 field of psychology. The list of honors and accomplishments associated with her name are endless. I'll give you a taste. She's been elected to the National Academy of Sciences. She is the recipient of the American Psychology Society's William James Fellow Award for Lifetime Achievement, the American Psychological Association's Award for Distinguished Scientific Contributions to American Psychology, the American Psychological Foundation's Gold Medal Award for Life Achievement in Psychology. She has served as the president of the Western Psychological Association, the Society for Research and Child Development, and the Consortium of Social Science Associations. Many of us know Eleanor Maccabee for her pioneering book co-authored with Carol Jacqueline in 1974, The Psychology of Sec Differences. It explored with a critical eye the research evident for differences between boys and girls, and then explained how these differences might develop. More recently, Eleanor Maccabee has collaborated on a large-scale longitudinal study of divorce and its repercussions, including a focus on custody and visitation arrangements four years after the divorce has occurred. And two resulting books are Dividing the Child, published in 1992, and Adolescents After Divorce in 1996, which explore the impact of divorce on children. Her 1998 book, The Two Sexes, Growing Up a Part, Coming Together, revisits the topic of sex differences that she explored in 1974. And actually, as testament to the continued respect with which she has held in her field, this book won for her the Eleanor Maccabee Book Award in Developmental Psychology, which was created in her honor back in 1993. It really is astounding when you think about it that someone who co-authored a book in 1957, Patterns of Child Rearing, is still critically examining how children develop over 40 years later. This tenacity of purpose could only be apparent in an individual who truly cares about the plight of children and the scientific field of child development. Please let me introduce to you Dr. Eleanor Maccabee of Stanford University. Thank you so much, Marie. This is wonderful for me to be back at Gustavus. I was here at one of these wonderful conferences in 1973, I believe it was. I remembered this campus as very shady. And I was startled to come back and see what the tornado had done, but also very impressed at the rebuilding and all the planting of the new trees. And it won't be long before it will look shady again. I must say also that what has happened over the years in the growth of the Nobel conferences, as impressive as can be, and I congratulate all of you who have been involved in this. But now let me talk about the nature of children and their nurture by parents. As you may know, I have been a lifelong student of child development. I know what that is. Yes. See if I can keep from knocking that off again. And particularly, I've been interested in children's social and emotional development as this emerges from the context of interaction with family members and peers. It's from this perspective that I want to consider the nature-nurtured debate. As all of you know, in the middle of the last century, the debate was quite subdued in that the field of psychology was heavily dominated by classical reinforcement learning theories. And child development was seen mainly as the process of learning a repertoire of habits and behavior patterns. Parents taught children to brush their teeth, tie their shoes, cover their mouths when they coughed, tell the truth, respect the property of other family members, and so on. With many repetitions, these things were expected to become nearly automatic, internalized. Child-rearing was seen mainly as a top-down process in which parents and teachers were the ones who set the agenda of what children were to learn and who guided the learning process by serving as models and by selectively providing rewards for desired behaviors and punishments for undesired ones. In those early days, nature was not entirely ignored. People did know about physiological maturation, including maturation of the brain. They knew that maturational processes produce some predictable changes with age in what children are capable of learning or in acting. The results of some of the early twin and adoption studies were also widely known, so that some genetic contribution to individual differences in certain of children's characteristics were acknowledged. And you would find those twin tables showing a comparison of identicals with fraternals in any introductory text. Still, these manifestations of innate factors were somehow kept in a separate intellectual compartment, and the major research efforts were devoted to exploring the connections between aspects of parental treatment of their children and the characteristics that the children had developed. Some connections were found. Alfred Baldwin, for example, in 1955, summarized the kinds of child behavior that were typically found in homes with different child rearing regimes and different household atmospheres. He reported that in homes where the families emotionally rejected their children, the children were often insecure, stubborn, non-compliant, and had adjustment difficulties. But in accepting harmonious and democratic households where discipline was consistent, firm, and fair, the children were usually well-adjusted and cooperative. Baldwin interpreted these findings as showing that the way children were treated by their parents had a major impact on how well-adjusted they became. A chorus of dissenting voices began to be heard in the late 60s and 70s, concerning the direction of effects when connections of this kind are found between parenting and children's outcomes. Perhaps Bell and others said such connections could reflect the effects of children on parents rather than vice versa. This dissent was only one of a number of converging newer theories and new bodies of data that began to influence the nurture nature debate. One important influence came from developmental psycholinguistics, especially Chomsky's claim that there must be an innate language acquisition device that enables children to learn language as rapidly as they do. Extending these ideas, people became more open to the idea that there might be other innate predispositions, some specific to certain domains, others more general, guiding children's development. Much more compelling, however, were the quantitative behavioral genetic studies which adopted more sophisticated methods of analyzing the data from studies of twins and adopted children. As Robert Ploman explained yesterday, behavior geneticists are interested in the variation among individuals and what causes this variation. For example, children differ greatly in how fast they grow and how tall they eventually become. I had decided ahead of time I wanted to use the example of height and I noticed that Robert did it too. And perhaps it's not surprising we both should choose this variable since we stand at exactly opposite ends of it. Do children differ because they are, some are raised in healthy environments with plenty of nutritious food and good medical care while others are undernourished and neglected? Or is it simply that some have tall parents and others have short parents and height is encoded in the genes that a child inherits? Obviously it could be some of both and behavior geneticists set themselves a task of quantifying how much of a variation in this or any other trait is due to heredity and how much is due to environment. While behavior geneticists certainly can ask and do about the heritability of physical characteristics such as height or weight or the incidence of certain diseases, they have focused extensively on psychological traits that is on the relative contribution of nature and nurture to individual variation in such attributes as intelligence, calm or excitable temperament, aggression or special talents such as musical ability. As you probably know, their analysis starts with the additive assumption. Heredity and environment they say together account for 100% of the variation in any human characteristic. Their logic then is if you can compute a quantitative estimate of the amount of variance accounted for by heredity, all