Survey of Expert Opinion on Intelligence: Causes of International Differences in Cognitive Ability

There's a series of ads running on TV during the NCAA Men's Basketball tournament, that featured Spike Lee, Charles Barkley, and Samuel Jackson, in one car. How much money would you bet that I can't find three guys famous enough to have their pictures common on the net who each look more like one of those guys than those guys look like each other, and are not black?

And height variations between populations are not, according to his research. See how that works? It's the global/local distinction again - common throughout science.

No, it's not. It was a quote I remembered from a New Yorker profile of the man and his research. He was properly careful - he did not say that genetics never made a difference at the population level in populations he had not studied, merely that his own research had dismissed the possibility in the populations he had studied. Which involved the major native and immigrant population groups of every inhabited continent except Australia.

Yes. He established, for example, that median height in a society was an excellent proxy for overall economic prosperity, useful for comparing the prosperity of populations widely separated in time and space - that the differences in genetic heritage could be ignored, because they had little to do with the variance in median height over time or space.

The point is this: an overlooked set of environmental variables was dominating something almost everyone assumed was an inherited trait of entire populations. And very likely part of the reason these obvious (once noticed) variables were overlooked was racial bias in the assessments - all these short people were some other race than the scientists observing them.

 

Log in or Sign up to hide all adverts.

I clearly said that I can summarize the arguments and findings of the article. The article contains tables with the names of the genes studied as well as their distribution between populations. I'm not arguing from authority I simply stated that Graves' article showed that IQ related genes showed no racial association. As far as his credibility as a source on this subject I asked him about his qualifications to review genetic research.

This is his reply. You can also review Graves' educational background and academic credentials here:

Here is my summary in my own words:

The article Race, Genomics and Intelligence: Slight Return by Joseph Graves is a review of the scientific literature on race and intelligence with special attention paid to the definition of biological race and the evidence against the idea that intelligence is associated with racial differences. Graves shows that the psychometric literature on race and intelligence relies on socially-defined racial categories that do not have biological reality. Graves uses Rushton as an example of a psychologist who did not adequately define race but claimed that there was an evolutionary basis to racial differences in IQ score. He provides a brief overview of the fallacies of Rushton's arguments including the misapplication of r/K selection theory and the fact that his racial matrix was based on arbitrary groups of people. Graves goes on to list the various ways in which the race concept has been defined: essentialist, taxonomic, population, and lineage. These models are all conceptually flawed and do not accurately describe human genetic variation.

Graves cites Ernst Mayr's definition of subspecies which he states is synonymous with biological races. A subspecies according to Mayr is population within a species that is differentiated from other populations to the point of being on the cusp of speciation. Graves cites empirical evidence on population subdivision using Sewall Wright's Fst threshold for classifying subpopulations (Fst> 0.25) to show that human populations do not meet Mayr's criteria for the classification of biological races. Graves further explains that the human genome contains genetic variants called single nucleotide polymorphisms (SNPs). These SNPs can be used to
study gene frequencies in geographic populations. Graves references a study that identified 11 genes associated with general intelligence. He calculated Fst from the SNPs reported in these genes.

Table 1 of Graves article reports the mean and standard deviation for Fst within each of these genes. Out of the 11 genes 9 have mean values well below Wright's threshold.

Please Register or Log in to view the hidden image!

According to Graves despite the problems with sampling these genetic variants across world populations the data does not support the position that there should be racially differentiated genetic variation for genetic variants associated with intelligence.

In Figure 1 of the article Graves shows a chart illustrating that some SNPs associated with variation in intelligence have high Fst. Once again the results do not show that there is a racial distribution between these genetic variants.

Please Register or Log in to view the hidden image!

Figure 2 shows the distribution of delta values for SNPs in 7 genes associated with cognitive performance between Sub-Saharan Africa and East Asia.

Please Register or Log in to view the hidden image!

Again the results do not show a racial association for the genes analyzed.

