RSS
Ron has re-posted his July 2012 post on Lynn and Vanhanen, together with all the comments that it raised, and says: “It provoked an enormous outpouring of responses all across the Internet, perhaps 99% of those hostile, often intensely so, but after over a dozen follow-up columns and responses, I believe I was proven correct in almost every particular.”
https://www.unz.com/runz/race-iq-and-wealth/
Many Unz readers agree with Ron’s evaluation, in that some commentators on my recent post on Scrabble as a measure of intellectual excellence commend “Ron Unz’s compelling demolition of many of the average IQ figures put forward by Lynn and Vanhenen.”
Is there anything to say about Ron’s post 5 years later which has not already been said in the many comments and replies, to which he has provided links, together with his own replies? It would be redundant to cover everything in that extensive debate, and much better to highlight relevant new developments. I think that the subsequent 5 years of research allow two general points to be made.
1 The Lynn database is being updated and re-edited, such that every reference is traceable back to the original publication, in all but a few cases, and any untraced publication is not included in country calculations. Furthermore, the basis on which Flynn corrections were calculated have been stated and three different ways of estimating it have been made explicit (the differences are not very large). Estimates about general country effects are based on actual country results, not estimates based on neighbouring countries. Anyone can now look through the database and carry out their own studies, according to how they rate the quality of the data. Further work is in hand to calculate country means by making due allowances for sample size and representativeness. To overcome the heterogeneity of psychometric tests, Becker has produced a subsample of Raven’s-Matrices-only results. Becker is aware of a further 830 sources of IQ data which have not been checked yet. At the same time, we are aware of other publications not in the database, and are trying to include them, with the agreement of other scholars who have collected them. Debates about Lynn’s figures should use the latest Becker versions for hypothesis testing, and the newest edition will be made public in roughly a week’s time. I will post up Becker’s conference talk on his work.
2 As far as I know, Ron did not conduct a re-analysis of the database, so we do not know if the general relationship between country intelligence and national wealth is affected by his comments. His post looks mostly at the European results, and compares different countries, different time frames, and the results of Europeans and others in the USA. His view is that the hereditarian position (50% genetics, 50% environment) cannot be sustained, because there is too much variability in European intelligence, a variability that is better explained by environmental and historical factors. Where Lynn and Vanhanen have carried out a global analysis with correlational analyses, Ron has done a largely Eurocentric comparison of country pairs, and then compared country of origin and later success of those immigrants in the US. The approaches are very different in type and range. Anomalies which Ron sees as evidence of cultural changes can be ascribed by Lynn to the noise generated by different tests, samples and testing dates; all details which wash out in a general correlational study.
Could Ron’s approach lead to testable conclusions? I think so. If, by improving the data quality, the stated correlations with country wealth are reduced, then it could be argued that the Lynn correlations had been inflated by unreliable intelligence test results. Furthermore, by looking at closely defined national or racial groups it would be possible to look at changes in intelligence levels before and after immigration. In the subsequent 5 years further papers have been published on the link between country IQs and wealth. In my view they strengthen the view that ancestry is part of the cause of group differences. Notice that I say “part”. For the purposes of argument, 50% of the difference. Here are some links to a few publications after 2012. This is a selection, no more, and seeks to buttress the case for ancestry being part of the cause of group differences.
First of all, here is Becker explaining the work on the Lynn database. This might be a little hard to follow because it is just the lecture slides, but I will come back to this work in subsequent weeks.
https://drive.google.com/open?id=0B3c4TxciNeJZQmwtYV9YR3lhSXM
Here is a 2014 paper by Rindermann giving the international gains and losses from immigration.
https://www.unz.com/jthompson/migrant-competence
Fuerst and Kirkegaard have studied a very simple measure: percentage European Ancestry. They have shown it is a good predictor of outcomes at the national, provincial and district level.
https://www.unz.com/jthompson/admixture-in-americas-european/
They have just published a paper on regional differences in Argentina
http://emilkirkegaard.dk/en/?p=6795
Individual papers do not resolve a debate, but I think that more papers like these will increasingly build a picture which supports ancestry as an important part cause of intellectual differences.
