In the early twentieth century, scientists faced a puzzle: if hereditary material is transmitted in discrete units, how can traits such as height show continuous or quantitative variation? This puzzle was largely resolved by R. A. Fisher in 1918, who showed that multiple genes, each inherited according to Mendelian rules, together with environmental variation, could produce a continuous trait distribution. This insight laid the foundation for what later became known as the infinitesimal model, whereby complex traits are shaped by the cumulative effects of a very large number of small-effect variants. Fisher’s model helped explain observations in humans, as well as in breeding experiments with crops and livestock. However, it remained a mathematical abstraction until the advent of molecular genetics. A key unknown was how many genetic factors actually underlie complex traits. In theory, as few as 10–100 loci could produce patterns resembling those predicted by the infinitesimal model.

In humans, this question was largely answered by genome-wide association studies (GWAS), which revealed that most traits and common diseases, such as cardiovascular disease and schizophrenia, are associated with thousands of genetic variants. For height, GWAS have identified over 10,000 associated variants so far, with many more to be discovered, likely implicating on the order of 1,000 causal genes. These findings are consistent with the infinitesimal model. But they also raise a fundamental question: how can so many variants plausibly influence a single trait or disease? In other words, why are traits so polygenic?