Source: TW
A VERY BASIC INTRODUCTION TO THE HEREDITARIAN POSITION ON RACE AND IQ.
Introduction
PART 1: Introduction.
Persistent differences in the average IQ of different groups (including races) have been the subject of study by scientists for decades, although most of this research has been conducted quietly and out of the view of the public due to its sensitive nature. The existence of gaps in average IQ between the races is now firmly established, and is no longer controversial among scientists who study human intelligence.
The biggest remaining controversy is over the reason for the gaps: Are they due to genetic differences between groups or to differences in the environments to which each is exposed? But even this central controversy can be more narrowly expressed now that the majority of scientists have come to believe that both genes and environment play at least some role in group IQ differences, and the question has therefore become:
How much of a particular gap — for example, that between Americans of African ancestry and Americans of European ancestry — is due to the influences of the environment (“nurture”), and how much to differences in genetic make-up (“nature”)?
Stated differently: Most scientists studying human intelligence believe race IQ gaps are due to a combination of both genetic and environmental factors. The only real controversy is over the share of each. The share of each may be anywhere from near-total to negligible.
Before going any further, I should state that these threads will be of more interest to intelligent laymen who have had limited exposure to the subject, and will be of less interest to those already possessing a strong familiarity or expertise.
Please note that because of its small population size, I’m probably going to be leaving out of my discussion Ashkenazi Jewish IQ, believed by intelligence researchers to be higher than the average of any other group, with estimates ranging from 108 to 118.+++(5)+++ This extraordinary Ashkenazi cognitive advantage no doubt explains why a group that comprises well under 1% of the world’s population has received about a quarter of all science Nobel Prizes, an astonishing accomplishment.
Also, a word about “Asians”. This obviously includes many different groups, with widely-varying average IQs (from 80 to 108), but in the US includes mostly northeast Asians (i.e., Chinese, Japanese, and Koreans) and south Asians (i.e., from the Indian subcontinent). Apart from Ashkenazi Jews, northeast Asians have the highest average IQ in the world (around 105). South Asians, on average, fare more poorly, but those who emigrate to the US (especially from India) tend to have IQs well above the south Asian average.+++(5)+++
Most of the discussion will relate to the IQ gap between Americans of European ancestry and Sub-Saharan African ancestry because that’s where most the data are and, as an American, this gap has greater social, cultural, political, and economic implications than any other. I probably won’t be discussing Hispanics — it’s a US Census category and not really a useful one for genetic ancestry, as it may include, for example, a Colombian-American with 90% European ancestry and a Dominican-American with 90% African ancestry.
Also discussed only very briefly will be world IQ scores, including Sub-Saharan African scores, which are the lowest in the world, lower than African-American scores. When the term “environment” is used in these threads I mean everything other than genetic influences — things like education, socioeconomic status, parental influence, life events, prenatal factors, dietary deficiencies, illnesses, random events, etc.
Note that I’m going to be using the terms “IQ” and “intelligence” interchangeably in these threads because the great majority of researchers believe that IQ tests measure intelligence accurately and also because using “IQ” uses fewer Twitter characters (no small thing).
Finally, I’ll state that although I retain an open mind, I believe that the arguments for genetic factors in the race IQ gaps are generally more convincing than those for environmental ones; therefore I believe genetics play a not-insignificant role in these differences. As a result, the threads will usually be framed from a hereditarian-leaning perspective. At the same time, however, I’m not uncritical of occasional shortcomings in the hereditarian explanation, and will be pointing these out.
What Is General Intelligence?
Before proceeding with these threads, it’s important to establish what it is that IQ tests measure. It’s called “general intelligence,” and it’s often just referred to as “g”. General intelligence is “the ability to reason, plan, solve problems, think abstractly, comprehend complex ideas, learn quickly.” Scientists have identified a variety of specific cognitive abilities, and “g” is “the pinnacle of the hierarchical model” of those abilities.
“[P]eople who do well on one mental test tend to perform well on all of them; that is, all mental tests correlate with one other. All measure mostly the same underlying ability factor [‘g’], no matter what their manifest content or purpose.”
“Mental abilities differ in their generality-specificity, and the general mental ability factor, g, is the most general of all.”
“[T]here is a strong consensus within psychology that g can be measured objectively using appropriate methods of factor analysis… g is a universally found statistical regularity.”
Advances in genomics and statistical modeling now offer scientists an opportunity to begin to directly find “g” at a genetic level — that is, identify an underlying reality, at a genetic level, between general [‘g’] and specific aspects of cognitive ability.
