Why Race Matters: Race
Differences and What They Mean
Author: Michael Levin
Westport, CT.: Praeger, 1997. Pp. X + 415.
The Western Journal of Black Studies
by L. Keita
Although the question of race has been an important sociological issue ever
since the development of anthropology as a study of different human groups,
contemporary philosophy has had relatively little to say about the topic.
Interestingly enough though, three of the luminaries of Western philosophy did
write about race as if those human groups that distinguished themselves both
geographically and phenotypically constituted natural kinds in terms of
temperament and intellect.
This amounted to the view held by Hume, Kant, and Hegel that the different
branches of humanity were distinguishable not only phenotypically but also
temperamentally and intellectually. Kant and Hume associated the dark
pigmentation of persons of African origin with cognitive deficiencies, and
Hegel wrote disparagingly of the natural temperament of Africans as
explanatory of their cultures. And even long before Kant's time the Greek
philosopher Aristotle argued about race along essentialist lines. Aristotle
inferred in a priori fashion that the dark pigmentation of the Africans of
ancient Egypt and Nubia as signaled cowardice. (Aristotle, "Physiognomics",
Discussions on the important issue of race, though ignored by contemporary
philosophers for the most part, have been an integral part of social science
discussion. The main debate centers on whether "race" is some sort
of natural kind or is a mere social construction. If race were just social
construction then essentialist arguments associating phenotypical
characteristics with behavior or cognitive dispositions would have no
ontological grounds to stand on. On the other hand, if races could be
established scientifically, that is, as natural kinds, then possible
inferences about dispositional and behavioral characteristics could be drawn
about the members of the different racial groups. This has been one of the
more popular approaches to the issue of race within the general context of
Western quantitative psychology and physical anthropology.
The interesting point about Michael Levin's Why Race Matters (the title
suggesting a play on the 1993 text Race Matters authored by social theorist
Cornel West) is that it is written by a philosopher and that it seeks to
extend the tenure of the old essentialist argument on race. Levin's thesis is
empirically observable phenotypical traits that differentiate the so-called
races macroscopically are casually connected with dispositional traits such as
intelligence and temperament. Of course, this thesis is hardly novel given its
long-standing tenure in the research paradigms of orthodox Western
quantitative psychology. Perhaps the best known of these efforts is that of A.
Shuey (1996) who
reported that the average measured IQ gap between "blacks" and
"whites" in American society is approximately 15 points, or in the
parlance of quantitative psychology, one standard deviation. It was claimed
that the average score of persons of European extraction is set at 100 points,
while those of African ancestry approximated 85 points. While some geneticists,
biologists, and psychometricians explain this average difference as due
primarily or exclusively to environmental causes, nativists such as Levin
offer a mainly biological explanation.
But merely to state this supposed fact would not satisfy epistemologically.
Further explanations are needed and this is what Levin suggests: "speculation
has long focused on the different pressures exerted by African and Eurasian
climates. Survival in the colder climates of Europe and Northern Asia requires
technologies unnecessary in Africa: clothing has to be fabricated, fires
sustained, food hunted and stored....
Planning is less adaptive in warmer climates where food is easier to get and
spoils when stored" (p. 136). Consider too: "Like the cheetah's
supple spine and the horse's hoof, the levels of intelligence of the different
races were responses to environmental pressures--as were the values embraced
by different groups" (p. 177).
Levin's reasoning here is causally essentialist and cannot be supported by the
facts. First of all the climatic argument is patently fallacious for the
following reasons. Neanderthal man, the European representative of Homo
erectus, though resident in Europe and other colder regions for at least
300,000 years (the archaeological claim is made that the Neanderthals became
extinct approximately 40,000 years ago), was not as cognitively evolved as
those members of Homo sapiens, who migrated from Africa to Europe and Central
Asia 30-40,000 years ago.