you have to do then is subtract that number from 100% and you have a good estimate of the magnitude of environmental effects on this particular characteristic. Let me say at the outset that I think this basic assumption behind their logic is flawed. Heredity and environment I believe are by no means entirely additive. They are also interactive and I'll come back to that point. For now, however, let me continue with the behavior geneticist's reasoning. Once they have estimated the environmental contribution, they subdivided into two components which they call shared and unshared environmental effects. Suppose that fraternal twins turn out to be quite similar more so than you would expect from their shared genetics or that adopted children are quite similar to their adoptive parents and their biologically unrelated siblings. These things would point to an effect of the environment that they share. It could mean that both children and a family regardless of how biologically related they are are put at risk if there's a great deal of conflict between the parents or if the parents are harsh and abusive toward the children or the similarity between siblings could mean that children benefit from having parents who are well educated or parents who have a harmonious relation with each other or who are warm and supportive toward the children but also effectively set limits. One of the most surprising findings from behavior genetics as Dr. Ploman showed yesterday is this, with respect to many characteristics, adopted children don't resemble the unrelated siblings they grow up with any more than they resemble children in other families and fraternal twins are surprisingly unlike one another. On many characteristics, fraternal twins or other siblings who differ in age appear to resemble each other only about as much as their genetic similarity would call for. The unlikeness of siblings has been very firmly established by Robert Ploman and his colleagues and I think the unlikeness of boys and girls and the same family or sisters or brothers would come as no surprise to most of you if you have raised more than one child. A number of people have interpreted these findings to mean that what parents do cannot matter very much. They can't do much to improve the functioning of a child who starts out with bad genes or do much to damage to children who start out with good genes. Consider the things we think of as risk factors for healthy development such as single parent families, poverty, bad neighborhoods, mental illness or drug addiction in the parents, harsh unsupportive parenting. The low shared environment effects have been interpreted by some to mean that these things actually have much less impact than we have supposed. Some of these interpretations come from respected scholars. Others come from more popular writings aimed at non-academic audiences. Ploman yesterday mentioned a book that I wanna say something about now. This is the 1998 book by Judith Harris called The Nurture Assumption in which she attacked the idea that the way parents bring up their children is effective in shaping them toward the outcomes that parents want. She based her case largely on the kind of behavior genetics findings that I have been describing, but also on a claim that the studies of parenting itself had shown only very weak connections between parental parenting practices and child outcomes. And as an alternative to parenting effects, she emphasized two things. First, the influence of the child's own genetic predispositions, which she claimed largely control the way parents treat their children. And second, that the power of peer groups to socialize children may be greater than the power of parents, at least when it comes to traits that will be enduring. Harris's book made quite a splash. There we go. This is the cover of Newsweek and based mostly on Harris's book, they raised this question, Do Parents Matter? Very provocative kind of title and that they hope would draw some readership for them. The New Yorker, which is usually very careful, came out with a very glowing review of the book with the same title, Do Parents Matter. These articles were essentially friendly reviews of Harris's book. Why were the media so fascinated by it? Partly, I think, because it's always fun to puncture the pretensions of experts and reveal their feet of clay. But also, Harris's message was seen as a life preserver for parents. The message was, you don't have to blame yourself if your children are not doing well. But there are also political implications. If parents don't matter, there's no reason to spend public money on programs that try to help parents be more effective in socializing their children. There have been some strong reactions to all of this from serious scholars. Child development scholars certainly do not reject some of the primary claims of behavior genetics. Twin and adoption studies do indeed show that individual variation among children is strongly affected by their genetic predispositions. But it's important to understand just what it means and what it doesn't mean. When we learn that some trait is highly heritable, does it mean that it's not influenced by environment? Not at all. Consider height once again, for example. Tall parents do have tall children and the children of short parents are usually short too. The heritability coefficient derived in twin studies is in the 90s, leaving very little room for environment to have any effect. Yet we know that Japanese American children are on the average nearly four inches taller than their grandparents living in Japan, from which we see that a human characteristic can be both highly heritable and highly susceptible to environmental influence. We get a similar picture from population changes over time. And here again, I'm using the same example that came to Bob Plummer's mind. We know that the shape of our bodies is to some extent heritable and that fatness does run in families and at the same time, there is a substantial increase in obesity in our country at the present time. And it's not happening because our genes are changing. It's a pure environmental effect superimposed on whatever dispositions to be fat or thin that an individual may have. It means that there may be an increase in the average level of a characteristic in a population without necessarily changing the rank order of individuals in terms of their individual vulnerability to obesity. I believe this is what we are seeing in studies of adopted children where it often turns out that some trait, let's say their IQs, is significantly correlated with the IQs of their biological parents, not their adoptive parents. At the same time, the average IQ of adopted children is higher than that of their biological parents. I've drawn a schematic here, this isn't real data. It says, though, each child gets a bonus in achieved intelligence from growing up in more advantaged circumstances than the biological parents would provide. So here's the biological parents IQ and here's the child's. And you will see that the children maintain the same rank order as their parents did. They're biological parents, but they are all higher than their biological parent. A large environmental intervention then, such as adoption, can be like a rising tide that lifts all boats. In the example illustrated here, the original rank order of the children based on their biological origins is not disturbed even though their mean score has risen. But of course, in real life, not all children will be benefited to the same degree and the rank order will be somewhat rearranged. Still, a significant correlation between the children and their biological parents can still be maintained while the children as a group are doing better than their parents. Please understand, all of this is not a criticism of behavior genetics. They study how and why a trait varies within a population, not how or why the average of the whole group goes up or down. I only mean to call attention to the fact that we can seriously underestimate the importance of environmental effects if we pay attention only to individual variation and not to mean changes. Consider that almost all parents succeed in teaching their children to tie their shoes. That's a strong