Lecture 7: Behavioral Genetics II

What this lecture covers

This lecture continues the course's look at behavioral genetics, moving from the classical twin and adoption studies covered earlier to the molecular techniques used to actually find genes linked to behavior, and finishing with a deep unpacking of what a heritability number really means.

You'll come away able to describe how researchers narrow a trait down to a genetic marker and then to an actual gene, recognize several real examples of genes tied to social and emotional behavior, and correctly explain why a high heritability score does not mean a trait is fixed by genes.

Key ideas

• Phenotype-first gene hunting: early behavioral genetics started with an observable trait difference (often a severe disease) and searched DNA for a shared stretch that tracked with it, producing a genetic marker rather than a single gene.
• Genetic marker: a stretch of DNA containing many genes that is statistically associated with a trait, not yet the specific responsible gene.
• Reverse strategy: starting from a gene already studied in animals (its function known), then checking whether humans have similar variants and whether those variants map onto behavior.
• Gene-environment interaction: the finding that what a gene "does" depends on the environment the organism is in, so the effect of a gene can reverse or vanish depending on conditions.
• Heritability: a statistic describing how much of the variability in a trait across a population is explained by genetic variation, not how much genes determine the average level of the trait.
• Environment inflation bias: because scientists typically study a trait in one tightly controlled environment, standard experimental practice systematically overstates how heritable a trait appears to be.
• Chance as a third factor: alongside genes and environment, random molecular events (such as Brownian motion during cell division) introduce variability that is neither inherited nor environmental.

Walkthrough

Recap and the shift to molecular genetics (3:03)

The lecture opens with a quick review of twin, adoption, and prenatal-environment studies from the prior lecture, then pivots to how, starting around the 1980s, researchers began combining classical behavioral genetics with molecular biology to actually locate genes. It walks through a ladder of starting points, from a clear phenotype difference down to differences in protein size, each used as a foothold for tracing genetic differences.

From genetic markers to disease genes (11:09)

Using diseases like PKU, Huntington's, and cystic fibrosis as early targets, the lecture explains how comparing DNA across affected and unaffected relatives in large families produced genetic markers, broad chromosomal neighborhoods rather than single genes. It covers the bioethics problems this raised, such as whether to test people for adult-onset diseases before symptoms appear, and how sequencing technology later let researchers pin down markers to specific genes and mutations.

Modern tools and finding genes tied to behavior (16:15)

The lecture introduces gene microarrays (credited to Stanford's Pat Brown) and quantitative trait loci analysis as tools for studying many genes at once rather than one at a time. It then flips the approach: starting from genes already understood in animal studies, such as the vasopressin receptor gene linked to monogamy in voles, and checking whether the same gene variants exist in humans and predict similar behaviors, including relationship stability and reading facial expressions. Similar examples follow for BDNF (linked to amygdala-driven anxiety), dopamine receptor D4 (linked to novelty-seeking), and neuropeptide Y (linked to anxiety and metabolism). Each of these genetic links is described as real but explaining only a small fraction of the variability in the trait.

What heritability actually measures (39:34)

This section reframes heritability as a measure of variability, not average trait level, and walks through the counterintuitive plant IQ example: if a gene fully explains differences in a trait within one environment, but the same gene shows no consistent effect once you add a second environment, heritability drops sharply. It shows how studying a trait in only one controlled environment (the standard, "careful" way to run an experiment) mechanically inflates the apparent role of genetics by removing environmental variability from the picture.

Counterintuitive cases and real interaction data (55:51)

Using the number of fingers on a hand (0% heritable, since variation is almost entirely from accidents) and 1950s earring-wearing (100% heritable, since it tracked sex almost perfectly), the lecture shows heritability can defy intuition entirely. It then covers real gene-environment interaction studies: PKU heritability dropping to near zero once diet is controlled, Caspi and colleagues' New Zealand study linking a serotonin-related gene to depression only in people with childhood stress exposure, and a parallel finding for a MAO gene variant and antisocial behavior tied to childhood abuse.

Environment overwhelming apparent genetic effects (1:22:11)

The lecture closes on gender differences in math and verbal test scores across 40 countries, showing that the size and even direction of these differences track a society's gender-equality index far more closely than they track sex itself, with the gap disappearing or reversing in more equal countries such as Iceland. This is used to reinforce that knowing someone's environment is often far more predictive than knowing their genotype.

Before you watch

• Review the earlier lecture on classical behavioral genetics (twin and adoption studies) since this lecture assumes familiarity with monozygotic/dizygotic comparisons and epigenetics.
• Be comfortable with the basic idea of a gene having different versions (alleles) and a promoter region controlling gene expression.

Check your understanding

1. What is the difference between a genetic marker and an identified gene, and why did early behavioral genetics studies typically find markers first?
2. Why does heritability measure variability rather than the average level of a trait, and what does that imply about interpreting a high heritability number?
3. Explain the finger-count and earring-wearing examples in your own words. What do they show about the relationship between genetic determination and heritability?
4. How did the Caspi study show a gene-environment interaction for depression, and why does studying a trait in only one environment tend to inflate its apparent heritability?
5. What did the cross-country study on math and verbal test scores suggest about the role of environment versus genetics in gender-linked cognitive differences?