Populations are not races

In a recent NYT op-ed, How Genetics is Changing Our Understanding of Race, geneticist David Reich challenged what he called “orthodoxy” about the use of race as a concept in genetics. A few days after this op-ed appeared, I wrote a short Twitter thread laying out some of my problems with it.

I’ve since gotten requests to expand on that thread in a blog post. I’d given up long-form blogging, but this topic can’t be avoided; so, here we are.

I don’t know David personally. Of my geneticist colleagues who do know him, none seem to think he was motivated by racism, crypto-racism, or even self-promotion (he has a book out). Most seem to think that – like his book – the op-ed was an honest, if counterproductive, attempt at outreach.

I don’t actually believe his motivation is important, only the consequences. However, he goes to great pains to distance himself from racism at several places in his text, and I do want to make it clear this isn’t meant as an ad hominem attack. The language and positioning of the piece, rather than the character of the writer, are what I want to engage with.

Race is a social construct

The piece opens by acknowledging that “race” is a social construct. As a genetic term, it’s very imprecise:

In the United States, historically, a person is “black” if he has any sub-Saharan African ancestry; in Brazil, a person is not “black” if he is known to have any European ancestry. If “black” refers to different people in different contexts, how can there be any genetic basis to it?

This point is reiterated later in the piece:

It is true that race is a social construct.

And again:

Self-identified African-Americans turn out to derive, on average, about 80 percent of their genetic ancestry from enslaved Africans brought to America between the 16th and 19th centuries.

And yet again:

…my laboratory discovered in 2016, based on our sequencing of ancient human genomes, that “whites” are not derived from a population that existed from time immemorial, as some people believe. Instead, “whites” represent a mixture of four ancient populations that lived 10,000 years ago and were each as different from one another as Europeans and East Asians are today.

So far, so good

The above quotes should make it clear that “race” is a concept defined by society, not by genes. It may be statistically correlated with ancestry, but never perfectly. Worse, it’s highly inconsistent: self-identifying as “black” in America accounts for only 80% of your genome (on average), self-identifying as “white” means you’re an unspecified mix of four very different populations, the terms “black” and “white” mean different things in different places at different times, and so on.

One reason for the ambiguity is what geneticists call admixture: migration and mixing cause once-disjoint ancestral populations to become mixed up. Very few individuals hail exclusively from one ancestral group. There may be more of one group than other in your DNA, but you’re a mix of multiple populations. No geneticist who has looked at the data would argue otherwise. (If you want an example, Figure 2 in the 1,000 Genomes Project is decent.)

Unfortunately, this is not a discussion among geneticists, but a widely-read NYT op-ed. And through the course of this op-ed, thanks to some unfortunately sloppy language and poorly-chosen use of ironic quotes, the distinct concepts of “races” and “populations” appear to become quite admixed themselves.

Triangulating with straws

One problem is that Reich appears to be trying to triangulate between, on the one hand, some specifically-named people who have expressed some seriously racist views (e.g. Nicholas Wade, Jim Watson – the links go to articles that critique their claims) and, on the other hand, anonymous colleagues who (Reich claims) are perpetuating some kind of “orthodoxy”:

I am worried that well-meaning people who deny the possibility of substantial biological differences among human populations are digging themselves into an indefensible position, one that will not survive the onslaught of science.

These “well-meaning people” aren’t named, which has the effect of ceding ground to the people who are. Those named people can now paint themselves as intellectually courageous holdouts without even identifying an opponent. Here’s Nicholas Wade doing just that.

In fact, the argument Reich is presenting as a counter to Wade et al is a straw man:

The orthodoxy maintains that the average genetic differences among people grouped according to today’s racial terms are so trivial when it comes to any meaningful biological traits that those differences can be ignored.

I simply don’t know any scientists who claim this. Scientists claim, rightly, that races are an imprecise proxy for ancestral populations (as shown above). They also claim, rightly, that average differences between ancestral populations are typically very small compared to the differences within those populations. But I don’t know anyone who says “therefore those differences should be ignored”. Perhaps he’s talking about non-scientists; but even so, who?