parenting effect, even if it cannot be revealed by studies that rely on variation and correlation. In the popular treatments of nature-nurture issues, claims about the strength of genetic influences and the weakness of environmental ones can be quite sweeping and they downplay important exceptions where influences of the home environment children share emerge quite clearly. Standard behavior genetic analyses reveal quite strong shared environmental effects for several traits, specifically anxiety in children, aggression, and the quality of a child's attachment to a parent. For these traits, then, certain aspects of the in-home environment that siblings share do indeed influence siblings in a similar way, pointing to an effect of the family environment rather than to the child's genetics. Thus, if a child shows an insecure pattern of attachment, the root of this problem are more likely to be found in a parental characteristic such as maternal non-responsiveness than in the child's genetic predispositions. And siblings are likely to resemble one another regardless of how genetically related they are. But now let us go back to the large body of twin and adoption studies which claim to find strong genetic effects and weak effects of the shared environment. As you might imagine, some strong critiques are coming from knowledgeable scholars concerning the theories, methods, and findings and the interpretation of results of behavior genetic research. First of all, critiques have raised questions concerning how much we can rely on a heritability coefficient that emerges from any given twin or adoption study. While there may be no doubt that there is a genetic contribution to an individual variation on a given trait, it can be difficult indeed to determine reliably how large this contribution is. For example, different studies of aggressive behavior have come up with estimates of variability that range from near zero to over 70%. The size of the heritability coefficient depends on many things. For example, twin studies typically yield higher heritability estimates than adoption studies do. Estimates of the heritability of a given trait depend on the cultural subgroup being studied, on the period of time when the study is done, on the sex of the children or adults in the sample, and their age at the time of measurement, and most especially on how wide a range of variation there is on the traits being studied within a given sample of people. The size of heritability estimates depends too on the source of the data. That is, parent ratings of their twins characteristically typically show higher heritability than do objective behavior observations of the same children on the same trait. And what this probably means is that parents of identical twins think their twins are somewhat more alike than they really are, and parents of fraternal twins think they're more different than they really are. If estimates of heritability for a given trait are variable and unstable, it follows that the size of the environmental effects estimated by subtracting their heritability estimate from 100% must be indeterminate as well. But of course there are certain characteristics where many studies come up with a similar heritability estimate. Let's assume for the moment that we do have a reliable estimate of heritability and have subtracted it from 100. We now have an estimate of total environmental effects and this reflects some total of any and all environmental effects including parenting and other aspects of the home environment, but also neighborhoods, schools, daycare centers and so on. We can subdivide this total environmental effect into shared and unshared effects. As I have said earlier, traditional behavior genetic analysis shows that many of the traits that have been studied have very little effect of the shared environment but a substantial effect of unshared environment. This might seem to mean that nothing about the home environment that children share where they grow up together in the same house with the same parents is having any effect, but it's important to be as clear as we can about the meaning of unshared environment. The term can refer to the fact that even when children do grow up in the same family, they can be exposed to different environmental inputs. An older child, for example, has to deal with a younger sibling while a younger child is continually exposed to a sibling who is larger and more competent. Within a given family, one child may be especially favored or one child may become the family scapegoat and receive more negative parenting than the other children. We should note that children are very sensitive to differential treatment by their parents and sometimes it's more the difference between siblings in the way they are treated than the absolute level of positive or negative parenting that they receive that is related to their adjustment. But a behavior geneticist would ask, why is one child in the family singled out for especially positive or especially negative treatment? Might it not be something about the child? Whether the child is especially attractive looking or especially homely, whether the child has an easygoing temperament or a jumpy irritable temperament that these things may trigger positive or negative reactions from the parents. Of course, this is not only a plausible explanation, it has been shown that these evocative effects of children on their parents' reaction, that has been clearly demonstrated. For example, we know that if an adopted child carries a genetic risk for antisocial behavior, that is the child's biological parent has a history of antisocial behavior, that child is more likely to receive negative coercive parenting from an adopting parent than an adopted child who carries no such risk. Presumably this difference arises because the genetically at risk child misbehaves more often and calls for stronger parental reactions. On the basis of such data, some geneticists are comfortable about assigning any correlation between the child-rearing activities of a parent and the behavior of a child to the child's genetics. I am not comfortable about this because it ignores the return feedback loop. And indeed, in the study I just described, negative coercive parenting had its own effect on the adopted children's adjustment independent of the child's genetics. The fact is that whatever the child elicits in the way of parenting responses from a parent, that parental response in its turn affects the child over and above the child's starting point. In conflictual situations, one can sometimes see an escalating cycle where parent and child each trigger higher and higher levels of anger in the other or a dampening cycle in which each in turn makes a positive move that helps to calm the other down. Parent-child interaction is highly reciprocal. When children initiate a bout of parent-child interaction by doing something that evokes a parental response, ignoring any effect of the ensuing feedback from the parent to the child is to underestimate an environmental effect. But this all simply underlines the fact that different children in a family can indeed be exposed to different home environments even though they're growing up in the same house. And presumably this can contribute to differences in the way two siblings develop. In some ways though, I think it's more interesting when some environmental effect event that the children actually do share such as parental divorce has a different effect on one child in a family than on the other. One child might become angry over this parental divorce and begin to act out aggressively toward family members or peers while another child in the same family might become sad and depressed or withdrawn. Both are behaving differently than they would have if the divorce had not occurred and therefore the divorce is having an effect. Oh, but ah, says the behavior geneticist, that's an unshared environmental effect. Anything that makes siblings more different than more alike, we call unshared. When behavior geneticists point to small effects of the shared environment, along with large effects of the unshared environment, some readers have interpreted this to mean that aspects of family life that siblings obviously share such