Richard Lynn

Christopher Brand (left) and Richard Lynn

Race Differences in Intelligence

An Evolutionary Analysis

Richard Lynn

WashingtonSummit Publishers Augusta, GA

A National Policy Institute Book 2006

© 2006 by Richard Lynn

For Joyce

Sei il mio amor e tutta la mia vita

(You are my love and my whole life)

Mimi - La Boheme, Act V

C O N T E N T S

1. The Meaning and Measurement of Intelligence

2. The Meaning and Formation of Races

3. Europeans

4. Africans

5. Bushmen and Pygmies

6. South Asians and North Africans

7. Southeast Asians

8. Australian Aborigines

9. Pacific Islanders

10. East Asians

11. Arctic Peoples

12. Native Americans

13. Reliability and Validity of Race Differences in Intelligence

14. Environmental and Genetic Determinants of Race Differences in Intelligence

15. The Evolution of Intelligence

16. Climate, Race, Brain Size, and Intelligence

17. The Evolution of Race Differences in Intelligence

Appendix: Intelligence Tests

References

E-Book

Chapter 1. The Meaning and Measurement of Intelligence

1.   Definition of Intelligence

2.   The Hierarchical Model of Intelligence

3.   The IQ

4.   Flynn Effects

Race differences in intelligence began to be analyzed scientifically in the middle years of the nineteenth century. In the 1830s, Samuel Morton (1849) in the United States assembled a collection of skulls, measured their volume, and calculated that Europeans had the largest brains followed by Chinese, Malays, and Native American Indians, while Africans and finally Australian Aborigines had the smallest brains. He concluded that these differences in brain size accounted for the race differences in intelligence. A similar view was advanced a few years later in France by Paul Broca (1861, p. 304): "in general, the brain is larger in eminent men than in men of mediocre talent, in superior than in inferior races." About the same time Francis Galton (1969) in England arrived at the same conclusion by a different route. He assessed the intelligence of the races by the numbers of geniuses they produced in relation to the size of their populations. He concluded that the Greeks of classical Athens were the most intelligent people, followed in descending order by the lowland Scots, the English, the Africans, and the Australian Aborigines.

In the twentieth century this question continued to be debated. The intelligence test was constructed by Alfred Binet in France in 1905. It was translated into English by Lewis Terman (1916) at Stanford University and later in the century a number of other intelligence tests were constructed. This made it possible to measure and compare the intelligence of the various races and by the end of the twentieth century many hundreds of studies had been published on this issue. Most of these have been concerned with the difference between blacks and whites in the United States, but studies have also been made of the intelligence of peoples in virtually every part of the world. For the difference between blacks and whites in the United States, the most authoritative studies are by Shuey (1966), who summarized all the studies from World War I up to 1965, Osborne and McGurk (1982), who updated this summary to 1980, Loehlin, Lindzey, and Spuhler's Race Differences in Intelligence (1975), Herrnstein and Murray's The Bell Curve (1994), and a series of publications by Jensen culminating in The g Factor (1998). There has been some interest in the intelligence of the Chinese and Japanese, which was reviewed by Vernon in The Abilities and Achievements of Orientals in North America (1982). A number of studies of the intelligence of Africans, Caucasians, and East Asians have been summarized by Rushton in Race, Evolution and Behavior (2000). All of these studies have been concerned with two problems. These are the evidence on race differences in intelligence, and the degree to which these differences are determined by genetic and environmental factors. It is widely accepted that race differences in intelligence exist, but no consensus has emerged on whether these have any genetic basis. All those named above have argued that there is some genetic basis for race differences. However, a number of authorities have concluded that there is no compelling evidence for genetic factors. This position has been adopted by Flynn in his Race, IQ and Jensen (1980), Brody in Intelligence (1992), and Mackintosh in IQ and Human Intelligence (1998).

The present book differs from previous studies in four respects. It is the first fully comprehensive review that has ever been made of the evidence on race differences in intelligence worldwide. Second, it reviews these for ten races rather than the three major races (Africans, Caucasians, and East Asians) analyzed by Rushton (2000). The races analyzed here are the Europeans, sub-Saharan Africans, Bushmen, South Asians and North Africans, Southeast Asians, Australian Aborigines, Pacific Islanders, East Asians, Arctic Peoples, and Native American Indians. Studies of these are presented in Chapters 3 through 12; Chapter 13 summarizes these studies and gives evidence on the reliability and validity of the IQs of the races. Third, Chapter 14 discusses the extent to which race differences in intelligence are determined by environmental and genetic factors. Fourth, Chapters 15, 16, and 17 discuss how race differences in intelligence have evolved over the course of approximately the last 100,000 years. These discussions are preceded by accounts of the nature of intelligence and the measurement of race differences given in this chapter, and of the concept of race in Chapter 2.

1. Definition of Intelligence

There is a widespread consensus that intelligence is a unitary construct that determines the efficiency of problem solving, learning, and remembering. A useful definition of intelligence was provided by a committee set up by the American Psychological Association in 1995 under the chairmanship of Ulrich Neisser and consisting of eleven American psychologists whose mandate was to produce a report on what is generally known and accepted about intelligence. The definition of intelligence proposed by the Task Force was that intelligence is the ability "to understand complex ideas, to adapt effectively to the environment, to learn from experience, to engage in various forms of reasoning, to overcome obstacles by taking thought" (Neisser, 1996, p. 1). This definition is generally acceptable, except for the component of effective adaptation to the environment. All living species are adapted effectively to their environment or they would not have survived, but many living species such as snakes and other reptiles cannot be regarded as intelligent. In economically developed nations, the underclass with its culture of long-term unemployment, crime, drug dependency, and welfare-dependent single mothers, is well adapted to its environment in so far as it is able to live on welfare and reproduce, but it has a low average IQ, as shown in detail by Herrnstein and Murray (1994), and is not intelligent in any reasonable sense of the word or as measured by intelligence tests.

A definition which avoids this misconception was proposed by Gottfredson and endorsed by 52 leading experts and published in the Wall Street Journal in 1994:

Intelligence is a very general mental capacity which, among other things, involves the ability to reason, plan, solve problems, think abstractly, comprehend complex ideas, learn quickly and learn from experience. It is not merely book learning, a narrow academic skill, or test taking smarts. Rather, it reflects a broader and deeper capability for comprehending our surroundings - "catching on," "making sense" of things, or "figuring out" what to do (Gottfredson, 1997, p. 13).

Intelligence conceptualized as a single entity can be measured by intelligence tests and quantified by the IQ (intelligence quotient). The theory of intelligence as largely a single entity was first formulated in the first decade of the twentieth century by Charles Spearman (1904), who showed that all cognitive abilities are positively intercorrelated, such that people who do well on some tasks tend to do well on all the others. Spearman devised the statistical method of factor analysis to show that the performance of all cognitive tasks is partly determined by a common factor. He designated this common factor g for "general intelligence."To explain the existence of the common factor, Spearman proposed that there must be some general mental power determining performance on all cognitive tasks and responsible for their positive intercorrelation.

2. The Hierarchical Model of Intelligence

Spearman also proposed that in addition to g, there are a number of specific abilities that determine performance on particular tasks, over and above the effect of g. In the 1930s an alternative theory was advanced by Thurstone (1938) that there are seven "primary abilities," which he designated reasoning, verbal comprehension, numerical ability, spatial ability, word fluency (the ability to produce a number of words as exemplars of a concept in a short period of time), memory, and perceptual speed. In the second half of the twentieth century, a general consensus emerged that both the Spearman and the Thurstone models were partially correct and that intelligence is best conceptualized as a hierarchical structure that can be envisioned as a pyramid in which there are some seventy narrow abilities at the base (Spearman's specific abilities), eight to ten second-order or group factors at the next level (Thurstone's primary abilities), and a single general factor (Spearman's g) at the apex. The leading contemporary formulations of this model have been set out by Horn (1991), Carroll (1993), and McGrew and Flanagan (1998). Their models are closely similar and propose that the eight to ten second-order factors consist of "fluid ability" (reasoning), "crystallized ability" (verbal comprehension), long-term memory, short-term memory, visualization (visual and spatial ability), numerical ability (arithmetic), mathematical ability, cultural knowledge, processing speed, and reaction time. This hierarchical model of intelligence is widely accepted among contemporary authorities such as the American Task Force on Intelligence (Neisser, 1996), Jensen (1998), Mackintosh (1998), Deary (2000), and many others. An extensive exposition of g and its structure, heritability, biology, and correlates has been presented by Jensen (1998) in his book The g Factor. He conceptualizes g as a construct or factor that he defines as "a hypothetical variable that 'underlies' an observed or measured variable" (p. 88). It is not possible to measure g directly, but the non-verbal reasoning IQs and scores obtained from intelligence tests and expressed as IQs (intelligence quotients) are approximate measures of g.

3. The IQ

The metric employed for the measurement of the intelligence of the races has been to adopt an IQ of 100 (with a standard deviation of 15) for Europeans in Britain, the United States, Australia, and New Zealand as the standard in terms of which the IQs of other races can be calculated. The mean IQs of Europeans in these four countries are virtually identical, as shown in Chapter 3 (Table 3.1), so tests constructed and standardized on Europeans in these countries provide equivalent instruments for racial comparisons. In Britain, Australia, and New Zealand, the intelligence tests have been standardized on Europeans, and this was also the case in the United States in the first half of the twentieth century. In the second half of the twentieth century American tests were normally standardized on the total population that included significant numbers of blacks and Hispanics. In these standardization samples the mean IQ of the total population is set at 100; the mean IQ of Europeans is approximately 102, while that of blacks is 87 and of Hispanics about 92 (see, e.g., Jensen and Reynolds, 1982). This means that when the IQs of other races are assessed with an American test standardized with an IQ of 100 for the total American population, 2 IQ points have to be deducted to obtain an IQ in relation to 100 for American Europeans. This problem does not arise with the only British test used in cross-cultural studies of intelligence. This is the Progressive Matrices, which has been standardized on British Europeans. The tests used in the studies of racial intelligence are identified by acronyms in the tables in which the results are presented. The full names of the tests and description of the abilities they measure are given in the Appendix.

In the summaries of studies of race differences in intelligence, IQs are given for general intelligence and, where possible, for the major primary abilities of reasoning, verbal comprehension, and visualization. IQs for general intelligence are obtained either from general intelligence tests that contain a mix of reasoning, verbal, visualization, perceptual, memory, and sometimes other items, or from tests of non-verbal reasoning ability such as the Progressive Matrices, which provide closely similar results to those of tests of general intelligence (Carroll, 1993; Jensen, 1998). A few studies are also available and summarized for race differences in immediate memory and musical abilities.

