Vox: Charles Murray is 80% right and we can't prove he's wrong about the other 20%

Here’s a 2017 blog post by me:

From Vox:

Charles Murray is once again peddling junk science about race and IQ

Podcaster and author Sam Harris is the latest to fall for it.

Updated by Eric Turkheimer, Kathryn Paige Harden, and Richard E. Nisbett May 18, 2017, 9:50am EDT

Eric Turkheimer is the Hugh Scott Hamilton Professor of Psychology at the University of Virginia. Twitter: @ent3c. Kathryn Paige Harden (@kph3k) is associate professor in the department of psychology at the University of Texas at Austin. Richard E. Nisbett is the Theodore M. Newcomb Distinguished University Professor at the University of Michigan.

Charles Murray, the conservative scholar who co-authored The Bell Curve with the late Richard Herrnstein, was recently denied a platform at Middlebury College. Students shouted him down, and one of his hosts was hurt in a scuffle. But Murray recently gained a much larger audience: an extensive interview with best-selling author Sam Harris on his popular Waking Up podcast. That is hardly a niche forum: Waking Up is the fifth-most-downloaded podcast in iTunes’s Science and Medicine category.

Getting worked up over Charles Murray being allowed on a podcast seems a little bizarre. (Here’s the podcast.)

Under the faux indignation and clickbait headline, however, this is about as good an attempt as any to shore up the Conventional Wisdom that the racial differences in average intelligence can’t be influenced by genetics at all. So I’ll go through a chunk of it, adding comments.

Interestingly, the article, when read carefully, is also about how Charles Murray is mostly so much more right than the Conventional Wisdom about IQ. But he’s still a Witch! The article is another one of these attempts to fight back against today’s rampant Science Denialism while not being accused of witchcraft yourself.

Here’s an important question: Do these triple bankshot approaches ever work?

They’re kind of like some prisoner of war being put on TV to denounce the Great Satan while blinking T-O-R-T-U-R-E in Morse Code? But what if nobody back home knows Morse Code anymore?

The basic problem is that the zeitgeist is continually dumbing down. We don’t worry about how to apply objective principles anymore to real world examples of human behavior, we just look for who are the Good Guys and who are the Bad Guys. And how can we tell? Just look at them: the cishet white males are the Bad Guys. What’s so complicated about that?

In this kind of mental atmosphere, will more than three Vox readers come to the end of this carefully coded article and say to themselves: “You know, Charles Murray is still as evil and stupid as I thought, but now I realize that most of what Murray says about IQ is Science and Good!”?

In an episode that runs nearly two and a half hours, Harris, who is best known as the author of The End of Faith, presents Murray as a victim of “a politically correct moral panic” — and goes so far as to say that Murray has no intellectually honest academic critics. Murray’s work on The Bell Curve, Harris insists, merely summarizes the consensus of experts on the subject of intelligence.

The consensus, he says, is that IQ exists; that it is extraordinarily important to life outcomes of all sorts; that it is largely heritable; and that we don’t know of any interventions that can improve the part that is not heritable. The consensus also includes the observation that the IQs of black Americans are lower, on average, than that of whites, and — most contentiously — that this and other differences among racial groups is based at least in part in genetics. …

(In the interview, Murray says he has modified none of his views since the publication of the book, in 1994; if anything, he says, the evidence for his claims has grown stronger. In fact, the field of intelligence has moved far beyond what Murray has been saying for the past 23 years.)

Eh … As I pointed out on the 20th anniversary of The Bell Curve, the world today looks even more like the world Herrnstein and Murray described.

The reality is that there haven’t been all that many revolutionary discoveries since then. The genomic research up through 2016 largely has panned out in the direction Herrnstein and Murray expected, although I’ve been told that a new preprint raises questions about Murray’s guess that the gene variants driving differences between the races are similar to the variants driving differences between individuals. If true, that would suggest that racial differences are in some ways more profound than Murray assumed, which would be ironic.

Turkheimer has gotten a lot of attention for a 2003 paper arguing that in one sample of poor people with lowish IQs, the heritability of IQ was lower than in better off populations, which is interesting but not hugely galvanizing. Emil Kirkegaard in 2016 asked “Did Turkheimer el al (2003) replicate?” I won’t try to adjudicate a question over my head.

