My letter to the Washington Post on race

October 22, 2023 • 9:15 am

About a week ago, the Washington Post published, starting on its front page, a long article arguing that race is a purely social construct without reality or utility, and thus should be eliminated. The author Sydney Trent, is a science journalist who covers social issues, and that may explain why the article was replete with scientific problems, among them the neglect of existing research on ethnic groups (my preferred term for “race”). You can see the article by clicking on the headline below. Since it’ll probably be paywalled if you subscribe, I found the whole article archived here.

Leaving aside the misleading “science says” (science doesn’t say anything, scientists do; and not all scientists agree that race isn’t real), I’ll show you three small excerpts of Trent’s piece:

Yet unlike in decades past, more ordinary Americans are coming to see “race” for what it is, [Carlos] Hoyt maintains. In interviews he conducted for his doctoral thesis and book, these people describe gradually awakening to the idea — through traumatic personal experiences with discrimination, through foreign travel or something they read — that they had been sold a bill of goods. “Race,” they decided, does not exist.

. . . The truth, [Adrian] Lyles, 37, said, is that “race has no quantifiable metric,” like socioeconomic status, for example, he said. “Where you have unreliable input, your data is trash.”

. . . In 2003, the completion of the Human Genome Project — which found that humans globally share 99.9 percent of their DNA — gave waste to the notion of “race” among the vast majority of scientists. But the public appears barely to have noticed. The idea still lives everywhere — in discrimination and criminal profiling, in the rise in hate speech and acts, in the recent Supreme Court decision ending affirmative action in college admissions, in the rhetoric of social justice advocates and the new capitalization of Black and White in the media. Racial categorization persists on job applications, medical forms, and most critically to Hoyt due to its high visibility, the Census.

Implicit in Trent’s effort to dethrone the term is the misguided idea that if you think “races” have any biological reality, then that buttresses racism.  That need not be true, but, historically, belief in races has been associated with the idea of a racial hierarchy in various traits (most often intelligence), and so I prefer to use “populations” or “ethnicity”, which doesn’t carry that historical taint.

Trent concludes that racial categories should be eliminated everywhere, especially on the census. The problem is that from the DNA figures above, she concludes that “racial categories”—the half-dozen or so “races” recognized in the past (white, black, Asian, and so on)—have no biological significance.  But she conflates “racial categories”, the named “boxes” above, with “race”, which I take to mean “a population that is genetically distinguishable from other populations of our species”.

Classical “races” were assumed to be absolutely demarcated geographically and morphologically, and to be separated by substantial genetic differences.  We know now that this conception of “race” isn’t true. There are no absolutely clear-cut categories into which everyone fits, genetic differences between even the “classical” races are not large, and there are “races within races”: populations that can be distinguished genetically from other populations often put into the same classical race. Again, that’s why I use “ethnicity”  or “population” to refer to such groups.

But there is no doubt that ethnicity, and even the “old fashioned” races, carry meaningful biological information and are genetically differentiable. If they weren’t, then you wouldn’t be able to pay companies like 23andMe to suss out your ancestry, or to trace the history of human migration by using genetic differences between populations. (23andMe told me from my DNA that I am 97.2% Ashkenazi Jew and 2.8% Eastern European, which matches perfectly with what I know from my family history.)

Ethnicity reflects evolutionary history, and if you use thousands of DNA sites (as you see in my letter below, even 99.6% identity between people—not 99.9% as Trent wrote—still leaves, in a genome of 3 billion base pairs, at least 12 million variable nucleotides. That variation is largely correlated with ancestry and geography, so that, for example, the DNA of most Europeans allows you to identify their birthplace to within 500 miles. Luana Maroja and I described the real situation in our recent paper (“The ideological subversion of biology“) in Skeptical Inquirer, a magazine published by the Center for Inquiry. (The race material is under point #5 of the paper.)

