Every now and then someone tries to give some genetic argument showing that population A and population B can’t possibly be significantly different in any trait that hurts someone’s feelings. They’re all bullshit.
The most famous one is probably Lewontin’s Fst argument. He said that since most (> 85%) genetic variation in humans is within-group, rather than between groups, human populations can’t be very different. Except for traits influenced by a few genes, like skin color, or a single gene that affects many traits, like EDAR, or a highly polygenic trait influenced by many genes, like height. That is, any kind of trait, with any kind of genetic architecture.
A related argument says that you only see differences in ‘superficial’ traits, like bone density, brain volume, or the shape of the skull. That’s sarcasm.
Another says that populations just haven’t been separated long enough to diverge much. Untrue: selection can move pretty rapidly. Amerindians have undergone noticeable genetic change (mostly immunological) over the past five hundred years. While some human populations been separated for much longer than that: often tens of thousands of years, sometimes (with Bushmen for example) hundreds of thousands of years. Plenty of time.
Lately I’ve seen a few people arguing that there can’t be big between-population differences in highly polygenic traits, those influenced by many alleles of small effect, because evolution of a highly polygenic trait is for some unknown reason very slow -” at least 100 millennia to evolve appreciably”. But that isn’t the case.
They’re thinking about intelligence – but height is also polygenic. Yet Pygmies and Tutsi exist. For that matter, natural selection over the past few thousand years has made southern Europeans shorter and northern Europeans taller.
I have no idea where this notion originated, because there’s nothing in genetic theory or practice that supports the idea that polygenic traits can’t change rapidly. Think of the Maltese elephants, isolated when sea level rose at the end of the ice ages: 3 feet tall in just a couple of thousand years.
While there are many counterexamples in agricultural genetics: you can breed for almost any trait* – appearance, conformation, size, disease resistance, flavor, behavior – and we have done so. Many of those traits are highly polygenic. In fact, Fisher’s infinitesimal model, fairly standard in evolutionary genetics, assumes that an infinite number of alleles influence a trait.
Most of these fallacies are addressed in the tome I’m perusing. The author correctly dismissed all of them. But then…
* except sex ratio. That’s nearly impossible.