Extraits

Chapter One Mirror, Mirror          In August 1939, in the tiny town of Piqua in northeast Ohio, an unmarried woman gave birth, slightly prematurely, to twin boys. The country as a whole was struggling through the last of the depression years, and times were still hard in Ohio, as in most other places in prewar America. This poor mother, a recent immigrant, simply had no way of supporting two newborn infants, so, like many others who found themselves in her position, she asked the hospital to help her place the boys for adoption. These were healthy, handsome babies, and they were taken quickly; both were adopted within a few weeks, both by middle-class working families. One of the families was from Piqua itself; the other was from across the state, in Lima. In what would be the first of a string of improbable coincidences involving these twins, both families named their newly adopted son James. But aside from these first few weeks in the same hospital, the two "Jimmies" would remain largely unaware of each other's existence, and they would not be brought together again for nearly forty years. Only then would they learn that they were genetically identical--that they were, in a sense, clones of one another.     Just a few years later, in a small city in western Michigan, another set of twin boys was born, again to an immigrant mother, but this time to one who was fortunate enough to have a husband and, even more fortunate, a husband with a job. It didn't pay much--he stitched and delivered burlap sacks to potato growers in the area--but it was enough to keep them all together. The twins came at the end of a string of four other children, and were seen as a mixed blessing for an already hard-strapped family. But they were welcomed; the family just squeezed a bit more tightly together in their modest house. Unlike the two Jimmies, these two boys would grow up in the same home, with exactly the same siblings and neighbors and pets, for nearly eighteen years. And the Michigan twins were different in another way: They were not identical, but fraternal.     These two sets of twins would never meet; they were never, in fact, aware of each other's existence. But they present those who study the development of human behavior with a natural experiment, an experiment of the type that could never be done with human subjects in a research laboratory. They have the potential to provide us with important insights into the role of genetics versus environment, of "nature versus nurture," in shaping who we turn out to be in life. The two fraternal twins, Tommy and Bernie, were genetically distinct, yet reared in the same environment. The two Jimmies, although genetically identical, were raised in different environments, with different parents, siblings, neighbors, and pets.     One of the two Jimmies became aware through his adoptive mother that he had a twin brother, and eventually set out to find him. He was successful, and when the two were finally reunited at age thirty-nine, and began to swap childhood stories with one another, they were astounded at the number of similarities in their lives. Some were trivial, and almost certainly nothing more than monumental coincidences. For example, each had adoptive siblings named Larry and pet dogs named Toy. Both preferred Miller Lite beer, and smoked Salem cigarettes. Each had married a woman named Linda, divorced, and then married a woman named Betty. One named his firstborn son James Alan, the other James Allen. Each loved stock car racing and hated baseball.     But as time went on, they would find additional, and from a biological point of view more intriguing, similarities in their separate early lives. Their personalities, as described by family members and as clearly indicated by standardized personality assessment tests, were remarkably similar. Each had developed sinus headaches at about age ten, a condition that eventually developed into recurring migraine headaches. Their descriptions of their symptoms, elicited by specialists, were virtually identical. Each had been good at math in school, while each struggled with English; their overall scholastic performances were remarkably similar. Each had a fondness for woodworking, and each developed the habit of biting his fingernails.     Tommy and Bernie, on the other hand, were as different as they could be from the very first day. Tommy was a peaceful baby, slept a great deal, and responded positively to being held and touched. Bernie was the opposite. He fussed or cried constantly; cuddling and caressing by family members didn't seem to console him. In a matter of just a few weeks, everyone began speaking of the twins in quite different terms. Tommy, by common consent, was "an angel"; Bernie was "difficult." Before long, it was obvious that the family interacted rather differently with each of the boys. The differences in their behavior continued to develop as the boys became toddlers and started interacting with other children in the neighborhood. Tommy seemed content to stay in the house and play with his toys; Bernie charged outside, where he would engage in play with other children. He seemed to have a penchant for rough-housing, and often came home bruised or cut. These differences carried through to the boys' early years at school. Tommy was adored by all his teachers; Bernie was sent home frequently because of his disruptive behavior in class.     