Ironic quotes are “bad”

Beyond the straw man of “well-meaning people who deny…”, there are a few lines in the op-ed that significantly undercut the key point that races and populations are different things.

One such line has Reich putting highly unfortunate ironic quotes around “races”:

I have deep sympathy for the concern that genetic discoveries could be misused to justify racism. But as a geneticist I also know that it is simply no longer possible to ignore average genetic differences among “races.”

You can see those are ironic quotes, right? He puts the quotes around “races” because he knows that “races” aren’t actually real.

I can’t emphasize this enough: Please, kids, do not use ironic quotes on the Internet. Especially for serious points! To reach for an analogy: say you’re at a party, you’ve had a few drinks, and you get this spontaneous, brilliant idea to juggle with fire, or swallow knives. It might seem like a great plan but, really, it’s just not worth the style points. It’s unsafe unless you’re a pro, and even then some idiot will probably try to copy you – with predictable, ugly consequences. Irony quotes are like jugglers’ torches or sword-swallowers’ knives. Best leave it.

I’ve seen some suggestions that these ironic quotes exonerate Reich: clearly he didn’t mean that races are real! Well sure; I’d be a whole lot more sympathetic to that if it wasn’t for the fact that “laying out a rational framework for discussing differences among populations” is the whole point of the op-ed. Otherwise, as Reich points out, we “leave a vacuum that gets filled by pseudoscience”. Um, yep; that also happens if we use the pseudoscientific terms ourselves. Sorry, I meant if we “use” the terms ourselves.


Reich sows additional confusion with lines like this, talking about his own research into hereditary risk factors for prostate cancer:

Did this research rely on terms like “African-American” and “European-American” that are socially constructed, and did it label segments of the genome as being probably “West African” or “European” in origin? Yes. Did this research identify real risk factors for disease that differ in frequency across those populations, leading to discoveries with the potential to improve health and save lives? Yes.

The conceptual-admixture is strong in this paragraph. Terms like “African-American” are socially constructed; labeling segments of the genome as “probably West African in origin”, on the other hand, is a reasonable exercise in genetic inference. Placing them side by side, in a paragraph whose overall point seems to be (paraphrasing) “we used ‘race’ as an imperfect proxy for ‘ancestry’ and we saved lives”, seriously undercuts the earlier points about ‘race’ and ‘ancestry’ being separate concepts.


The op-ed appeared 11 days ago, an eternity in Internet Time. Since then, there’s been a rebuttal from social scientists that has ruffled some geneticists’ feathers for introducing additional technical errors into the discourse. (Unfortunately, I think we geneticists ourselves bear much of the blame for this, as a coherent collective response has been lacking.) David Reich wrote a follow-up one week later, which tries to clarify his position.

Meanwhile, however, the damage has been done. People like Nicholas Wade have claimed the article as cover:

At last! A Harvard geneticist, David Reich, admits that there are genetic differences between human races, even though he puts the word race in quotation marks.

See what I mean about ironic quotes? As a shield against misinterpretation, they suck.

Other commentators have seized on David Reich’s comments about genetics and intelligence from the same op-ed. I haven’t even gone near that, since it’s a whole other hornets’ nest… in brief, I find this another case where scientists (especially when speaking publicly) should really break down vague terms such as “intelligence” into more precise constituent concepts.

Furthermore, we really lack the data to predict any form of “intelligence” (ironic quotes, O NOES!!!!) from genes, at least at the present time. This thread by Antonio Regalado does a decent job of summarizing this latter point about lack of data (though I wish he’d avoid using “race” as his example, for reasons I hope are obvious at this point). This post from Graham Coop, using tea-drinking in Britain and France as a toy example, is really good too.

So how should we talk about this?

“Race” is not only a socially constructed term, but an incredibly loaded one. The United States was second only to Nazi Germany in eugenics. Authority figures spout racist, xenophobic, bigoted tropes. Scientific racism is a lurking menace. It’s incredibly naive to think that geneticists can use terms like “race” in public discourse without having those words misappropriated and abused by racists.