But, anyway, the last big scientific finding to raise major questions about the Jensenist view was the Flynn Effect in the 1970s-1980s, which Herrnstein and Murray didn’t exactly ignore: they named it in The Bell Curve.

Murray’s premises, which proceed in declining order of actual broad acceptance by the scientific community, go like this:

1) Intelligence, as measured by IQ tests, is a meaningful construct that describes differences in cognitive ability among humans.

2) Individual differences in intelligence are moderately heritable.

3) Racial groups differ in their mean scores on IQ tests.

4) Discoveries about genetic ancestry have validated commonly used racial groupings.

5) On the basis of points 1 through 4, it is natural to assume that the reasons for racial differences in IQ scores are themselves at least partly genetic.

Until you get to 5, none of the premises is completely incorrect. However, for each of them Murray’s characterization of the evidence is slanted in a direction that leads first to the social policies he endorses, and ultimately to his conclusions about race and IQ. We, and many other scientific psychologists, believe the evidence supports a different view of intelligence, heritability, and race.

We believe there is a fairly wide consensus among behavioral scientists in favor of our views, but there is undeniably a range of opinions in the scientific community. Some well-informed scientists hold views closer to Murray’s than to ours. …

Let’s take Murray’s principles one at a time.

Intelligence is meaningful. This principle comes closest to being universally accepted by scientific psychologists. …

But observing that some people have greater cognitive ability than others is one thing; assuming that this is because of some biologically based, essential inner quality called g that causes them to be smarter, as Murray claims, is another. There is a vibrant ongoing debate about the biological reality of g, but intelligence tests can be meaningful and useful even if an essential inner g doesn’t exist at all.

Indeed. So what is the relevance of g to this debate?

The question of g is fascinating and also quite difficult. But it’s not absolutely relevant to this debate other than that poor Stephen Jay Gould got all hung up on g, fulminating: “The chimerical nature of g is the rotten core of Jensen’s edifice …”

As I’ve pointed out before, for example, Harvard requires applicants to take the SAT or ACT, both of which correlate considerably with IQ. The goal is to supplement the GPA with a measure that gives additional insight into brainpower. Say the g factor doesn’t exist and that there is zero correlation between an SAT math score and an SAT verbal score. Harvard would still favor students who score well on both measures over those who score well on only math or verbal. In the real world, there is a lot of correlation between SAT Math and SAT Verbal scores, just like the g factor theory implies. But, I suspect, we would still be having this IQ and Race debate if there weren’t.

Intelligence is heritable. To say that intelligence is heritable means that, in general, people who are more similar genetically are also more similar in their IQ. Identical twins, who share all their DNA, have more similar IQs than fraternal twins or siblings, who only share half. Half-siblings’ IQs are even less similar than that; cousins, still less.

Heritability is not unique to IQ; in fact, virtually all differences among individual human beings are somewhat heritable. … Heritability is not a special property of certain traits that have turned out to be genetic; it is a description of the human condition, according to which we are born with certain biological realities that play out in complex ways in concert with environmental factors, and are affected by chance events throughout our lives.

Okay!

This is a pretty funny example of the rhetorical strategy of much of this article. It’s designed to get readers to say to themselves: “That nasty moron Murray thinks the heritability of intelligence is partly genetic, when smart people know it’s really a … description of the human condition!”

An awful lot of this article consists of the three professors agreeing with Murray, but phrasing their endorsement of various Bell Curve assertions in such a way that Vox readers will think it’s actually a crushing takedown of Murray. The whole thing is full of these kind of trick maneuvers.

Do these kind of Secret Decoder Ring articles ever work? Does anybody ever finish the article and say to themselves, “Yes, Charlie Murray is just as evil and stupid as I previously believed, but now I’m aware that 80% of what Murray says about IQ is Science and Good!”

I dunno …

The basic problem is that the zeitgeist is just getting dumber and dumber as the dominant way of thinking gets more childish: Good Guys vs. Bad Guys. (And you determine who are the Good Guys and who are the Bad Guys not by something complicated like what they do, but by something simple: who they are.) So the likelihood of this kind of devious triple bankshot approach actually smartening people up doesn’t seem all that likely. But what do I know?

