Review of Metapedia’s Race and Intelligence review

Wikipedian battlegrounds

For those who don’t know, Wikipedia is a common battleground for the ideological part of the race and intelligence debate. One can see this in the talk pages of these articles. See also the discussion of the phenomenon here.

The most active pushers of environment-only right now are WeijiBaikeBianji and maunus. WBB got a temp ban for it recently, but it has been lifted, so he is back in business. Probably the easiest indication of his extreme bias is his own compiled list of good sources. A search reveals that there is not a single citation of Jensen, the most prominent researcher in this area. There are also 0 for Richard Lynn, and Phil Rushton, i.e. the three grand old men of the hereditarian side. On the other hand, one will find 9 references to Flynn, 17 to Sternberg and 2 to Nisbett.

Not surprisingly, due to the boringness of engaging in edit warring and because the Wikipedia source code is free, others have set up Wikipedias more suitable to their ideologies. Perhaps the most comical is the ultra-conservative creationist Wikipedia Conservapedia. Of more interest is RationalWiki, a Wiki centered on rationality and pseudoscience. The general content of the Wiki is good, e.g. on NLP or dowsing. However, it is awful on matters that the American left-wing does not like, including race and intelligence. Unsurprisingly, the combination is even worse, for instance in the article on The Bell Curve:

The Bell Curve is a highly controversial 1994 book by Richard Herrnstein and Charles Murray. It purports to show that intelligence is the most dominant factor in the trajectory of each person’s life, and it serves to predict such things as socioeconomic status and tendencies towards criminal behavior. The book has served as a platform for many modern-day racists, giving them an “intellectual” basis and source of data to support many of their beliefs. Quite a bit of the research compiled within The Bell Curve is not disputed, but the conclusions drawn from it are considered by many to be bunk, and it has been criticized for aiding racist ideologies.

One of the admins of the website, Krom, is currently engaging John Fuerst in discussion at the OpenPsych forum. The thread is up to >34 pages by now.

All in all, RationalWiki is an okay source but mind the topic. Alternatives include TalkOrigins (for creationism), SkepticalScience (for climate denialism), as well as the Skeptic’s dictionary which covers many areas.


Some time ago I noticed that someone had set up another Wiki that does not fret about the race and intelligence stuff. I decided to give the article a read since I’m an expert in this area. To my surprise, it is fairly updated. While not quite as good a review of the facts as John Fuerst’s old 2012 review, it is more up to date, even covering findings published at HumanVarieties.

When I read the Race and Intelligence page I had a lot of comments, but just kept them in my head. However, this seemed like a bit of a waste. Hence, I decided to write them down, forming a review post. I could have instead edited the page, but I’m too busy. The page seems to be mostly written by an anonymous swede.

Review of Metapedia’s Race and Intelligence

Quotes are from the article unless otherwise stated.

Race differences in intelligence was historically a common view. For example, Muslim writers stated low intelligence among Blacks.[1][2]

Early scientific research started in the nineteenth century and included methods such as skull and brain measurements.[3]

The first IQ test was created in 1905. By the end of the twentieth century many hundreds of studies on racial IQ differences had been published.[3]

Race research in general, including also race and intelligence research, become increasingly a taboo subject after WWII. During this time the Pioneer Fund was influential in keeping some research and debate alive.

Galton is not mentioned but should be. He made the first quantitative estimates of racial differences in intelligence. They were surprisingly precise despite Galton not having any actual mental tests (only some imprecise chronometric tests in the end). Galton traveled widely in Africa and formed his judgment on that basis. Jensen wrote a lengthy review of Galton’s lasting influence.

In 1969 Arthur Jensen caused great public controversy with the article “How Much Can We Boost IQ and School Achievement?” in which he argued for genetics being an important explanation for the measured differences.[4]

Here they should cite the original article. Because the article was so influential, they should do a direct quote of his words. Jensen did not state his case quite as strongly as the writing makes it seem.

Richard Lynn in his 2006 book Race Differences in Intelligence reviewed the literature on worldwide IQ testing and calculated the average IQs for different races based on earlier IQ test results (citing hundreds of different studies testing the average IQ of different races).


This should be in a table.

They should mention the debate about the Sub-Saharan African IQs with Wicherts, Lynn, Meisenberg, and Rindermann. See citations in Rindermann, H. (2013). African cognitive ability: Research, results, divergences and recommendations. Personality and Individual Differences, 55(3), 229-233.

The average IQ of the world as of 2000 has been estimated to be 90 based on estimated average country IQs and country population sizes.[8]

A histogram of the IQs seems in order here.

The US group “Hispanics” is a diverse group that may have European, Amerindian, and Sub-Saharan African origins in varying proportions. Most are of mixed Amerindian/European origin.

In the US the tested average IQ of “Hispanics” is typically intermediate between that of Blacks and Whites.[12] Both the above mentioned 2001 meta-analysis and the book Race Differences in Intelligence found an average IQ of 89.[3]

Probably mention the results from Fuerst’s 2014 review: Fuerst, J. (2014). Ethnic/Race Differences in Aptitude by Generation in the United States: An Exploratory Meta-analysis. Open Differential Psychology.

