Protests against Charles Murray inadvertently prove the points he made in The Bell Curve?

I’m wondering if “Protesters Disrupt Speech by ‘Bell Curve’ Author at Vermont College” (nytimes) inadvertently proves Charles Murray correct. Here’s the summary of The Bell Curve from the best minds of American journalism circa 2017:

Hundreds of students at Middlebury College in Vermont shouted down a controversial speaker on Thursday night, disrupting a program and confronting the speaker in an encounter that turned violent and left a faculty member injured.

Laurie L. Patton, the president of the college, issued an apology on Friday to all who attended the event and to the speaker, Charles Murray, 74, whose book “The Bell Curve,” published in 1994, was an explosive treatise arguing that blacks were intellectually inferior to whites because of their genetic makeup.

I listened to an abridged version of the book about 13 years ago and noted in a posting that the book was not in fact about race. The Wikipedia page on the Charles Murray contains the following summary:

The Bell Curve: Intelligence and Class Structure in American Life (1994) is a controversial bestseller that Charles Murray wrote with the Harvard professor Richard J. Herrnstein. Its central point is that intelligence is a better predictor of many factors including financial income, job performance, unwed pregnancy, and crime than one’s parents’ socio-economic status or education level. Also, the book argued that those with high intelligence (the “cognitive elite”) are becoming separated from the general population of those with average and below-average intelligence, and that this was a dangerous social trend. Murray expanded on this theme in his 2012 book Coming Apart.

Much of the controversy erupted from Chapters 13 and 14, where the authors write about the enduring differences in race and intelligence and discuss implications of that difference. While the authors were reported throughout the popular press as arguing that these IQ differences are genetic, they write in the introduction to Chapter 13 that “The debate about whether and how much genes and environment have to do with ethnic differences remains unresolved,” and “It seems highly likely to us that both genes and the environment have something to do with racial differences.”

In other words, you have to get through 12 chapters of a 22-chapter book before coming to the two chapters that the New York Times says the book is about! (And, in any case, the Bell Curve authors were mostly citing research on the Heritability of IQ, not conducting it.) Certainly the New York Times focus on “blacks were intellectually inferior to whites” says more about the New York Times than about Charles Murray. The chapters in question could just as easily have been summarized with “whites were intellectually inferior to Asians.” Somehow the paper that styles itself the Great Friend of the Colored Races can’t resist running story after story about how Americans with darker skin do poorly in school (see New York Times to employers: Toss resumes from applicants who went to school in poor neighborhoods).

What the book is actually about seems consistent with our Age of Rage (concerning inequality). As Wikipedia notes:

those with high intelligence (the “cognitive elite”) are becoming separated from the general population of those with average and below-average intelligence, and that this was a dangerous social trend.

If you like to fret about inequality, the sidelining of less-than-brilliant workers in favor of robots, etc., why wouldn’t you love Charles Murray? I’m wondering if the incident at Middlebury College shows that America’s brightest millennials are unable to read a book before showing up to protest. Then of course we have the fact that their counterparts in Asia spent those hours studying. In my own 2004 posting I said that the silver lining of the otherwise depressing book was that old people like me might still be able to find work because “if the book is right most [future young Americans] will be dumb as bricks.”

Related:

26 thoughts on “Protests against Charles Murray inadvertently prove the points he made in The Bell Curve?

  1. “Hundreds of students at Middlebury College in Vermont shouted down a controversial speaker on Thursday night,”

    So long as the campus administration (i.e. bureaucrats) do not call out the riot squad to stop the violence it will continue. The radicals got exactly what they wanted: they shut up people who disagree with them. I think the college administrators sympathize with the radicals so they will not take any effective steps to stop them and any opposing voices to the left will be silenced.

    Look up the speech by Stanford provost John Etchemendy on the left wing monoculture at colleges.

  2. As I understand it, the students also attacked a professor (NO, NOT A PROFESSOR?!?) and bashed the escape car. All just some fun for the antifa lads, officer. Oh what’s that, I wasn’t looking (wink).

    This thing with private death squads, er police forces, is an interesting quirk of American universities. Also the private amateur courts, even for crimes that would lead to prison in the outside world. Someone could probably write an interesting book about that.