And, as I say in my published letter, one study showed that if you ask people to self-identify their “old fashioned” race (they used 3,636 Americans who self identified as either African American, white, East Asian, or Hispanic), and then independently look at their DNA in a blind study, you find that when you compare the DNA with the self-identification, you find a 99.84 percent match! That means that even the widely-reviled “classical” races are genetically differentiable using cluster analysis.  This is not surprising because these groups evolved in different parts of the world, and for much of their history they evolved in semi-isolation, leading to the accumulation of differences in the DNA by either genetic drift or natural selection.

At any rate, the Post‘s article was scientifically misleading, and so I set out to correct it by writing a letter to the paper. Mirabile dictu, they published it, and you can find it by either clicking on the screenshot or by simply reading my letter reproduced below the headline.

They edited it fairly heavily for length, so I had to leave out stuff like locating someone’s birthplace from their genes. Still, I think I did make the point that there is substantial genetic variation among people and diagnostic genetic variation among ethnic groups, and that this variation is useful in several ways.  If I could make one change, it would be to re-insert something that was cut and that I missed when I reviewed the edits: I would have inserted “large differences in” at the point where I put an asterisk in the letter below.

If you get the paper version of the Post, the letter is on page A27; if you have an online subscription, it’s here (or see my letter and another one by clicking the headline below). They changed the title of Trent’s article after the online version was published to what you see above.

The letter:

The Oct. 19 front-page article “A categorical no to the concept of race” argued that human “race” is a social construct without biological meaning. But there is important scientific data showing that race is indeed associated with diagnostic and useful biological differences.

Scientists have long rejected the simplistic view of races as groups distinguishable by appearance, geographic origin and * genes. But rejecting this view of “racial categories” and arguing that humans “share 99.9 percent of their DNA,” misses important genetic differences between populations. In fact, while humans share 99.6 percent of their genome, our genome has more than 3 billion base pairs, leaving more than 12 million DNA sites that vary among people.

This variation is correlated with ancestry and geography. It is used, for example, by genealogy services to tell people about their ancestry. In forensics, it’s used to identify criminals and bodies. And DNA variation is used to map the location of genes causing disease, an effort with great medical promise because the frequency of genetic diseases such as schizophrenia varies among populations.

Jerry A. CoyneChicago
The writer is emeritus professor of ecology and evolution at the University of Chicago.

There was also another letter making a different point about racial designations, and I’ll add that, too:

The front-page article “A categorical no to the concept of race” explained how treating race as objective rather than socially constructed has led to demographic confusion while shoehorning people into categories at variance with how they view themselves. As an example, the article mentioned the recent custom of uppercasing “Black” and “White” in American news media. The Post ought to champion a more nuanced standard, perhaps by using lowercase for how people (whose self-identification might be unknown) are viewed by others and uppercase for how they view themselves. The majuscule would then carry the same connotation of intentional membership as it does in the distinction between “Republican” and “republican” or between “Deaf” and “deaf.”

Charles H. BennettCroton-on-Hudson, N.Y.

Thanks to a reader who encouraged me to write my letter.

41 thoughts on “My letter to the Washington Post on race

  1. Social justice warriors get it exactly backwards. They say that race is biologically a fiction, but socially it should be the most important thing to care about in a person. The truth is that race is of considerable biologic significance, but should be the least important thing to care about in a person.

    1. Well-stated indeed! One of the many absurdities of this obsession, amongst the “woke-Jihadis/woke-Jacobins” with the racializing of everything, is that striving for “color-blindness” (as in “content of character” being more important than “color of skin”) is now regarded by many, if not most, of the “wokerati” as irredeemably racist.

    2. Well said. The “backwards” insight parallels my thoughts. There is fodder for a Dr. Suess story here.

      1. Just a spontaneous thought :

        Sex can be discerned in viewing faces. It’s quite robust, but not perfect.

        Race can be discerned with viewing faces as well.

        I find that interesting, somehow – that these two properties/subjects are of such extraordinary interest to woke adepts.

