Why can the assumption that MZ and DZ twins are raised in equally similar environments be problematic?

Three books published in the period 2018-2020 have helped promote the idea that human behavioral differences are largely caused by hereditary factors—a false idea we can now view in the context of the 2020 pandemic crisis and the accompanying greater appreciation of the common biology we all share. The first book was published by psychologist Robert Plomin, perhaps the world’s leading behavioral geneticist, in 2018. Its title is Blueprint: How DNA Makes Us Who We Are.1 In a 2019 article, I showed that Plomin’s arguments in favor of genetics do not hold up, that he airbrushed genetic research and eugenics out of the pre-1970s history of his chosen field of psychology (see also 1:48-2:00 of this 4/30/2020 interview), and that his claims about the relatively new polygenic score (PGS) method must be evaluated in the context of his 40-year history of making unfulfilled claims and predictions in the area of behavioral molecular genetic research.2 (Problems with the PGS method are discussed HERE, HERE, HERE, HERE and HERE.) Plomin’s arguments in Blueprint were based largely on twin studies.

The other two books are neuroscientist Kevin J. Mitchell’s 2018 Innate: How the Wiring of Our Brains Shapes Who We Are, and political scientist Charles Murray’s 2020 Human Diversity: The Biology of Gender, Race, and Class.3 Because I have already reviewed Plomin’s Blueprint, here I will show that although Mitchell and Murray cited twin studies as the main source of evidence in support of their claims, these studies are based on a long-controversial assumption which they were unable to defend in their books, or on Twitter. (Plomin was not active on Twitter in the first half of 2020.)

Mitchell believes that molecular genetic studies have already identified some genes that play a role in behavioral differences, and Murray believes that polygenic scores “already explain significant proportions of the variance in many traits” (p. 279). Molecular genetic studies of behavior, however, have been characterized by the publication of false-positive results followed by non-replication, systematic error, the need to produce “discoveries” that please research funding sources, publication bias in favor of positive findings, and a reliance on questionable assumptions, twin studies, and heritability estimates (more below). Behavioral and psychiatric gene discovery claims have been appearing since the 1960s, sometimes accompanied by sensational reports in the mass media, yet rarely if ever hold up. Based on a half-century of false-alarm behavioral gene discovery claims, the appropriate reaction to new claims should be extreme skepticism and caution.

Before examining how Mitchell and Murray cited twin studies in support of their views (they also cited adoption studies), I will briefly discuss the underlying logic of twin studies, and the main reasons why we should call this logic into question.

Twin Studies

The two main methods that use twins for behavioral research purposes have produced (1) thousands of publications based on the twin method, which uses reared-together MZ (monozygotic, identical) and reared-together same-sex DZ (dizygotic, fraternal) twins, and (2) six published twins reared apart or “TRA” studies, also known as “separated twin studies,” which study MZ (and sometimes DZ) twins who supposedly were (but in most cases weren’t) separated at birth and reared apart from each other in different family environments.

The twin method, which was developed in the 1920s, compares the behavioral resemblance, concordance rates, or psychological test score correlations of MZ versus same-sex DZ pairs. MZ pairs share a 100% genetic resemblance, whereas DZ pairs share an average 50% genetic resemblance. Twin method results usually show that MZ pairs behave more similarly or correlate higher than same-sex DZ pairs at a statistically significant level—a finding that I will designate “MZ > DZ.” Genetic interpretations of MZ > DZ are based on twin researchers’ acceptance of the “equal environment assumption,” also known as the “EEA.”

The EEA states that MZ and DZ pairs grow up experiencing roughly equal environments, and that the only behaviorally relevant factor distinguishing these pairs is their differing degree of genetic relationship to each other (100% vs. 50%). Critics argue that the EEA as it relates to behavioral twin studies is obviously false, since when compared with same-sex DZ pairs, MZ pairs grow up experiencing (1) much more similar treatment by parents and others, (2) much more similar physical and social environments, (3) more similar treatment by society due to their sharing very similar physical features, and (4) identity confusion and a much stronger level of emotional attachment.

I put together a table showing the results of studies that assessed levels of identity confusion and psychological attachment experienced by MZ and DZ pairs (seen HERE), and these levels are much higher among MZ pairs than among DZ pairs.