  3. Siddartha Mukhergee, in his current book “The Gene, An Intimate History” concludes that “intelligence”, “g”, and “IQ” (the ratio of intelligence to age where correspondence to age equals 100 and a tested intelligence of , say, 106, implies 6% higher than the average of the age group) are HERITABLE but not DIRECTLY INHERITABLE (passing from generation to generation intact).

    Single-gene defects like sickle-cell anemia disease and Tay-Sachs (prolific in Afro-Carribbeans / Indians and Ashkenazi Jews respectively) are legitimately “genetic diseases” and can predicted by analyzing an individual’s genome.

    The various “intelligences” are products of the structure and context of the tests used to measure them and the results among races are almost meaningless, especially in severely impoverished populations.

    Measuring intelligence is a cultural construct much too complex to correlate with a gene or combination of genes. There is much more genetic diversity WITHIN races and tribes than BETWEEN races and tribes. The nature of evolution is that “genetic variations antedate the separation into continents, and perhaps even the origin of the species, less than half a million years ago…there has therefore been too little time for the accumulation of substantial divergence.”

    It’s difficult to summarize; read the book if you can. It is much more contemporary science than Murray’s work.

  4. IQ is of course the third rail of America — the reason that all men are not created equal, at least in terms of intelligence. Until recently this didn’t particularly matter — roads to be built, rails to be laid, machines to be operated. But now a lot of that is mechanized and IQ matters. The other Donald, supra. has some gibberish about why IQ is learned (and helpfully capitalizes some words so those of us reading impaired can understand what he has to say) — but i am dubious. The eagerness of him, people like him and Dr. Mukhergee to reach a conclusion on not a lot of evidence makes me suspicious — that their objective is to reach a certain conclusion rather than to test reality in an objective way. If Dr. Mukhergee said the contrary he would be unwelcome in the pages of NYT and the New Yorker and be hounded the same way as Mr. Murray. Maybe a democracy like the US can’t look at these issues truthfully — because if the answer is all men are not created equal then where do you go and what do you do?

  5. @the other Donald

    “There is much more genetic diversity WITHIN races and tribes than BETWEEN races and tribes.”

    Right, in dog breeds, 65% of genetic variation is withing a breed, therefore, there’s very little difference between a Great Dane and a Chihuahua according to your logic, it’s merely a social construct. By the same token, there are really no cat breeds where intra/inter genetic variation is similar to those of humans: 84% vs. 16%.

  6. My impression of the IQ discussion is that intelligence is largely heritable (correlation 0.6-0.8; wikipedia cites 0.75); that it consists of a single factor, Spearman’s g (somewhat surprisingly to me); and that g can, if so desired, be tested in culturally invariant ways, e.g., Raven’s matrices from the 1930s.

    Also, other aspects of personality, for example conscientiousness, are also heritable at about 0.5 (wikipedia). That means educational attempts to teach or rely on ‘grit’ are likely to have less effect than hoped for.

    Finally, there is also the fact that even if two samples or populations overlap, this does not imply that they are indistinguishable. But that I learned from the undergrad statistics course.

    https://en.wikipedia.org/wiki/Heritability_of_IQ
    https://en.wikipedia.org/wiki/G_factor_(psychometrics)
    https://en.wikipedia.org/wiki/Raven%27s_Progressive_Matrices
    https://en.wikipedia.org/wiki/Big_Five_personality_traits

  7. Talk of DNA and genes in the context of race is almost a red herring. Objective phenotypic measures obviously cluster and are obviously highly heritable. You don’t need DNA or genomics for race to be real and useful. The genomics people have proven to be surprisingly useless for any practical application over the last 20 years despite billions of dollars in funding.

    Talking about intelligence as a social and cultural construct is silly because we live in a particular culture where the metrics we use are clearly valid. We live in a world where ability to do mathematics and understand complex texts is rewarded, and that’s what our IQ tests measure. There might be some more valid measures of intelligence in a very different culture, but that doesn’t matter because we don’t live in it.

  8. ukie07: “therefore, there’s very little difference between a Great Dane and a Chihuahua according to your logic”

    The last time I checked, humans aren’t breed like Chihuahua.

