One common and rather expected comment is that what I’m talking about is “pseudoscience”. This is the most straightforward to address. The components of my argument, the pillars of it, are based on valid and sound arguments based in mainstream and accepted science. There is a lot evidence supporting those. The conclusions based on that evidence are not in the realm of whether or not they are supported, but by “how much”.
General cognitive ability is heritable. This is not in question for anyone even remotely aware of the facts and evidence available from modern science.
Even direct experience, (although anecdotal) and having as small a data set as a single person could discover in their daily life, supports this conclusion. We would likely consider it “common sense.” While this debate went on for decades, there is no longer a reasonable conclusion that general cognitive ability is entirely, mostly, or in the majority based on environment. While environment is a factor, neither we nor the enormous body of research suggests that is not the case. The question now in research is “how much” impact it has.
What we’re discovering is that environment matters less than you think it would. The
heritability of intelligence is high. Arguments against this tend to involve moving goalposts, redefining terms, or are fallacies called “arguments from ignorance” or “of the gaps.” Let’s not spend too much time on that. The heritability of general cognitive ability is at least fifty percent, with evidence that heritability increases to between seventy and eighty percent on reaching adulthood. Again, this is material taught at at the very least at the graduate level in current university courses. Hopefully at the undergraduate level.
The next point is that if the trait is heritable, that it would vary within a population. This
should be obvious, both from any basic understanding of genetics and from direct observation.
This is so common-sensical I hardly fathom why I have to mention it. Whether it were one allele variant, ten, one hundred alleles with multiple variants, or one thousand that
contributed to some trait, the combinations that contribute to the expression of that trait
differ in their distribution throughout a population.
We do know that for some cases, a single gene is responsible for differences in expression in the phenotype. Just one. For other traits, the expression of that trait is moderated by one or more other genes, or masked by them. It is the case that sometimes differences in a single gene are responsible for distinct and important phenotypic differences–ones that then are more or less prevalent in a group or population, as well as more or less prevalent in one population or another.
So, even though observation of a single SNP, a single allele variant, or a single chromosome might not tell us enough to know which of two (or more) groups a subject fit into, it does tell us something. Odds, for example, of being in this or that group, but with no certainty.
What we do know, is that when we sample many many SNPs from across the genome of two subjects, we can tell with more than 99% accuracy what that subject’s ancestry is. We can tell a great deal about the subject, including what geographic ancestral populations are present or absent in their heritage.
Oh, oh, you say–that doesn’t matter, because reasons. No, no, we’ve just outlined above why is does matter. If you don’t see that yet, let’s take a look at it.
We can identify membership in a population at a very coarse and at a fine level as well. At a regional level, there is plenty of mixing at boundaries, to be sure, and mixing that has gone on due to historical migrations, and due to modern travel. However, it is very clear from the current science that we can still identify the component historic populations and ancestors.
Otherwise, 23andMe and Ancestry.com wouldn’t be able to do what they do.
One critic for some reason mentioned that a variant for sickle cell could show up across
populations. While it could, there are two different variants that arose, and having one vs the other tells us something about the likelihood of having ancestry from one population or the other. That plus a selection of other gene variants and we can be rather sure what the subject’s heritage is.
In short, the more of the genome and the more variants you examine for a given subject, the more certain you can be what the various ancestral contributors were. You know the regions. If we leave it at region and population, and not that other word that is so charged and distracting, perhaps that will help.
The distinctions we make, the labels we apply to these populations are not arbitrary in the sense that they don’t have descriptive, explanatory, or predictive power. We can determine many things about a subject regarding health and risks for diseases, and other likely and potential present properties. This is crucial to understand. This destroys the social construct argument. If you don’t see how, start over.
Not all members, because of variation, will have the traits that are more prevalent. Hopefully it’s clear that I’m not saying that–I don’t think anyone is saying that. Membership in a population is not deterministic for these traits. It just means the probability is higher as the traits are found more often in the population, as those gene variants are more often present, whether one variant or more in combination. This is well understood and is why it’s not a “social construct” when applied in health care. Your ancestral population or populations matter.
An attempted critique was that “at the level of an ethnic group, this breaks down.” I found this extremely peculiar as a claim for several reasons. Most importantly that the science of genetics shows almost exactly the opposite. We can identify these smaller groups very well, and make the same sort of predictions based on membership in that group.
Second, although my video did address culture, I would not have suggested that an “ethnic group” was a coherent population from a genetic perspective. Multiple genetically distinct populations can be members of the same ethnic group. This would actually be a fascinating thing to study.
Now, it may not be just one gene or one variant, as we’ve discussed, that an expressed trait is coded by. The same apparent or effective result might occur because of more than one variant. Something you observe might not have the same cause across one population. Some of those variants might be present in another population, but not be expressed in the phenotype due to the presence of other genes or variants thereof. In this way, you can have variation that’s larger in one population, and have some smaller number of those variants in another population, but still have distinct differences in phenotypic expression and criteria to distinguish the populations from one another. This is key. The populations can still be classified differently. They are distinct.
The classifications into major groupings are not arbitrary. They are not just labels, as I’ve
already said. They are labels, yes, but we label many things and the labels have utility. They are useful. They have descriptive power, predictive power. Life and death predictive power, in the field of health care.
Given what we’ve discussed, hopefully you see that the there is no sane way to accept the criticism that “at a genetic level, there is no such thing as this categorization of these populations.” That is absurd.
All that we have discussed so far is solid science. Valid and sound.
It is the next bit, the fusion of the two, that seems to rankle and rile up our detractors the
most. We know that intelligence is a heritable trait. I would hardly suggest that Steven Pinker is an evil racist Nazi. At a popular science level, he tells us, “All traits are heritable. Environment matters less than you think.”
Since we know from study of the genome that there is variation between populations, it is possible for intelligence to vary between populations. And the evidence shows that it, in fact, does. Whether we do or don’t know if it’s one gene (it’s not), ten, one hundred, or one thousand, with dozens of variants each, the fact is that intelligence varies between
populations of different geographic ancestry. The average level of cognitive ability is
different between different groups. The most generous estimate available for the anti-hereditarians from studies, primarily from pre-adult subjects, is that only fifty percent of the variance is due to environment. The other fifty percent, in this case, is due to heredity. I personally find the evidence of a heritability of .7 to .8 on reaching adulthood to be more compelling, which makes things rather worse for the environmental side of the argument. Fluid when young, more fixed when older. Not really surprising.
We know intelligence is heritable. We know that it varies between populations (as well as within them–of course).
I’d hope the critics would not argue that intelligence has no impact on culture. Clearly if a population has a larger number of bright and brilliant individuals that will have an effect on the cultural achievement of that population.
Some populations do have more intelligent members than others, as those populations have a higher average intelligence. I don’t have to know what genes, how many variants, or anything like that, because for our purposes all we need to know at the moment is that it’s heritable.
It is. There is no question. Genes influence intelligence. Average intelligence varies between populations. Intelligence influences culture. Therefore, a population’s genetic makeup influences their culture. To some extent. How much remains a question. But genes influence culture. Q.E.D.
There is no pseudoscience here. No religiosity. No a priori racism. Just the truth. What we do with this truth is an entirely different matter.