The Journal of the American Medical Association has published a 6-page article about how to incorporate race and ethnicity into medical reporting. It’s not bad as far as it goes; in fact there’s only one thing wrong with it, but to me it seems like a big thing. Read by clicking on the screenshot below; you can also download a pdf file at the site.
They first define “race” and “ethnicity” by using the Oxford English Dictionary, which is the way I’d go about it. Here are their definitions:
The Oxford English Dictionary currently defines race as “a group of people connected by common descent or origin” or “any of the (putative) major groupings of mankind, usually defined in terms of distinct physical features or shared ethnicity” and ethnicity as “membership of a group regarded as ultimately of common descent, or having a common national or cultural tradition.” For example, in the US, ethnicity has referred to Hispanic or Latino, Latina, or Latinx people.
Although these definitions are overlapping, since they both incorporate people of “common descent”, one (or at least I) tend to think of “race” as the assertion, now known to be wrong, that humanity is divided up into a finite number of physically and genetically well-demarcated groups. (This claim historically went along with the assertion that there’s more genetic variation between “racial” groups than within those groups, we now know that that is absolutely wrong; within-group variation hugely predominates). Nevertheless, one can show by using data from many genes and gene sites, and clustering algorithms, that humanity can be shown to form genetic clusters that correspond to geography, which of course corresponds to evolutionary history.
But the issue is that there are clusters within clusters within clusters, and where you draw the line and say “this cluster” is a “race” is purely subjective. That’s why I don’t like the term “race”, as it’s too freighted with biological misconceptions as well as social assumptions and, of course, the use of “race” as a way to divide and rank people. .
“Ethnicity” is a different matter, as it’s not freighted, and although the definition above conflates ancestry with “cultural tradition”, they’re often connected. But for biological purposes I’d stick with ancestry, which of course refers to shared genes.
What I object to in the JAMA article is this sentence (I’ve put it in bold):
Race and ethnicity are social constructs, without scientific or biological meaning.
This is not so much flat wrong as grossly misleading. For example, I just cited the paper of Rosenberg et al., which shows that the genetic endowment of human groups correlates significantly to their geographical location (for example, if you choose to partition human genetic variation into five groups (how many groups you choose is arbitrary), you get a pretty clear demarcation between people from Africa, from Europe, from East Asia, from Oceania, and from the Americas. (To show further grouping, if you choose six groups, the Kalash people of Asia pop up). This is one reason why companies like 23 And Me stay in business.
This association of location with genetic clustering (and these geographic clusters do correspond to old “classical” notions of race) is not without scientific meaning, because the groupings represent the history of human migration and genetic isolation. That’s why these groups form in the first place. Now you can call these groups “ethnic groups” instead of “races”, or just “geographic groups” (frankly, you could call them almost anything, though, as I said, I avoid “race), but they show something profound about human history. The statement in bold above could be used to dismiss that meaning, which is why I consider that statement misleading.
As I said, there are groups within groups. Even within Europe, a paper by Novembre et al. reported, using half a million DNA sites, 50% of individuals could be placed within 310 km of their reported origin and 90% within 700 km of their origin.. And that’s just within Europe (read the paper for more details). Again, this reflects a history of limited movement of Europeans between generations. Finally, in terms of “self identification”, Tang et al., using just 326 markers, performed a genetic cluster analysis and identified four groups that matched nearly perfectly with the “racial” self-identification of people given four choices (white, African-American, East Asian, and Hispanic). Here’s what they found:
Of 3,636 subjects of varying race/ ethnicity, only 5 (0.14%) showed genetic cluster membership different from their self-identified race/ethnicity. On the other hand, we detected only modest genetic differentiation between different current geographic locales within each race/ethnicity group. Thus, ancient geographic ancestry, which is highly correlated with self-identified race/ ethnicity—as opposed to current residence—is the major determinant of genetic structure in the U.S. population.