as a divorce or whether they're living in a two-parent or single-parent family or the level of their parents' education, that these things cannot be affecting the children's adjustment because twin and adoption studies have found very little effect of shared environment. No, this is not what the behavior genetic analysis means. It only means that if these shared events do have an effect, the effect is different on different children in the family and doesn't serve to make the children more alike. Clearly, we need to think of two kinds of cases in which behavior geneticists have labeled as unshared environmental effects. The case where two children in a family are actually exposed to different environmental inputs as when their parents treat them differently and in the other case, when they share exposure to the same environmental input, such as high parental conflict, but perceive it differently and or react to it differently. I believe it's unfortunate that these things have been called by the same name because it fosters the idea that environmental effects, events that the children in a family are actually all exposed to are having no effect. Whereas, in fact, a shared event can have a distinct effect on one child but not the other or on both children, but in different ways, and these are clear environmental effects that emanate from environments that children share. They need to be acknowledged as such whether or not they make children more alike or more different. Critics remind us that in traditional twin and adoption studies, there are no direct measures of environmental conditions. Environmental estimates are merely residuals, what's left over after the genetic influences have been subtracted out. Critics have been urging that we need to include measures of environmental factors directly and in recent years, there has been a substantial accumulation of studies in which direct measures of the rearing environment, including the quality of parenting, are available for children whose degree of genetic relatedness to their siblings or their parents is known. One of the most intriguing of these is a recent study by Dieter Deckard and colleagues. He's a young man at the University of Oregon. They selected 61 pairs of identical twins. It's not easy to find 61 pairs if you just from your community, but they did. All of these children were close to three and a half years of age. They observed each twin separately in interaction with the mother. Part of the time, the mother was simply playing with a child but in addition, the mother guided the child in using a toy that calls for cooperation between mother and child. Have any of you seen this toy called Etch-a-Sketch? Each person turns a knob in order to make a line go in a certain way. We know that on the average, identical twins are treated more similarly by their parents than our other sibling pairs. Even so, it turned out that there were differences in how the mothers treated the two children. In most families, one twin was treated more positively with more praise and warmth while the other was treated more negatively with more criticism and harshness. Mothers did differ on how great this discrepancy was. Obviously, these differences in maternal behavior could not have been evoked by any differences between the twins that had their roots in genetic characteristics. But here is the punchline. The differences in maternal treatment were related to how different from each other the twins became. The twin treated more negatively, developed more behavior problems and was less compliant and responsive toward the mother than the more positively treated twin. I regard this as an unambiguous demonstration of a parenting effect that could not have been driven by an initial predisposition of the children involved. There are studies now that include both identical and fraternal twins, siblings and half-siblings or adopted children, which also include measures of parenting or other important aspects of the children's home environments. These studies have continued to show us substantial genetic effects and they show environmental ones as well. But more interesting, I find, is the fact that the effect of an environmental event may differ for children with different genetic predispositions. In other words, nature and nurture interact. As an example, let us consider a large-scale study of adopted children done in Sweden. This is by Bowman and colleagues. They studied two groups of adopted children. One group had a biological parent with a criminal record. The other group did not. One could say then that one group may have carried a biological predisposition to engage in criminal activity. Did this predisposition manifest itself in children who had never lived with their biological parents? Yes, it did, but only in some of them. Some of the children engaged in acts of petty criminality as they grew toward adulthood. But whether they did so depended only slightly on what their biological parents had been like if this factor was considered in isolation. 12% of them did show petty criminality. As they grew up, even though they had no, when they had a biological risk. The children's behavior depended slightly on conditions in the adoptive home as well. If one of the adoptive parents became depressed or alcoholic or drug addicted, or if there were other risk factors present such as divorce, this did not affect all children equally. It had very little effect on a child's petty criminality if the child carried no biological risk. But the strong effects were seen when both biological and environmental risks were present. For these children, something like 40 some percent, I don't have the side labels on it, did engage in petty criminality. A very similar pattern of results was found in a study of adopted children carried out in Finland in which some of the children had a biological mother with schizophrenia while others did not. This study is especially interesting because schizophrenia has been very firmly shown by Seymour Kettie and others to be quite highly heritable. That is to say, it has a 50% concordance, as I think Bob Plummer's slide showed yesterday. In the children, schizophrenia was manifested almost exclusively in children who carried both the biological risk of the disease and were exposed to an adverse rearing environment in their adoptive homes. Children who were raised in a favorable environment seldom manifested the disease. In other words, an environmental trigger was required for this genetic factor to be expressed. Would you attribute criminality or schizophrenia in these cases to heredity or environment? Obviously, it's to the combination of both. Traditional behavior genetic analysis doesn't take such interactions into account. These interactions are simply omitted from the calculations of genetic and environmental effects resulting, I believe, in overestimating the strength of both nature and nurture as separate unrelated causal elements. By now I suspect some of you must be thinking, well if we want to know whether the environment's children grow up and make a difference in how they develop. Surely there must be some way we can study these effects more directly without having to make the inferences that are involved in twin and adoption studies. Well yes, there are ways. In science, the gold standard is doing an experimental intervention in which you arrange for one group of children to be treated in one way and another group to be treated in another way. Then we can see whether they differ later on in ways that are related to the treatment they received. Of course, you have to make sure if you can that the two groups didn't differ in the first place and as most of you know, the best way to do this is to assign people randomly to the different treatments or to a no treatment control group. If you have enough cases, you can be sure within known limits that the average scores of the people in these groups didn't differ initially on any attribute that might confound your results. Obviously we can't assign children at random to different sets of parents or different kinds of home environments. However, we do sometimes find an experiment of nature that comes close. Researchers in France, and the leading person in this