Hogben vs. R. A. Fisher

Tue Apr 26 10:52:06 EDT 2011

Below is a section of my book, The Probability Structure of Scientific Information, on the assault of my intellectual hero, Lancelot Hogben, on the statistics developed by the British biometric school, more specifically those of R.A. Fisher.  I may not agree with everything Hogen writes but he does clarify the intellectual and philosophical bases of modern inferential statistics.  He also makes very clear that all this scientometric stuff we do stems from eugenics, and, when we evaluate scientists, institutions, nations, etc., we stand upon the road that leads to Auschwitz.

As a native of Wisconsin and alumnus of University of Wisconsin at Madison, I found Hogben's observations on the UW and small town Wisconsin life very insightful.  It should be noted that Hogben's son, Adrian, received his medical degree from the UW.  It should also be noted that R. A. Fisher's son-in-law, George Box, developed the UW's statistics program and that, when I was an undergraduate there, R. A. Fisher himself often visited Madison to see his daughter Joan Fisher Box, whose book on her father I often cite.    

Stephen J. Bensman

LSU Libraries

Louisiana State University

Baton Rouge, LA   70803

USA

notsjb at lsu.edu

The above concepts were the focus of the ferocious critique of the inferential system developed by the British biometricians in the book Statistical Theory by Hogben (1957).  In his autobiography Hogben (1998) stated, "In Statistical Theory, 1957, I repudiated the still almost universally accepted assumptions of Fisherian statistics" (p.124), and in the book itself he described the concept of "a random sample of some infinite hypothetical population of possible values" as "the kingpin of the theory of statistical inference expounded by R. A. Fisher" (p. 98).  Hogben's case against the biometricians was political, scientific, and philosophical.  Politically he did not like their eugenics.  Part of this was personal.  As the son of an impoverished family, he had attended Cambridge on a scholarship system, which Major Leonard Darwin, son of Charles Darwin and Fisher's mentor, together with other eugenicists denounced as dysgenic and a waste of public resources, because students from poor families were genetically inferior (Sarkar, 1996).  Hogben was not only opposed to discrimination on the basis of class but also on the basis of race.  He was compelled to cut short his stay as professor of zoology in Cape Town due to his openly anti-apartheid attitudes and actions (Wells, 1978, pp. 194-198).  Hogben (1940) once described his tenure at Cape Town as a "four years' sojourn among the chromatocracy of South Africa" (p. 47), and in his memoirs Hogben (1998) stated, "By the beginning of 1929, the prospect was not at all healthy for university staff who openly opposed the Government's racial policy" (p. 114).  He was also a steadfast opponent of Nazism, helping Jewish scientists in Germany in the late 1930s, and in 1940 he was caught by the German invasion of Norway, when he went to lecture at the University of Oslo on genetic credentials of Nazi race theory.  Since the Blitzkrieg had cut him off from England, returning home entailed one-year odyssey via Sweden, the Soviet Union, Japan, and the United States, where he accepted an offer of a one-semester visiting professorship to teach mathematical genetics at the University of Wisconsin in Madison and gained a "glimpse of one-horse-town life" (Hogben, 1998, p. 177).  Here it should be pointed out that Hogben (1998) was not only anti-Nazi but anti-Communist, finding "Marxist dogma dished out as dialectical materialism deeply distasteful and wholly incompatible with my own criteria for intellectual integrity" (p. 129).  To him, it was just another "brand of religious fanaticism," and he dismissed the Communist scientific bible-Anti-Dühring by Friedrich Engels-as a "hodgepodge of Teutonic mysticism, phrenology, and mid-nineteenth century naturalism" (p. 128).  

            In his book Statistical Theory Hogben (1957) revealed that his anti-eugenicist stance was an important component of his hostile attitude toward the British biometric school.  Here he described Galton as "the father of the political cult variously named eugenics or Rassenhygiene" (p.106) and his views as "Galton's racialist creed" (p. 325).  Considering Karl Pearson as "Galton's disciple" (p.106), he declared that "K. Pearson's flair for ancestor worship had ample scope for self-expression in his partnership with the founder of the Eugenic cult" (p. 176).  However, Hogben reserved his greatest contempt for R. A. Fisher.  In a passage discussing the downfall of the Nazi Party in his autobiography excised from the published version by the editors but quoted by Tabery (2008), Hogben has this to say about Fisher:  

   ...After the war, the Nuremberg justices of the peace had Rosenberg 

   hanged. If I believed in hanging people for their opinions, the only 

   extenuating circumstances I might enter with a clear conscience as a 

   plan for mercy on behalf of the late Sir R. A. Fisher would be that he 

   did not occupy a government post with responsibility for 

   implementing his convictions.   pp. 753-754.