  2. Glad to see this – I’m still sorting through this, but I didn’t see a definition of “social construct” from Sydney Trent – and I don’t think PCC(E) relies on “social construct”, though I must read later (I’m in haste).

    IMHO “social construct” means that any thing or ideacan be claimed to be the product of society – such that different social inputs will give differently constituted things or ideas.

    And yes, I draw this from James “Conspiracy Theorist” Lindsay’s New Discourses analysis/wokish dictionary, and yes, I see “social construct” as wokecraft.

    1. Also “social construct” makes possible hermetic alchemy to exert control over anything. “Man” can then make society, which can then make “man”.

      The Ouroboros symbolizes this – Marx wrote about “man” making “man”. I.e. I’m using that old language of “man”.

    2. Every word used is a social construct. The Romans used “V” to represent the number “five.” We use “5.” There is no reason why five can’t be represented by “23” or “m” or “#.” Society decides what symbol to use. To dismiss ideas on the basis that they are “social constructs” is postmodern inanity.

      And no, objectivity and social constructs are not at odds with one another. “Sound” was once defined as “that which is heard.” Under that social construct the question “Does a tree make a sound when it falls with no one around to hear it?” was a conundrum. After the famous experiment in which an electric bell in a glass jar failed to make a sound when the air was removed, “sound” was re-defined to mean “vibration in the air.” An objective social construct replaced the subjective. And yet postmodernists insists there is no objectivity. More inanity. And a bias against science.

      Postmodernists see society as a constant struggle for power. But it is inane to think all power bad. The mother telling her children “Don’t run with scissors!” expresses power. Are her children obligated to call her an oppressor? Get real!

      1. Isn’t the designation of a sea vs an ocean, or pond vs lake, a social construct; in the sense that an arbitrarily arranged measurement for size/volume is applied? Is there some sophistry involved if someone were to say that “lakes are a social construction,” and therefore the term is not useful for discussing geography?

        1. I wouldn’t call it sophistry after I learned about the dialectic, gnosticism, and Hermeticism.

          The dialectic works here by taking that mundane thing we all understand – society sort of generally produces knowledge – and adds some thing to it – alchemically – to make the gnostic claim of a deeper insight that the mundane interpretation misses.

          The alchemy of the “social construct” then depends on something – hmmm, what is that thing? It is the gnostic themselves who are making the claim.

          Thus, the dialectic has given power to the gnostic, over people without the “insight” (which of course is a conjuring trick).

          Is my take, anyway.

      2. What you are pointing to is the difference in what structuralist call “signifier” (word) and “signified” (thing pointing to the word).

        Or as Shakespeare put it in Romeo and Juliet, “A rose by any other name would smell as sweet”. Meaning that changing signifier does not necessarily lead to a change of signified. But post-modernists think it does.

    3. Did you perhaps mean to write

      IMHO “social construct” means any thing or idea that can be claimed to be the product of society

      ? Because that would actually be correct.

  3. If race was a purely social construct without reality or utility, then it would be legal to choose any racial identity. If gender choice is allowed, why not race?

    1. Because society would have to construct it correctly – which is to illustrate the incoherence of “social construct”.

  4. Wouldn’t “imagining a future where categories don’t matter” involve imagining people saying “I don’t see race” and judging people on the “content of their character?” I did notice that the author of the article scornfully dismisses the “rhetoric of social justice advocates,” so the apparent contradiction with Critical Social Justice Theory might be intentional.

    If, in a desire to smash all categories, those who insist that sex is a social construct are jumping on board the view that race is also a social construct, they probably need to rethink that. It would allow people to identify in and out of races based on psychological and cultural factors. Even bringing that up is a high crime of transphobia.

    1. Your comment immediately struck me :

      Hegel used his alchemical dialectic to examine categories, such as apples – red, green – to find an essential appleness without (perhaps) Aristotelian accidentals (I’m still trying to understand that part).

