Nature, Nurture, and Identical Twins (with David Bessis)

Episode Summary

Executive Summary: EconTalk features mathematician David Bessis arguing that claims of strong genetic determinism—especially from twin studies of IQ—are overstated. He contends that the famous “twins reared apart” literature is based on small, biased, and poorly controlled samples, and that human outcomes reflect a complex mix of genes, family, environment, chance, and idiosyncratic life experience. He also connects the debate to mathematics, parenting, and how people understand their own agency.

Main Topics: Critique of genetic determinism (Priority: 5/5): Bessis rejects the idea that genes fully determine intelligence or mathematical talent, arguing instead for a nuanced view in which genes matter but do not dominate destiny. Problems with twin studies (Priority: 5/5): He argues that studies of identical twins raised apart are too small, not truly randomized, and often lack proper controls, making their results much less decisive than popularly portrayed. Heritability, correlation, and interpretation (Priority: 4/5): The discussion explains how percentages like 30%, 50%, or 80% heritability are often misread as destiny rather than as noisy statistical associations with substantial room for variation. The role of environment and idiosyncratic life paths (Priority: 5/5): Bessis emphasizes Turkheimer’s third law: much of human variation comes from factors beyond genes and family, including chance events, developmental dynamics, and personal history. Mathematical talent as learned practice (Priority: 4/5): Drawing on his book Mathematica, Bessis argues that great mathematical achievement comes from method, persistence, and cognitive style, not just innate brilliance. Implications for parenting and education (Priority: 4/5): The conversation explores why parenting and schooling still matter even if their effects are noisy and hard to scale, and why deterministic thinking can discourage useful action. Why people care emotionally about nature vs. nurture (Priority: 3/5): Russ Roberts and Bessis reflect on why these debates are so personally charged: they affect how people understand merit, agency, identity, and self-worth.

Key Arguments: Genes matter, but they are not destiny; the real dispute is whether genetics is one factor among many or the whole explanation. The famous twin studies of identical twins raised apart do not constitute the clean natural experiment they are often claimed to be. Twin samples were small and non-random, with serious selection and environment biases, including shared gestation and often delayed separation. Proper interpretation requires a control group of fraternal twins reared apart; without it, the conclusions are overstated. Heritability percentages are frequently misunderstood: even a 50% or 80% genetic contribution leaves room for major environmental, developmental, and stochastic effects. Human outcomes depend on a unique developmental journey, not just family background or measured social-economic variables. Bessis argues that his own mathematical development and that of elite mathematicians can be explained by practice, methods, and long-term cognitive training, not only innate gifts. The debate matters practically because it influences how people think about parenting, education, and human agency.

Data Points: Date of episode: December 22, 2025 - Opening introduction by Russ Roberts Prior episode on Mathematica: October 2025 - Roberts notes Bessis was previously on EconTalk discussing his book Twin-study sample size (identical twins): About 56–58 pairs - Roberts and Bessis discuss the Bouchard study of twins reared apart Twin-study sample size (fraternal twins): About 30 pairs - Control group mentioned as too small by the researchers Separation timing: About 5 months after birth on average - Bessis says the twins were not literally separated at birth Gestation period: 9 months - Bessis notes twins share the womb, making prenatal environment a confounder Heritability examples: 30%, 50%, 80% - Used to illustrate how different correlations leave varying room for non-genetic factors High-end genetic model example: 100% - Bessis describes a hypothetical perfect linear relationship between polygenic score and phenotype Probability framing for genius: 1 in 1,000,000 or 1 in 1,000,000,000 - Bessis uses these rough figures to describe rare mathematical genius Falconer formula year: 1960 - Roberts and Bessis reference the classic twin-study approximation

Pivotal Quotes: "When you separate two identical twins at birth, raise them in two different, two random families... the two results are barely more divergent than two different tests of the same person." — Russ Roberts: Roberts summarizes the popular interpretation of twin-study claims before Bessis critiques it "The third law says there is a huge amount of variability that is neither coming from your genome nor your family." — David Bessis: Bessis invokes Eric Turkheimer’s third law to explain unexplained human variation "It is super important. If we say everything is decided on the day you were born, we kind of cancel that conversation." — David Bessis: Bessis explains why rejecting strict determinism matters for agency, parenting, and education

Implications: Listeners should treat genetic claims with caution: twin studies are informative but not decisive. The debate affects how we think about education, parenting, talent, and personal agency, and it argues for humility about any simple nature-versus-nurture formula.