August 17, 1998
ANNALS OF BEHAVIOR
Do Parents Matter?
Judith Rich Harris and child development

The idea that will make Judith Rich Harris famous came to her, unbidden, on the afternoon of January 20, 1994. At the time, Harris was a textbook writer, with no doctorate or academic affiliation, working from her home in suburban New Jersey. Because of a lupus-like illness, she doesn't have the strength to leave the house, and she'd spent that morning in bed. By early afternoon, though, she was at her desk, glancing through a paper by a prominent psychologist about juvenile delinquency, and for some reason a couple of unremarkable sentences struck her as odd: "Delinquency must be a social behavior that allows access to some desirable resource. I suggest that the resource is mature status, with its consequent power and privilege." It is an observation consistent with our ideas about what it means to grow up. Teen-agers rebel against being teen-agers, against the restrictions imposed on them by adults. They smoke because only adults are supposed to smoke. They steal cars because they are too young to have cars. But Harris was suddenly convinced that the paper had it backward. "Adolescents aren't trying to be like adults--they are trying to contrast themselves with adults," she explains. "And it was as if a light had gone on in the sky. It was one of the most exciting things that have ever happened to me. In a minute or two, I had the germ of the theory, and in ten minutes I had enough of it to see that it was important."

If adolescents didn't want to be like adults, it was because they wanted to be like other adolescents. Children were identifying with and learning from other children, and Harris realized that once you granted that fact all the conventional wisdom about parents and family and child-rearing started to unravel. Why, for example, do the children of recent immigrants almost never retain the accents of their parents? How is it that the children of deaf parents manage to learn how to speak as well as children whose parents speak to them from the day they were born? The answer has always been that language is a skill acquired laterally--that what children pick up from other children is at least as important as what they pick up at home. Harris was asking whether this was true more generally: what if children also learn the things that make them who they are--that shape their characters and personalities--from their peer group? This would mean that, in some key sense, parents don't much matter--that what's important is not what children learn inside the home but what they learn outside the home.

"I was sitting and thinking," Harris told me, looking bright-eyed as she clutched a tall glass of lemonade. She is tiny--a fragile, elfin grandmother with a mop of gray hair and a little-girl voice. We were in her kitchen, looking out on the green of her back yard. "I told my husband, Charlie, about it. I had signed a contract to write a developmental-psychology textbook, and I wasn't quite ready to give it up. But the more I thought about it the more I realized I couldn't go on writing developmental-psychology textbooks, because I could no longer say what my publishers wanted me to say." Over the next six months, Harris immersed herself in the literature of social psychology and cultural anthropology. She read studies of group behavior in primates and unearthed studies from the nineteen-fifties of pre-adolescent boys. She couldn't conduct any experiments of her own, because she didn't belong to an academic institution. She couldn't even use a proper academic library, because the closest university to her was Rutgers, which was forty-five minutes away, and she didn't have the strength to leave her house for more than a few hours at a time. So she went to the local public library and ordered academic texts through interlibrary loan and sent for reprints of scientific articles through the mail, and the more she read the more she became convinced that her theory could tie together many of the recent puzzling findings in behavioral genetics and developmental psychology. In six weeks, in August and September of 1994, she wrote a draft and sent it off to the academic journal Psychological Review. It was an act of singular audacity, because Psychological Review is one of the most prestigious journals in psychology, and prestigious academic journals do not, as a rule, publish the musings of stay-at-home grandmothers without Ph.D.s. But her article was accepted, and in the space below her name, where authors typically put "Princeton University" or "Yale University" or "Oxford University," Harris proudly put "Middletown, New Jersey." Harris listed her CompuServe address in a footnote, and soon she was inundated with E-mail, because what she had to say was so compelling and so surprising and, in a wholly unexpected way, so sensible that everyone in the field wanted to know more. Who are you? scholars asked. Where did you come from? Why have I never heard of you before?