A study (pre-print) conducted in 2019 was the first to seriously make the attempt, and it found that “a genetic g factor accounts for 58.4%… of the genetic variance in the cognitive traits, with trait-specific genetic factors accounting for the remaining 41.6%.”
Intelligence research quality
The quality of published scientific research into human intelligence is arguably higher than in any other area of psychology. Two of the most reliable indicators of research quality — statistical power and replicability — are higher in published intelligence research than they are in other areas of psychology. A meta-analysis found that the median power in intelligence research is considerably higher than the median power across psychology as a whole (53% versus only 23%), as well as considerably higher than in the social and behavioral sciences as a whole (24%). That same meta-analysis also found that the area of intelligence research focusing on group (e.g., race) differences is the most high-powered of all (62% median power). Another large meta-analysis found that the field in psychology which focuses most on genetic influence of psychological traits (behavioral genetics) had by far the most adequately-powered findings in all of psychology (over 70%).
Sample sizes in IQ studies (especially when using proxy data like standardized test scores) are larger than in any other research field in psychology, sometimes containing tens of thousands of individuals. All together, there are millions of data points across IQ literature. Unlike other research areas in psychology and other social sciences that employ tiny samples and experimental designs and which generate anemic effects that don’t replicate, most of the well-known findings in IQ research seem to endlessly replicate with robust effects.
Replication of research study findings is also usually higher in other (non-psychometric) areas of intelligence research — for example, research in the field of cognitive psychology had a higher replication rate (50%) than that in the field of social psychology (25%).
Steven Pinker:
“The research replicability crisis in psychology doesn’t apply to IQ research — these studies have huge sample sizes and replicable results. But people hate the message.”
This rigor is important to remember because left-wing activists often resort to calling any research finding that doesn’t fit their blank slate model “pseudo-science.” Intelligence research is among the least “pseudo” in all social science.
Rigor is also important because activists have
“often asserted that, when it comes to taboo topics like race, genes and IQ, scholars should be held to higher evidentiary standards… because of the harm that could result if their findings became widely known.”
Intelligence research draws from a variety of sciences, including behavioral genetics, psychometrics, neuroscience, molecular and evolutionary biology, population genetics, and cognitive, behavioral, developmental, educational, and evolutionary psychology. Intelligence theory
“is more fully developed and more mathematically sophisticated than almost any other psychological construct. More is known about the underlying cognitive, genetic, and brain processes for intelligence than for any other complex psychological construct.”
One well-known scientist:
“IQ tests are among the most reliable, predictive measures in psychology — one of the field’s crowning achievements. If IQ isn’t a valid concept, no concept in psychology is valid.”
Black white gap survey
Based upon results from a survey published in the scientific journal Intelligence in 2020, it appears that most intelligence researchers lean toward the view that genes have more influence on the US black-white IQ gap than the environment.
15 respondents estimated that genes account for 50% of the gap, 37 estimated it’s responsible for more than 50%, and 34 less than 50%. In other words, 60% of the respondents believed that genetic differences accounted for half or more of the black-white IQ gap. The results clearly expose the lie of activists on the left who claim scientists don’t believe genes play a role in the US black-white IQ gap. More than 60% of respondents here say it accounts half or more of the gap. My feeling is that these survey results may understate the actual support for a genetic hypothesis — due to the sensitivity of the issue, I suspect those with the more politically-acceptable pro-environment views were more likely to respond to the survey.
Wondering why you didn’t know this before? It’s because the self-protective code of conduct in intelligence research is to keep politically volatile findings and views inside a “circle of trust” — it’s “safe” to discuss them with colleagues and peers, but not with the “public”.+++(5)+++ Scientists who have ventured outside this “circle of trust” have been met with mobs (virtual and actual); formal condemnation; loss of employment, contracts, and grant money; physical violence; and removal from boards and societies.
Significantly, the political views of the experts in the survey clearly lean heavily left. So, it appears exceptionally unlikely that their science is driven by “white supremacist” views, as left-wing activists often absurdly charge.
Finally, it’s worth noting that the three scientists conducting the survey are respected in the field and had earlier jointly published other surveys of intelligence experts in peer-reviewed scientific journals.
I’ll quickly mention the results of a 1984 survey of experts — conducted before the hereditarianism had gained its recent momentum and before the human genome had been mapped, and which was marred by many respondents’ unwillingness to respond to certain questions. The clear majority of those committing to a response, however, indicated they believed that both genes and the environment played a role in black-white IQ differences. Unlike the more recent survey, no opportunity was given for a more precise response (e.g., “50/50”).
