How can you research the science of race when bad actors are eager to distort your conclusions?


There are interesting questions in the population genetics and evolution of different human groups, and it would be nice if there weren’t wretched ideologues who will happily misinterpret every difference between two groups of people, or even two people, to turn a description of differences into a ranking of superiority. It’s the Jordan Peterson problem of turning everything into evidence of a hierarchy.

Jedidiah Carlson provides some specific examples of how the right wing mangles research. It’s easy to see when the current fad is for murderous mass shooters to provide manifestos with their interpretation of the science; they are happy to name the credentialed scientists who provide fodder for their delusions.

The Buffalo shooter’s scientific bibliography has clear echoes to a similar citation scandal that arose in the late 1970s and early 1980s. During this era, the National Front (NF), a neofascist political party in the UK that had been steadily growing throughout the 1970s, distributed a series of pamphlets with articles referencing mainstream academic research. Their goal was to justify the organization’s platform of ethnic nationalism, white supremacism, and eugenics using contemporary science. The first wave of NF propaganda proclaimed, “scientists say that races are born different in all sorts of ways, especially in intelligence. This is because we inherit our abilities genetically.” Here, the NF cited the work of Hans Eysenck and Arthur Jensen, two of the most vocal proponents of the hereditarian theory that genetics could explain IQ differences between racial groups. Steven Rose, a champion of radical science and coauthor of Not in Our Genes with Richard Lewontin and Leon Kamin, lambasted Eysenck and Jensen in a 1978 letter to the editor of Nature, calling upon them to “publicly and unequivocally dissociate themselves from the National Front and its use of their names in its propaganda.” Eysenck and Jensen both complied with Rose’s request, albeit without a hint of apology for the societal harm their research precipitated. Eysenck asserted that he was “absolutely opposed to any form of racism” and claimed that “No-one familiar with Professor Jensen’s or my own writings could possibly misinterpret our arguments about the mean differences between various racial and other groups with respect to intelligence as implying the kind of policies advocated by the National Front.” Jensen echoed this self-absolving and patently false sentiment but also took the opportunity to lash out against his leftist critics for being, as he believed, as guilty as the far right in their desire “to promote and to gain public acceptance of a particular dogmatic belief about the nature of racial differences.”

That’s fascinating. Jensen actually tried to argue that oh no, he’s not a racist!, while producing some of the most outrageously bad pseudoscience defending racist discrimination. This is an ongoing problem in recognition, because it is common for racists to deny they are racist, while promoting awful garbage that they will never deny. As the Southern Poverty Law Center points out, “Jensen worked hard to develop a reputation as an objective scientist who “just never thought along [racial] lines,” and to portray critics of his racist conclusions as politically motivated and unscientific.” Right. That’s why he has a long entry at the SPLC.

Jensen is way, way out there, and it’s patently obvious that he was a screaming bigot manipulating the data to support an evil conclusion. But there have also been other scientists, less aggressive about their racism, who have been quietly smuggling bad science into the literature. How about kindly old Grandpa EO Wilson, who, after his death, was found to have been supporting all kinds of openly racist ideas? On the one hand, we’re supposed to objectively evaluate scientific ideas, but on the other, we’re supposed to somehow ignore the biased presuppositions that have led to those ideas, which makes no sense. People regarded sociobiology with suspicion when it first came out, because we were supposed to consider only the limited set of facts presented within it, but somehow we should overlook the fact that it quickly acquired a following among the worst kinds of people, the ones who wanted a racist conclusion and could read between the lines and see that sociobiology was a tool to reach that conclusion? Only racists are allowed to see the obvious interpretations, critics are “politically motivated and unscientific”, which provides a useful ratchet to make sure only the racist perspective gets widely disseminated.

So what do we do about subjects like sociobiology or evolutionary psychology, which promote, with the authors’ open consent and approval, bad ideas like genetic reductionism or determinism? I don’t know. I don’t like the idea of censorship, so perhaps a better idea would be if the various channels of scientific communication, the journals and blogs and so forth, were more proactive in rejecting work that is so clearly constructed around fallacious premises? Good luck enforcing that. The gatekeepers seem to have mostly bought into the bad ideas, since they’re typically privileged beneficiaries of the biases.

And then even work in which the authors were not advocating racism (near as I can tell) will be chewed up and twisted by malicious actors to arrive at a malicious conclusion. There’s no avoiding that.

