摘錄

Chapter One Making Connections Alice laughed: "There's no use trying, "she said; "one can't believe impossible things." "I dare say you haven't had much practice," said the Queen. "When I was younger, I always did it for half an hour a day. Why, sometimes I've believed as many as six impossible things before breakfast." --Lewis Carroll           Back in 1980, the Schlitz company tried to convince people that its beer was better than the major rival brands. The company raised prices while using more corn syrup and less barley malt, the ingredient that gives beer its flavor and body. Offering customers a less tasty, more expensive product backfired, with sales going down dramatically. The disastrous course could eventually be reversed, but only if the public were convinced that Schlitz had become not only comparable to, but better than, the competition.     The key to commercial redemption lay in clever advertising, specifically a razzle-dazzle TV extravaganza involving blind beer-tasting tests done in what appeared to be a football stadium with cheering crowds and a referee. Seated at a table were either confirmed Bud or Miller drinkers, people who drank at least two six-packs of their preferred beer weekly. In front of them were two identical opaque mugs, one containing Schlitz, the other either Bud for Bud drinkers or Miller for Miller drinkers.     The results? In just about every test, roughly half the Bud or Miller drinkers picked Schlitz. It would seem that Schlitz must be the better beer. What else explains why so many loyal Bud or Miller drinkers switched to Schlitz? Well, the conclusion is questionable because the so-called beer-tasting test left out a key element: Schlitz drinkers. What if roughly half of them had switched to Bud or Miller? That would have confirmed the alternative hypothesis that Schlitz was now essentially no different from the other brands. By leaving out Schlitz drinkers, the Schlitz taste test was really only half a test--the hypothesis was only half-asked--and therefore no test at all because the hypothesis of Schlitz's superiority was never in danger of refutation.     Likewise, by ignoring certain kinds of evidence, social psychological theories often avoid the danger of refutation. The usual parent-child correlations, no matter how consistent with social theorizing, leave out much information that supports alternative explanations of a person's behavior. Given that parents P are a dominant presence, while children C are little and impressionable things, how readily we assume that parenting explains much of the unique psychological development of a child--that P explains C . Unfortunately, such PC theory ignores or minimizes new, more definitive evidence of critically important genetic and other inborn factors ( X, Y, Z ) that have little or nothing to do with parenting.     The power of such PC theorizing is evident when otherwise sensible people seem oblivious to the obvious, for example, when the usually sensible historian Will Durant suggests quite inexplicably that, "Henry Purcell's genius was in large part a product of social heredity--i.e., adolescent environment. His father was master of the choristers at Westminster Abbey; his uncle was a `composer in ordinary for the violins to his Majesty'; his brother was a composer and a dramatist, his son and his grandson continued his role as organist in the Abbey." Is it not just as obvious--and, given the evidence, most likely true--that Purcell's genius was the product of genetic heredity? The Nature of the Evidence For novelist John Irving, "It's a no-win argument--that business of what we are born with and what our environment does to us. And it's a boring argument, because it simplifies the mysteries that attend both our birth and our growth." Irving is surely right that the argument, being essentially statistical--applying to the generality rather than to any one person--must render human experience less real, less complex, less poignant; most arguments over nature and nurture really do seem rather boring. Irving is also right that the usual evidence, being essentially correlational--confounding cause and effect, nature and nurture--must render scientific explanations less clear, less certain, less definitive.     However, research with twins and adoptees really has moved us beyond sterile debate, allowing us to test hypotheses that lead to interesting findings. Such testing has yielded results whose implications are rather clear for certain traits, and interesting not just for showing the strength of genetic influence, but in raising serious questions about the power of parental influence. The key to all of this is the quality of the methods and the credibility of the findings, and this is as true for behavior genetic research as for the social science kind.     Strictly speaking, the behavior genetic method isn't truly experimental, for like the rest of us, twins and adoptees don't get randomly assigned to rearing environments. Most twins are reared by the same parents who have given them their genes, which confounds the influences of nature and nurture. With adoptees, including identical twins reared apart, there are other potential problems, for example, they generally get placed with parents who are above average in many ways intellectually and emotionally, and who are unlikely to be abusive. All this violates the experimental ideal of having newborns randomly assigned to homes representing the full range of rearing environments, from the healthiest to the most abusive. Adoptees may even be selected on the additional grounds of having ethnic, religious, social, intellectual, or even physical characteristics that match those of the adoptive parents--nothing random about this. Finally, twins reared apart are sometimes brought up by different relatives, and there's nothing random about this, either.     