group, his name is spelled D-U-Y-M-E. I don't know how it's pronounced in French. I think doom, doom, something like that. He took advantage of a kind of natural experiment with a group of children who were adopted at a later than usual age. These children had all been removed in infancy from their parental homes because of abuse or neglect and had subsequently lived in foster homes or institutions. At about the age of four, these children were given IQ tests. The sample of children chosen for study included only those whose IQs were in what we call the dull normal range. Their average IQ was 78. Between the ages of four and six, they were adopted into homes that varied considerably in terms of the resources the family could provide for the children. Their IQs were measured again a number of years later when they were between the ages of 11 and 18. And here is what had happened to their IQs. The group that went into the most advantaged families had gained 19 points in IQ. The next group, 12, I believe, and the lowest group, eight. Please note that the three groups of children who went into three different kinds of homes didn't differ initially at the time they were placed for adoption. There's no reason to believe that the more promising children were placed selectively in the more affluent homes. This seems to me to be quite a clear demonstration then that the resources of adopted families make a substantial difference in their children's outcomes, even in something like IQ. And that would be normally a shared environmental effect if you use a different definition. Even in something like IQ, which is thought to be highly heritable. Of course, we don't know exactly what it is about a high SES home that was beneficial to these children. Was it that the parents were better educated, or providing more stimulating environments or more sensitive caregiving, or that they lived in better neighborhoods and sent their children to better schools? All we can say is that it was something about the environments these parents provided that was beneficial to the children. Can we do experimental work that focuses more specifically on parenting itself? So even though we can't randomly assign children to different kinds of parents, we can establish an intervention program designed to help mothers and fathers improve their parenting. Then we can randomly assign some parents to a control group that remains untreated on a waiting list, perhaps. And others to an intervention group which receives parenting training of one kind or another. If there is an effect of the intervention on the way parents deal with their children, then the next step is to assess the children to see whether any improvements in parenting practices have proved beneficial to the children. Such studies are difficult indeed to do, but there are a few instances in which brave and careful researchers have actually done it. An example comes from work in Holland, Van Den Bohn located 100 low income Dutch families whose children had been identified as irritable when they were only two weeks old. Beginning at age six months, half the families were randomly assigned to an intervention program in which the mothers were trained in attentive, sensitive, responsive caregiving while the control group received no such training. The training occurred over a three month period when the infants were between six and nine months of age. At age one year, the trained mothers were indeed more responsive and sensitive in dealing with their irritable infants. And their infants were much more likely to have developed a secure attachment relationship with their mother than the infants in the control group. The study was carried on, the children were followed and at age three, the intervention infants continued to be more secure and were also more cooperative with their mothers and more positive in their interaction with playmates. There is now a considerable body of other studies that involve random assignment to a program in which parents are worked with in a variety of ways to try to improve their parenting while a comparison group does not receive this treatment. Some of these are small clinical trials, others are more comprehensive. For example, there's a large research program at Berkeley done by Phil and Carolyn Cowan in which parents are randomly assigned to one of two kinds of intervention groups or to a control group. The Cowans have shown that an intervention focused on the quality of the marital relationship can have a strong impact on the kind of parenting practices that a couple employ. They have also shown that improvements in marital and parent-child relationships are associated with improvements in the children's academic performance and a reduction in behavior problems that the children show in kindergarten, first grade and fourth grade. So it seems to me, let me see how I'm doing. Well, let me add then, in recent years Head Start programs have been expanding to include infants and toddlers and three-year-olds originally they were mainly four-year-olds. Some Head Start programs are entirely home-based in which visitors provide support and training to the mothers. Some are entirely based on working with the children in childcare centers and others are mixed involving intervention with both the mothers at home and the children at a daycare center. The results of studies in which children have been randomly assigned to a Head Start program or a waiting list are shown showing that children make the best progress in the mixed programs and indicating that providing support to the parents adds a significant increment to the benefits children derive from being enrolled in a Head Start center. It seems to me that if we add up the research that I've been telling you about, it offers clear evidence that children's home environments can and do make a significant difference in how well they fare. In view of these findings, you may wonder why there are still voices emphasizing genetics so heavily. First of all, we know, I believe that both genes and home environment can play a substantial role, but we may have overemphasized genetics partly because of the excitement over the race to complete the mapping of the human genome. Scientists everywhere have hoped and expected that now that this process is complete and the human genome is unlocked, it will be possible to find clues to the causes and processes underlying a vast array of human diseases and psychological vulnerabilities, as well as an array of human skills and exceptional talents. Our new knowledge from molecular genetics has paid off notably in a number of instances, but some of the findings have been sobering too. There's a long path between the gene and its phenotypic expression and many factors intervene along the way and the effect a gene will have can depend on certain other genes with which it may or may not co-occur, but most important for our concern, some genes need a specific environmental trigger before they are expressed. Here we see additional evidence that nature and nurture act jointly, not as separate cause of entities. Gene environment interactions loom large in the findings from molecular genetics and we saw that yesterday in Eric Candel's wonderful talk. Our current heavy emphasis on genetics has also been fed by the upsurge of work on the human brain on how the brain functions as we think, perceive, remember, plan, and behave. Neuroimaging has enabled us to see which portions of the brain light up when we are doing these things and there are a number of exciting studies in which connections have been found between genes and brain functioning. For example, people respond, people differ with respect to which allele they carry on the serotonin transporter gene. It turns out that which allele we carry affects the way a locus in our brains, that is the amygdala, reacts to a stressful environmental input. We can infer from this that certain emotional reactions that are mediated by the amygdala will differ according to which allele on the serotonin transporter gene an individual carries. Naturally, findings like these lead many people to infer that what happens in our brains is hardwired, genetically controlled, and that behavior linked to brain functioning must be