Psychology › Section G · Thinking and intelligence › Topic G.6
The source of intelligence
Why do people differ in measured intelligence? The honest answer is that genes and environments both matter, in ways that interact and change across life. This topic explains what twin and adoption studies show, what heritability does and does not mean, and how researchers interpret differences between groups.
Asking the right question
"Is intelligence caused by genes or by upbringing?" sounds simple but cannot be answered for a single person, because every ability grows from genes and environment acting together from conception onwards. Researchers ask a different question: across a population, how much of the variation in scores is linked to genetic differences between people, and how much to differences in their environments? Family, twin and adoption designs, introduced in the human genetics topic, address it, and their answers can change when environments change.
Evidence from families
The closer the genetic relationship, the more similar people's IQ scores tend to be. Summaries of many studies give approximate correlations of about 0.85 for identical twins raised together, about 0.6 for fraternal twins raised together and about 0.45 for ordinary siblings raised together. Identical twins raised apart, studied most famously in the Minnesota Study of Twins Reared Apart from 1979, correlate at about 0.7 to 0.75. Adoption studies add that adopted children's scores in adulthood resemble those of their biological parents more than those of their adoptive parents. Genes clearly contribute, but so does environment, since identical twins are far from perfectly correlated.
What heritability means
Heritability is the proportion of variation in a trait, within a particular population at a particular time, that is associated with genetic differences between its members. Estimates for adult intelligence in Western samples are often around 0.5 to 0.8. Heritability does not say what share of one person's intelligence comes from genes, which is meaningless. It does not mean a trait is fixed: height is highly heritable, yet average height rose markedly across the twentieth century as nutrition improved. And because it depends on how much environments vary, the same trait can show different heritability in different populations.
Worked example: reading twin correlations
Suppose identical twins raised together correlate at 0.85 and fraternal twins raised together at 0.6. 1. Both kinds share a home, but identical twins share all their genes and fraternal twins about half of their varying genes. 2. So the higher identical-twin correlation points to a genetic contribution. 3. A rough estimate, Falconer's formula, doubles the difference: 2 × (0.85 − 0.6) = 0.5, suggesting a heritability of about 0.5. 4. Caveats: identical twins may be treated more alike, twins raised apart were often placed in similar homes, and twins share a womb. 5. Conclusion: the evidence supports a genetic contribution, but any single number is an estimate.
Heritability shifts with age and context
Heritability of intelligence rises with age, from roughly 0.2 in infancy to around 0.4 in middle childhood and higher in adulthood, perhaps because people increasingly choose environments that suit their tendencies. It can also vary with circumstances. Eric Turkheimer and colleagues reported in 2003 that heritability was much lower among very poor US families than among better-off ones; later studies find this pattern more consistently in the United States than in Western Europe or Australia. At the molecular level, intelligence is influenced by thousands of genetic variants, each with a tiny effect; there is no single gene for intelligence.
Environments that move scores
Environmental effects are well documented. A 2018 analysis of many studies estimated that each extra year of schooling raises IQ by about 1 to 5 points. In French studies, children adopted after early deprivation gained more IQ points when placed in better-off families than in poorer ones. Iodine deficiency, lead exposure and severe malnutrition all lower scores, and correcting them helps. Poverty brings chronic stress, poorer schools and fewer learning resources. The Flynn effect, rising scores over a few decades, is too fast to be genetic and shows how much environments can shift the averages of whole populations.
Group differences in scores
Average test scores differ between some groups, such as socioeconomic and racial or ethnic groups in the United States, and researchers widely agree on several points. The gap in average scores between Black and white Americans narrowed by several points in the last decades of the twentieth century. A 1996 American Psychological Association task force concluded that there was no direct evidence for a genetic explanation of that gap, and no such evidence has been established since. Race is a social category that does not map neatly onto genetic variation. Differences in income, schooling, discrimination, health and test familiarity are plausible contributors, and the size of each is still studied.
Worked example: within versus between
The geneticist Richard Lewontin offered this thought experiment. 1. Take a sack of genetically varied seeds and split it at random into two handfuls. 2. Plant one in rich soil and one in poor soil, with everything else equal within each field. 3. Within each field, height differences are almost entirely genetic, because the soil is the same for every plant: heritability is high. 4. Yet the gap between the two fields is entirely environmental, caused by the soil. 5. Conclusion: knowing that a trait is heritable inside groups does not, by itself, tell you the cause of a difference between groups.
Controversy, courts and stereotype threat
In 1969 Arthur Jensen argued that genes probably contributed to racial gaps in scores; the claim drew fierce scientific and ethical criticism, and direct evidence for it has not emerged. In Larry P. v. Riles, filed in 1971 and decided in 1979, a federal court barred California from using IQ tests to place Black children in classes for the "educable mentally retarded", as the tests had not been validated for that use. Claude Steele and Joshua Aronson reported in 1995 that stereotype threat, worry about confirming a negative stereotype, lowered test performance. Later large replications suggest the effect is real in some settings but smaller and less consistent than first thought.
Learning disabilities are different
A learning disability, called specific learning disorder in current manuals, is a persistent difficulty with one area of skill that is not explained by overall intellectual ability, and it often occurs in people of average or higher intelligence. Dyslexia, the most common, mainly involves difficulty linking letters to speech sounds, which makes reading slow and effortful; reversing letters such as b and d is common in all beginning readers and is not its defining sign. Dysgraphia affects handwriting and written expression, and dyscalculia number sense and arithmetic. Only a qualified professional can identify one.
Common mistakes to avoid
First, applying heritability to individuals: it describes variation in a population. Second, equating heritable with unchangeable: heritable traits such as height respond to environment. Third, assuming within-group heritability explains between-group gaps, which Lewontin's seeds show it cannot. Fourth, presenting genetic explanations of racial gaps as established: there is no direct evidence for them. Fifth, citing stereotype threat as a large, settled effect: replications show it is smaller and inconsistent. Sixth, confusing learning disabilities with intellectual disability, or treating letter reversals as proof of dyslexia.
Key takeaways
Genes and environment both shape intelligence, and the useful question is what explains variation in a population. IQ correlations rise with genetic relatedness: about 0.85 for identical twins raised together, 0.7 to 0.75 raised apart, 0.6 for fraternal twins. Heritability, often 0.5 to 0.8 in adults, applies to populations, not individuals, rises with age and can fall in deprived settings. Schooling, adoption into better-off homes and good nutrition raise scores. High heritability within groups cannot explain gaps between them, and there is no direct evidence for genetic explanations of racial gaps. Learning disabilities are specific difficulties, not low overall intelligence.
Flashcards
Flashcard
What explains variation in a population, not in one person
Test your understanding
Question 1 of 5
A study reports that the heritability of IQ in a population is 0.6. Which interpretation is correct?
Question 2 of 5
Identical twins raised together correlate at 0.80 and fraternal twins raised together at 0.55. Using Falconer's formula, what is the rough heritability?
Question 3 of 5
What does Lewontin's seed thought experiment show?
Question 4 of 5
Which statement best reflects current scholarship on average IQ differences between racial groups in the United States?
Question 5 of 5
How does a specific learning disorder such as dyslexia differ from intellectual disability?