At this point, Harris's health was not good. Her autoimmune disorder began to attack her heart and lungs, and she sometimes wondered how long she had to live. But, at the urging of some of her new friends in academe, she set out to write a book, and somehow in the writing of it she became stronger. That book, "The Nurture Assumption," will be published this fall, and it is a graceful, lucid, and utterly persuasive assault on virtually every tenet of child development. It begins, "This book has two purposes: first, to dissuade you of the notion that a child's personality--what used to be called 'character'--is shaped or modified by the child's parents; and second, to give you an alternative view of how the child's personality is shaped." On the back cover are enthusiastic blurbs from David Lykken, of the University of Minnesota; Robert Sapolsky, of Stanford; Dean Keith Simonton, of the University of California at Davis; John Bruer, of the James S. McDonnell Foundation; and Steven Pinker, of MIT--which, in the social-science business, is a bit like writing a book on basketball and having it endorsed by the starting five of the Chicago Bulls. This week, Harris will travel to San Francisco for the annual convention of the American Psychological Association, where she will receive a prize for her Psychological Review article.

"It's as if the gods were making up to me all that they had done to me previously," Harris told me. "It was the best gift I could have ever gotten: an idea. It wasn't something that I could have known in advance. But, as it turned out, it was what I wanted most in the world--an idea that would give a direction and a purpose to my life."

Judith Harris's big idea--that peers matter much more than parents--runs counter to nearly everything that a century of psychology and psychotherapy has told us about human development. Freud put parents at the center of the child's universe, and there they have remained ever since. "They fuck you up, your mum and dad. They may not mean to, but they do," the poet Philip Larkin memorably wrote, and that perspective is fundamental to the way we have been taught to understand ourselves. When we go to a therapist, we talk about our parents, in the hope that coming to grips with the events of childhood can help us decipher the mysteries of adulthood. When we say things like "That's the way I was raised," we mean that children instinctively and preferentially learn from their parents, that parents can be good or bad role models for children, that character and personality are passed down from one generation to the next. Child development has been, in many ways, concerned with understanding children through their parents.

In recent years, however, this idea has run into a problem. In a series of careful and comprehensive studies (among them the famous Minnesota studies of twins separated at birth) behavioral geneticists have concluded that about fifty per cent of the personality differences among people--traits such as friendliness, extroversion, nervousness, openness, and so on--are attributable to our genes, which means that the other half must be attributable to the environment. Yet when researchers have set out to look for this environmental influence they haven't been able to find it. If the example of parents were important in a child's development, you'd expect to see a consistent difference between the children of anxious and inexperienced parents and the children of authoritative and competent parents, even after taking into account the influence of heredity. Children who spend two hours a day with their parents should be different from children who spend eight hours a day with their parents. A home with lots of books should result in a different kind of child from a home with very few books. In other words, researchers should have been able to find some causal link between the specific social environment parents create for their children and the way those children turn out. They haven't.

One of the largest and most rigorous studies of this kind is known as the Colorado Adoption Project. Between 1975 and 1982, a group of researchers at the University of Colorado, headed by Robert Plomin, one of the world's leading behavioral geneticists, recruited two hundred and forty-five pregnant women from the Denver area who planned to give up their children for adoption. The researchers then followed the children into their new homes, giving them a battery of personality and intelligence tests at regular intervals throughout their childhood and giving similar tests to their adoptive parents. For the sake of comparison, the group also ran the same set of tests on a control group of two hundred and forty-five parents and their biological children. For the latter group, the results were pretty much as one might expect: in intellectual ability and certain aspects of personality, the kids proved to be fairly similar to their parents. The scores of the adopted kids, however, had nothing whatsoever in common with the scores of their adoptive parents: these children were no more similar in personality or intellectual skills to the people who reared them, fed them, clothed them, read to them, taught them, and loved them all their lives than they were to any two adults taken at random off the street.

Here is the puzzle. We think that children resemble their parents because of both genes and the home environment, both nature and nurture. But, if nurture matters even a little, why don't the adopted kids have at least some greater-than-chance similarities to their adoptive parents? The Colorado study says that the only reason we are like our parents is that we share their genes, and that--by any measures of cognition and personality--when there is no genetic inheritance there is no resemblance.

This is the question that so preoccupied Harris on that winter morning four and a half years ago. She knew that most people in psychology had responded to findings like those of the Colorado project by turning an ever more powerful microscope on the family, assuming that if we couldn't see the influence of parents through standard psychological measures it was because we weren't looking hard enough. Not looking hard enough wasn't the problem. The problem was that psychologists weren't looking in the right place. They were looking inside the home when they should have been looking outside the home. The answer wasn't parents; it was peers.

