In their 2013 article for Current Directions in Psychological Science, Tucker-Drob, Briley, and Harden propose a transactional model for the heritability of cognitive ability. The basis for this model is the well-documented biological phenomenon, the basis of phenotypic variation, the regulation of gene expression by environmental stimuli. The authors apply this finding to the characteristic of cognitive ability and support it using evidence from twin studies. They propose that, through self-selection of stimulating environments and experiences, individuals with inherited high cognitive ability cause the expression of more cognitive ability genes, which in turn feeds the selection of more stimulating environments. However, beyond this straightforward and reasonably supported model they also provide a guiding interpretation which places heavy emphasis on the significance of genetic heritability of high cognitive ability. They ignore shortcomings in the data and make claims which are both unnecessary and unsupported by the evidence. The most significant of these are presented, with rebuttal.

Having claimed that cognitive ability is undeniably a heritable trait (“no longer a question of serious scientific debate”) (Tucker-Drob, Briley, and Harden, 2013; 1), the authors propose that this trait is subject to the same regulatory forces of expression as affect everything from the production of digestive enzymes to the melanin content of the skin. Although undoubtedly a simplification, this claim seems to hold true. The evidence brought to support the heritability, on some level, of cognitive ability is reasonably robust. In fact, it follows naturally from the fundamental theoretical basis underpinning the whole diversity of individual characteristics: the regulation of gene expression. As a trait which varies between individuals, changes over the lifetime, and is subject to heritable factors, it is reasonable (if not inevitable) that cognitive ability should have some form of expression regulation mechanism in the genome. This proposed mechanism is sound and well founded. The authors then propose that heritability of cognitive ability is environmentally influenced, and further that high cognitive ability is more heritable than low cognitive ability.

However, the shortcomings of their specific model begin to show as early as the first example. The authors discuss the average educational attainment of Norwegians, which increased during the 20th century due to social and regulatory changes favouring education, in the context of measured heritability of educational attainment increasing during roughly the same period. The educational attainment data reports numbers for 1960 vs. 2000, while data on the heritability thereof reports “before 1940” and after, while itself being published in 1985. Ignoring the obvious flaw of comparing temporally un-matched samples of population data, the example provides no evidence of what the authors claim: educational attainment is not a measure of cognitive ability, but of the strength of social norms and reach of educational programs. Given Norway’s system of public education (10 years compulsory, 3 years ‘optional’ in name only, 3–8 years optional university level), the c. 2000 attainment of 11.86 years average demonstrates that 91% of individuals complete the first 13 years. Completion of compulsory public education is not a matter of individual aptitude or achievement, but of social norms and education system policy. Nine in ten completion of compulsory education is not indicative of high cognitive ability, and the data on heritability of cognitive ability are too narrow to offer a comparison.

The next section of the piece, which introduces the underlying framework of “gene-environment correlation” before explaining the authors’ proposed transactional model of cognitive ability specifically, serves as evidence against the very application which the authors make. They put it simply: “a broad array of presumably ‘environmental’ experiences—such as negative life events, relationships with parents, and experiences with peers—are themselves heritable” (Tucker-Drob, Briley, and Harden, 2013; 2). Clearly a broad range of experiences and qualities can be inherited. While the authors propose this to support the heritability of cognitive ability, it just as easily supports the heritability of academic inclination, curiosity, and even academic performance, factors which are often measured as indicators of cognitive ability. In fact, these very listed examples of heritable experiences are known to influence cognition themselves, as are other factors which fit the same category, such as parent attachment (Fryers & Brugha, 2013; 9). This gets into the base of the authors proposed model: that gene-environment correlation acts in concert with gene expression regulation to promote the achievement of cognitive ability potential in individuals with a genetic disposition for it. In particular, the authors posit that the natural outcome of this process is as observed: individuals with high cognitive ability are likely to have inherited it. In fact their evidence shows that, rather than cognitive ability itself being heritable, the indicators often measured for cognitive ability are heritable as are many influences on the development of cognitive ability.

