Unz on Race/IQ – Response to Lynn and Nyborg
Richard Lynn has now produced a lengthy and detailed rebuttal to my article Race, IQ & Wealth questioning his theories, as has Helmuth Nyborg, another leading IQ expert and strong supporter of Lynn. Their analyses have been published or highlighted on several prominent racialist websites, and I am herein providing my own rejoinder.
First, I will admit to being a bit confused about Lynn’s overall position. Although he often seems to be endorsing my viewpoint in its generality, he seems to strongly dispute nearly all the specific details.
For example, Lynn claims that he has never denied the substantial role of educational and other environmental factors in determining the IQ results of different populations, and that he is simply suggesting that genetic factors may also influence “some” of the IQ differences he reports. As it happens, this is almost exactly the central thesis of my own article, in which I argue that the evidence overwhelmingly refutes what I call the “Strong IQ Hypothesis,” but that a “Weak IQ Hypothesis” might very well be correct. Essentially, I am proposing that the enormously large differences in population IQ reported by Lynn are primarily due to factors of social environment—poverty, education, rural deprivation—but that even if these external differences were completely eliminated, there still might remain a much smaller residual genetic effect, though its size and direction is somewhat speculative.
But although Lynn seems to be affirming my “Weak IQ Hypothesis” in the general case, he seems to oppose it in every given particular, which I find perplexing. For example, the word “education” appears only once in the index of his 300 page book IQ and the Wealth of Nations, and in that reference he provides a few sentences citing academic articles claiming a strong educational role in IQ, but then follows with a paragraph debunking and refuting the notion that education has any fundamental or permanent impact on IQ. Meanwhile, there seems to be no discussion anywhere of related factors such as “culture” or “rural deprivation”, and I am not aware that he raises any of these possibilities in his reporting and analysis of specific national IQ scores, even when these seem to show massive fluctuations over time. He certainly does recognize the important role of various biological factors such as nutrition and health, but even these merit only three or four pages out of 300.
In his 2006 sequel, he discusses these issues at greater length, but with little more clarity or consistency. For example, he raises the possibility of socio-economic factors substantially impacting both IQ and educational attainment, but concludes “This explanation cannot be correct” based on adoption studies (p. 44). Although at points in the book he vaguely seems more open to the role of “social environment” as a factor, he almost never seems to consider it when he discusses specific differences in national IQs. I might speculate that he is providing himself some “plausible deniability” by accepting non-genetic explanations in the vague abstract while almost always ignoring or dismissing them in each specific case. If I am mistaken, and he does indeed support “the Weak IQ Hypothesis,” I would be very glad to know this.
Numerous others might also benefit from such clarification. Not a single one of the vast outpouring of critical remarks I have received from Lynn admirers has ever suggested that I was misrepresenting Lynn when I characterized his position as essentially IQ-determinism.
As to the specifics of Lynn’s rebuttal, one significant problem I quickly encountered was a high degree of what seems to be serious factual error. For example, he directly and repeatedly quotes me as claiming “the European peoples are genetically indistinguishable” and “all the European peoples are genetically indistinguishable,” devoting several long paragraphs to refuting this claim. However, I never made any such totally absurd and scientifically ridiculous statements, and when I asked him for his source, he was unable to locate it anywhere in my writings. Admittedly, I did claim that East Germans and West Germans were indeed “genetically indistinguishable” and also that Greeks and Turks are quite genetically similar (as Lynn himself has stated on various occasions), but obviously such statements are entirely different from claiming the same is also true for Swedes, Basques, and Greeks. As Lynn himself explains in rebutting my (non-existent) claim, Nordics tend to be fair and blue-eyed while Sicilians are very rarely so, rendering it rather unlikely those two groups would be “genetically indistinguishable.” If Lynn wishes to portray me as a total ignoramus on genetic matters that is certainly his privilege, but I would prefer he read my writing more carefully and avoid inventing spurious quotations to buttress his case.