Several studies of the IQ of Gypsies, a people of South Asian origin living in Europe since several centuries but with little intermarrying with other groups, have found average IQs ranging from 70 to 83.[16]

See also Gypsies: Intelligence.

They should cite the results from the meta-analysis presented at the London Conference of Intelligence. The presentation is here.

The 2015 book The Nature of Race stated that “the cognitive ability scores of international migrants tend to correlate with the cognitive ability scores of those from the regions of origin. That is, to some extent, contemporaneous migrants carry their region of origin abilities with them and the differences brought persist at least until the second or third generation (Carabaña, 2011; De Philippis, 2013; Fuerst, 2014; Kirkegaard, 2015).”[18]

They should not cite long quotes like this (in line) and not giving the references for the cited literature. Readers can’t know what “(Kirkegaard, 2015)” refers to without looking up that study. I don’t even know which study it is.

The book The Nature of Race stated that “As Baten and Juif (2013) note, the international cognitive ability differences are not new and they precede the event of mass schooling. As such, Baten and Sohn (2013) found that Korea, China and Japan had high numeracy levels in the 1600s; Juif and Baten (2013) found that Spanish and Portuguese had higher numeracy levels than Amerindian Incans in the 1500s. Juif and Baten (2013) also found that 1820 cohort ethnic/national cognitive ability levels predicted 21st century national levels.”[18]

They should show a scatter plot of the Age Heaping x IQ scores. See my earlier post.

There have been some estimates of the IQ:s of monkeys, apes, Homo habilis, and Homo erectus based on how far they progress or have been estimated to progress on Piaget’s cognitive stages of development. Monkeys have been estimated to have an IQ of about 12, apes about 22, Homo habilis about the same as apes, and Homo erectus about 50.[3] Some apes have in captivity been taught limited language skills. However, sceptics have doubted that the claimed language abilities are real.[22][23]

Lynn’s speculations probably do not warrant inclusion. The last two references should not be cited.

Those arguing for a genetic explanation, sometimes referred to as “hereditarians”, are frequently subjected to various forms of ad hominem personal attacks. This may include accusations of being “racist” (in some extremely negative sense), associations with claimed “racists”, claimed “racists” using the research etc. Obviously ad hominem personal attacks are not scientifically valid arguments regarding what causes the racial IQ differences.

These claims probably warrant some examples. RationalWiki is full of them. One can also cite stuff from Nisbett and Sternberg’s writings.

Denying the existence of races may be used as an attempted argument against race and IQ research. However, it should be noted that even if races in the sense of subspecies were proven to be incorrect, then this does not actually make the genetic debate disappear. Blacks and Whites would still differ genetically regarding, for example, the genes for skin color and the genetically determined prevalence of sickle-cell anemia. So they could differ also regarding IQ genes.

Furthermore, IQ is likely affected by a very large number of genes. This means that even if the population differences regarding the population frequencies of individual gene variants affecting IQ are all small, but these population frequencies correlate, then the total effect of many such small but correlated differences may be that the population differences regarding genetic effects on IQ are very large.

It is perfectly possible to study the role of genetics as an explanation for differences between groups that are not subspecies. Current examples would include the enormous amount of medical research regarding the genetic differences between those having a certain disease and those not having this disease.

Zero references are given for this section. Including a few as well as perhaps a figure would be in order. Something like this perhaps

IQ values such as the 54 for Bushmen have been by some as implausible low since it would a give a diagnosis of mental retardation in European countries. However, Europeans with more severe forms of mental retardation often have genetic diseases that cause many other problems beside the low IQ. A better comparison is with European children with comparable IQ. An IQ of 54 is equivalent to the average IQ of European 8 years old children. These can learn to read, write, and do arithmetic. Historically the great majority European children worked productively at this age. This is also the case today for many children of this age in developing nations.[3]

It is worth mentioning that there are very few studies of these groups. Lynn is not known for being diligent with his exact numbers, see Malloy’s review, e.g. for Thailand. I seem to recall the Bushmen value was incorrect.

Controlling for different average socioeconomic status (SES) of Blacks and Whites only reduces the Black-White IQ gap by a third or 5 points. Furthermore, if the Black-White IQ gap is in part caused by genetics, then this number is overstated since the Black-White SES gap is partially caused by the Black-White IQ gap.[4] Not considering such effects is one example of The sociologist’s fallacy.

Here it is also worth citing Sesardic’s excellent book.

A 2014 genetic study, although not studying racial IQ differences, found that “using a new technique applied to DNA from 3000 unrelated children, we show significant genetic influence on family SES, and on its association with children’s IQ…our results emphasize the need to consider genetics in research and policy on family SES and its association with children’s IQ.”[30]

The citation of only this study makes it seem like this is a new idea or new finding. In fact Rowe et al reported a finding like this in the late 1990s. The idea is fairly obvious and was first promoted widely by Herrnstein in the 1970s. There are also multiple other newer studies also using GCTA finding narrow h2 values based on common variants around .20.