Today we can also study genes and behavior more directly by analyzing people’s DNA. These methods have given scientists a new way to compute heritability: Studies that measure DNA sequence variation directly have shown that pairs of people who are not relatives, but who are slightly more similar genetically

Such as members of the same race?

Much of the brain fog that besets Vox-level discussions of this question is due to Americans forgetting that race is deeply related to the question of who your relatives are. American intellectuals seldom think in terms of family trees, even though biological genealogy is just about the most absolutely real thing there is in the social realm. The simple reality is that people of one race tend to be more closely related in their family trees to people of the same race than they are to people of other races. But almost nobody notices the relations between race and genealogy in modern American thinking.

, also have more similar IQs than other pairs of people who happen to be more different genetically. These “DNA-based” heritability studies don’t tell you much more than the classical twin studies did, but they put to bed many of the lingering suspicions that twin studies were fundamentally flawed in some way. Like the validity of intelligence testing, the heritability of intelligence is no longer scientifically contentious.

In other words, “the heritability of intelligence is no longer scientifically contentious.” Nor is “the validity of intelligence testing.”

The new DNA-based science has also led to an ironic discovery: Virtually none of the complex human qualities that have been shown to be heritable are associated with a single determinative gene!

It’s almost as if the genetics behind the most complex object in the known universe, the human brain, are also complex.

There are no “genes for” IQ in any but the very weakest sense. Murray’s assertion in the podcast that we are only a few years away from a thorough understanding of IQ at the level of individual genes is scientifically unserious. Modern DNA science has found hundreds of genetic variants that each have a very, very tiny association with intelligence, but even if you add them all together they predict only a small fraction of someone’s IQ score.

And that fraction goes up year by year as larger and larger sample sizes are assembled.

The ability to add together genetic variants to predict an IQ score is a useful tool in the social sciences, but it has not produced a purely biological understanding of why some people have more cognitive ability than others.

Indeed, “it has not produced a purely biological understanding.” But the biological understanding is improving annually.

This is the usual debate over whether a glass is part full or part empty. What we can say is that each year, the glass gets fuller.

Most crucially, heritability, whether low or high, implies nothing about modifiability. The classic example is height, which is strongly heritable (80 to 90 percent), yet the average height of 11-year-old boys in Japan has increased by more than 5 inches in the past 50 years.

True. I write about height a fair amount in part because the effects of nurture on height are so clear. Thus, it’s plausible that the effects of nurture on intelligence probably exist too, even though they are hard to document.

As a non-scientist, I’m more of a nurturist when it comes to IQ than most actual scientists in the field. The scientists emphasize that that the half or so of the influences on IQ that aren’t nature aren’t what we normally think of as nurture, such as having a lot of books in the house growing up. Instead, what gets lumped under nurture appears to be mostly random bad luck that we don’t really understand.

But I’m more cautious on this than most researchers. I’m not convinced that they’ve figured out what drives the Flynn Effect over time, so I’ll hold open the possibility that more traditional nurture may play a considerable role.

But, please note, the Japanese remain one of the shorter nationalities despite a couple of generations of first world living standards. They’ve been surpassed in average height by the South Koreans, for example. The tallest Europeans on average include the wealthy Dutch and the much less wealthy Serbs, Croats, Bosnians, and Albanians. So, height differences among ancestral groups appear to be part nature, part nurture.

A similar historical change is occurring for intelligence: Average IQ scores are increasing across birth cohorts, such that Americans experienced an 18-point gain in average IQ from 1948 to 2002.

Indeed, the Flynn Effect is extremely interesting, as I’ve often pointed out.

And the most decisive and permanent environmental intervention that an individual can experience, adoption from a poor family into a better-off one, is associated with IQ gains of 12 to 18 points. …

There was a small French study of cross-class adoption with a sample size of 38. Despite the tiny sample, I find its finding that nature and nurture are about roughly equally influential (with nature a little stronger) quite plausible. (My general presumption before studying any interesting question is that we’ll end up around fifty-fifty.)