Much of the scientific community’s outrage in the aftermath of Buffalo centered around the shooter’s citation of a paper colloquially known as the “EA3” study (Lee et al., published in 2018 in Nature Genetics). This study, carried out in over 1.1 million individuals of European descent, identified hundreds of genetic variants associated with “educational attainment” (often abbreviated to EA)—i.e., the number of years of school completed, often taken to be an “easy-to-measure” proxy for intelligence. The shooter’s reference to the EA3 study came in the form of a screenshot of a plain-looking document (figure 1) proclaiming, “The latest findings on genetics and intelligence show that biological factors contribute to the gap in intelligence between European and African populations.” Beneath this image, the shooter weighed in with his own interpretation, punctuating his earlier claims that “whites and Blacks are separated by tens of thousands of years of evolution, and our genetic material is obviously very different.”

Many variations of this table can be found throughout the internet, but the earliest version can be traced back to a thread on 4chan (an anonymous and largely unmoderated online forum) timestamped to September 15, 2018, barely six weeks after Lee et al. was published online (on July 31, 2018). The original post that initiated this thread (figure 2) is a perfect example of what sociologist Aaron Panofsky calls “citizen scientific racism”: an individual, having come across the EA3 study, collected the top EA-associated variants from a supplementary table of the paper, annotated these variants with the allele frequencies in European and African populations using publicly available data from the 1000 Genomes Project, and curated a set of EA-associated variants with the greatest differences in population frequency to argue that Europeans are genetically predisposed to higher intelligence.

The responses to this thread rapidly crystallized into a simple propaganda strategy: turn these “findings” into a standalone unit of easily-digestible visual information—or a meme, for lack of a better term—and let it organically spread across other online spaces. Shortly thereafter, another user took these suggestions to task and independently reproduced the original post’s analysis, presenting the results in a table similar to that shown above. Within hours, this image began to circulate in other 4chan threads and mutate into alternate versions, often accompanied by zealous calls for diffusing these memes throughout the internet. “SPREAD THESE IMAGES LIKE WILDFIRE,” encouraged one user. “This is the new IOTBW” said another, referring to the racist slogan, “It’s OK to be white.” The meme was even passed on to a cabal of popular alt-right bloggers and Youtubers who “have several PhDs and can give you a hand…plus they’re fantastic propagandists.” This collective enthusiasm for propagandizing the EA3 study appears to have been wildly successful. Altogether, variations of this meme have been posted over 5,100 times on 4chan and regularly appear on more mainstream social media platforms like Reddit, Twitter, and Quora. Contrary to the scientific community’s prevailing narrative that the shooter was an isolated extremist who happened to stumble upon the study,20 these data demonstrate that the EA3 study has been a significant force in empowering far-right extremists for years, virtually since the day it was first published.

(Note that Carlson article includes many figures that illustrate the point he’s making, but he’s flagged all of them with a “do not replicate” watermark. They often come from places like 4chan, so I agree, let’s not promote these vile sources.)

One step forward that Carlson promotes is the revitalization of activist-scientists. We need to speak up on all fronts, rather than passively sitting by while nonsense gets published in multiple outlets.

Weaponized science continues to threaten far more than the public image of scientific authority. Today, it has morphed and evolved to find new victims and modes of victimization, and exploits whatever platforms and resources are at its disposal to promote its message. Synthesizing the lessons learned from past radical science movements provides us with a path forward: our collective response to weaponized science must be fiercely multimodal and operationally diverse, taking place in the pages of scientific journals, the digital streets of social media, and the physical spaces of our institutions and cities.

He also gives us three challenges.

First, we must further educate ourselves on the ecosystem of weaponized science. Second, we must actively resituate our appetite for scientific progress towards the service and liberation of our communities. Finally, we must channel this knowledge and desire for change towards the development and implementation of creative strategies to disarm weaponized science, inoculate against its normalization, build resilience and solidarity, and spread those ideas like wildfire.

All right, I think I’ve been doing the first. I’m depressingly familiar with the bad science that gets published in all kinds of outlets. I’ve been involved in the second already, too, as one of those people who strongly believes that science should be serving a larger social purpose. The third…I’m not sure about what creative strategies I could implement, beyond just telling all of you what sucks about some of our modern science.

Comments

  1. says

    Let’s hypothesize for a moment that there is a gene complex that produces higher scores on IQ tests. That would be necessary for IQ test-based racism to work. So why hasn’t it been found? Wouldn’t that be within the capabilities of current DNA searching and statistics? Well, where is it?

    (I know IQ tests are bullshit, which is why I used that somewhat awkward construction)

  2. moarscienceplz says

    @Marcus,
    The evil libruls have suppressed that knowledge, along with proof the moon landings were faked and how to make engines that run on water.

  3. raven says

    How do they get from “Blacks have lower IQs” to “It is a good idea to go to a supermarket and shoot and kill Black people?”

