The Latin square

A few years ago, R. A. Fisher (who has been derscribed as “the greatest Darwinist since Darwin”, and as “the founder of modern statistics”) fell victim to one of the waves created by the Black Lives Matter campaign, and a window commemorating him in his Cambridge college was removed (undoubtedly without due process: just two days elapsed between the proposal and the fait accompli). I wrote about this, and there were quite a few comments: see here.

Now I don’t have to defend Fisher any more, since A.W.F. Edwards has done it so much better in a book, The Latin Square, published by Cam Rivers Publishing last year. (The title was given because the window illustrated a Latin square which occurs on the dust-jacket of Fisher’s book The Design of Experiments.)

The book is a collection of essays written by Edwards at various times both before and after this incident. He does a good job of dispelling the misunderstandings, sharp practice and simple lies which fed the campaign to discredit Fisher.

Because of its origin, the book is somewhat repetitive. I will be briefer and just refer to the book. I say a few words about three allegations against Fisher.

1. Fisher was a racist. It is hard to see how this could be true, given that on his death the Indian Statistical Institute closed for a day as a mark of respect.

Fisher’s view, in his words:

Mankind as a whole certainly constitutes a single family, and it is an old ideal and certainly not a dead one to treat all mankind as our brethren.

Briefly, how did the allegation arise? Possibly because he objected to one statement in a UNESCO report on race in 1951:

Available scientific knowledge provides no basis for believing that the groups of mankind differ in their innate capacity for intellectual and emotional development.

As Edwards points out, this is a “null hypothesis”, a concept invented by Fisher, who was clear about the fact that a null hypothesis can never be proved but may be disproved by further research. In Fisher’s time the best verdict on it was probably “unproved”, but advances in genetics have surely disproved it now (despite the protestations of people like Stephen Jay Gould). Edwards also explains how its failure can come about simply as a result of genetic drift, without need for selection pressure. Needless to say, this statement can fall but does not take away our responsibility to treat all of the human family without prejudice.

2. Fisher was a eugenicist, in the disagreeable sense of the term in the USA and Germany. Again refuted by Fisher’s expressed view:

… anything so big as eugenic aims must be controlled by the personal choice of individuals acquainted with their own individual needs and circumstances …

Fisher was in favour of child allowances, to encourage “clever” people to have more children: not given by the State, but funded by contributions rather as occupational pensions are now. Ironically, it was William Beveridge, a fellow member of the Eugenic Society, who introduced state-funded child benefit. Imagine the storm if they were done away with now!

The source of the trouble may the phrase “elimination of defectives”, but for Fisher this means “defective genes”, not “defective individuals”, and he hoped it could be done by a combination of genetic counselling and voluntary sterilisation, although in one of his papers (misunderstood by historians) he proved that this would be a slow process.

3. Fisher’s interest in statistics and genetics was driven by his enthusiasm for eugenics. This is simply false, and in my view the real disgrace here is that historians could make this accusation without looking at either Fisher’s publications or his biography. Edwards remarked that after his degree he had the choice of a permanent position in Pearson’s laboratory or a temporary position at Rothamsted Experimental Station; he chose the latter.

I will end with two remarks arising from comments on my previous post on Fisher.

First, a good point in the other direction which needs to be addressed. As one correspondent pointed out, how would I feel if I were of a diffent “race” and had to eat every day in a hall illuminated by a Latin square constructed by a racist? Even if Fisher were not a racist (as seems to be the case), as long as he is perceived as a racist this point still has validity. But in that case, the blame falls not on Fisher but on those people who created and pushed the myth. Now myths can be very powerful; and if they are supported by Fellows of the Royal Society and former Regius Professors of History, it takes a certain amount of courage not to be affected by them.

Second, I have to reveal my own prejudices. Another commenter quoted extensively from Lancelot Hogben. I do not warm to Hogben. I was given one of his books as a school prize, and although it was a maths book I found it absolutely unreadable. Later I encountered G. H. Hardy’s account of his views (in A Mathematician’s Apology), which were that the only mathematics worth doing is useful mathematics (which according to Hardy’s view meant roughly school mathematics). Unnnervingly close to the views that Stalin’s thugs took of people like Egorov, Luzin, Cebotarev, and Bronstejn. And, incidentally, the flashpoint between Fisher and Hogben was likely to have been politics: Hogben was on the left.