Born That Way

by William Wright

  The head of the Pioneer Foundation, Harry Weyher, would not comment on how they choose projects to fund, on the grounds that such statements might appear an effort to influence research. Taking a worst-case supposition, that the foundation still fosters the racial, eugenicist views of its founder, they probably operate on the conviction that genetic differences between the races do exist; therefore any research into genetic influence on personality and behavior will advance this assumption. To believe that Bouchard would publish only those conclusions that reinforced whatever racial ideas Pioneer might have, or in any way to alter the direction of his research, is to believe that the Minnesota team of eighteen scientists, working with graduate students and scores of twins and in the full glare of a major university, could conspire to tailor results to the suspect beliefs of the Pioneer Foundation, just one of several money sources.

  Another backer of the Minnesota studies was the Koch Foundation. In describing how this came about, Bouchard said, “I got a letter one day from a man named David Koch. I’d never heard of him. The letterhead was something like the Koch Technical Group. He wrote that he’d read a little about the study of twins reared apart by Sir Cyril Burt and that the study had been exposed as a fraud. He said he was curious about it and could I give him any information? I went through my files and pulled out a bunch of clippings and wrote him a single-page, single-spaced letter outlining what I knew about the Burt business. I ended by saying if he wanted any more information to let me know.

  “A month later I got a letter thanking me, saying it was exactly what he’d wanted. Then he wrote, ‘By the way, I have a small foundation. We occasionally fund research projects. Why not submit a proposal?’ I ran to the research office and asked them what is the Koch Foundation. They looked it up and said it generally provides educational scholarships for offspring of their company’s employees, mostly small grants of three or four thousand dollars. I thought that it didn’t make sense for me to pursue a grant unless I got around fifty thousand dollars. Not worth the time. I wrote a proposal and asked for eighty thousand dollars. Two weeks later I had a check in the mail for eighty thousand dollars!”

  Bouchard roared at the recollection. Later he learned that David Koch was the billionaire scion of an Oklahoman oil and gas dynasty, a still-young businessman with scientific interests who had previously supported the anthropological work on man’s African origins of Donald Johansen, the discoverer in 1974 of “Lucy.” Looking back on the Koch windfall, Bouchard said, “This country is great for having so many different sources of money. It’s not like most countries, where everything comes from the government. Here, if the government turns you down, you can go elsewhere.”

  FINALLY, IN 1984 Bouchard’s twin study produced a number of papers on their findings, although the researchers chose to bury their most significant finding, a concordance of .73 on I.Q., in the back of a paper about something else. All the research documented significant heritabilities for the traits measured. Also in that year, for a book entitled Individuality and Determinism, which was edited by Sydney W. Fox, Bouchard contributed a chapter that was primarily a defense of previous twin I.Q. studies. Here again he downplayed his own findings and instead addressed the criticisms of the earlier studies, systematically refuting each of the arguments against them in what appeared to be a warning to the critics-in-waiting not to use on him the exhausted, discredited arguments with which they had flailed earlier studies.

  He takes on all the critics of I.Q. studies, paying particular attention to the most extreme, Leon Kamin, who wrote in his 1974 book, The Science and Politics of I.Q.: “Whatever the ‘experts’ may say, there is no compelling evidence that the heritability of I.Q. is 80 percent, or 50 percent, or 20 percent. There are not even adequate grounds for dismissing the hypothesis that the heritability of I.Q. is zero. The evidence is clearly inconsistent with a high heritability.”

  In rebuttal, Bouchard first stated the most persuasive evidence turned up by his study: the mean correlation of I.Q. between reared-apart twins of .72, as compared with .52 for fraternal twins reared together. (His numbers changed slightly as the number of pairs tested increased.) As might be expected, identical twins reared together had an even higher correlation—.86. Although this MZT figure is considerably larger (.14) than the MZA figure, it is not as large as the difference between the mean correlations of MZAs and DZTs (.20). This said very clearly that identical twins reared in different homes are closer in I.Q. than fraternal twins reared in the same home.

  Bouchard goes to great pains to methodically address each one of Kamin’s stabs at discrediting twin data. Having addressed his complaints about fraternal-identical comparisons, Bouchard went on to dispute the criticisms of reared-apart twin studies, most notably those in the 1981 book by Susan Farber, and painstakingly answers the points of each, one by one.

  In his rebuttals, Bouchard did not cite one way of looking at his statistics that make nonsense of the critics’ most frequent complaint: that separated twins are usually adopted into very similar homes, thereby homogenizing their test performances. In all his measures, the reared-apart identical twins measured consistently more concordant than reared-together fraternal twins. This makes it unavoidable that, regardless of the similarity of the two twins’ home environments, they could not be as similar as the one home that reared-together fraternal twins shared.