4. Flynn Effects

A problem with the quantification of race differences in intelligence is that IQs have been increasing since the 1920s in many parts of the world. These secular increases were first shown by Smith (1942) in Hawaii and have been confirmed in several subsequent studies such as that of Cattell (1951) in Britain. They have become known as the Flynn effect following their documentation by James Flynn (1984, 1987). When results are reported for the IQs of populations an adjustment needs to be made for Flynn effects, as otherwise populations obtain spuriously high means when they are scored on norms obtained from Europeans a number of years previously. The magnitude of the Flynn effect varies with different tests. Mean IQs on the Wechsler tests increased in several countries by approximately 3 IQ points per decade from the mid-1930s to the 1990s, but the Verbal IQ increased by approximately 2 IQ points per decade and the Performance IQ by approximately 4 IQ points per decade (Flynn, 1984, 1998; Lynn and Pagliari, 1994). For the Standard Progressive Matrices, the British mean IQ increased at a rate of approximately 2 IQ points per decade from 1938, when the test was constructed, up to 1979, when the last British standardization on children was carried out (Lynn and Hampson, 1986; Flynn, 1987). IQs on the Goodenough Draw-a-Man Test in the United States increased by 3 IQ points a decade between 1955 and 1968, calculated from the Harris (1963) and the United States Department of Health, Education and Welfare (1970) standardizations. The same rate of increase on this test has been found for blacks in South Africa from 1950 to 1988 (Richter, Griesel, and Wortley, 1989). Adjustments for Flynn effects have been made in all the figures for IQs presented for the populations in subsequent chapters. Where tests have been used for which the magnitude of the secular increase is not known, an increase of 3 IQ points per decade has been assumed.

There is no general consensus regarding the causes of the Flynn effect. A number of different theories by leading experts are presented in Neisser (1998). Some, including Flynn (1987) himself, believe that there has not been any significant increase in what may be called "real intelligence" and that the increases must be due to improvements in test taking skills. Others such as Greenfield (1998), Mackintosh (1998), and Williams (1998) have argued that the increases are genuine and that a number of factors are likely to be responsible, including a generally more cognitively stimulating environment, especially from television, computer games, improvements in education, and the increased education of parents. The presence of the Flynn effect in the development of infants measured by tests such as the age at which an infant is able to stand up makes these factors unlikely. It is probable that there has been some genuine increase in intelligence as a result of improvements in nutrition that have produced increases in height, in brain size, and probably in the neurological development of the brain during the twentieth century (Lynn, 1990a, 1998b).

Chapter 2. The Meaning and Formation of Races

1. The Formation of Races, Varieties, and Breeds

2. Varieties in Non-human Species

3. Taxonomies of Races

4. Race Differences in Diseases

5. Do Races Exist?

A b concerned with race differences in intelligence needs to define both intelligence and race. In the last chapter intelligence was defined and in this chapter a definition is offered of race. A simple and straightforward definition of race is that it consists of a group that is recognizably different from other groups. A fuller definition is that a race is a breeding population that is to some degree genetically different from neighboring populations as a result of geographical isolation, cultural factors, and endogamy, and which shows observable patterns of genotypic frequency differences for a number of intercorrelated, genetically determined characteristics, compared with other breeding populations. Geographical contact zones between races generally contain racial hybrids, who show intermediate values of gene frequencies from the more central distributions of the breeding groups. These hybrid and mixed race populations are known as clines.

1. The Formation of Races, Varieties, and Breeds

It is a general principle of evolutionary biology that when populations of species become isolated from one another they evolve into two or more sub-species. These are generally termed varieties, strains, or breeds. In the case of humans these different varieties are called races. These different varieties evolve as a result of the four processes of founder effects, genetic drift, mutation, and adaptation. The founder effect is that when a population splits and one group migrates to a new location to form a new population, the group that migrates will not be genetically identical to the one left behind. Hence the two populations differ genetically. The genetic drift effect is that gene frequencies change over time to some extent as a matter of chance and this leads to differences between populations. Drift continues with time and leads to increasing differences between races. The mutation effect is that new alleles (alleles are alternative forms of genes) appear through chance in some populations and if they are advantageous for survival and reproduction will gradually spread through the population. An advantageous new allele may appear as a mutation in one race, but not in others. The adaptation effect is that when a population migrates to a new location some alleles will be advantageous that were not advantageous in the old location. Individuals possessing advantageous alleles in the new location have more surviving offspring, so their alleles will be selected for and will gradually spread though the population. New varieties of several species have evolved as adaptations when populations have migrated into arctic environments. Some of these, such as foxes, bears, and hares, have evolved white fur to give them camouflage so they are not so easily seen by predators or prey. In all these cases mutations for white fur have appeared and spread through the population because they have given the animals possessing them a selective advantage. Eventually the new advantageous alleles entirely replace the less advantageous alleles and are then said to have become "fixed."

In many cases it is uncertain why different strains have evolved different characteristics. For instance, the fur of the European squirrel is red while that of the North American squirrel is grey. Possibly one of these colors confers a selective advantage and appeared by chance in one of these populations through a genetic mutation.

2. Varieties in Non-human Species

It has long been recognized that most species have several varieties or what in humans are called races. Early in his career Charles Darwin noted the different varieties of turtles on the Galapagos Islands and it was this that set him thinking how these had evolved. Later in his book The Variation of Animals and Plants under Domestication (1868) he described the different varieties of a number of species such as pigeons, each of which have their own distinctive manner of flight, movement, and cooing.

There are a number of different varieties or races among the apes. There are four races of chimpanzee. These are the true chimpanzee (Pan satyrus verus) indigenous to West Africa between Guinea and Nigeria, the bald chimpanzee (Pan satyrus satyrus) of Cameroon and Gabon, the pygmy chimpanzee (Pan satyrus paniscus) of north central Zaire, and the Schweinfurth chimpanzee (Pan satyrus schweinfurthi) of northeast Zaire. These races differ in physical appearance, distribution of blood groups, and the cries they utter. Different races have evolved among animal species in accordance with the same principles as among humans. For instance, there are two races of gorilla. These are the mountain gorilla (Gorilla beringei) native to the mountains around lakes Edward and Kivu in eastern Zaire, Rwanda, and western Uganda, and the coast gorilla (Gorilla gorilla) of the forests of Cameroon and Gabon. The two races are geographically isolated from one another by about a thousand miles and have evolved differences in physical appearance and blood group. The mountain gorilla has a narrower skull, shorter arms, longer legs, thicker hair, and blood group A, while the coast gorilla has a broader skull, longer arms, shorter legs, thinner hair, and blood group B (Baker, 1974). Some of the differences between the two races have evolved as adaptations to their different environments. The mountain gorilla inhabits a colder and open environment while the coast gorilla inhabits a warmer and densely forested environment. The mountain gorilla has developed thicker hair than the coast gorilla as a protection against the cold. The coast gorilla has developed longer arms to swing from tree to tree. There is no obvious explanation for why the mountain gorilla has a narrower skull, longer legs, and blood group A. These differences may have arisen through founder effects, genetic drift, or chance mutations, or they may confer some unknown advantage.

There are also a number of varieties among domestic animals. These are normally called breeds and have been bred by humans to serve a variety of useful purposes. Frequently they have been bred for greater size or, in the case of cattle, milk yields. In some cases they have been bred to adapt better to certain environments. For instance, varieties of hardy sheep have been bred that flourish on mountains and differ from lowland sheep. Humans have bred as many as seventy-nine different breeds of dogs for a variety of abilities, such as retrievers for retrieving game, sheep dogs for rounding up sheep, rottweilers for guarding premises, cocker spaniels for house pets, and so on. These breeds differ in their general intelligence, their specific abilities, and the ease with which they can be socialized and made obedient (Coren, 1994).

3. Taxonomies of Races

Biologists and anthropologists began to analyze and classify races in the middle years of the eighteenth century. The first taxonomy of races was advanced by the Swedish biologist Carl Linnaeus in 1758. In his System Naturae he proposed that there are four races which he designated Europaeus (Europeans), Afer (black Africans), Asiaticus (Asians), and Americanus (Native Americans). In 1776 the German physician Johann Friedrich Blumenbach added a fifth race and proposed a classification based principally on skin color. He designated these five races the Caucasian (white), Mongolian (yellow), Ethiopian (black), American (red), and Malayan (brown). These taxonomies were based on the clustering of morphological features and coloration in different races such as the Europeans' white skin, straight hair, and narrow nose, the sub-Saharan Africans' black skin, frizzy hair, and wide nose, the Mongolians' (East Asians) black hair, yellowish skin, and flattened nose, the Native Americans' reddish skin and beaky nose, and the Malaysians' brown skin. Morton (1849) used Blumenbach's five-race classification when he made the first analysis of brain size in relation to race.

In the early twentieth century data were collected on differences in the frequencies of blood groups in various populations throughout the world. Hirszfeld and Hirszfeld (1919) showed that the frequencies of a number of blood groups are consistent with race differences in coloration and morphology. For instance, blood group A is present in 41 to 48 percent in Europeans but in only about 28 percent of sub-Saharan Africans, while blood group B is present in between 10 and 20 percent of Europeans and about 34 percent of sub-Saharan Africans. Native Americans have virtually no A or B blood groups and almost all of them have the O blood group.

The accumulation of data on the distribution of the Rhesus (Rh) blood groups was used by Boyd (1950) to advance a five-race taxonomy consisting of (1) Europeans with high frequencies of blood groups Rh cde and cde; (2) Africans with very high frequencies of Rh cde; (3) East Asians with high frequency of B and virtually no cde; (4) American Indians with very high frequency of O, absence of B, and few cde; and (5) Australids with high A, negligible B, and cde. This analysis showed that blood-group distributions were consistent with the morphological and coloration racial taxonomies of classical anthropology.