But these African members of Homo sapiens were themselves evolutionary
descendants of African varieties of Homo erectus. If the colder climates of
Europe and northern Asia were more evolutionarily challenging than the warmer
climates of Africa then one would have expected Homo sapiens to emerge in
those environments and not in the tropical ecologies of Africa. Furthermore,
if the northern ecologies and climates of the world were more challenging and
evolutionarily selective then one would have expected that the European and
north Asian descendants of resident Homo erectus (Neanderthal and others)
would have been the colonizing migrants into Africa and other tropical regions
of the globe. But the reverse has occurred.
Again, the argument that the colder climates of Europe and northern Asia
produced more cognitively evolved branches of humanity than those derivative
from the wanner climates cannot be sustained given that the world's first
civilizations (on the assumption of the Eurocentric definition of civilization)
emerged in tropical and sub-tropical regions. And we have no proof that the
originators of such civilizations migrated from the colder climates of the
Eurasian landmass into tropical and subtropical regions. The African
civilizations of ancient Nubia and ancient Egypt developed in tropical and
subtropical regions, so too the civilizations of Mesopotamia and Harappan (the
The only possible exception to this trend is the civilization of China. But
even here the site of origin of this civilization is not fully in the
Levin also errs in his explanatory thesis in that there are regions of Africa
where Africans have lived for several thousand years where the climatic
conditions are very clearly subtropical and even temperate at certain times of
the year. The climates of southern Africa extend beyond the tropic of
and approximate the climates of European and west Asian territories such as
Greece, Turkey, and Iran. Thus if cold climatic conditions constitute the
necessary conditions for environmental challenge - the ultimate
leitmotiv of human evolutionary change - then
one would have to argue that some members of the African branch of
humanity were also subjected to climatic conditions similar to those of Europe
and temperate-zone Asia. Levin expresses an ignorance of geography when he
fails to recognize that there are parts of Africa within the tropical zone
where the temperatures are quite temperate on account of altitude. One refers,
for example, to the highlands of Kenya, Ethiopia and the plateau regions of
Putting aside for the moment questions about the nature of intelligence and
whether IQ tests do measure intelligence, I want to demonstrate now that
Levin's claim that there is a strict causal connection between IQ and race -
as defined by Levin--is false. Levin puts it as follows: "When `race' is
operationalized geographically, generalizations about races acquire clear
empirical meaning.... To say the mean intelligence of whites exceeds that of
blacks is to say that the mean intelligence of people of European descent
exceeds that of people of African descent. Every such generalization may be
false, but they are uniformly meaningful" (pg. 21). For Levin "intelligence"
is determined by
quantitative scores on IQ tests regardless of the environmental or genetic
components of such scores. This is curious given that most of the discussion
on IQ testing is concerned with examining the basis for interracial or
interethnic differences in reported scores. Levin writes: "should the
black and white
intelligence polygene turn out to be identical, and there is a mean difference
in intelligence due entirely to environmental factors, all that would follow
is that blacks would be on average as intelligent as whites if both were
raised identically.... It would not follow that blacks are as
intelligent" (pg. 38). In fact this approach to the question of
intelligence as a phenotypical trait is tantamount to equating "intelligence"
with "education" or "training." Yet most of Levin's text
is concerned to argue in favor of "intelligence" not only as
phenotype but as phenotype linked causally to genotype. There is a world of
difference between these two positions.
But I return to my original counter claim against Levin: race and intelligence
are not causally connected. I refer to international comparisons of IQ scores
as reported by psychometric testing. According to arch-nativist Richard Lynn
(1978) the IQ scores of southern Europeans is a full standard deviation lower
than that of the average score of northern Europeans. According to Lynn the
reported average IQ of Spaniards in Spain is 87. We are also informed that in
Yugoslavia and Greece respectably some tested school children scored 89 in
both instances. Lynn also reports that Italian immigrants to the United States
scored 84 while their Swedish counterparts scored 102. To compound the
north-south IQ gap in Europe we are informed again that Portuguese immigrants
United States scored 83, a score lower than that registered by African
Americans in general. Of similar interest too is the fact that when children
of northern and southern European immigrants were tested with non-language
tests along with other children of European ancestry the IQ gap remained the
same. Children of northern European descent scored 97, those of southern
scored 85, while those of less recent European ancestry scored 98.