      Queer Theory is obsessed with this in the modern era, seeking to obliterate categories, to reject discernment – because discernment is what makes reality comprehensible. That somehow obstructs “a identity without an essence” (D. Halperin).

      Ok that took longer than I thought, but TL;DR :

      hermetic alchemy, Hegel, the dialectic, same in kind – different in degree.

      as above, so below

  5. As a small tangential PS, the Metropolitan Police in London have stopped recording the ethnicity of drivers who are stopped by officers, just as other police forces start doing it.

  6. “…and so I prefer to use “populations” or “ethnicity”…” – J. Coyne

    Most scholars draw a distinction between race and ethnicity (ethnie).

    “The concept /race/ is distinct from the concept /ethnicity/. /Ethnicity/ is a cultural concept. Its conditions make essential reference to items such as language, nationality, religion, outlook, habits, norms, styles, or skills. The concept of ethnicity is also cultural in that human subjectivity figures constitutively in it: in contrast to the identity of a racial group, the identity of a “mature” ethnic group (an ethnic group that has a developed self-conception) is determined at least in part by its /members’ conception/ of what it is to be a member of that group.

    Although distinct, the concepts /race/ and /ethnicity/ do not differ in every respect. The concept /ethnicity/ resembles the concept /race/ structurally in making essential reference to ancestry. Each ethnic group shares a common ancestry.
    Inasmuch as the concept of ethnicity includes the concept of ancestry, it contains a biological “moment.” We therefore cannot say flatly that the concept of race is biological and the concept of ethnicity is not. Ethnicity is a biocultural notion. Race is a noncultural biological notion.
    In contrast to /race/, the concept /ethnicity/ makes no essential reference to patterns of visible physical differences. There is no tension in the idea that individual members of two distinct ethnicities may be visually indistinguishable.”

    (Hardimon, Michael O. /Rethinking Race: The Case for Deflationary Realism./ Cambridge, MA: Harvard University Press, 2017. pp. 40-1)

    “[W]e arrive at the following definition of the term /ethnie/: ‘a named human population with myths of common ancestry, shared historical memories, one or more elements of common culture, a link with a homeland and a sense of solidarity among at least some of its members.'”

    (Hutchinson, John, and Anthony D. Smith. /Ethnicity./ Oxford: Oxford University Press, 1996. p. 6)

    “[A]n ethnic group is a community formed by common descent and sharing cultural features (language, religion, etc.) that mark it off from neighbouring communities.”

    (Miller, David. /On Nationality./ Oxford: Oxford University Press, 1995. p. 19)

    1. In light of the above, plus other historical facts about the concept of “race”, it’s open to argue that ethnicity is real but race is not. If a concept from its very beginning has been weighed down by too many incorrect assumptions, it is perhaps better discarded in favor of other ways of talking. I suggest that this applies to race (as construed from about the 17th century onward).

      1. True enough, except that activists in non-white races don’t want to stop talking about race because it gives life to their “anti-racism” (i.e., anti-white) political agenda. Their epithets of “Oreo”, “banana” and “apple” accusing betrayal would have no sting if race stopped being a thing.

  7. I think we should just stick to the word “race” for race, since that’s what everyone uses (and, as above, “ethnicity” strictly means something different than just shared-inheritance clusterings). Yes, races are not clearly distinct and not discrete (since it’s a branching pattern, and groupings are fuzzy edged since humans interbreed), but that’s never been central to the concept of “race”. And yes, there’s then a layer of social construction on top (the American concept of “Hispanic” is more social than biological).

    The early scientists were far more sensible on this than they’re usually given credit for. For example, Blumenbach (1752 to 1840) was the first to scientifically categorise humans (inventing the term “Caucasian” and others), but he’s quoted as saying:

    “All national differences in the form and colour of the human body [. . .] run so insensibly, by so many shades and transitions one into the other, that it is impossible to separate them by any but very arbitrary limits.”