unlearned. Nothing could be further from the truth as we learned yesterday. Eric Candel told us repeated experiences in our environment bring about change and growth in synapses as long-term memories are stored. He gave us the example of how playing a stringed instrument brings about changes in a part of the motor cortex that controls the movements of the left hands, and the number of neurons involved and their connections grew. And I had a flashback to what he told us about when Gaul thought that the parts of the brain that are active in romantic ideas, for instance, swell, and you can feel them on the outside of the head. Well, it turns out that if you play a stringed instrument, something in your side of your head is getting bigger. It's just that you can't feel it as a bump. So the inputs we get from our environment, including being trained and educated in certain ways, help to shape how our brain capacity is deployed and what connections will be formed as learning occurs. And of course, there are constraints by the brain's own architecture that will determine also or produce constraints on what kind of changes these can be. And both of these things change with age. The brain matures so that there is an increasing participation of the frontal cortex in performing certain tasks. At the same time, a vast accumulation of knowledge is occurring based on cumulative experience with the environments and our brains store this knowledge and make use of it as we make our way through the enterprises of our daily lives. So the nature of our brains and the environmental inputs provided to our brains jointly act not as two separate causal factors. As I said earlier, the nature and nurture debate recently took the form of claims that children are strongly influenced by their genes but not much influenced by the way they're brought up by their parents. So far I've been presenting evidence that children's home environments do indeed matter, but I haven't said very much about parenting itself. First of all, let me say that I believe the motivation to care for and protect our offspring is deeply encoded in our genes. But what is required of parents in order to carry out this biological imperative has changed greatly over the history of our species. And the process of raising children always co-occurs with an array of other necessary tasks. In the modern world, parents differ greatly in how salient child rearing is in the total mix of their activities and goals. We now have an extensive body of research in which parents and children are observed interacting with each other. We find that from infancy on, interaction is a highly reciprocal process. Parents and children typically respond to one another's moods with congruent moods. Each is more likely to react to a cross or complaining initiative by the other with a negative response than a positive one and positive approaches tend to engender a cooperative response. Parent-child pairs have hundreds, thousands of interactions over years of time. Each pair develops a unique relationship in which each knows intimately what to expect of the other, how fully the partner can be trusted, when to back off, and when to make an attempt to influence the other. Indeed, both parents and children sometimes react to one another more in terms of their stereotypes about what the person is likely to do than to the specific thing that the partner has just done. But the parent-child relationship cannot be symmetrical in the way that it could be between two intimate adults who develop a joint agenda and try to act in the other's best interests as well as their own. The parent-child relationship is essentially unbalanced because the parent has responsibility for the child's welfare and socialization while the reverse is not true. So that of necessity, the parent and child frequently don't have the same goals and parents of necessity have more authority. But also, the parent is a grown-up, and this usually, although not always, means that the parent has better impulse control than the child, has better ability to think of long-term goals than only responding to immediate pressures, better ability to understand the child's point of view and the child's needs than the child has to understand the parents. So any parent-child relationship is a melange involving differential authority, unequal resources, but also reciprocity. Families differ greatly in how these ingredients are combined. There are societies in which parents actually seldom talk with their children, and where children are not bargained with or their wishes consulted. Our society is one in which these things do turn out to matter. A modern point of view about parenting is that it still requires parents to teach necessary habits and impose restraints when the child's behavior calls for it. But especially important is the development of what might be called a good socializing relationship between the parent and child. And relationships are more than the sum of their parts. They develop characteristics that aren't exactly the same thing as the characteristics of each of the participants. In a relationship, parents are sensitive, in such a relationship, the parents are sensitive to the child's emotional states and point of view. They manage to make interaction with parents interesting and pleasant and helpful enough on the whole so that the child comes to adopt a positive attitude toward cooperating with the parent's agenda. When parents lay down the groundwork for a good relationship of this kind while a child is young, they have an easier time maintaining mutual trust and communication with the child at a later age. We know, for example, that a parent's ability to keep track of an adolescence, whereabouts and friends and activities depends heavily on the adolescence being willing to talk openly to the parents about these things. And the child's openness depends on a history of a good socializing relationship. There is now reason to believe that a good socialization relationship may be a better predictor of the child's eventual adjustment and accomplishments than the behavioral qualities of either parent or child taken individually. Here are a few main thoughts that I want to leave you with. First, I want to make a bow to the behavior genetics movement. It has been very helpful in forcing us developmentalists who study parenting to come to grips with the differences among siblings. We always studied one child in a family. We never really looked at this matter of sibling differences. Next, I would stress that both genes and environment can have powerful effects at the same time. They are not a zero sum game. Then I want to urge that genes and environment are not two separate forces. They act together at every step of the way from the moment of conception to the way we function as adults. We ignore their interactions at our peril. And my final point is that influence between a child and parent is necessarily mutual. Claims have been made that when parents treat children differently, this reflects the fact that children who are genetically different in temperament or intelligence or whatever, elicit different parenting from their mothers and fathers. Is this true? Of course it's true. Does this mean that their interaction is all a function of the child's genetics, that how parents respond to their different children has no effect? Of course not. The continuing relationship between two people who interact frequently with one another is always reciprocal. Each constitutes an important element in the social environment of the other and each must adapt to the other. To assert that the parent-child relationship is a one-way street with only the child having influence seems to me bizarre. But to insist as I have been doing that influence flows both ways is not enough. We need to know, what we know now is that we have to pay close attention to the emotional valence of these interactions. Almost every parent-child relationship contains some elements of conflict involving pressure and criticism or discipline from the parent's resistance or sullenness or delay on the part of the child. But most parent-child relationships also contain elements of comforting, humor, helpfulness, tenderness, play, and fun. What matters is