When Hogben applied for the position of professor of social biology at the London School of Economics to escape apartheid South Africa, one of his competitors for the post, according to Fisher's daughter, J. F. Box (1978, p. 202), was Fisher.  Hogben's candidacy was promoted by the Marxist Harold Laski, and in his memoirs Hogben (1998) surmised that Laski's main reason was the following:

               ...Laski's main concern in inveigling me into taking the chair of Social

               Biology was that the brass hats of the Eugenics Society were already 

               congratulating themselves on the prospect of one of their co-religionists

               getting the job.  At the time, the Society was truculently anti-socialist and

               a stronghold of racial prejudice.  Unlike so many other leftists in the 

               academic world of that time, Laski did not underrate the sinister significance

               of racist doctrines upholstered with bogus biology then gaining ground

               in Germany.  p. 121.

The above surmise was not far off the mark, for the correspondence of R. A. Fisher (M. Keynes, 2001-2002, 44 (March/June 2002), p. 1; Bennett, 1983, pp. 112-113) reveals that he was being encouraged in his candidacy by the Leonard Darwin, President of the Eugenics Society and Hogben's bête noire. 

            In his autobiography Hogben (1998) provided the following evaluation of human genetics as a science when he assumed the post of Professor of Social Biology at the London School of Economics:

                  At that time human genetics was a morass of surmise and superstition.

               It had as yet no sufficient theoretical foundation for firm conclusions about

               the results of matings necessarily beyond the range of experimental control.

               In short, no advance could materialise without further mathematical exploration

               of the postulates of experimentally established principles....  The rationalisation

               of race prejudice by appeal to biological principles was then plausible only

               because human genetics was so immature.... p. 122.

Hogben crossed scientific swords with Fisher on the issue of nature vs. nurture shortly taking over the social biology professorship.  Tabery (2008) has analyzed the origins and development of this controversy, summarizing it in the following manner:

   ...R. A. Fisher, one of the founders of population genetics and the

   creator of the statistical analysis of variance, introduced the biometric

   concept as he attempted to resolve one of the main problems in the biometric

   tradition of biology - partitioning the relative contributions of nature 

   and nurture responsible for variation in a population. Lancelot Hogben, 

   an experimental embryologist and also a statistician, introduced the 

   developmental concept as he attempted to resolve one of the main 

   problems in the developmental tradition of biology - determining the 

   role that developmental relationships between genotype and environment 

   played in the generation of variation....  p. 717.

According to Tabery, whereas the eugenicists emphasized hereditary variation and the anti-eugenicists emphasized environmental variation, the crucial variability for Hogben was a third type that resulted from the combination of a particular genetic constitution with a particular environment.  Tabery considered this controversy as a precursor of the one that erupted later in the century over the relationship of IQ to race.  The origins of the controversy can be traced back to the landmark paper by Fisher (1918) entitled "The Correlation between Relatives on the Supposition of Mendelian Inheritance," in which he reconciled biometry and Mendelism.  At the end of this paper Fisher acknowledged his "indebtedness to Major Leonard Darwin, at whose suggestion this inquiry was first undertaken, and to whose kindness and advice it owes its completion" (p. 433).  Fisher (1918) stated his main conclusion thus:

   By means of the fraternal correlation it is possible to ascertain the

dominance ratio and so distinguish dominance from all non-genetic causes, 

such as environment,...[and] to calculate the numerical influence not only of 

dominance, but of the total genetic and non-genetic causes of variability.  An

examination of the best available figures for human measurements shows

that there is little or no indication of non-genetic causes....  pp. 432-433.

Hogben (1933a) subjected Fisher's paper to critical analysis in an article entitled "The Limits of Applicability of Correlation Technique in Human Genetics."  This article was incorporated by Hogben (1933b, pp. 91-121) as a chapter entitled "The Interdependence of Nature and Nurture" in his book entitled Nature and Nurture, which comprised the William Withering Memorial Lectures he delivered at the University of Birmingham Faculty of Medicine.  In the article Hogben (1933a) characterized Fisher's paper as an attempt at "a synthesis between the particulate theory of inheritance and the problem of nature and nurture as it had been formulated by Galton and his successors" (p. 379).  He summarized the conclusions of his critique in the following three points:

      (1) The technique of correlation can be used to draw attention to the

existence of genetic differences or of differences due to environment

provided the selection of data is appropriate to the kind of differences

we wish to detect.