Also surveyed were the political orientations of the experts, and, like the 2013 survey, the results indicated a left-leaning orientation. That 1984 survey became the basis of a book published in 1988, in which the authors pointed out that the results
“contrast greatly with the coverage of these views as represented in the media, where the reader is led to draw the conclusion that “only a few maverick…”… ’experts’ support the view that genetic variation plays a significant role in individual or group difference.”
They wrote that few scientists are willing to risk personal attack by publicly speaking about the field’s mainstream (but non-PC) scientific findings.
David reich
Thread by @PaoloShirasi on Thread Reader App – Thread Reader App Paolo Shirasi
PART 7: David Reich and Population Genetics
Underlying much research into group (e.g., race) intelligence differences is the following general principle: Any trait influenced by genetics — cognition being a leading example of one — will vary across large populations (such as racial groups). Harvard’s David Reich is perhaps the world’s leading population geneticist. He wrote:
“Since all traits influenced by genetics are expected to differ across populations… the genetic influences on behavior and cognition will differ across populations, too.”
Reich again:
“[A]s a geneticist I also know that it is simply no longer possible to ignore average genetic differences among ‘races’.”
David Reich, one more time:
“I am worried that well-meaning people who deny the possibility of substantial biological differences among human populations are digging themselves into an indefensible position, one that will not survive the onslaught of science.”
Reich chooses his words very carefully: “Substantial biological differences.”
In summary: There are genetic variations across populations of humans, which are caused (as in every other species) by natural selection, and some of these variations are “substantial,” and do indeed overlap with our idea of race. Reich is telling us that our most powerful, advanced, and reliable scientific tools and technologies for analyzing genes are showing us that the genes of humans cluster in ways that approximate notions of continental and sub-continental ”race.”
Reich is producing ground-breaking work at a blistering pace that is unsettling to IQ environmentalists in the race and IQ debate.
NYT: “Most scientists felt lucky if they published one or at the most two Nature papers in a lifetime. Reich [is] publishing three or four a year.”
Human populations likely differ in psychological and cognitive traits for the same reason they differ in other traits: They faced different selective regimes. The human brain is not somehow immune to selective forces.
So, to sum up from recent threads:
(A) There are significant and agreed-upon racial IQ gaps;
(B) most researchers see a strong genetic role in these gaps; and,
(C) population genetics anticipates differences in traits (including cognition) across different population groups.
Individual differences
No one disputes wide-ranging differences in intelligence between individuals. There is now a near-consensus view among scientists that INDIVIDUAL adult variation in intelligence is primarily explained by genetic differences between individuals.
By mid-adulthood, it’s estimated that anywhere from 50% to nearly 90% of the individual variance in intelligence is attributable to heritable factors. No major psychological trait is as heritable as intelligence — a stunning statistical correlation up to 0.88 in later life. A 1994 statement by 52 scientists played it safe in declaring that “genetics plays a bigger role than does environment in creating IQ differences among individuals.” Since that date, the scientifically-estimated influence of genes on IQ has, if anything, only increased.
Cognition seems more generally heritable than other traits. In the largest medical study ever conducted (45 million people), cognitive conditions were the likeliest of all to have genetic influences, and the least likely to have shared environmental influence. The full effect of heritabilty occurs by mid-adulthood. This is due to something called the “Wilson Effect” — a tenet of behavior genetics that, as people age, their genes exert more influence over IQ, and the influence of environmental factors decreases.+++(5)+++ When children are younger, environmental factors have a more significant influence on IQ, but especially after puberty this influence starts to recede and genes move toward dominance.
The heritability of intelligence increases linearly from around 20% in infancy to 40 to 50% in adolescence, then 60% in young adulthood, upward to 80% (or even higher) in later adulthood, but probably declines somewhat after age 80.+++(5)+++
There is a majority view among scientists
“that IQ is significantly heritable beginning at least at age 7.. that by age 10 genetic variance is larger than shared environmental variance… [and] heritability increases with age until late adulthood.”
The increase in the heritability of intelligence is particularly significant between early childhood and a person’s teenage years.
“The findings supporting the Wilson Effect are highly robust…
They do not depend on a single design,… are much the same across a variety of Western industrialized countries,..
[and] many of the samples are quite comprehensive so their results probably apply broadly.”
The Wilson Effect points to a genetic explanation for intelligence.
If the environment is the main engine driving intelligence then we would expect that its importance would increase with age due to its cumulative effects. But, in fact, the opposite happens.