As the two Jimmies moved into their young adult years, during which they had no contact, they continued developing similar habits and personality traits. They expressed themselves alike, and used similar slang phrases. Both suddenly put on about ten extra pounds at roughly the same point in life. In the years preceding their ultimate reunion, both encountered assorted stresses that led to chest pains and high blood pressure. Both had difficulty sleeping, and both were taking Valium for general nervousness. Both held clerical jobs, and each had developed a fascination with police work; each had become a volunteer sheriff's assistant in his respective community.     Tommy's and Bernie's lives took very different courses. Tommy was a middling student, but went on for two additional years after high school at the local community college. After a short stint in the military, part of which was spent in Vietnam, he entered a Catholic seminary and became a priest. Bernie dropped out of high school, got into a series of scrapes with the law, and has been in and out of prison a good deal of his adult life. His social interactions have remained extremely difficult, and he never married. In spite of their extraordinarily different personalities, they have stayed in close touch over the years.     The behavior of these fraternal twins highlights a subtle but very important point about the interaction of individuals with their environment. We commonly think of the impact of the environment on the individual, but for human beings this interaction is actually a two-way street. Although Tommy and Bernie grew up in exactly the same environment, they manipulated that environment differently. This was not necessarily a willful or even conscious manipulation on their part. These were two genetically quite different individuals, and some of their differences were expressed as personality differences that were immediately perceived by the people around them. Based in turn on their own personalities, those people responded in different ways to Tommy and Bernie. It was not the environment that made one quarrelsome, and the other mild-mannered. But their differences elicited markedly different responses from those around them, and in that sense they cannot be said to have grown up in exactly the same environment.     We will talk a good deal about "environment" as we proceed through this book, because it is impossible to talk about the role of genes in behavior without taking the environment into account as well. So let us take just a moment to think about what we mean when we refer to the environment. For most animals, the environment means what we might call the physical or ecological environment, the natural surroundings in which the individual "behaves": competes for resources to stay alive, to find a mate, and to produce offspring. These basic behaviors are the same for humans as for animals, but we pursue them in two quite different sorts of environments. We, too, function within the context of an ecological environment; we, too, must eat, stay warm, find food and a mate. But unlike animals, we also function within the context of a cultural environment. Culture consists of a wide range of abstract ideas, social customs, rituals, creative works, and institutions that are largely made possible by language. The cultural environment shapes human beings every bit as forcefully as the ecological environment; in fact, it can be argued that culture is now a more important factor in human genetic evolution than is the ecological environment. So as we proceed through the chapters that follow, we should always bear in mind not only the role of environment in determining behavior in a general sense, but also the unique role of the cultural environment in determining human behavior in particular. The interaction between our genetic selves and our cultural selves is very complex indeed.     The Biological Nature of Twins     Twins have always fascinated us. They are a mystery, and so it is perhaps not surprising that they had become bound up in mythology long before they became an object of medical investigation. They are feared in some cultures, and revered in others. Twinning is much more common than live births would suggest. Although only about four sets of fraternal twins, and one set of identical twins, are produced per 1,000 live births in the United States, sophisticated sonographic studies of early human pregnancies suggest that as many as one in eight conceptions--and possibly more--produces multiple embryos. The vast majority of these are lost in the first few weeks of pregnancy, and under normal circumstances their existence is undetected by either the mother or her physician.     The most intriguing twins--identical twins--are created when a single, fertilized egg splits into two parts shortly after fertilization. This can occur at several different stages in development; the later it occurs, the more similar the twins. Once fertilization has taken place, the egg starts dividing and begins its journey through the Fallopian tube toward the uterus. The immediate product of the union of a sperm and egg is called a zygote, and so twins arising from a single fertilization event are also called monozygotic twins. Once the zygote starts dividing, the developing cell mass is called an embryo; when the embryo is clearly recognizable as human (after about five weeks), it is referred to as a fetus.     Amazingly, for a number of rounds of cell division after an egg is fertilized, entire individuals can be formed from only a portion of the cells in a developing embryo. None of the cells in the evolving cell mass has yet become specialized, and all retain the potential to produce an entire individual. Physical splitting of the cell mass, or "partitioning" as it is sometimes called, can occur at various stages in the early stages of the gestational process. This part of the twinning process is not well understood. It is not at all clear why such a coherent cell mass would break apart, or, once it did, why the parts would not simply rejoin, since embryonic cells are very sticky. Early mouse embryo cell masses can be readily separated in the laboratory; but great care has to be taken to prevent the separated parts from rejoining. The stable partitioning of an early embryonic cell mass is such an unlikely event that many obstetricians wonder whether it might be a type of birth defect, albeit an entirely harmless one.     