the balance between the two, especially that the negatives should not outweigh the positives. This is the heart of a good socializing relationship. Both parent and child contribute to it and both benefit from it. Thank you. At this point, I would invite our panelists to come up for some question and answer. If you have questions for Dr. McAbee, please write it on a piece of paper and send it to our ushers and they'll bring them up front. Dr. Plowman here. Ah, there we are. Okay. We're one more here. All right. Your name isn't up there now. That's it. I bet people know who you are. Why don't you move over? Yeah. Well, let's begin our question and answer session. Would anyone on the panel like to offer a comment or have a question? Dr. Rappaport. Well, I think what's been so exciting about the last decades of research have been that there's been more elaboration of the complexities starting out with the assumption that of course both genes and environment are important. And so my comment would be that there are several interesting studies that I think make the complexity even more hard to understand and that I think deserve to be addressed in a debate and attention that we'll go on past all of our lifetimes, I'm sure. One is some studies. There are several, but one is by Dr. Plowman, in fact. Forgive me for presenting this. In which it's shown that the cognitively adopted children as they reach adolescence resemble their adopting, their biological parents more than they did when they were younger. That is, as they get older, they resemble their biological parents more, suggesting that there are certain genes that probably act later in development. Similarly in child depression, there's some evidence that children with a very early onset, pre-pubertily, are less likely to be genetically loaded for depression than children with adolescent onset. I'm just saying that there's a whole host of studies developmentally that suggests that it's far more complex and that certain genetic influence can in fact come in later. This is not a contradiction so much as just to say that I think this fabric of the tension between the two needs to be examined over a long period of time and perhaps picking each variable specifically as you go in anticipating where the balance and this fabric of influences is going to play. Yes, I think that is important and it seems to me that some parenting effects, much depends on what your outcome measures are. If you look, for example, at what kind of a career a child takes up, parents have influence on that and that can last for a lifetime if you train yourself to be an engineer and engineer all your life because your parents got you into engineering school. That's a heavy lifelong parenting effect but I think that there are some of the things of the sort you're describing that are perhaps deep temperamental things which have a heavy genetic component that parents have more control over their manifestation while the parents are there in the same house with the child and the child then gets released out into a larger environment with a whole array of other important controlling things. You also have, I believe, the fact that some things that parents have an influence on like the nature of attachments within the family may not manifest themselves so very much when the children are outside their natal family but may show themselves once again when they marry and have children. We're only beginning, I think, to have any information about those delayed effects or the variations in the time course of a person's life when influences are the most important. Dr. Bloman. I'd just like to say how much I enjoyed the talk and really agree with almost everything that you said which is an amazing change really in developmental psychology where it isn't a matter of a confrontational issues. I thought that Professor McAvie presented the results in a very fair and honest way. I mean, it wasn't a polemic at all. It was really honest in pointing out the problems with the research that's been done so far and that she agreed genetic influence is important, that non-shared environment is important with some interesting interpretation issues about when the shared things become non-shared which is, Judith just said, is an important direction for future research. The importance of GE interaction which I didn't have time to talk about yesterday but I agree that those are issues that we really wanna get at and especially GE correlation, that genes and environment are correlated and like other correlations, it's not genetics or environment, it's both and that's important to understand. And I would even agree with her last point that there's too much emphasis on genetics now. I mean, especially in the medical sciences or psychiatry, I mean, you'd think schizophrenia is a genetic disorder but you know it's not. The concordance of identical twins is 50%. The news 20 years ago was the concordance of identical twins is 50% as compared to 1% in the population. That's part of the story of genetic influence but it doesn't mean it's all genetic. If there is significant genetic influence, it doesn't mean it's a genetic disorder because the story now is that identical twins are only 50% concordant. That means half of the time these clones are different in terms of their schizophrenia outcome and you can't explain that any other way than to say it's environmental and who's studying the environmental component of that or even worse, yesterday I mentioned the example of breast cancer where you'd think breast cancer is a genetic disease given what you hear about these two genes that have been identified but 85% of the time identical twin women are discordant for breast cancer. One has it and the other doesn't life term. So by far the most important cause of breast cancer is not genetic and people aren't studying that because of all the excitement about the human genome project and DNA. So I would really, second what I think is the most important message from her talk is that both nature and nurture are important and we ought to be studying both and especially we ought to be studying both together. Dr. Kagan? Yeah. Nope. All right, we'll go to the questions then. Are you ready? How about? Here we go. Is there evidence that poor parenting done early in a child's development can be corrected later in life? I think it depends upon how much later you're talking about. I certainly do know that some of the intervention studies have dealt with parents of say six-year-olds where there's a particularly interesting one in Oregon where the mothers are all single parents and they're all dealing with a six-year-old boy who has been acting out and defying her and the training is to teach her how to not let herself be coerced by that child without becoming violent toward that child. And they are making inroads on this with mothers who have had this child for six years. Whether you can heal parenting and the parent-child relationship when you've had bad things happening for 13 years and the child is now an adolescent, I don't know. Does anybody else know? We look to you for expertise in this. All right, here's a more technical question. Does restricting the range of environmental effects such as similar nutrition within a culture produce spuriously large estimates of heritability? Goes on as the example of changing height as a function of diet in Japanese kids in association with large heritability, point nine, raises the possibility that a restriction of range on environmental factor, similar nutrition in a culture, may lead to spuriously large estimates of heritability effects. Is that a possibility? I think it is and what I'm thinking about is the fact that in many studies of adopted children, all the families that they're adopted into are pretty well-functioning families so they wouldn't have been selected as adoptive homes so that you get a restricted range of environments and that probably does lead to an underestimate of possible environment effects but I'd like to hear your comment about that. Yeah, I think at the technical level it's interesting. The basic idea is that these are descriptive statistics, they describe what is in a population with that population's environmental and genetic variation. So