This narrowed range of environments experienced by the twins and adoptees limits what can be said about family influence from such research. No doubt, greater emotional differences than are usually found for identicals would be found if one twin were reared by loving parents, the other by abusive parents. Bigger psychological differences would mark twins reared in radically different cultures, but again, the interesting question is rather how much some of the psychological similarities expected for identicals would yet survive such environmental differences.     While we can't answer that question, we can answer the more realistic one about differences for identicals reared apart in social environments that differ in the usual, still quite divergent ways that mark ordinary families in a multicultural society. Twin and adoption studies may each be flawed, but they are each flawed in a different way. Each has holes, but stacked one on top of the other, the holes don't go through. What does come through, though, is consistent evidence of genetic reasons for individual differences. And the evidence is convincing because it comes from studies that disentangle the relative influence of heredity and environment. It is the difference between observations that two things go together and controlled observations that really tell us if one of those things causes the other. This is why even one flawed adoption study is worth more than countless conventional family studies.     The power of adoption research can be illustrated here with the so-called birth-order effect on personal development. The question is this: Why might first-borns be brighter, more conservative, and less radical than second borns? Social science offers three explanations. One is that during their earliest, most formative years, first-borns get higher-quality intellectual stimulation. After all, first-borns have just their parents, who are smart adults, while second-borns must contend with parents and a sibling. Another says that first-borns become intellectually more sophisticated by acting as teachers of their second-born siblings. A third says that conflicts inevitably arise when a second-born sibling must compete for parental favors already acquired by the first-born child. Out of these conflicts comes a more liberal personality, one that is relatively more open to new experiences.     Such theorizing about family influence seems reasonable enough, but only by ignoring that birth order is as much physiological as social, that it reflects which sibling is senior in the family (social standing), but also which came from a first or later pregnancy. Psychologists Jeremy Beer and Joe Horn have disentangled the effects of these two normally confounded factors by comparing first-born adoptive children reared either as a social senior with a younger sibling or as a social junior with an older sibling. Thus, they compared first-born social seniors with first-born social juniors.     With such a comparison, none of the expected IQ and personality differences showed up; the expected connection between psychological differences and social seniority simply wasn't there. The clear implication is that the birth-order effect is uterine rather than social, physiological rather than familial. And so it may be asked, with the application of a more definitive scientific method, how many other presumed social influences, including those involving rearing, would likewise prove to be biological? The answer, we will see, is as sobering as it is illuminating. Parenting, Intelligence, and Academic Achievement Good parents care about their children's academic achievement, but where does academic achievement come from? The usual answer, rearing, really can't explain why children's academic achievement correlates more strongly with their parents' intelligence than with their parents' attitudes and rearing styles. In experimental fashion, hold parenting constant--select parents with similar parenting styles--and variation in parental intelligence still predicts their children's academic achievement in adolescence. Now hold parental intelligence constant--select parents with similar intelligence--and variation in parenting doesn't predict that achievement. Here, then, is a good example of how one salient thing (parenting style) can seem so important, when a more subtle thing (intelligence) is the real influence.     But why does intelligence rather than parenting style predict a child's achievement? One social explanation is that intelligence translates into educational practice; for example, brighter parents stimulate their children more, increasing their abilities and therefore their ultimate achievements. Sounds perfectly reasonable, as do many social explanations, but here is a simpler one: Parents transmit their genetic potential to their biological offspring. Even within the same family, children who get the genetic potential for higher intelligence score higher on IQ tests and show greater academic and socioeconomic achievement. In short, intelligence predicts intelligence, whether it is parents' IQ and their offspring's IQ, parents' IQ and their offspring's achievement, or, for that matter, a person's IQ and achievement.     