Haldane and the Emergence of Modern Evolutionary Theory 73In the dispute between Fisher and Wright, Provine sided with Wright. But
there was the broader dispute, that between the genetical and naturalist narra-
tives mentioned at the beginning of this paper. In that dispute, both Fisher and
Wright and, for that matter, Haldane, Kimura and the other theoretical popula-
tion geneticists, are jointly pitted against Mayr and his followers. By emphasizing
the significance of the Wright-Fisher dispute in the history of evolutionary theory,
Provine implicitly denies Mayr’s reconstruction. There will be much more of that
below but, meanwhile, the relevant question here is why Provine’s later accounts
also ignores Haldane. Part of the answer may lie in Haldane’s consistent refusal
to take sides in the dispute between Fisher and Wright: each of them found Hal-
dane to be closer in position to the other. Provine, following Wright, presumably
placed Haldane with Fisher and hence doubted his significance.^33 But the main
reason is Provine’s fundamental decision to present the Fisher-Wright dispute as
the most central issue in the history of evolutionary theory. A perhaps inevitable
consequence of this decision, in spite of all the achievements recorded earlier in
this paper, is the elision of Haldane’s role in that history.
To turn to the naturalist account, already in the 1950s, the genetical account
began to be challenged. The first to explicitly challenge it was C. H. Waddington.
In 1952, in Oxford, at the Symposium of the Society for Experimental Biology,
Waddington argued that the prestige of the mathematical work of Fisher, Haldane
and Wright was unwarranted. It
did not achieve either of the two results which one normally expects
from a mathematical theory. It has not, in the first place, led to any
noteworthy quantitative statements about evolution. The formulae in-
volve parameters of selective advantage, effective population size, mi-
gration and mutation rates, etc., most of which are still too inaccurately
known to enable quantitative predictions to be made or verified. But
even when this is not possible, a mathematical treatment may reveal
new types of relation and of process, and thus provide a more flex-
ible theory, capable of explaining phenomena which were previously
obscure. It is doubtful how far the mathematical theory has done this.
Very few qualitatively new ideas have emerged from it [1953, 186].Waddington [1957] republished these remarks verbatim in 1957, inThe Strategy
of the Genes. For him, the only contributions of mathematical population genetics
were (i) the demonstration that ordinary Mendelian genes “would respond to the
process of natural selection” (p. 61); and (ii) the demonstration that continuous
variation can be accommodated within a Mendelian framework, that is, Fisher’s
work from 1918.
(^33) Moreover, almost all the population genetic models that Haldane had ever constructed were
deterministic. They assumed large (in principle, infinite) populations. They were models in
which selection would generally prevail over population structure; certainly, they had no room
for stochasticity. Here, even from a mildly Wrightian point of view, Haldane and Fisher clearly
belonged together.