Modern WisdomWhy Is Behavioural Genetics A Hated Science? - Dr Stuart Ritchie
CHAPTERS
- 0:00 – 0:41
Behavioural genetics’ credibility problem: the rise-and-fall of candidate gene claims
The episode opens with a blunt example of how behavioural genetics earned skepticism: candidate gene studies that tied single genes to complex traits like depression or memory. Ritchie explains how a huge body of published work and career capital was later shown to be largely non-replicable.
- •Candidate gene era produced thousands of high-profile papers with large claimed effects
- •Most findings failed when tested in larger samples
- •Careers and funding were built on results that turned out to be mostly noise
- •Sets the stage for why distrust persists around genetics-and-behavior claims
- 0:41 – 2:31
Why behavioural genetics is distrusted: immutability fears and politics-by-association
Ritchie outlines the main misconception driving hostility: that genetic influence implies traits are fixed, unchangeable, and politically justificatory. He argues behavioural genetics describes variation in current populations, not a moral endorsement of social hierarchies or an argument against intervention.
- •Genetic influence is often misheard as ‘unchangeable’ or ‘predetermined’
- •People infer researchers’ politics from the mere mention of genetics
- •Behavioural genetics aims to explain variation, not justify inequality
- •Genetic effects can vary across contexts, times, and regimes
- 2:31 – 5:45
Genes and environments interact: Estonia’s communism-to-market transition example
Using a study from Estonia, Ritchie discusses how political and social environments can change the measured genetic contribution to educational attainment. The takeaway is that environments can suppress or enable the expression of genetic differences, undermining simplistic ‘genes are destiny’ narratives.
- •Polygenic scores used to compare cohorts born before vs after communism
- •Heritability of educational attainment appeared higher after communism ended
- •Interpretation: freer environments can allow ‘genetic potential’ to manifest more
- •Illustrates gene–environment dependence rather than pure genetic determinism
- 5:45 – 11:10
The ‘height vs IQ’ double standard and what heritability really means
Chris and Stuart explore why people accept genetic explanations for traits like height and eye color but resist them for intelligence or education. Ritchie emphasizes that heritability doesn’t imply immutability, using examples like myopia corrected by glasses and height affected by nutrition.
- •Public accepts genetics for non-politicized traits but rejects it for intelligence/personality
- •Heritable traits can still be heavily modifiable by environment (glasses, nutrition)
- •Complexity isn’t unique to ‘social’ traits—‘simple’ traits become complex under scrutiny
- •Resistance often reflects discomfort with measurement and quantification
- 11:10 – 15:41
Uneven starting lines and moral conclusions: Rawls, Paige Harden, and policy implications
The conversation shifts to the ‘uneven playing field’ problem: genetic variation makes equal opportunity difficult to conceptualize in a meritocratic culture. Ritchie argues there isn’t one inevitable political conclusion; acknowledging genetic differences can justify more support for those who start with disadvantages.
- •Genetic differences can feel ‘unfair’ within meritocratic narratives
- •Common intuition: parents with multiple children observe large innate differences
- •Rawls’ ‘veil of ignorance’ and Harden’s ‘genetic lottery’ framing
- •A progressive implication: more resources for struggling kids, not fatalism
- 15:41 – 19:34
IQ testing’s origins: from helping struggling children to later political misuse
Ritchie pushes back on the idea that IQ and early intelligence research were inherently reactionary. He describes early intentions—objective identification of needs and talent outside privilege—while acknowledging later eugenic interpretations and policy failures like unequal grammar vs secondary modern outcomes.
- •Early intelligence testing aimed to identify and support struggling students (Binet)
- •Historical example: Godfrey Thompson’s egalitarian resource view
- •Butler Education Act and the grammar school system’s intentions vs outcomes
- •Later association with eugenics shaped public suspicion, but isn’t inherent to the science
- 19:34 – 22:28
How heritable is intelligence—and why heritability rises with age
Ritchie explains the robust finding that the heritability of IQ tends to increase over the lifespan. He references twin evidence (e.g., Vietnam Era Twin Study) and clarifies what heritability means: the portion of variance associated with genetic differences in a given population and context.
- •Heritability estimates for IQ rise into adulthood (often ~60–80%)
- •Environmental influence appears stronger earlier in life, then wanes in relative terms
- •Twin studies provide long-running, replicable evidence for these patterns
- •Heritability is about variance in populations, not fixed outcomes for individuals
- 22:28 – 31:35
Replication crisis hits behavioural genetics: candidate genes to GWAS and polygenic scores
Ritchie describes how behavioural genetics was ‘smashed’ early by replication failures in candidate gene work, prompting a methodological shift. The field moved toward genome-wide association studies and polygenic scores, recognizing that complex traits are influenced by many variants of tiny effect.