    [He was of course, a product of his time, for example thinking that Noah’s Ark landed somewhere in the Caucasus mountains and hence that Caucasians are closest to those who stepped off the Ark; but overall he got a lot right regarding “race”.]

  8. I do realise there are races, this is obvious to anyone not blinded by ideology. You don’t really need biological studies (worthy as they are) there.
    I did my best to blur it though. All of my four children are of either Asian or African descent (and my Caucasian input, of course).
    In South Africa, particularly the Western Cape, there is no way to determine race in at least half (a low estimate) of the population. There is Caucasian, Khoe, Malasian and Bantu inputs.

    1. Yes – and nobody can choose their life – this is the Geworfenheitflungness of Heidegger. This is the condition Rousseau meant by (paraphrased) “Man is born free, but everywhere we see him in chains” – the chains, or prison inscribed on our bodies (Butler) sought to be destroyed and escaped by Hegel’s dialectic – Hermetic alchemy.

      THOSE are the centuries-old ideas that “social construct” is riding on. Dawkins referred to this in a recent podcast, that we DO choose our “race” already, by conventional reproduction (to put it biologically).

      … ok sorry, I just get in an excited state by this…

  9. I had read that article and was stunned by the fact that biology barely entered the picture.

    The question I have is this: Does race have anything to do with skin color and does skin color have anything to do with genetics?

      1. Thank you for response. It’s astounding, really, that something so obvious is being deemed as non-existent. I wonder how far a society can wonder from reality before reality comes knocking at the door.

  10. I had heard that the meaning of “race is an unscientific concept” is that while genes for someone’s appearance have certain geographical clusters, genes for some other aspect of their body might cluster in a different way.
    People whose ancestors are from a region where a lot of wheat and other gluten grains are grown are less likely to have celiac disease, for example. So those genes are distributed according to the traditional agricultural practices in the area.
    There are different selection pressures on different aspects of the person, resulting in different patterns of geographical variation.
    So, the argument goes, race refers specifically to someone’s appearance, therefore our concept of race singles out only a small part of their genetics – but a socially important part.

    1. Although, people tend to interbreed with others of a similar appearance, so the genes that determine appearance would likely be especially important for the clustering of people’s genetics.

      1. I suspect that most of the DNA variation that differentiates human groups has nothing to do with appearance. In fact, it’s likely that most of the genes are nonfunctional DNA that differentiated through genetic drift. All that’s required for genetic differentiation is physical isolation of populations, and that can lead to differential selection but also differential genetic drift. The nucleotides used by 23andME, for instance, likely have nothing to do with physical appearance, much less any physical trait. They are CORRELATED with physical traits, of course, because genes differentiate in a correlated way.

  11. Your letter is very good, but there’s one point it overlooked that I think is of great importance. It’s about this statement: “In 2003, the completion of the Human Genome Project — which found that humans globally share 99.9 percent of their DNA — gave waste to the notion of “race” among the vast majority of scientists.” This statement about the Human Genome Project is referring to something specific, and this statement also has a specific history, which I’ve researched.

    First, here’s the original piece of research that this and similar statements have been based on. In 2004, the National Human Genome Center at Howard University held a conference to determine what the Human Genome Project revealed about the meaning of race in the genomic era. This conference resulted in a team of experts in sociology, anthropology, history and genetics assembling an authoritative set of statements on what racial divisions mean. The statements were published as a special issue of Nature Genetics in October 2004.

    Francis Collins, the director of the Human Genome Project, gave the following summary in this paper:

    “Well-intentioned statements over the past few years, some coming from geneticists, might lead one to believe there is no connection whatsoever between self-identified race or ethnicity and the frequency of particular genetic variants. Increasing scientific evidence, however, indicates that genetic variation can be used to make a reasonably accurate prediction of geographic origins of an individual, at least if that individual’s grandparents all came from the same part of the world. As those ancestral origins in many cases have a correlation, albeit often imprecise, with self-identified race or ethnicity, it is not strictly true that race or ethnicity has no biological connection. It must be emphasized, however, that the connection is generally quite blurry because of multiple other nongenetic connotations of race, the lack of defined boundaries between populations and the fact that many individuals have ancestors from multiple regions of the world.”

