their money and success still left them struggling for meaning and happiness. In one of his first extended write-ups of the Grant Study, Vaillant approvingly quotes the conclusions of another longitudinal project. “No especially blessed individual turned up in this assessment,” the scientists wrote. “The luckiest of the lives here studied had its full share of difficulty and private despair.”121

This depressing truth—everyone struggles, everyone suffers—led to Vaillant’s first revelation, which is that our mental health is defined by how we cope. The body responds to injury by relying on a raft of automatic defenses—when we bleed, the blood clots; when we are cut, the skin scars. Vaillant believed that the mind had its own mechanisms of protection, especially when it came to handling stress and trauma. In Freudian theory, these mechanisms are known as adaptations. Some adaptations are psychotic—we might become paranoid or start hallucinating—while others are immature and neurotic and manifest themselves as hypochondria and addiction. The healthiest defenses, however, are those Vaillant places in the “mature” category, such as humor, sublimation, and altruism.122 Instead of drowning our sorrows in whiskey, we cheer ourselves up by helping others or writing a poem about the sadness. “The basic distinction is quite straightforward,” Vaillant says. “The mature defenses are all about the other. They help you help other people. The immature defenses, in contrast, might make you happy in the moment, but they totally screw up your life and relationships.” Vaillant then pauses, as if he’s about to deliver a line he knows I’ll like: “The essence of love is in realizing that someone matters more than you. In the short run, that’s the hard part. In the long run, that’s the fun part.”

This theory of mental life raises an obvious question: If psychiatric health is simply the process of adaptation, what determines our adaptive style? Why do some people get back pain while others make art? Why do some people flee into affairs while others find solace in their marriage? Here Vaillant’s research has been most revealing, providing yet another layer of evidence in support of attachment theory. While Freud argued that our defense mechanisms were shaped by the sexual tensions of childhood, Vaillant has come to believe that they’re actually determined by our relationships with other people. In particular, Vaillant says, it is the experience of loving and being loved that most closely predicts how we react to the hardships of life; human attachments are the ultimate source of resilience. “The seventy-five years and twenty million dollars expended on the Grant Study points, at least to me, to a straightforward five-word conclusion,” Vaillant writes. “ ‘Happiness equals love. Full stop.’ ”VI 123

The line seems too romantic to be true. Yet, Vaillant insists that the numbers can’t be denied. When he began working on the Grant Study, Vaillant was barely familiar with the research of Bowlby and Ainsworth. “I thought of him [Bowlby] as someone who studied juvenile delinquents,” Vaillant says, “and it wasn’t until I was asked to write a short biography of Bowlby [for the American Journal of Psychiatry] that I discovered that he was big on the same things that I was big on. And I think what we both came to see is that you really can’t do anything without love.”

The power of love begins at the beginning, just as Bowlby and Ainsworth suspected. Based on the initial interviews with the Harvard men, Vaillant assessed the warmth of their early childhood. Did the subjects feel loved by their mother and father? How often did they eat family meals together? Did they stay in close contact with their brothers and sisters? The answers to these questions turned out to have sweeping consequences, even decades later. When Vaillant compared the lives of men in the Cherished category—those with the most secure attachments in early life—to those in the Loveless category, he found that the Loveless were three times more likely to have been diagnosed with a mental illness, five times more likely to be “unusually anxious,” and four times less likely to rely on mature adaptations when dealing with adversity.124 A third of the men without a warm maternal relationship suffered from dementia in old age, a rate 2.5 times higher than those with a warm relationship.125 (Interestingly, a loving mother was far more protective against memory loss than “vascular risk factors,” such as low cholesterol.)VII The variation in childhood warmth even predicted success in the workplace, as subjects from the most loving homes earned 50 percent more money over their careers than those from the bleakest homes.126 Early attachment is more predictive of achievement than any other variable measured in the Grant Study, including IQ scores.127

But love doesn’t only matter at the start of life; the need for attachment is not just a phase of development. As the men entered their fifties, Vaillant increasingly focused his interviews on their closest relationships. (Vaillant likes to compare longitudinal studies to good wines—they tend to get better with age.) He asked about their marital lives and “patterns of entertaining”; about their oldest friends and their parenting habits. This mass of biographical data allowed Vaillant to rank the Grant subjects based on the quality of their relationships in middle age. According to the data, nothing counts more than our attachments. Nothing even comes close. The loneliest subjects in the Grant study were ten times more likely to suffer from a chronic illness before the age of fifty-two, five times more likely to be diagnosed with a mental illness, and eight times more likely to rely on immature defense mechanisms. These men often feigned self-reliance, pretending not to need other people, but Vaillant discovered that they actually lived life in terror. They were three times as likely to become heavy users of alcohol and tranquilizers.128

