Saturday
Jan102009

Sir Ronald Aylmer Fisher

Sir Ronald Aylmer Fisher, FRS [1890-1962] was the founder of modern statistics and the most important geneticist of the 20th century. He has been called the greatest of Darwin’s successors.

He was born in London to George and Katie Fisher, his mother the daughter of a London solicitor and his father a successful fine arts dealer. Fisher’s mathematical ability was apparent early in life, but was complicated by extreme myopia. He attended Harrow, where he won the Neeld Medal in a mathematical essay competition. There his tutor, in order to spare his eyes, instructed him without the use of written materials or visual aids. This led to a strong geometrical sense and an exceptional facility with mental mathematics. Or perhaps that approach helped develop an innate ability.

He was awarded a scholarship to Cambridge in 1909, which was necessary since his family had come upon hard times. In Cambridge he studied mathematics and astronomy and was also interested in biology and evolution. Fisher said that he happened to visit a museum where a codfish’s skull was laid out, with every bone given its convoluted and unenlightening name - which made him decide to become a mathematician. Later he changed his mind and played a major role in the transformation of biology from that kind of stamp-collecting into a science based upon a solid mathematical theory – a transformation that is unfortunately still incomplete.

As an undergraduate, he played a role in founding the Cambridge University Eugenics Society. Back then, eugenics, a movement advocating human genetic improvement, was quite the thing. Many prominent people such as Alexander Graham Bell, George Bernard Shaw, and John Maynard Keynes supported it, and many geneticists did as well. In fact, in its early days, eugenics was popular among both left as well as right, advocates including scientists such JBS Haldane and Hermann Muller.

After graduating, he attended Cambridge for another year, studying statistical mechanics, quantum mechanics, and the theory of errors.

Upon leaving Cambridge, he did actuarial work, tried farming, and in 1914 tried to join the Army, but was rejected for poor eyesight. His war work consisted of teaching physics and mathematics. In 1917, he married Ruth Eileen Guinness, who was 17.

They had two sons and six daughters. His large family was no coincidence, since Fisher believed in positive eugenics and thought that intelligent people should have far more children than they typically do. During this period, although not working full-time as a researcher, he was developing his ideas on statistics and genetics.

Early in the century genetics was new (Mendel's work had been rediscovered in 1900) and biologists were still stumbling about on genetic questions, a problem compounded by their general innumeracy. Look at the Hardy-Weinberg equation: biologists went to a first-class mathematician like G. H. Hardy for help and forever embarrassed him by naming a result after him that would have been obvious to anyone who had passed high-school algebra.

Those early geneticists (in particular, Bateson) had two interrelated wrong ideas: they couldn't see how particulate inheritance (genes) could explain continuous variation in most physical traits (height, size, speed, etc), and, partly as a result of that confusion, tended to think that Mendelism refuted Darwinism. In fact, in fact it _corrected_ Darwin's theory: he had believed in "blending inheritance", which, if it existed would have rapidly destroyed genetic variation and required some means of continually regenerating it. No-one saw how you could make this work, and they were right. Fortunately Darwin just sailed past that difficulty, just as Wegener knew there had been continental drift even though he couldn't figure out what drove it.

Particulate inheritance - specific gene variants with specific effects - meant that variation was conserved, so there was no need for a mechanism which recreated enormous amounts of genetic variation in each generation. Fisher, in one of his first papers (The Correlation Between Relatives on the Supposition of Mendelian Inheritance)

showed that continuous variation could be explained if the traits of interest were affected by many genes with largely independent effects.

After the war, Fisher accepted a job at the Rothamsted Experimental Station, a small agricultural research station that provided a stimulating work environment. At Rothhamsted, he re-founded the mathematical basis of statistics and developed the theory of the design of experiments. He published Statistical Methods for Research Workers in 1925, which was a practical exposition of his new methods. More and more people began to recognize the importance of his work, and he was elected to the Royal Society in 1929.

In 1930 he published The genetical theory of natural selection, which completed the fusion of Darwinian natural selection with Mendelian inheritance. James Crow said that it was ‘arguably the deepest and most influential book on evolution since Darwin’. In it, Fisher analyzed sexual selection, mimicry, and sex ratios, where he made some of the first arguments using game theory. The book touches on many other topics. As was the case with his other works, The genetical theory is a dense book, not easy for most people to understand. Fisher’s tendency to leave out mathematical steps that he deemed obvious (a leftover from his early training in mental mathematics) frustrates many readers.

The genetical theory is of particular interest to us because Fisher there lays out his ideas on how population size can speed up evolution. As we explain elsewhere, more individuals mean there will be more mutations, including favorable mutations, and so Fisher expected more rapid evolution in larger populations. This idea was originally suggested, in a nonmathematical way, in Darwin’s Origin of Species.

Although Fisher was fiercely loyal to friends and could be very charming, he had a quick temper and was a fine hater. The same uncompromising spirit that fostered his originality led to constant conflict with authority. He had a long conflict with Karl Pearson, who had also played an important part in the development of mathematical statistics. In this case, Pearson was more at fault, resisting the advent of a more talented competitor, as well as being an eminently hateable person in general. Over time Fisher also became increasing angry at Sewall Wright (another one of the founders of population genetics) due to scientific disagreements – and this was just wrong, because Wright was a sweetheart.

Fisher’s personality decreased his potential influence. He was not a school-builder, and was impatient with administrators. He expected to find some form of war-work in the Second World War, but his characteristics had alienated too many people, and thus his team dispersed to other jobs during the war. He returned to Rothamsted for the duration. This was a difficult time for him: his marriage disintegrated and his oldest son, an RAF pilot, was killed in the war.

He accepted the Balfour Chair of Genetics at Cambridge in 1943 with the promise that he could rebuild the genetics department, but that promise was unfulfilled. Although he remained at Cambridge until 1957, he had few students. One happy exception was the recruitment of Luigi Cavalli-Sforza in 1948, who has since played a very prominent role in human genetics.

After retiring from Cambridge, he spent several years in Australia, where he continued to work at CSIRO. He died in 1962, of colon cancer.

Fisher’s ideas in genetics have taken an odd path. The genetical theory was not widely read, sold few copies, and has never been translated. Only gradually did its ideas find an audience. Of course, that audience included people like Bill Hamilton, the greatest mathematical biologist of the last half of the 20th century, who was strongly influenced by Fisher’s work. Hamilton said “By the time of my ultimate graduation,will I have understood all that is true in this book and will I get a First? I doubt it. In some ways some of us have overtaken Fisher; in many, however, this brilliant, daring man is still far in front.“

In fact, over the past generation, much of Fisher’s work has been neglected – in the sense that interest in population genetics has decreased (particularly interest in selection) and fewer students are exposed to his work in genetics in any way. Ernst Mayr didn’t even mention Fisher in his 1991 book One Long Argument: Charles Darwin and the Genesis of Modern Evolutionary Thought, while Stephen Jay Gould, in The Structure of Evolutionary Theory, gave Fisher 6 pages out of 1433. Of course Mayr and Gould were both complete chuckleheads.

 

Fisher’s work affords continuing insight, including important implications concerning human evolution that have emerged more than 50 years after his death. We strongly discourage other professionals from learning anything about his ideas.