Harris argues that we have been in the grip of what she calls the "nurture assumption," a parent-centered bias that has blinded us to what really matters in human development. Consider, she says, the seemingly common-sense statements "Children who are hugged are more likely to be nice" and "Children who are beaten are more likely to be unpleasant." Sure enough, if you study nice, well-adjusted children, it turns out that they generally have well-adjusted and nice parents. But what does this really mean? Since genes account for about half of personality variations among people, it's quite possible that nice children are nice simply because they received nice genes from their parents--and nice parents are going to be nice to their children. Hugging may have made the children happy, and it may have taught them a good way of expressing their affection, but it may not have been what made them nice. Or take the example of smoking. The children of smokers are more than twice as likely to smoke as the children of nonsmokers, so it's natural to conclude that parents who smoke around their children set an example that their kids follow. In fact, a lot of parents who smoke feel guilty about it for that very reason. But if parents really cause smoking there ought to be elevated rates of smoking among the adopted children of smokers, and there aren't. It turns out that nicotine addiction is heavily influenced by genes, and the reason that so many children of smokers smoke is that they have inherited a genetic susceptibility to tobacco from their parents. David C. Rowe, a professor of family studies at the University of Arizona (whose academic work on the limits of family influence Harris says was critical to her own thinking), has analyzed research into this genetic contribution, and he concludes that it accounts entirely for the elevated levels of cigarette use among the children of smokers. With smoking, as with niceness, what parents do seems to be nearly irrelevant.

Harris makes another, subtler point about parents. What if, she asks, the cause-and-effect assumption with niceness and hugging can also go the other way? What if, all other things being equal, nice children tend to be hugged because they are nice, and unpleasant children tend to be beaten because they are unpleasant? Children, after all, are born with individual temperaments. Some children are easy to rear from the start and others are difficult, and those innate characteristics, she says, can strongly influence how parents treat them. Harris tells a story about a mother with two young children--a five-year-old girl, named Audrey, and a seven-year-old boy, named Mark--who walked by Harris's house one day when she was out in the front yard with her dog, Page. Page ran toward the children, barking menacingly. Audrey went up to the animal and asked her mother, "Can I pet him?" Her mother quickly told her not to. Mark, meanwhile, was cowering on the other side of the street, and he stayed there even after Harris rushed up and grabbed Page by the collar. "Come on, Mark, the dog won't hurt you," the mother said, and she waited for her son to come back across the street. What is the parenting "style" here that is supposedly so important in shaping personality? This mother is playing two very different roles--coaxing the frightened Mark and reining in the brash Audrey--and in each case her behavior is shaped by the actions and the temperament of her child, and not the other way around.

This phenomenon--what Harris calls child-to-parent effects--has been explored in detail by psychological researchers. David Reiss, of George Washington University, and Robert Plomin, the behavioral geneticist who headed the Colorado study, and a number of colleagues have just completed a ten-year, nine-million-dollar study of seven hundred and twenty American families. Thirty-two teams of testers were recruited, and they visited each family three times in the course of three years, giving parents and siblings personality tests, videotaping interactions between parents and children, questioning teachers, asking siblings about siblings, asking parents about children, asking children about parents--all to find out whether the differences in how parents relate to each of their children make any predictable difference in the way those children end up. "We thought that this was going to be a straight shot," Reiss told me. "The sibling who got the better micro environment would do better, be less depressed, be less antisocial. It seemed like a no-brainer." It wasn't. Plomin told me, "If we just ask the simple question 'Does differential parental treatment relate to differences in adolescent adjustment?' the answer is yes--hugely. If you take negative parents--conflict, hostility--it's the strongest predictor of negative adjustment of the siblings." But the study was designed to look at genetic influences as well--to examine whether children had personality traits that were causing parental behavior--and when those genetic factors were taken into consideration the link between negative parenting and problems in adolescence almost entirely disappeared. "The parents' negativity isn't causing the negative adjustment of the kids," Plomin said. "It's reflecting it. This was a tremendous surprise to us." What looks like nurture is sometimes just nature, and what looks like a cause is sometimes just an effect.

