Most behaviour geneticists will react with pleasure and recognition of their programmatic approach captured in Uchiyama et al. Figure 2. As Uchiyama et al. note, behaviour genetics developed G × SES models (Purcell, Reference Purcell2002), integrating these with measured moderators (Tucker-Drob & Bates, Reference Tucker-Drob and Bates2016), testing these between cultures (Bates, Hansell, Martin, & Wright, Reference Bates, Hansell, Martin and Wright2016; Bates, Lewis, & Weiss, Reference Bates, Lewis and Weiss2013), including non-WEIRD (western, educated, industrialized, rich, and democratic) cultures (Hur & Bates, Reference Hur and Bates2019). It also developed methods and theory to visualise changes in heritability over time (e.g., Briley, Harden, Bates, & Tucker-Drob, Reference Briley, Harden, Bates and Tucker-Drob2015). This is exemplified in, for example, cross-cultural + longitudinal + genetically informed work on dyslexia (Samuelsson et al., Reference Samuelsson, Byrne, Olson, Hulslander, Wadsworth, Corley and Defries2008) and continues with non-transmitted gene measures revealing parental competence and social-network niche construction as envisaged in the target (Bates et al., Reference Bates, Maher, Medland, McAloney, Wright, Hansell and Gillespie2018). In this sense, Uchiyama et al.'s superbly curated encyclopaedia of cutting-edge behaviour genetic, evolutionary and cultural research shows Uchiyama et al.'s future is “already here, just not evenly distributed.” Wider use of this approach would yield significant benefits (Sherlock & Zietsch, Reference Sherlock and Zietsch2018) (note: since the 70s, Eaves and colleagues presaged and implemented much of this programme from cultural transmission and G × E to inspiring non-transmitted genetic models) (Eaves, Reference Eaves2017; Eaves, Pourcain, Smith, York, & Evans, Reference Eaves, Pourcain, Smith, York and Evans2014; Heath et al., Reference Heath, Berg, Eaves, Solaas, Corey, Sundet and Nance1985). Similarly integrative figures exist in evolution and in culture research of course, for example, Lumsden and Wilson (Reference Lumsden and Wilson2005).

Applications and tests of bio-psycho-sociocultural models require samples that are unnecessarily scarce. For instance, there is only one test of G × SES for intelligence in Africa (Hur & Bates, Reference Hur and Bates2019). We agree with Uchiyama et al. that many, many more, not fewer, genetically informative twin and molecular studies in more diverse global samples are urgently needed, if the target article's ambitions are to be achieved. Uchiyama et al. thus, highlight exciting opportunities in macro, micro, and occulted cultural clusterings combined with genetics. Moving beyond postcode and parental education, and doing this globally with measured genetics needs investment but would repay it richly. The 11-year-old in a poor, neighbourhood but who edits Wikipedia using creative commons MIT chemistry lectures, and who collaborates on zero-to-one endeavours with friends on discord is invisible under typical SES metrics. So too the canalisation or invariance of potential and capability under challenging environments is too seldom examined.

Finally, I applaud the authors for their focus on intelligence. Birthed in equal parts in education, intervention, genetics, I/O psychology, sociology, and, in part because of controversy, in methodological rigour and innovation, intelligence research reflects much of what is desired by the target article. Perhaps more than at any time in the past 40 years, however, research in this area is under pressure and this is likely a tragic cultural error. Intelligence (alongside equally under-researched traits of goal-focussed ambition and creative zeal) is essential to the origin of the wonders of invention inspiring Uchiyama et al. – from vitamin-D supplements to co-opting our cellular machinery using consciously designed mRNA, masking genetics in precisely the ways the target article highlights.

Three questions for the research programme: For how long could a population thrive if furnished with all of today's inventions and institutions, but shorn of ability-associated genetic polymorphisms? Genetic causes of ability are likely crucial to “keeping the mask up” and emphasising this seems valuable. Second, does the theory reserve sufficient space for the intransigence of phenotypes? Vitamins provide clean examples of efficient masking of genetics by invention. However, precisely because of their genetic complexity, many more phenotypes have resisted reengineering, reflected in unyielding mental diseases such as schizophrenia or depression, and, glaringly, the resolute unwillingness of IQ to yield up durable and deep increases above the effects of school exposure (Ritchie, Bates, Der, Starr, & Deary, Reference Ritchie, Bates, Der, Starr and Deary2013), or, even more glaring, the absence of any heritability suppressing large effect vitamin-C style intervention craved in education (e.g., Li & Bates, Reference Li and Bates2019, Reference Li and Bates2020). As Ceci (Reference Ceci1996) noted, by realising potential, education per se opposes rather than advances egalitarian goals. Does the theory speak to this?

Studies corroborate the ability of environment to amplify, reduce, or even reverse G × E effects on intelligence (Tucker-Drob & Bates, Reference Tucker-Drob and Bates2016). What they have not shown is similar sign changes on genetic main effects. It is now possible to directly test this with measured variants. So a third question: Regarding sign-invariance of genetic effects rather than effect size changes, what fraction of DNA variants associated with traits such as cognition or reading skill do Uchiyama et al. predict will reverse their effects under conditions which raise mean educational outcomes? This has not been seen to date (Samuelsson et al., Reference Samuelsson, Byrne, Olson, Hulslander, Wadsworth, Corley and Defries2008), with results supporting conventional mostly additive models in genetics (Hill, Goddard, & Visscher, Reference Hill, Goddard and Visscher2008).

This brings us to a final point. It often seems that rather than the Elephant that is culture being ignored, it is the less visible DNA-based machinery maintaining the galactically complex machinery of the body and brain that is ignored. Moreover is often implied that because genetic variance is always above zero, malleable, and dynamic, then perhaps genes place no limits on the phenotype. Can the authors speak to how they articulate the difference between the complex learning machinery that is the “blank slate” provided by genetics and what is written on it? Asked by a friend how he should educate his child, the originator of the blank slate idea, Locke, recognised that because knowledge was so much more taught rather than discovered by children, that precisely what was being taught was crucial to their later attainment and character. But it is less clear that this follows for the “slate” itself, as opposed to what is stored on it. A useful direction, then, would be to incorporate a well-specified distinction between changes of state in the sense of what is stored in the system, from changes to the nature of the system in the sense that vitamin-C changes the nature of the biological system, or an engineered molecule may create a de-novo capacity: for instance, allowing the immune system to destroy a cancer cell. Can the theory capture such distinctions? For example, school changes knowledge about the topics taught (storage), but much of the capacity to operate on that knowledge seems orthogonal to this (Engelhardt, Briley, Mann, Harden, & Tucker-Drob, Reference Engelhardt, Briley, Mann, Harden and Tucker-Drob2015; Ritchie, Bates, & Deary, Reference Ritchie, Bates and Deary2015). Widespread understanding and acceptance that genetics is at least as diverse and impactful as culture in individual differences is, perhaps, the challenge of this decade and certainly crucial to understanding many practical issues in education and society. The target article is a step towards meeting this challenge and we commend it.