A few months ago I was posting on R.A. Fisher's Genetical Theory of Natural Selection. I stopped because I was a bit confused as to what to do about the chapter on dominance, which was basically an exposition on a theory which has been falsified by the preponderance of data. In sum, R.A. Fisher contended that the dominance of an allele emerged via evolutionary processes, while his primary interlocutor Sewall Wright contended that this phenomenon was an emergent property of physiological dynamics. Fisher noted that "Wild Type" phenotypes are invariably dominant. He suggested that deleterious mutants are initially semi-dominant and so reduce fitness in heterozygotes, but modifier alleles on other loci arose in response to selection to mask their phenotypic effect. So, over time a mutant allele would manifest recessiveness where the trait would only emerge in the non-heterozygote state (which is very rare because these alleles were generally extant at low frequency). Wright's explanation was more straightforward: a biochemical process may require only a threshold of protein product, and so one copy of the allele may be sufficient to result in function. The reality is that on many traits there is a discernible difference between Wild Type homozygotes and heterozygotes, but to the first approximation this is not evident. But in any case, the fact that the latter explanation is what most students of biology are taught attests to the fact that the elucidation of molecular mechanisms in genetics over the past few generations has resulted in the vindication in general of Wright's hypothesis.
But there is a bigger philosophical difference at work here I believe. Fisher was at heart a theoretician who applied his sharp mathematical mind to biological questions. Wright was originally an experimentalist, an applied researcher in physiological genetics, who became interested in abstract questions later because he wished to model the systems of breeding within which he worked. This difference in background led to a parting of the ways in regards to the manner in which the two great theoretical minds in population genetics viewed the various forces in evolutionary dynamics. Fisher was the arch-selectionist, and his explanation of the origin of dominance highlights his general recourse to the overarching logic of selectionist processes in shaping genetic architecture and phenotypic expression. In contrast, Wright was originally a physiological geneticist (in our day he would likely have started out a molecular geneticist) who was more conscious of the proximate processes which constrained and counteracted the force of natural selection. Given an infinite amount of time the inefficiencies of mechanistic dominance might be abolished by the emergence of modifier genes, but the reality is that by that time the species and populations in question would likely be at another peak of metastability. In regards to the question of dominance we have here an illustration of the weakness of Fisher's tendency to focus on grand ultimate processes at the expense of minute details of implementation.
Related: Robert Skipper has a post which addresses this controversy between Fisher & Wright in more detail.
- Log in to post comments
Fisher noted that "Wild Type" phenotypes are invariably dominant.
Huh? In any case of an autosomal dominant genetic disease, isn't the wild-type phenotype recessive relative to the mutant phenotype?
Huh? In any case of an autosomal dominant genetic disease, isn't the wild-type phenotype recessive relative to the mutant phenotype?
i meant in terms of overwhelming numerical preponderance (he was looking at mice and drosophila genetics at the time). i guess "invariably" is too strong of a word....
But isn't it pretty obvious that an unfavorable dominant mutation is not going to become common?
Fisher at his best may have been an absolute genius, but it looks like he really dropped the ball on this one. If you look at genes where there is "dominance", for example eye color, it's not like you have a gene for a brown pigment and a gene for a blue pigment and the brown is somehow stronger. Rather, there's a gene for a brown pigment, and eyes without this pigment are blue. And yes, eyes are somewhat browner with two pigment-producing copies. I presume it's pretty much the same situation for fruit flies, except their eyes are red and white.
i meant in terms of overwhelming numerical preponderance (he was looking at mice and drosophila genetics at the time). i guess "invariably" is too strong of a word....
If he was looking at mice, his hypothesis is even more inexplicable. I don't think there were that many allelic series known for mice in 1930, but one of the classics is the nonagouti locus. Of the 5 nonagouti alleles known in 1928 (see Dunn, L.C., 1928, PNAS 14:816-819), two of them, are dominant relative to the "wild type" allele.
Wright agreed that *advantageous* mutations might evolve to become more dominant (through the selection of modifiers). The dispute was over whether *disadvantageous* mutations, initially neither dominant nor recessive, could evolve to become recessive. Wright argued that the second-order selection to make them recessive (selection on modifer genes rather than direct selection against the mutations as such) would be so weak that it would be swamped by any other effects (plus or minus) of the modifier genes on fitness. Wright's argument is plausible enough, but it is ultimately an empirical mattter. There is good evidence that dominance is not *always* due to Fisher's proposed mechanism, but I don't think Fisher ever claimed that it was.