English
Arabic
Chinese (Simplified)
Dutch
French
German
Hebrew
Hindi
Italian
Japanese
Korean
Persian
Polish
Portuguese
Russian
Spanish
Swedish
Turkish
Vietnamese
One thing that I think gets too little attention is that heritability is not some kind of magic, universal number (yes, I know Dr. Thompson and many others here understand this, it just isn’t mentioned often enough IMHO, especially in popular writing). It must be expressed for a given population with its own composition of genetic and environmental variability. Typically these studies are done for a relatively genetically uniform population (e.g. Americans of European ancestry). It is likely we would see different responses (percent variance explained by genetics and environment) for varying balances of genetic and environmental variability within the study population.
This issue is only increased for between groups comparisons where typically both genetic and environmental variability are increased. It might be helpful to derive measures of environmental and genetic difference between populations as an attempt to account for this. One example would be using genetic distance and difference in (log?) GDP (perhaps a bit circular). Other examples would be Piffer’s polygenic IQ score and quality of life metrics. It might be instructive to try different metrics and see which best explain the IQ and wealth variation observed empirically.
One important challenge is that the environmental influences on IQ are not well understood and may be counterintuitive (e.g. IIRC the non-effect of the WWII Dutch famine on later IQ). There may be interactions of effects which make it harder to draw broad conclusions. For example, the nutritional status of mothers is likely to make a substantial difference in the impact of breast feeding. It might be both beneficial in a well nourished population while detrimental in a malnourished population. Quality of formula (e.g. DHA content) used also matters of course.
Given that the main actionable response to any of this research would be to try to improve the local environment (policy changes are secondary IMHO) I think this is a critical issue.
In Ron’s comparison of European countries (e.g. West Germany and East Germany during communism) we have two countries which are likely to have a relatively large environmental difference compared to their genetic difference. It is only natural to expect a larger relative contribution from the environment there.
From Ron’s post:
The correspondence also supports other explanations like improving conditions (e.g. more efficient use of national resources, not increased GDP yet) improving both IQ and wealth over time. Not to mention the various possible virtuous cycles involving these factors. Looking at time lags of the various measures might be helpful here.
P.S. I am mostly concerned with the interpretation quoted initially. I think Ron’s analysis is valuable, but I think it is more usefully considered as an elaboration of the hereditarian view (whatever that actually means) than as a refutation. I think that quote is an overreach.
Thanks for your comment. Yes, a heritability estimate should only be as good as the particular sample and historical circumstances from which it is derived. However, if lots of those heritability estimates, drawn from different samples, geographies and historical times, come up with roughly the same estimates, then a meta-analytic summary is justified.
I think that time lag studies would be informative, so long as we can decide a priori what those lags should be. One generation (25 years) or two generations (50 years)? Rindermann and others have looked at whether the same predictions from ancestry hold for 19th century wealth measures, and they do hold, suggesting that there is a deep underlying trend.
Robert Tombs “The English and their history” shows that by medieval times British wages were at a level not seen in much of the Third world until the 1990s.
One further point I’d like to emphasize (though already mentioned somewhere in my 30,000-odd words of writing on the subject) is that during the first few decades of the 20th Century, all IQ tests conducted on Americans of Italian ancestry produced results in the IQ=75 range (always relative to a mainstream white population of IQ=100), and the same was approximately true of most other Americans of Southern or Eastern European ancestry such as Greeks or Slavs.
As far as I can tell, these IQ test results were published by the leading academic scholars of the field, who seemed to make every scientific effort to scrupulously rule out the contaminating impact of language-fluency or testing bias. Yet within two or three generations, all these groups had approximately converged to the white average.
Add to these cases, the much more recent examples of the Ireland Irish or the East Germans rising 15-odd Flynn-adjusted IQ points within a single generation or so, and I think what I characterize as “the Strong IQ Hypothesis” becomes quite difficult to defend.
https://www.unz.com/runz/race-iq-and-wealth/#related-series
https://www.unz.com/runz/unz-on-raceiq-response-to-lynn-and-nyborg/
I agree that any rapid rise in intelligence scores in any group weakens the strong IQ hypothesis (which I think should be renamed the “Mainly Genetic Hypothesis”). To progress this, we should look at the biggest gains and falls in the whole data set, try to see if there is a consistent pattern (rather than just measurement error due to unreliable tests and small or biased samples) and then see how that affects the overall global correlation between country intelligence and country wealth.
Once Becker has finished the next iteration of the database we can try interesting him in that analysis.
I don’t really understand what this hypothesis is supposed to mean.