    If we are going to shoot lower IQ people, half of all white people would end up dead.
    Because half of all white people have IQs that are…less than 100, which is the median.

    (This is a rhetorical question. I have no idea how their attempts at reasoning work or if they even are trying to reason things through.)

  4. says

    We do a GWAS, of course, and then we find not a locus, but a sprawling cloud of innumerable components that contribute in some small way to the score, and also have complex interactions with each other that make the specific contribution of each component uninterpretable.
    Then we publish a paper that says IQ is genetic.
    Easy.

  5. moarscienceplz says

    “Second, we must actively resituate our appetite for scientific progress towards the service and liberation of our communities.”
    I don’t really get how this is supposed to work. Is PZ supposed to stop studying spiders unless he can show his work fights human racism?

  6. robro says

    PZ @ #4 — You left out the part about “submit a proposal for a research grant” to some awful billionaire’s foundation looking for ways to justify his/her bigotry.

  7. birgerjohansson says

    Do not forget the correlation between leprechauns and unemployment (it played a major role in Oops, Apocalypse ).

  8. birgerjohansson says

    What we carelessly call intelligence has several subsets.

    There are plenty of anecdotes about Australian aborigines having a knack for orienting themselves and finding their way in unfamiliar places and towns.
    Living in the most hostile inhabited continent for 50,000 years with a flat featureless landscape – it is at least plausible that there has been a meaningful selective pressure…

  9. birgerjohansson says

    There is a non-human population with interesting properties: Bowhead whales can live for more than 200 years.
    Recently, at least one of the genes that contribute to their longevity has been identified.

    I suggest we do a bit of CRISPR on Bannon and the other crooks (please, please, include Farage and Boris) while we work out the details.
    The result may be long-lived ten-ton crooks that need to be housed in tanks like the Guild Navigators, but some failures among the way are inevitable.

  10. moarscienceplz says

    @#6
    Well PZ, your response makes it seem more like the answer to my question is actually ‘yes’, but I take your point that there are many angles. An astronomer would probably have a hard time designing a research project that supported antibigotry in and of itself, but they could try to look for non-white and non-male astronomers who could be collaborators, or otherwise find ways to amplify those astronomers.

  11. says

    Any IQ studies in the U.S. have to account for the lead poisoning that resulted from lead based paint deterioration in low income urban housing where the Black population was confined by redlining and racism. Known symptoms of lead poisoning as a consequence of lead chemically mimicking calcium in the brain include lower IQ, shrinking of the forebrain that controls inhibition, and damaging the myelin sheathing of brain neurons that produces a ‘buzzing’ brain. This produces the known symptoms of irritability, violence, and inability to learn among those lead poisoned at an early age. It has also been shown to be a cause of the high violence in the US from lead based gasoline that is now declining since it was banned. A large part of illegal drug use has been linked to using drugs to self-medicate the ‘buzzing’ brain from lead poisoning.

  12. david says

    As I read it, the EA3 study showed that there are genetic factors that predispose people to do well in a system that is racist, and which gives preferential treatment to people from groups that carry those factors.

  13. chrislawson says

    david@14 and others–

    The essential problem with EA3 and similar studies is that they crunch hundreds of thousands of gene snippets and correlate against dozens of socioeconomic variables. Of course there are going to be some strong matches just by chance. It doesn’t indicate anything about causation. And socioeconomic status is also highly heritable, which makes it even more ridiculous to pull out causal inferences. Then, as zoniedude points out, there are known non-genetic causes of IQ differences in socioeconomic and racial groups such as childhood lead exposure which cannot be accounted for in these study designs.

    We know from population-based studies that Poland has higher rates of several genetic conditions. Polish names also have higher rates of certain letters. Using the same logic as using EA3 data to suggest genetic causes for IQ difference, we could just as easily argue for a “letter Z theory” of male pattern baldness. Not picking on Poles here. The point is this strategy can be applied to any subgroup of people, even spurious ones like “people with prime number birthdates”. Just find two markers correlated with that subgroup and infer causation. The appeal to race and gender essentialists is obvious.

  14. microraptor says

    Raven @3: These groups are sustained by hate. If they somehow managed to kill all the black people off, they’d go out and find a group of white people that was “inferior” due to hair color or eye color or being too tall or too short or being liberal instead of conservative and then keep this up until they had to start turning on each other due to a lack of any external enemy they could blame.

