The popular version of genetics goes like this: genes exist, genes determine traits, traits determine the person. Real population genetics is considerably more complicated — and considerably more interesting.
In 1990, when the Human Genome Project was launched, many researchers anticipated that decoding the full sequence of human DNA would unlock the answers to most of biology's central questions. There would be a 'gene for intelligence', a 'gene for depression', a 'gene for aggression'. Diseases would be explained by a small number of specific mutations. Character and ability would turn out to be inscribed in the genetic code, readable if only the code could be read.
More than thirty years have passed. The genome has been sequenced. The genomes of millions of living and long-dead people have been analysed. The picture that has emerged differs radically from the expectations of 1990 — not because genes are unimportant, but because the relationship between genes and traits turned out to be fundamentally different: more complex, more non-linear, more dependent on context than anyone had anticipated.
The vast majority of interesting human traits — height, body weight, cognitive ability, temperament, and the risk of most common diseases — are polygenic. This term simply means that they are influenced not by one or a few genes, but by thousands of genetic variants, each contributing a very small effect. To appreciate what this means in practice, consider height.
Height is one of the most thoroughly studied polygenic traits. Large genetic studies involving hundreds of thousands of participants have identified more than ten thousand genetic variants associated with height. Together, these variants explain roughly 40 percent of the variation in height we observe between people. The remaining 60 percent is explained by nutrition, illness in childhood, stress and other environmental factors, as well as interactions between genetic and environmental influences that have not yet been fully characterised. Height is not simply 'tall parents produce tall children', though parental height does matter. It is thousands of genetic variants, each contributing a tiny amount, in constant interaction with a person's environment and history.
For cognitive ability, the picture is more complex still. The genetic score for cognitive performance — calculated by summing up the effects of thousands of associated variants — explains roughly 10 to 15 percent of the variation in cognitive test scores in a population. The remaining 85 to 90 percent is everything else. This does not mean that genes are irrelevant to intelligence. It means that a 'gene for intelligence' is as meaningless a concept as a 'gene for height'. What exists is a diffuse cloud of thousands of variants, each with a tiny effect, interacting with each other, with education, with nutrition, with experience.
Even if we knew every genetic variant a person carries, that would still be insufficient to predict their characteristics accurately. The reason is epigenetics — a layer of biological regulation that controls which genes are active and to what degree, without changing the underlying DNA sequence itself.
Epigenetic regulation works through chemical tags attached to the DNA or to the proteins around which DNA is wrapped. These tags can switch genes on or off, or modulate how actively they are expressed. Crucially, these tags are sensitive to the environment: to what a person eats, to the stress they experience, to illnesses, to exposure to various substances.
A vivid illustration: identical twins — who have exactly the same DNA sequence — develop noticeably different patterns of gene activity by middle age. Their identical genomes are being read differently, because their different life experiences have left different epigenetic marks. 'Same genes' does not mean 'same person'. Furthermore, some epigenetic marks can be passed to the next generation — a phenomenon called transgenerational epigenetics. This adds yet another layer of complexity to the already complex relationship between genes and traits.
One of the most important lessons from ancient DNA research concerns the mechanisms of evolutionary change. For a long time, a simplified version dominated: traits change because natural selection operates — individuals with more advantageous variants survive and reproduce more successfully, so those variants gradually become more common.
This is true as a principle. But the scale and prevalence of this process turned out to be considerably more modest than expected. Studies of how genetic variant frequencies have changed in European genomes over the past ten thousand years show that the majority of changes are explained not by natural selection, but by gene flow — the movement of people. When people with different genetic profiles move into a region and have children with the local population, the frequency of variants changes simply because new people arrived, not because any selection pressures favoured those variants.
To make this concrete: when Anatolian farmers arrived in Europe roughly eight thousand years ago, they carried higher frequencies of the genetic variant associated with lactase persistence — the ability to digest milk sugar as an adult. The frequency of this variant increased in Europe. Part of this increase reflects genuine selection: farming societies that kept cattle gained a real nutritional advantage from being able to drink fresh milk, and there is good evidence that selection for this variant intensified particularly between five and eight thousand years ago. But part of the early change in frequency was simply demographic — new people with different variant frequencies arrived and had children with local hunter-gatherers.