A more detailed taxonomy of races was advanced by Coon, Garn, and Birdsell (1950), who proposed seven major races, each of which was subdivided into two or more subraces. These were (1) Caucasoids, subdivided into Nordics of Northwest Europe, Slavs of Northeast Europe, Alpines of Central Europe, Mediterraneans of South Europe, North Africa, and the Near East, and Hindi of India and Pakistan; (2) East Asians, subdivided into Tibetans, North Chinese, Classic East Asians (Koreans, Japanese, Mongolians), and Eskimos; (3) Southeast Asians, subdivided into South Chinese, Thais, Burmese, Malays, and Indonesians; (4) American Indians, subdivided into north, central, south, and Fuegians; (5) Africans, subdivided into East Africans, Sudanese, West Africans, Bantu, Bushmen, and Pygmies; (6) Pacific Islanders, subdivided into Melanesians, Micronesians, Polynesians, and Negritos; and (7) Australian Aborigines, subdivided into the Murrayian peoples of southeastern Australia and the Carpentarian people of northern and central Australia. A closely similar seven-race taxonomy was proposed by Baker (1974) comprising the five major races of Blumenbach and the Khoi Bushmen, consisting of the Hottentots and Bushmen of southwest Africa and the Kalahari desert, and the Australids, consisting of the Australian Aborigines and Melanesians.

In the 1980s and 1990s Nei and Roychoudhury (1993) and Cavalli-Sforza, Menozzi, and Piazza (1994) developed a new method of classifying humans into races on the basis of a number of genetic polymorphisms (polymorphism means that a gene has more than one allele or alternative form). The technique is to take a number of polymorphic genes for blood groups, blood proteins, lymphocyte antigens, and immunoglobins, and tabulate the different allele frequencies in populations throughout the world. These tabulations are then factor analyzed to find the degree to which the allele frequencies are associated to form clusters of populations that are genetically similar to one another. The Nei and Roychoudhury data for 26 populations have been factor analyzed by Jensen (1998) to show the existence of six major groups of humans that correspond closely to the races proposed by classical anthropologists. Using the traditional terminology, these are (1) Africans of Sub-Saharan Africa (Pygmies, Nigerians, Bantu, Bushmen); (2) Caucasoids (Lapps, Finns, Germans, English, Italians, Iranians, North Indians); (3) East Asians (Japanese, Chinese, Koreans, Tibetans, Mongolians); (4) Southeast Asians (Southern Chinese, Thais, Filipinos, Indonesians, Polynesians, Micronesians); (5) Amerindians (North and South Native American Indians and Inuit); and (6) Australian Aborigines (Australian Aborigines and New Guineans).

The same technique has been used by Cavalli-Sforza, Menozzi, and Piazza (1994) to analyze a larger data set of 120 alleles for 42 populations. These data were used to calculate the genetic differences between each population and every other population. From these they calculated a genetic linkage tree that groups the populations into what they called "clusters." They have found ten major clusters. These are (1) Bushmen and Pygmies; (2) sub-Saharan Africans; (3) South Asians and North Africans; (4) Europeans; (5) East Asians; (6) Arctic Peoples; (7) Native American Indians; (8) Southeast Asians; (9) Pacific Islanders; and (10) the Australian Aborigines and the Aboriginal New Guineans. It is apparent that this classification corresponds closely to the racial taxonomies of classical anthropology based on visible characteristics of color of skin, hair, eyes, body shape, limb length, and the like but for some reason Cavalli-Sforza, Menozzi, and Piazza (1994) prefer the term "clusters."

4. Race Differences in Diseases

There are race differences in a number of diseases that have a genetic basis including cystic fibrosis, PKU (phenylketonuria), hypertension, stroke, diabetes, prostate cancer, breast cancer, obesity, myopia, and schizophrenia. These differences have arisen through the processes of founder effects, genetic drift, mutation, and adaptation. There is such an extensive body of research on these that it would take a book to summarize it. The differences are illustrated here by the gene frequencies of cystic fibrosis and PKU in Europeans, sub-Saharan Africans, and East Asians (Orientals) given by Bodmer and Cavalli-Sforza (1976). These are shown in Table 2.1. The figures represent the gene frequencies (percentage prevalence rates) in the population. It will be seen that the gene frequencies of cystic fibrosis in Europeans are four or five times higher than in sub-Saharan Africans and East Asians, while gene frequencies of PKU are slightly more than twice as high in Europeans than in the other two races. The lower half of the table shows that the gene frequencies of the two diseases are quite similar in different European populations as widely dispersed as Austria, Australia, Canada, England, and the United States.

Table 2.1. Gene frequencies (percentages) of cystic fibrosis and PKU in Europeans, sub-Saharan Africans, and East Asians

5. Do Races Exist?

From the eighteenth century until the middle years of the twentieth century all anthropologists, biologists, and social scientists accepted that the human species contains a number of biologically distinct races. Thus, in the 1920s the British anthropologist Sir Arthur Keith wrote:

So clearly differentiated are the types of mankind that, were an anthropologist presented with a crowd of men drawn from the Australoid, the Negroid, East Asian or Caucasoid types, he could separate the one human element from the other without hesitation or mistake (Keith, 1922, p. xviii).

Curiously, this seemingly indisputable observation began to be disputed from the middle decades of the twentieth century, when a number of anthropologists began to assert that races do not exist. One of the first to adopt this position was the anthropologist Ashley Montagu (1945a.) in his book Man's Most Dangerous Myth: The Fallacy of Race. The title suggests that the concept of race is a myth and therefore that there is no such thing as race. However, in the book Montagu made it clear that he believed that races do exist. He wrote:

In biological usage a race is conceived to be a subdivision of a species which inherits the physical characteristics serving to distinguish it from other populations of the species. In the genetic sense a race may be defined as a population which differs in the incidence of certain genes from other populations, with one or more of which it is capable of exchanging genes across whatever boundaries (usually geographic) may separate them. If we are asked whether in this sense there exist a fair number of races in the human species, the answer is that there do (p. 6).

It is clear from this that race is neither a "myth" nor a "fallacy." Considering that Montagu evidently accepted that races exist it seems strange that he should have given his book such a misleading title.

Later in the second half of the twentieth century a number of anthropologists and geneticists came to assert that there is no such thing as race. In 1962 the anthropologist F. B. Livingstone (1962) published a paper "On the non-existence of the human races" in which he declared "There are no races, there are only clines" (p. 279). Clines are hybrids between two pure races. Clines invariably appear at the junction between races who interbreed and produce mixed-racial hybrids. Thus, in Latin America there is a large population of Mestizos, who have European and Amerindian ancestry and can be considered a cline. Similarly, the Pacific Islanders are a mixed race cline derived from the interbreeding of Southeast Asians and East Asians. It has often been asserted that the existence of intermediate forms, clines, or hybrids invalidates the concept of races. This is obviously not the case. Among dogs, clines and hybrids are called mongrels, but the existence of mongrels does not mean that there are not pure breeds.

However, in the next decade the geneticists Walter Bodmer and Luigi Cavalli-Sforza (1976, p. 698) were to write of "the existence of many different racial groups in man" and that the "races could be called sub-species if we adopted for man a criterion from systematic zoology. The criterion is that two or more groups become sub-species when 75 percent or more of all individuals constituting the groups can be unequivocally classified as belonging to a particular group." They go on to say that when human races are defined broadly, it is possible to identify the race of many more than 75 percent of the population. Hence races certainly exist among humans. Some twenty years later this same Luigi Cavalli-Sforza opted to go with the flow and we find him writing of the "scientific failure of the concept of human races" and that "the concept of race has failed to gain any acceptance" (Cavalli-Sforza, Menozzi, and Piazza, 1994, p. 19). However, they write "we can identify 'clusters' of populations." These clusters turn out to be the same as the races of classical anthropology and later in their book we find the authors using the classical racial terminology. For instance, they write that Africa "is inhabited by two aboriginal groups, Caucasoids in the north almost down to the southern borders of the Sahara, and Negroids in sub-Saharan Africa" (p. 167). Evidently they had forgotten their previous assertion that the "scientific failure of the concept of human races human species can only be divided into 'clusters'" (a transparent euphemism for races). Only six years later this same Luigi Cavalli-Sforza apparently changed his mind again because he pronounced that races do exist and that a race can be defined as "a group of individuals that we can recognize as biologically different from others" (Cavalli-Sforza, 2000, p. 25). It appears that he has made a resolution to deny the existence of race but every now and then he forgets and the r— word slips out.

By the beginning of the twenty-first century the denial of the existence of races became increasingly frequent. In 2004 the American Anthropological Association announced on its website that "race is not a scientifically valid biological category." "There are no biological races," asserts Jefferson Fish (2002, p. xii), a professor of psychology at St. John's University in New York, but he does not explain the grounds on which he makes this assertion. Graves (2002, p. 2-5), a biologist at the University of Arizona, also asserts that "biological races do not exist" and writes that "the term race implies the existence of some nontrivial underlying hereditary features shared by a group of people and not present in other groups," and that this is not true for human races. Contrary to this assertion, there are a number of "hereditary features" that are present in some races and absent in others. For instance, the genes for black skin are present in Africans and absent in Europeans, East Asians, and American Indians, while the genes for the epicanthic eyefold are present only in East Asians, Arctic peoples, and in some American Indians. Furthermore, the concept of race need not imply that there are some alleles (alleles are alternative forms of genes) that are only present in some races but are absent in others. It is sufficient that there are differences in allele frequencies between different races. There are a number of alleles for which this is the case. For example, the allele for sickle cell anemia is much more frequent in Africans than in other races, while the allele for cystic fibrosis is much more common in Europeans (Table 2.1, p. 12).

Graves (2002, p. 5) writes "The majority of geneticists, evolutionary biologists and anthropologists agree that there are no biological races in the human species." Cohen (2002, p. 211) likewise asserts "Almost all anthropologists agree that races in the popular sense do not exist and never have existed." These assertions are incorrect. A survey of the views of American anthropologists carried out in 1985 found that the existence of races was accepted by 59 percent of biological and physical anthropologists and about one third of cultural anthropologists (Lieberman and Reynolds, 1996).

Despite the denials of the existence of race by a number of American anthropologists, the reality of race is widely accepted throughout the rest of society. Medical journals contain numerous papers on race differences in a variety of diseases and disabilities, including the prevalence of HIV infection. There is a journal Ethnicity and Health devoted to racial differences in the prevalence of diseases. In the social sciences there are two journals devoted to race differences (Race and Class and Ethnic and Racial Studies) and other journals contain numerous papers on race differences in intelligence, educational attainment, earnings, socio-economic status, unemployment, prejudice, discrimination, alcohol consumption, tobacco use, drug addiction, sexual experience, longevity, crime, and mental retardation. Corporations promote equal opportunities for the races in their employment. Employees sue corporations for racial discrimination and frequently obtain substantial compensation awarded by juries who have no problem in understanding the meaning of race. Many universities exercise positive discrimination in favor of black and Hispanic applicants. Judges pronounce that racially segregated schools are unconstitutional. Citizens in many countries state their race in census returns and these are analyzed by sociologists and demographers. In Britain there is a Race Relations Commission whose task is to promote racial equality and prosecute employers for racial discrimination. Neither the people responsible for this work nor the general public has any difficulty in understanding what race means and no doubt would be amazed to learn that many American anthropologists assert that race does not exist.