Graves concludes his article by noting that the claims in The Bell Curve were thoroughly discredited. Despite this he says that racialist thinking in psychometry is alive and well. As an example he mentions the comments of Noble prize winning geneticist James Watson who stated that he was "inherently gloomy about the prospect of Africa" because "all our social policies are based on the fact that their intelligence is the same as ours – whereas all the testing says not really." Graves states that there is overwhelming evidence that human genetic variation does not match our socially defined categories, but despite this some scholars seem to continue their desire to maintain racial thinking in sociology, biomedicine, and psychometrics. He says that new genetic research has further established that these claims are absurd. As an example he points out that claims of the genetic superiority of European and East Asians over Sub-Saharan Africans run afoul of the fact that these populations contain more deleterious genes. Graves further points out that genetic hypotheses for phenotypic variance require that genotypes are reared in equal environments for multiple generations to equalize maternal effects. He notes that this never happens in racially stratified societies.

Graves offers ideas on how to improve the conditions of demographic groups that have been historically discriminated against by providing equal education, nutrition, health care and freedom from exposure to toxic materials such as lead. He says that the more complex a trait is the more it is likely to be disrupted by poor environment. Finally Graves states that we will never be able to launch a credible research program on the genetic differentials of intelligence associated with genomic variation associated with ancestry, until this social justice is achieved. However this has never really been the aim of people most invested in this research.

 

Log in or Sign up to hide all adverts.

Thank you! You put that to bed succinctly, and even supported my previous post's suggestion concerning education, nutrition and environmental exposure to toxins.

 

Log in or Sign up to hide all adverts.

Yes, we've all heard the "there was discrimination in the past, it's that" argument. Your rendition of the experience of Negroes in US slavery in no way explains the global pattern. Your story about the twins is pointless. Do you have any data showing slave conditions cause a persistent IQ drop? I doubt it. You are just making up stories. Flynn himself has said that Blacks have made little gain in general intelligence. Your evidence for Black IQ in the USA is an e-mail from Nisbett? The Brooks-Gunn study was done on children where heritability estimates are unreliable. Furthermore "environmental variables" such as number of books in the house are just a proxy for parental IQ.

 

What groups is he using to calculate Fst? Why do you think an Fst below 0.25 means no between group variation, when it obviously doesn't? How do you explain genes over 0.25 Fst meaning no between group variation? Graves seems to have taken Templeton's mispreresentation of Sewall Wright's Fst statistic applied to subspecies, and applied it to single genes as if it means no difference, when any non-zero Fst means a difference. What nonsense!

BTW, where was the article published?

 

There is also racial oppression and bigotry in the present, and hangover effects of similar in the past - such as stereotype threat, and neurotoxin exposure, and complexity of childhood environment, and so forth.

You have to control for these variables, and many others, or you cannot compare test results between populations. You also have to control for the obvious possibility of a genetic influence on vulnerability to these variables.

It's not an "argument" so much as an observation: these things affect the results of tests of cognitive ability, and they are unevenly distributed between populations. So you have to control for them. If you don't - and nobody has - you cannot draw valid conclusions about a genetic basis for any observed variance in cognitive abilities between populations.

You brought up the example of height - take a lesson from it. You don't even have a measurement problem with height, as you do with "cognitive ability".

 

A claim has been made that IQ genes are evenly distributed among races and I am addressing that claim, among others. When these claims have been addressed I will address others, and perhaps make my own claims. There is quite a lot of counter argument here so forgive me if I don't address all of it at the same time, but I'll get to it. Be patient!

 

In history, how smartz were the rest of the world in comparison to Mesopotamia?

Please Register or Log in to view the hidden image!

 

Of course, many racists believe that there is no discrimination today. Not that Phill is racist; he just likes to support racist positions with common racist positions and tactics.

 

Well the, that's crazy right there. FIrst, who would ever take IQ to be reliable? Second, who would ever take it to be the only measure of intelligence? Third, who would think that there would ever be genes related to a specific human test? That's like thinking that there are SAT genes or Understanding-A-David-Lynch-Film genes.

There have to be a lot of genes involved in intelligence. And in order to work, these genes have to produce a number of complex proteins. These proteins can only be formed as part of processes that are highly dependent on the cellular environment, this cellular environment can very from cell to cell, this environment is also dependent on the overall environment of the organism.