Thus it would seem that the IQ gap is not based on race after all, but on
something more akin to exposure to levels of modern technology and education.
According to convention all Europeans are classified as belonging to the
so-called Caucasoid race, yet there are significant differences between the
scores obtained from northern Europe and those from southern Europe. The main
difference observable between northern and southern Europe is not race but
rather level of industrialization and modernization in the technical sense.
Lynn's article not only deals with the IQ scores of Europeans but also with
those of Asians and Africans. For example, Lynn states that "In India,
there is a considerable literature on intelligence testing.... All the mean
IQs lay in the range from 81 to 94, the overall mean being about 86" (p.
269). Yet "a small sample of 25 postgraduate students at the University
of Calcutta, who took Raven's Test produced an incredibly low mean IQ of
75" (pg. 269). Curiously enough the inhabitants of India are considered
"caucasoid." For persons of African origin Lynn reports that the
scores range from 75 (Ghana, Jamaica) to 88 (Uganda, Tanzania.) One might want
to compare such scores with those of West Asians: Iraq (80) and Iran (low
80's) (Lynn, p. 269).
The scores form East Asia are to be somewhat qualified given that no scores
are reported from mainland China where the vast majority of East Asians live.
As Lynn put it: "Little is known about the intelligence levels of
Mongoloids in their homelands. The majority of studies have been made on
Chinese and Japanese immigrants to the United States" (pg. 272). Lynn
reports a score of 99 for
Chinese ethnics from Hawaii and scores of 107 and 114 for Chinese and Japanese
subjects in Vancouver respectively. But there is the current belief among some
psychometricians (Murray, Rushton, et al.) that East Asians are naturally more
intelligent that Africans and Caucasoids. But I have pointed out that the vast
majority of East Asians have not been subjected to IQ tests. Furthermore, the
Eskimos who are considered members of the same racial complex as east
Asians score between the ranges of 70 and 85. This is hardly proof of
Mongoloid intellectual superiority.
There are two other well-discussed aspects of the IQ controversy that Levin
discusses with little epistemological care. These two topics are the
"Flynn effect" and "identical twin testing." The Flynn
effect (1987) is based on the research conducted by James Flynn which states
that there have been great intergenerational increases in IQ reported over the
recent years. For example, Flynn noted that between 1949 and 1974 the IQ
scores of French persons increased 21 points. Similar kinds of increases were
noted for Japan, Germany, and Austria. In fact, Flynn remarked on the same
phenomenon for fourteen nations. Clearly, genetic factors could not be at work
here. We can attribute the increases in IQ purely to environmental changes (schooling
and other kinds of cognitive exposures).
Levin does recognize the Flynn effect (pg. 128) but dismisses an environmental
explanation for the "black-white IQ gap" with the unproven claim
that since "black" and "white" environments have been
converging the racial IQ gap may well be genetic. But Levin has not proven
that the sociological environments of blacks and whites have been converging.
Levin even argues that if the IQ gap does not change over time for both blacks
and whites this difference means "that the race gap is due to genes"
(p. 129). It is difficult to follow Levin's reasoning here. What is at stake
here in the discussion is the impact of the environment in determining IQ
scores. The explanation offered for intergenerational differences in IQ scores
is necessarily environmental, that is, over time environments change for the
same racial group. But then Levin illogically rejects this possible
explanation for the black-white IQ gap by claiming that there are no
environmental differences between the sociologies of blacks and
But we know that there are evident environmental and sociological differences
between blacks and whites in the United States and elsewhere. In fact the only
way in which possible cognitive differences between blacks and whites could be
properly evaluated is for sufficiently large samples of black and white
monozygotic (identical) twins to be randomly adopted by black and white
households across all socioeconomic levels. Thus each pair of black and white
twins will be adopted at birth individually by one randomly chosen black
family and one randomly chosen white family. But again, even if such an
experiment were possible, what guarantees do we have that the identical twins
have all been randomly distributed throughout society?
This brings me to the very issue of research done on identical twins reared
apart as a way of determining the effect of the environment on IQ scores.
