28-1 What evidence points to a genetic influence on intelligence, and what is heritability?

Do people who share the same genes also share mental abilities? As you can see from FIGURE 28.1, which summarizes many studies, the answer is clearly Yes.

Identical twins who grow up together have intelligence test scores nearly as similar as those of the same person taking the same test twice (Haworth et al., 2009; Lykken, 1999). (Fraternal twins, who typically share only half their genes, differ more.) Even when identical twins are adopted by two different families, their scores are very similar. Estimates of the heritability of intelligence—the extent to which intelligence test score variation can be attributed to genetic variation—range from 50 to 80 percent (Calvin et al., 2012; Johnson et al., 2009; Neisser et al., 1996). Identical twins also exhibit substantial similarity (and heritability) in specific talents, such as music, math, and sports. Heredity accounts for more than half the variation in the national math and science exam scores of British 16-year-olds (Shakeshaft et al., 2013; Vinkhuyzen et al., 2009).

Scans of identical twins’ brains reveal that gray- and white-matter volume is similar, and areas associated with verbal and spatial intelligence are virtually the same (Deary et al., 2009; Thompson et al., 2001). Their brains also show similar activity while doing mental tasks (Koten et al., 2009).

Although genes matter, there is no known “genius” gene. One worldwide team of more than 200 researchers pooled their data on the DNA and schooling of 126,559 people (Rietveld et al., 2013). No single DNA segment was more than a minuscule predictor of years of schooling. Together, all the genetic variations they examined accounted for only about 2 percent of the schooling differences. The gene sleuthing continues, but this much seems clear: Intelligence is polygenetic, involving many genes (Bouchard, 2014). Wendy Johnson (2010) likens it to height: 54 specific gene variations together have accounted for 5 percent of our individual differences in height, leaving the rest yet to be explained.

Where environments vary widely, as they do among children of less-educated parents, environmental differences are more predictive of intelligence scores (Rowe et al., 1999; Tucker-Drob et al., 2011; Turkheimer et al., 2003). To see why, consider humorist Mark Twain’s fantasy of raising boys in barrels until age 12, feeding them through a hole. Let’s take his joke a step further and say we’ll give all those boys an intelligence test at age 12. Since their environments were all equal, any difference in their test scores could only be due to heredity—thus, heritability would be 100 percent. But what if a mad scientist cloned 100 boys and raised them in drastically different environments (some in barrels and others in mansions)? In this case, heredity would be equal, so any test-score differences could only be due to environment. The environmental effect would be 100 percent, and heritability would be zero.

Adoption studies help us assess the influence of environment. Consider:

Seeking to untangle genes and environment, researchers have also compared the intelligence test scores of adopted children with those of their family members. These include (a) their biological parents (the providers of their genes) and (b) their adoptive parents (the providers of their home environment). What do you think happens as the years go by and adopted children settle in with their adoptive families? Would you expect the family-environment effect to grow stronger and the genetic-legacy effect to shrink?

If you said Yes, behavior geneticists have a stunning surprise for you. Mental similarities between adopted children and their adoptive families lessen with age, dropping to roughly zero by adulthood (McGue et al., 1993). Genetic influences—not environmental ones—become more apparent as we accumulate life experience. Identical twins’ similarities, for example, continue or increase into their eighties (Deary et al., 2009). In one massive study of 11,000 twin pairs in four countries, the heritability of g increased from 41 percent in middle childhood, to 55 percent in adolescence, to 66 percent in young adulthood (Haworth et al., 2010). Similarly, adopted children’s verbal ability scores over time become more like those of their biological parents (FIGURE 28.2). Who would have guessed?