Interpreting the findings of behavior genetics studies is fundamentally a quest for meaning. This is certainly true in a field like education where the purpose of research is ultimately to improve the provision of education, and by extension, improve the outcomes attained by students. The bold concluding claim of Uchiyama et al., “Nothing in behavioral genetics makes sense except in the light of cultural evolution” (sect. 6, para. 3), should prompt us to reexamine the meaning of decades of educational behavior genetics studies. If heritability estimates are consistently high for educationally relevant traits what should this information mean? And, how should it be applied to the real world of students, teachers, schools, and education systems?

Uchiyama and colleagues identify education as a prime example of a culturally transmitted phenomenon. Indeed, access to universal education is so embedded in western societies (the locus of much behavior genetics research) that it is difficult to imagine a world in which it did not exist. It may be self-evident to point out that educational contexts are not static: They continuously evolve both within and between systems, influenced by local and global cultural and policy shifts (e.g., Addey, Reference Addey2017; Vadeboncoeur, Reference Vadeboncoeur and Richardson1997). For example, development of theories of learning, constantly changing administrative structures, variable recommendations on instructional practices, and frequent reform movements are features of many educational systems (Rose, Reference Rose2006; Stanovich & Stanovich, Reference Stanovich and Stanovich2003; Woods, Reference Woods2021). For readers who come to the behavior genetics literature from the field of education research, discussion of heritability estimates may seem to put an overabundance of emphasis on quantifying the genetic influence on phenotypes at the expense of examining the cultural variation within and between school systems, or over time.

In behavior genetics, educational cultural contexts are encompassed by the terms shared and nonshared environment. The complexity and richness of culture is, perhaps necessarily, stripped out in the textbook definitions for shared environment, “all nongenetic influences that make family members similar to one another,” and nonshared environment, “all nongenetic influences that are independent for family members, including error of measurement” (Plomin, DeFries, Knopik, & Neiderhiser, Reference Plomin, DeFries, Knopik and Neiderhiser2013, p. 96). In behavior genetics, the environment is represented as the reverse of the genetic coin: whatever remains in the absence of genetic effects. While this dichotomy is a simplification of a more complex reality, the emphasis on genes versus environments, rather than genes situated within environments, tends to obscure both cultural evolution within contexts, and differences between educational systems that may affect educational provision and outcomes.

The fact that heritability estimates are population statistics is well understood and not trivialized among behavior geneticists. Nonetheless, the central importance of interpreting heritability as a population statistic (e.g., Smith, Reference Smith2011) is often glossed over in discussions about “genes for” traits like reading and mathematics ability in educational studies. Inferences are made that by measuring the heritability of traits we have somehow found out something definitive about the “genetics” of that trait. The central contention of Uchiyama and colleagues is that such overarching conclusions are not supportable given the presence of cultural homogeneity within twin samples, and continually evolving educational cultures.

Uchiyama and colleagues make a compelling case that high heritability estimates of educational phenotypes may not indicate optimal educational environments, as is often claimed (e.g., Kovas, Tikhomirova, Selita, Tosto, and Malykh, Reference Kovas, Tikhomirova, Selita, Tosto, Malykh, Kovas, Malykh and Gaysina2016). High heritability may instead reveal something about the cultural homogeneity of environments from which twin samples are drawn. Heritability may be equally high in poor education systems as in excellent education systems, so long as the system is relatively homogeneous, and all students obtain the same poor or excellent education. For example, heritability estimates for reading skills are not remarkably divergent in educational systems where overall student attainment is relatively high (e.g., Hong Kong Chinese; Chow, Ho, Wong, Waye, and Bishop, Reference Chow, Ho, Wong, Waye and Bishop2011) compared with systems where average attainment is consistently relatively lower (e.g., Australia, see Byrne, Olson, & Samuelsson, Reference Byrne, Olson, Samuelsson, Kilpatrick, Joshi and Wagner2019; Thomson, Hillman, Schmid, Rodrigues, & Fullarton, Reference Thomson, Hillman, Schmid, Rodrigues and Fullarton2017). This point alone is a problem for the context-free interpretations of behavior genetics findings in education and should prompt some hesitation about the broad extrapolations that are made about the genetic basis of academic abilities.

The definitional and interpretational difficulties described above lead next to questions around the applicability of behavior genetics research to educational policy and practice. There is disagreement about the extent to which behavior genetics research can and should be extrapolated to educational policy and practice recommendations (Asbury & Wai, Reference Asbury and Wai2020; Byrne et al., Reference Byrne, Little, Olson, Larsen, Coventry and Weymouth2020; Panofsky, Reference Panofsky2015). Nonetheless such recommendations are made. One of the recurrent themes of educational behavior genetics is the idea that understanding the genetic etiology of traits will lead to improved interventions (Shero et al., Reference Shero, van Dijk, Edwards, Schatschneider, Solari and Hart2021), redesign of school systems (Asbury & Plomin, Reference Asbury and Plomin2014), and personalized education (Kovas et al., Reference Kovas, Tikhomirova, Selita, Tosto, Malykh, Kovas, Malykh and Gaysina2016). Such claims, however, elide the fuzziness of interpretation inherent in behavior genetics. Turkheimer (Reference Turkheimer2015) argued that the “layer of theory between data and their interpretation is thicker and more opaque [in behavior genetics] than in more established areas of science” (p. s32). That is, interpretations of behavior genetics research rely on an accepted theoretical understanding of heritability estimates (e.g., Harden, Reference Harden2021), which underlies suggested applications in the real world. However, if heritability estimates in educational studies are, (a) inflated to an unknown extent (e.g., Coventry & Keller, Reference Coventry and Keller2005; Keller & Coventry, Reference Keller and Coventry2005), arguably by sample-specific cultural homogeneity, and, (b) can be confounded by cultural evolution, then the suggested applications to policy and practice could easily be wrong. The value of Uchiyama and colleagues to the field of educational behavior genetics is in their articulation that accepted explanations of heritability estimates are indeed contestable theoretical positions and can (should?) be interpreted differently via the lens of cultural homogeneity and evolution.