Levin does refer to the different studies done on monozygotic twins reared
apart (p. 97) but does not shed much analytical light on the issue. The most
extensive of these tests (as cited by Levin) are those carried out by Pedersen
(1992), Burt (1966) and Bouchard (1990). The average estimate of [h.sup.2] for
these studies is .80 with the average difference between IQ scores being put
at 7 points.
All this is interesting but it does not shed much light on the questions
concerning race and IQ. After all, if monozygotic twins reared apart are
reared in environments that are sociologically similar then the significance
of reported IQ differences loses explanatory importance. What is significant
about monozygotic twin studies are the reported individual ranges between
tested twins. I argue that if there are a significant numbers of monozygotic
twin IQ scores that demonstrate tested gaps of approximately 15 points -- the
approximate black-white gap in the United States and some parts of Africa,
then the hereditarian thesis is cast in doubt.
Psychometrician Arthur Jensen (1972) offers some data on this in the paper
"IQs of Identical Twins Reared Apart". Jensen offers details on 4
twin studies. One of the studies discussed is that of Burt's (1966), but this
research has been mired in controversy so I shall not use it in my analysis.
sixty-nine (69) pairs of twins tested by Shields (38), Newman et al. (19) and
Juel-Nielsen (12), ten (10) pairs of the sixty-nine (69) had gap differences
of at least 15 points. This number amounts to approximately 14 percent of the
total. This is significant. If one restricts the analysis to Shields and
Newman the percentage increases to 18 percent approximately. We note that the
Juel-Nielson study was carried out in Denmark where social equality is an
ideal aimed at. This explains why in this study the largest IQ gap registered
is 12 points.
Levin's thesis, though mainly about the issue of intelligence and race is not
exclusively so. He extends his racial thesis to the question of values and
human behavior. Levin puts it thus: "The races simply differ in abilities,
behavior, and standards of evaluation" (p. 163). Levin's sociobiological
approach to human decision making is once again evident when he writes:
evolution of blacks and whites, which appears to have produced cognitive and
temperamental differences, makes it possible, indeed likely, that behaviors
and norms pathological for whites are not pathological for blacks, and that
identical behaviors and norms have different functional significance for the
two races" (p. 186).
In the section on values Levin seeks to expound on this thesis with references
to interracial differences in criminality and cultural attitudes. The goal
here is to argue once again for an essentialist theory of race not only in
terms of cognitive abilities but also in terms of values. But again Levin,
though avowedly empiricist in orientation, fails to be exhaustive in his
analysis. Obvious proof of my claim here is that there are societies of
persons of African origin where crime and acts of violence are extremely rare.
The crime rate is extremely low in rural communities of persons of African
ancestry in Africa and the Americas. It is the sociological argument that
explains why the descendants of peaceful rural African Americans are often
involved with the law in the urban, high unemployment areas of North America.
In the discussion on values Levin once again appeals to the environment to
explain "the strong individual dominance drive" among Africans (p.
140) and a supposedly cooperative and democratic European temperament. Levin
writes, "Recall the hypothesis that conditions in northern Eurasia
strongly favored cooperation" (p. 168). In fact Levin's analysis is
totally erroneous: it is African society that has been criticized, as being
incompatible with the individualism required of market economies. It is in
Africa and other areas of the African world rather than in Europe that the
idea of the extended family, with its concomitant principles of cooperation
and altruistic obligation, is held to be widespread. But I do not make an
essentialist argument here. Human social relations are determined maximally by
the principle of sociological contingency: individualism and anomie are
rampant in large urban areas while cooperation and altruism are more common in
the rural areas of whatever continent. So again, Levin's thesis fails for lack
of persuasiveness in terms of empirical evidence and analysis. In this regard
Levin's speculations on the supposed differences in free will capacity between
"blacks" and "whites" cannot be supported either
scientifically or theoretically.
