IQ, Clusters, and Francisco Gil-White
Tyler Cowen shares his thoughts on the idea that it is important to try to preserve the average level of a society’s IQ, as though this is some kind of precious public good:
I don’t assign special status to The Conservation of IQ for two reasons. The first is the Flynn effect, or the fact that measured IQs have been rising steadily over time. This implies some combination of a) IQ gains come naturally under conditions of progress, and b) IQ statistics are to some extent phony and don’t measure real intelligence. We can debate the mix, but either deflates fears that IQ is somehow especially scarce or endangered. These data also suggest that IQ is an artifice to be unpacked rather than a primary category.
Second, defenders of the IQ view tend to read evolutionary biology and intelligence research. My roots are in cultural history. Clusters of amazing achievement come and go pretty quickly, usually through some mix of environmental effects and luck. Look at Venetian painting. It was much better centuries ago, but I doubt if Venetian IQs have been falling. Once we see how such enormous differences can be explained by non-IQ factors, I again don’t obsess over the variable.
I find this pretty persuasive. The most stimulating thing I’ve read about IQ lately is the chapter on the topic in Francisco Gil-White’s webbed book on Resurrecting Racism: The modern attack on black people using phony science, which is mostly an attack on John Entine’s book Taboo: Why black athletes dominate sports and why we’re afraid to talk about it. I did not know, for example, that the IQ test was invented by Binet and Simon specifically as a test of the cognitive skills middle/upper-class French public schools were trying to develop in their students, and explicitly not as an assessment of raw, genetic, mental processing power. I also didn’t know that Sir Cyril Burt was a huge fraud. Gil-White’s theory an adapted cultural capacity that helps accounts for cultural and cognitive variation (a theory he shares with Boyd, Richerson, and Henrich, among others), seems to me to make good biological sense, to help make sense of the Flynn effect, and to pose a big problem for the Murray-Herrnstein Bell Curve thesis. Reading Gil-White changed my p on the existence of genetically inherited general intelligence from above .5 to below. I get excited when that happens!
Brief aside on clusters of achievement: I want better explanations! What accounts for 4th c. BCE Athens? Late 18th-early 19th c. Scotland? Early 20th c. Vienna? Philosophy mostly just coasts along on these efflorescence’s of genius. What’s going on? Why aren’t we having having one of these NOW, in the US? One idea is that our system of supporting our top intellectuals through the huge, geographically dispersed university system practically guarantees that no single place will develop enough of a critical mass of talent to create a world-historical outburst of brilliance. What if the world’s best economists, philosophers, psychologists, etc., were clustered in the same place in the same way the world’s best software engineers are clustered in Silicon Valley?
Fun facts about Gil-White: (1) he was canned from the Penn psychology department he says for his controversial views about recent history regarding several ethnic conflicts. I have developed no opinion on Gil-White’s opinions about these matters. I just know he’s a fascinating and intellectually solid theorist in biological anthropology. (2) He is the son of Francisco Gil Diaz, who was Secretary of Finance for Mexico under Vincente Fox.
March 1st, 2007 10:24
You may want to consider APA Task Force on Intelligence: Knowns and Unknowns before moving your priors very far.
March 1st, 2007 17:39
Will: Even if the book gives you some worries about the Binet test, don’t all the different IQ tests correlate with each other pretty strongly? If IQ is nonsense, why is it a better predictor of success than is GPA?
My prior on intelligence being primarily heritable (with the remainder mostly being random noise) is about 0.8. Will read the Gil-White chapter though.
March 1st, 2007 18:22
Ok, reading Gil-White.
1. Even if Binet is not cross-culturally comparable, there are intelligence tests that are culturally neutral: Raven’s progressive matrices, for example. I don’t think that we see big differences in results when looking at IQ as measured by those tests, and results from all tests correlate strongly with one another.
2. The article is almost libelous in its description of Jensen
3. There are many, many other studies of heredity and IQ — see the APA task force report noted by Brent, above. IQ really looks to be highly hereditary.
I’m really rather surprised that you so revised your estimates on the basis of the Gil-White paper. If anything, the weakness of the critique would make me revise my prior upwards rather than down.
March 1st, 2007 22:30
What of the possibility that these “clusters of achievement” are largely a construction of us looking back in history, and were not in reality as impressive as we remember them. It is after all very difficult to view intellectual/cultural history as a mass of interconnected ever slowly creeping transformations and strains of thought, and instead view history as a bunch of “this is doctrine X, it was invented by miraculous genius people in location G! things were the same for a while and then successor doctine Z was invented 200 years by later super awesome genius H!”.
Because thats how we learn the intellectual history only to find out later that all of these ideas had close antecedents in various texts written by authors we’ve never heard of in places we thought nothing was going. (Take as an example the Wealth of Nations in your Scottish Enlightenment case - I seem to be constantly reading that was Smith wrote was said earlier and better.)
I ain’t saying that these clusters are all imagination. There after all only a finite number of historical instances where societies have enough resources, access to prior materials, dumb historical luck in having their work preserved/appreciated. But maybe beyond those obvious factors most of the complications owes to stuff that happened after not during these historical moments
March 2nd, 2007 04:36
I also didn’t know that Sir Cyril Burt was a huge fraud.
Heritability of IQ according to Cyril Burt ~70%
Heritability of IQ according to the Minnesota Twins study ~70%
For a fraud, Burt’s findings were remarkably accurate.