  • Rowe, D. C., Vesterdal, W. J., & Rodgers, J. L. (1998). Herrnstein’s syllogism: Genetic and shared environmental influences on IQ, education, and income. Intelligence, 26(4), 405-423.
  • Marioni, R. E., Davies, G., Hayward, C., Liewald, D., Kerr, S. M., Campbell, A., … & Deary, I. J. (2014). Molecular genetic contributions to socioeconomic status and intelligence. Intelligence, 44, 26-32.

An environment only explanation for the Black-White IQ gap predicts that the IQ gap should be smaller at higher levels of parental SES since these children should be less exposed to the environmental factors lowering IQ. However, the gap is actually equal or larger at higher parental SES levels.[4] In contrast, hereditarians can explain this by regression to different racial genetic averages (see the section “Regression to the mean” below).[31] Another explanation is Black parents having lower average genetic IQ than White parents despite having similar SES. This may be due to factors such as affirmative action causing discrimination against Whites in education/employment.[32]

That does not have to be the case. But yes, a simple environment only model based only on AAs receiving more negative environment factors that are only present in substantial amounts in the lower S levels of societies would imply this. Such a model is falsified.

One early view was that the US Black-White IQ gap was caused by the segregated schools. However, the 1954 Supreme Court decision against segregated schooling and the consequent nationwide program of school busing did not cause the gap to disappear. Furthermore, the Coleman Report found little support for the schools being an important explanation for the Black-White IQ gap or IQ results in general. Negligible, and in some cases, negative correlations were found between IQ and variables such as pupil expenditure, teachers’ salaries, teachers’ qualifications, student/teacher ratios, and the availability of other school professionals. Also, IQ group differences are found also in European countries with desegregated schools. Hereditarians have furthermore argued that the Black-White IQ gap is equally large for 3-year-old children who have not yet started school.[4][3]

Coleman report is mentioned but not cited properly. No references for the European results or their exact nature.

Similar diagram based on the 1994 book The Bell Curve.

No actual source given for this figure.

Certain factors that are common in developing countries like iodine deficiency during pregnancy/childhood and certain tropical diseases like malaria are known to affect IQ negatively. However, these factors are very rare in developed countries and thus cannot explain for example the US Black-White IQ gap.

That malnutrition would be more common among Blacks than Whites in the US is argued to be excluded the absence of height differences and nutritional studies.[3]

They keep citing Lynn 2006 for all kinds of claims. This is not proper. They should cite primary literature showing that these diseases are not important factors. I don’t know such studies, maybe they exist. A recent Faroe Island study found that prenatal mercury poisoning may have a small effect on later IQ (2.2 IQ points per 10 fold increase), even controlling for maternal IQ.

  • Debes, F., Weihe, P., & Grandjean, P. (2015). Cognitive deficits at age 22 years associated with prenatal exposure to methylmercury. Cortex.

A 2013 study examined to what degree the average country IQ differences are caused by poor living conditions at or near the test-takers’ time of birth and stated that “The paper finds that the impact of living conditions is of much smaller magnitude than is suggested by just looking at correlations between average IQ scores and socioeconomic indicators…As far as IQ and the wealth of nations are concerned, causality thus appears to run mostly from the former to the latter. The test-takers’ region of ancestry dominates the regression results. While differences in average scores worldwide can thus be plausibly viewed as being influenced by genetic differences across world regions, it is also possible that score differences are influenced by regional differences in culture that are independent of genetic factors. Differences in average IQ across world regions may change in the years ahead insofar as the strength of Flynn effects may not be uniform, but some regional differences in average g levels seem likely to continue indefinitely.”[33]

Another long quote that is given in-text instead of block. This presentation suboptimality is very common. I shall not mentioned it more times.

There may be problems with testing US immigrants or other persons who are not native English speakers if using English verbal IQ tests. On the other hand, such groups may be tested with non-verbal tests such as Raven’s Progressive Matrices. Hereditarians have argued that studies have shown that IQ test scores predict school grades and job performance equally well for Africans as they do for non-Africans[4][10]

The scores from some countries may be uncertain due to factors such as only small studies being available. On the other hand there are also very large scale international student assessment tests that avoid many of these problems. See the article Countries and intelligence.

They fail to cite the most important book in this area, Jensen’s 1980 book.

They cite my study about item-level Raven’s. They could cite that for lack of bias here as well. The item-difficulty scores had very high cross-sample correlations with a mean of .88.

Stereotype threat is an argued fear that a person’s behavior will confirm an existing stereotype of a group to which the person belongs. This may in turn lead to an impairment of the person’s performance. This has been seen as one explanation for the racial gaps.

Early laboratory experiments finding a stereotype threat effect has been greatly misreported, in both popular and academic literature, as showing that stereotype threat explains the whole gap.[34] An unpublished “meta-analysis of 55 published and unpublished studies of this effect shows clear signs of publication bias. The effect varies widely across studies, and is generally small. Although elite university undergraduates may underperform on cognitive tests due to stereotype threat, this effect does not generalize to non-adapted standardized tests, high-stakes settings, and less academically gifted test-takers. Stereotype threat cannot explain the difference in mean cognitive test performance between African Americans and European Americans.”[35][36]

They should cite prominent authors invoking this explanation. Surely Flynn, Sternberg and Nisbett will provide the necessary text material.