Since the mid-1960s, research and common sense have converged on the conclusion that MZ pairs grow up experiencing much more similar environments, and are treated much more similarly, than are DZ pairs. This means that twin researchers and their critics don’t have to argue anymore about whether MZ and DZ environments are different, since almost everyone now agrees that they are different. Twin researchers nevertheless continue to maintain that the EEA is valid on the basis of three main arguments, two of which I described and deconstructed HERE. Below I briefly summarize these first two arguments and the fallacies they are based upon, and later I will describe the third argument.

Argument A

Supporters of what I have called Argument A recognize that MZ pairs grow up experiencing more similar environments than experienced by DZ pairs, but they maintain that the EEA is valid because MZ pairs “create” or “elicit” more similar environments and parental treatment for themselves because they behave more similarly for genetic reasons. Therefore, they argue, environmental influences on twins’ behavioral similarity should be counted as genetic influences. One of many examples of Argument A is found in a 2000 article by genetic researchers writing in support of the “validity of twin studies.” Although they recognized that “there is overwhelming evidence that MZ twins are treated more similarly than their DZ counterparts,”

“…the more similar parental treatment of MZ vs. DZ twins occurs in response to the greater similarity of actions initiated by MZ pairs.…It seems…likely that the increased similarity in treatment of MZ twins is a consequence of their genetic identity and the more similar responses this elicits from the environment” (italics in original).4

Argument A, however, is a circular one because the conclusion that genetic factors explain MZ > DZ is based on a premise that assumes the very same thing. Twin researchers invoking Argument A fallaciously refer to the genetic premise in support of the genetic conclusion, and then refer back to the genetic conclusion in support of the genetic premise, in a circular loop of faulty reasoning. One observer wrote that circular reasoning is “empty reasoning in which the conclusion rests on an assumption whose validity is dependent on the conclusion.”5 That, in a nutshell, is the fallacy of the Argument A defense of the EEA and the twin method.6

Argument B

Twin researchers using Argument B also recognize that MZ pairs experience more similar environments than DZ pairs, but claim that the EEA remains valid until critics are able to identify the “trait-relevant” aspects of the environment that cause MZ pairs to behave more similarly. An example of the Argument B definition of the EEA is found in a 1993 publication by genetic researcher Kenneth Kendler and his colleagues:

“The traditional twin method, as well as more recent biometrical models for twin analysis, are predicated on the equal-environment assumption (EEA)—that monozygotic (MZ) and dizygotic (DZ) twins are equally correlated for their exposure to environmental influences that are of etiologic relevance to the trait under study” (italics added).7

Supporters of Argument B attempt to shift the burden of proof from themselves onto critics for showing that MZ and DZ pairs differ in their exposure to “trait-relevant” environmental factors. However, a basic principle of science is that the burden of proof falls squarely on the people making a claim, not on their critics. Therefore, given that they recognize that MZ and DZ environments are different, behavioral twin researchers using Argument B—and not their critics—are required to identify the specific and exclusive trait-relevant environmental factors involved in the behavioral characteristic or psychiatric disorder in question. After accomplishing this, they then must show (1) that MZ and DZ twins did not experience such factors, or (2) that MZ and DZ twins experienced such factors to roughly the same degree. In most cases, behavioral twin researchers have been unable to identify specific and exclusive “trait-relevant” environmental factors. Until they are able to identify such factors, and until they are able to subsequently determine that MZ and DZ pairs were similarly exposed (or not exposed) to these factors, Argument B defenses of the EEA fail completely.

***

Because Mitchell and Murray based their arguments largely on studies using the twin method, and surprisingly not so much on TRA studies such as the very problematic “Minnesota Study of Twins Reared Apart,” the focus of this article is on the way these authors discussed, or failed to discuss, the validity of the twin method and the EEA. (Plomin did not discuss the EEA at all in Blueprint.)8

Kevin Mitchell’s 2018 Innate: How the Wiring of Our Brains Shapes Who We Are

The word “innate” means “determined by factors present in an individual from birth.” A major theme of Mitchell’s book by that name is that, due to heredity plus pre-natal developmental wiring, “We are different from each other in large part because of the way our brains get wired before we are born,” and that “it is mainly genetic variation affecting brain development that underlies innate difference in psychological traits” (p. 7). For Mitchell, potential post-natal environmental influences on behavior count for very little. “Our common experience,” he wrote, tells us that “at some level, people just are the way they are—that they’re just made that way” (p. 1). Although Mitchell at times seemed to soften his stance when using terms such as “behavioral tendencies,” “development,” “brain plasticity,” “stochastic,” and “interactions,” the core message of Innate is that our genes are the main factor determining how we behave, and who we become.