March 2nd, 2007 11:48
Jesus H. Tapdancing Christ. This post proves that even the brightest bulbs flicker a bit sometimes. Will, just paying attention to Gil-White’s tone should set off alarm bells that he has a strong ideological bias on this subject. Tread very, very carefully.
Straight-up: Gil-White’s comprehension of modern psychometrics is just barely above the level of an ignoramus, he’s sometimes downright dishonest (e.g. in his treatement of Jensen), and his observations on the history of intelligence testing and behavioral genetics are noteworthy but do not speak to the current state of the science. Yes, Cyril Burt disgraced himself by making some shit up. But this is neither here nor there with regard to the genetics of intelligence except as a historical note. Gregor Mendel massaged his data pretty seriously, but this doesn’t falsify Mendelian genetics.
March 2nd, 2007 16:10
Gil-White’s clumsy hatchet job also fails to mention that Liam Hudson, a professor of educational psychology at Edinburgh University and one of Burt’s most ardent opponents, had Burt’s housekeeper destroy Burt’s papers almost immediately after Burt’s death, and that the “missing” research assistants have been found.
For more details, consult Ronald Fletcher’s _Science, Ideology, and the Media_ (New Brunswick, N. J.: Transaction Publishers, 1991), a book that concludes the case against Burt is “not proven”.
March 2nd, 2007 17:57
Jo, that depends on your standard of proof. Frankly the sheer improbability of it is enough for me to convict him of at least some fabrication. See David Burbridge’s recent paper for what I think is a pretty level-headed analysis from a guy who doesn’t have a horse in this particular race.
March 2nd, 2007 17:59
Oops, screwed that one up. Here’s the actual link.
March 2nd, 2007 22:59
Burt’s good name has been raked through the mire for the good part of a half a century now. Smearing the good name of a dead man has all been good fun but enough is enough. The “missing” researchers have been found. And other exculpatory evidence was destroyed under the guidance of one of Burt’s staunchest academic critics almost before the body was cold, an act that should strike even the most ardent Burt critic as … irregular.
I invite anyone who is tempted to join the hate Burt bandwagon to first read Fletcher’s book, along with a similar independent work by Robert Joynson.
Fletcher addresses the Arguments From Statistical Improbability in his book. And since Burt’s findings are consistent with the findings of other independent researchers with more robust data sets, the character assassination campaign that evidently continues to this day is moot anyway.
March 3rd, 2007 22:50
‘The Greatest Poet That Has (N)ever Existed’…
Will Wilkinson has an interesting post discussing a rather odd remark made by Charles Murray in a recent interview: “I……
March 4th, 2007 20:57
Variation in IQ is largely the result of either heredity or environment. If variation in IQ is largely the result of environment then we would expect to see a lower average IQ among the blind than among the sighted, after all the blind are deprived of an entire category of exposure to the environment. But in fact, this is not the case. The blind and the sighted have the same average IQ.
March 6th, 2007 03:29
I only looked at the last chapter, but I don’t see how you can take this guy seriously. It seemed to be all conspiracy theory and moral outrage.
March 10th, 2007 19:42
“Philosophy mostly just coasts along on these efflorescence’s of genius. What’s going on? Why aren’t we having having one of these NOW, in the US?”
They’re all studying neuroscience. Did you get the memo?
I really enjoy that word, efflorescence, but I don’t know if it’s a good way to visualize some of the history’s most successful ideas, for example Islam, Freudianism and Marxism (all of which have pretty winning but dubious track records at the start of the 20th century and are by no means squeezed today).
I’m only dropping these particular Ideas because, like neuroscience, they’re good at answering questions and they address man’s misery. Neuroscience will do that. And prove that we’re smarter than other people. And help us build robots that use guns responsibly.
March 14th, 2007 10:51
I fail to see the controversy. You can take any test, get a bunch of people to take it. Hopefully, it will generate a nice bell curve. Then you can say how far you deviate from the average. It could be controversial, if you think it measures more than someone’s ability to take a test and do good on it. I remember in college totally freezing up on a quiz and forgetting everything. Now if that were an IQ test, I’d be thrown in remedial education. Or if I was wealthy I’d attend a private college that allows one to have oral exams in addition to written ones and passed.
March 16th, 2007 09:23
You could have learned those things about IQ tests from reading Gould’s _The Mismeasure of Man_ many years ago. (There’s good reason to think the Minnisota Twins studies have more than a bit of book-cooking in them too. At the very least they seriously mis-describe the cases.) Gil-White is a funny guy, and by funny I mean crazy, as in serious conspiracy theory, the jews are out to get me type crazy. I think this was the sourse of his tenure problems at Penn, a place where it’s very hard to get tenure anyway. (One grant of tenure in 10+ years in the philosophy dept., for example.) He’s crazy, but right in this case.
March 28th, 2007 15:34
(There’s good reason to think the Minnisota Twins studies have more than a bit of book-cooking in them too. At the very least they seriously mis-describe the cases.)
No there isn’t, and no they don’t.
May 31st, 2007 03:48
Will,
You really need to study up more on IQ … especially from more level-headed sources than the conspiracy theorist Francisco Gil-White. IQ is a very important topic that sheds light on all the subjects you are interested in. You should be reading experts like Jensen and Flynn, not crackpots like Gil-White.
Steve