In another example, Lynn implies my review of national IQs was careless in that I relied upon the data in his 2002 and 2006 books, and ignored the more recent and extensive data presented in his 2012 book. However, since this latter book was actually published on July 16, 2012, six days after my own article had been sent to the printers, I do not believe I can reasonably be faulted for failing to incorporate his new material. Furthermore, once I managed to obtain an electronic copy of his new 2012 book, I discovered it contained at least several puzzling anomalies, which Lynn has been unable to clarify for me.
For example, Lynn refutes my evidence for a low Ireland IQ during the 1970s by referring me to the more extensive data in his latest book, saying it debunks my claim. However, when I examined the Ireland IQs in that book (p. 402), I discovered that he had inexplicably failed to include the massive 1972 study of 3,466 students which established an Irish IQ of 87 and which had appeared in all of his previous books. When I asked him why he had excluded the largest early Irish IQ study, he said he had no answer, and that perhaps “this omission was a mistake.” As it happens, nearly all of his other Ireland data tends to be from around 1990 or later, and simply reinforces the strong evidence of a rapidly rising Irish IQ which I had already discussed in my own analysis.
As another example, I had pointed out that Austrians and Croatians were quite genetically similar, and that the huge 11-13 point IQ gap reported by Lynn (2002) seemed wildly implausible as a primarily genetic effect, probably instead reflecting the depressed socio-economic condition of Croatia at the time, and hence likely to rapidly shrink under economic development. Lynn “refutes” my claim by citing the more recent data in his 2012 book which shows the Austrian/Croatian IQ gap has now narrowed to just a single point, mostly due to a huge rise in Croatian IQ; this would seem the exact prediction of my own model. Strangely enough, his 2012 data allegedly includes a Croatian sample from the large Buj (1981) collection, which had never appeared along with the other 19 Buj results in any of his previous two books, and when I asked him why he had never previously included this particular Croatian result, he had no explanation.
Sometimes Lynn construes my slightly loose phrasing as serious error. For example, I mention that Lynn’s 2010 research showed that today the Southern Italian IQ “was as low as 89.” He claims I misquoted him, since only the Sicilian IQ was 89, while other parts of southern Italy were in the 90-92 range. Since his other studies had placed Northern Italian IQ at 102-103, I had described this gap of 10-14 points between Northern and Southern Italians as being “almost a full standard deviation” while he argues it is closer to being two-thirds of a standard deviation. Since the IQ gap between Southern Italians in Italy and those in America is so wide, he suggests that only the very smartest Southern Italians immigrated here, but this seems totally implausible, especially when we consider that mean reversion would have drastically reduced any initial difference. Furthermore, all the IQ data from the 1920s which I quoted placed Italian-American IQ back then in the 78-85 range, which directly contradicts Lynn’s hypothesis.
Lynn also challenges my Balkan analysis. His 2012 book places the Bulgarian IQ at 92.5 and the Romanian IQ at 91, with the largest and most recent Romanian study coming in at only 88. Just I claimed, these figures tend to be 10 points or more below such northern European peoples as the Germans, Dutch, or Swiss, and I find a strictly genetic explanation of this huge gap far less plausible than the obvious social differences between Europe’s wealthiest and its poorest countries. Lynn claims these very low Balkan IQs are due to substantial African and Middle Eastern ancestry, but I have never heard of this being the case for Balkan Slavs, and would like to see some evidence.
Lynn’s 2012 book also places the Greek IQ at 92, and he challenges my assertion that Americans of Greek or South Slav ancestry are actually well above the white American average in income, IQ, and other indicators of ability. However, I obtained this exact result from the GSS dataset by running RACE=WHITE/ETHNIC against the WORDSUM, EDUC, and CONINC variables. Both Greeks and Yugoslavs have Wordsum-IQ and years of education significantly above the white average, while the family income of both these ethnicities are among the highest of any white group, $122,700 for Greeks and $113,500 for Yugoslavs against just $97,900 for the average white American. These results would seem implausible if Lynn’s very low IQ figures for Greeks and Balkan Slavs were largely innate.