It may be wondered why a number of American anthropologists reject the concept of race. The answer has been given by two Polish anthropologists, Kaszycka and Strkalj (2002, p. 334). They write:

Americans have become very sensitive to race, and the term has acquired strongly sensitive connotations. Many American scientists have opted for the non-existence of human races. Furthermore, the growing demands of "political correctness" militate against the use of the term in and outside science.... Few scientists dare to study racial origins, lest they be branded racists simply for being interested in the problem.

The reason for the rejection of the concept of race by a number of American anthropologists is apparent from the title of Montagu's book Man's Most Dangerous Myth. Montagu evidently believed that people's consciousness of race is dangerous because it tends to foster racial antagonisms that can escalate into conflict. To prevent this it would be better for the concept of race to be suppressed. In Europe most anthropologists accept the validity of the concept of race. Thus, a survey of Polish anthropologists carried out in 2001 found that 75 percent agreed with the proposition "There are biological races within the species Homo sapiens" (Kaszycka and Strzalko, 2003). It is mainly in the United States that the existence of race has come to be denied by a number of anthropologists and a few biologists and social scientists who have sacrificed their scientific integrity to political correctness.

Chapter 3. Europeans

1. Intelligence of Indigenous Europeans

2. Europeans outside Europe

3. European University Students

4. Brain Size

5. The Heritability of Intelligence in Europeans

The europeans have been recognized by all the classical anthropologists as one of the major races. Linnaeus (1758) described them as Europaeus. They have frequently been designated Caucasians or Caucasoids because of the belief that they originated in the Caucasus. A number of anthropologists have categorized them together with the South Asians and North Africans in a single Caucasoid group. However, the Europeans are distinguishable from the South Asians and North Africans by their lighter skin color and, in the northern Europeans, blonde hair and blue eyes. The distinction between the Europeans and the South Asians and North Africans has been confirmed by Cavalli-Sforza, Menozzi, and Piazza (1994) in their classification of the human races on the basis of a number of genetic markers. This has shown that Europeans represented by Italians, Danes, English, and Basques comprise a homogeneous "cluster" differentiating them from other races. Coon, Garn, and Birdsell (1950), Cole (1965), and a number of other anthropologists have sub-divided the Europeans into seven sub-races consisting of the Mediterranean peoples of Spain, Italy, and southeast Europe; the Alpine peoples of France and central and southern Germany; the Nordic peoples of England, the east of Ireland, and Scotland, the Netherlands, Belgium and Northern Germany, Denmark, Norway, Sweden, and Western Finland; the Celtic peoples of Wales, the west of Ireland, and the western highlands of Scotland; the Dinaric peoples of east-central Europe; the Slavic peoples of northern Poland, the Baltic states, and Russia west of the Urals; and the Basques of northern Spain and southwest France. The Nordic peoples have lighter skin color, blonde hair, and blue eyes, while the central and south Europeans more typically have darker skins, darker or black hair, and dark eyes.

1. Intelligence of Indigenous Europeans

Studies of the IQs of Europeans in Europe are summarized in Table 3.1. These IQs are calculated in relation to a British mean of 100 and standard deviation of 15. Twenty-one of the studies were carried out by Buj (1981) on samples of adults from major cities. Most if the remainder are derived from one of the three versions of the Progressive Matrices (CPM, SPM, and APM). Row 61 giving an IQ of 89 for Serbia is probably a shade too low because the sample is described as being from "predominantly lower or lower middle class families" in and around Belgrade (Moyles and Wolins, 1973, p. 372). The range of IQs of the Europeans is from 87 for one of the studies in Ireland and 88 for one of the studies in Greece to 107 for one of the studies in Germany and the Netherlands. There are also some inconsistencies in the same countries, where the IQs typically differ by two or three IQ points and in the cases of Portugal and Poland by as much as 13 and 14 IQ points. These differences are partly caused by sampling errors and are partly genuine, arising from differences in living standards and possibly from sub-racial differences in Europe. Sampling errors in studies of the intelligence of national populations arise in the same way as in opinion polls on voting intentions, where normally several polls carried out at the same time give results that differ by a few percentage points. We should not search for the meaning of differences of a few IQ points between studies when in many cases these are simply sampling errors. The important thing is to look for general patterns.

The only significant general pattern of the IQs in Europe appears to be that IQs are a little lower in southeast Europe than in the remainder. In the Balkans IQs are 94 for Romania, 92.5 (the average of the two stud ies) for Bulgaria, 90 for Croatia, 89 for Serbia, and 92.5 (the average of the four studies) for Greece. The probable explanation for this is that the Balkan peoples are a hybrid population or cline, comprising a genetic mix between the Europeans and South Asians in Turkey. Hybrid populations or clines arise in the borderlands between two races as a result of interbreeding. In the Balkans such a cline evolved because of the close geographical proximity between southeast Europe and Turkey, and the occupation of large territories in southeast Europe by Turkey for a number of centuries during the time of the Ottoman empire. This has brought about a mixing of Turkish and European genes with the result that contemporary Turks and Greeks are genetically quite similar. This has been shown by Cavalli-Sforza, Menozzi, and Piazza (1994) in their genetic linkage tree, in which Greeks are shown to be more closely related to Iranians and other southwest Asian peoples than to Italians, Danes, and English. This genetic similarity is also apparent for intelligence, for which the IQ of 90 in Turkey is closely similar to those in the range of 90 to 94 in Greece, Romania, Bulgaria, and Croatia in southeast Europe. Because the peoples of southeast Europe are a cline it is considered appropriate to exclude these in estimating the European IQ. The median IQ of the remaining countries is 99 and is considered the best estimate of the IQ of Europeans.

Apart from the lower IQs in the Balkans, there are three other countries with IQs somewhat lower than the European average. The first is Lithuania, with an IQ of 90-92. These low figures may be sampling errors because they are rather lower than in neighboring Russia (97), Poland (99), and Estonia (99). The second is Ireland, for which the mean IQ of the four studies is 92. The most probable explanation for this is the long history of emigration in which there has been some tendency for the more intelligent to migrate, leaving the less intelligent behind. This has also occurred in Scotland, where the average IQ is 97, and in Corsica, where the average IQ is lower than in mainland France (Lynn, 1979, 1980). The third country with a slightly depressed IQ is Portugal, for which the two results are IQs of 101 and 88, which can be averaged to 94.5. The depressed IQ in Portugal is consistent with its having the lowest per capita income in western Europe and its modest intellectual achievement. The Portugese have only won one Nobel Prize for science out of the 346 awarded during the period 1901-2003. This was awarded in 1949 to the neurosurgeon Antonio Moniz for the innovation of the operation of prefrontal leucotomy as a treatment for mental illness, and is not now considered a desirable therapy. It may be that intelligence in Portugal has been depressed by the admixture of sub-Saharan Africans in the population. Portugal was the only European country to import black slaves from the late fifteenth century onwards for agricultural and domestic work. According to Du Bois (1939, pp. 132-133), in the sixteenth century blacks outnumbered whites in Lisbon and in the plantations of the Algarve in the south of the country. This may be an exaggeration, and it may be that the proportion of blacks has declined in succeeding centuries. Nevertheless, if the present population of Portugal contains 20 percent of African descent and the IQ of the Africans is 70, this would be expected to produce a population with an IQ of 94.

It may be surprising that there does not appear to be much difference between IQs in the twelve former communist countries of Eastern Europe, among which the median IQ is 96, and the 14 countries of western Europe, among which the median is 98.5. The difference is small and not statistically significant, so it seems that although the former communist countries have had much lower living standards for some sixty years following the end of World War II, this has not impaired the intelligence of the populations.

Table 3.1. IQs of indigenous Europeans

2. Europeans outside Europe

Europeans have migrated to many parts of the world. Studies of the intelligence of these populations are summarized in Table 3.2. Rows 1 and 2 give IQs of 93 and 98 for Argentina. Row 3 gives an IQ of 97 for Australia based on a standardization of the American Otis test. Row 4 gives an IQ of 100 for Australia derived from the administration of the SPM to National Servicemen (the IQ of this sample was 102, but because men obtain higher

Table 3.2. IQs of Europeans outside Europe

mean IQs than women by approximately 5 IQ points on this test (Lynn and Irwing, 2004), the figure has been reduced to 100). Row 5 gives an IQ of 98 for a sample of young Australian children. Row 6 gives an IQ of 95 for European children in Brazil from Sao Paulo. Row 7 gives an IQ of 97 for Canada obtained from a sample of 7 to 12 year olds. Row 8 gives an IQ of 100 for Canada obtained from the standardization of the WISC-111 on a representative sample of 2,200 6-16 year olds.

Row 9 gives an IQ of 99 for Chile based on a study finding that European students at the Universidad Catolica de Valparaiso had the same IQ as Austrian students (n=320). Row 10 gives an IQ of 95 for European children in Colombia. Row 11 gives an IQ of 98 for European children in Baja California in Mexico. Row 12 gives an IQ of 99 for New Zealand obtained from a standardization of the Otis test in the 1930s. Row 13 gives an IQ of 101 derived from the standardization of the Progressive Matrices. Row 14 gives an IQ of 102 obtained from the Christchurch Child Development Study. Row 15 gives an IQ of 94 for European 16-year-olds in Natal in South Africa. Rows 16 through 21 give six IQs in the range between 99 and 103 for Europeans in theUnited States compared with those in Britain. The IQ of 100 given in row 20 is derived from the standardization of the WAIS-3 in Britain. Row 22 gives an IQ of 96 from a standardization of the Progressive Matrices in Uruguay. Row 23 gives an IQ of 100 for European 7-year-olds in Zimbabwe.

The median of these IQs is 99, the same as that of Europeans in Europe. The results show that even in the quite poor countries of Latin America (Argentina, Brazil, Colombia, Mexico, and Uruguay), which have per capita incomes about one third of those in North America and Western Europe, the IQs of Europeans are only fractionally below those in affluent nations. This confirms the results in Europe, where the much poorer former communist countries have about the same IQs as the affluent Western countries.