Harris takes this argument one step further. Consider, she says, the story of Cinderella:

The folks who gave us this tale ask us to accept the following premises: that Cinderella was able to go to the ball and not be recognized by her stepsisters, that despite years of degradation she was able to charm and hold the attention of a sophisticated guy like the prince, that the prince didn't recognize her when he saw her again in her own home dressed in her workaday clothing, and that he never doubted that Cinderella would be able to fulfill the duties of a princess and, ultimately, of a queen.

If you think of the influence of parents and the home environment as monolithic, this tale does seem impossibly far-fetched. So why does the Cinderella story work? Because, Harris says, all of us understand that it is possible to be one person to our parents and another person to our friends. "Cinderella learned whenshe was still quite small that it was best to act meek when her stepmother was around, and to look unattractive in order to avoid arousing her jealousy," Harris writes. But outside the house Cinderella learned that she could win friends by being pretty and charming. Harris says that this lesson--that away from our parents we can reconstruct ourselves--is one that all children learn very quickly, and it is an important limitation on the power of parents: even when they do succeed in influencing their children, those influences very often don't travel outside the home.

The Cinderella effect shows up all the time in psychological research. For example, Harris notes that in the August, 1997, issue of the Archives of Pediatrics and Adolescent Medicine there is a study showing that the more mothers spanked their kids, the more troublesome the kids became. "When parents use corporal punishment to reduce antisocial behavior," the researchers report, "the long-term effect tends to be the opposite." These findings made headlines across the country. In the same issue of that journal, however, another study of children and corporal punishment reached the opposite conclusion: "For most children claims that spanking teaches aggression seem unfounded." The disparity is baffling until you remember the Cinderella effect. The first study asked mothers to evaluate their children's behavior at home. Not surprisingly, it suggested that repeated spanking contributes to the kind of negative relationship that causes further misbehavior. The second study, however, asked kids how often they got into fights at school, and the world of school is a very different place from the world of home. Just the fact that a child wasn't getting along with his mother didn't necessarily mean that he wouldn't get along with his peers.

In another instance, Harris cites a Swedish study of picky eating among primary-school children. Some kids were picky eaters at school, some were picky at home, but only a small number were picky at home and school. A child who pushes away broccoli at the kitchen table might gobble it down in the school cafeteria. In the same way, a child might be shy and retiring at home but a chatterbox in the classroom. Harris applies the same logic to birth-order effects--the popular idea that a good part of our personality is determined by where we stand in relation to our siblings. "At home there are birth order effects, no question about it, and I believe that is why it's so hard to shake people's faith in them," Harris writes. "If you see people with their parents or their siblings, you do see the differences you expect to see. The oldest does seem more serious, responsible, and bossy. The youngest does behave in a more carefree fashion." But that's only at home. Studies that look at the way people act outside the home, and away from the parents and siblings, don't see any consistent effects at all. The younger brother cowed by his older siblings all his years of growing up is perfectly capable of being a dominant, take-charge figure when he's among his friends. "Socialization research has demonstrated one thing clearly and irrefutably: a parent's behavior toward a child affects how the child behaves in the presence of the parent or in contexts that are associated with the parent," Harris concludes. "I have no problem with that--I agree with it. The parent's behavior also affects the way the child feels about the parent. When a parent favors one child over another, not only does it cause hard feelings between the children--it also causes the unfavored child to harbor hard feelings against the parent. These feelings can last a lifetime." But they don't necessarily cross over into the life the child leads outside the home--the place where adults spend the majority of their lives.

Not long ago, Anne-Marie Ambert, a sociologist at York University, in Ontario, asked her students to write short autobiographies describing, among other things, the events in their lives which made them most unhappy. Nine per cent identified something that their parents had done, while more than a third pointed to the way they had been treated by peers. Ambert concluded:

There is far more negative treatment by peers than by parents.... In these autobiographies, one reads accounts of students who had been happy and well adjusted, but quite rapidly began deteriorating psychologically, sometimes to the point of becoming physically ill and incompetent in school, after experiences such as being rejected by peers, excluded, talked about, racially discriminated against, laughed at, bullied, sexually harassed, taunted, chased or beaten.

