South Australian Medical Heritage Society Inc

R. A. Fisher's Life and Death in Australia, 1959-1962

John Ludbrook is a Professorial Fellow in the University of Melbourne , and Professor Emeritus in the University of Adelaide . Mailing address: 563 Canning Street , Carlton North, Victoria 3054, Australia ( E-mail: ludbrook@ bigpond.net.au). I am most grateful to Henry Bennett, Oliver Mayo, and Donald Simpson for their reminiscences about Fisher's stay in Adelaide . My daughters, Alice and Geraldine, photographed the memorials to Fisher in St Peter's Cathedral, Adelaide . I am also grateful to the Editor and four reviewers for their useful and constructive comments and suggestions.

In retirement, Fisher went to live in Adelaide , South Australia , at the invitation of the statistician E. A. Cornish and the geneticist J. H. Bennett. He died in Adelaide , following an operation for colon cancer, on July 29, 1962 . During his life, Fisher engaged in vigorous controversy with Karl Pearson, Jerzy Neyman, and W. S. Gosset , to name but a few. After Fisher's death, his family donated his book copyrights and other intellectual and personal material to the University of Adelaide . This has resulted in the republication of his major books and scientific correspondence, and in a Web site from which can be downloaded most of his published papers.

KEY WORDS: Adelaide ; J. Henry Bennett; E. A. Cornish; W. S. Gosset; Jerzy Neyman; Karl Pearson.

1. The false story of Fisher's death

This essay was prompted by the mis-statement of David Salsburg (2001, p. 180) that: "RA Fisher had died of a heart attack on the boat returning him to Australia."  I am told that Salsburg corrected this error in later reprintings.

Salsburg made other errors of fact in his book. For instance, he set the episode of his title ("The Lady Tasting Tea") at a genteel dons' tea party at Cambridge University (Salsburg 2001, p. 1). Joan Fisher Box gives the true story. It took place, of course, in the tea room of the Rothamsted Experimental Station in the early 1920s (Box 1978, p. 134). A Dr. Muriel Bristol claimed she could tell whether the tea or the milk was added first in a cup of tea. A crude experiment was done to test her claim. Fisher obviously remembered this episode when he described the "thought experiment" that led to what came to be known as Fisher's exact test. .(Fisher 1935).

2. The true story of Fisher's death

The fact is that Fisher died in the Queen Elizabeth Hospital, Adelaide , South Australia , at 7.50 p.m. on July 29, 1962, at the age of 72. I know the details because I have perused his case notes, and I have the personal testimony of a (then junior) surgeon who was in attendance on him. His death followed an operation by my predecessor in the Chair of Surgery in the University of Adelaide , the late Professor R. P. Jepson. The operation was for a three-centimeter-diameter cancer of the sigmoid colon, invading muscle but without lymph node involvement. Post-operatively, he developed an anastomotic leak, a pericolic abscess, septicemia and, finally, repeated pulmonary emboli. Though the abscess was drained, appropriate antibiotics were administered, and the pulmonary emboli were recognized and treated with the anticoagulant heparin, he suffered a final, massive, pulmonary embolus. Another inaccuracy about Fisher's illness was perpetrated by Gehan and Lemak (1994), who said that Fisher underwent his sigmoid colectomy "under local anaesthesia." Not true. The late Dr. Maurice Sando gave a general anesthetic.

There is a memorial brass plaque to Fisher on the floor of the right-hand aisle of St Peter's Anglican Cathedral in Adelaide (Fisher was a lifelong "scientific" Christian). The plaque overlies his ashes. It is inscribed "Ronald Aylmer Fisher Kt. ScD. FRS . 1890-1962." And on the end of the adjacent pew, in which Fisher is said to have sat on most Sundays, an inscription is carved. It consists of the coat of arms of Gonville and Caius College, Cambridge and the words "In Memory of Sir Ronald Fisher FRS 1890-1962."