Race differences in average IQ score. People who identify as black or Hispanic in the US and elsewhere on average obtain lower IQ scores than people who identify as white or Asian. That is simply a fact, and stating it plainly offers no support in itself for a biological interpretation of the difference. To what extent is the observed difference in cognitive function a reflection of the myriad ways black people in the US experience historical, social, and economic disadvantage — earning less money, suffering more from chronic disease, dying younger, living in more dangerous and chaotic neighborhoods, attending inferior schools?

Okay, but let’s think about African-American height for a moment, since we were just talking about Japanese height. There’s this guy you may have heard of named LeBron James.

He’s really tall.

In fact, there are a lot of tall, healthy African-Americans currently dominating the NBA playoffs. In terms of height, African-Americans don’t appear to be a malnourished, beaten down population like, say, Guatemalan Indians.

Similarly, the last 72 men to qualify for the finals of the Olympic 100 meter dash, from 1984 through 2016, have been at least half black.

Now you could say, like James Flynn, that contemporary African-American culture is detrimental to the full development of African-American cognitive functioning, that black Americans focus too much on basketball and gangsta rap.

I think that’s highly possible.

But, who exactly is responsible for that? Charles Murray?

This is another triple bankshot approach: if we can just punch Charles Murray enough (metaphorically or literally), then inner city blacks will realize they should stop listening to gangsta rap and instead become patent attorneys. Or something.

… Race and genetic ancestry. First, a too-brief interlude about the biological status of race and genetic ancestry. The topic of whether race is a social or biological construct has been as hotly debated as any topic in the human sciences. The answer, by our lights, isn’t that hard: Human evolutionary history is real; the more recent sorting of people into nations and social groups with some degree of ethnic similarity is real; individual and familial ancestry is real. All of these things are correlated with genetics, but they are also all continuous and dynamic, both geographically and historically.

Our lay concept of race is a social construct that has been laid on top of these vastly more complex biological realities. That is not to say that socially defined race is meaningless or useless. (Modern genomics can do a good job of determining where in Central Europe or Western Africa your ancestors resided.)

And since “modern genomics can do a good job of determining where in Central Europe or Western Africa your ancestors resided,” they can, of course, also do the easier job of determining whether the bulk of your relatives were from Europe or sub-Saharan Africa.

However, a willingness to speak casually about modern racial groupings as simplifications of the ancient and turbulent history of human ancestry should not deceive us into conjuring back into existence 19th-century notions of race — Caucasoid, Negroid, Mongoloid, and all that.

Funny how the Obama Administration spent 8 years heartily enforcing policies based on categories called whites (i.e., Caucasoid), blacks (Negroid), and Asians (Mongoloid) and all that. It’s almost as if the Obama Administration believed that such categories are good enough for government work.

Murray talks about advances in population genetics as if they have validated modern racial groups. In reality, the racial groups used in the US — white, black, Hispanic, Asian — are such a poor proxy for underlying genetic ancestry that no self-respecting statistical geneticist would undertake a study based only on self-identified racial category as a proxy for genetic ancestry measured from DNA.

Okay, but the implication of that argument is 180 degrees backward from what Turkheimer et al are rhetorically implying. Isn’t it obvious that IQ studies that use self-identified race, as most do, are going to find a slightly lower correlation between race and IQ than ideal studies that use actual genetic ancestry?

For example, both Barack and Michelle Obama self-identified on the 2010 Census solely as black, but Barack clearly has a higher IQ than Michelle. The Vox authors in effect complain that studies based on self-identification would lump both together as purely black, ignoring Barack’s substantial white ancestry. That’s a reasonable methodological complaint, but its implications are the reverse of what they imply.

Similarly, there is an obvious correlation in the U.S. among Hispanics between white ancestry and educational attainment that gets blurred if you rely purely on self-identification.

Black Harvard professors Henry Louis Gates and Lani Guinier complained in 2004 that a very large fraction of Harvard’s affirmative action spots for blacks go to applicants, like Barack, with a white parent and/or foreign elite ancestry instead of toward genuine descendants of American slaves, like Michelle. (They sort of dropped the topic after the rise of Barack later that year).