It’s generally understood that heritability of a trait must be defined with respect to a specified set of environments. Certainly if one includes environments as disparate as those across countries, or across eras, or, in some cases, across levels of assimilation, then the heritability will go down — perhaps far down below .5. But if one includes only those environments which obtain in, say, modern day America, and within groups which have had time to assimilate, the heritability will go up, perhaps being distinctly in excess of 0.5.
So are we to say that the Strong IQ hypothesis is false in the first sort of case, but true in the second?
What applies to one particular case may not apply to another.
So, just because IQ testing proved to be wrong about Irish and Italians doesn’t mean that it’s wrong about everyone.
For example, take American blacks and African blacks. Suppose American blacks grew up playing modern sports and have a good hang of it. In contrast, black Africans grew up chucking spears at hippos and jiving, BUT they never played any kind of modern sports.
Suppose someone comes up with AQ using modern sports. Suppose American blacks score high on, but African blacks do poorly. One might be tempted to conclude that American blacks have higher innate athletic aptitude than African blacks. But it could be attitudinal. American blacks are so accustomed to modern sports that they have a ‘natural’ feel for it. Even if cultural biases are removed from AQ testing, African blacks just have no feel for it since they never engaged in such activity. Their idea of athleticism is just chucking spears at hippos, hollering at lions, and running from angry baboons. So, even though they are just as strong and fast as American blacks, when they are given athletic tasks that require modern-sports-mentality they just don’t get it.
Now, suppose these African blacks are raised with modern sports all around. So, they become accustomed to modern sports mentality and do quite well on AQ, indeed as well as American blacks.
Now, based on this finding, would the same apply to Mexicans, Chinese, Hindus, and etc?
No. The difference between American blacks and African blacks was attitudinal and ‘feel of the sports’. American blacks had the ‘feel’ for the game whereas African blacks had no idea as to modern athletics. But when African blacks were given the chance to develop the same kind of ‘feel’ or ‘hang’ for the game, they did just as well.
But even if non-blacks are imbued with the same ‘feel’ for the sports, they will continue to score lower on AQ because they are innately inferior in raw athletic ability.
I think the same goes for IQ. The problem with IQ testing is it’s impossible to remove all cultural biases. The very notion of testing is culturally biased. Many cultures never had such, especially among the poor. So, the very notion of taking a test is a daunting idea for some folks.
Also, some environments are not mentally stimulating. If you’re a bumpkin growing up in a farm, your tasks may be same routine over and over. So, you become accustomed to doing those tasks and never develop knack for something else. So, if you’re given a test, even one that tries to be culturally unbiased, you feel confused and disoriented. As poor Irish spent most of their time eating potatoes and southern Italians spent most of their time hiding from the mafia, they never had much time of mental activity, especially of the institutional-intellectual kind.
So, naturally when they were given IQ tests, they did very poorly. Even if testers tried to make the test as culturally unbiased as possible, the very notion of doing such a task would have been alien to the Irish and Italians and Greeks, who spent most of their time screaming and pulling their hair.
This is why giving IQ tests, no matter how well-devised to be unbiased, is a fool’s errand in societies where mental tasks are not part of the culture. So, global IQ results are highly suspect.
But as with the AQ among African blacks and American blacks, it doesn’t follow that just because Irish and Italians caught up with other whites in IQ with the change of environment, influence, and stimulation, the same will be true of other peoples.
I think…
if you take an average Chinese guy and give him lots of athletic training, he will do better in sports than average black guy given no such training.
and
if you take an average African guy and give him lots of intellectual training, he will do better in mental activity than an average Chinese guy whose entire life is about being a bumpkin hopping over dung.
but, all things being equal, if both African guy and Chinese guy were given same athletic training, the black guy will do better
and
if both African guy and Chinese guy were give same intellectual training, the Chinese guy will do better.
There is much room for improvement among all races with more training, but the problem with Africa isn’t only lower IQ but higher aggression and unruliness. If blacks were lower in IQ but had the personality of Forrest Gump, they might sit down and learn something. But they are naturally unruly and be bored by ‘lame’ schooling and be flipping out over slightest stimuli that gets them excited. So, if a black girl in class gets up and starts shaking her booty, the whole class be clapping, jiving, and going wild. Blacks have lots of energy but their racial personality prefer to expend it on funky stuff.
However, if lots of those heritability estimates, drawn from different samples, geographies and historical times, come up with roughly the same estimates, then a meta-analytic summary is justified.
