Some of the complexities involved in gene-environment interplay

Tabery’s review of the history of disputes over geneenvironment (G x E) interaction provides a splendid, and persuasive, account of the need to consider the fundamental, philosophical differences between a variationpartitioning approach and a mechanism-elucidation approach—the former being exemplified by Fisher and the latter by Hogben. Fisher is rightly described as a giant of 20th century science but, in relation to G x E, it is relevant that his understanding of some aspects of genetics was poor. This is well exemplified by his attack on the claims from Richard Doll that smoking might cause lung cancer. 1 This was based on the factual evidence that there is a substantial heritability to the propensity to smoke cigarettes. Fisher’s mistake was to ignore the fact that the argument that there was no environmental effect of smoking on cancer had to be based on evidence that the liability to smoking was highly correlated with the liability for lung cancer, and this was neither tested nor shown. Tabery says that Hogben was not so well known as Fisher, although he made an equally lasting contribution to science. It is probably true that Fisher was better known among scientists but it is probably equally true that Hogben was better known to the general public. His book published in 1936 Mathematics for the Million was a wonderful example of how the magic of mathematics could be made both understandable and of riveting interest. 2 Albert Einstein