Race, genes, & intelligence, part 1

B. The Hereditarian Hypothesis and plausibility

The issue being discussed is called the hereditarian hypothesis (HH from now on).(8) The hypothesis is that some of the well established statistically average ethnoracial differences in cognition (9, 24, 33, 35, 50, 52, 75, 76, 99) have a partial genetic etiology and that these differences result in some of the pervasive average performance (and behavioral) differences that we find in our society (24, 52, 53, 75, 76). Refer g and group differences.

Specifically, according to the hypothesis as often formulated, some of the average performance differences between (and within) groups result from biological differences related to general intelligence which develop, to some extent, as a result of genotypic difference. Here, general intelligence (the g factor) refers to the latent trait that IQ and others cognitive tests measure and is, while closely related, conceptually distinct from the term “intelligence” as generally used; it is conceived as being a psychometric construct with neurophysiological functional and structural correlates.

Templer, Tomeo, Arikawa, Willians, 2002. Asian-Black differences in aptitude and difficulty of chosen academic discipline (75)

Roth, Huffcutt, Bobko, 2003. Ethnic Group Differences in Measures of Job Performance:
A New Meta-Analysis (76)

For the Hereditarian Hypothesis to be robust 5 prerequisites must be fulfilled:

1) The said populations must represent different average genetic populations and there must be plausible mechanisms by which the differences in intelligence could have developed (I-V). 2) Tested differences in psychometric intelligence, specifically in g, must represent measurable, functionally important differences within and between populations (VI-VII). 3) Intelligence must be heritable within populations and map neurophysiologically (VIII-X). Finally, 4) the differences must be relatively consistent across time and nation, given the parameters of population genetics (XI).

It’s a rather audacious hypothesis! With regards to the Black-White difference in the US, 1-3 (I-IX) are no longer in serious question. 4 (X) is generally supported. I will outline these below:

The reality of race

I. The HH presupposes that there are descendant populations, which there are. Humans split from their most recent primate cousin ~300,000 years ago; Eurasians split from west SubSaharans ~100,000 years ago and left Africa 40-80,000 year ago; West Eurasians split from East Eurasians 30-40,000 years ago (3, 21, 22, 64, 71).

Campbell and Tishkoff, 2009. The Evolution of Human Genetic Review and Phenotypic Variation in Africa

II. It presupposes that 40-100,000 years was sufficient time for evolving intrahuman differences and that socially significant differentiation did occur; 40-100,000 years was enough time and differentiation did occur (1, 2, 13 15, 20, 21, 22, 42, 62, 63, 66), some of which is socially significant. This is why race (or regional ancestry) is medically relevant (22, 28, 69). Human evolution has, in fact, rapidly accelerated over the last 40,000 years (48). And selection is still ongoing (61).

Coop, Pickrell, and Novembre, 2009. The Role of Geography in Human Adaptation

Voight, et al., 2006. A map of recent positive selection in the human genome

III. It presupposed that descendant populations could theoretically differ genetically in some cognitive dispositions on average (refer to Lewontin’s second fallacy), which there is every reason to believe they could (7, 13, 16, 14, 30, 45, 59, 60, 65, 66). (See Appendix 1: Major theories supporting the evolution of population specific difference in cognition). And which there is direct evidence that they, in fact, do in some manner (13, 20, 44). For example, there has been recent intense differential selection for genes that code for the NRG–ERBB4 pathway, a pathways which is involved in for neural development and which is associated with various psychiatric phenotypes (44). Additionally, as in the case of DRD4, there is direct evidence that population specific alleles affect social behavior and academic performance (64-68). As for alleles that are associated with intelligence, two SNPs of DTNBP1 (rs1018381 and rs2619522) have consistently been found to influence general cognitive ability (109); the frequencies of these alleles vary by regional ancestry.  Other genes that affect cognition and have alleles that vary by regional ancestry include PKU (110) and APOE (111). 

Genes under recent selection for nervous system, brain function and development; language skills and vocal learning
CDK5RAP2, CENPJ, GABRA4, PSEN1, SYT1, SLC6A4, SNTG1, GRM3, GRM1, GLRA2, OR4C13, OR2B6, RAPSN, ASPM, RNT1, SV2B, SKP1A, DAB1, APPBP2, APBA2, PCDH15, PHACTR1, ALG10, PREP, GPM6A, DGKI, ASPM, MCPH1, FOXP2 (see: 59-64)
Laland, Odling-Smee, Myles, 2010. How culture shaped the human genome: bringing genetics and the human sciences together (13)

Wang, et al., 2006. Global landscape of recent inferred Darwinian selection for Homo sapiens