You did not get res’s point.
So he was wrong assuming that you would get it.
One can imagine environments where measured heritability will be high and environments (societies) where heritability measured by the same method will be low. So meta-analytic studies will be pointless as there is no “some kind of magic, universal number” to be found.
I think this whole discussion underscores the considerable benefits of sometimes examining old books…
For example, when I sat down and actually read the books written nearly 100 years ago by America’s leading IQ researchers, I was very surprised to discover several things.
First, although the scholars had been endlessly smeared as “incompetent racists” by generations of critics, sometimes even including more recent IQ researchers, they actually seemed scrupulously fair and scientific-minded, reporting the results they found and forming various different hypotheses to explain these, as well as means of evaluating those hypotheses.
Second, their findings suggested a fascinating result, that as far as I can tell, nobody had previously noted.
For decades, Richard Lynn and others had always refuted suggestions that cultural/economic deprivation might strongly impact IQ scores by pointing to the East Asian nations, whose IQ scores were usually around 100 despite experiencing (back then!) some of the world’s deepest levels of poverty. This argument had always seemed reasonably persuasive to me, and (I assume) many others as well.
However, Lynn was apparently unaware of the history of East Asian IQ scores in America during the early 20th century. At the same time that the impoverished American-born children of Italian, Greek, or Slavic families usually had IQs 20-30 points below that of mainstream whites, children from even more impoverished Chinese or Japanese immigrant families generally had IQs of around 100.
As I pointed out in a sidebar to my big Race/IQ article, some unknown mix of cultural and genetic factors seems to almost entirely insulate East Asian IQs from the impact of “deprivation,” while almost every European population is extremely vulnerable:
https://www.unz.com/runz/the-east-asian-exception-to-socio-economic-iq-influences/
How good are the samples from the early 20th century? By the early 20th century, there were millions of Italians and Slavs in America, mainly in the Northeast. There were far fewer Asians, and most of them were in Hawaii and California, far from the Northeast where most academic research was taking place.
Ron, I managed to miss that article, thanks for highlighting it! Can you recommend any particularly interesting “old books” about IQ?
For some of the countries (e.g. Japan) I would have guessed at a “counterintuitive environment” type explanation given that, even when consumed at levels we might think of as somewhat deprived, their diet seems very healthy (cf. longevity, I note your comments in the article about that as well). But that does not apply across all of the countries, or even more importantly across the Asian populations in the US. Your article raises interesting questions about some of the healthy diet studies as well (e.g. Okinawan longevity).
Has anyone else taken a look at your ideas in the five years since that article?
One thing, I am having trouble reconciling the numbers and assertions in your article with the Flynn Effect findings in Asia given in the introduction to Lynn’s 2013 article here: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3834612/
Any thoughts?
Another thought, I wonder if the IQ deprivation resistance idea fits into Rushton’s Asian > White > Black observation for so many traits. That would actually be an encouraging comment on African IQ potential (both in Africa and the US) if so.
P.S. Chuck has a good comment at your link (#7 ) about Rindermann doing time lagged studies looking at Wealth and IQ: https://www.unz.com/runz/the-east-asian-exception-to-socio-economic-iq-influences/#comment-595467
Mr. Unz,
COMT Val158Met could possibly be the explanation (my emphasis):
– https://www.unz.com/jthompson/genetics-of-racial-differences-in-intelligence-updated/#comment-1903682
I’d have to dig out the books I used from five years ago, and those merely quoted the studies in question from question, originally published in various academic journals. But as I recall, they seemed pretty solid, and the authors quoted them matter-of-factly. And actually, I think due to Terman at Stanford, the West Coast was an important hotbed of IQ research, with much of the testing of Italians, etc. being from there, or from Midwestern states rather than the East Coast.
Admittedly, the studies of Italians, Greeks, Portuguese, and Slavs were more numerous. But those almost uniformly showed an IQ deficit of something like 20-25 points relative to Anglo-Saxons. Meanwhile, most of the studies of Chinese and Japanese studies showed them to be fairly close to Anglo-Saxons.
Interestingly enough, since all the Italians and Portuguese tested very close to Mexicans, they were quite logically grouped together as “Latin peoples.” Back then, all these groups were considered “sort-of white” in much the same way.