3. European University Students

Studies of the intelligence of European university students are summarized in Table 3.3.

All the samples have IQs of 100 or above, as would be expected, and the median IQ is 105. The principal interest of the results is for comparison with university students in Africa and South Asia, where IQs are typically about 10 to 20 points lower.

 Table 3.3. Intelligence of European university students

4. Brain Size

We noted in Section 1 that IQs are lower in Southeast Europe and in the Iberian Peninsula than in the remainder of Europe. We would expect that these differences would also be present in brain size because of the correlation between brain size and intelligence of 0.40 (Vernon, Wickett, Bazana, and Stelmack, 2000). We look now at differences within subpopulations of Europeans to see whether this is the case. The data on brain sizes of a large number of populations collected by Jurgens, Aune, and Pieper (1990) are shown in Table 3.4 together with IQs. Row 1 shows that Europeans in North America have the largest brain size and IQ. Row 2 shows that these are followed by Europeans in North, Central, and Eastern Europe. Row 3 shows slightly smaller brain size and IQ in Spain and Portugal. Row 4 shows a continuation of the downward trend with smaller brain size and IQ in Southeast Europe. Row 5 shows a further continuation of the downward trend with smaller brain size and IQ in the Near East obtained from samples of South Asians from Turkey and Iraq. Row 6 shows the lowest brain size and IQ in South Asians in India. Details of the IQs of the South Asians in Turkey, Iraq, and India are given in Chapter 6.

Table 3.4. Brain size (cc) and intelligence in Europeans and South Asians

5. The Heritability of Intelligence in Europeans

The heritability of intelligence is the extent to which differences in intelligence are determined by genetic factors. We are interested here in the question of the heritability of race differences in intelligence, but before discussing this we need to consider the heritability of individual differences in intelligence within countries. There are three sources of evidence on this problem. These consist of studies of identical twins reared apart, a comparison of identical and non-identical twins reared in the same families, and a comparison of unrelated adopted children reared in the same families. All three kinds of evidence show that the heritability of intelligence for adults is approximately 0.80, or 80 percent. This means that if all individuals were reared in identical environments, the differences between individuals would be reduced to 80 percent of the actual differences.

Studies on the heritability of intelligence for adults and children have been summarized by Bouchard (1993, p. 58). For adults, the evidence from identical twins reared apart is based on five studies for which the average correlation weighted by sample size is 0.75. This figure needs to be corrected for test reliability (correction for attenuation), for which a reasonable figure is about 0.9 (Bouchard, 1993, p. 49; Mackintosh, 1998). This correction increases the correlation to 0.83. This is a measure of the heritability. The evidence from a comparison of the degree of similarity between identical twins and same-sex, non-identical twins brought up in the same families is that there is a correlation of 0.88 for identical twins and 0.51 for same-sex non-identicals. Correcting the correlations for the reliability of the tests and adopting a reliability coefficient of 0.9, the corrected correlations become 0.98 for identicals and 0.56 for same-sex non-identicals. The heritability can be calculated by Falconer's (1960) formula consisting of doubling the difference between the correlations of identical and same-sex non-identicals. The difference between the two correlations is 0.42, and doubling this difference gives a heritability of 0.84.

A third method for estimating the heritability of intelligence is to examine the correlation between the IQs of unrelated children adopted and reared in the same families. The magnitude of the adopted family environmental effect (the "between family effect") is expressed by the correlation between the twin pairs. The summary of the research literature by Bouchard (1998) concludes that among adults the correlation is 0.04, indicating a heritability of 0.96. However, this method underestimates the environmental effect because it does not take into account effects operating on one child but not on the other, such as prenatal and perinatal effects. The two twin methods yielding heritabilities of intelligence of 0.83 and 0.84 are more accurate. These figures are very close to the estimate of approximately 0.85 given by Jensen (1998, p. 179).

The heritability of intelligence among children is considerably lower, at approximately 0.42 among 4-6 year olds and 0.55 for the age group 6 to 20 (Bouchard, 1993, p. 58; Jensen, 1998, p. 179). The reason for this is probably that parents exert environmental effects on children that progressively wear off during adolescence. It is by including the lower heritability figures derived from children with the higher figures for adults that some scholars put the heritability of intelligence at between around 0.40 to 0.80. For instance, in a statement drawn up by Gottfredson (1997, p. 14) and endorsed by 52 experts, it is stated that "Heritability estimates range from 0.4 to 0.8, most indicating that genetics plays a bigger role than environment in creating IQ differences among individuals." Most of the studies from which these high heritability figures are obtained come from Europeans in affluent western nations. However, a study of 144 identical and non-identical twin pairs in Russia yielded a heritability of 0.78, which corrected for test unreliability is increased to 0.87 (Lipovechaja, Kantonistowa, and Chamaganova, 1978).

The conclusion that intelligence has a high heritability implies that there are genes that determine intelligence. The first of these in normal populations was discovered in the late 1990s by Chorley et al. (1998). It lies on chromosome 6, and possession of one of the alleles of this gene contributes about 4 IQ points to an individual's intelligence.

Chapter 4. Africans

1. Intelligence of Africans in Sub-Saharan Africa

2. University Students in Africa

3. Africans in the Caribbean and Latin America

4. African Americans in the United States

5. Africans in Britain

6. Africans in the Netherlands

7. Africans in Israel

8. Short-Term Memory and Perceptual Speed Abilities of Africans

9. Musical Abilities

10. Reaction Times

11. Brain Size

12. African-European Hybrids

13. Heritability of Intelligence in African Americans

14. Environmental and Genetic Explanations of African-European IQ Differences

15. Estimation of the Genotypic African IQ

The term africans is used here for the principal indigenous peoples of sub-Saharan Africa. They should be distinguished from the North Africans, indigenous to Africa north of the Sahara; from the pygmies; and from the Bushmen or Khoisans, the other race in sub-Saharan Africa, of whom only a few tens of thousands now survive, principally in the Kalahari Desert and as Hottentots in South Africa. A variety of terms have been used for the African peoples, including Afer (Linnaeus, 1758), Ethiopians (Blumenbach, 1776), and Negroids (Cole, 1965). Whatever the name used, the Africans have always been regarded as one of the major races in the taxonomies of classical anthropology, including that of Linnaeus (1758), Blumenbach (1776), and Coon, Garn, and Birdsell (1950). Cavalli-Sforza, Menozzi, and Piazza (1994) have confirmed the distinctive genetic characteristics of the Africans in their classification of humans into genetic "clusters," in which these peoples were represented by west Africans of the region west of Nigeria, Nilotics of the upper Nile in southern Sudan, Ethiopians, and Bantus, a large group present in most of sub-Saharan Africa from Nigeria in the west to Kenya. The most distinctive features of Africans are their very dark skin, dark eyes, broad nose, thick everted lips, and woolly hair. Their blood groups differ from Europeans in having a lower frequency of group A, which is present in about 27 percent as compared with around 46 percent in Europeans, and a higher frequency of group B, which is present in about 34 percent as compared with around 14 percent in Europeans.

1. Intelligence of Africans in Sub-Saharan Africa

The first attempt to estimate the intelligence of Africans was made by Galton (1869) on the basis of his own experience of them during his travels in southwest Africa and the accounts of other travelers. He constructed a scale of grades of intelligence in which one grade was equivalent to 10.425 IQ points on the IQ scale. He estimated that Africans were about two grades below the English, giving them an IQ of 79. Subsequent studies of the IQs of general population samples of Africans in sub-Saharan Africa have shown that this estimate overestimated the African IQ by slightly over one grade.

Studies of the IQs of Africans in sub-Saharan Africa are summarized in Table 4.1. Explanations of the results set out in the table are given when appropriate. Row 1 gives an IQ of 64 for Cameroon for adult workers. Row 2 gives an IQ of 64 for the Central African Republic for young men applying for a technical training course at a college in the city of Bangui during the years 1951-1955. Rows 3 through 5 give IQs of 64 for samples from Congo-Brazzaville collected at the same time in the cities of Brazzaville and Pointe-Noire. Rows 6, 7, 8, 9, and 10 give IQs of 64, 68, 62, 68, and 65 for Congo-Zaire. Row 11 gives an IQ of 59 for Equatorial Guinea. Row 12 gives an IQ of 80 for adults in Ghana. The IQ is exceptionally high for sub-Saharan Africa, possibly because the sample came from the capital city of Accra; the people in capital cities typically have higher IQs than those in the rest of the country, probably because there is a tendency for more intelligent individuals to migrate to the capital; IQs in London and Paris are higher than in the rest of Britain and France (Lynn, 1979, 1980). Row 13 gives an IQ of 62 for a representative sample drawn from the whole

Table 4.1. IQs of Africans in sub-Saharan Africa

of Ghana. Rows 14 and 15 give IQs of 63 and 70 obtained in two studies for Guinea. Rows 16 through 21 give IQs of 69, 75, 69, 76, 89, and 63 for Kenya. The IQ of 89 in row 20 for a sample of 7-year-olds tested in 1998 is much higher than the other figures and the IQ of 76 found by the same investigators in their 1984 study (row 19) and than any other IQ in sub-Saharan African populations. Its disparity from the other studies makes its validity questionable because the IQ of 75 given in row 17 is obtained from a standardization of the same test for the whole of Kenya carried out in the same year, and the IQ of 69 given in row 19 was also obtained in the same year. These two IQs are typical of those obtained throughout sub-Saharan Africa and are credible, but they cast doubt on the IQ of 89. Furthermore, the gain of 15 IQ points from an IQ of 76 to 89 over the 14-year period is uniquely high in studies of the secular rise of IQs and cannot be accepted as credible. Further, it is difficult to believe that children in Kenya can have a higher IQ than African Americans in the United States, where the IQ has remained constant at approximately 85 since the 1920s but where the living standards and nutrition of Africans are much higher than in Kenya. For these reasons the reported IQ of 89 for Kenya is considered unreliable. The IQ of 63 given in row 21 is an average of 65 for 6-year- olds at school and 61 for those not at school, suggesting that the effect of schooling is to raise the IQ by 4 points.