A headline announces that more breast-feeding may facilitate eventual academic achievement. But does a correlation between the number of months children are breast-fed and their academic achievement really mean that breast-feeding has a salutary effect on academic motivation and capability? If so, then mothers should breast-feed longer; but if not, what then? Reading further reveals that the mothers who breast-fed longest were better educated and wealthier--two good indicators of relatively high intelligence. If brighter mothers breast-feed longer, then their children's higher achievement is likely a matter of heredity more than of breast-feeding.     The media are replete with such ambiguous and misleading findings, which often suggest that rearing profoundly affects a child's intellectual development. Yet closer inspection suggests the opposite, that the correlation represents a child's inborn potentials and a parent's limited influence. According to one headline, for example, "Harsh Discipline Hurts IQ." The implication is that, because harsh discipline goes (correlates) with a 12-point IQ disadvantage, punishment causes lower IQ. But what if low-IQ children cause their parents to punish them more? Or what if low-IQ parents tend to be more aggressive and therefore to have more punitive relationships? In either case, less punitive discipline will not appreciably increase a child's IQ.     According to another headline, "Poverty before Age 5 Leaves Children with Lower IQs." The implication is that, because poverty goes (correlates) with low IQ, poverty causes lower intelligence. But what if poverty mostly reflects the low intelligence of parents whose low-IQ children would grow up to be relatively low-IQ adults, poverty notwithstanding? Then reducing poverty will not make low-intelligence kids score appreciably higher on IQ tests.     We read that the number of words an infant hears each day predicts later intelligence, academic achievement, and social competence. Researchers looked at 42 children born to professional parents, working-class parents, or low-income parents. During their first 2 1/2 years, the children's every word and interaction with a parent was recorded for an hour each month. At age 3, standard intelligence testing showed that children of professional parents scored highest, and that spoken language seemed to be a key variable. But was it?     Rather, spoken language may have been an outward expression of what was the actual key difference, a genetic one. Professional parents talked three times more to their infants and gave much more positive feedback to their children than was the case for the low-income parents, but can there be any surprise in this? We are informed that affirmative feedback is important, that a child who hears "What did we do yesterday?" or "What did we see?" will listen more to a parent than will a child who always hears "Stop that!" or "Come here!" Yes, but which parent likely has the higher IQ, and therefore which one will transmit more genes for IQ?     Animals reared in enriched environments have larger brains with more neurons and more connections. Surely this fits what many people have always suspected: that good parents enrich their children's lives, right from infancy, that they provide as much cognitive stimulation as possible along with loving-kindness and emotional support. But do such findings with animals really apply to humans? Even poor households in modern societies may be the psychological equivalent of enriched environments that make laboratory animals develop more brain cells. If so, then most children are already about as stimulated and big-brained as they are likely to be.     Now suppose the findings really did indicate that the more stimulating the human environment, the bigger the effect on a developing brain. What, in the practical sense, would all those extra neurons and new connections really mean? There might be no noticeable effect on potentials and on behavior. On the other hand, children might be brighter for a short while, after which the gains might disappear as inner-directed brain reorganization occurs during puberty, with millions of neurons pruned and billions of neural connections lost or rearranged. Extra growth of neural tissue in animals exposed during infancy to enriched environments disappears when they are transferred to ordinary environments. So it seems fair to question the permanence of early-enrichment effects.     Let's say that the extra stimulation imposed early in life really did translate into reliably brighter kids who could go on to be more successful adults. Then would the effect be big or small relative to other influences? What if the effect explained only 5 percent of all the variation in adult achievement compared to, say, 75 percent from genetic differences? Would the effort justify the financial and emotional cost of a new educational push that would inevitably be demanded by parents, politicians, and the education industry?     