I'm not saying that it's impossible to find super-important, swamp-any-environmental-effect intelligence genes, but it would be really hard to find these genes and really unlikely.

We already know that we can cause huge differences in intelligence through environment. We can create huge differences in intelligence just by using hindu-arabic numerals rather than roman numerals. Just by building the technology we have, both physical and conceptual, we have increased the mental abilities of humans far beyond what was possible two hundred years ago.

 

It wasn't meant to. I used the example of the history of African-Americans in the United States as evidence that environmental differences can have major effects on the nurturing of intelligence. Also I didn't just mention slavery but institutional racist discrimination (segregation, Jim Crow etc) which lasted for 100 years after slavery as a basis for claiming that the environment between Black and White Americans has not been equal throughout America's entire history.

There are historical reasons for environmental inequality across the globe as well.

No, it's not pointless. The point was to show you that even with identical genes two people can have major differences in intelligence due to environmental factors. This is a basic principle of experimental quantitative genetics. In order to infer genetic causality from phenotypic differences exhibited by different genotypes you have to raise those genotypes in the same environment. Otherwise you can not conclude that there are differences in genetic potential based on different phenotypes.

Different genotypes can have the same genetic potential. This also applies to groups. You can have genetic variance between groups of people such as geographic populations and still have the same genetic potential for certain traits.

No, I gave you an example of how environmental differences can stunt IQ in a population. American slavery and discrimination causing environmental differences between Black and White Americans is just a fact. We know that factors like social discrimination, malnutrition and environmental toxicity can affect IQ scores. The disproportionate poverty rates of African-Americans was caused by slavery and institutional racist discrimination. Slavery caused the African-American population to be poor from the beginning. Discrimination enforced by the government for 100 years ensured that many of them stayed poor.

Despite environmental improvement in the last few decades due to laws against racist discrimination we are still seeing the effects of past racist discrimination today. The Civil Rights Movement created new opportunities for Blacks but it did not eliminate inequality. More work needs to be done to improve the environment for all people.

Did he? According to the study I cited Flynn said that Blacks have made significant gains in g (general intelligence).

Please Register or Log in to view the hidden image!

Table 2 shows that the estimated g gains of Blacks on Whites were 91.13% of the comparable IQ gains on the WAIS (2.57/2.825 .9113) and 94.73% of the comparable IQ gains on the WISC (4.67/4.935.9473). The average of the two is 93%. Multiplying that value by the IQ gain indicates that between 1972 and 2002, Blacks made a g gain on Whites equivalent to 5.13 points.

Source: Black Americans Reduce the Racial IQ Gap Evidence From Standardization Samples Psychological Science Volume 17—Number 10 (2006)

Yes. Nisbett has done reliable research on this subject. I asked him this last year because the article where he said the IQ gap was about 10 points is 10 years old. I have read that Black test scores improved in recent years (Obama Effect) so I wanted to know by how much.

This is what he said:

Apparently there is no reliable IQ data on race any more. We can look at achievement data for ourselves to see if they verify Nisbett's claim.

That doesn't matter. We are not talking about heritability but rather the influence of environment on IQ. Studies on children give reliable indicators of how the learning environment affects people.

You could say that controlling for environmental variables simply controls for the intelligence of the parents and therefore this research only represents the upper limits of genetic potential in the Black population however the abstract clearly says that maternal education does not reduce the ethnic difference further.

This is some of the most reliable research we have that racial IQ gaps are caused by the environment.

 

Continued.....

The groups he used came from 50 populations varying from regions identified as Africa (sub-Saharan Africa), Europe, Asia (Middle East and Eurasia), East Asia, Oceania, and the Americas. Fst is a measure of population substructure and is most useful for examining the overall genetic divergence among subpopulations. What Graves did is analyze genes related to IQ and calculate their Fst values to see how much genetic divergence they exhibited between populations. He used Sewall Wright's threshold of 0.25 designed to determine fixation of different alleles in subpopulations. Most of the SNP's analyzed did not meet the threshold therefore their divergence does not meet criteria for identification of a trait (intelligence) that is differentiated between populations.

