Educability and group differences (Arthur Jensen, 1973)


Why read a book from 1973? 41 years old? Well, it was a good read indeed! It is interesting how much of the evidence for racial differences were in place already in 1973, and it has more or less only become stronger since then. Basically, the book is a shorter and less technical (but not that much!) version of his major book The g Factor.



This distinction between the individual and the particular gene pool from which the unique combination forming his genotype was derived extends beyond his family to the racial group with which he is identified and to the social status into which he is born. You are not your race; you are not your group. You are you. That is, if you are talking genetics. If you are talking sociology or politics, that may be another matter. You may be psychologically tied to and influenced by whatever groups you happen to identify with. If you are either elated or depressed about yourself because of such identification, don’t attribute this to genetics. It in fact contradicts this kind of typology which compels so many persons to identify with various groups as if the statistical attributes of the group determined their own characteristics. Racism and social elitism fundamentally arise from identification of individuals with their genetic ancestry; they ignore individuality in favor of group characteristics; they emphasize pride in group characteristics, not individual accomplishment; they are more concerned with who belongs to what, and with head-counting and percentages and
quotas than with respecting the characteristics of individuals in their own right. This kind of thinking is contradicted by genetics; it is anti-Mendelian. And even if you profess to abhor racism and social elitism and are joined in battle against them, you can only remain in a miserable quandary if at the same time you continue to think, explicitly or implicitly, in terms of non-genetic or anti-
genetic theories of human differences. Wrong theories exact their own penalties from those who believe them. Unfortunately, among many of my critics and among many students I repeatedly en­
counter lines of argument which reveal disturbing thought-blocks to distinguishing individuals from statistical characteristics (usually the mean) of the groups with which they are historically or socially identified. I know professors, for example, who cannot bring themselves to discuss racial group differences when any persons from different racial groups are present, and the fact that I am
able to do so perhaps makes me appear insensitive in their eyes. I was once bothered by this too. I got over it as I studied more genetics and came more and more to appreciate its real implications.


Well written! I had the same idea, namely that what unites racists and ‘antiracists’ is their collectivism, their focus on the properties of groups instead of individuals.



The important distinction between the individual and the

populationmust always be kept clearly in mind in any discussion

of racial differences in mental abilities or any behavioral charac­

teristics. Whenever we select a person for some special educa­

tional purpose, whether for special instruction in a grade-school

class for children with learning problems, or for a ‘gifted’ class

with an advanced curriculum, or for college attendance, or for

admission to graduate training or a professional school, we are

selecting an individual, and we are selecting him and dealing

with him as an individual for reasons of his individuality.

Similarly, when we employ someone, or promote someone in

his occupation, or give some special award or honor to someone

for his accomplishments, we are doing this to an individual.

The variables of social class, race, and national origin are

correlated so imperfectly with any of the valid criteria on which

the above decisions should depend, or, for that matter, with any

behavioral characteristic, that these background factors are irre­

levant as a basis for dealing with individuals – as students, as

employees, as neighbors. Furthermore, since, as far as we know,

the full range of human talents is represented in all the major

races of man and in all socioeconomic levels, it is unjust to allow

the mere fact of an individual’s racial or social background to

affect the treatment accorded to him. All persons rightfully must

be regarded on the basis of their individual qualities and merits,

and all social, educational, and economic institutions must have

built into them the mechanisms for insuring and maximizing

the treatment of persons according to their individual behavior.


As written by a true racist, or something…



A common misconception often arises in connection with standards

such as the following from an article by Dreeben (1969): ‘First,

genetic forces and environmental forces operate on two distinct

dimensions of time. Genetic effects are established when an ovum

is fertilized – at one moment in time; environmental effects extend

over time.’ This is often erroneously believed to mean that although

individuals may be endowed with different genotypes at the moment

of conception, all change and differentiation that take place

thereafter are the result of environmental forces. But this interpreta­

tion overlooks the fact that the genes exert a continuing influence

on developmental processes. Many genetic effects are manifested

phenotypically only in later stages of development. As an obvious

example, patterns of baldness are genetically determined but do

not show up until middle age. Behavioral characteristics associated

with maturational processes, like mental development, variously

manifest genetic effects increasingly as the individual grows from

infant to adult. This is clearly seen in the gradually increasing

degree of correlation between the mental abilities of parents and

their biological children from infancy to late adolescence, which

occurs even when the children have never had contact with their

biological parents after infancy and have been reared by adoptive

or foster parents (e.g., Honzik, 1957). Under a normal range of

environmental conditions, an individual’s phenotypic IQ, from

infancy to maturity, converges toward its genotypic value.