Despite the strong impression suggested by magazine articles, not a shred of evidence supports the notion that vigorous efforts to promote psychological enrichment, especially during the so-called critical period, will make an otherwise normally reared child bigger brained and, therefore, brainier. Parenting and Personality Fifteen centuries ago Saint Augustine challenged the belief that personality comes from the planetary alignment at a person's birth. The great fifth-century theologian had his hands full, for as historian Will Durant notes, people accepted, for example, that "persons born under the ascendancy of Saturn would be cold, cheerless, saturnine; those born under Jupiter, temperate and jovial...." Thus, to have a proper horoscope, to observe the hour, required knowing the precise moment of birth and the precise position of the stars.     Nonsense, countered Augustine, for if astrological alignments were truly important, how can fraternal twins born at almost the same moment be as different as any two siblings born in different seasons? Citing the biblical case of Esau and Jacob, he wrote: "Now there was such a difference in their lives and characters, such divergence in their actions, such disparity in the affection of their parents, that these discrepancies turned them into mutual enemies. [Furthermore,] one of those twins was a paid servant, the other was not; one was loved by his mother, the other was not; one lost that position which in those days was counted a great honor, while the other acquired it. Besides, what a vast difference between their wives, their sons, the whole setting of their lives!"     The moral of the story: Even if we can't do experiments--even where there are no experimental data--we can at least think experimentally, for example, about parent-child correlations, some of which are as misleading or illusory as are personality-astrological correlations.     Media stories about parents and children are legion, but what do they really tell us about parental influence? A headline says that "Family Rituals Promote Better Emotional Adjustment." Apparently, because reading bedtime stories and other family rituals go (correlate) with good feelings, such family rituals cause better emotional development. But what if family rituals merely reflect happier, better adjusted family members who like to be with each other? Then adopting more family rituals will not appreciably affect the emotional adjustment of children as they grow up.     A second headline says that "Teens Who Feel Loved Do Better." Apparently, because feeling loved goes (correlates) with doing well socially--say, in the sense of not using drugs--love causes teens to do better. But what if lovable teens who grow up to be lovable adults are simply loved more? Then trying to love teens more will not make them do appreciably better.     A correlation in social behavior has been found, but only for parents and their biological children. Biological children with fewer behavioral problems tend to have more supportive parents; the correlation is a modest 23. It is, however, only 7--just about nonexistent--for adoptive children and their adoptive parents. Apparently, without the usual 50 percent genetic overlap that permits parent-child resemblance, almost no correlation exists between typical parents' behavior and their children's emotional adjustment.     Lacking adoption methods, social scientists are typically reduced to using less definitive methods. With one, they classify parents' self-reported attitudes and approaches to rearing their children--for example, as warm or hostile, restrictive or permissive. They also ask parents to categorize their children as healthy, neurotic (anxious, sensitive), or antisocial. Healthy children are outgoing, independent, friendly, and creative. Neurotic children are quarrelsome, unhappy, timid, and socially withdrawn. Antisocial children are immature and aggressive, noncompliant, or delinquent. Not surprisingly, the expected parent-child associations are found: Parenting associated with healthy children tends to be warm, authoritative, and permissive while parenting associated with neurotic and antisocial children tends to be hostile, authoritarian, and either restrictive or overly permissive.     Good parenting does tend to go with good outcomes, bad parenting with bad outcomes, but does "go with" represent causality from parent to child? If so, why are there so many normal children with neurotic parents or antisocial children with nurturing parents, and why are such questions so often not asked? Even a strong connection between rearing and traits would support various interpretations, in particular, the usual one that parents determine their child's development, and a variant of this, that children determine their parents' behavior. We need only ask ourselves this: Where do parenting practices come from? The obvious answer is parental personality . Regardless of a child's behavior, parents tend to express themselves in certain ways that, we will see, are substantially heritable. We are also responsive to situations and react accordingly, which means that children's traits can drive their parents' behavior. And this includes parents reacting maladaptively to the eccentric or socially disruptive behavior of their own or another parent's child.     Thus, causation can go from child to parent, as when a hyperactive, autistic, or excessively timid child makes a parent aggressive, intrusive, or withdrawn. The parent of a highly active child is in some danger not only of being blamed, but of accepting blame. After all, conflict and family strife seem almost inevitable when even a temperamentally normal parent interacts with a highly active child. Parents of active children are likely to engage in power struggles that can make for disruptive, unsatisfying relationships. Unfortunately, such parental reactions--such parent-child correlations--are often mistaken for causes. When tempted to conclude, no wonder little Billy is so neurotic; just look at his parents, we should consider the reverse: No wonder his parents are so neurotic; just look at little Billy.     