There is also convergent evidence supporting the hypothesis that populations differ in cognitive genetics. Beal et al. (1984) found that cranial capacity varied across populations and that cranial volume was positively correlated with temperature (92). Bailey and Geary (2010) found that human brain evolution was driven by population density, variation in paleoclimate, and temperature variation (72); Odokuma et. al. (2010) found that average Nigerian cranial capacity was less than average European cranial capacity (73); Rushton and Ankney (2009) and Jensen (1998) report numerous studies which consistently found that cranial capacity correlates with general intelligence (74). McDaniel (2005) conducted a meta-analysis on all MRI studies of brain size and intelligence and found a .30 correlation (103). Connecting this convergent evidence back to the findings above, ASPM and MCPH1, two genes under recent positive selection, have been found to be associated with sex-specific increases in brain size (106). For an elaboration of this paradigm refer to Ash and Gordon’s Brain Size, Intelligence, and Paleoclimatic Variation (77) and Jensen’s Population Differences In Intelligence: Causal Hypotheses (84).

IV. It presupposes a social understanding of ‘race’ which refers to groups with different average population ancestry, which we have (11):

According to the Journal of Philosophy account, the logical core of the ordinary concept of race is the concept of a group of human beings:
H1. Who are distinguished from other human beings by visible physical features of the relevant kind
H2. Whose members are linked by common ancestry
and
H3. Who originate from a distinctive geographical location.
Hardimon, 2009. Wallis Simpson was Wrong

See also A semantic defense of the ordinary race concept for an elaboration of this.

V. It presupposes that ‘race,’ in fact, refers, in part, to descendant populations, which it does (17,28, 43).

Zakharia, et. al. 2009.Characterizing the admixed African ancestry of African Americans

Nelis, et al., 2009. Genetic Structure of Europeans: A View from the North–East

The meaningfulness of intelligence

VI. The HH presupposes that there are experimental means of measuring individual cognitive differences, which there are (4, 10, 11, 23), and that these measures are fairly well defined, which they are.
McGrew, 2009. CHC Theory and the human cognitive abilities project: Standing on the shoulders of the giants of psychometric intelligence research (15)

VII. It presupposed that differences in intelligence within and between populations are functionally important, which they are (7, 10, 19, 24, 33, 35, 46, 50, 52, 75, 76, 99) — though the scope of the importance is still in debate.

The heritability and biological basis of intelligence within populations

VIII. The HH also presupposes that intelligence has neurophysiological correlates which, in fact, it does (49, 102, 103).

Gläscher, et al. 2009. Lesion mapping of cognitive abilities linked to intelligence

IX. It presupposes that some of the various forms of cognitive processing (4, 5, 15) are heritable, and that differences between individuals within a population could be attributed, in part, to genes; which they can (12,20, 36).

Within population heritability of General Intelligence across age.

Bouchard, 2009. Genetic influence on human intelligence (Spearman’s g): How much?

From: Deary et al. 2010. The neuroscience of human intelligence differences

“Basic genetic influences on intelligence Investigation of the presence of genetic influences on general intelligence dates back to the nineteenth century, when Francis Galton published two papers concluding that mental abilities were transmitted by heredity from one generation to another15. Despite an intermittently hostile political reception, many studies since then — based principally on twin and adoption samples — have replicated this observation, and none has contradicted it16. estimates of how much of the total variance in general intelligence can be attributed to genetic influences range from 30 to 80%. General intelligence factors, in the form of latent traits from which measurement error has been removed, fall at the high end of this range. Broad domains of cognitive ability — such as verbal and perceptual– organizational abilities — generally show similar amounts of genetic influence, although the genetic influence on memory tends to be somewhat smaller.”

From: Hayworth, et al., 2009. Generalist Genes and High Cognitive Abilities

“We found strong support for the Generalist Genes hypothesis for high cognitive abilities in that genetic ‘group’ correlations were substantial between g, reading, math and language. The average genetic correlation of 0.58 is comparable to genetic correlations found in multivariate genetic studies in unselected samples, which are about 0.60 between g and learning abilities and about 0.70 between cognitive abilities. A direct comparison can be made with our previous analyses of the entire distribution using the same sample, measures and methods as in the present analysis of high cognitive ability (Haworth et al. 2009b). In that report, we found that the average genetic correlation was 0.68 for the entire distribution.”

From Bouchard, 2009. Genetic influence on human intelligence (Spearman’s g): How much?

“The history and conceptual background of the heritability statistic is briefly discussed. The construct of heritability is embedded in the method of structural equation modeling widely used in modern population genetics and in human behavior genetics. The application of structural equation modeling to behavioral phenotypes is shown to be a useful and informative analytic tool, as it implements the research strategy of ‘strong inference’. I describe the application of ‘strong inference’, via the use of structural equation models in the domain of human intelligence, and demonstrate its utility as a means of refuting well formulated scientific hypotheses. The construct of Spearman’s g is shown to be a strongly confirmed scientific hypothesis. Genetic and environmental influences are shown to influence g differentially over time, with shared environmental influences predominating early in life, but dissipating to near zero by adulthood. The hypothesis of substantively significant genetic influence on adult g is documented by multiple lines of evidence and numerous replications.”