Row 22 gives an IQ of 82 for Madagascar. Although usually counted as part of sub-Saharan Africa, the population of the island includes a significant number of Southeast Asians originally from Indonesia who migrated to the island about the first century AD (Cole, 1965). The population also contains Africans and hybrids of the two races. The proportions of the three groups in the population are not precisely known although it is believed that African ancestry predominates. The mean IQ of 82 is higher than that of any of the samples of the Africans in sub-Saharan Africa given in Table 4.1 except for the questionable 89 for Kenya given in row 20. The population's IQ is intermediate between that of around 65-70 of Africans and around 87 of Southeast Asians (see Chapter 7), although somewhat closer to that of Africans, as would be expected for a Southeast Asian and African mixed race population in which African genes predominate. There is no apparent environmental explanation for why the IQ of the population of Madagascar should be higher than that throughout mainland sub-Saharan Africa.

Row 23 gives an IQ of 64 for Mozambique. This sample had a mean of 3.5 years schooling and included some individuals from Transkei and Malawi. Row 24 gives an IQ of 70 for Nigeria obtained for children (age not given) attending schools in the town of Zaria. An elite sample of boys at grammar school (number=179) in the same town obtained an IQ of 81. The test was the Leone Test and is described by the author as "devised by an African for African children" (Farron, 1966, p. 53). The result belies the assertion often made that Africans are handicapped on tests constructed by Europeans. Rows 25 and 26 give IQs of 64 and 69 for two further studies in Nigeria.

Rows 27 and 28 give IQs of 64 for two samples of adults in Sierra Leone. Row 29 gives an IQ of 65 obtained in the first study of the intelligence of Africans in South Africa, carried out in the 1920s. Rows 30 through 41 give IQs in the range between 58 and 77 obtained in twelve later studies. Row 35 gives an IQ of 71 for prison warders who had had between 9-12 years of education. Row 36 gives an IQ of 77, which is the highest for general population samples in sub-Saharan Africa but was obtained from the Draw-a-Man test, which is a rather poor test of general intelligence. Row 39 gives an IQ of 58 for a large sample of 16-year-old Africans in school who had completed approximately ten years schooling. The comparison is with South African Europeans. Row 40 gives an IQ of 69 for a sample of adults described as "competent men, all in long standing employment in a sophisticated environment...." (Nell, 2000, p. 27). Row 41 gives an IQ of 68 for adolescents with a few years of school­ing in the Northern Transvaal. Rows 42 and 43 give IQs of 67 and 64 for samples of adolescents at school in Soweto. Row 44 gives an IQ of 67 for third-grade Zulu school children in Natal.

Row 45 gives an IQ of 69 for Sudan obtained for Shilluk children and adolescents described as "one of the primitive Nilotic Negro tribes" (p. 164) in the Southern Sudan. The IQ given is the mean of four tests: the Goddard and Porteus Mazes, Alexander Passalong, and Draw-a-Man. Rows 46, 47, and 48 give IQs of 64, 74, and 73 for three further studies in the Sudan.

Row 49 gives an IQ of 78 for a sample of secondary school students in Tanzania and is exceptionally high for African samples. The author of the study explains that the reason for this is that the sample was highly selected because "the number of places in secondary school is extremely limited and eligibility is determined by competitive examination" (Klingelhofer, 1967, p. 207). The high IQ of this sample cannot be regarded as representative. The result is informative in so far as it shows that an elite sample at secondary school has an IQ of 78 and this suggests that the IQs in the range of 65-72 typically found in sub-Saharan Africa are valid. Rows 50 and 51 give IQs of 65 and 72 for two more representative samples in Tanzania and are consistent with those typical in sub-Saharan Africa. Row 52 gives an IQ of 80 for Uganda for a selective sample of schoolchildren described by Vernon (1969, p. 182) as "much superior to the East African population in general." This explains why the IQ is higher than that of representative samples in sub-Saharan Africa. Row 53 gives an IQ of 73 for a large and representative sample of Ugandan children. Rows 54 and 55 give IQs of 77 and 64 for Zambia. Rows 56 and 57 give IQs of 61 and 70 for Zimbabwe obtained by Zindi, an African psychologist at the University of Zimbabwe.

The most striking feature of the IQs of Africans in sub-Saharan Africa is that they are consistently so much lower than those of Europeans set out in Table 3.1 of Chapter 3. The median IQ is 67 and is adopted as the best estimate of the IQ of Africans. With the exceptions of the IQ of 82 for Madagascar, which is the highest in the table because of the Southeast Asian element in the population, and the IQs of 78 for the elite secondary sample in Tanzania (row 48) and 80 for the elite secondary sample in Uganda (row 51), and the questionable 89 for Kenya (row 21), all the IQs fall in the range of 59 to 77, while all the European Caucasoid IQs fall in the range of 87 to 105. There is no overlap between the IQs of the two populations. The variations of the African IQs do not appear to vary by geographical location and are probably attributable to sampling and measurement errors. The IQ of Africans has not shown any change since the first study published by Pick (1929) obtained an IQ of 65 for Africans in South Africa. The four most recent studies of Africans in South Africa carried out in the 1990s found virtually identical IQs of 69 (Nell, 2000), 68 (Sonke, 2000), 67 and 64 (Skuy et al., 2001).

2. University Students in Africa

Twelve studies have been reported of the intelligence of African university students in South Africa. Some of these also give IQs of European students tested at the same time. The studies are summarized in Table 4.2. Row 1 gives an IQ of 75 for African students at Legon University in Ghana tested with the Block Design (Kohs Blocks) test from the Wechsler Test. All the remaining rows give results for South Africa. Row 2 gives an IQ of 84 for African and 103 for European university students calculated in relation to American adult norms given in Raven, Court, and Raven (1994). Rows 3 and 4 give results for students on the Blox test and gives the IQs of Africans in relation to South African European student norms of 100. Row 5 gives results for the WAIS-R for students with an average age of 25 years at the African universities of Fort Hare, Zululand, the North, and the Medical University of South Africa. The Verbal IQ was 78 and the Performance IQ 73, showing once again that the Africans have low IQs in all major cognitive abilities and disconfirming the claim sometimes made that Africans are handicapped in language tasks. Row 6 gives an IQ of 100 for science students at the University of the North. Row 7 gives an IQ of 77 for students at a less prestigious African university. Row 8 gives an IQ of 83 for students at the University of the Witwatersrand and the Rand Afrikaans University in Johannesburg. Row 9 gives an IQ of 82 for African students at the Venda University in the Northern Transvaal. The comparison European group was at the University of Tilberg in the Netherlands. Row 10 gives an IQ of 81 for psychology students at the University of the Witwatersrand. Row 11 gives an IQ of 93 for first year engineering students at the University of the Witwatersrand. Row 12 gives an IQ of 99 for a slightly reduced number of the same students who took the Advanced Progressive Matrices 16 months later. Both Africans and Europeans obtained IQs approximately 6 points higher on the second testing, probably as a practice effect. Row 13 gives an IQ of 101 for a further sample of African engineering students at the University of the Witwatersrand and shows that the African students scored 15 IQ points lower than the European whites.

The mean IQs of general student samples shown in rows 1 to 5 and 7 to 9 all fall in the narrow range of 72 to 84 with a median of 81. The IQs of 100 in row 6, 93 in row 11, and 99 in row 12 are higher than the others because they are for science and engineering students who were admitted to the universities on the basis of their performance in entrance tests of mathematics and physics, and these normally have higher reasoning ability than students in most other academic disciplines. For instance, in Iran 18-year-olds studying math scored 10 IQ points higher than those studying literature (Mehryar, Shapurian, and Bassiri, 1972). In Britain, education students with degrees in science scored 9 IQ points higher than those with degrees in arts (Heim, 1968). The IQs of European students in South Africa are in the range between 100 and 105 and are about the same as those of European students in other countries (see Chapter 3, Table 3.3). The interest of these results is that they show that typical African students who have had some 12 years of school and have gained entry to university obtain IQs in the range of 72-84. Since these are an African cognitive elite, these results suggest that the IQ of about 70 for the general population is valid and about right. The results also show that IQs of African students in South Africa are on average about 20 IQ points lower than those of European students, and that a considerable gap between the IQs of Africans and Europeans remains when they are matched for years of education, African university students have had ten to twelve years of formal education but apart from those studying math and physics, obtain IQs in the range of 72-84. Their IQs are some 10 to 12 IQ points higher than the African average because they are a select group.

Table 4.2. IQs of African and European university students in Africa

3. Africans in the Caribbean and Latin America

Studies of the IQs of Africans in the Caribbean and Latin America are summarized in Table 4.3. Row 1 gives an IQ of 80 for children in Barbados; this figure has been calculated from the IQ of 83 of well-nour ished children arid 68 of malnourished children reported in the study, weighted by the results of a 1968 survey finding a prevalence of moderate and severe malnutrition in preschool children in Barbados of 16.5 per cent (Galler, Ramsay, Solimano et al., 1983). Row 2 gives an IQ of 70 for Africans in Brazil attending school in a favela (shanty town) in Brasilia. Row 3 gives an IQ of 64 for the mothers of these children. Row 4 gives an IQ of 71 for Africans in Sao Paulo in Brazil. Row 5 gives an IQ of 67

Table 4.3. IQs of Africans in the Caribbean and Latin America

for 3-year-old African children in Dominica. The low IQ of these infants suggests that poor education is not a factor responsible for the low IQs of Africans in the Caribbean. Rows 6 through 12 give IQs from seven studies of the IQ in Jamaica in the range of 60-75 with a median of 67. Row 13 gives an IQ of 60 for 4-year-olds in St. Lucia and row 14 an IQ of 70 for children in St. Vincent. The median of the fourteen studies of intelligence of Africans in the Caribbean and Latin America is an IQ of 71. This is slightly higher than the median IQ of 67 of Africans in sub-Saharan Africa. The explanation for this may be that Africans in the Caribbean and Latin America have some admixture of genes from Europeans. It has been estimated that the proportion of European genes in the African population of Jamaica is 6.8 percent (Parra, Marcini, and Akey, 1998).