They cite Wichert’s conference presentation, but not the actual published meta-analysis, this does not make sense. They should also cite work by Lee Jussim on this topic. Steve Sailer is an inappropriate source if cited alone.

  • Flore, P. C., & Wicherts, J. M. (2015). Does stereotype threat influence performance of girls in stereotyped domains? A meta-analysis. Journal of school psychology, 53(1), 25-44.
  • Jussim, L. (2012). Social perception and social reality: Why accuracy dominates bias and self-fulfilling prophecy. Oxford University Press.

In the middle of the twentieth century a large number of early childhood intervention programs, such as the Head Start program, were tried with one expectation being that these would eliminate or substantially reduce various IQ gaps including the racial IQ gaps. Large initial IQ gains were also found but the initial enthusiasm declined as it become apparent that the IQ or achievement tests gains soon faded away as the children grew older. For example, a 1995 review of 36 such early intervention programs found no consistent pattern of lasting effects on IQ or achievement tests.[37]

A few of the programs have found longer lasting effects contrary to this general pattern. The most well-known may be the Abecedarian Early Intervention Project which found limited IQ gains lasting to adulthood. However, there have been various criticisms. One is that there is evidence for the intervention and control groups being dissimilar due to pure chance (e.g., sampling error) or non-random attrition of participants. The other few claimed exceptions have been criticized due to poor methodology, “teaching the test”, and even a conviction of misuse of federal funds. A 2014 article stated lack of good evidence for anything except a null (or small) long-term effect from intervention programs (as well as adoptions) on the g factor.[37][38][39]

Here they fail to cite Nijenhuis and my paper on the g-loadedness of Headstart IQ gains.

Furthermore, a large meta-analysis of these intervention studies showed that the favorite studies cited by sociologists are statistical flukes, i.e. it is citation bias. The authors who collected the studies did however not do the publication bias analysis, or they didn’t publish it. But I did it and published it just on Twitter. It is worth publishing of course, but I have been too busy. Someone else can write it up and add me as senior author.

  • te Nijenhuis, J., Jongeneel-Grimen, B., & Kirkegaard, E. O. (2014). Are Headstart gains on the g factor? A meta-analysis. Intelligence, 46, 209-215.

Authors don’t list the data in the study, but they were so nice as to share it with me. Remember to cite their study if you use it!


The Minnesota Trans-Racial Adoption Study studied 265 children adopted by White upper-middle-class parents with an average IQ of 120. Despite this similar environment, consistent racial differences were found on IQ, school grades, class ranks, and aptitude tests. At age 17 Whites scored 106, mixed race 99, and Blacks 89. 89 was also the average score for Blacks in general in Minnesota. The same difference between mixed race and Black children occurred also in in some cases in which the adopting parents wrongly thought that mixed race children had two Black parents. These results caused considerable debate. Non-hereditarians have raised objections such as the Black and mixed race children having psychological problems due to identity issues, possibly being placed in relatively poorer homes, and being adopted later and having more prior foster home placements both of which is associated with lower IQ. Hereditarians have argued that none of these are convincing explanations.[10][4]

Should probably also cite discussion in Jensen 1998 as well as Loehlin’s 2000 book chapter.

I also think Flynn has discussed it but not sure which books.

There are also three studies of adopted East Asian Children who in some cases were malnourished and adopted late. Despite this, and presumably also identity issues, they scored highly on IQ tests.[10]

Cite the primary study. It’s a very small study by Lynn, n=22 or something.

A study comparing 83 German White children with 98 mixed race children born to post-WWII German mothers and “Black” soldiers found only very small IQ differences. As for adoption studies on young children, this study has been criticized for not having any follow-up when the children were older. 20-25% of the “Black” soldiers were from French North Africa. The soldiers from the US almost certainly had higher IQ than the average Black due to Army General Classification Test excluding 30% of Blacks.[10] Furthermore, the results for the White children differed greatly for the boys and the girls. The expectation would be similar results and the large difference may be an indication of methodological/sampling problems with the study.

There is also a study finding a significant IQ difference between mixed race children born to White mothers versus Black mothers. This is argued to support an environmental explanation. Again it has been criticized for only testing the children when young. The White mothers had longer education and thus likely a higher IQ. The two groups also scored intermediately between the average IQs of the Black and White children in the study.[10]

Cite primary literature, not reviews by R&J for the 100th time.