Further complicating the authors’ model is their poor handling of shortcomings in their base evidence, twin studies. Due to restrictions on data retention and access, international adoption, research funding, and scope, the majority of twin studies showing heritability of cognitive ability sample from within the same country, with the same social norms, educational policy, and even similar socioeconomic context. The authors, however, insist that socioeconomic status is a predictor of both the heritability and level of cognitive ability. This is faulty: if socioeconomic status predicts cognitive ability and socioeconomic status is itself significantly heritable, then it will appear that socioeconomic status influences heritability of cognitive ability. This is demonstrated in the cited data, which break down outside of the US. In particular in social democracies such as Sweden (and Norway) where, the authors admit, strong social programs even the playing field and give all individuals of diverse backgrounds access to the same pool of potential environmental stimuli to select. In these states, it seems that the socioeconomic factors typically having a deleterious effect on  educational access and achievement (and thus, cognitive ability) are substantially reduced. As a result, the heritability of cognitive ability is low. This indicates that the authors’ model of positive experience selection for cognitive ability is fundamentally flawed: a socioeconomically- and gene-environment-linked deficit model better explains inheritance of cognitive ability. Instead of cognitive ability itself being inherited, it is a positive environment, with access to stimulating and diverse education and other experiences (among many factors), which is inherited. It is the absence of this environment and various opportunities of experience which reduces an individual’s achieved cognitive ability. This is clearly shown in the countries where their model does not hold, which have strong social programs.

The trouble with the authors’ model is not that it is based on a flawed mechanism, but that its predictions do not hold. There is no reason that high heritability of cognitive ability should correlate with high cognitive ability itself, absent the other factors. Their model provides no explanation for this claim, yet it is a clear component of their position. The authors themselves have set up evidence and arguments which can be used to make a completely different point: cognitive ability shows high heritability because environmental factors that influence it—such as social norms around education, parenting style, household stability, and academic aptitude and drive—are themselves heritable.

Although the authors identify and describe a compelling mechanism for the influence of environment on cognitive ability, they seem to mistake the causal direction of its operation in their interpretation. Rather than an abundance of stimulating experiences and environments improving an individual’s actualisation of their inherited potential, as is implied, a more parsimonious explanation is that a shortage of these influences suppresses achievement of genetic potential. Their base assumption is made clear on p.3: “The ‘end state’ of this transactional process—high levels of and high heritability of cognitive ability—is therefore expected to differ depending on the quality and availability of environmental experiences.” Emphasis here is on the the end state, particularly the high levels of cognitive ability. In other words, those who demonstrate the higher heritability of cognitive ability also demonstrate higher levels of that ability. High cognitive ability is more closely linked to parent performance than low cognitive ability. If this end state were high heritability of cognitive ability alone, then their argument would hold. However, the association of “high levels of” cognitive ability in this outcome does not follow from their model, and indicates that the observations could be better explained as a default outcome; Ockham’s Razor is well applied. This critical flaw in their proposal is laid bare in their only addressed counter-argument, that their model breaks down in countries with stronger education systems and social programs. Rather than individuals’ ability to self-select unique experiences reinforcing a pattern of gene expression and experience selection to cyclically maximise expression of inherited cognitive ability, it is a much simpler explanation that instead a paucity of these experiences suppresses the expression of inherited cognitive ability.

References

Fryers, T., & Brugha, T. (2013). Childhood Determinants of Adult Psychiatric Disorder. Clinical Practice and Epidemiology in Mental Health : CP & EMH, 9, 1–50. http://doi.org/10.2174/1745017901309010001

Tucker-Drob, E. M., Briley, D. A., & Harden, K. P. (2013). Genetic and Environmental Influences on Cognition Across Development and Context. Current Directions in Psychological Science, 22(5), 349–355. http://doi.org/10.1177/0963721413485087