Clinical research on infants reinforces the point. Roughly 4 percent of children have extremely difficult temperaments. Their sleeping, eating, and energy levels are markedly irregular, unpredictable, and frustrating to parents. So, too, are their anxiety and irritability, and their tendency to withdraw from or react against new stimuli. Their reactions are unusually intense and have an all-or-nothing quality.     As temperament researchers Thomas, Chess, and Birch explain it, "such a child will tend to cry loudly when he is hurt whether he has a minor scratch or a deep and painful gash.... He may fuss as intensely when he makes a spelling mistake in a minor bit of homework as he may when he makes an error in an extremely important project in which he is engaged. When frustrated, such children may destroy whatever they are working on whether it be a minor construction with tinker toys or a complicated model which has been worked on for several weeks."     Not surprisingly, the parents tend to be irritable, anxious, punitive, or distant, but also guilt-ridden, frustrated, and helpless. Yet before the birth of their difficult child, they no more readily experienced such emotions than did a comparable group of parents whose children turned out to be normal. Rather, they had the usual positive expectations about being a parent and positive attitudes about how to rear children. The drastic change embarrasses the usual assumption that a parent's attitudes and behavior strongly influence not only a child's temperament, but that child's eventual development. Rather, it suggests just how much children are in charge of their own socialization.     A newspaper story has reported that parental use of corporal punishment actually made children more antisocial in later years. The actual finding was that the self-reported number of spankings parents used on their 6 to 9-year-olds correlated with the amount of lying, cheating, and disruptive behavior the children displayed in school 2 years later. Spanking can aggravate bad behavior in specific situations, but does it really increase the disposition to bad behavior that, for reasons of inborn potential, might get worse regardless of what parents do? Perhaps spanking is as much an effect as a cause, a reaction to socially disruptive behavior that a frustrated parent cannot imagine controlling in any other way.     The parent-personality connection can be looked at profitably by asking about sources of self-esteem. Many people with low self-esteem were exposed to parenting that stressed how much they were inferior to others. A child's self-esteem could be undermined by a mother's constant criticizing, while his or her sense of self-importance could be injured by a father's emotional distance. People can thus come to see themselves as unworthy and unlovable.     Well then, does self-esteem come from how people think about themselves, or does how they think about themselves come from their self-esteem? Either view could be argued from standard social science correlations. Perhaps the better answer is that self-esteem simply depends on how good one feels, which depends on one's temperament more than upbringing. After all, with no special change in external condition or personal history, a few little pills of serotonin booster can cause lighter mood and greater self-confidence.     Does the pill create in some people what comes naturally to others because of their sunnier temperament--that is, self-esteem and thoughts about one's self therefore normally arise from an inherited tendency, regardless of circumstances, to feel good or bad, happy or sad, secure or insecure? This idea can explain people who have low self-esteem and a tendency to feel unhappy despite good luck and fortunate circumstances. It can also explain all those people with high self-esteem and a sense of well-being despite bad luck and difficult circumstances.     An even better question: What would happen if newborns were separated from their biological parents and reared by adoptive parents? Would the adoptive offspring born of high self-esteem adults develop the active, self-assertive, high self-esteem qualities of their biological parents? And is it not likely that their active, self-assertive (high self-esteem) traits would cause the adoptive parents to be more active and assertive?     We need no reminding that parents and their biological children show the kinds of modest correspondences called family resemblance. Parents who are characteristically warm, demanding, respectful, and authoritative might simply pass the genes for these admirable qualities to their children--no surprise, then, that these children wind up with the relatively high self-esteem that their parents showed as kids and who, when they, too, become parents, are similarly warm, demanding, respectful, and authoritative with their children. Parents surely influence their children's self-esteem, but just how much and to what extent does that influence carry over into adulthood? On this question, social science correlations, being fundamentally ambiguous, are merely arguable. Divorce, Child Abuse, and Personality Compared to children of single-divorce parents, children of parents who divorce and remarry more than once experience more anxiety, depression, and a sense of vulnerability. They have elevated risks for engaging in precocious sexual behavior, for alcohol and drug abuse, and for committing suicide. They also have more problems with school adjustment, including lower academic achievement and a relatively high risk of falling below the poverty line. Eventually, they have more troubled marriages and a higher divorce rate.     