US Blacks have on average a low degree of European ancestry. If the partially genetic explanation is correct, then one would expect that those with more European genetics would have higher IQ and brain weight. Studies using skin color as an indirect measure of the degree of European genes of Blacks have found weak such correlations. The weak correlation is argued to be explained by skin color in African Americans being only a weak indicator of the degree of African Ancestry. The results are argued to support that 50-75% of the IQ gap is explained by genetic factors. Non-hereditarians have argued that not all studies have found this correlation but non-hereditarians have argued that this was due to these studies being too small. Another argument is that the correlation may be due to societal advantages causing higher IQ for Blacks with lighter skin color. Hereditarians have argued that this is unlikely, East Asians have been discriminated against but still do not score low on IQ tests, and a study regarding if Blacks with darker skin color were more discriminated found no or contradictory results.[43][44][4]

The best resource here is the meta-analysis Fuerst did, see the presentation mentioned earlier. There is a small but positive relationship between brighter skin and outcomes, as expected by genetic hypothesis.

There are also several other kinds of empirical evidence against “colorism” (skin color differences as the cause of group differences due to factors such as racism). For example, darker skin color not being associated with more negative outcomes after controlling for IQ differences.[45]

Here the article cites an entire category of posts on HV. That’s not right. Cite specific posts. Better, write a review article.

Another method is by examining the relationship between the degree of European blood groups and IQ. Two studies from the 1970s argued that there were no such relationship. These have been criticized for using genetic makers that would have been unable to detect a relationship.[10][46]

Here it is best to cite Loehlin et al’s early review book of the evidence. They discuss it further. The books appears to be mostly forgotten by now. Google Scholar doesn’t have a proper citation for it, Amazon neither.

  • Loehlin, J.; Lindzey, G., and Spuhler, J.. (1975). Race Differences in Intelligence.

A study from the 1930s using self-reported degree of European ancestry found a small negative correlation with IQ. However, self-report has been criticized as being very uncertain and in particular regarding possible race mixing during the slavery period.[10]

Cite primary study. Also, cite Nisbett’s discussion of it for environmentalist view.

  • Nisbett, R. E. (2009). Intelligence and how to get it: Why schools and cultures count. WW Norton & Company.

Hereditarians have pointed out more recent studies in the US, Brazil, and South Africa finding that the average IQ of the populations of mixed Black and White origin is intermediate between that of Blacks and Whites. In the case of the US study explanations based on social class and “discrimination based on skin tone” were argued to be ruled out or controlled for. In certain US areas where Blacks have a low degree European ancestry their average IQ is unusually low.[10][4][3]

Again, primary sources needed.

Also, an actual study of AAs European% and NAEP scores did show expected relationships: African% -.31, Euro% .28, Amer .14. On the other hand, for Hispanics, one of the relationships was in the wrong direction: African .27, Amer -.24, Euro .11. Because there are generational changes with Hispanic scores, the AA ones weigh stronger.

A 2014 article found 31 genetic admixture studies which reported, for individuals residing in the Americas, associations between continental ancestry (e.g., European, Amerindian, Sub-Saharan African, East Asian, and Pacific Islander) and some index of educational attainment or socioeconomic status. None of the associations went in a direction opposite to that predicted by the average IQ scores of the ancestral populations. The results were argued to “support a racial hereditarian hypothesis along with others that predict a fairly internationally consistent association between continental ancestry and cognitively correlated indices of socioeconomic status such as education, income, and job prestige”.[47]

They cite Fuerst’s early post on the admixture project for individual level data. It is better to cite the presentation since it contains more studies (48). It’s also not an article, it’s a blogpost.

Regarding immigrant cognitive ability it has been stated that “The matter, of course, is complicated by migrant selectivity, ethnic identification attrition, differential breeding patterns, non-trivial environmental influence on measures, and so on. Yet, were a racial hereditarian position correct, one would expect to find, when looking across numerous countries, a robust statistical association between region of origin scores and migrant scores.” Studies on the average cognitive ability of various immigrant groups are argued to find similarities with region of origins for at least several generations after the immigration which have been argued to support a partially genetic explanation.[18]

West Indians of African origin that emigrate to the US are more successful than US Blacks. This has been seen as possible evidence for some environmental factor affecting US Blacks negatively. Another explanation is that these emigrants are not random sample of West Indians of African origin but a selected group with unusual characteristics. One 2008 study wrote that “West Indian success can be attributed entirely to the greater talent and ambition of those who choose to move. Similarly, the subset of African Americans who are voluntary internal migrants are better off than their less venturesome counterparts. Once this point is clear, it is easy to see why West Indian success offers no lessons for African American improvement.”[49]

Instead of just citing John’s summary of the spatial transferability hypothesis, it is better to cite all the studies looking at immigrants by country of origin. They invariably show the expected pattern for socioeconomic outcomes. Cognitive data is missing, so it cannot be used. However, I recently obtained permission to use the Danish draft test data, so we will see what I find.

Another factor which could possible cause a change in average cognitive ability score for a group (immigrant-derived or not) is if an increasing number of mixed-race individuals tend to identify themselves as members of this group (such as mixed Black-White individuals tending to identify themselves as Black). However, such a cognitive ability score change could reflect the average genetics of the group changing.

This is possible to investigate if one can get a sample with both generation and admixture. For instance for US Hispanics. Such a dataset may exist.