Such numbers are mere distillates, a scientific summary of anonymous men. But there is still value in knowing the individual stories. (As Vaillant notes, “The Old Testament isn’t around today because it was full of useful statistics.”) While Vaillant’s writing is dense with correlations, it’s also filled with compelling case studies. In these miniature biographies the sweep of the Grant Study becomes clear. “I’ve come to know these men,” Vaillant says. “I’ve eaten at their tables. I’ve met their kids. I’ve been asking them questions for forty years. And you know what? They’re still surprising me.”

Consider the sad tale of Oliver Kane, a subject in the Grant Study. Kane’s childhood was framed by loss. His father died when he was one, followed by the death of his mother fourteen years later. Nevertheless, Kane relied on his unusual intelligence—Vaillant describes him as “perhaps the most intelligent man in the study”—to succeed at Harvard and embark on a lucrative career as a management consultant. (In the early 1960s, Kane was making more than $70,000 a year, which is equivalent to more than half a million dollars today.) But the money proved meaningless.129 Instead of buying a home, Kane chose to live in residential hotels, using a men’s club as his mailing address. He traveled constantly for work—his final interview with Vaillant took place in the Admirals lounge at O’Hare Airport—and he made few lasting relationships.130 “I came to see him as a man who tried to live without love,” Vaillant says. “Whether that was because of his shitty childhood, or because he just didn’t want to be bothered, I don’t know. But he was very much an experiment.”

The experiment had a terrible ending. What Kane learned—a lesson that came too late—is that success means nothing if it’s not shared. When he was fifty years old, Kane flew his small plane into a mountain. The authorities couldn’t say for certain if it was suicide, but Vaillant notes that Kane had spent the days before the crash revising his will. And then there’s his last note to the Grant Study, written a year before Kane’s death. The words are steeped in despair: “Ironically, as I have acquired more external success, I have more and more doubts that I have chosen a way of life that really means anything.”131

In his writings, Vaillant counterposes the tragic life of Oliver Kane with the story of Anna, a woman in the “Genetic Studies of Genius.” Begun in 1921 by the psychologist Lewis Terman, the study attempted to track the impact of intelligence in the classroom. It focused on outliers and included 1,528 children who scored in the top few percent on the IQ test. Although the initial plan was to end the longitudinal project after a few years, Terman’s subjects enthralled him, and he continued to follow them for the rest of his life.

When the “Termites” were in their late seventies, Vaillant asked for permission to interview ninety of the women. He wanted to see if the patterns he’d observed among the Harvard men applied to both genders. That’s when he met Anna, a woman with “short gray hair, dimming vision, too much weight, and arthritic hands.”132 Born into a poor family in rural Colorado, Anna spent her career as a math teacher in a public high school. She enjoyed teaching, but struggled to balance the demands of her job with the needs of her four children. (Anna’s husband had been incapacitated by recurrent bouts of illness, which meant that the family largely relied on her income.) Most evenings, Anna would put the kids to bed, make a big pot of coffee, and then grade geometry papers until her eyes gave out. When Vaillant interviewed Anna, she was living alone in a small apartment; her husband had died a few years before. “You might look at her and think, ‘She’s had a rough go of it,’ ” Vaillant says.

But Anna was not sad or defined by her regrets. She insisted on telling Vaillant the happy story of her forty-year marriage, which had only grown “closer and closer” over time. She talked about the consolations of faith and the pleasures of driving her bright yellow Volkswagen Rabbit to the grocery store. She kept circling back to her children and grandchildren, pointing to their framed pictures. She bragged about their accomplishments before catching herself: “That’s pride. And the church says we’re not supposed to be proud.” But then, Vaillant writes, “she would go right back to being proud.”133

What can be learned from all these archived lives? When asked to describe the single most important finding of his longitudinal study, Vaillant focuses on the crucial importance of our closest relationships.134 To make his case, Vaillant then describes a study he oversaw on the relative happiness of lottery winners and paraplegics. The newly rich, it turned out, weren’t any happier than the paralyzed. Everyone eventually habituated to their condition; money was the most fleeting reward. Instead, Vaillant found that as the Grant subjects entered their late eighties—nearly half of them were still alive—only one variable predicted their life satisfaction: the capacity for loving relationships. “I wrote once that when we are old our lives become the sum of everyone we have loved,” Vaillant says. “That’s still true. I believe it more than ever.”