So, it was known that heritability increases already in 1957 (or 1973). I thought it entered common knowledge in 1994 with McGue M, Bouchard TJ, Jr, Iacono WG, Lykken DT. 1993. Behavior genetics of cognitive ability: A life-span perspective.



Probably the best evidence for the threshold hypothesis would

be the finding of significantly higher heritability in groups that

are above average in SES and environmental advantages than in

groups of low SES.13 No one has ever done this systematically.

The gifted children in Terman’s study came mostly from the

higher SES levels and unquestionably had considerably better

than average environmental advantages for intellectual develop­

ment. The mean IQ of their siblings was 123 and the correlation

between the IQs of the gifted and their siblings, estimated from

the sibling regression, is 0-44, which, when corrected for attenu­

ation, is close to the genetically predicted sibling correlations of

0*5 (with random mating) or 0-6 (with an assortative mating

coefficient of 0-5), and does not differ much from sibling correla­

tions reported in the general literature. The gifted group as adults

were, on the average, of higher SES than their own parents. Thus

the offspring of the gifted probably enjoyed even greater environ­

mental advantages. The narrow heritability of IQ in this group,

estimated from the midparent-midchild regression, is 0-85. This

is significantly higher than the best estimate of narrow’ heritability

(0*71) given by Jinks and Fulker (1970, p. 342) on the basis of

Burt’s data, which includes a wride range of SES in the English

population. It is also higher than the midparent-midchild correla­

tion (0-69 + 0-03) found in a largely rural population sample in

Vermont in 1920, with environmental advantages presumably

much below those provided by the Terman gifted and their

spouses (Jones, 1928, p. 69). These heritability findings, then, are

consistent with the threshold hypothesis. But the total evidence

for the hypothesis must still be regarded as quite ambiguous. A

clear finding of an appreciable difference between h2 in the Negro

and white populations, however, would be consistent with the

hypothesis depicted in Figures 7*5 and 7*6. It could mean, in

effect, that the scale of environmental effects differs for the bulk

of the two populations and not simply that the two populations are

distributed about different means on the same additive (i.e., equal

interval) scale of environments. So now we must examine what

meager evidence exists on the estimation of h2 in Negro populations.


With all the talk about h2 x SES interactions, Jensen was there 40 years ago too. :p



A statistical test could be applied to determine if the lesser

variance of the Negro IQ distribution is an artifact of the scale or

a ‘fact of nature’. One would determine, for both Negro and white

population samples, separately and together, whether there is any

significant correlation (both linear and non-linear relationships

should be sought) between family means (based on fraternal twins

or siblings7) and within-family variances. Since the total variance

( V T) of a subpopulation is comprised of the between-families

variance ( VB) plus within-families variance ( V w), we should

determine if two subpopulations which differ in VT differ in VB

or Vw or in both. If they differ only in VB, this suggests a ‘fact of

nature’ rather than an artifact of scale, and this interpretation is

strengthened if it is found that there is no significant correlation

between family means and within-family variances. A correlation

between within-family variances and family means suggests a

scale artifact which might be eliminated by a transformation of

the scale. These tests, however, would not be worthwhile unless

performed on quite large and representative samples of the sub­

populations in question. If it is found that the most adequate scale

from all these standpoints shows marked differences in IQ variance

for Negroes and whites, and if the heritabilities of IQ were either

closely comparable in both populations, or smaller in the Negro

population, the genetic uniformity hypothesis would be very

untenable. It would indicate less genetic variance in the Negro

population. (The results could, of course, go in the opposite

direction, but the evidence based on the existing scales of mental

ability indicates less variance in the Negro samples.) Smaller

variance, with the consequence of a lesser proportion of the

subpopulation having higher values on the intellectual ability

scale, even if the mean were the same as in the general population,

would have important social consequences for the subpopulation

with the lower variance in terms of the proportion of its members

who are able to compete successfully in those endeavors in which

proficiency is most highly correlated with intellectual ability.

J. B. S. Haldane (1965, pp. xcii-xciii) noted that ‘For cultural

achievements high variability may be more important than a high

average. . . . When we say the ancient Greeks were great mathe­

maticians we are in fact thinking of about 20 men. We know

nothing about the average Greeks in this respect.’


Why should two populations have different genetic variances?