Finally, the relationship between self-identification and racial ancestry has been investigated via DNA a lot recently, and the results are pretty much that, for whites and blacks, the government’s categories for self-identification are good enough for government work. In 23andMe studies, people who self-identify as non-Hispanic whites are overwhelmingly over 90% white by ancestry. People who identify as non-Hispanic African-Americans are largely at least 50% black.

23andme found among their clients, by my calculations:

If the average self-identified black is 73.2% black and the average self-identified white is 0.19% black, then the average black in America is 385 times blacker than the average white. That doesn’t seem very murky to me.

This was all predictable from the workings of the One Drop System.

Some of this will change in newer generations raised under somewhat different rules, but the basic reality discovered by genome studies is that in America, individuals who self-identify as non-Hispanic whites or as non-Hispanic blacks tend to be quite different by ancestry.

Genetic group differences in IQ. On the basis of the above premises, Murray casually concludes that group differences in IQ are genetically based. But what of the actual evidence on the question? Murray makes a rhetorical move that is commonly deployed by people supporting his point of view: They stake out the claim that at least some of the difference between racial groups is genetic, and challenge us to defend the claim that none, absolutely zero, of it is. They know that science is not designed for proving absolute negatives, but we will go this far: There is currently no reason at all to think that any significant portion of the IQ differences among socially defined racial groups is genetic in origin.

“No reason at all” is pretty silly. A much more reasonable suggestion would be that Occam’s Razor currently favors the hypothesis that some of the IQ gap is genetic in origin, but the subject is extremely complicated and it could turn out to be different.

It’s also possible that there is something we don’t understand at present about this dauntingly complex subject that makes a reasonably final answer not possible, a little bit like how Gödel’s incompleteness theorems came as a big surprise to mathematicians and philosophers such as Bertrand Russell.

In any case, we’ll learn a lot more about this subject over the next couple of decades due to the ongoing advances in genomics. …

But, right now, it sure seems like the wind has mostly been blowing for a long, long time in Murray’s direction and there’s not much reason to expect it to suddenly reverse in the future.

Toward the end of the Vox article:

Liberals need not deny that intelligence is a real thing or that IQ tests measure something real about intelligence, that individuals and groups differ in measured IQ, or that individual differences are heritable in complex ways.

But liberals must deny that racial differences in IQ could possibly be heritable in complex ways.

But isn’t the upshot of this article that Charles Murray is more correct than the Conventional Wisdom about 80% of what’s at issue?

Why isn’t this article entitled, for example: “Charles Murray is mostly right and Stephen Jay Gould was mostly wrong”?

And that leads to a meta-point: Instead of liberals attempting to imply, using all their rhetorical skills, that only horrible people like Charles Murray think there is any evidence at all for a genetic influence on differences in average IQs among races, shouldn’t they be spending more time explaining why, if Murray turns out to be right, that wouldn’t be The End of the World? Right now, we get told over and over about how unthinkable and outrageous this quite plausible scientific finding would be and how only bad people, practically Hitlerites, think there is any evidence for it at all.

This conventional wisdom strikes me as imprudent.

Personally, I think, this seemingly horrifying potential scientific discovery ought to be easily endurable, just as the NBA has survived the rise of the popular suspicion that the reasons LeBron James and other blacks make up most of the best basketball players include genetic differences.

I’ve long argued that The Worst that liberals can imagine about the scientific reality isn’t actually so bad. Murray’s world looks an awful lot like the world we live in, which we manage to live in. But I don’t have the rhetorical chops to reassure liberals that life will go on. I’m an official Horrible Extremist.

But that raises the question: Who does have the rhetorical skills to undermine the increasingly hysterical conventional wisdom and package the mature point of view about genetic diversity in the old soft soap that will go over well with Nice People?

Clearly, even Charles Murray doesn’t have the eloquence to reassure liberals.

Fortunately, there is this guy who is obsessed with genetic diversity in sports, having read David Epstein’s HBD-aware The Sports Gene, And he is really good at public speaking to liberals. And he doesn’t have that much else on his plate at the moment: Barack Obama.

So if Mr. Obama ever reads this, let me ask him to think about taking on the public service of deflating the Science Denialist hysteria over race and genetic diversity.

P.S. This article’s junior co-author, Paige Harden, had some more respectful things to say about Murray back in March.