Vaillant struggled to live out that belief. After having four children with his first wife, Vaillant filed for divorce in 1970. He soon married again, had another child, and then, twenty years later, left his second wife for a colleague. Five years later, Vaillant asked his second wife to take him back. She did, but it didn’t work out. Vaillant has since married for the fifth time, to the psychiatrist Diane Highum. In old interviews, Vaillant compared himself to King Lear, a distant father at the head of a dysfunctional family.135 (Several of his children have been estranged from him for long periods.) While Vaillant’s personal difficulties could be seen as undermining his scientific message—how can he preach about the importance of love when he can’t even stay married?—I’ve come to believe that, in a strange way, his flaws have drawn him even deeper into the work. He studied the Grant men to see where he’s been, but also to know where he’s going; to understand what he has, and also what he lacks. “Use this gently,” he tells me. “But I’ve often thought that writing Adaptation [Adaptation to Life was Vaillant’s first book about the Grant subjects] was a plea for help. That I was saying, ‘Won’t someone please be attached to me?’ I guess you could say that I learned about the importance of attachment in my own life by watching it unfold in my subjects.”

It’s such a melancholy thought—that this affirming research has been a cry for closeness, that Vaillant only discovered the importance of love from afar, as a statistical proof, distilled from the lives of other people. And now the Grant men are dying and will soon all be gone. I wonder what Vaillant will do then. It must have been hard to lead a study when the results put your own life in such stark relief, when the wisdom you speak is what you’ve always sought, but never found.

At least, never found until he neared the end of his life. Two years after our first interview, I contacted Vaillant again. I wanted to make sure I’d got things right. He read a draft of this chapter and said he had a “critical coda” to add to my description of his life and work. Over the last few years, Vaillant’s wife had “befriended” all of his children and helped him reconnect with them. The man who had once compared himself to Lear was now the patriarch of a “solid family unit.” To celebrate Vaillant’s eightieth birthday, the entire extended family—all fifteen of them, including his kids, grandkids, first wife, current wife, and stepdaughter—went on a five-day cruise to Bermuda. I ask Vaillant what it feels like to have finally repaired his closest relationships. “I find the language of attachment theory to be quite apt,” he says. “It provides me with a sense of security. Tranquillity. Peace, even.”

As Vaillant notes, this unexpected turn is echoed in the lives of some of his scientific subjects. In his final account of the Grant men, Vaillant describes a number of them who were saved by their later marriages. These men had struggled, like Vaillant, to form lasting relationships with family and friends; they had wrestled with intimacy and lost. But then, often because they found a stubborn partner, a wife who refused to accept their failings, they discovered in old age the pleasures of secure attachment. “She [Diane, his wife] has made this possible,” Vaillant says. “Her kindness, her unconditional positive regard. It’s because of her.”

Such is the nature of human salvation. Vaillant always knew the correlations. He has written so eloquently about the sway of love. But the words were never enough; the numbers didn’t change his life. It took another person to do that.

Our last conversation is brief; Vaillant seems eager to get off the phone. I’m happy for his happiness, even if it’s arrived in his last act. Because it’s never too late—love never loses the capacity to transform what it touches. The most recent data from the Grant study makes this clear. Among the men who never formed intimate relationships, roughly 13 percent have survived into their late eighties and early nineties. However, among subjects who were better at attachment, the survival rate is closer to 40 percent.136 “The capacity for love turns out to be a great predictor of mortality,” Vaillant tells me. “I cannot tell you why that is. I can only tell you that it’s so.”

And so the science has come full circle: the feeling that once seemed so immeasurable has, in the lives of all these old men, become the only thing worth measuring.