X. While most contemporary hereditarians argue that the said performance differences result from genetically encoded differences for general intelligence, it’s worth noting that the psychometric and biological existence of g, and the g-loadedness of group difference, is not a necessary precondition for the hereditary hypothesis to be true. As the environmentalist James Flynn noted:

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. I attribute no intent or motive to Gould, it is just that you cannot rebut arguments if you do not acknowledge and address them. (Flynn 1999a, 373)

The existence of g just makes the implications of the hereditarian hypothesis, if true, unavoidable. As Jensen (2000) pointed out, g “lies at the heart of the whole problematic nexus involving the nature of group differences, the merits of meritocratic selection in a diverse society, the legitimacy of using tests, their adverse impact on certain groups, and its redress by group preferences in college admissions and employment.” It also significantly strengthens the hypothesis. The existence of g loaded differences makes implausible a number of environmental arguments (including virtually all purely sociological ones, such as test bias, motivation, stereotype effect, etc.) and it allows for a number of hereditarian arguments. This is now a moot point, of course, because it is now known that g is not a psychometric artifact, that g has a robustly biological dimension, and that the said differences (at least in the US) are g-loaded (4, 5, 15, 36, 84).

Deary and Johnson, 2010. The neuroscience of human intelligence differences

As for g, it’s psychometrically structurally similar across populations, sexes, ages, and cultures it’s structurally similar (101). The basic structure of g can be found in other primates (102). Structurally and functionally, it correlates with neural organization, the volume of white and grey matter, total brain size, rate of cerebral glucose metabolism, and nerve conduction velocity, resting EEG and average evoked potentials, cerebral glucose metabolic rate, speed and efficiency of brain functioning inferred from reaction time (103).

The relative consistency of ethnoracial differences

XI. It presupposes that ethnoracial populations differences are relatively consistent, from the perspective of population genetics, across time and nation. There is support for this (54, 55 — but see also 56, 57).

Internationally, the basic pattern below can be seen (54, 55, 56, 79, 80, 81, 82). The intelligence of nations is intercorrelated with technological advancement, academic achievement, GDP, patent rates, numbers of scientists, rates of HIV, etc. National intelligence also predicts the productivity of individual immigrants coming from a particular nation (112). (Global IQ table as summarized by Steve Sailer).

Rindermann, 2007. Relevance of education and intelligence at the national level
for the economic welfare of people

The highest values for the smart fractions are found in East Asia …followed by Western and Eastern European and North American countries, by South European countries, Arab or Muslim and Latin American countries and finally by sub-Saharan countries.
Rindermann, Sailer, Thompson, 2009. The impact of smart fractions, cognitive ability of politicians and average competences of peoples on social development

When it comes to causality and national differences, the causal pathway runs from cognitive capacity out:

Rindermann and Thompson, 2011. Cognitive Capitalism : The Effect of Cognitive Ability on Wealth, as Mediated Through Scientific Achievement and Economic Freedom

Within highly mixed nations, with no explicit racial classes, the expected pattern predicted by Hereditarians can also be seen between individuals with different proportions of ancestral mixture. Villarreal (2010) found them in Mexico:

…Despite this ambiguity, I found evidence of profound social stratification by skin color in contemporary Mexico. Individuals with darker skin tone have significantly lower levels of educational attainment and occupational status, and they are more likely to live in poverty and less likely to be affluent, even after controlling for other individual characteristics. Differences in socioeconomic status between Mexicans of different skin tones are indeed large. Although measurement differences preclude precise cross-country comparisons, the differences between Mexicans in the three color categories used in this study, and particularly between individuals classified as white and non-white, are comparable to the differences between African Americans and non-Hispanic whites in the United States……
Villarreal, 2010. Stratification by Skin Color in Contemporary Mexico

The differences are ubiquitous. The Black (West Sub-Saharan African)-White (European) difference, for example, can be seen in Britain (87, 88), Canada, Brazil (89), the Caribbean, the Netherlands, and throughout Africa (56, 57). Of course, it matters what specific ancestral African and European (or East Asian or Mixed Amerindian) populations one is comparing. (See: Lynn, 2006. Race differences in intelligence)

Overall, the global pattern of differences is as predicted by the Hereditarian Hypothesis. As Wendy Johnson noted, when commenting on Richard Lynn’s The Global Bell Curve, “Lynn’s data are essentially correct and do reflect the general state of the world.”