  My brilliant editor, Victoria Wilson, on reading this argument, pointed out that fraternal twins growing up together may strive to be dissimilar. Subtle as this point is—and it is the sort of skewing possibility for which researchers must constantly be alert—it is overcome by the impossibility of willing one’s I. Q. higher or lower or the unlikelihood of making oneself shy because of an aggressive twin, excitable because of a phlegmatic one, and so on. Results showed, however, that traits amenable to conscious choice, like the last two mentioned, measured no more concordant than those, like I.Q., that are not. Even though Bouchard’s paper is about I.Q., he once again unobtrusively works his own findings into his discussion of all I.Q. studies instead of trumpeting what many considered the definitive data on the controversial subject.

  Summarizing his other findings, Bouchard writes: “In the domain of personality, estimates of heritability are somewhat lower [than in I.Q. and physiological traits] but are still substantial, somewhere in the area of .50 to .60.” While he admits his concordances are somewhat higher than those of earlier efforts to measure the heritability of personality by adoption studies, he says, “both the twin studies and the adoption studies do, however, converge on the surprising finding that common family environmental influences play only a minor role in the determination of personality.” This was quite a gauntlet to toss down at the heavyweights of psychological thought like John B. Watson, B. F. Skinner, and Dr. Spock, for whom family environment was pretty much everything. Even in Freud’s constructs, family influences were far from “minor.”

  IN 1986 the Minnesota study published a paper on homosexuality in the British Journal of Psychiatry. The paper began by reviewing six earlier studies of homosexuality that compared MZs with DZs and that consistently found far higher concordances among the identicals, in one case 100 percent. There were, the Minnesota authors allowed, potential flaws in each of these studies. In all of the six pairs studied by the Minnesota group, two male pairs and four female, at least one twin had a history of homosexuality. Of all the pairs, one male pair was definitely concordant in that both twins had been actively homosexual well before reuniting. The other male pair was less clear and illustrated the pitfalls of self-reporting.

  One brother said he was fully homosexual; the other considered himself heterosexual, but when he gave his sexual history, however, he said that between the ages of fifteen and eighteen he had had an affair with an older man. Now in his mid-thirties, he said that he is happily married to a woman but added that they made love infrequently. In the eyes of skeptics, this history did not qualify him as an award-winning heterosexual. And if h
is heterosexuality was disallowed, the male concordance in this tiny sample would have been 100 percent.

  With the female twins, the homosexuality paper was on only slightly firmer statistical ground in that it included four pairs. In three pairs, one twin was leading an active lesbian life. A twin in the fourth pair had had a long and intense affair with another woman but was currently involved with men, so the Minnesota team designated her bisexual. The identical twin sisters of all four of the homosexually active women were exclusively heterosexual. Bouchard put this discordance forward as strong evidence that lesbianism is not genetic, that it appears to be environmentally determined. While unlikely, a behavioral trait could be genetic in one sex and not in the other; it is not impossible. Although this was just one of many traits, it is one that could spell the difference between a happy life and a difficult one, so with Bouchard’s findings, the concerned, involved parent was still in business.

  Several things are unusual about this paper. For most of the six years the Minnesota Twin Study had been in existence, Bouchard and his colleagues had emphasized in their papers the dangers of drawing conclusions from small samples. So it is surprising that they were so quick to draw conclusions about male homosexuality (“it is hard to deny genetic factors”) based on only two sets of male MZAs. In their previous papers, the Minnesotans had been cautious about drawing any conclusions, even when their data was weighted toward a particular inference. Now, with a small sample in which the orientation of one subject was unclear, they were speculating broadly on the causes of lesbianism.

  With a study that pointed to the opposite result—no genetic component to a trait—and based on a sample that was unusually meager, they almost trumpeted their shaky conclusions. The lesbian paper’s final sentence reads: “If this remains a constant finding, it will be the strongest evidence known to us which attributes a major behavioral complex exclusively to environmental factors.” Researchers will never admit to anything but the purest scientific considerations in the presentation of their data, but my suspicion is that the lopsided lesbian paper reveals a feeling of relief in Minneapolis that genes turned out to be less than all-powerful. See, it seemed to be saying, we are not as genes-giddy as you thought.

  If relief was the feeling, it turned out to be unjustified. The far more comprehensive study of homosexuality done in 1993 at Northwestern University by Michael Bailey and Richard Pillard confirmed Bouchard’s conclusions about male homosexuality—although with a far lower, but still significant, concordance between twenty-three MZA pairs of .52. But a follow-up study on female homosexuals contradicted Bouchard’s four pairs when .48 of thirty-five pairs of identical female twins were both gay, almost half, as opposed to .16 of eighteen pairs of fraternal twins and .14 of nontwin sisters. So it appears the Minnesota team was overly eager in announcing the discovery of a major trait in which genes played no part.

  FIVE

  MINNESOTA’S TRIUMPHS

  LATE IN 1986, the Minnesota group had completed assembling its data for a major paper, the concordances of a broad range of personality traits between sets of separated twins, now totaling forty-four. The paper also included data on 331 reared-together twins, 217 identical and 114 fraternal. A trait that showed one of the highest levels of heritability, .60, was traditionalism, or a willingness to yield to authority. The only one of these traits to show a higher heritability was social potency, which included such characteristics as assertiveness, drive for leadership, and a taste for attention. A surprisingly high genetic component—a .55 concordance—was found in the ability to be enthralled by an esthetic experience such as listening to a symphonic concert.