Of interest too are Levin's attempts to justify his arguments about the
cognitive abilities of individuals of African ancestry by claiming that there
is no evidence of creative, intellectual production from such persons in
history. Levin writes that "the absence from Africa of advanced material
culture is more
than an accident is confirmed by the failure of post-colonial Africa to
sustain the technology left by whites" (p. 120). Levin is wrong in his
claims about "the absence of material culture in Africa" and the
fact that there are problems with technology transfers in Africa cannot be
attributed to genetic causes. There are places in Europe where there are
serious problems with technology transfers and maintenance as in countries
such as Greece, Albania, and Bulgaria. And there are countries in the African
world where basic technologies such as electricity, telecommunications, roads
and so on are efficiently maintained. Ready examples are Kenya, Senegal,
Barbados and Botswana.
The problems of technology facing the African world have nothing to do with
genetics but with the complex of relations between the industrialized nations
and their ex-colonies within the context of the capitalist world order.
Levin's ignorance of the history of civilizations is again evident when he
writes "no important discovery, invention or world leader emerged from
Africa. The art, music, architecture, literature, and political history of
Eurasia owe virtually nothing to Africans" (p. 194). In response to those
who would claim that the architectural, cultural, and technological influences
of ancient Egypt and Nubia, as African civilizations, on the Eurasian world
easily refute the above assertion, Levin states without any elaboration and
with reference only to some obscure pseudo-anthropologist (Baker, 1974) that
"the Egyptians were not black" (p. 194).
But the originators of the world's first qualitatively path-breaking and
influential technological civilizations were the ancient Nubians and Egyptians,
both of African racial origin. Levin also fails to recognize that the Moorish
architecture of southern Spain is of African origin, and that the originating
and creative impulses for contemporary Euro-American music and art are also of
But I want to elaborate further on Levin's lack of knowledge of Africa's
anthropological and historical past with respect to the ancient Egyptians and
Nubians. Levin's illogical reasoning on this issue goes something like this:
Persons of African origin are incapable of producing any form of genuine
civilization. The ancient Egyptians produced forms of genuine civilization.
Ergo, the ancient Egyptians were not persons of African origin. The problem
with this unsound argument is that Levin's first premise is empirically false.
The Greek historian Herodotus specifically refers to the ancient Egyptians as
"black-skinned and woolly haired" both sufficient phenotypical
characteristics for membership in the African race. Other Greek writers such
as Aristotle also make reference to the physical characteristics of the
ancient Egyptians and Nubians, contrary to Levin's assertions. Levin writes:
"Africanists cite scattered reference to blacks in Herodotus to support a
Nubian origin of Greek religion, but ignore Aristotle's silence about Africa.
Why should Aristotle have lied, but not Herodotus?" (p. 195). Despite the
strangeness of this proposition, Aristotle did not lie for we find in his
"Physiognomics" the following:
"Too black a hue marks the coward as witness Egyptians and Ethiopians and
so does also too white a complexion as you may see from women" (Vol. VI,
812a). In Book XIV of Problems Aristotle makes reference to the hair form of
Egyptians and Ethiopians: "Why are the Ethiopians and Egyptians
bandy-legged? Is it because the bodies of living creatures become distorted by
heat, like logs of wood
when they become dry? The condition of their hair supports this theory; for it
is curlier than that of other nations, and curliness is as it were crookedness
of the hair." (Book XIV, p. 317)
What Levin fails to recognize is that just as there are no reasons why
eyesight and hearing capacities might differ on average between individuals of
the different geographical racial groups, so too with the human cognitive
faculties. Once the biological threshold of Homo sapiens was reached in Africa
there was no further need for evolutionary pressures to yield groups of
individuals with significantly differing cognitive capacities wherever they
migrated on earth. The cognitive capacities of Homo sapiens Africanus were
adequate enough to ensure survival not only in tropical Africa but also in
temperate Eurasia and the frigid Arctic. I repeat the proof of this thesis
that I established above: Homo sapiens Africanus was in no way cognitively
disadvantaged with respect to the Neanderthal human types that had been
resident in the cold climates of Eurasia for at least 300,000 years -- a time
span much longer than the entire period for which Homo sapiens has existed.