Mitchell’s Definition of the EEA

In support of his “people are made that way” theme, Mitchell claimed that “twin and adoption studies” are able to “distinguish [the] possible effects of nature from those of nurture” (p. 12). “Twin studies,” he wrote, “compare people who have the same degree of shared environment, but differ in how similar they are genetically” (p. 13). He used the traditional definition of the EEA that reared-together MZ and DZ pairs “grow up under similar conditions” (p. 13), and “share a family environment similarly” (p. 17). Later in the book he qualified/contradicted this position with the (as we have seen circular) Argument A claim that “much of the variance in how parents treat their children is indeed driven by the child’s own behavior and genotype,” which supposedly explains why “pairs of MZ twins have significantly more similar life experiences than pairs of DZ twins” (p. 96). Again, this argument is circular because Mitchell assumed that children’s behavior is “driven by their genotype” in order to conclude that twin studies show that children’s behavior is driven by their genotype.

Based on his endorsement of the EEA, Mitchell repeated the standard behavioral genetic claim that twin studies “allow us to estimate how much of the variance in a trait is due to genetic differences between people, and how much is due to differences in family environments, and, importantly, how much is unexplained even when those two factors are taken into account” (p. 18). The disputed claim that we can determine “how much of the variance of a trait is due to genetic differences between people,” and that such a determination is meaningful, is presented in behavioral genetics in the form of a “heritability estimate” (0%-100%). I place “heritability estimate” in quotation marks because, as a pair of critics wrote, “the term ‘heritability,’ as it is used today in human behavioral genetics, is one of the most misleading in the history of science.” (More on the “heritability” fallacy HERE.)

A major problem with Mitchell’s EEA definition is that, as we have seen, most twin researchers have recognized for decades that MZ pairs grow up experiencing much more similar environments than DZ pairs experience. In a 2014 article (ironically) written in defense of the twin method and the EEA, for example, criminology twin researcher J. C. Barnes and colleagues recognized, “Critics of twin research have correctly pointed out that MZ twins tend to have more environments in common relative to DZ twins, including parental treatment…closeness with one another…belonging to the same peer networks…being enrolled in the same classes…and being dressed similarly.”9

According to Mitchell, “If the environment you grow up in were the only thing that mattered for some trait, then the similarity between pairs of MZ twins should be about equal to that between pairs of DZ twins” (p. 13). This is wrong. Because MZs experience much more similar environments than DZs, a behavioral trait caused entirely by environmental factors will also produce MZ > DZ. The key question is how we should interpret MZ > DZ. Twin researchers and popularizers of their work such as Mitchell argue in favor of a genetic interpretation, whereas critics often argue (1) that MZ > DZ can be explained largely or entirely by environmental influences, or (2) that MZ > DZ is uninterpretable because the twin method is unable to disentangle the potential behavior-shaping influences of genes and environments.

Mitchell Fails to Respond to Challenges to the EEA

On 3/2/2020, I asked Mitchell on Twitter (@WiringTheBrain) to defend his claim that MZ and DZ environments are equal, and that adoption studies are able to disentangle potential nature and nurture influences.10 This thread can be found HERE. Mitchell was active on Twitter, and I was asking him to defend the validity of the EEA in my role as a recognized critic of twin research.11 He did not respond.

Mitchell, who believes that “intelligence is humanity’s defining characteristic” (p. 155), mentioned in Innate that the average 1970s-era IQ score in his native Ireland was 85, “a massive difference from the average of 100 seen in the United Kingdom at the same time.” However, he wrote, due to “increasing urbanization, industrialization, and prosperity, with concomitant increases in nutrition, health, and average length of education,” IQ now averages a “massively” higher 100 in Ireland (p.167). When I asked him on Twitter how this observation is consistent with his claim that human behavioral differences are “innate,” he responded, “That might give you a clue that I don’t in fact hold or argue for the position you think I do.” Indeed, in the book he wrote that intelligence is not “an immutable, innate trait” uninfluenced by environmental factors (p. 165). Very well, but as one Twitter commenter noted, instead of Innate, the title of Mitchell’s book could have been the far less marketable “Malleable.”

