Home > Positioning > Persons > Fisher

R. A. Fisher (1890–1962)

Fisher was a statistician and geneticist who, more than any other single figure, built the mathematical foundations on which the Modern Synthesis rests. The Genetical Theory of Natural Selection (1930) demonstrated that Mendelian inheritance and Darwinian natural selection are not merely compatible but mutually reinforcing — that continuous variation in a population can arise from the combined effects of many genes, each of small effect, and that natural selection acting on this variation is a powerful and directional force. He was also a founder of modern statistics — analysis of variance, maximum likelihood, experimental design — and a committed eugenicist. The two sides of his career are not separable: his statistical innovations and his eugenic commitments grew from the same intellectual programme.

Life

Born 17 February 1890 in East Finchley, London. Severe myopia from childhood, uncorrectable at the time; his mathematics tutor at Harrow trained him to work problems without pen and paper, developing an exceptional capacity for geometric and spatial reasoning. BA in mathematics from Gonville and Caius College, Cambridge (1912), with a further year studying statistical mechanics and the theory of errors.

Worked at an investment house in London (1913–15), then as a statistician for the City of London (1915–19) while also teaching physics and mathematics. Unable to serve in the First World War due to his poor eyesight. Statistician at Rothamsted Experimental Station (1919–33), where he developed the analysis of variance, experimental design, and much of what became the modern statistical toolkit — working on agricultural field trials. Galton Professor of Eugenics at University College London (1933–43), succeeding Karl Pearson. Balfour Professor of Genetics at Cambridge (1943–57). After retirement, moved to Adelaide, Australia, at the invitation of E. A. Grubb, and worked at the CSIRO Division of Mathematical Statistics until his death on 29 July 1962.

Elected Fellow of the Royal Society (1929). Knighted (1952). Royal Medal (1938), Darwin Medal (1948), Copley Medal (1955). Founding fellow of the International Biometric Society and the Genetics Society.

Population genetics

Fisher’s central achievement was to demonstrate mathematically that the Mendelian theory of discrete inheritance produces exactly the continuous variation that biometricians observed in natural populations — and that natural selection acting on this variation is sufficient to drive adaptive evolution.

The fundamental theorem of natural selection. Published in The Genetical Theory of Natural Selection (1930): the rate of increase in the mean fitness of a population at any time is equal to its additive genetic variance in fitness at that time. The theorem gives natural selection a quantitative expression. Its interpretation has been debated since publication — Fisher’s own formulation was narrower than many subsequent presentations assumed, applying to the additive genetic component of fitness rather than to total fitness — but its significance as the first rigorous mathematical statement of selection’s power is not disputed.

Runaway sexual selection. Fisher developed the theoretical account of how a preference for a trait and the trait itself can become genetically correlated in a positive feedback loop. The peacock’s tail is the paradigm case: once a female preference for long tails exists, males with longer tails are selected and the preference-trait correlation amplifies across generations, producing elaborate ornaments far beyond what natural selection alone would sustain. Darwin had identified sexual selection; Fisher gave it its mathematical logic.

The Fisher-Wright tension. Fisher and Sewall Wright represented two emphases within population genetics. Fisher emphasised large populations, additive gene effects, and the directional power of natural selection. Wright emphasised small populations, gene interaction (epistasis), genetic drift, and the “shifting balance” theory in which drift moves small populations between adaptive peaks. The tension between Fisherian and Wrightian emphases structured population genetics through the mid-twentieth century and has never fully resolved.

Statistics

Fisher’s statistical work, developed primarily at Rothamsted from 1919 to 1933, laid the foundations of modern frequentist statistics:

Analysis of variance (ANOVA). A method for partitioning observed variation into components attributable to different sources — treatment effects, environmental variation, interactions. Developed for the analysis of agricultural field trials; became the dominant tool for experimental analysis across the sciences.

Maximum likelihood. A general method for estimating parameters by finding the values that make the observed data most probable. Fisher developed the theory, proved its asymptotic efficiency, and established it as the standard approach to statistical estimation.

Experimental design. Randomisation, replication, blocking, factorial design — the principles of how to structure experiments so that valid inferences can be drawn. Fisher’s The Design of Experiments (1935) codified these principles; the randomised controlled trial follows from them.

Significance testing. Fisher formalised the use of p-values as a measure of evidence against a null hypothesis. The p < 0.05 threshold, which became ubiquitous across the sciences, was Fisher’s suggestion — originally offered as a convenient convention, not as a rigid rule.

The statistical and genetic work are deeply connected. Fisher’s earliest statistical innovations (the method of maximum likelihood, the distribution of the correlation coefficient) emerged from questions about inheritance and variation. Rothamsted gave him both the data and the problems; genetics gave him the motivation.

Eugenics

Fisher was a committed eugenicist throughout his life. He served as chairman of the University of Cambridge Eugenics Society, was a founding member of the Eugenics Society (later the Galton Institute), and the second half of The Genetical Theory of Natural Selection is devoted to eugenic arguments — including the claim that the decline of civilisations is caused by the differential fertility of social classes, with the upper classes reproducing less than the lower.

Fisher’s eugenics was not incidental to his scientific work. His interest in genetics was motivated from the start by eugenic concerns — understanding the inheritance of human traits in order to improve the human stock. His appointment at UCL was to the Galton Chair of Eugenics, a post explicitly designed around this programme. He provided a statistical defence for the sterilisation policies of the Third Reich in a 1939 review.

The post-war reckoning with eugenics has reached Fisher’s legacy directly. In 2020, Gonville and Caius College removed Fisher’s name from a stained-glass window; the same year, the R. A. Fisher Committee was established to address Fisher’s eugenic legacy within the statistical community. UCL’s history of eugenics review has examined the institutional context in which Fisher and Pearson worked. How Fisher’s scientific contributions stand alongside his eugenic commitments is a live question — debated within statistical societies, addressed by Cambridge colleges, and worked through in the history of biology. The commitments are part of the record either way.

Where Fisher stops

Fisher’s programme is mathematical and populational. It works at the level of gene frequencies in large populations, with natural selection as the dominant force. What it does not address is the organism — the developmental, physiological, and ecological reality that lies between genotype and fitness. Fisher’s models treat the phenotype as a mathematical function of the genotype; the biology of how genes produce organisms is abstracted away. The tension between Fisher’s gene-frequency mathematics and the organism-centred biology of Mayr and the naturalists was a constitutive tension within the Modern Synthesis itself.

Key works

See also: Darwinism · Darwin · Mayr · Dawkins