Generally the poorest people are considerably more likely to be also less intelligent, in general, than the people who at least manage to provide themselves a better economic security. Therefore, it is very difficult to find a large number of very intelligent people, in a general way, and also very poor. In the same way as finding “stupider” people [or less intelligent people] who are very rich. Because the mundanely smarter ones are more likely to quickly secure a better family environment for themselves and for their family or inherit a good standard of living from their parents, then it is difficult for many of them to turn out to be very poor.
It is like the relationship ” skin color x climate ”. Because the vast majority of human populations that evolved to have lighter skins live in temperate zones, this does not mean that they could not have evolved into lighter skin even in intertropical areas. It is a primarily logical and constant relationship: temperate climate selected for lighter skin such as poverty causes reducing of intelligence, only always taken as ” causal ”.
The only way to know if poverty really has a capacity [itself] to reduce intelligence would be to put very intelligent people in poor environments, force them to procreate one another, and to prevent them from improving their environment, that is, to act intelligently.
But seems very unlikely to prove that ”poverty” cause less intelligence, but sencodarily, poor environment can act as accumulative or somatic force to increase this intelectual penury for example, a empoverished family, come from middle class, and half of their descendents marry a clearly ”low-quality” women or men in intellectual terms [even this type of event also seems to be rare].
Ron, I know you must be busy now but More Unz articles!
I would like to hear ypour thougts on this and other topics.
I wonder what Mr. Unz’s thoughts would be as to the lowest genotypic group mean and/or when it might be actualized into a phenotypic mean.
Just playing devil’s advocate, I would say the ancient Greco-Roman world had peerless art and epics. Of course, favorable climate and geography. Italy was also really the center of the Renaissance. Some of this might just be smart fraction and large population, but it at least makes it plausible.
Ireland also had epics. Not as perfectly awe-inspiring as Homer or Virgil, but still the Tain is pretty good. Ireland is quite genetically close to the UK, not to mention, today, Ireland’s per capita is usually given as higher than the UK’s, despite it being a colony for many hundreds of years and the Penal Laws. And capital does primarily move freely.
So, just because IQ testing proved to be wrong about Irish and Italians doesn’t mean that it’s wrong about everyone.
IQ testing did not prove to be wrong. IQ testing results proved that the ideas some people like to entertain about the IQ tests are completely wrong. You can’t sweep the fact that some ethnic groups tested 20-25 lower at some point of their history under the carpet. One would think that in a scientific process there would occur a correction, however the process clearly must be not scientific as the correction did not occur and the purveyors of the IQ business continue as if nothing has happened. The question must be raised what is the driving force of the purveyors of the IQ business. Clearly it is not the search for the truth. What is their mission and who pays for it? Follow the money.
“if you take an average Chinese guy and give him lots of athletic training, he will do better in sports than average black guy given no such training.”
Which sports? Somatype matters, so talking about the average black and Chinese guy, you’re talking Meso for the black guy and endo for the Chinese guy (whites are endo as well but less so than Asians).
https://notpoliticallycorrect.me/2017/03/25/racial-differences-in-somatype/
“First, although the scholars had been endlessly smeared as “incompetent racists” by generations of critics, sometimes even including more recent IQ researchers, they actually seemed scrupulously fair and scientific-minded, reporting the results they found and forming various different hypotheses to explain these, as well as means of evaluating those hypotheses.”
Mr. Unz, care to tell us the name of the books? Thanks.
Sure. Probably the most useful was the monograph “Intelligence and Immigration,” published in 1926 by Clifford Kirkpatrick. I think there was also another book written on the intelligence of various immigrant groups published a few years earlier, but I can’t quite remember the name or author.
Also, “American Ethnic Groups,” published in 1978 and edited by Thomas Sowell, summarized a lot of the data for Asian groups from the first half of the 20th century.
Thanks for the recommendations! I just ordered copies.
This from 1922? Mental differences in certain immigrant groups: psychological tests of south Europeans in typical California schools …
It should be out of copyright, but I was unable to find a PDF.
The author (Kimball Young) also had journal articles on this topic.
Probably worth mentioning that there are two versions of the Thomas Sowell edited 1978 book:
Here is a review of the first: https://fee.org/articles/book-review-american-ethnic-groups-edited-by-thomas-sowell/
The latter has an additional 160 pages of data.
It was?