In sum, Levin's text should be understood as not much more than a repetition
of the traditional arguments on the issue of "race differences" in
intellect, temperament and physiological capacities. I do not deny that the
world's environments have selected for gross human physiological traits such
pigmentation, hair form, epicanthic eye folds and so on, but once Homo sapiens
Africanus emerged with the capacities for language (all languages necessarily
derive from the first African languages), and conceptual thought (necessary
for the first art works and Neolithic technology) the selecting influences of
the world's environments were rendered redundant with respect to human
What this means is that Levin's argument concerning racial differences with
respect to intelligence, temperament and value choice cannot be supported. His
quasi-inductivist thesis that the present average IQ scores of some groups of
persons of African descent is explained by a supposed dearth of African
creativity over time, I have shown to be fallacious. Levin's text, in essence,
is just another instance of that persistent ideological strain in Western
thought which claims that persons of African origin are deficient in the
important intellectual characteristics that define the human species.
The reason for the persistence of this specific ideology is that it is
required to maintain the idea of the Eurocentric racial caste system
established at the dawn of the modern era to justify the economic division of
labor required by the captivity of Africans in the Americas, the captivity and
of Native Americans in the Americas, and the colonization of Africans in
Africa. The forced labor of Africans in the Americas and Africa had to be
justified by an ideology that claimed that persons of African origin were less
cognitively capable than Europeans. If it were admitted that IQ scores reflect
only sociological differences between groups then the social stability of the
racial caste system invented by modern Eurocentric discourse would be
seriously undermined from an intellectual standpoint. Something similar was at
work in the European invention of the concept of the naturalness of
hierarchies of ancestry required for the stability of Europe's feudal orders:
aristocrat and serf were
distinguishable ancestrally purely on the spurious concept of "blood."
The politico-economic situation of the post-Civil Rights era in the United
States and that of post-colonial Africa are highly unstable in terms of the
division of labor established at the origins of Modern Europe. The economic
"success" of modern Europe springs from the exploitation of African
labor (in the Americas and Africa) and the resources of Africa by the
entrepreneurial administration of European capitalism. The racial ideology
that made the wealth and success of Europe possible is now being challenged.
It is the defense of that racial ideology that explains the popularity and
notoriety of theorists such as Eyesenck, Jensen, Shockley, Murray, Lynn,
Murray, and Levin.
I have argued above that the racial ideology emphasizing the cognitive
limitation of persons of African descent is fallacious. I established the
1) The idea that the colder climates of Eurasia were more challenging than
those of Africa thereby leading to a more cognitively evolved branch of Homo
sapiens is false. Neanderthal man existed in Europe for at least 300,000 years
but was no more cognitively evolved than the incoming Homo sapiens Africanus.
2) IQ test results do not establish cognitive differences based on race since
Southern Europeans, West Asians, South Asians, Eskimos, and others register
scores similar to those of persons of African descent. It is evident that IQ
scores reflect particular sociological environments than otherwise.
As a final note: the apparent seriousness of the text is marred by a frivolous
"hypothetical address by the President of the United States of America to
a Joint session of Congress and the American People" supporting Levin's
theses on racial differences in intelligence and temperament, and three
nonsensical appendices. There is also an inexcusable grammatical error in the
first line of page 147.
Aristotle. "Physiognomics" in Minor Works, trans. W.S. Hett. London:
Heinemann, 1963. It should be noted that the editor of Aristotle's Minor Works
and Problems expresses doubt whether these works were authored by Aristotle
himself. In any case they do reflect in this instance empirical observative
the Greeks made of the African phenotype of the Egyptians and Nubians.
Problems, trans. W.S. Hett. London: Heinemann, 1970. Flynn, J. 1987.
"Massive IQ Gains in 14 Nations: What IQ Tests Really Measure."
Psychological Bulletin 101: 171-191.
Jensen, A. 1970. "IQs of Identical Twins Reared Apart." Behavior
Genetics 1. 133-148.
Lynn, R. 1978. "Ethnic and Racial Differences in Intelligence:
International Comparisons." In Human Variation: The Biopsychology of Age,
Race, and Sex, eds. R. Travis Osborne, Clyde Noble, and Nathaniel Weyl. Eds.:
Shuey, A. 1966. The Testing of Negro Intelligence, 2nd ed. New York: Social