 

3. Why was Fisher in Adelaide ?

In his retirement from the Arthur Balfour Chair of Genetics in the University of Cambridge, he was invited to Adelaide by Dr. E. A. Cornish, Head of the Division of Mathematical Statistics in the Commonwealth Scientific and Industrial Research Orga­ nization (CSIRO) and Professor of Statistics in the University of Adelaide, supported by J. Henry Bennett, Professor of Genetics in the University of Adelaide. Fisher accepted this invitation in 1959, when his Presidency of Gonville and Caius College Cambridge expired. Fisher had visited Adelaide in 1953, as a guest of EA Cornish and CSIRO. Cornish had studied under Fisher in the late 1930s, and they have two joint statistical publications (Cornish and Fisher 1937; Fisher and Cornish 1960). From 1954 to 1956, Bennett was a Fellow of Gonville and Caius College . Cambridge , and studied under Fisher. In 1956 he was appointed Professor of Genetics in the University of Adelaide . Fisher accepted the invitation, for several reasons it seems. An important one was that he had liked the Adelaideans. He arrived in Adelaide in 1959 for a six-month stint as a Senior Research Fellow in CSIRO, but stayed there (with intermittent excursions to other Australian cities, New Zealand , Europe , the USA , Japan , France . Italy , India and the UK , where he was feted and received numerous honours and honorary degrees ( Box 1978 ), until his death.

Fisher was a person who sometimes seemed to go out of his way to make enemies. One was Karl Pearson, with whom he had a major dustup over his (Fisher's) demonstration that Pearson had got his degrees of freedom wrong in evaluating his chi- squared statistic (Fisher 1922). Later, they squared off publicly over scientific inference (see Inman 1994). The culmination of this feud was Fisher's memorial paper after the death of Pear­son, in which he suggested that Pearson had fudged some of his data (Fisher 1937). But Fisher had said the same about Gre­ gor Mendel—or his gardener (Fisher 1936)! Another feud was with Jerzy Neyman, whom he accused repeatedly of being out of touch with the real world of experimentation (Fisher 1960; Box 1978, p. 262 ff; Neyman 1961). Fisher also strongly criticized Neyman's confidence limits and argued strongly (but wrongly) in favour of his own fiducial limits (Fisher 1960; Neyman 1961; Zabel] 2001). He even argued with his long-time friend W. S. Gosset ("Student"). Though the foregoing suggests that he was a cantankerous man, Henry Bennett (personal communication) swears that his sometimes aggressive manner was only a reflec­tion of his pursuit of the truth. But other colleagues describe him as having "a contentious spirit" and "a quick temper" (Yates and Mather 1963).

Fisher's sojourn in Adelaide was remarkable for the friends that he made. I have spoken recently with Henry Bennett and Oliver Mayo (with whose family Fisher stayed for some time and who went on to become a distinguished geneticist), who re­ member him with enormous affection. I also spoke with neuro­ surgeon Donald Simpson. It was his son Matthew and daughter Kate, together with a gaggle of other children, who followed Fisher as if he were a Pied Piper on a walk through a suburb of Adelaide . This was only one of many walks through the suburbs of Adelaide accompanied by adoring children. In short, while he was in Adelaide he made only friends, never enemies.

In 1961, Fisher moved his residence to St Mark's College of the University of Adelaide , where he was President of the Senior Combination Room at the time of his death. Fisher was accompanied to Adelaide by his daughter, Joan, until she married the statistician George E. P. Box and went to live in Madison , WI .

4. Fisher's legacy to Adelaide

Fisher died intestate, but his family donated his intellectual property to the University of Adelaide . This has resulted in Henry Bennett having edited and published so much of Fisher's work (Bennett 1965, 1971-1974, 1983, 1990; Fisher 1990, 1999). Even better, Henry Bennett and the Barr Smith Library of the University of Adelaide have two Web sites, from which one can download most of Fisher's published papers and books, scientific correspondence, and a remarkable amount of other ma­ terial. The Web sites are: www.library.adelaide.edu.au/digitised/ fisher/index.html and www.library.adelaide.edu.aukial/special/ fisher.html.

5. Fisher's legacy to statistics

Thus far I have stuck to facts. But it is my very strong opinion that Fisher should be regarded as the father of modern biostatistics. Not because he was, or wished to be, a theoretical statistician of the ilk of Jerzy Neyman and others of the West Coast schools in the United States, though he had great mathematical and conceptual skills, especially in multidimensional geometry. It was Fisher's good fortune to start his statistical career at the Rothamsted Experimental Station. His initial commission was to analyze past experiments in agriculture, but he soon graduated to designing new experiments. This led to his invention, validation and application of a variety of statistical techniques. These included the conversion of Student's t distribution into a test of significance; analysis of variance; maximum likelihood estimation; the concepts of sufficiency, consistency and efficiency; and, most importantly, his recognition of the importance of ran­ domization in experimental design and tests that result logically from that process: randomization (permutation) tests on contin­ uous variables (Kempthorne 1955). Then there was Fisher's now famous "exact" test on 2 x 2 tables of frequency, alluded to in the title of Salsburg's (2001) book, and described by Fisher in The Design of Experiments (Fisher 1935, p. 11 ff.). These tributes, and many others, were written by professional statisticians from both sides of the Atlantic (Fienberg and Hinkley 1960; Yates and Mather 1963; Savage 1976; Kruskal 1980; Zabell 2001). Indeed, Zabell says "Ronald Aylmer Fisher, arguably the greatest statistician of this (or any other century)...."