The g factor (general factor) is often seen as the underlying general mental ability that is measured more or less well by different cognitive tests. A test’s g loading, or a subtest’s g loading, refers to how well it correlates with the g factor. The g factor has been argued, based on evidence such as twin studies, to be largely genetically determined. Black-White IQ differences are largest on those tests and subtests having the highest g loadings. This has also been seen as evidence for a genetic explanation.[10] See also the sections “Spearman’s hypothesis” and “Significance of IQ and g“.

They need to cite a review of heritability of cognitive ability. Also I would avoid the phrasing “genetically determined” to avoid fueling the fire of people who hate genetic determinism — a position pretty much no researcher believes in.

The Flynn effect refers to the observed worldwide slow increase in average IQ test scores. A variety of explanations have been proposed such as increased familiarity with taking tests (thus not reflecting genuine intelligence changes) and explanations such as improved nutrition during childhood which affects the developing brain (thus reflecting genuine intelligence changes). This has been seen as evidence for that IQ can be changed significantly by environmental factors and that the racial IQ gaps may eventually disappear. However, at least in the US and some other developed nations the gains from the Flynn effect correlate negatively with g loadings and inbreeding depression. This argued to show that that at least in those countries the environmentally caused Flynn effect will not significantly narrow the largely genetically determined Black-White IQ gap.[15] A 2013 meta-analysis concluded that “It appears that the Flynn effect and group differences have different causes.”[50]

There is a newer meta-analysis with the Jensen method and the Flynn effect.

  • Woodley, M. A., te Nijenhuis, J., Must, O., & Must, A. (2014). Controlling for increased guessing enhances the independence of the Flynn effect from g: The return of the Brand effect. Intelligence, 43, 27-34.

Regression to the mean refer to the tendency for an exceptional result, such as getting all 6s or all 1s when rolling several dice, to be followed by less exceptional results (regression) that is closer to the average (mean) result. Hereditarians argue that the relatives of Blacks and White with exceptional IQs, low or high, will show predictable differences in regression due to the Black and White populations having different genetically determined average IQs. These predictions are argued to be confirmed. The same effect may possibly happen also due to environmental factors that behave similarly to IQ genes but hereditarians argue that this is unlikely and no such factors have been presented.[10][4]

I don’t know why regression phenomena would be evidence in favor of genetic models. See also Brody discussion:

  • Brody, N. (2003). Jensen’s genetic interpretation of racial differences in intelligence: Critical evaluation. The scientific study of general intelligence: Tribute to Arthur R. Jensen, 397-410.

The reaction time in response to a stimuli can be measured and tested on a variety of different tasks. Differences in reaction time are argued to be due to neurophysiological differences in the brain’s ability to process information which is also what IQ tests measure. Due to the unusual nature of the testing it is unlikely to be influenced by practice or education. Non-hereditarians have dismissed it as having a low and uncertain correlation with IQ. The related measure decision time has been similarly dismissed. Hereditarians have argued that this only applies when only one task is used. When the results from different tasks are combined, as is also done in IQ tests, the correlation between IQ and reaction time is 0.6-0.7. Although not all studies completely agree, overall racial differences are found consistent with those from IQ testing. Just as for IQ these racial differences are largest on the tasks that best measure the g factor.[10]

Primary lit. needed. In this case, they can be found via Jensen’s 2006 book:

  • Jensen, A. (2006). Clocking the Mind.

Hereditarians have argued that brain size is highly heritable and when reviewing numerous studies, then indirectly measured brain size (such as from skull measurements) have a 0.2 correlation with IQ and MRI measured brain size have a 0.4 correlation with IQ. If using the method of correlated vectors to distill g from the subtests of an IQ test, then the correlation was on average 0.63. One study found a correlation of 0.89 between g loadings and number of gray matter clusters.[53]

This review is outdated now. The numbers were artificially high due to publication bias. The best uncorrected estimate is about .25. Perhaps .30 after correction. Note however that overall volume may not be the best measure to use. Cortical surface seems to be better. Better yet, one can combine multiple measures and get a much better estimate.

  • Pietschnig, J., Penke, L., Wicherts, J. M., Zeiler, M., & Voracek, M. (2014). Meta-Analysis of Associations Between Human Brain Volume And Intelligence Differences: How Strong Are They and What Do They Mean?. Available at SSRN 2512128.
  • Ritchie, S. J., Booth, T., Hernández, M. D. C. V., Corley, J., Maniega, S. M., Gow, A. J., … & Deary, I. J. (2015). Beyond a bigger brain: Multivariable structural brain imaging and intelligence. Intelligence, 51, 47-56.

Differences in brain size between different species are associated with differences in various musculo-skeletal traits. This can partly be explained as adaptations to an increasingly larger brain. The same musculo-skeletal differences are seen between different human races which has been argued to further support the existence of brain size differences and also make an environmental explanation much more difficult.[56][57]

This I have never heard of. I will research that.

See also the research by Heitor Fernandes.