Differences in gene frequencies and in the degree of assortative

mating are the chief causes.8 A difference in gene frequencies for

a given characteristic will cause different means and variances,

although if the number of gene loci is large, the difference in

variances will be relatively less than the difference in means. If

the genetic means in both populations are equal, the most likely

explanation of unequal genetic variances is differences in degree

of assortative mating. That is, the tendency for like to mate with

like with respect to a particular trait. It is known that there is a

high degree of assortative mating for intelligence in the white

population. (There are no published studies of assortative mating

for intelligence in non-white populations.) Assortative mating

increases the total genetic variance in the population; it also

increases the between-families variance relative to within-families

variance. Some 15 to 20 percent of the total variance in the white

population is attributable to assortative mating for intelligence.

Assortative mating per se has no effect on the mean, so if both the

genetic means andvariances differ between two populations, we

can suspect differences in gene frequencies as well as differences

in assortative mating.


This is interesting. Especially if the assortative mating might be increased due to internet dating. This could have big social implications. With IQ 100, SD 15, 2.3% are >130. But if we increased variance 10% cuz of stronger assortative mating, it wud be 3.45% above. The effects are huge when we get farther out. Ofc, on the other hand, it will also give me imbeciles. Im willing to take the trade. :p


As for the unequal variances between races with africans having less. This seems implausible in the light of the fact that africans have higher variance in all genes. So, it wud be odd if they had lower variance in the g-genes.



Since variation in skin pigmentation, because of its social-

environmental consequences, is controlled in this research design,

any direct biochemical connection between degree of skin pig­

mentation and intelligence must be either ruled out or, if such a

relationship is established, its consequences for the present design

must be assessed. The possibility of a biochemical connection

between skin pigmentation and intelligence is not totally unlikely

in view of the biochemical relation between melanins, which are

responsible for pigmentation, and some of the neural transmitter

substances in the brain. The skin and the cerebral cortex both arise

from the ectoderm in the development of the embryo and share

some of the same biochemical processes.


And it now makes sense that Jensen later wrote his comment in: Jensen, Arthur R. “Comments on correlations of IQ with skin color and geographic–demographic variables.” Intelligence 34.2 (2006): 128-131.


Templer and Arikawa emphasize that they regard skin color only as a climatic variable, a multigenerational reflection of climatic history. And this may well be theoretically adequate for their present purpose. But we should not let it mislead us to dismiss completely other possible, and presently causal, connections between skin color and IQ—an idea the authors, perhaps too cautiously, called “absurd.” This stance overlooks the probability of the genetic phenomenon of pleiotropy acting as at least a partial cause of the IQ × skin color correlation in present day populations. (Pleiotropy is the condition of a single gene having two or more phenotypically quite different effects. For example a single gene could affect both IQ and skin color.)



But the whole notion of equating for SES, in the first place,

involves what has been called the ‘sociologist’s fallacy’. This fallacy

is seen in full bloom in one sociologist’s criticism of studies of

Negro-white IQ differences which equated the groups for SES

or other environmental factors: ‘Actually in most of the studies he

[Jensen, 1969a] reports on, the most important environmental

variable, the IQ of the parent, has not been equated at all’

(Stinchcombe, 1969, p. 516). Apart from the strictly environmental

effect of parental IQ ,1 it is obvious that, since IQ variance contains

a large genetic component, equating groups for parental IQ means

equating them for genetic factors more than for environmental

factors. The same is true, though to a lesser degree, when we

equate for SES. When typical Negro children are equated with

white children on some index of SES, one is comparing a majority

of the Negro population with some lower fraction of the white

population.2 The white comparison group, therefore, is not

genetically representative of the entire white population but is

genotypically (as well as environmentally) lower by some sub­

stantial degree. Thus, if one supposes one is equating only for

environmental influences, equating on SES equates too much.The

method would be a proper control of environmental factors if all

children had been placed in their SES categories completely at

random, in the nature of a true experiment. But as it is, SES

classification is more a result than a cause of IQ variance.


quite possibly the first formulation of the sociologists fallacy.



Data on two white populations show that fetal loss (Fx genera­

tion) in matings of the parental generation ( P J increases

cumulatively by approximately 2-5 percent to 3 percent with

each additional country of birth in the great-grandparental

generation (Px). A dependent relation shows that increased fetal

loss is also related to greater distances between birthplaces of

mates within the Px generation. Conversely, low fetal loss is

encountered with a small number of countries in the background

and shorter distance between birthplaces. It is suggested that a

large number of countries of birth represents a larger number

of Mendelian gene pools and that with increased mixture of

these gene pools, fetal loss increases proportionately. An animal

model is cited in support of this contention, (p. 24)


Never heard of this effect! It somewhat offsets hybrid vigor.