4. African Americans in the United States

There have been many hundreds of studies of the intelligence of African Americans in the United States. The most important of these are summarized in Table 4.4. Row 1 gives results of the first major study based on military conscripts in World War I tested with the combined Army Alpha and Beta tests that measured non-verbal and verbal IQs and from which the later Wechsler tests were constructed. The number of Europeans was 93,973. Row 2 gives results for military conscripts in World War II and row 3 the results of military conscripts for the Vietnam War. It is noteworthy that the mean IQ of 77 of Africans is lower in World War II and the Vietnam War than in World War I, and is also lower than the average IQ of 85 that is generally given for the mean IQ of African Americans in the United States. Rows 4 through 7 give results of Shuey's compilation of all American studies. Row 4 gives an IQ of 87 derived from 17 studies of pre-school children. Row 5 gives an IQ of 85 derived from 26 studies of primary school children using individual tests such as the Stanford-Binet. Row 6 gives an IQ of 85 for primary school children derived from 103 studies for group tests of verbal ability and 41 studies of group tests of non-verbal ability. Row 7 gives an IQ of 85 for high school students. Rows 8, 9, and 10 give the results of Osborne and McGurk's (1982) updated summary of American studies published during 1976 through 1980. Row 8 gives an IQ of 80 derived from 66 studies of preschool 3-5 year olds. Row 9 gives an IQ of 87 derived from 126 studies of primary school children and row 10 an IQ of 87 derived from 17 studies of high school students.

Rows 11, 12, and 13 (Broman et al., 1975) give results for large samples not included in the Osborne and McGurk review. Row 11 gives an IQ of 85 for African mothers tested in the National Collaborative Perinatal Project

Table 4.4. IQs of African Americans in the United States

and rows 12 and 13 give IQs of 87 for their children at the age of 4 years and 7 years. Row 14 gives a g IQ of 86 for Africans from the standardization sample of the WISC-R. Row 15 gives IQs of 81 for g, 86 for verbal, and 84 for visualization for employed individuals collected by the United States Employment Service. Row 16 gives African-American IQs of 85 for g, 87 for verbal, and 86 for visualization for a sample in California. Row 17 gives IQs of 85 for g, 87 for verbal ability, and 86 for visualization abil­ity obtained from the standardization sample of the WAIS-R. Row 18 gives an IQ of 82 from the AFQT. Row 19 gives an IQ of 93 from the standard­ization sample of the K-ABC. Row 20 gives a vocabulary IQ of 85 from the standardization sample of the Peabody Picture Vocabulary Test. Row 21 gives an IQ of 83 from the standardization sample of the Stanford-Binet-4; in this sample African Americans obtained a short-term memory IQ of 89 consistent with a number of other studies finding they do relatively well on short term memory. Row 22 gives an IQ of 85 for 3-year-olds from the standardization sample of the Stanford-Binet-LM. Row 23 gives an IQ of 83 calculated from the first principal component as a measure of g obtained from military personnel. Row 24 gives an IQ of 85 from the standardization sample of the WISC-3. Row 25 gives an IQ of 88 from the standardization sample of the Kaufman Adolescent and Adult Intelligence Test. Row 26 gives an IQ of 81 for a sample of employed individuals collected by the United States Employment Service. Row 27 gives a visualization IQ of 85 derived from the block design subtest of the WISC-R obtained from the national NHANES III sample.

Row 28 gives an IQ of 85 for infants aged 3.0 to 3.4 years from the standardization sample of the Stanford-Binet-4. Row 29 gives an IQ of 85 for a representative sample aged 70 and older from the continental United States (i.e., excluding Alaska and Hawaii). Row 30 gives an IQ of 90 for vocabulary for African adults obtained in the NORC surveys for 1990-96 from a representative sample from the continental United States. This unusually high figure is attributable to the shortness of the test, consisting of defining the meaning of ten words. Row 31 gives an IQ of 87 from the 1992 National Adult Literacy Survey, a test consisting of verbal comprehension and arithmetic administered to a representative sample from the continental United States.

There are five conclusions to be drawn from the studies of the intelligence of African Americans. First, the median IQ is 85 and is widely accepted as the best estimate of the African-American IQ. This estimate is close to the 83.5 obtained by Roth, Bevier, Bobko, Switzer, and Tyler (2001) from a meta-analysis of 105 studies based on 6,246,729 individuals. The variations in the means obtained in different studies are probably due to sampling, measurement errors, and differences in the abilities measured in different rests. It has been shown in many studies that Africans do relatively well in rests of memory, so the size of the African-European difference reflects to some degree the extent to which memory tests are represented in the IQs. For instance, one of the higher IQs in the table is the 88 obtained in Kaufman's KAIT. This test contains seven subtests, of which one is a memory for faces test that requires the identification of the faces of famous people. On this subtest Africans obtained a mean IQ of 92.5.

Second, the African-American IQ of approximately 85 appears in children aged 3, as can be seen in rows 22 and 28. These results tell against the theory often advanced by environmentalists that poor education and racism are responsible for or contribute to the low IQ of Africans. Even among 2-year-olds Africans have an IQ of 92 (row 19). This is not so low as in the other studies because African infants mature earlier than Europeans up to the age of two years (Lynn, 1998d; Rushton, 2000). It is not until their third year that their IQs fall below that of Europeans and only in their fourth year that their IQ declines to reach their IQ of approximately 85, as shown in rows 22 and 28.

Third, the IQ of approximately 85 of African Americans is substantially higher than the average IQ of 67 of Africans in sub-Saharan Africa. Two factors can explain this difference. The first is that American Africans enjoy a better environment than Africans in Africa in a number of respects, includ­ing much higher living standards and better nutrition and health. The second is that African Americans have on average about 25 percent of European ancestry and this increases their IQs above that of Africans in Africa (Reed, 1969; Chakraborty, Kamboh, Nwanko, and Ferrell, 1992).

Fourth, in the five studies giving verbal and visualization IQs, American Africans score one or two points higher on the verbal IQs. The verbal IQs appear to be more culturally biased, so this tells against the theory often proposed by environmentalists that Africans perform poorly because the tests are biased against them. This confirms the conclusions reached by McGurk (1953a, 1953b) that African Americans are not more impaired on what were considered culturally biased general information problems and by Jensen (1980) that tests are not biased against African Americans.

Fifth, there appears to have been no improvement in the IQs of African Americans over the course of the twentieth century. Thus, the median IQ of the fourteen studies carried out from 1980 to 1998 is 85, the same as that of the earlier studies. This conclusion is confirmed by the absence of any tendency for the African-American-European difference to be smaller in younger age groups. African-European IQ differences at different ages have been reported by Reynolds, Chastain, Kaufman, and McLean (1987) for the WAIS-R standardization sample collected in 1978. The African IQs are 86 in 16-19-year-olds, 85 for 20-34 and 35-54-year-olds, and 86 for 55-74-year-olds. It has also been shown in bi-yearly data that there has been no difference in African-American-European intelligence over the period 1974-1996 (Lynn, 1998e). Finally, in the standardization sample of the KAIT (Kaufman Adolescent and Adult Intelligence Test, Kaufman et al., 1994) there was no significant difference between the youngest and oldest age groups. In fact the youngest age group, born between 1980 and 1991, had a slightly lower IQ of 83 compared with an IQ of 88 of the oldest age group, born on average in 1921.

5. Africans in Britain

Africans began to migrate to Britain in substantial numbers shortly after the end of World War II. The first immigrants came mainly from the Caribbean and in the last quarter of the twentieth century a number came from Africa. From the 1960s studies were published of the IQs of African immigrants. The results of these are given in Table 4.5. Row 1 gives an IQ of 88 for what is believed to be the first published result of the children of West Indian Africans and is for a sample of Caribbean children in London, where the majority of these immigrants settled. Row 2 gives an IQ of 82 calculated by Vernon (1969, p. 169) for another sample in London in the 1960s. Row 3 gives a reasoning IQ of 88 and a vocabulary IQ of 82 for West Indian children compared with European English children attending the same secondary school in the district of Haringey in London; the district is poor and the European children will have scored below the national average, thereby inflating the IQs of the West Indians. To adjust for this the IQ of the Europeans is assumed to be 95. Row 4 gives an IQ of 89 for a sample of children in London. Row 5 gives an IQ of 86 for samples of children in Birmingham and in Deptford, London.

Row 6 gives an IQ of 104 for 9 African children taken into institutions as infants because their mothers were unable to look after them. In the same study the IQs of mixed race children and white children also taken into institutions were measured, with the results that the mixed race had an IQ of 110 (n=15) and the whites an IQ of 104 (n=36). The results are out of line with the other results in the table, all of which show African children in Britain have IQs well below whites. Moreover, it would normally be expected that the IQs of the children would be below average intelligence because the mothers were predominantly unskilled and put them into institutions, and would probably have been of below average intelligence. The results that these children had IQs above average are remarkable and need replication. If they can be confirmed as valid, they suggest that black mothers do not provide such a good environment as the white foster parents who reared these children, but there is little evidence to support this inference. The number of children (9) was very small and possibly this is just a fluke result.

Row 7 gives an IQ of 86 for a national sample of Afro-Caribbean children in Britain. Rows 8 and 9 give IQs of 73 for children in Britain born in the Caribbean and of 82 for those bom in Britain. The IQ of 73 for those born in the Caribbean is closely similar to that of 71 of indigenous Caribbean children given in Table 4.2. Row 10 gives a verbal IQ of 86 for West Indian children tested with the English Picture Vocabulary Test. Row 11 gives an IQ of 85 for West Indian children at a comprehensive school in the town of Ilford in Essex; the IQ of 85 is lower than that of Indian subcontinent children in the same school who obtained an IQ of 91; this is the first of a number of studies in Britain finding that Caribbean immigrants have lower IQs than Indian immigrants from the sub-continent. Row 12 gives a vocabulary IQ of 78 for all West Indian children at maintained (public) schools in an education authority in the Midlands.

Row 13 gives an IQ of 86 derived from a reading test on a very large national sample of 12,530 15-year-olds. Row 14 gives an IQ of 85 obtained on a vocabulary test by West Indian children in the north of England compared with 851 Europeans attending the same schools. Row 15 gives an IQ of 86 obtained by West Indian children at school in a town in the Midlands; Indians from the Indian sub-continent attending the same schools obtained an IQ of 96, showing once again that South Asians in the same environment as Africans obtain higher IQs. Row 16 gives an IQ of 90 for a sample of West Indian children in London. Row 17 gives an IQ of 87 for a sample of West Indian 4-year-olds.

Row 18 gives an IQ of 89 for a national sample of Caribbean children drawn from the whole of Britain born in 1958 and who had been in Britain for more than 4 years; a further group of 39 who had been in Britain for fewer than 4 years obtained an IQ of 83, suggesting that residence in Britain raises the IQs of Caribbean children by around 6 IQ points. It has sometimes been suggested that many of the recent immigrant children from the Caribbean spoke a form of Creole West Indian English that made it hard for them to understand the teachers, but the fact that immigrant West Indians performed about the same on non-verbal reasoning tests as on verbal comprehension makes this unlikely. Row 19 gives an IQ of

Table 4.5. IQs of Africans in Britain

92 for a national British sample born in 1970; the high IQ of this sample may indicate that the IQ of Caribbean children has increased slightly, but the subsequent studies in the table show no improvement in the IQs of African children from the 1960s through the 1980s, so this may be a chance result.