I, and many of my fellow biomedical investigators, are deeply indebted to Fisher's advice on how to design experiments and how best to analyze the results. I would be branded a liar if I were to state that we have always followed Fisher's precepts. But some of us, at least, have tried.

REFERENCES

Bennett, J. H. (ed.) (1965), Experiments in Plant Hybridisation: Mendel's Original Paper in English Translation with Commentary and Assessment by R. A. Fisher, Edinburgh, Oliver and Boyd.

___________ (1971-1974), Collected Papers of R. A. Fisher (vols. 1-5), Adelaide, University of Adelaide .

__________ (1983), Natural Selection, Heredity, and Eugenics: Including Selected Correspondence of R. A. Fisher with Leonard Darwin and Others, Oxford : Clarendon Press.

__________ (1990), Statistical Inference and Analysis: Selected Correspondence of R. A. Fisher, Oxford . Clarendon Press.

Box, J. F. (1978), RA Fisher: The Life of a Scientist, New York , Wiley. Cornish E. A.. and Fisher, R. A. (1937). "Moments and Cumulants in the Specification of Distributions," Revue de l'hinsitut International de Statistique. 5, 307-320.

Fienberg, S. B.,and Hinkley, D. V. (1960), RA Fisher: An Appreciation. Lecture Notes in Statistics I. , New York . Springer-Verlag.

Fisher, R. A. (1922), "On the Interpretation of X 2 From Contingency Tables, and the Calculation of P:Journal of the Royal Statistical Society. 85,87-94.

_________ (1935). The Design of Experiments. Edinburgh . Oliver and Boyd.

__________ (1936), "Has Mendel's Work been Rediscovered?" Annals of Science. 1,115-137.

__________ (1937), "Professor Karl Pearson and the Method of Moments," Annals of Eugenics, 7,308-318.

_________ (1960), "Scientific Thought and the Refinement of Human Reasoning." Journal of the Operations Research Society of Japan , 3,1-10.

_________ (J .H. Bennett, ed.) (1990), Statistical Methods, Experimental Design, and Scientific Inference, Oxford : Oxford University Press.

___ (1999), The Genetical Theory of Natural Selection: A Complete Variorum Edition. Oxford : Oxford University Press.

Fisher. R. A., and Cornish E. A. (1960). "The Percentile Points of Distributions Having Known Cumulants," Technometrics. 2.209-225.

Gehan. E. A., and Lemak, N. A. (1994), Statistics in Medical Research: Developments in Clinical Trials, New York : Plenum Medical Book Co.

Inman, H. F. (1994), "Karl Pearson and RA Fisher on Statistical Tests: A 1935 Exchange from Nature," The American Statistician, 48, 2-11.

Kempthorne, O. (1955), "The Randomization Theory of Experimental Inference." Journal of the American Statistical Association, 50,946-967.

Kruskal. W. (1980) "The Significance of Fisher: A Review of RA Fisher: The Life of a Scientist," Journal of the American Statistical Association, 75,1019— 1030

Neyman. J. (1961), "Silver Jubilee of my Dispute with Fisher," Journal of the Operations Research Society of Japan . 3.145-154.

Salsburg. D. (2001), The Lady Tasting Tea: How Statistics Revolutionized Science in the Twentieth Century. New York : Henry Holt and Company. Savage, L. J. (1976). "On Rereading RA Fisher:' The Annals of Statistics, 32, 441-500.

Yates. F., and Mather, K. (1963), "Ronald Aylmer Fisher," Biographical Memoirs of Fellows of the Royal Society of London . 9.91-120.

Zabell. S. L. (2001). "Ronald Aylmer Fisher:' in Statisticians of the Centuries. eds. C. C. Heyde. and E. Seneta. New York : Springer. pp. 389-397.

 

Reproduced with permission from THE AMERICAN STATISTICIAN, Volume 59, Number 2, May 2005, © 2005 by the American Statistical Association. All rights reserved.164-165

-o0o-