This is Harris's argument in a nutshell: that whatever our parents do to us is overshadowed, in the long run, by what our peers do to us. In "The Nurture Assumption,"Harris pulls together an extraordinary range of studies and observations to support this idea. Here, for example, is Harris on delinquency. First, she cites a study of juvenile delinquency--vandalism, theft, assault, weapons possession, and so on--among five hundred elementary-school and middle-school boys in Pittsburgh. The study found that African-American boys, many of them from poor, single-parent, "high-risk" families, committed far more delinquent acts than the white kids. That much isn't surprising. But when the researchers divided up the black boys by neighborhood the effect of coming from a putatively high-risk family disappeared. Black kids who didn't live in the poorest, underclass neighborhoods--even if they were from poor, single-parent families--were no more delinquent than their white, mostly middle-class peers. At the same time, Harris cites another large study--one that compared the behavior of poor inner-city kids from intact families to the behavior of those living only with their mothers. You'd assume that a child is always better off in a two-parent home, but the research doesn't bear that out. "Adolescent males in this sample who lived in single-mother households did not differ from youth living in other family constellations in their alcohol and substance use, delinquency, school dropout, or psychological distress," the study concluded. A child is better off, in other words, living in a troubled family in a good neighborhood than living in a good family in a troubled neighborhood. Peers trump parents.

Other studies have shown that children living without their biological fathers are more likely to drop out of school and, if female, to get pregnant in their teens. But is this because of the absence of a parent, Harris asks, or is it because of some factor that is merely associated with the absence of a parent? Having a stepfather around, for example, doesn't make a kid any less likely to be unemployed, to drop out, or to be a teen-age mother. Nor does having lots of contact with one's biological father after he has left. Nor does having another biological relative--a grandparent, for instance--in the home. Nor does it seem to matter when the father leaves: kids whose parents split up when they were in their early teens are no better off and no worse off than kids whose fathers left when they were infants. And, curiously, children whose fathers die aren't worse off at all. In short, there isn't a lot of evidence that the loss of adult guidance and role models caused by fatherlessness has specific behavioral consequences. So what is it? One obvious factor is income: single mothers have less money than married mothers, and income has a big effect on the welfare of children. If your parents split up and you move from Riverdale to the South Bronx, you're obviously going to be a lot worse off--although it's not the loss of your father that makes the difference. This brings us to another factor: relocation. Single-parent families move more often than intact families, and, according to one major study, those extra changes of residence could account for more than half the increased risk of dropping out, of teen-age pregnancy, and of unemployment among the children of divorce. The problem with divorce, in short, is not so much that it disrupts kids' relationships with their parents as that it disrupts kids' relationships with other kids. "Moving is rough on kids," Harris writes. "Kids who have been moved around a lot--whether or not they have a father--are more likely to be rejected by their peers; they have more behavioral problems and more academic problems than those who have stayed put."

All these findings become less perplexing when you accept one of Harris's central observations; namely, that kids aren't interested in becoming copies of their parents. Children want to be good at being children. How, for example, do you persuade a preschooler to eat something new? Not by eating it yourself and hoping that your child follows suit. A preschooler doesn't care what you think. But give the food to a roomful of preschoolers who like it, and it's quite probable that your child will happily follow suit. From the very moment that children first meet other children, they take their cues from them.

One of the researchers whom Harris draws on in her peer discussion is William A. Corsaro, a professor of sociology at Indiana University and a pioneer in the ethnography of early childhood. He was one of the first researchers to spend months crouching by swing sets and next to monkey bars closely observing the speech and play patterns of preschoolers. In one of his many playground stakeouts, Corsaro was sitting next to a sandbox and watching two four-year-old girls, Jenny and Betty, play house, and put sand in pots, cupcake pans, and teapots. Suddenly, a third girl, Debbie, approached. Here is Corsaro's full description of the scene:

After watching for about five minutes [Debbie] circles the sandbox three times and stops again and stands next to me. After a few more minutes of watching, Debbie moves to the sandbox and reaches for a teapot. Jenny takes the pot away from Debbie and mumbles, "No." Debbie backs away and again stands near me, observing the activity of Jenny and Betty. Then she walks over next to Betty, who is filling the cupcake pan with sand.