Again, Lynn suggests that only the absolutely smartest Greeks and South Slavs came to America, and again I point out there is no evidence for this, that the IQ differences results from such selective migration would have substantially regressed after the first generation, and that the 1920s American IQ tests for Greek immigrants placed them at just 83.
In attempting to rebut my GSS evidence of a very rapid rise in American-born Mexican-American Wordsum-IQ over the last few decades, Lynn cites a 2010 paper by Ang et. al which he says refutes my claim of such a Super-Flynn Effect. However, he is incorrect. The study he references aggregates all Hispanics, while my result was solely for American-born Mexican-Americans, a much smaller subset. Indeed, if the foreign-born Mex-Ams are included, the rise in Wordsum-IQ largely disappears.
Furthermore, I have also pointed out that between 1975 and 2011, there was a five-fold rise in the percentage of Hispanics taking the SAT, while the gap between white and Hispanic scores remained essentially unchanged. Given that the Hispanic results were dipping so much deeper in their ethnic ability-pool, I think this can only be explained by a large rise in general Hispanic academic performance.
Lynn also disputes my claim of an anomaly in the pattern of East Asian IQ scores. Yet across all of his 50-odd such IQ studies, there is just a single case of East Asians scoring below 100, despite enormous poverty and deprivation. Meanwhile, most of his Southern European IQ studies yield results closer to 90 than to 100. I am deeply skeptical that the East Asian/Southern European ability gap is actually so enormously large. Furthermore Lynn never addresses the historical fact that virtually all the Southern and Eastern European immigrant groups in America had tested IQs of around 80-85 during the 1920s, while their Chinese-American and Japanese-American immigrant counterparts usually scored 100 or above. I suspect huge factors of cultural, educational, and socio-economic deprivation were responsible for the absurdly low immigrant European IQ scores, and Lynn completely ignores this important evidence, perhaps because it would be very difficult to reconcile with his strongly genetic IQ model.
Finally, Lynn closes his rebuttal by repeating his boilerplate disclaimer that he has “never maintained that IQ is overwhelmingly determined by genetics,” although this seems to be his clear reasoning in every single particular example he discusses. However, if I am mistaken and he is not being disingenous on this point, I cannot possibly understand why he chooses to oppose the overall conclusions of my article. As mentioned above, I had emphasized that absolutely none of my analysis would rule out “the Weak IQ Hypothesis,” in which some residual European IQ differences might indeed be due to genetics rather than environment. If—as Lynn seems to be suggesting—this were actually his own position, why would he have written a heated 6,000 word rebuttal instead of simply thanking me for agreeing with him?
I find much less to discuss in Helmuth Nyborg’s rather brief rebuttal.
Nyborg criticizes me for relying solely on Lynn’s 2002 book, but as I have repeatedly made clear, I supplemented this with his 2006 book as well as some of his other books and writings. I obviously did not use the data in his 2012 book, since it appeared in print simultaneously with my own article. Nyborg also cites the contrary IQ evidence of twin/adoption studies, which I had already addressed at length in my own article.
He describes “the Strong IQ Hypothesis” as a “straw man” position, an “imposter” which “obviously runs counter to all behavior genetics evidence,” and I would certainly agree with those charges. But just because a hypothesis is patently absurd does not mean it may not be widely held or promoted in certain circles, and my entire article was merely intended to demonstrate that absurdity. Meanwhile, I pointed out that “the Weak IQ Hypothesis” seems perfectly consistent with all evidence, and if that is actually the position of Lynn and Nyborg, then we should have no major disagreement.
I am also very glad that Nyborg emphasizes that Lynn and his co-author stand as “undisputed and widely respected leaders of the field” of IQ studies. Some of my critics have grown very concerned over the major anomalies I have highlighted in Lynn’s data and have begun suggesting that by focusing my attention on Lynn’s work, I was unfairly attempting to discredit the entire field of IQ research.
Finally, I do regret the nasty personal attacks and misrepresentations which Nyborg, Lynn, and many of their allies have endured. But given the many hundreds of caustic insults and harsh denunciations I have recently received from Lynn’s energetic admirers, I would suggest that there might be two sides to this story.
(cross-posted at www.ronunz.org)