Templeton by the way did not misrepresent Wright's Fst statistic and neither did Graves. Graves explained to me by email the significance of Wright's threshold for subspecies classification.

The issue here is not that there is absolutely no difference in the frequency of genes but how much statistical difference is there that we can determine to be significant. If the difference is statistically very low then the evidence does not point to a racial association for a given trait. What population genetic theory would predict is that a polygenic trait like intelligence would be represented by genes that are scattered throughout the population. So the genes that determine high intelligence would be present in some individuals more than others and because intelligence is highly heritable these genes would run in families but they would not be concentrated in one population over another. That's what the data that Graves presented indicates.

The article was published as a chapter of a book on the role genes and environment play in influencing IQ.

Full Citation: Graves, J.L, Race, Genomics, and IQ: Slight Return for Intelligence Quotient: Testing, Role of Genetics and the Environment and Social Outcomes, Ed. Joseph Kush, Nova Scientific Publishers, pp. 69 –86, 2013.

 

Can you reference any other geneticist using Fst 0.25 to dismiss a gene as insignificant? I mean your email where Graves reviews himself is nice and all, but to me applying an Fst measure to dismiss genes just seems like some ad hoc contrived nonsense, based on nothing. So anyone else using Fst at all to dismiss genes would add some nice support to the method. I mean you do understand that Fst is just a distribution or differentiation measure, which has no connection to significance of effect sizes? You do get that, right?

Incidentally, have you read this: A review of intelligence GWAS hits: Their relationship to country IQ and the issue of spatial autocorrelation

 

How do you square this with us all being genetically the same height as you assert below? If we consistently see the same patterns between ancestry and outcome, is not the default hypothesis genetics? It would be for any other trait. Populations can be well defined, can you explain how they are not? What variables specifically are you concerned about? The magnetic field of Jupiter as it passes? Are you sure your vague objections about all possible variables are not applicable to any study, but you only demand these impossible standards for race related studies?

Really? Could you explain how it contradicts Visscher's work in the area? You can find references to it in this excellent paper: Still Missing, by Eric Turkheimer. "We are all genetically about the same height". It's probably just my biases that are causing me to laugh. I'm sure his reasoning is sound, after all, you agree with him.

Not at all. It's really very simple. Sample people all over the world, take their genotypes, identified only by a random number, and then plot them according to similarity (PCA is good for this). You'll see the Blumenbach partition: East Asians, Caucasoids, SS Africans, and Native Americans. Of course you'll get a lot of mixed people scattering between, but those major clusters will be very distinctive. So if you want to partition human variation to operationalise it, that's your first partition. And then of course you can break it down further, to Europeans, South Asians, etc. Nobody ever puts India and China in the same cluster, based on phenetics, genes or appearance. That's because a natural division simply isn't there. And then post facto, after making the clusters, you can see that all of the people in or near the SS African cluster are what we call Blacks or Negroes in the "social race" concept. Of course Americans have a tendency to call mulattoes Black, but I think you get the point. So the clustering is independent of any cultural information. It just happens to match quite nicely "social race", so the same labels could be used, although Caucasoid race would often be called West Eurasian population, but they are the same thing.

 

Graves doesn't use Fst to dismiss genes only to analyze their genetic divergence within subpopulations. I haven't read that article before although I have heard of David Piffer who has published articles in Mankind Quarterly, a journal notorious for publishing studies regarded as Scientific Racism. I see that this one was published in Intelligence. I will take a look at it and try to get Graves or a geneticist familiar with this type of research to look at it.

 

That's exactly what he does. Why not get a geneticist to look at Graves' paper?

 

I can do that as well.

 

Please do.

Source

Surely dismissing every gene not differentiated by the large amount of Fst 0.25 would dismiss every gene contributing to a highly polygenic trait. It's clearly something Graves made up to dismiss race IQ genes, and has been applied nowhere else. In short, the Graves paper is garbage.