No doubt multiple divorce adversely affects children, especially in the short run. Many will experience apprehension, resentment, alienation, and loneliness, and many will have abandonment fears and fantasies. In some children, many of these psychological changes, like physical scarring, will last a lifetime. Yet an association between the behavior of parents and the personality of their offspring does not prove that one is an important cause of the other. Each may reflect something common to both, which would explain why the parents of temperamentally difficult children have an elevated risk of divorce.     That divorce is an outward indication of hidden dispositions shared by parents and children comports with what twin research suggests, namely that genetic influence partly determines whether marriages succeed. Parents can transmit to their children the genetic qualities of troubled temperament, poor judgment, and self-centered immaturity that figure in the psychology of multiple divorce. It is therefore no surprise that men who divorce their spouses are many times more likely to have a diagnosable antisocial personality, which is a highly heritable personality trait. And the converse: The children of divorced parents are likely to be genetically different from the children of parents who do not divorce. All this can explain the children's emotional and academic problems, even those they display many years before their parents divorce, and eventually their own elevated divorce rate. Genetic factors controlling parental dispositions are likely transmitted to children, who then express these dispositions in their own way and when grown up are likely to behave with their spouses and children as their parents did with them. Can it be assumed that the cycle of family violence from generation to generation is strongly genetic rather than merely shaped by upbringing? No doubt the behaviors of parents and children interact, but the relative power of the underlying dispositions--in particular, the child's disposition--can be resolved only by adoption studies that have yet to be done.     Adults who report having been abused as children often have problems of aggressiveness, antisocial behavior, drug abuse, sexual promiscuity, depression, and poor self-esteem. Many seem to wind up abusing their children or spouses, but why? Perhaps it stems from personal experience with abusive rearing.     Child abusers often report that they themselves were abused as children; estimates range from 7 to 70 percent. We can take the average to be approximately 40 percent, but what does any such figure really mean? Conventional social theorizing offers two possibilities. One is that abusive parents model socially undesirable behaviors, through which their children become badly educated about what is permitted--about violence toward loved ones being a normal part of parenting. The children, as parents themselves, apply that lesson with their own family members.     The other possibility is that abusive parenting creates an antisocial personality that itself causes abusive behavior not only with spouses and children, but with non-family members as well. Through neglect, punishment, and traumatic experience, children learn that the world is a dangerous place. They become vigilant and sensitive, even to the point of minor paranoia. They have problems of aggressiveness, antisocial behavior, drug abuse, sexual promiscuity, depression, and poor self-esteem.     One problem with all this theorizing is that most abused children grow up to be nonabusive parents, and many abusive parents were not abused as children. Another problem is that the child abuser's telltale personality traits and behavior problems are heritable, meaning that, at least for some people, an abusive disposition may express one or more genetic potentials that can develop in many rearing environments, not just the abusive kind. If the abuser potential is at least partly inborn, the pertinent question is this: How much does experiencing suboptimal parenting, including the abusive kind, enhance the risk of becoming a child abuser beyond what comes from inborn potential?     Until we have a definitive answer, anyone's assumption is as good as any other's--for example, the skeptical position taken by neuroscientist J. Allan Hobson, who argues that hundreds of thousands of people have "troublesome problems like anxiety, depression, and neurosis. Society readily assumes that all these people must have had some history of psychological stress or trauma that has caused them to be this way. I say no. They rant and rave, or flit about nervously, or fail to show up for work, or carry Valium in their purses because there is a functional disorder of the state of their brain-mind. They may well have been abused as children or have lost their self-esteem, and this can cause real emotional distress. But it does not cause their actual anxiety, depression, neurosis, or any other of a long list of problems."     