Very early partitioning (up to three or four days post-fertilization) of human embryos results in identical twins that have separate placentas. Twins arising from partitioning events occurring after about four days (roughly 70% of monozygotic twin pairs) usually share a common placenta. There is some evidence that identical twins attached to separate placentas may develop slightly differently, and there has been a lengthy--and largely unresolved--debate about how this might affect the future development of the individuals involved. On occasion, partitioning may occur up to as late as ten days post-conception, and in these cases separation may be only partially complete, leading to various degrees of the condition known as Siamese twins. All three types of twins--shared placenta, individual placentas, and Siamese--are monozygotic and genetically identical, but it is conceivable that their different developmental patterns may result in subtle differences between the twins as adults.     More than two genetically identical individuals may arise from a single fertilization event, although this is quite rare. The developing embryonic cell mass may split into three or even more parts, resulting in multiple monozygotic individuals. The Dionne quintuplets, born in Canada in 1934, were genetically identical and thus monozygotic. It is also possible to have mixed multiple births; it is not at all unusual for quadruplets to consist of one pair of identical twins and one pair of fraternal twins, for example.     We must introduce a note of caution here about use of the term "genetically identical" with regard to monozygotic twins. Although monozygotic twins start out life from a common genetic blueprint, the development of a fully grown adult from the early zygotic stage is not a perfectly controlled process. Potentially mutagenic errors are regularly detected and removed in the line of cells giving rise to sperm or eggs in each generation, but this process is considerably less strict in production of those cells that form the soma (all of the rest of the cells in the body) during the development of a single individual. Normally, mutations that arise in somatic cells do not spread far in the body, because most somatic cells give rise to only a few progeny in their lifetimes, particularly in behavior-generating tissues such as the brain. Nevertheless, the accumulation of somatic errors throughout life is a potential source of genetic differences in otherwise identical twins.     Even in the development of two individuals from completely identical genetic blueprints, the developmental pathways might not be exactly the same. Particularly in development of the nervous system, which is at the center of all behavior, there is a good deal of randomness in the generation of varying portions of the brain and peripheral nerves. During embryonic and fetal life, newly forming nerve cells toss out fibers more or less randomly into their immediate vicinity. These will by chance form connections with other nerve cells or with nearby muscle cells. Nerve cells failing to make a connection die off; those that establish a connection retain that connection essentially for life. But even two genetically identical twins will develop slightly different patterns of nerve cell connections, and these differences could well be a basis for differences between monozygotic twins. Detailed analyses of the brains of monozygotic twins have in fact revealed small but potentially important neuroanatomical variations.     Unlike identical twins, fraternal twins are conceived when two separate sperm fertilize two separate eggs, and they are thus referred to as dizygotic twins. They never share the same placenta. The resulting embryos share the same womb at the same time, but they are no more alike genetically than any two children born to the same parents through different birth events. That makes them far more alike genetically (50% alike, on average) than two children selected from the population at random, but still rather far from complete genetic identity. Identical twins (with extremely rare exceptions) are always of the same sex; individual members of fraternal twin pairs each has a random chance of being male or female.     Until well into the present century, same-sex twins were judged to be identical or fraternal largely on the basis of appearance. Occasionally, fraternal twins may seem so alike that it would be easy to mistake them for identical twins. On rare occasions, identical twins may seem slightly different in appearance. Unless twins are of the opposite sex, they are now routinely tested for blood group proteins or DNA markers that allow doctors and parents to make an unambiguous determination of their genetic status.     What Twins Can Tell Us about the Role of Genes in Human Personality     Scientists interested in the genetics of human behavior are interested first and foremost in variability. The question really is not whether genes underlie human behavior. Ultimately every aspect of the existence of every biological organism is determined by its genes; humans are no different in this respect. The real question is to what extent variability in the genes affecting behavior contribute to the variability in human behavior we see around us, and to what extent this variability is determined by differences in the environment--the home in which the individual was raised, churches and schools attended, and the community in which the individual lives and works throughout his or her life.     