Astonishingly, some “common factors” in cognition, such those mediating certain “executive functions,” have been found to be as much as 99% heritable (i.e., “nearly entirely genetic in origin”). These traits can interact with and/or underlie general intelligence (“g”). The heritability of intelligence is basically the same for all people, from the dimmest to the brightest —
“a recent study of 11000 twin pairs found that the top 15% of the intelligence distribution was as heritable (0.50) as the rest of the distribution (0.55).”
The fact that intelligence is one of the most stable behavioral traits, yielding a correlation of 0.63 in a study of people tested at age 11 and then again at age 79, shouldn’t be surprising given its predominantly genetic basis.
However: Just because there’s an large genetic basis to individual differences in IQ does not necessarily mean that any genetic basis to group (e.g., racial) differences would be as large. Individual and group IQ differences are related but separate issues. But as Richard Haier, a leading intelligence scientist, wrote:
“Since there is overwhelming evidence that genes influence the former [individual differences], it would not be unreasonable to hypothesize that genes at least partially influence group differences.”
Genes
PART 9: Identifying the Genes — GWAS and PGS
There are still educated people who say that intelligence is a function of hard work and effort, or, as some on the left claim, purely the product of privilege or oppression, despite the fact that scientists have known for decades that IQ is mostly genetically-influenced. In the previous thread, I described the scientific consensus that differences in intelligence between individuals are primarily due to genetic differences between them — by mid-adulthood, heritability explains about 60-80% of the IQ differences between individuals.
Now, just in the last several years, there have been scientific advances that have enabled scientists to identify some of the specific genetic variants that are associated with intelligence — the single biggest development in intelligence research in decades. In other words: Scientists can now identify precisely those intelligence-influencing genetic variants which blank slate fundamentalists have been telling us for decades can’t possibly exist.
Well over a thousand specific genetic variants associated with intelligence have been identified over the past few years using Genome-Wide Association Studies (GWAS), and then had their predictive value quantified by polygenic risk scoring (PGS or PRS). In GWAS you take DNA samples from many individuals and see which gene variants tend to be correlated with the variable of interest (e.g., intelligence).
The genetic component of IQ is spread over hundreds of DNA regions, each making a tiny contribution (“many genes each to small effect”). By analyzing the specific genes influencing intelligence, scientists can calculate “polygenic risk scores” for predicting IQ.
“Each [genetic] variant identified by [GWAS] as statistically associated with [IQ] can be assigned an effect size, representing the estimated magnitude of the increase in the trait… or liability for the trait… that is associated with possession of a copy of the variant.”
“[A] polygenic score for a [person’s] genome represents an aggregation, usually in the form of a sum, of the estimated effect sizes of the genetic variants in [that person’s] genome.” “Whereas 20 years ago, researchers were working with sample sizes of a few hundred… today genomics allows for an analysis of hundreds of thousands of genetic markers in hundreds of thousands of individuals…”
The largest GWAS/PGS study have several million individuals. Rather than measuring the impact of genes in the aggregate, scientists are identifying specific genes that affect specific traits. GWAS/PGS are becoming so powerful that in just over a couple of years polygenic scores for educational attainment (which is closely related to intelligence) became among the most powerful predictors in the behavioral sciences.
Think about that: A predictor of intelligence can incorporate the identity of some of the ACTUAL GENETIC VARIANTS associated with intelligence. These predictors are unique because they can “predict outcomes just as well from birth as later in life,” and they are “causal predictors in the sense that nothing in our… environment can change the differences in DNA sequence that we inherited…”
Also, “correlation doesn’t necessarily mean causation” doesn’t apply with GWAS. Because GWAS is a fully longitudinal study (your genome doesn’t change over the course of your life), correlation DOES equal causation (the genes associated with a trait do in fact cause that trait).+++(5)+++
Robert Plomin:
“Polygenic scores are causal predictors in the sense that nothing in our brains, behaviour or environment can change the differences in DNA sequence that we inherited from our parents.”
One cannot overstate the progress GWAS has made. Between 2013 and 2018 the number of genetic variants associated with intelligence increased from 3 to 1,271 (and the number will continue to increase). Companies are beginning to use the science to create tests to help parents predict the IQ of pre-born children, and tests for adults should soon follow as GWAS/PGS improves.
In summary: GWAS has the potential to revolutionize how scientists do research on race, genes, and intelligence.
Source: TW
This might be a good time to revisit GWAS, an advance in genomics which has identified the specific genetic variants associated with IQ and then had these variants’ IQ-predictive values quantified by polygenic risk scoring (PGS or PRS). GWAS is going to revolutionize how scientists do research into race and intelligence. Ultimately, as more genetic variants become associated with IQ, predictive models (such as PGS) improve, and sample sizes become even larger and more diverse, scientists may be able to use these tools to settle the race and IQ debate once and for all.