Even John Watson would come to understand the power of love. In 1935, Watson’s young wife, Rosalie, died of pneumonia. The loss devastated Watson. As his son James would remember, Watson spent the night weeping; it was the only time his children ever saw him cry.137 Although he couldn’t bear to tell his sons about their mother—they learned she was dead from the cook—Watson stood in the doorway and put his arms lightly around their shoulders. “That moment was the one occasion in which Watson’s children experienced a genuine expression of intimacy and affection from their father,” writes Kerry Buckley, Watson’s biographer.138

After Rosalie’s death, Watson began drinking heavily, consuming a quart of whiskey every day according to a friend. He worked as an advertising executive for another decade—he never lost his talent for sales—before moving to a vast estate in the hills of western Connecticut. He took care of the apple trees and the dogs, but rarely took visitors. He refused to talk about his dead wife, not even with his sons.139

In 1957, the American Psychological Association chose to honor Watson at its annual meeting for sparking a “revolution in psychological thought.” Although Watson hadn’t worked in a science lab in decades, his theories remained influential. (Bowlby was still little known outside the field of child psychiatry.) Watson had driven down to New York City to receive the award, but at the last minute refused to enter the ballroom. He was afraid that he would cry onstage, that the “apostle of behavior control” would not be able to handle his own feelings.140

The following year, Watson’s health began to deteriorate. In the days before his death, Watson gathered his strength for one final act. He collected all of his papers—a lifetime of manuscripts, letters, and research notes—and carried them to the fireplace, where he began throwing them into the flames. His secretary asked him what he was doing. Watson’s reply was cryptic: “When you’re dead, you’re all dead.” And then he turned back to the fire and watched his work burn.141

I. Watson’s first job after quitting academia was in advertising, as he sought to apply the techniques of behaviorism to the marketing of consumer products. It was a lucrative shift; Watson was good at selling things. In a campaign for Maxwell House, he coined the term coffee break; he created an early infomercial for Pebeco toothpaste; he even convinced the queen of Spain to endorse Pond’s facial cream. “It can be just as thrilling to watch the growth of a sales curve of a new product as to watch the learning curve of animals or men,” Watson later wrote.

II. Bowlby was also moved by stories from foundling hospitals, institutions established in the nineteenth century to care for orphaned children. Although these hospitals supplied infants with adequate nutrition, they struggled to keep them alive. A 1915 review of ten infant foundling hospitals in the eastern United States, for instance, concluded that up to 75 percent of the children died before their second birthday. The best hospital in the study had a 31.7 percent mortality rate. Robert Karen, Becoming Attached (Oxford: Oxford University Press, 1994), 19.

III. One of the persistent criticisms of attachment theory is that, at least in the early decades, researchers focused solely on the role of the mother, making it seem as if she were the only parent who mattered. This is a mistake that Bowlby, toward the end of his career, took pains to correct. “Looking after babies and young children is no job for a single person,” Bowlby declared in a 1980 lecture. “If the job is to be well done and the child’s principal caregiver is not to be too exhausted, the caregiver herself (or himself) needs a great deal of assistance. . . . In most societies throughout the world these facts have been, and still are, taken for granted and the society organized accordingly. Paradoxically it has taken the world’s richest societies to ignore these basic facts.”

IV. In his early writings on separation, Bowlby imagined the child mind as a homeostatic system, wired to seek out a balance between two opposed goals: safety and learning. His thinking on the subject was strongly influenced by World War II military technology, which used rudimentary computers to control the behavior of machines. By monitoring changes in the external environment, these “cybernetic” devices were able to adjust their own responses. This new approach helped lead to the invention of “smart bombs”—they altered their own trajectory based on radio signals—and gun mounts that automatically focused on the target. Phillip Shaver and Mario Mikulincer, Attachment in Adulthood: Structure, Dynamics and Change (New York: Guilford Press, 2007), 219.

V. The precise mechanisms behind this dramatic increase in illness remain unclear. One likely possibility is that troubled attachment styles in childhood increase the incidence of psychosomatic illness. For instance, a 2006 study led by Robert Waldinger at Harvard Medical School found that middle-aged adults who experience insecure attachments in early life are far more likely to experience “somatization,” which is the tendency to experience bodily symptoms in response to psychological stressors. Robert J. Waldinger, Marc S. Schulz, Arthur J. Barsky, and David K. Ahern, “Mapping the road from childhood trauma to adult somatization: The role of attachment,” Psychosomatic Medicine 68.1 (2006): 129–35.

VI. As Vaillant notes, Virgil came to a similar conclusion a few thousand years ago, when he declared, “Omnia vincitamor,” or love conquers all. Unfortunately, Virgil “had no data” to back up his poetry. George Vaillant, Triumphs of Experience (Cambridge, MA: Belknap Press, 2012), 52.

VII. The importance of the maternal relationship doesn’t mean fathers don’t matter—it was merely a by-product of the age, as the scientists largely neglected to study the typical father-child relationship.