Here’s an idea: if we want to lay out a rational framework for discourse, to defeat pseudoscience… why not try being consistently clear with terminology? How about not using pseudoscientific terms, not even in an ironic way? If “race” is primarily a social concept and very flaky as a genetic concept, then geneticists should have no truck with it. Call in the social scientists!

I actually think David Reich did a better job of articulating this in his follow-up, where he lists “six truths”. Unfortunately, those six truths are buried at the end of his article. I’m going to repeat them here (yes I realize this is also at the end of my blogpost, oh well):

1. “Race” is fundamentally a social category — not a biological one — as anthropologists have shown.

2. There are clear genetic contributors to many traits, including behavior.

3. Present-day human populations, which often but not always are correlated to today’s “race” categories, have in a number of instances been largely isolated from one another for tens of thousands of years. These long separations have provided adequate opportunity for the frequencies of genetic variations to change.

4. Genetic variations are likely to affect behavior and cognition just as they affect other traits, even though we know that the average genetic influences on behavior and cognition are strongly affected by upbringing and are likely to be more modest than genetic influences on bodily traits or disease.

5. The genetic variations that influence behavior in one population will almost certainly have an effect on behavior in others populations, even if the ways those genetic variations manifest in each population may be very different. Given that all genetically determined traits differ somewhat among populations, we should expect that there will be differences in the average effects, including in traits like behavior.

6. To insist that no meaningful average differences among human populations are possible is harmful. It is perceived as misleading, even patronizing, by the general public. And it encourages people not to trust the honesty of scholars and instead to embrace theories that are not scientifically grounded and often racist.

This, I think, is a more reasonable style guide for talking about genetics and race. It still has problems (point 6 repeats the straw man) but it is an improvement. Most importantly, I think we need to emphasize point 1 (race is a social category) and be ready with multiple data-founded talking points to back up point 3 (race is only partly correlated to population).

If we aren’t clear that admixture happens between populations, or about the other reasons why “race” is flawed as a genetic concept, then admixture will occur between the concepts of “race” and “population”. Genetics will be abused to reify discredited racist stereotypes. This could happen all too easily. The knock-on effects of this, as DNA sequencing becomes more pervasive (as it inevitably will), would be more damaging than anything Cambridge Analytica or the Internet Research Agency could inflict on our society.


14 thoughts on “Populations are not races

  1. “I simply don’t know any scientists who claim this”

    See http://judgestarling.tumblr.com/page/5

    “They also claim, rightly, that average differences between ancestral populations are typically very small compared to the differences within those populations”

    This isn’t obviously true. Height is an example where the variation between is as large as variation within.


    • Thank you for these links.

      Graur is making a specific claim that population sizes and separation times are too small to permit 74 alleles evolving in concert. Mitchell is claiming that a specific quantitative trait (IQ) is mostly due to environmental factors. Neither is “denying the possibility of substantial biological differences among human populations”, which is the claim I (still) haven’t seen scientists making.

      Yes, height is indeed an example where inter-group variation can rival intra-group variation. I didn’t say there were no such cases, or that the consensus position of science is that there are no such cases. What I wrote is that this is *typically* not the case for quantitative traits of interest. “Typically” is not the same as “universally”. Height is an exception. Of course it depends on what quantitative traits you find interesting.


      • Graur argues that no meaningful differences in complex traits could have evolved, which is effectively the same as “denying the possibility of substantial biological differences among human populations”.

        Mitchell is talking about IQ specifically, but his claim is similar to Graur’s in that he makes the very strong claim that we can rule out differential selection for cognitive ability, even though there is little to no research that supports his position.

        What reason do we have to regard height as a special exception rather than as an example where genetic differences are particularly obvious?


      • You seem to be willfully misinterpreting both Graur and Mitchell to generalize their claims. There’s plenty of evidence that differences in IQ (for what that’s worth) are bigger within populations than between. I think we’re done here.