If gene variants associated with IQ significantly vary between groups (as we already know they do between individuals) — e.g., certain of these genes have a higher frequency in some races than others — then this would support a genetic hypothesis for race IQ gaps.
A hereditarian’s view:
“[GWAS/PRS] lets us expect that it is only a matter of time before geographically differentiated patterned distributions of coupled gene variants will be identified to further support the existence of biologically based individual and group differences.”
“Populations differ in frequency of SNPs… associated with intelligence, in types of y-DNA chromosomal haplogroups and in genetic distance measures…. [research] results will most likely [defeat] …claims of gene-free population non-variance for [IQ].”
At least one researcher (Davide Piffer) has already attempted to use PGS (with a novel approach) to predict group intelligence. However, many researchers, including some sympathetic to the hereditarian position, appear to believe that such efforts may be premature. Whether premature or not, Piffer found “population differences in PGS which correlated with estimates of average population IQ and students performance on standardized tests,” exactly what you would predict with a genetic explanation for race IQ gaps.
Another 2019 study also shows exactly what a hereditarian model would predict: “Genetic correlations computed using GWAS summary statistics are very similar to phenotypic correlations."
In other words, those observable characteristics (like race and measured IQ) that scientifically correlate well, also genetically correlate well. Part of the problem that early PRS investigators are having is that there are fewer genetic data on persons of African descent. The medical research establishment, which uses GWAS/PGS to conduct important research has become particularly vocal about this problem. The attenuation of predictive power when blacks are PGS subjects is fairly significant — up to 85% in one large, rigorous study on genes and educational attainment. (But of course this attenuation also supports the argument of meaningful genetic differences between races.)
Another (and possibly more important) problem for IQ researchers is the expectation that those institutions which possess the genetic data necessary to conduct investigations into race and IQ may be unwilling to release them because of the sensitive nature of the research. I’ll be discussing the political and cultural context of intelligence research in some detail in later threads. Suffice to say for now that pressure from left-wing activists has successfully impeded much of the investigation into the subject since the 1960s.
Twin studies
PART 10: Twin and Adoption Studies and the Genetic Basis of Individual Differences in IQ
Previously I discussed how genome-wide association studies (GWAS) and polygenic risk scoring (PGS or PRS) were revolutionizing research into human intelligence. GWAS is identifying some of the actual genetic variants responsible for individual differences in IQ. But before GWAS, how were researchers so certain that individual differences in intelligence are predominantly influenced by genes? (Remember that the current estimates of genetic influence on individual IQ differences existed before GWAS/PGS.)
Before GWAS/PGS, it was twin studies and adoption studies more than anything else that supported a genetic explanation for most of the individual variation in IQ. Significantly, not only do twin studies and adoption studies provide insight into INDIVIDUAL IQ differences, they also provide powerful evidence for the reasons behind for GROUP (race) differences in intelligence. (This will be discussed in a later thread.)
The largest meta-analysis ever of twins (14,558,903 pairs), found that substantial twin resemblance in cognition and most other traits was
“solely due to additive genetic variation… the data are inconsistent with substantial influences from shared environment…”
Identical twins tend to have the same IQ, even when raised apart — the correlation of adult IQ between them if raised apart is .75 to 0.80. The correlation between two adults of different biological parents raised in the same household is almost zero in adulthood. Studies have found childhood IQ correlations of up to .88 for identical twins of the same sex raised in the same family.
But siblings who are raised in the same family who are not related to each other have a correlation of only .04 by adulthood.+++(5)+++ A general principle of intelligence research:
“The resemblance between general cognitive abilities increases proportionally with genetic proximity (i.e., the more similar the genetic makeup of two people is).”
Adopted children will much more closely resemble their biological (i.e., “birth”) parents in cognitive abilities than their adoptive parents. The discrepancy becomes fairly remarkable as times goes on — as the influence of the environment wanes and that of genes increases.
The correlation between adopted brothers and sisters (genetically unrelated people raised together) shows that the IQ correlation between them falls to zero in adulthood — suggesting shared childhood environment has no impact on intelligence in adulthood.+++(5)+++
Another graphical representation showing the dominance of genes. You can see the correlation between the adoptive parents’ IQ and the adopted child’s IQ gradually drops to zero by the age of 12, while that between parent and biological child increases until 12.