  Good news for the environmentalists was that one important trait, one of the species’ most appealing, appeared to have little genetic basis: the need for social intimacy and loving relationships. The heritability figure for this trait was .29, putting it just under Bouchard’s somewhat arbitrary threshold of genetic significance. If you were warm and loving, you probably came from a warm and loving home. But here it should be remembered that if a trait has a high heritability, say above .60, that still leaves room for many people who came by the trait not by their genes but by environmental influences.

  Other noteworthy numbers emerged. In five of the categories tested—stress reaction, aggression, control, traditionalism, and absorption—the separated twins tested slightly more concordant than the identical twins reared together. As this oddity is not mentioned in the paper, the Minnesota team appeared to give it little significance. No trait showed zero heritability, and most were in the .45 to .60 range.

  In this comprehensive paper on personality, the Minnesota team could not resist taking shots at earlier psychological studies that measured similarities in twins reared together and showed the twins to be far more similar to each other than they were to nontwin siblings. Oblivious to the possibility of genetic influence, the studies attributed those similarities to environmental factors such as twins’ influences on each other, more similar treatment by others, and greater expectations from families of similar behavior from twins than from brothers and sisters. All of these explanations were wiped out when the twins studied were raised separately.

  It is understandable that Bouchard, who was usually diplomatic and nonconfrontational in putting forth his orthodoxy-shattering conclusions, would sound a note of exasperation about these gene-blind studies. He knew that the researchers behind them, all trained psychologists who knew of the high level of genes shared by twins (50 percent or 100 percent), were so grounded in the prevailing dogma of their profession that they never thought to look outside the environment for explanations of why twins were more similar in personality than other siblings. Bouchard made clear that he found this an astounding misassumption. By ignoring the possibility of genetic influences on personality in setting up these studies, by placing genes off-limits, psychology’s policy-makers assured none would be found.

  Minnesota’s paper on personality traits repeatedly reproached earlier studies aimed exclusively at the environment, saying that because they focus on “social class, child-rearing patterns and other common-environment characteristics of intact biological families, [these studies] cannot be decisive because they confound genetic and environmental factors.” The phrase cannot be decisive was a polite way of saying they tell us nothing. It was as though scientists undertook studies of the makeup of water with the proviso that only hydrogen be looked at; oxygen was off bounds.

  The personality paper concludes with a subdued summation of the Minnesota team’s startling findings: “It seems reasonable, therefore, to conclude that personality differences are more influenced by genetic diversity than by environmental diversity.” The paper also asserted that shared family environment turned out to have a negligible effect on personality. Social closeness was the only trait that appeared to be mainly a product of family environment.

  WHEN THIS IMPORTANT PAPER was concluded, it was sent out for review prior to publication in the Journal of Personality and Social Psychology. The review process by other, unrelated experts in the field is a standard procedure for scientific papers and in complex studies like Minnesota’s about personality measurements can take up to a year to complete. Participants in the review process often discuss their findings with colleagues at other universities; only in unusual situations is tight secrecy maintained. Still, it is rare for the results to reach the press until the paper is reviewed and published.

  A prolonged incubation period prior to publication was not to be the case with Minnesota’s findings on personality. The results leaked out and were proclaimed in the lead story of the “Science Times” section of the New York Times on December 2, 1986, a full year before the paper itself was published. The article ran with the headline MAJOR PERSONALITY STUDY FINDS THAT TRAITS ARE MOSTLY INHERITED. Daniel Goleman, who later wrote Emotional Intelligence, started his article with a straightforward declaration of the findings’ principal import: “The genetic makeup of a child is a stronger
influence on personality than child rearing, according to the first study to examine identical twins reared in separate families.”

  Goleman referred to 350 pairs of twins studied by the Minnesota group and later added that, of these, forty-four pairs were identical twins raised in different homes. He described the study’s design and gave samples of the questions aimed at measuring personality (“True or False? When I work with others I like to take charge”). Early in the piece, Goleman stated the overall results: “For most of the traits measured, more than half the variability was found to be due to heredity, leaving less than half determined by the influence of parents, home environment and other experiences in life.”

  In a box running beside the article’s first column, Goleman listed the eleven traits examined and gave the heritability figure for each. The highest were social potency (.61) and traditionalism (.60). Five were between .50 and .60: stress reaction, absorption, alienation, well-being, and harm avoidance. Three were between .40 and .50: aggression, achievement, and control. The last, social closeness, just squeaked above the significance level with a revised figure of .33. A pair of warm, loving twins must have arrived in Minneapolis to boost the earlier figure.

  The article included quotes from other experts that underscored the significance of the findings. Jerome Kagan, the distinguished Harvard psychologist, was quoted as saying, “If in fact twins reared apart are that similar, this study is extremely important for understanding how personality is shaped.” This was a generous comment since Kagan himself, with thirty years of developmental research behind him, had been a leader in the struggle to understand how personality was shaped.