Charles Murray’s 2020 Human Diversity: The Biology of Gender, Race, and Class

Charles Murray was a co-author of the very controversial and widely publicized 1994 book The Bell Curve, where he and psychologist Richard Herrnstein argued that intelligence is “significantly heritable,” and that some portion of the observed mean IQ-score difference between white and black Americans is caused by genetic factors. (Important rebuttals to The Bell Curve can be found HERE and HERE; more on Herrnstein’s history in the IQ-genetics debate can be found HERE). Like Mitchell, Murray’s arguments in Human Diversity were based largely on twin method results. (For a critical review of Human Diversity, see Michael R. Jackson’s 2020 article “A Hereditarian Edifice on a Foundation of Sand.”)

Of the three books I discuss in this review, only Murray acknowledged that criticism of twin studies even exists, and he recognized the reality that MZ pairs grow up experiencing more similar treatment and social environments. Nevertheless, on the basis of two faulty arguments, he claimed that the twin method produces roughly valid “heritability estimates.”

Murray Invokes Argument B

Murray’s first argument in defense of the EEA was that the greater environmental similarity of MZ pairs does not introduce a major bias in twin studies. Murray invoked the Argument B “trait-relevant” position when he cited the opinion of a pair of twin researchers, who concluded that greater MZ versus DZ environmental similarity “is a violation of the EEA only if these aspects of the environment are etiologically relevant to the phenotype of interest.” Murray wrote that these researchers found that, apart from “a few exceptions,” it doesn’t seem to make a difference whether MZ environments are more similar, a finding he claimed applied to intelligence, personality, schizophrenia, and other areas of behavior (p. 217).

Adding to my earlier description of the fallacy of Argument B, Murray’s acknowledgement that MZ pairs grow up experiencing more similar environments than DZ pairs provides enough information to invalidate genetic interpretations of MZ > DZ, because twin pairs sharing a more similar genetic relationship to each other also share more similar environments. In his evaluation of family studies showing that behavioral characteristics (traits) “run in the family,” Kevin Mitchell concluded that even though siblings share a certain degree of genetic similarly, the fact that they also share a similar family environment means that, in a family study, “we cannot distinguish possible effects of nature from those of nurture” (Mitchell, p. 12). Very true, but regardless of what Mitchell, Murray, and Plomin think, the same conclusion holds true for twin studies.

Murray Invokes Argument C

Murray’s other argument in defense of the EEA was that, whereas unequal MZ and DZ environments might lead to an overestimation of heritability, assortative (non-random) mating leads to an underestimation of heritability. “Assortative mating” has been defined by others as “the tendency for people to choose mates who are more similar (positive) or dissimilar (negative) to themselves in phenotype characteristics than would be expected by chance.” Murray “summed up” his position by invoking what we might call the Argument C defense of the EEA, which holds that violations of the “no assortative mating” and “equal environment” assumptions roughly cancel each other out in favor of genetics. According to Murray,

“Twin studies have come under criticism for overstating the role of genes. The reality is that violations of the random mating assumption are common and lead to modest understatement of the role of genes, whereas violations of the equal environments assumption have even more modest effects in the other direction and are uncommon. Overall, heritability as estimated by twin studies appears to be accurate, with errors tending on net to slightly underestimate heritability rather than overestimate it.” (p. 217)

To identify only one problem with this argument, let’s suppose that the environmental null hypothesis—which states that there are no genes for behavior—is true. In this case there would be no assortative mating bias because mating patterns would have no direct genetic influence on human behavior, and MZ > DZ would be completely caused by non-genetic factors. Murray’s claim that assortative mating patterns lead to a “modest understatement of the role of genes” circularly assumes in advance that the environmental null hypothesis is false. A twin study, however, is an experiment designed to test whether the environmental null hypothesis is false. The findings of this experiment cannot be based on a built-in assumption that it is false.12

As we see, Argument C represents just another illogical failed attempt to support the validity of the EEA.

Murray Highlights a Fatally Flawed 2015 Twin Study Meta-Analysis

Murray (pp. 221-224) discussed a 2015 twin study meta-analysis (analysis of combined studies) by Dutch twin researcher Tinca Polderman and her colleagues. He described this study as a “mammoth undertaking, effectively covering all twin studies from 1958 to 2012” (p. 221), and he highlighted the study’s heritability estimates as well as its supposed support for the idea that “the shared environment usually plays a minor role in explaining personality, abilities, and social behavior” (p. 8). Polderman and colleagues pooled 2,748 twin studies and then calculated heritability estimates, based on MZ > DZ, for more than 17,000 physical, medical, and psychological characteristics (traits). Amazingly, they failed to defend or to even mention the EEA, that is, the assumption upon which their genetic interpretations of all 2,748 studies were based! In a 2015 article, I reviewed this study and concluded,