“Unrelated siblings’ IQ correlation decreases from .31 in childhood to .19 in adolescence and (based on a summary of the literature) from .25 in childhood to .00 in adulthood.”
“On a broad range of intelligence tests, monozygotic twins reared apart are approximately as similar to each other as one individual tested twice. The genetic component is massive. Adopted children score closer to their birth mothers than their adoptive mothers.”
Transracial adption
PART 11: The Hereditarian Explanation –- Transracial Adoption Studies
Over the next few threads, I’ll talk about some of the findings and areas of research which hereditarians cite in advancing their argument that genetic differences explain much if not most of the race IQ gaps. In the previous thread, I discussed how adoption studies have helped establish that individual differences in IQ are mostly influenced by heritable factors.
Adoption studies are also a good starting point for talking about race differences in IQ. Adoption studies that include racial variations of adopted children and adoptive parents — such as transracial adoption studies — are particularly compelling.
In transracial adoption studies, such as the the Minnesota Transracial Adoption Study, the familiar average IQ gaps between children of different races typically remain even after their adoption by high-IQ white upper-middle class parents able to provide optimal environments.
Studies show that by the end of childhood, Asian adoptees end up with higher average IQs than white adoptees, white adoptees with higher IQs than mixed-race (mixed black-white) adoptees, who end up with higher average IQs than black adoptees.+++(5)+++
Note the familiar linearity of the results, and remember that the children in the study were all raised in optimal environments due to the education level and wealth of adoptive parents. The Minnesota Transracial Adoption Study is the largest and best-designed study.
Some environmentalists argue that the pre-adoption environments of the black children may have caused their lower IQ outcomes at age 17, but multiple studies involving extreme cases of even severely malnourished infants and toddlers counter this argument.
It’s worth noting that East Asian adoptees from the 1960s through 1980s were more likely to suffer early deprivation and malnourishment than US black adoptees, and yet quickly outperformed black and white children once established in their adopted families.+++(5)+++
“Korean and Vietnamese children adopted into white homes, even though as babies many had been hospitalized for malnutrition, nonetheless grew to have IQs 10 or more points higher than their adoptive national norms.”
Environmental theories cannot explain this finding. There are a few other transracial adoption studies that have less clear-cut results, but these have tended to be smaller and less rigorous and the IQ testing occurred at ages before much of the Wilson Effect had kicked in.
Brain studies
PART 12: The Hereditarian Explanation –- Brain Studies
Hereditarians are quick to point out that one of the most reproducible findings in the science of human intelligence, across dozens of studies, is the association between brain size and intelligence.
Bigger brains (or larger cranial capacity or volume of the interior of the cranium) associate with higher intelligence, on average. Larger brains contain more neurons and synapses and process information faster. T his finding is important in the race and IQ debate because average brain size has been found to vary by race — northeast Asians have the largest brains, followed by Europeans, and then Sub-Saharan Africans. The mean brain size of races correlates with their average IQs.
A massive 2018 study using UK Biobank data from 13,608 adults, which was controlled for sex, age, height, socioeconomic status and race found a “robust association between total brain volume and fluid intelligence… consistent with previous findings.”
Another even larger (29,004 individuals) and more recent age- and sex-corrected study from 2019 also found an association between total brain volume and general intelligence. These are just two of the more recent large well-controlled brain size studies. There are many others.
The correlation between brain size and intelligence in MRI studies is typically a “moderate” one between .30 and .40. It’s important to note that modern brain studies use advanced imaging and scanning technologies to accurately measure brain size, and study sample sizes can be dozens or even hundreds of times larger than in studies conducted only decades ago.
Many of the specific genes associated with brain size have now been identified in a large genome-wide association study (GWAS). Others will likely be shortly identified. Even more intriguing, a 2019 GWAS meta-analysis, one of the largest ever conducted (involving over 50,000 individuals), identified 346 new genes associated in the overlap between brain volume and intelligence.
Estimates from twin studies indicate that genes contribute from 50% to 90% of the variance to both cranial capacities (based on external head size measures) and to brain volume measured by MRI.
“[M]any specific brain structures [are] strongly heritable (i.e., 80% or more)… [there is] a strong heritability for brain volume (greater than 80%).”
Significantly, recent studies, drawing upon multiple data sets, find that the correlation between bigger brains and higher intelligence is a causal one. As previously mentioned, these findings are significant because average brain size appears to vary by race, with most studies finding NE Asians having the largest cranial volume, followed by whites, and then sub-Saharan Africans -– the same order as in their average IQs.
Brain size differences between races are present at birth, which “implies some prenatal and biological mediation.”
