Does Haldane’s Dilemma pose a problem for human evolution?

The short answer

No. It might pose a problem in the future as better estimates are made of some relevant variables, but we are not there yet.

The longer answer

I. What is Haldane’s dilemma?

An article by John Haldane (1957) estimated provisionally1 that gene substitutions would become fixed by natural selection only after an average of 300 generations per substitution. Some creationists say this rate of gene substitution is too slow to account for the number of gene substitutions that occurred in humans since the chimpanzee-human split.

II. How accurate was Haldane’s estimate?

It is unclear how accurate Haldane’s estimate was. Haldane himself stated candidly, “I am quite aware that my conclusions will probably need drastic revision” (Haldane, 1957: 23). Some of the subsequent literature agrees and substantially lowers the number of generations required for natural selection to fix each gene substitution, but the matter appears not to be settled. If you are interested in a lot more detail on this point, I refer you to a discussion by Robert Williams.

III. How many gene substitutions become fixed by natural selection?

According to  Ian Musgrave (2007), it generally is accepted that most gene substitutions are the result of neutral mutations, which become fixed by means other than natural selection. Haldane’s estimate concerns only the rate of fixation for gene substitutions by natural selection, so it does not place a limit on the rate of gene substitution as a whole. To know whether Haldane’s dilemma poses a potential problem for the evolution of humans requries first that one first determine how many positively-selected gene substitutions actually have occurred since the human-chimpanzee split. Musgrave (2007) sets a generous upper bound of around 960 such substitutions, with the actual number likely somewhere between 340 and 580.

IV. Was there enough time for the substitutions?

Estimates for the human-chimpanzee split range anywhere from 4 million years ago (e.g. Hobolth 2011) to 12 million years ago (e.g. Moorjani et al. 2016). Human generation lengths have been estimated anywhere between 20 and 30 years (Tremblay 2000). Combing the extremes of these two ranges with Haldane’s limit, we get a range of between 444 and 2000 relevant gene substitutions. There is overlap between this range and Musgrave’s estimate of the number of actual substitutions. Narrower ranges for the above variables might reveal a problem, but they also might demonstrate once and for all that there is no problem; only time will tell.


Haldane JBS. 1957.  The cost of natural selection. Journal of Genetics 55:511-524.

Hobolth A., Dutheil JY, Hawks J, Schierup MH & Mailund T. 2011. Incomplete lineage sorting patterns among human, chimpanzee, and orangutan suggest recent orangutan speciation and widespread selection. Genome Research 21:349–356.

Moorjani P., Amorim CEG, Arndt PF & Przeworski M. 2016. Variation in the molecular clock of primates. Proceedings of the National Academy of Sciences of the United States of America, 113:10607–10612.

Musgrave I. 2007. Haldane’s non-dilemma. Last accessed 30 Jun 2018.

Tremblay M. & Vézina H. 2000. New estimates of intergenerational time intervals for the calculation of age and origins of mutations. American Journal of Human Genetics, 66: 651–658.