The study of such questions in humans is accompanied by a number of restraints. In subsequent chapters, we look at the role of genes in causing variations in behavior in a wide range of animal species, from the simplest single-cell organisms, to fruit flies and roundworms and to mammals such as rats and mice. Our information comes from a variety of experiments that are simply not possible in humans. Laboratory animals can be selectively bred to reveal inheritance patterns from generation to generation. In many of these species, dozens of generations can be produced in a single year; humans require over a dozen years to produce a single generation. If we suspect that a particular form of a given gene causes a particular behavioral variation in an animal species, we can very often insert that gene variant into one of them to see just what it does.     We cannot do any of these things in humans. We can only observe from the outside whatever it is that humans do naturally and of their own free will. Thus one of the oldest (and still used) methods of studying the role of genes in human behavior is to look for patterns of behavioral inheritance in families. Traditionally, this involved applying certain behavioral assessment tests to as many members of as many generations of as many families as possible, and then applying statistical tests to the results to determine whether the transmission of these traits seemed to be heritable. While this approach has been very important in spotting potentially heritable human traits, it has suffered from questions about both the validity of the behavioral assessment tests used and the statistical methods used to define heritability. In later chapters, we see how modern molecular genetics has greatly improved the power of family lineage studies, but admittedly many questions remain.     The study of behavior in twins, and in children adopted into genetically unrelated families, has also greatly strengthened family lineage studies of the heritability of behavior. The basic strategy in twin studies, for any given variable behavior such as personality, is to look at the variability in that behavior when comparing pairs of monozygotic twins reared together or apart, as well dizygotic twin pairs reared together or apart, and to compare both of these with variability in the same trait among non-twinned biological and adopted siblings. Monozygotic twins reared apart provide additional controls for the influence of genes and environment on behavioral variability; they are essentially identical genetically, but they grow up in different cultural environments. Dizygotic twins reared together test the effect of different genetic constitutions in the same environment.     The results of such studies must also be interpreted by statistical means, and to be meaningful, many pairs in each category must be tested. When such comparisons are carried out on a large enough scale it is possible, as we will see shortly, to dissect out genetic influences from environmental influences, based on the degree to which the subjects under study share common genetic backgrounds versus common environmental influences. In working out the relative contributions of genes and environment to the differences we see between individuals, geneticists refer to a fairly straightforward formula:     V = G + E s + E n     This formula says simply that the variability ( V ) we see between two individuals should be accountable for by a combination of differences in their genes ( G ), plus differences in their environment ( E ). Environmental differences (whether cultural or ecological) can in turn be divided into those differences that are shared between two individuals ( s ), and those that are not shared ( n ). For example, two children raised in the same home would share certain environmental factors relating to parents and other family members, but would not share certain things in their external environment--different school experiences, for example, and different friends. Particularly when looking at variability involving monozygotic twins, we could also add an additional term to this equation to reflect the possible differences in the detailed aspects of fetal development discussed earlier.     We use the term "intrascale correlation," or simply "correlation," when we discuss comparisons of various individuals with respect to a given trait. Correlation in this sense is a complex parameter, dependent on sophisticated statistical arguments. For our purposes it means simply the following: a correlation of 1.0 indicates that the performance between tested pairs, carried out with large numbers of pairs, is absolutely identical; a correlation of 0 indicates the performance between tested pairs was completely random. In reality, correlations of 1.0 and 0 do not occur. Because even the same individual tested on different days rarely has a correlation of more than 0.9 on most tests, a score of 1.0 between two different individuals would not be considered real, and when the correlations of large numbers of pairs are averaged, a score of 1.0 is simply not seen. Two randomly selected individuals would by chance show some degree of correlation on most tests, although the average correlation over many pairs would seldom exceed 0.05 to 0.10. Particularly when randomly selected individuals are tested for several variables simultaneously, the overall correlation usually falls below statistical significance, but for that reason we can never say it is zero. (Continues...)