    Here is a quote from another paper, by Lynn B Jorde and Stephen P Wooding, published in the same journal issue:

    “Genetic variation is geographically structured, as expected from the partial isolation of human populations during much of their history. Because traditional concepts of race are in turn correlated with geography, it is inaccurate to state that race is “biologically meaningless.” On the other hand, because they have been only partially isolated, human populations are seldom demarcated by precise genetic boundaries.”

    These aren’t difficult to understand statements. But most lay people don’t get their science information directly from academic journals, so it’s up to journalists and science educators to accurately communicate these conclusions to the public. The first major newspaper to report on this journal issue was the Times of London, which covered it in an article titled, “Gene tests prove that we are all the same under the skin”. The article in the Times said the following:

    “The popular notion that skin colour can indicate physical or mental differences between groups of people has been demolished by a new analysis of the human genome, which declares race to be a biologically meaningless concept.”

    Notice the difference. The Nature Genetics papers concluded that “it is inaccurate to state that race is ‘biologically meaningless'”, but what was reported in the Times is that this journal issue “declares race to be a biologically meaningless concept”. For the remainder of 2004, most of the subsequent media coverage of this journal issue was based on the Times article, rather than on the original papers.

    If you do a search at Google or Google books for “Human Genome Project” combined with the phrase “biologically meaningless”, you can find dozens of more recent publications, including some from academics, repeating the London Times’ inaccurate summary of this journal issue. These later summaries are usually cited to secondhand, thirdhand, of fourthhand summaries of the Nature Genetics papers. It can take a lot of work to follow the chain of citations back to the original papers these statements are purporting to summarize, but I’ve done that several times, and I can provide examples here if other people want them.

    Over the past 19 years, it has gradually become “common knowledge” that the Human Genome project reached this conclusion, and the author of the recent Washington Post article might not have even been aware of what set of papers this statement was originally purported to be based on. I’ve seen creationists do something similar to this, in which they repeatedly cite their own inaccurate summaries of a paper, gradually becoming further and further removed from the original paper and forgetting what it actually said. These summaries of what the Human Genome Project concluded about race are possibly the most severe non-creationist example I’ve seen of the same practice.

    1. Thank you for the comment. Very enlightening two points, though. You take my letter to task for overlooking a correction of one of Trent’s statements. However, I could not add any material to that letter, which was already over length and edited heavily by the newspaper. It’s not surprising that Trent didn’t know this, as she is not even a science journalist but a “social issues” reporter.She was clearly way out of her depth with this one.

      Futher, your comment is an essay, not a comment: it is nearly 800 words long, while the suggested limit in the posting Roolz is 600 words (see rules on sidebar). Please keep comments shorter in the future, thanks!

      1. I’m not really faulting your letter for the fact that it didn’t mention this issue. I’ve also had letters published in newspapers and magazines, so I’m familiar with their length limitations. (One of mine was about a somewhat similar issue–I’m the author of the letter on this page that cites David Reich: ). I’m mentioning this just because I’ve been aware for a long time of how the conclusions of the Human Genome Project with respect to race have been distorted to something 180 degrees reversed from what they actually were, and this seems like something you should also be aware of.

        In the future, if there’s a piece of information related to one of your posts that I think you should know, but that’s too complex to describe it in 600 words or less, would it be better to email it instead?

  12. Eliminating racial categories on the census would be a disastrous idea. Race is strongly correlated with so many things, from economic status and school performance to health and incarceration, that eliminating the categories would seriously damage our ability to understand and change the world.

  13. The problem seems the conflation of race in the sense of empirical genetics with race in the sense of social role — in Hume’s terms, the difference between “is” versus “ought”.

Leave a Comment

Your email address will not be published. Required fields are marked *