   (2) The belief that a comparison between observed correlations of

relatives and correlations based upon purely genetical assumptions

provides us with a measure of the influence of nurture is not justified,

because of the close relationship between the distribution of gene 

differences and differences due to environment in populations of viviparous

animals which live in families, especially when, as with human populations,

the environment of different families may differ greatly.

   (3) A balance sheet of nature and nurture, if it has any significance

in the light of modern experimental concepts, does not entitle us to set

limits to changes which might be produced by regulating the social or

physical environment of a human population.

In his lectures Hogben (1933b) capped these conclusions by pointing to "the danger of concealing assumptions which have no factual basis behind an impressive façade of flawless algebra" (p. 121). 

            However, perhaps Hogben's biggest disagreements with the British biometricians were not political and scientific but philosophical, and he set forth the bases of these disagreements in his book, The Nature of Living Matter.  In the first place, as described above, here Hogben (1930) divided reality into the public world of scientific discourse, where ideas had to be communicated and validated, and the private world of unverifiable opinion.  According to him (p. 127), the failure to recognize that biology no less than physics was an ethically neutral science was a result of the religious hostility evoked by Darwin's theory of evolution in the mid-19th century, which forced biologists to focus not on the logical structure of the new theory but its apparent implications for social philosophy.  The purpose of experimental biology of his generation was to free biology of this and to reduce the problems of organic evolution to an exact science.  From this perspective, Hogben criticized the eugenicists for injecting their private opinions into the public discourse of science, hurting the development of biology as an ethically neutral science.  Thus, he wrote that eugenic propaganda in England had been dominated by an explicit social bias making it unpalatable to a section of the community assuming the role of a governing class, declaring, "The greatest obstacle to the spread of a sane eugenic point of view is the eugenists themselves" (p. 214).  According to Hogben, the biologist should make clear when he is speaking as a professional biologist and when he is speaking a private citizen, and he drove the point home in the following manner:

               ...As a private citizen the biologist is entitled to his own opinions 

               concerning the merits of sterilizing the unfit, just as he is entitled

               to his own opinions on the Single tax or the advantages of capital 

               punishment.  Such opinions usually belong to his private world.  In

               his public capacity, as a biologist, he is primarily concerned with

               sterilizing the instruments of research before undertaking surgical 

   operations upon the body politic.  p. 215. 

On the whole,  Hogben comes across as a voice of sanity and reason in a world rapidly going insane.

            In the second place, as has been seen above, Karl Pearson in his The Grammar of Science consciously based science and British biometrics on the idealistic precepts of Bishop Berkeley and Immanuel Kant.  For Pearson, science dealt with sense impressions, and its field was essentially the contents of the mind.  From this perspective, the statistician deals not with material reality, which he cannot know, but with mental constructs of reality.  In his book Statistical Theory Hogben (1957) described such type of thinking as "Platonism," and it has been seen that he traced its origins in statistics back Quetelet,   He considered both Galton and Pearson disciples of Quetelet, calling the latter "the architect of a system of values and of an epistemology later inflicted by Karl Pearson on a generation still surviving" (p. 109).  Hogben had run into this type of thinking at the 1929 South Africa meeting of the British Association in the ideas of the astrophysicist , Arthur Eddington (1929), who in a book entitled The Nature of the Physical World described this world as "mind-stuff."  Hogben (1930) had countered this in his book The Nature of Living Matter with his "behaviourist" approach, by which he meant that the work of the Russian physiologist, Ivan Pavlov, on the conditioned reflex had destroyed the distinction between "mind" and "matter," making such questions moot.  For Hogben the question shifted from one of ontology or existence to one of epistemology or what was communicable and knowable in the public world of scientific discourse.  In Statistical Theory Hogben (1957) considered Fisher's concept of "a random sample of some infinite hypothetical population of possible values" as emblematic of the Platonic underpinnings of the inferential statistics developed by the British biometric school.  Linking Fisher's hypothetical universe with Quetelet's average man and the normal paradigm, Hogben identified "the angelic choir in the Platonic empyrean of universals with an infinite population of the Normal Man" (p. 180), and he denounced as "Platonic constructs" the concepts of "the infinite hypothetical population, the normal man and the normal environment" (p. 476).  There is truth in Hogben's charge, for in Fisher's world we do appear to be prisoners in Plato's cave, trying to divine the nature of our infinite conceptual or hypothetical populations from the shadows (statistics) cast upon the wall by the Ideas or Forms (parameters) of these populations.  However, there still remains the question of whether what Hogben considered as one of the major weaknesses of British biometrics is actually one of its major strengths.  

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.asis.org/pipermail/sigmetrics/attachments/20110426/f0ae4f08/attachment.html>