Research by Graham Coop and colleagues, published in 2009 and 2012, and a large body of subsequent work, demonstrated that many of the apparent 'signals of strong natural selection' identified in human genomes by earlier analytical methods turned out, under more careful scrutiny, to be artefacts of population mixing rather than evidence of genuine selection. The two patterns can look superficially similar in the data, and distinguishing them requires explicit modelling of migration history. This does not mean selection never acts on human traits — it clearly does. Lactase persistence is a well-documented example. Adaptations to high altitude in Tibetan populations are another. But these are relatively few examples against a very large dataset.
Popular claims to the effect that 'such-and-such a nation has genes for aggression, or laziness, or industriousness' rest on two compounded misunderstandings. First, there are no sharp genetic boundaries between nations — only gradual transitions, as the first article in this series explains. Second, even if such boundaries existed, character traits are polygenic, environmentally sensitive and epigenetically regulated. There is no 'gene for industriousness'. What exists is a diffuse network of thousands of variants influencing neurotransmitter systems, which interact with culture, upbringing and economic conditions in ways that cannot be reduced to a genetic category.
A large study published in Nature Human Behaviour in 2022 analysed data from four million people across 33 countries, examining the genetic associations with persistence — used as a measurable proxy for what is sometimes loosely called 'work ethic'. The genetic component explained approximately 11 percent of the variation. Cultural and social context explained incomparably more. Again: this does not mean genes are irrelevant to how people behave. It means their role is to set a range of possibilities, not to determine an outcome.
None of this means genetics is unhelpful when planning a family. Quite the opposite. There are areas where genetic knowledge is critically important and acts reliably. Monogenic diseases — conditions caused by a single gene, such as cystic fibrosis, spinal muscular atrophy, Huntington's disease, and phenylketonuria — are ones where one gene (or one pair of gene variants) determines disease risk in a very direct way. Carrier screening before conception allows prospective parents to understand this risk and make informed decisions. Chromosomal conditions such as Down syndrome are well detected through preimplantation genetic testing. Inherited cancer risk variants such as BRCA1/2 are another area where genetic testing changes clinical decisions in ways that are well established.
These are specific, medically validated applications of genetics — ones that work reliably and have clear clinical value. They are very different from ancestry percentages or speculations about ethnic character. Understanding the difference allows genetic information to be used where it is genuinely powerful, rather than over-interpreted where it is not.
'Genes explain everything' is a popular notion, not a scientific one. Genes matter enormously. But most traits people would like to predict from DNA — intelligence, character, temperament, 'national tendencies' — are determined by thousands of genetic variants interacting with environment, upbringing and chance in ways that do not reduce to a simple genetic prediction. Population-level changes in human genomes over thousands of years are explained primarily by gene flow — migration and mixing — rather than by pervasive strong selection. Genetics as a tool for reproductive planning is powerful where it is precise: in identifying specific recessive disease variants and chromosomal abnormalities. Where it is imprecise — in predicting complex polygenic traits — it offers ranges of probability, not guarantees and not sentences.
Polygenic trait — a characteristic influenced by the combined effect of a large number of genetic variants, each contributing a small amount individually.
Genome-wide association study (GWAS) — a type of research that scans the genomes of many thousands of people to identify genetic variants associated with particular traits or conditions.
Polygenic score — a numerical estimate of a person's genetic predisposition to a particular trait, calculated by summing the effects of thousands of individual genetic variants.
Gene flow — the movement of genetic material between populations through migration and interbreeding; one of the primary mechanisms by which the genetic composition of populations changes over time.
Epigenetics — the study of changes in gene activity that do not involve changes to the DNA sequence itself. Epigenetic marks can be influenced by environment and experience, and some can be transmitted to subsequent generations.
Population stratification — a potential source of error in genetic research, arising when groups being compared differ in their ancestral background. Can create apparent signals of selection that are actually the result of population mixing.
Thousands of people are already building families on their own terms.
Browse profiles