    If there’s a difference in average intelligence between races, the difference is small (vastly smaller than the average body-size difference between Great Danes and a Chihuahua).

    And the bell-curves (population distribution) of different races largely overlap.

    What that means is that race is a poor predictor of the intelligence of an individual.

    Keep in mind that averages are a property of a population and averages are often mentioned without any sense of the variability (“standard deviations”).

    If you know that there’s a (small) difference in average intelligence between races, what do you do with that information?

  9. > if there’s a difference in average intelligence between races, the difference is small

    We have 100 years of data on this and it all says there is a persistent IQ gap of one standard deviation. You are welcome to argue interventions might close the gap but we’ve been trying that for many decades at great expense and it hasn’t worked. So “try harder/spend more” isn’t really a valid answer. I’m open to the idea that the gap could be narrowed over the course of another century or two, but it is simply not going to change much in my lifetime.

    Pretending there isn’t a gap is a huge problem. Everyone needs to be aware of this reality. When we deny the gaps, differences in outcomes are attributed to racism and poverty, when we know that’s not true. We waste billions and billions of dollars (trillions since the 60s?) on fixes we now know won’t work. Instead people need to make peace with the reality that different groups will have different outcomes in society.

  10. @davep

    “The last time I checked, humans aren’t breed like Chihuahua.”

    And yet, despite artificial selection pressures with dogs, the intra breed F_st (measure of genetic diversity) with dogs is still higher than the inter breed F_st. Humans are closer to cats in that respect where landraces (natural breeds) exhibit similar intra/inter genetic variation to those of humans. One may still argue that there’s no difference between an Abyssinian and the Norwegian Forest cat, based on his/her flawed interpretation of genetic variability (see Lewontin fallacy), but that would rather be a religious discussion rather than a scientific one.

  11. @bobbybobbob

    Well, 23andme results were quite accurate wrt my ancestry, so not quite as useless as one might imagine 🙂

    There are interesting results in for example evolutionary genetics, e.g. Neanderthal admixture, etc, that general public may not be aware of due to the politicized nature of the genetics related subjects and associated fear of some scientists to even engaged in this sort of research and then being ostracised (see the OP).

    Also, see the work of Cavalli-Sforza on related matters.

  12. We don’t have 100 years of data. What little genuine like-to-like comparisons there are show a much smaller than 1SD gap, sometimes no gap in some testing samples. And then there’s just obvious issues IQ fanatics dismiss like the wildly different look and behavior of black American kids who test “retarded” (<70 IQ) and white American kids who test "retarded".

    HBD, which is basically a weird religion around IQ testing and IQ testing as RPG (a lot of the "test results" are nonexistent suppositions that are at best derived from looking at countries where people were tested and going "I guess those guys down the road would test this way if anyone did it", and at worst just made up out of whole cloth), never considers the possibility that if we humans are "bio-diverse" racially, then the testing instruments devised by one racial group would of necessity be of limited utility across other racial groups.

    Not utterly useless, but probably best used as they were originally meant to be used– as part of individual tailoring of lessons for individual children.

  13. bobbybobbob: “We have 100 years of data on this and it all says there is a persistent IQ gap of one standard deviation.”

    There doesn’t appear to be a consensus about there being a “persistent one standard deviation gap”.

    Even so, that difference isn’t a reliable predictor of intelligence.

    You don’t teach an average anyway (an average is a property of a group, not an individual).

  14. We do have 100 years of good data from sources such as the US army, the bar associations, and so forth. Murray covers all this ground in The Bell Curve in the two chapters on race. I’m not talking about the hand-wavy international data sets; we have high quality data from American institutions.

    > the testing instruments devised by one racial group would of necessity be of limited utility across other racial groups.

    But IQ is not of limited utility. It has more predictive power than any other metric regardless of race.

    This willful self delusion would be cute if it didn’t have such serious ramifications. We are destroying institutions with affirmative action and destroying lives with lawsuits regularly because of these delusions. We can’t afford this anymore.

  15. bobbybobbob “But IQ is not of limited utility. It has more predictive power than any other metric regardless of race.”
    The IQ of an individual might have predictive power (I wasn’t talking about individual measurements). The average IQ of a race doesn’t have any predictive power for an individual of that race.