if you change that environmental variation by restricting it, then the differences that are left are relatively more due to genetic differences but it's actually a more profound issue. If you think of say intelligence for example and genetic influence on intelligence, what this means is that if you restrict the range of environmental variation, you increase the extent to which the differences that are left are due relatively to genetic factors. So the bizarre thing is say in terms of educational opportunity, if we actually did equalize educational opportunity for all children, we made it the same, we'd basically be restricting that aspect of our environmental variation and thus creating higher heritabilities as a result of environmental equalization. So it's an interesting notion to consider both at the technical level, as Eleanor says, adoptive families aren't representative necessarily of all families as a whole and they may not be representing all the environmental variation but then also at this larger conceptual level of it makes us think again perhaps about some of the things just assume are necessarily good. I'm not saying I don't have any particular value on this but it is interesting to think that if we equalize educational opportunity and make that the same for everybody, the differences that are left are more a function of heritability. Another question, could you comment on situations in which a young child acquires power to control the parent-child relationship? What causes this imbalance and what would be the results if it occurs? It's not a good thing. And what causes it I think mainly, oh it has a variety of elements. Some parents honestly are reluctant to be in authority over their children. It's not consonant with some of our ideology. We have democratic views and we think everybody has a right to be heard and sometimes we carry that a little too far when we're dealing with a two and a half or three year old who doesn't have as much sense as we have. And we are sometimes not, don't have somehow some mothers lack the self-confidence to say, listen I know what's right and I have the right to tell you what to do. So what you sometimes find is mothers or seldom fathers, more often mothers, who simply back off. And we have done studies where we have looked over a period of time with mother-child pairs and we find that a child who's difficult and quickly gets angry and if the mother backs off six months later that child is beginning to be in control whereas if she doesn't back off the child is cooling down. So it's a matter of training mothers somehow if they need it, most mothers have the good sense not to let themselves be coerced by a child. But the earlier you can get at this and hold your ground without getting angry and punishing severely, that is the trick. All of your examples of mother-child interactions, cohort, see all of your examples of mother-child interactions involve mothers. What about the importance of the father-child relationship? I think that fathers are infinitely important. I also think that mothers and fathers normally, that's not the right word usually, have a different role in the family and a different relationship with children. Nowadays fathers are clearly becoming more involved in the day-to-day interaction with children than they used to be. There is also a companion phenomenon called father's flight from fatherhood. Fathers are spending a larger and larger proportion of their lifetime in households with no children. Divorce is partly responsible for this single parent. Households are almost entirely mothers, not always but very largely so that the father is living somewhere else, he's not with his children, marriage is delayed and fewer children are born to families. The drop in fertility is absolutely astonishing. I was just recently hearing about what's happened in Quebec where the drop has been from seven and a half children per family down to 1.7 or so. That's the largest drop anywhere in the world and Japan the drop is huge. So both sexes in a sense are withdrawing from parenthood. Marrying late, having fewer children, trying to keep other enterprises in their lives going more actively, especially for the women. You may, I have seen for a long time as a feminist the objective that men and women should have exactly the same responsibility for involvement with children. We women have made fantastic gains over the last 50 years as I have seen them in terms of openness of the workplace at all levels to us and the kind of education that's available to us. We have not made that kind of gains with respect to the balance between caregiving by mothers and fathers. We've made some gains, but really they've been very slow. It's a big social experiment and we still don't know how far that movement can go. I have a gut feeling that women are gonna continue to be more involved with the nurturing and care of very young children. I noticed that when women are freed from the home by sending their children to a childcare center, who's taking care of the children in the childcare center? It's women and that is the case the world over. So I think that it's time for us feminists to get that on to our agenda to try to understand and come to grips with the fact that child rearing is what many women really want to spend some of their lives doing and want to do it right while at the same time maintaining all the other things that are possible to us now with our long lives. There's a question. Would you talk about rigid and flexible parenting styles and the effect that they might have on children with specific genetic makeup? Ah, I'm gonna get out of my field of real expertise here because I don't know a lot about the parenting of children with specific genetic problems. I'm thinking that for some children, I believe some autistic children and some hyperactive ones, it's important to have a fairly inflexible environment that is a predictable one where the child is not overstimulated, the environment is controlled so that the child has only the things they can deal with. Perhaps an inflexible parenting style, that's what you wanna call it, would help those children. That's what they try to achieve in the centers where they offer daycare for such children. Which children benefit most from flexibility? Teenagers, I think. Ha ha ha ha. The extension of this, what is your take on parents who want to be their children's pal or best friend? Oh, fashioned in this respect, I feel that parents have an authority role and I know that this does not preclude at all having periods of time where you're acting as a pal when you go camping together and have fun that doesn't change the fact that it is the parent who's responsible and in charge. Now, of course, during adolescence, what you have to do is gradually let that go and move more, if you can, toward a relationship of greater equality. Throughout childhood, however, I think that it benefits everybody if certain authority lines are understood. Let's have one last question here. Should parents be required to carry a license to parent that reflects at least minimal competence to do the job? No. No, when I imagined a kind of world where we would have tests of the sort that we now have pervading our entire school system of competence and parenting, I shiver. I think that it's wonderful that we have a huge variety of parenting. I do know that there is parenting happening that's damaging to children. We ought to have social programs to either try to help those parents do differently or protect the children. And we do have such things going on. But considering the wide range of parenting that doesn't get into that kind of dangerous category, we want variety. We want variety of outcomes. And I think that you were saying yesterday, Jerry, about the importance of making sure that children who are highly reactive, that their special talents and possibilities are protected. The same thing is true of parenting that resonates nicely with such a child and doesn't resonate nicely with another child. We want both kinds to be there for children who need it. Well, thank you very much for your attention. Thank you very much, Dr. McAgey. We'll adjourn until one o'clock. And we'll hear from Dr. Rappaport. No, I think this is...