Critics have pointed to the average brain size differences between men and women and argued that there is no IQ difference which would indicate that brain size is an unreliable measure. Hereditarians have argued that some recent studies do have shown small average IQ differences between the sexes and that regardless there are clearly proven large differences between the sexes regarding narrower abilities. For example, men have on average greater spatial abilities which may be very computationally demanding and require the on average larger male brain areas. Thus, the average brain size differences between the sexes are argued to be significant regarding cognitive differences.[10]

It’s a weak argument. It’s possible for there to be a gender interaction for volume size that does not interfere with the cross-population pattern.

Also, men have greater chronometric abilities.

Another criticism is that the racial brain size differences only can explain a small part of racial IQ differences since the correlation between brain size and IQ in individuals is argued to be relatively low and that calculating the “explained variance” gives an even lower value. One response is that the genes affecting IQ must not necessarily affect brain size but may likely have various effects including non-brain size effects such as on neuronal connections, neuronal metabolism, neuronal insulation, the biochemical environment surrounding neurons, and so on. But even if racial brain size differences only explain a small part of the racial IQ differences, then this may still be very problematic for a 100% environmental theory.

There is at least one study showing that there is genetic correlation for brain size and IQ scores.

  • Posthuma, D., De Geus, E. J., Baaré, W. F., Pol, H. E. H., Kahn, R. S., & Boomsma, D. I. (2002). The association between brain volume and intelligence is of genetic origin. Nature neuroscience, 5(2), 83-84.

This is an old article. Check the citations of it to see if there are newer studies.

Direct genetic evidence basically requires two things. One is knowledge about how genes are distributed in different races. This is being rapidly achieved by continued technological developments which have dramatically lowered the costs for analyzing a person’s DNA. Several science projects have already completed analyzing or are in the process of analyzing the DNA of persons from different populations worldwide. One example is the 1000 Genomes Project which is analyzing the DNA of 2600 people from 26 different populations worldwide. 1700 had been completed as of March, 2012 and made publicly available.[64]

There is also ALFRED, used by Piffer often.

The 2015 book The Nature of Race stated that this “model has interesting theoretical and empirical support. Regarding theory, in (at least some) non-human species, climate is associated with between population variation in cognition, brain size, and heritable neural functioning (see, for example: Roth et al., 2010; Roth et al., 2012; Roth et al., 2013); cold evolved populations are, apparently, sharper. For humans, models which assume a simple relationship between selection conditioned on cognitive ability and climatic harshness over the last 60,000 years reasonably predict current global cognitive capital (see: Hart, 2007; relatedly: Grall, 2012). Regarding empirical findings, climate by way of cranial size explains a non-trivial portion of the National IQ variance (see: Meisenberg and Woodley, 2013). Generally, cognitive and cognitively related somatic differences are in agreement with the cold weather model; this model is also in agreement with the literature regarding other species.”[18]

Again, citing primary lit. is needed.

New genetic mutations, which could include ones causing higher IQ, would be more likely to arise in large populations. This can be combined with the cold temperature theory. This would explain why Arctic People who live in very cold areas, but only as small populations, did not evolve a very high IQ. Europeans and East Asian who had large populations and lived in relatively cold areas evolved a higher IQ.[3]

Plausible, but a simulation should be done to check this hypothesis. The difficulty is presumably obtaining reliable population count/density data for older times.

Studies in several countries have found that IQ vary on a north–south gradient inside the countries as predicted by the climate theories.[81][82][83]

Two counties are mentioned, one of them twice. However, it has also been found in some other counties, such as Vietnam.

Also, it was not found in India. I don’t recall whether it was found in China or not. r=.33

Another is that in relatively recent European and East Asian societies three key elements are argued to have existed: 1. Class differences that reflect differences in intellectual performance. 2. A higher level of reproductive success in higher social classes than in lower ones. 3. No barriers to downward social mobility. The lower classes are argued to have been gradually replaced by people of higher social origin (and IQ). This is argued to have caused a relatively recent increase in IQ in these societies. Other societies are argued to have lacked stratification, or been too rigidly stratified, or favoring other characteristics than IQ as causes of social mobility.[85]

Primary lit. Cite Clark’s book, not Frost’s blogpost.

The g factor (general factor) is the underlying general mental ability that is measured more or less well by different cognitive tests The existence of g does not exclude the existence also of narrower cognitive abilities (that correlate with one another and g as explained in the IQ article). g has sometimes been criticized for reasons such as being a claimed statistical artifact. Supporters argue that such attempted criticisms have a long history but that all have ultimately failed. For example, there have been many unsuccessful attempts to find important forms of intelligence that do not correlate with g. Furthermore, g is argued to have a very high genetic heritability, to be unchanged by training such as taking repeated IQ tests, to account for almost of all of the predictive ability of cognitive tests, and findings in neurobiology “establish a biological basis for g that is firmer than that of any other human psychological trait”. Furthermore, successfully discrediting g as a statistical artifact would change the situation much less than some may expect. Different races would continue to systematically differ on numerous different tests of various correlated cognitive abilities as well as to continue to systematically differ on various correlated life outcomes and achievements and there would still continue to be correlations between cognitive tests and life outcomes and achievements.[87][51]

The article needs to be rewritten in various places. Genetic heritability? It’s a tautology.