One fairly recent study found pre-natal racial differences in head circumference “evident at 21 weeks of gestation.” Brain size differences between races cannot be adequately explained by poverty and nutrition because they were found even when these variables were dramatically worse in China and Korea than in Europe and the US (which demonstrated smaller brain sizes than NE Asians).
Not only do whites have larger brains, on average, than blacks, but also different regions of the brain vary in size by race — for example, “the OFC and amygdala appear to occupy a significantly greater proportion of the total cerebral volume” of blacks.
Brain structure has also been found to be different between whites and northeast Asians.
Admixture studies
PART 13: The Hereditarian Explanation –- Admixture Studies
So far I’ve talked about how the overwhelmingly genetic basis of individual differences in IQ, along with findings from transracial adoption studies, brain studies and population genetics, tend to support a hereditarian (genetic) explanation for the IQ gaps between races.
In this thread, I’ll discuss yet another argument used by hereditarians: Racial admixture studies which show that mixed-race populations (i.e., children of one black and one white parent) have, on average, IQs between the averages of white and black populations.
You’ll remember from the previous thread on transracial adoption studies how mixed-race children adopted into upper middle-class white families fare better on average than the black adoptees, but worse on average than the white adoptees on IQ test performance. A well-known example of an equidistant racial admixture study conducted in South Africa found that the average IQ of the mixed-race population averaged about 85, perfectly intermediate to the black average of 70 and the white average of 100. Side Note: You’ll notice the low black South African IQ.
Hereditarians argue that US blacks have a higher average IQ than sub-Saharan Africans at least partly because of their racial admixture (about one-quarter of the gene pool of African-Americans is European). +++(5)+++
“Data from large, publicly available, archival data sets… show that groups of mixed-race individuals have mean scores intermediate to unmixed groups of Blacks and of Whites,”
according to a 2002 study.
Relatedly, the wide range of average IQs among South American nations also provides some insight into racial admixture. As one would expect under the hereditarian hypothesis, those nations comprised of mostly-white admixture (e.g., Argentina, Chile) have higher average IQ scores than those comprised of mostly-non-white admixture and population (e.g., Bolivia, Ecuador).
Another interesting finding from admixture research is that
“[c]hildren wrongly believed by adoptive parents to have had two black biological parents [they actually had mixed black-white parentage] had the same IQs as children correctly classified as of mixed race.”
A 1961 admixture study finding that out-of-wedlock mixed-race children of African-American GIs and German women had the same IQ as out-of-wedlock children produced by white American GIs and German women is often cited by environmentalists. However, this study is severely flawed for several reasons:
- First, the children were very young, tested before puberty, mostly before the full “Wilson Effect”;
- second, a quarter of the “black” fathers were actually North Africans (not sub-Saharan African); and…
- …third (and most importantly), the black GIs were unrepresentative of the average African-American due to “rigorous pre-induction selection” by the US Army which was based upon — you guessed it — the results of intelligence testing. This testing (i.e., the Army General Classification Test) removed the lowest-scoring blacks for selection into the service (about one-third of all blacks taking the test). Only about 3% of whites were removed. This had an obvious equalizing effect.
Finally, while GWAS investigation into admixture and IQ is just beginning, a decent-sized 2019 study involving an admixture analysis found that “European ancestry was a consistent predictor of cognitive ability,“ consistent with the findings discussed above. The study’s bottom line interpretation of the findings supports an estimate of roughly 60% of the black-white gap being genetic, which is what some hereditarians have long suspected it to be.
Misc findings
PART 14: The Hereditarian Explanation –- Miscellaneous Findings
Over the last few previous threads, I’ve discussed some of the more well-known scientific findings and areas of research cited by hereditarians to argue for a genetic basis for racial differences in IQ. Now I’ll very briefly discuss some less well-known ones.
Here’s one: Hereditarians assert that the consistency of race IQ patterns across countries, as well as testing regimens, strongly suggests a genetic influence. For example, northeast Asians have average higher scores on IQ tests than whites everywhere — in Asia, US and Europe.
Another one: Scientists have found that racial differences in traits (not just cognition) are most pronounced with respect to those abilities that are more heritable, and IQ is the most heritable of all psychological traits.
Even within the hierarchy of all cognitive traits, the more heritable a particular cognitive ability is, the larger the black-white gap in that ability tends to be. This strongly suggests a genetic explanation.
Mean black–white differences in IQ are greater on IQ subtests that are more responsive to heritable differences than they are on more “culturally malleable” subtests.