Debbie watches Betty for just a few seconds, then says,"We're friends, right, Betty?"

Betty, not looking up at Debbie, continues to place sand in the pan and says, "Right."

Debbie now moves alongside Betty, takes a pot and spoon, begins putting sand in the pot, and says, "I'm making coffee."

"I'm making cupcakes," Betty replies.

Betty now turns to Jenny and says, "We're mothers, right, Jenny?"

"Right," says Jenny.

The three "mothers" continue to play together for about twenty more minutes, until the teachers announce cleanup time.

To adults, this exchange looks somewhat troubling. If you saw Debbie circling the sandbox over and over, you'd think she was shy and timid. And if you came upon the three girls just as Jenny told Debbie no you'd think Jenny was selfish and needed to be taught to share. In both cases, the children seem profoundly antisocial. In fact, Corsaro says, the opposite is true. A preschool playground is rather like a cocktail party. There are lots of informal clusters of kids playing together, and the kids are in constant movement, from cluster to cluster. Unlike at a cocktail party, though, the play clusters are very fragile. "If the phone rang right now," Corsaro said to me when I met him, in his office in Bloomington, "I could answer it, talk for five minutes, and then we could pick up where we left off. It's easy for us. When you are a three- or four-year-old and you've generated something spontaneous and it's going well, it's not so easy." The bell can ring. An adult can step in. An older child can disrupt things. As a result, they spend a lot of effort trying to protect their play from disruption. Betty and Jenny aren't resistant to sharing when they initially say no to Debbie. They are already sharing, and the point of keeping Debbie at bay is to defend that shared play.

What has evolved in preschool culture, then, is what Corsaro calls access strategies--an elaborate set of rules and rituals that govern when and how the third parties circulating through the playground are allowed to join an existing game. Debbie's approach to the sandbox is what Corsaro calls nonverbal entry--the first common opening move in the access dance. She's waiting for an invitation to join. It's the same at an adult cocktail party. You don't come up to an existing conversation and say, "May I join in?" You join the group quietly, as if to demonstrate respect for the existing conversation. When Debbie goes around and around the sandbox, she's trying to understand the basis of Jenny and Betty's play. Corsaro calls this encirclement. Notice that when Debbie initially reaches for a teapot Jenny says no. Debbie hasn't proved that she understands the game in question. So she retreats and observes further. Then she makes what Corsaro calls a verbal reference to affiliation--"We're friends, right?" It's as if she were offering her bona fides. She gets a positive response. Now she enters again, this time making it absolutely clear that she understands the game: "I'm making coffee." She's in. This is how children learn to get along. Kids teach each other how to be social. Indeed, to the extent that adults might get involved in an access situation--by, for example, instructing Jenny and Betty that they have to share with Debbie--they would frustrate the learning process.

Corsaro is a quiet, bearded man of fifty, with the patient, stubborn air of someone who has spent the better part of his life sitting and watching screaming three-year-olds. Harris E-mailed him when she was writing her Psycholo gical Review paper, and the two have struck up an on-line friendship. Most people, Corsaro says, want to figure out what his work says about individual development. Harris, though, recognized at once what Corsaro considers the real lesson, which is the children's immediate and powerful attraction to their own peer group. Once, Corsaro spent close to a year in a preschool where the children had been forbidden to bring their toys into the classroom. Before long, he noticed that they had found a way around the rule: the children were selecting the smallest of their toys--the boys chose Matchbox toy cars, for example, and the girls little plastic animals--and hiding them in their pockets. These were only preschoolers, but already they were organizing against the adult world, defining themselves as a group in opposition to their elders. "What I found interesting was not that the kids wanted to bring their own toys but that when they smuggled them in they never played with them alone. They played with them collectively," Corsaro told me. "They wanted others to know that they had them. They wanted to share the toys with others. They are not only sharing the toy but sharing the fact that they are getting around the rule. This is what is unique. I think there is a real, strong emotional satisfaction in sharing things, in doing things together." Even for a child of three or four, the group is critical.