  16. davep: So the average retrieving propensity (“RQ”) of Golden Retrievers as a group doesn’t have any predictive power for a Golden Retriever puppy? We would expect the individual Golden to have the same RQ as a Chihuahua? Or we would expect all dogs, regardless of race (breed), to have the same RQ?

  17. “never considers the possibility that if we humans are “bio-diverse” racially, then the testing instruments devised by one racial group would of necessity be of limited utility across other racial groups.”

    On the contrary. Consider for example Raven’s Matrices. (Link above.)

  18. Group differences in humans are/were the product of natural selection, as much as in animals. The seemingly obvious idea seems to be revolting to many, thus the denial and substitution of faith for science, as in:

    “Descended from apes! My dear, let us hope that it is not true but if it is, let us pray that it will not become generally known.”

  19. @Ivan: And your point is…?

    I believe it would be more useful if could please provide a few extra links to substantiate your statements, since I’d really want more detail, **blush**

    It looks like I might need to learn basics of the subject (grad level math & statistics are OK). Thanks in advance!

  20. philg: “So the average retrieving propensity (“RQ”) of Golden Retrievers as a group doesn’t have any predictive power for a Golden Retriever puppy? We would expect the individual Golden to have the same RQ as a Chihuahua? Or we would expect all dogs, regardless of race (breed), to have the same RQ?”

    Why are people talking about dogs? I’m not talking about dogs.

    So, it seems you would expect that every individual black person that shows up for an interview must have an IQ that is lower by 1 standard deviation than every white person. Actually, you should waste you time interviewing white persons because Asians are smarter (on average). This is what your dog analogy implies.

    philg: “Or we would expect all humans, regardless of race (breed), to have the same IQ?”

    That would be racist. I know that, even within race, intelligence isn’t the same. It would be dumb to expect otherwise.

    It seems odd that you are suggesting treating individuals based on the intelligence you expect they have rather the intelligence they actually have.

    The basic failure is treating individual human beings based on a property of a population.

    No one cares if you do that to your dogs (or your slaves).

    There’s a basic principle of statistics (especially, as applied to humans) that you and others here don’t get.

  21. People are confused by dog breeding.

    If you selectively (actively) breed humans to be dumb (rejecting humans who happen to be born smarter than you like), you aren’t “expecting” them to be dumber. You are *choosing* them to be dumber.

  22. davep: I think you’re winning an argument against a straw man here. Nobody has suggested shaking hands with a stranger and saying “Because you appear to be of Race X, I expect that you must have IQ Y.” Least of all Charles Murray! The book was about long-term social trends for populations and sub-populations (i.e., he was dealing with a minimum of millions of people at a time).

    As with the New York Times (see original posting), might your focus on black Americans say more about you than about Charles Murray or the other folks who’ve commented here? There are plenty of other subpopulations to consider within the U.S. and around the world. Who else on this thread has expressed an interest in black American IQ other than yourself and the New York Times? (The Practical Conservative mentions black Americans, but in the context of correct testing; she does not express an opinion about what their actual IQ scores might be.)

  23. philg: “davep: I think you’re winning an argument against a straw man here. Nobody has suggested shaking hands with a stranger and saying “Because you appear to be of Race X, I expect that you must have IQ Y.””

    You did, pretty much directly:

    philg: “So the average retrieving propensity (“RQ”) of Golden Retrievers as a group doesn’t have any predictive power for a Golden Retriever puppy?”

    If dogs are not an analogy for humans, why are you (or others) taking about dogs?

    philg: “As with the New York Times (see original posting), might your focus on black Americans say more about you than about Charles Murray or the other folks who’ve commented here? “.

    You and others here use the words “race” and “breeds”. Maybe, you and others don’t know what those words mean.

  24. The assumption that Raven’s is “culture-blind” is itself a clear case of selection bias on the cultural front. As well, the idea that people who agree to join the Army are representative is an assumption. Perhaps valid, perhaps not, but it doesn’t mean you’re comparing like-to-like when you look at all people who wanted to join or who agreed to take the exam.

  25. IQ tests do not deal with literature and culture. As far as I recall they stress pattern finding, symbolic and geometric. What they do not evaluate is creativity, risk taking and drive.

Comments are closed.