Third: A statistical concept called “regression to the mean” is also often cited as evidence to support a genetic explanation. An important characteristic of IQ scores of individuals is that they tend to ultimately move toward the IQ average of their parents. “The sum of all these individual regressions is going to add up to regression towards the mean of the source population” — for example, a person’s race.
So, for example, black and white children of parents with IQs of 110 move toward different averages — whites ultimately toward 100 and blacks ultimately toward 85. Hereditarianism can explain this phenomenon much more easily than any environmental theory.
One meta-analysis concluded that environmentalists
“cannot predict these results but must argue that cultural factors somehow imitate the effect theoretically predicted by genetic theory, which have also been demonstrated in studies of physical traits and in animals.”
When Arthur Jensen tested the theory in a large sample of black and white siblings “the regression line showed no significant departure from linearity throughout the range of IQ from 50 to 150, as predicted by genetic theory but not by [environmental] theory.”
A fourth pro-hereditarian finding is “Spearman’s Hypothesis,” the observation that the more “g-loaded” a test is — i.e., the better it captures general intelligence (“g”) — the greater the group gap between blacks and whites.
“The rationale is straightforward. If g is the main source of between- and within-group differences, then there should be a positive relationship between a given test’s g loading and the mean Black–White group difference on that test.”
This is in fact the case. A 2013 meta-analysis:
“Black/White differences in mean IQ have been clearly shown to strongly correlate with g loadings, so [there will be] large group differences on subtests of high cognitive complexity and small group differences on subtests of low cognitive complexity.”
Black-white differences, for example, are larger on the highly g-loaded Backward Digit Span test than on the less g-loaded Forward Digit Span test. This positive correlation between a race IQ gap and “g-loadings” is what one would expect if IQ differences were predominantly genetically conditioned rather than culturally conditioned.
If IQ was socially/culturally conditioned by environmental forces, we would expect the black-white gap to be greatest on the most culturally-loaded tests, but, in fact, that’s not what occurs.
Unsurprisingly, the same pattern occurs when comparing whites to northeast Asians: The more g-loaded the test, the greater the mean group differences are between northeast Asians and whites (in favor of Asians).
Fifth: Another theory used to support a genetic hypothesis relates to an eerie symmetry in race differences in other (non-IQ) “life-history” traits: East Asians and blacks consistently fall at two ends of a continuum with whites intermediate on 60 different measures. These measures include things like personality, maturation, social organization, and reproduction. For example, black children sit, crawl, walk, and put on their clothes earlier than whites, and whites earlier than East Asians.
Finally: There’s a statistical argument based upon d-value. Given the 15-point IQ gap between US whites and blacks, “the environmental hypothesis requires a mean environmental difference between groups to be at least d = 1.414.”
This does not seem plausible since socioeconomic factors (which bundle most of the important environmental variables) average a total of about d = 0.658, far short of 1.414. (Also, it’s important to note that a large part of socioeconomic status is itself related to genetics.)
Summary
PART 17: The Hereditarian Explanation — Summary
In summary, even apart from GWAS/PGS, a compelling case for a genetic explanation for race IQ gaps can be made from the overwhelming evidence for the role of genes in IQ differences between individuals, the consistent large gaps in IQ and academic testing performance, brain size studies, physical brain characteristics, racial admixture studies, evolutionary and population genetics principles, brain imaging studies, transracial adoption studies, regression to mean findings, consistency of patterns in worldwide IQ, the prominence of racial differences regarding the most heritable abilities, Spearman’s Hypothesis, and the symmetry of race differences in other (non-IQ) “life-history” traits.
Taken together, all of these findings make a strong case for a genetic explanation. Genetics studies using GWAS and other tools remain the most likely current path to settle the race and IQ debate once the necessary genetic and IQ data are available and some of the technical, statistical, and methodological issues can be worked out.
Arthur Jensen: “For a question so complex… no one fact,… nor indeed any single line of evidence, can hope to be determinative. Rather, resolving the issue requires examining several independent lines of evidence.”
The question should always be: Where does the preponderance of the evidence take you? More toward an environmental or a genetic explanation?
However, some hereditarians argue that it’s the COMPARATIVE WEAKNESS OF THE ARGUMENTS AND REBUTTALS MADE BY ENVIRONMENTALISTS that tips the scale convincingly toward a genetic explanation for racial IQ gaps. Perhaps the biggest selling point of hereditarianism is that it provides a theoretical framework which can account for millions of data — exactly what a successful theory should do — while environmentalism seems incapable of that task.
Over the next several threads I’ll discuss the problems with the claims of IQ environmentalists, and will rebut their arguments.