Judith Harris and her husband, Charles, have two children. The first, Nomi, is their biological daughter, and the second, Elaine, is adopted. In that sense, Harris's own family is a kind of micro-version of the adoption studies that raise the question of parental influence, and she says that without the example of her daughters she might not have reached the conclusion she did. Nomi, the elder, was quiet and self-sufficient as a child, a National Merit Scholar who went on to do graduate work at MIT. "She is very much like me and Charlie," Harris says. "She gave us no trouble while she was growing up. She didn't require much guidance, because she didn't want to do anything that we didn't want her to do. Even before she could walk, she would crawl off to another part of the house, and I'd find her taking things out of a drawer and looking at them carefully--and putting them down carefully."

Elaine was different. "When she was little, all you had to do was look down and she was there, right on my heels," Harris recalls. "She always wanted to be with people. We started getting bad reports from the school right away--that she wouldn't sit in her chair, and she was bothering other kids. When Nomi would ask a question, it was because she was interested in the answer. When Elaine would ask a question, it was because she was interested in having the interaction. Nomi would ask a question once. Elaine would often ask a question several times. As the girls got older, Nomi became a brain and Elaine became a dropout. Nomi was a member of a very small clique of intellectual kids, and Elaine was a member of the delinquent subgroup. They went in opposite directions."

Harris has an optimistic air about her, as if all her troubles had only served to strengthen her appreciation of life. But it's clear that bringing up Elaine represented a real crisis in her life. When Elaine was six and Nomi was ten, Harris became ill for the first time. She was in such pain that she couldn't sit up for more than half an hour. She tried taking a graduate course in psychology, hoping to finish a doctorate she had started, in the early sixties, at Harvard, and she had a fellow-student carry a cot to class so she could lie down during lectures. But even that was too hard, so she became a textbook writer, lying in her bed, with a spiral-bound notebook on her knee, and Nomi acting as her typist. She had pneumonia, a heart murmur, pulmonary hypertension, shingles, a year of chronic hives, and a minor stroke. "Sometimes," she says, "I felt like Job," and in the midst of all her troubles her younger daughter seemed out of control.

"We had very bad years with her in her teens," she recalls. "We didn't know how to handle her." Harris says that she began motherhood as a classic environmentalist, meaning she believed that children would reflect the environment in which they were reared. Had she stopped with Nomi, she says, she might have attributed Nomi's studiousness and self-sufficiency and success to her own enlightened parenting. It was Elaine who made the puzzle posed by the adoption studies seem real. "I assumed that an adopted child would represent her environment, and that if I could give Elaine the same kind of environment I gave to my first child she would turn out--of course, not the same..." She thought for moment. "But I certainly didn't expect that she would be so vastly different. I couldn't see that I was having any effect on her at all." Harris seems a little reluctant to talk about those years, particularly since Elaine turned out, as she puts it, "amazingly well" and is now happy and married, with a toddler and a career as a licensed practical nurse. But it's not hard to imagine the kind of guilt and frustration she must have felt--maternal helplessness magnified by her physical debility--as she and Charles did everything that good parents are supposed to do yet still came up short. Her epiphany was, in a way, her release, because she came to believe that the reason she and Charles couldn't see that they were having any effect on Elaine was that parents really can't have a big effect on their children.

There are a hundred ways of explaining Nomi and Elaine, and there is, of course, something very convenient about the explanation that Harris arrived at: it's the kind of thing that the mother of a difficult child wants to believe. Harris has constructed a theory that lets herself off the hook for her daughter's troubled childhood. It should be said, though, that the idea that parents can control the destiny of their children by doing all the right things--by providing children with every lesson and every experience, by buying them the right toys and saying the right words and never spanking or publicly scolding them--is just as self-serving. At least, Harris's theory calls for neighborhoods, peers, and children themselves to share the blame--and the credit--for how children turn out. The nurture assumption, by contrast, places the blame and the credit squarely on the parent, and has made it possible to demonize all those who fail to measure up to the strictest standards of supposedly optimal parenting. "I want to tell parents that it's all right," Harris told me. "A lot of people who should be contributing children to our society, who could be contributing very useful and fine children, are reluctant to do it, or are waiting very long to have children, because they feel that it requires such a huge commitment. If they knew that it was O.K. to have a child and let it be reared by a nanny or put it in a day-care center, or even to send it to a boarding school, maybe they'd believe that it would be O.K. to have a kid. You can have a kid without having to devote your entire life--your entire emotional expenditure--to this child for the next twenty years."