- •Candidate gene effects were implausibly large and mostly unreplicable
- •GWAS examines hundreds of thousands of variants, improving reliability
- •Complex traits are highly polygenic: many tiny effects add up
- •Remaining scientific concerns: assortative mating, sample diversity, prediction portability
- 31:35 – 35:48
General intelligence vs ‘multiple intelligences’: what replicates and what’s branding
The discussion turns to intelligence structure: while people show specific strengths, there is a highly replicable general factor (g) explaining much shared variance across cognitive tests. Ritchie critiques Howard Gardner’s multiple intelligences as lacking empirical grounding despite major cultural influence.
- •IQ tests contain multiple subtests, but they positively correlate (g factor)
- •g explains roughly 40–50% of variance across diverse cognitive tasks
- •Gardner’s ‘multiple intelligences’ theory is portrayed as conceptually appealing but data-light
- •Training specific tasks (e.g., working memory games) shows limited general transfer
- 35:48 – 39:05
Is EQ real? Emotional intelligence, grit, and the jingle/jangle problem
Ritchie argues EQ often repackages existing constructs like personality and cognitive ability. He uses meta-analytic findings suggesting EQ predicts outcomes only until IQ and Big Five traits are included, and extends the critique to ‘grit’ as largely conscientiousness with a catchier name.
- •EQ correlates with job performance in isolation, but adds little beyond IQ + personality
- •Many ‘new’ constructs are re-descriptions of older ones (jingle/jangle fallacies)
- •Grit strongly overlaps with conscientiousness (very high correlation)
- •Pop-psych branding can drive adoption more than incremental scientific contribution
- 39:05 – 45:39
Dying psychological concepts: social priming’s collapse and mindset overreach
Ritchie recounts famous social priming results (elderly words causing slower walking; ‘thinking outside the box’) and how better-controlled replications failed. Growth mindset, by contrast, appears to have small effects, but earlier claims were vastly inflated and culturally over-implemented.
- •Social/behavioral priming produced implausible claims that often failed replication
- •Replication improved measurement (e.g., sensors vs stopwatches) and erased effects
- •Growth mindset likely yields small, context-dependent benefits, not life-changing transformations
- •Overclaiming drove widespread school adoption without commensurate evidence
- 45:39 – 51:52
Placebo and expectation effects: why controls and truth-seeking matter
Chris raises the ‘expectation effect’—beliefs shaping outcomes—and whether exaggeration might be pragmatically useful. Ritchie insists that science’s job is truth, and that weak controls and poor blinding let expectation effects masquerade as real mechanisms, contaminating findings across domains.
- •Expectation effects can create real-seeming changes without the claimed mechanism
- •Poor control groups (doing nothing vs intervention) inflate apparent effects
- •Better blinding and active controls are difficult but essential
- •Distinguish pragmatic persuasion in practice from scientific inference about causality
- 51:52 – 58:22
SSRIs, the serotonin story, and research bias: small effects, messy evidence
Ritchie addresses the high-profile challenge to the ‘chemical imbalance/serotonin’ explanation of depression. He argues mechanism uncertainty doesn’t automatically negate efficacy; however, publication bias, outcome switching, and inconsistent definitions of depression complicate interpretation of SSRI trial evidence.
- •Recent reviews question simple serotonin-deficit explanations for depression
- •SSRIs have noticeable side effects, implying biological activity even if mechanism differs
- •Trial literature shows severe publication bias and selective reporting/spin
- •Meta-analytic conclusions vary: critics call effects trivial; others argue small effects matter at scale
- 58:22 – 1:07:55
What is depression, anyway? Latent disease vs symptom-network models
Ritchie explores debates about whether ‘depression’ is a single underlying entity or a label for interacting symptoms (insomnia, low mood, irritability) that reinforce one another. He notes symptom-network approaches may better capture heterogeneity and could explain inconsistent treatment effects across studies.
- •Depression measurement varies widely (multiple inventories/scales)
- •Network view: symptoms cause and amplify each other, not necessarily one latent cause
- •Heterogeneity across people and phases of illness makes trials noisy
- •Better definitions and measurements may improve treatment research clarity
- 1:07:55 – 1:18:15
How to be effectively skeptical: open science, post-publication review, and cultural norms
In the closing stretch, Ritchie lays out how to question consensus without falling into conspiracism. He advocates higher evidentiary standards, open science practices (preregistration, shared data/materials), and stronger norms of critique—plus tools like PubPeer and independent expert commentaries.
- •Raise standards rather than selectively trusting ‘your side’
- •Use Science Media Center-type expert reactions to contextualize headline studies
- •Open science: preregistration, open data, open materials, open peer review
- •PubPeer and fraud detection (e.g., image manipulation) as post-publication safeguards
- •Science benefits from adversarial critique; excessive politeness can undermine rigor
- 1:18:15 – 1:18:51
Where to follow Ritchie: writing, Substack, and ‘Science Fictions’
Ritchie shares where to find his ongoing work and writing about research quality and replication. The conversation ends with pointers to his online presence and his book focused on scientific reliability problems.
- •Twitter handle: StuartJRitchie
- •Long-form writing on Substack
- •Book: ‘Science Fictions’ on replication and research reliability