But there would have been a sizable population back then of Italians, Greeks, Portuguese, and Slavs that spoke very bad and broken English, while there would have been a few Chinese that spoke decent English while most would have been living in Chinatowns and would have had virtually no English. Presumably they didn’t test Chinese who could speak no English at all.
It seems plausible that the samples would have been very different. The sample for the Italians, Slavs, etc. would skew towards a population with limited English, while the sample for Chinese would skew towards the small minority that spoke good English.
I really need to focus on my big software project, which should shortly be ready for general testing. But I might as well add one more point.
As I mentioned, there seems to be a great deal of empirical evidence that East Asian IQs are “insulated” from the impact of economic/cultural deprivation in a way European IQs aren’t. The reasons aren’t entirely clear, but here’s a plausible hypothesis, at least regarding the Chinese.
For a milennium or two, ordinary Chinese lived under massive selective pressure, with business skills/intelligence being a an absolutely crucial survival/reproductive factor, quite unlike the situation of ordinary Europeans. During much of this period, they were also extremely poor, often existing on the edge of starvation.
Given these two factors, it wouldn’t be surprising if there were heavy selection toward insulating Chinese intelligence against the ill-effects of poverty and deprivation.
I discussed all of this in a 2013 article:
https://www.unz.com/runz/how-social-darwinism-made-modern-china-248/
Just a raw empirical observation: In Europe, in primarily Northern-Western countries, the wait-staff who are indigenous or second generation but educated in-country more than a decade ago, can still accurately add a dinner-tab for a table and, in many cases if necessary, accurately split it, without the aid of a calculator or the cash register. I haven’t seen that skill in evidence in North America for some time. IQ almost certainly has a hereditary component, but I would wager environment is the dominant factor. As for first generation immigrants, no matter what language they speak, all but a few demonstrate a marked cognitive deficiency. The European who thinks they are getting doctors and engineers at bargain prices is demonstrating social and emotional quotients well below their IQs.
It’s strange that you forward an evolutionary, genetic explanation. I would prefer a cultural explanations. What about that Chinese were less likely to come from illiterate families than Slavs or Italians? What about that knowing Chinese ideographs predisposes one to solving some type of puzzles one encounters in IQ tests? What about that taking tests in Chinese society is several thousand years old tradition and it is understood as a mean of improving ones station in life, so Chines put more effort into test taking unlike more fatalistic and passive aggressive Slavs or Italians?
Just as one example, Dean Martin was born one hundred years ago this month (6/7/1917), in Steubenville, Ohio, to immigrant parents from Italy. Young Dino Crocetti did not begin to speak English until he started school, in 1922, despite his being a native-born American citizen, living in an Eastern industrial city, in 20th Century America.
But southern and eastern Europe were a lot poorer than the US a century ago. Italian, Greek, Slavic, etc. immigrants to the US would have been less deprived than their compatriots back home. Do you suppose that southern and eastern Europe’s IQ was even lower than the 20 to 25 points lower you posit that southern and eastern European immigrants had relative to Anglo-Saxons? It does not seem very plausible that southern and eastern Europe would have had an IQ 30 or more points – 2 or more standard deviations – lower than Anglo-Saxons did a century ago.
Furthermore, if we go back in history, all Europeans would have been more deprived than immigrants to the US a century ago were. Do you suppose Europeans had IQs 2 or more standard deviations lower than Anglo-Saxons do?
Seeing as how Mexicans have huge amounts of Amerind admixture, that sounds highly illogical to me….
Language and part of culture, namely catholicism that make latin americans and latin european-iberians similar.
Yes, but, using that logic, we could also group Black and White Americans together under the Anglo-American rubric. But that would mean overlooking the fact that Black Americans are approx 80% Sub-Saharan African in terms of ancestry….
Perhaps fecal transplants are the secret to boosting IQ:
http://www.newser.com/story/244642/cyclists-might-turn-to-poop-doping.html
Hitler’s physician was treating him with “poop”.
Yes. Do you have data to the contrary?
Yes I know for this comparison it’s don’t make sense.
don’t understand what surprises you. first i don’t think Lynn ever said there no impact of deprivation. second early 20 century IQ test were only invented and were not accurate. also samples how large and unbiased are they? Sicily Greece and some Slavs have IQ below or equal 90. WASP in USA have IQ of 105. now at that time probably poorest peasants immigrated from those countries so their IQ must be lower. so makes sense. those Europeans could be less fluent at English back than also.