“Because twin method MZ-DZ comparisons are based on the false assumption that MZ and DZ pairs experience equal environments, it doesn’t matter whether researchers pool together the results of 5 twin studies, 500 twin studies, 2,748 twin studies, or a million twin studies. Like the individual studies, the pooled results for behavioral characteristics can be completely explained by the non-genetic (environmental) influences experienced to a much greater degree by MZ versus DZ twin pairs. Two wrongs don’t make a right, and 2,748 environmentally confounded twin studies pooled together don’t make a genetic finding, at least as it relates to human behavioral differences” (italics in original).

Behavioral geneticist Eric Turkheimer wrote approvingly in a 5/28/2015 blog posting that the Polderman study “represents an inconceivable amount of work. And the meta-analysis itself is beautifully executed. The graphs are striking, the numerical analysis is sophisticated.” They are indeed impressive, but if they are all based on the same false assumption, then Murray’s, Polderman’s, Turkheimer’s, and others’ genetic interpretations of the results are massively misleading, because we could just as easily interpret these results in favor of environmental factors.

Murray’s Response on Twitter

On 2/15/2020, I tweeted arguments (HERE) against Murray’s (@charlesmurray) defense of the twin method and the EEA, adding a link to one of my own online articles for a more detailed analysis. The only response Murray provided was to my original Twitter post, where he wrote, “The assumptions underlying twin studies have been exhaustively studied. Happily, you can examine the debate in detail with full documentation if you buy a copy of Human Diversity.” Of course I already had a copy of Human Diversity, where in Chapter 10 (pp. 215-217) Murray presented a one-sided “debate” which consisted of his citing a few researchers who concluded that the twin method did not contain major environmental biases. On 2/24/2020 I tweeted a follow-up request to Murray to defend his positions, but he did not respond.

On 4/10/2020, I asked Mitchell and Murray on Twitter to comment on a table showing that MZ pairs experience much higher levels of identity confusion and attachment than experienced by DZ pairs:

“To @WiringTheBrain & @charlesmurray: The main arguments in your recent books are based on the assumption that MZ and DZ twins experience ‘equal environments.’ Could you please explain how this assumption is consistent with the results in the table below[?]”

Neither Mitchell nor Murray responded to this post. Although I posted it well into the 2020 COVID-19 pandemic crisis, both authors were very active on Twitter throughout February, March, and April of that year, and both used Twitter to promote their books during this period.

Conclusions

In the greatest negative “test” of the EEA ever performed, nearly a century of research has shown us that twin pairs experiencing similar environments and high levels of identity confusion and attachment—MZs—behave much more alike than do pairs experiencing less similar environments and lower levels of identity confusion and attachment—DZs. Because MZ > DZ is explainable on environmental grounds, genetic interpretations of MZ > DZ in behavioral twin studies past, present, and future must be rejected outright. And yet, influential contemporary authors such as Plomin, Mitchell, and Murray, not to mention most social and behavioral science textbook authors, continue to cite these studies in support of genetic theories of behavior.

Regardless of the authors’ views or intent, false claims of behavioral “blueprints” and behavioral “innateness” have major social and political implications. These claims support the idea that changing or improving people’s environments won’t accomplish very much, and can lead to the persecution of people and groups seen as “inferior” or biologically predestined to behave in socially disapproved ways. They can even influence political decisions about how to deal with a pandemic. In other words, though completely false, these claims are dangerous.

To varying degrees, the arguments found in the recent books by Mitchell, Murray, and Plomin are based on the validity of the controversial MZ-DZ equal environment assumption, yet apart from one Murray endnote (p. 426), none of these authors mentioned the name or cited the work of a critic of twin research. Of the three, only Murray made an (albeit weak) attempt to defend the EEA. I am not aware of any valid argument in support of the EEA, and Mitchell, Murray, and Plomin appear unable to supply one. For this and many other reasons, the major claims found in their recent books must be rejected.

***

Mad in America hosts blogs by a diverse group of writers. These posts are designed to serve as a public forum for a discussion—broadly speaking—of psychiatry and its treatments. The opinions expressed are the writers’ own.

Why is the study of identical twins important to determining in and how environmental factors affect gene expression?

What is the equal environments assumption in twin studies?

Why are twin studies problematic?

Will identical twins raised in different environments have the same traits?