Definitive answers to such questions won't come from mere comparisons of parents and children because, as we have seen, correlations cannot disentangle the influence of genes and rearing. Currently, the only way to get something like a definitive answer would be with adoptees, because the correlation of adoptive children with their adoptive parents can estimate influence of rearing, while the correlation of those same adopted children with their biological parents can estimate the influence of heredity.     A simple study would compare two groups: the adopted-away offspring of divorced parents or child abusers and the adopted-away offspring without such biological parents. All the adoptees, of whatever genetic background, are reared from birth by normal parents. Upon reaching adulthood, all become parents, but what kind? Here are two scenarios, each supporting a different view.     In the first, the adoptees of both groups wind up with a comparably low rate of divorce or of abusing their own children. Apparently, a normal rearing prevents either potential, which implies that divorce or abusive rearing engenders the potential. But wait. In the second scenario, only those adoptees born to divorced or abusive parents have the elevated rate usually observed for people who have experienced divorce or who were abused by their natural parents. Now the opposite conclusion must be drawn: A disposition to divorce or abusiveness reflects a genetic potential that, when sufficiently strong, resists the influence of normal upbringing. Parenting, Smoking, and Obesity Cigarette smoking is 2 to 4 times greater in the offspring of smoker parents than the offspring of nonsmoker parents. Conventional social explanations invoke modeling, the promotion of behavior by example, and who provides more salient and compelling examples than a parent? Peers do, but can peer pressure, as parenting, really explain why just some children graduate from the experimental to the habitual use of cigarettes?     Children choose, and are chosen by, peer groups. People with the inborn potential to become smokers may prefer and be preferred by others with the same potential. In this way, inborn potentials could be the real prime movers that shape social conditions. Therefore, why not assume that inborn potential rather than parental modeling or even peer pressure is the fundamental reason for who becomes a smoker and who does not?     A comparison of twins suggests that over 40 percent of the smoker-nonsmoker difference is genetic. Furthermore, the parent-child correlation for smoking is just about 0 if the genetic connection is 0, as when the parent-child relationship is adoptive. When it comes to smoking, it seems that inborn potential strongly constrains not so much who will experiment with smoking, for this is largely a matter of peer pressure, but who will become a smoker, and that parental influence has little or nothing to do with it.     The same argument can be made for body weight. Children and their parents correlate about 25; siblings about 40. Clearly, body weight runs in families, but why? Many people believe that the answer lies in how parents manage food and eating during the formative years, and also what children perceive in their siblings. If that is so, why does body weight correlate 0 for adoptees and their adoptive siblings, while it correlates over 70 for identical twins reared apart? Why is the correlation for parents and biological children the same even if those children are adopted away at birth?     If the body weight of siblings and children can indeed be predicted from the weight of biological relatives but not from the weight of adoptive relatives, then body weight runs in families largely for genetic reasons. Turns out, the heritability of body weight is about 70, meaning that about 70 percent of individual differences from thin to fat comes from genetic differences.     Such findings question conventional assumptions about parental influence. They suggest that, to a greater extent than is often assumed, the usual parent-child resemblances reflect a common genetic quality often more than shared experiences. This could explain why children's psychological problems are more reminiscent of problems their parents displayed as children than of problems the parents display as adults--why, for example, preadolescent behavioral disorders are even more strongly correlated with a father's childhood behavioral disorders than with his adult substance use. Likewise, an 8-year-old child's aggressiveness correlates better with the aggressiveness his or her parent displayed at the same age than the aggressiveness the parent currently displays. Rather than imitation of parents and others, a heritable aggressive self-expressiveness running through the family seems to be the rule.     The idea that certain traits run in families for strong genetic reasons has enormous implications, one of which concerns the historical tradition of blaming parents, especially mothers, for their child's emotional problems and mental illness. Parent blaming reflects a common inclination to exaggerate parents' responsibility for a child's ways of being--likewise parents' inclination to accept responsibility--even where such inclinations are scientifically unjustified. Recognizing the genetic factor in behavior provides a much-needed antidote to this inclination. Copyright © 1999 David B. Cohen. All rights reserved.