Nature or nurture? Or both?
Proponents of the argument over what primarily shapes intelligence—what “intelligence” actually means–bring salient points to the discussion, and in neuroscience it is generally considered that nature and nurture work are in many ways indistinguishable and work together. Neither operates in isolation to make intelligence what it is.
From the psychophysiological perspective, our brains are the product of genetic coding, as our other organs and indeed everything about our bodies. However, based on identical twin studies done over time, even with identical genetics different individuals have different exposures and experiences in life, and these modify one’s biology and development, and help inform and develop thoughts, thought processes and perhaps thinking itself. Assessing intelligence in the context of genetics is, from a practical standpoint, very much a study of genotypes vs. phenotypes – an individual’s heredity (genotype) and how it relates to the actual observed traits of that individual (phenotype).
How much of intelligence is inherited?
Scientists know it is substantial; humans are not born a blank slate, and limits, with great variance, are proscribed. But can these hard-wired differences yet be traced to specific genes?
Looking at very large cohort studies, including tens of thousands of participants in some cases, researchers have found certain similarities in brain regions that tend to correlate with certain genes. For example, researchers have found that two brain structures, the putamen and hippocampus, are particularly associated with genetic inheritability—as far as their volume is concerned. To not read into the research data more than actually exists, these results simply mean that the size (volume) of these brain structures is similar between those who are linked by inheritance. The researchers observed that there are “several common genetic variants underlying variation in different structures within the human brain. Many seem to exert their effects through known developmental pathways including apoptosis, axon guidance and vesicle transport.”
For general intelligence (known in the parlance as ‘g’—this is what is tested for in classic IQ tests), there are data to suggest, as researchers from New York recently, concluded, “about half of the variance in g is accounted for by common genetic variation among individual. We conclude that the molecular genetics of psychology and social science requires approaches that go beyond the examination of candidate genes.”
Likewise, other researchers have found that associations in common genetic variants are linked with cognitive phenotypes. Not all research agrees with this line of logic either, and in the case of this particular study, the researchers used a proxy phenotype instead of a direct measure of intelligence or cognitive performance—self-reported years of schooling (educational attainment). Certainly some methodological issues contribute to uncertainty in these findings—self-reported years of schooling assumes that 1) ‘self-reporting’ isn’t subject to respondent bias (which it is) and 2) that ‘years of schooling’ is a reliable proxy for intelligence (which is unlikely; there are overwhelming factors such as demographics, access to education, etc.).
Though no research has come to a conclusive end on this topic insofar as it relates to genetic analysis, there have been several converging lines of research which suggest that general intelligence (g) is about as heritable (a term that means traits which are inherited) as height. This does not imply that intelligence and height are related, just that it seems that the amount of explainable variance in height among families is about the same as the explainable variance in intelligence.
Think of intelligence as you think of race horses. The semen from the fastest horses sells at a huge premium, which is why many Triple Crown level horses are often retired at such a young age and put out to stud. Their semen will breed many stars, but also their share of duds—because speed, like height, can vary by the potluck of how genes mix. Similarly, If you are born to a family to Inuit Indians (Eskimos), it is unlikely that one of your children will grow up to play center in the National Basketball Association.
The same holds true for intelligence; if your parents have high IQs, you more than will likely do too; children born from parents with low IQs are more than likely to have low intelligence. There is huge individual variance of course; a champion race horse sometimes emerges from lackluster breeds; short parents occasionally have a tall child; and the offspring of dullards may turn out to be quite bright. That’s genetic variability. But genes do proscribe possibility. The nature/nurture story is complex. Speed, intelligence and height are termed ‘complex traits’ because they involve a multitude of factors (socioeconomic status, experiences, nutrition, etc.) and are very much phenotypes (the sum of genetic code plus environment and experiential exposures—nature and nurture).
So while genetic associations with intelligence are likely to be real, and broad qualitative estimates of inheritance suggest that this is true, there is a lot of work to still be done. Genetic associations have been found (as described above) with volume of certain brain regions—but brain volume itself isn’t necessarily associated with intelligence. Proxies (such as educational attainment) need to be used to see if certain phenotypes of cognitive performance (intelligence) are related to genetic predispositions—and these proxies have moderate-to-high error variance.
And even those studies which suggest certain genes show relationships with intelligence have detractors who feel that many of these signals could be false positives. Further to this, some lines of research suggest that 50 percent—or less—of intelligence is explained by the genetic models—which may mean that many of these constructs is in error (and formally unaccounted-for).
Also, these (very) large cohort studies has laid bare one of the problems of null hypothesis significance testing: With so many data points, it becomes much easier to detect an effect (easier to identify smaller differences); This also means that statistical significance can be reached in cases where there is no practical significance, making false positives more likely.
Ben Locwin, PhD, MSc, MBA is a Contributor to the Genetic Literacy Project and is an author of a wide variety of scientific articles for books and magazines. He is also a researcher and consultant for a variety of industries including behavioral and psychological, aerospace, food and nutrition, pharmaceutical, petrochemical, and academic. Follow him at @BenLocwin.