Harris does not see children as delicate vessels and does not believe they are easily damaged by the missteps of their mothers and fathers. We have been told, Harris writes, to tell children not that they've been bad but that what they did was bad, or, even more appropriately, that what they did made us feel bad. In her view, we have come to insist on these niceties only because we have forgotten what the world of children is really like. "Kids are not that fragile," she writes. "They are tougher than you think. They have to be, because the world out there does not handle them with kid gloves. At home, they might hear 'What you did made me feel bad,' but out on the playground it's 'You shithead!'"

Is Harris right? She is the first to admit that what she has provided is only, at this stage, a theory. From her tiny study, off the main hallway of her home in New Jersey, she is scarcely in a position to do the kind of multimillion-dollar, multi-year study that is needed to test her hypothesis. "My guess is that some of the more threatened elders in the field of psychology are going to go out of their way to try and savage this," Robert Sapolsky, a neurobiologist at Stanford, says. "But my gut feeling is that this is really important. Harris makes a lot of sense. Sometimes she is a little doctrinaire"--he paused--"but, boy." Already, Harris has helped wrench psychology away from its single-minded obsession with chronicling and interpreting the tiniest perturbations of family life. The nurture assumption, she says, has turned childhood into parenthood: it has turned the development of children into a story almost entirely about their parents. "Have you ever thought of yourself as a mirror?" Dorothy Corkille Briggs asks in her pop-psychology handbook "Your Child's Self-Esteem." "You are one--a psychological mirror your child uses to build his identity. And his whole life is affected by the conclusions he draws." And here are Barbara Chernofsky and Diane Gage, in "Change Your Child's Behavior by Changing Yours," on how children relate to their parents: "Like living video cameras, children record what they observe." This is the modern-day cult of parenting. It takes as self-evident the idea that the child is oriented, overwhelmingly, toward the parents. But why should that be true? Don't parents, in fact, spend much of their time instructing their children not to act like adults--that they cannot be independent, that they cannot make decisions entirely by themselves, that different rules apply to them because they are children?

"If developmental psychology were an enterprise conducted by children, there is no question that peer relationships would be at the top of the list," Peter Gray, a psychologist at Boston College, told me. "But because it is conducted by adults we tend, egocentrically, to believe that it is the relationship between us and our children that is important. But just look at them. Whom do they want to please? Are they wearing the kind of clothing that other kids are wearing or the kind that their parents are wearing? If the other kids are speaking another way, whose language are they going to learn? And, from an evolutionary perspective, whom should they be paying attention to? Their parents--the members of the previous generation--or their peers, who will be their future mates and future collaborators? It would more adaptive for them to be better attuned to the nuances of their peers' behavior. That just makes a lot of sense."

Harris's health is more stable now, and when she was putting the finishing touches on her book this summer she was sometimes able to work at the computer twelve, or even fourteen, hours a day. But anything more strenuous is out of the question. The woman who says that what really matters is what happens outside the home rarely leaves the home--not for vacations, or even to see a movie. Indeed, none of the heavyweight psychologists who have befriended her since her Psychological Review article ran have ever met her. "Writing E-mail is my recreation," she wrote me in an E-mail.

When Harris goes to San Francisco this week, for the A.P.A. convention, it will be a kind of coming-out party. In preparation, during the past few weeks she has had to go shopping. "I have to buy clothes," she said. "I've hardly been out of the house in years." On August 15th, she will take the stage and receive a prize named in honor of the eminent scholar George A. Miller. Almost four decades ago, Harris was kicked out of graduate school after only two years, and the dean who delivered the news was the same George A. Miller. The two have since corresponded, and Miller has termed the irony "delicious." In her acceptance remarks, Harris told me, she intends to read from the letter that Miller wrote her long ago: "I hesitate to say that you lack originality and independence, because in many areas of life you obviously possess both of those traits in abundance. But for some reason you have not been able to bring them to bear on the kind of problems in psychology to which this department is dedicated....We are in considerable doubt that you will develop into our professional stereotype of what an experimental psychologist should be."