  2. I see myself so strongly in this article, in your complaints. If he said two sort of true things, he’s not “sowing confusion,” the subject matter is, the world is. You’re right of course, the very overlap that works for that medical story works for Nazis, and that sucks. And I’m not happy that we’re stuck trying to control everything we say because some Nazi will find a way to twist it, they always will anyway, it’s their fault, not ours. Caveat: all I got is this article, I know no other particulars of this story, I am barging in not knowing who is on whose “side.”


  3. “But I don’t know anyone who says “therefore those differences should be ignored”. Perhaps he’s talking about non-scientists; but even so, who?”

    – me, ultra-liberals, SJWs, people concerned with fighting inequality, people who would prefer life sans Nazis . . . gender is “real,” but equality demands laws that insist we do not let it rule everyone’s life.


  4. “I am worried that well-meaning people..[ ]. ” which you rebut with “I simply don’t know any scientists who claim this…”

    Why do you narrow his meaning to only scientists? Do you really not think there are many people, both influential and not scientists, who deny the exact things Reich claims “well-meaning” people deny?

    Essentially it looks like you’ve substituted “scientists” for “well-meaning people” and then spent much of your time rebutting that version…which looks very much like a strawman, however unintentional….



    • First, it is not a straw man if it’s intentional, and when you say I “spent much of [my] time rebutting it”, let’s be clear that we are talking about exactly one paragraph of my post, containing five sentences – of which the last is “Perhaps he’s talking about non-scientists; but even so, who?”

      If you would have preferred I devote slightly more of that paragraph to the latter premise, that he is worried about non-scientists, then fair enough; except that (a) in that case I have even less of an idea which people he means, and consequently even less to say about it, (b) it’s far from clear to me that this *is* in fact what he means. The build-up to that point starts with him describing a scientific “consensus” based on Richard Lewontin’s work that has supposedly morphed into “orthodoxy”, but he then talks about how we are “sticking our heads in the sand”.

      I agree that one possible reading of this is that he means that geneticists believe one thing but, with “heads in the sand”, they are allowing “well-meaning” laypeople to succumb to pseudoscience. It’s also a possible reading that he thinks geneticists (or some of them) have succumbed to the orthodoxy. I genuinely don’t know, and the real problem is that the text is woefully unclear on this point, because “Some people say…” and “I am worried about people who think…” are terrible rhetorical devices. If you are making a counter-argument then you should first clearly cite the argument. So this goes to my more fundamental criticism, which is that a geneticist doing outreach on race MUST take extra pains to be crystal clear, and avoid vague unidentified counter-arguments (“I am worried about well-meaning people…”) as well as things like pseudoscientific terms in quotes.


      • Thanks for the reply.

        I realize you hedged by saying “maybe he’s referring to the public”, but I don’t think that’s enough to make it a strawman.

        The most plausible, parsimonious – and most charitable – reading is that he means “well-meaning people” as a weighted reference to the public. The next sentence after the one you quote makes it almost explicit.

        “…….I am also worried that whatever discoveries are made…. will be cited as “scientific proof” that racist prejudices and agendas have been correct all along, and that those well-meaning people will not understand the science well enough to push back against these claims.”

        Again – “those well-meaning people will not understand the science well enough”

        Given that his talk of “orthodoxy” was preceding all that, the most parsimonious reading is again that it’s a weighted reference to the public. But I definitely take your related criticism of the piece that he could have been more explicit.

        And I also take your general point that it’s slippery move for Reich to always refer generically – instead of specifically or with names – to those he’s arguing against.



      • We disagree on what’s the most plausible reading. My interpretation is backed up by point 6 of his second op-ed, where he clearly attributes this idea to scientists (“To insist that no meaningful average differences among human populations are possible is harmful. It is perceived as misleading, even patronizing, by the general public”).

        Either way, a good-faith disagreement where the possibility of misinterpretation is explicitly acknowledged is in no way a strawman.


    • Note also that in his follow-up op-ed, he repeats the straw man more clearly (point 6), and in this repetition he is quite clear that he means scientists:

      “To insist that no meaningful average differences among human populations are possible is harmful. It is perceived as misleading, even patronizing, by the general public. And it encourages people not to trust the honesty of scholars and instead to embrace theories that are not scientifically grounded and often racist.”


Comments are closed.