That is, there is almost perfect correspondence between what “race” (or ethnic group) Americans consider themselves to be and the identification of groups using observed genetic differences. Because these are Americans, and move around more, the genetics reflect ancestry more closely than geography, though in Europe geographic origin is also important.
I needn’t point out that the morphological traits that we use to distinguish people from different areas also reflect genetic differences, including facial characteristics, hair color and texture, eye shape, and, of course, pigmentation. These are based on genetic differences. Of course this doesn’t mean there is a “Caucasian race” distinguishable by morphology, but simply that the way we have divided up humanity is not without biological meaning. (What these differences mean, and how they evolved, is of course, obscure.) Again, there is biological meaning in ethnicity, if you see ethnicity as reflecting groups having common evolutionary descent.
Finally, as we all know, different ethnic groups have different incidence of genetic diseases, and these reflect genetic differences among groups. West African blacks and their descendants in the U.S. are, for example, more prone to sickle-cell disease than those of other groups. Ditto for Tay-Sachs disease and Ashkenazi Jews. Of course these diseases are not found only within the ethnic groups, but there are significant difference in the incidence of diseases, and thus in the gene forms causing those diseases, among these groups. If you consider West African or Ashkenazi Jews as “ethnic groups”, which they are according to the definition above, then yes, ethnicity has a biological meaning, which reflects evolutionary history and common ancestry (if you don’t know the sickle-cell story, you should look it up).
The article goes into all kinds of nuances about how to report race (self-report seems to be the best way), and I don’t have much of a beef with their classification or how it’s to be used. EXCEPT that they recommend recording race not as a way to aid in diagnoses and genetic counseling (if two Ashkenazi Jews came to an obstetrician, she’d probably recommend they be tested to see if they were Tay-Sachs carriers, but she wouldn’t recommend the same for blacks) but as a way to ensure “equity” of treatment and accurate reporting of the incidence of diseases.
The article’s underlying rationale for recording race at all is that although (as they claim) race or ethnicity has nothing to do with biology, it does have to do with socioeconomic factors like racism, “disparities and inequities” and “intersectionality”, and those factors may play a role in disease. Note that class is not even mentioned, even though that surely plays a big role as well. But to ignore ethnicity except insofar as it (supposedly) closely correlates with health-related socioeconomic conditions is to not only overlook genetic data correlated with disease, but also to make unwarranted assumptions—that all members of an ethnic group are likely to share socioeconomic factors causal in disease.
I guess what bothers me the most about this article, besides the ignoring of genetic factors in favor of socioeconomic ones, is the claim that there is no biological significance of “race” or “ethnicity”. Depending on how you define these terms, that’s misleading. And if you use a “common ancestry” definition of either word, it’s just wrong. The claims that race and ethnicity are social constructs having nothing to do with biology overlooks a whole world of genetics, evolution, and demographics. It’s a phrase that hides what interests many of us about variation among groups in the human population. (For another take on genetic differences between groups that reflect evolution via different local adaptations, see this short note by Sarah Tishkoff, which shows that many interesting and important adaptations vary among ethnic groups.)
In other words, what bothers me is the idea, reflected in the statement in bold above, that all humanity is genetically the same. This is a mantra of much of the Left, reflecting a repugnance towards biological determinism and a sense that humans are almost infinitely malleable. But humans are genetically not the same, whether you’re talking about differences between ethnic groups or between men and women. It’s time that we face the data and admit this, and also realize that recognizing group differences is not at all the same as admitting that some groups are “better” than others. The assumption that the recognition of differences will automatically lead to ranking and then to bigotry is the mistaken conflation that produces articles like the one in JAMA.
As Richard Feynman said in another context (the Challenger disaster), “For a successful technology, reality must take precedence over public relations, for Nature cannot be fooled.” .
Flanagin A, Frey T, Christiansen SL, AMA Manual of Style Committee. Updated Guidance on the Reporting of Race and Ethnicity in Medical and Science Journals. JAMA. 2021;326(7):621–627. doi:10.1001/jama.2021.13304