  15. raven says

    In Realityland, if you want to look at racial/ethnic groups that almost always score low on IQ tests, that is…the Irish. Historically, the Irish have been the poster group for dumb people.

    https://www.theamericanconservative.com/raceiq-irish-iq-chinese-iq/
    Ron Unz

    First, Lynn was hardly unique among leading IQ experts in characterizing the Irish as being low IQ. For example, Hans Eysenck, one of the foremost IQ researchers of the 20th century said exactly the same thing in his 1971 book “Race, Intelligence, & Education,” claiming that the Irish IQ was very close to that of American blacks, and that the Irish/English IQ gap was almost exactly the same size as the black/white gap in the U.S., being roughly a full standard deviation. Eysenck’s stated position unsurprisingly caused a considerable furor in the British media, including all sorts of angry responses and even (facetious) threats of violence. So the huge and apparently well-designed 1972 study of 3,466 Irish schoolchildren which placed the mean Irish IQ at just 87 hardly seems an absurd outlier.

    Well, OK the Irish for many decades scored low on IQ tests.
    I guess for the racists, that means they can find Irish supermarkets and randomly shoot anyone that looks Irish.

    This rapid convergence between Irish and British IQs should hardly surprise us. According to the GSS, the Wordsum-IQs of (Catholic) Irish-Americans rank among the very highest of any white ethnic group, with a value almost identical to that of their British-American ethnic cousins.

    WAIT!!! Hold on here. Don’t start getting your AR-15s ready yet.

    It turns out that lately, Irish IQs have been rising and are now equal to the British. They are equal to WASPs in the USA.

    The point that can easily be made over and over again is that IQ is very changeable and mallable.
    It is effected by many variables such as socio-economic status, early childhood nutrition, early childhood upbringing, education quality, culture, and so on.
    We can say that there is a heritable component to IQ but it isn’t really all that high.

  16. raven says

    From the article above.

    One of the many surprises I’ve encountered when reading the dozens of web pages and many hundreds of comments attacking my Race/IQ analysis is the overwhelming focus of these critics upon my Irish data.

    Although I discuss similar ethnic IQ evidence regarding the Greeks, Balkan Slavs, Southern Italians, Dutch, Germans, and various other European peoples, it sometimes seems like the attacks on my Irish analysis are…

    At various times, European groups including the Greeks, Slavs, Italians, Dutch, Germans, Spanish, Irish, Portuguese, etc. have scored low on IQ tests and have been considered inferior humans.

    At other times and later on, these same groups have scored much higher than other groups on IQ tests.

    So what does this tell anyone?
    Collective IQs of human subgroups are very changeable and mallable and there is a very large environmental component to this score.

  17. chrislawson says

    raven@18–

    Best predictor of racial group held to be genetically stupid/violent/lazy = most recent immigration wave.

  18. lotharloo says

    WAIT!!! Hold on here. Don’t start getting your AR-15s ready yet.

    It turns out that lately, Irish IQs have been rising and are now equal to the British. They are equal to WASPs in the USA.

    I’m confused. Do we shoot all the British and white people now?

  19. erik333 says

    @1 Marcus Ranum
    Such structures would necessarily have to exist for human intelligence (or lack thereof) to be explainable by evolution. IQ test just don’t work.

  20. says

    …i.e., the number of years of school completed, often taken to be an “easy-to-measure” proxy for intelligence.

    Um, no, that’s not a proxy for “intelligence,” it’s a middling-to-good proxy for how much opportunity one had to learn certain things in a school setting. Which affects what one did learn overall, but which also has absolutely ZERO genetic component, and which should therefore have absolutely no say in determining any sort of genetic determinant of anything.

  21. cvoinescu says

    Marcus Ranum @ #1

    Let’s hypothesize for a moment that there is a gene complex that produces higher scores on IQ tests.

    erik333 @ #22

    Such structures would necessarily have to exist for human intelligence (or lack thereof) to be explainable by evolution.

    Clearly they exist, but they’re not the same thing. The set of variants that explain most differences in cognitive ability (or whatever flawed proxy you choose for that fuzzy, multifaceted trait) between present-day modern humans is likely very different from the set of variants that explain the differences between modern humans, on one hand, and archaic humans and/or other present-day species, on the other hand. By definition, the former would vary between present-day human individuals, but I’d expect the latter to be pretty much fixed (not variable in modern humans), with little overlap between the two sets.

    Regarding the set correlated with differences between individuals, not between species, I expect it to be a vast tangle of variants of very small and non-uniform effect, each of magnitude and even direction conditional on many other combinations of variants, overlapped by another set of what would be considered, to various levels of agreement, risk variants for various pathologies, from mild to almost lethal, with the latter being thankfully rare but with dramatic deleterious effects. What I don’t think exist at all are single variants, or small subsets of variants occurring together, with large positive effects on cognition.

  22. cvoinescu says

    Just to clarify: what makes you smarter than some other guy is not the same as what makes people smarter than chimps.