Also the anti-hereditarian IQ researcher James Flynn has rejected attacking g in order to discredit racial IQ differences and has stated regarding Stephen Jay Gould’s book The Mismeasure of Man (which attacked the hereditarian Arthur Jensen) that “Gould’s book evades all of Jensen’s best arguments for a genetic component in the black-white IQ gap, by positing that they are dependent on the concept of g as a general intelligence factor. Therefore, Gould believes that if he can discredit g, no more need be said. This is manifestly false. Jensen’s arguments would bite no matter whether blacks suffered from a score deficit on one or 10 or 100 factors.”[51]

Flynn seems to have changed his mind. See the recent discussion with Gottfredson and Turkheimer:

Differences regarding average IQ is one explanation for differences between different regions regarding early achievements such as the creation of civilizations. However, non-hereditarians have pointed out that the earliest civilizations often occurred in regions which today do not have a very high average IQ. One response is that hereditarians have never argued that IQ (or genetics) is the only explanation for differences between human groups. The earliest civilizations occurred in very fertile river valleys which at the early stages of technological development likely were the only regions which allowed the high population density necessary for the development of civilization. In contrast, factors such as the long, harsh winters and the very hard clay soils in Northern Europe for a long time prevented a high population density. Technology had to advance greatly before this changed, one example being that the technologically much more advanced heavy plough necessary to take advantage of such clay soils was first introduced in the Medieval period.[88][3]

Regarding history and HBD, the best source is:

  • Hart, M. H. (2007). Understanding human history: An analysis including the effects of geography and differential evolution. Washington Summit Publishers.

Morphological evidence based on for example statues, paintings, skeletal remains, or mummies have often been used to argue that various ancient populations or cognitive elites were quite racially different from the current populations living in the same area. Even if historical records are available, practices such religious conversion are often accompanied by changing personal and family names which often makes it difficult to identify the correct race from the name of persons. The earlier presence of racially different cognitive elites or internal dysgenetic changes in a population would likely be very difficult to detect by genetic studies of the current populations living in these areas.

No references given.

Another question is why areas such as the Sub-Saharan Africa never developed a civilization despite technology and civilizations eventually spreading to many parts of Eurasia and there being many contacts with for example the Egyptian, the Roman and the Islamic civilizations. In contrast, Amerindians (including the Maya who lived in a jungle region) eventually created civilizations (although lacking in several key aspects) despite having no contact with the civilizations and technology in Eurasia.[92]

The claims are too strong. There were pre-colonial civilizations in Africa.

Race differences in intelligence can explain why most conquests of people(s) by another people throughout human history have involved a northern people conquering southern people(s). This despite the northern regions usually being less populous due to harsher climates. For example, China was never threatened by southerners but repeatedly conquered by northerners. Similarly India was repeatedly conquered by northerners. Europeans conquering various southern peoples but not East Asians, etc.[93]

This sounds like a claim that needs to be quantified.

Lynn has argued that East Asians, despite having slightly higher average IQ, have produced much less creative discoveries and innovations in the arts and sciences than Europeans. One possibly explanation is that East Asians have lower average creativity than Europeans. Lynn argues that this is supported by Northeast Asians scoring lower on the personality trait openness to experience.[94]

Better source needed. There are country-level personality measures, but they aren’t very good.

  • Meisenberg, G. (2015). Do We Have Valid Country-Level Measures of Personality?. Mankind Quarterly, 55(4), 360-382.

A significant part of the debate following The Bell Curve was regarding how important the IQ group differences were for the future achievements of the groups in the US. The book argued for the strong importance of IQ for numerous factors such as future educational achievements, employment, income, divorce rates, and crime.

A summary of the findings are in order. One could e.g. use my image summaries of the book’s findings.

Furthermore, a high IQ has been argued to be more important for group outcome differences than for individual outcome differences. A 2011 study presented “four different channels through which intelligence may matter more for nations than for individuals: (i) intelligence is associated with patience and hence higher savings rates; (ii) intelligence causes cooperation; (iii) higher group intelligence opens the door to using fragile, high-value production technologies; and (iv) intelligence is associated with supporting market-oriented policies.”[95]

A claim which is strongly born out by my studies of group-level correlates for immigrants in north European countries and for regions and states. E.g. correlation of national S and IQ is .86 (weighted using sqrt(pop), identical without weights).

There are also some countries that are racially different from the surrounding countries and that perform very differently from these surrounding countries on various variables. Examples include Haiti (predominantly Sub-Saharan Africans) and Singapore (predominantly East Asians).

Israel is another example.


In general, the article is fairly comprehensive, but lacks many primary sources. Right now it is mostly a summary of claims made by Rushton and Jensen’s two review articles, Lynn’s books and Fuerst’s book. The text needs a proofreading as there are a number of confusing uses of terminology as well as many errors. The writer is not a native.

I hope that my comments here can lead to an improved article. Or better, someone should write an updated review of the race and intelligence question. Since a lot of research has been done in the last year or so by Fuerst, Dalliard, Malloy, Hu and myself, it would probably have to be written by one or more of us.