Row 20 gives an IQ of 88 for a sample of African schoolchildren in schools in London, the majority of whom had been born in Britain. Row 21 gives an IQ of 92 for a sample of African schoolchildren in Cambridgeshire. The IQs are in relation to indigenous British children attending the same schools and these are likely to be below national norms because the British of higher socio-economic status tend not to send their children to schools where there are appreciable numbers of immigrants. The effect of this will be to inflate the IQs of the Africans. There are no national norms for the test so the IQ of the Africans in relation to British cannot be accurately calculated. Probably the IQ of the British in this study was about 95, and hence the IQ of the African sample in relation to British national norms will have been about 87 and therefore about the same as other samples of Africans in Britain. Row 22 gives an IQ of 89 for a sample of 65-75-year-old Africans in London obtained in 1996-98 compared with a national sample of 5,379 indigenous British.

The results of the studies of the intelligence of Africans in Britain raise three points of interest. First, the median IQ of the studies is 86 and is almost exactly the same as the average of 85 of Africans in the United States. These figures are substantially higher than the median IQ of 67 of Africans in sub-Saharan Africa and of 71 in the Caribbean, from where most Africans in Britain have come in the post-World War II decades. Second, the higher IQ of Africans in Britain is attributable to the better environment. This effect is shown in the study by Yule, Berger, Rutter, and Yule (1975) given in rows 7 and 8 showing IQs of 73 for those born in the West Indies and 82 for those born in Britain, suggesting that residence in Britain raises the IQs of Caribbean children by around 9 IQ points. This result is confirmed by the Mackintosh and Mascie-Taylor (1985) study shown in row 18, where the West Indian children from the Caribbean who had been in Britain for more than 4 years had an IQ of 89, while the IQ of a further group of 39 who had been in Britain for fewer than 4 years obtained an IQ of 83, suggesting that residence in Britain raises the IQs of Caribbean children by around 6 IQ points. The two results are quite similar, suggesting that being raised in Britain increases the IQs of Africans by 7-8 IQ points. This increase is probably largely a result of better nutrition and health care and perhaps education, although there seems to be no evidence that education in the West Indies is poorer than in Britain and is sometimes asserted to be better. The effect of improved nutrition for West Indian immigrants was shown by the Yule et al. (1975) study that found that West Indian Africans born in Britain are taller than those born in the Caribbean who had come to Britain some time during childhood, a difference of 0.67d (standard deviation units). Third, the IQ of 87 for a sample of West Indian 4-year-olds given in row 17 is virtually exactly the same as that obtained by older West Indian children at school and shows that the low IQs of West Indian children cannot be blamed on schools, prejudice of teachers, difficulties understanding teachers' spoken English, and so on. This result confirms those found in the United States, that the low IQ of Africans is present in pre-school children.

6. Africans in the Netherlands

During the second half of the twentieth century a number of Africans migrated to the Netherlands from the former Dutch colony of Surinam in the northeast of South America and from the Netherlands Antilles, the former Dutch colony in the Caribbean. Studies of their intelligence are summarized in Table 4.6. Row 1 gives an IQ of 86 for the children of immigrants from Surinam. The test used and the age of the sample are not given. The population of Surinam consists of 35 percent Creoles of mixed African-European ancestry, 10 percent Africans, 33 percent Asian Indian, 16 percent Indonesian, and 3 percent American Indian. The IQ of 86 is about what would be predicted from this racially mixed population because the largest group, the Creoles, would be expected to have an IQ about midway between Africans in Africa (67) and Northwest Europeans (100), and the second largest group, the Asian Indians, should have an IQ of approximately 82 (see Chapter 6). Row 2 gives an IQ of 84 for a sample of the children of first generation immigrants from Surinam and the Netherlands Antilles. Row 3 gives an IQ of 88 for a sample of the children of second-generation immigrants from Surinam and the Netherlands Antilles. These children have a four-IQ-point gain compared with the children of first generation immigrants shown in row 2. This confirms the studies in Britain showing that second generation immigrants obtain higher IQs than first generation. Row 4 gives an IQ of 85 for a further sample of the children of immigrants from Surinam. The test used and the age of the sample are not given. Row 5 gives an IQ of 83 for another sample of immigrants from Surinam and the Netherlands Antilles. Row 6 gives an IQ of 85 for adult immigrants from Surinam. Row 7 gives an IQ of 85 for immigrants from the Dutch Antilles, whose population is 85 percent African and mixed African-European. The IQs obtained in the studies are closely similar, with a median of 85, the same as that of Africans in the United States.

Table 4.6. IQs of Africans in the Netherlands

7. Africans in Israel

A number of African Ethiopian Jews migrated to Israel during the closing decades of the twentieth century. The intelligence of a sample of 250 15-year-olds was assessed by Kaniel and Fisherman (1991) using the Standard Progressive Matrices. The authors write:

The Ethiopian Jews were tested one year after they arrived in Israel. All the subjects were tested in groups in their schools, using standard procedure. Each group was shown the first practice item of the test and solved it together. Special care was taken to make sure the Ethiopian Jews understood how the test was organized, to ensure their ability to fill out the answer sheet. There was no time limit (p. 28).

The authors made errors in the calculation of the IQ of the Ethiopians. I have given the correct figures in Lynn (1994b). The mean score on the test was 27, equivalent to the first percentile on the British 1979 standardization norms and to an IQ of 65. It is assumed that the Israeli data were collected in 1989 and that the British IQ increased by 2 IQ points between 1979 and 1989. To adjust for this increase, the IQ of the Ethiopian Jews needs to be reduced to 63.

A second study of the IQ of Ethiopian Jews has been published by Kozulin (1998). These were 14-16-year-olds who had been in Israel four or more years, were attending Israeli boarding schools, and were tested with the Progressive Matrices. Their mean IQ was 65. These results suggest that education in western schools does not benefit the African IQ.

8. Short-Term Memory and Perceptual Speed Abilities of Africans

Hitherto African intelligence has been considered in terms of g (gen­eral intelligence), reasoning, verbal, and visualization abilities. We now consider studies on the short-term memory and perceptual speed ability of Africans. Short-term memory is typically measured by the Digit Span test, consisting of the ability to recall a series of numbers either in the order in which they are presented (forward Digit Span) or in reverse order (backward Digit Span). Perceptual Speed is typically measured by the Coding and Digit Symbol subtests in the Wechsler tests that require accurate and rapid scanning of visual information. These studies have shown that Africans have relatively strong short-term memory and perceptual speed abilities. The results are summarized in Table 4.7. Row 1 gives IQs of 75 and 76 for African 10-12-year-olds (n=l,123) compared with Europeans (n=l,489) obtained for non-verbal reasoning and for verbal ability measured by the Lorge-Thorndike test, a much higher IQ of 90 for short-term memory measured by Digit Span and a remarkable IQ of 102 for Perceptual Speed. The authors comment: "given a test that involves only speed but no appreciable cognitive factor, the Negro children perform as well as or better than the European children" (Jensen & Rohwer, 1970, p. 60). Row 2 gives a typical IQ of 85 for the verbal and performance scales of the WISC-R obtained for 622 African 5-11-year-olds compared with 669 Europeans and a short-term memory IQ of 94 as the average of forward (IQ 96) and backward (IQ 92) digit span. Row 3 gives IQs for African 12-18- year-olds ob tained from the Project Talent data set and shows a relatively high IQ of 94 for immediate memory as compared with 89 for abstract reason­ing. Row 4 gives an IQ of 94 for short-term memory for 5-9-year-old African Surinamese immigrants in the Netherlands (n=183) compared with European children; the test consisted of the presentation of ten drawings, each of which was given an arbitrary name, and the task was to remember as many of the names as possible. Row 5 gives IQs for African 6-16-year-olds (n=711) compared with Europeans (n=776) of 82 for verbal ability, 78 for visualization ability, and a higher IQ of 90 for short term memory measured by Digit Span. Row 6 gives an IQ of 94 for short-term memory for Africans obtained from a meta-analysis of 31 studies of children and adults. Row 7 gives IQs for a number of primary abilities from South Africa from the study of 1,093 African and 1,056 European 16-year-olds tested with the Junior Aptitude Test, a test constructed in South Africa that provides measures of Abstract Reasoning (AR), Verbal Reasoning (VR), Verbal Comprehension (Verb), Visualization (Vis), Short-Term Memory (STM), and Perceptual Speed (PS); the sample also obtained a Mechanical Ability IQ of 68. It will be seen that the African short-term memory IQ (79) and the perceptual speed IQ (69) are both higher than their Abstract Reasoning Ability (58) and their Verbal Reasoning Ability (63), confirming the American studies. In this sample the visualization and mechanical abilities are also all stronger than abstract and verbal reasoning ability. Row 8 gives a short-term memory IQ of 74 for a sample of 196 10-year-olds in Jamaica, compared with 67 entered as the median of the seven studies given in Table 4.3.

Jensen (1998) has interpreted these and other results as showing that the African-European differences in intelligence are largely differences in Spearman's g. According to this theory, short-term memory and perceptual speed are weak measures of g, so Africans do relatively well on them. The theory has received considerable support, summarized by Rushton (2003), but has also attracted some criticism from Dolan and Hamaker (2001).

Table 4.7. Primary abilities of Africans

9. Musical Abilities

It has often been considered that Africans have good musical abilities and a particularly strong sense of rhythm. The conclusion appears to have been first suggested in the fourteenth century by the Arab writer Ibn Butlan who wrote that if an African "were to fall from heaven to earth he would beat time as he goes down" (Lewis, 1990, p. 94). Musical abilities are associated with intelligence, so it is interesting to consider whether Africans have the good musical abilities often attributed to them, or poor musical abilities consistent with the low IQs they obtain on intelligence tests.

Musical abilities are measured by simple tests such as the identification of pitch change (identifying whether the pitch of one tone is higher or lower than that of another; in the initial items the difference between the tones is great but as the test progresses the tones become closer until it is extremely difficult to distinguish which is higher); memory (a tune is played twice and on the second playing one note is altered; the task is to identify the altered note); chord analysis (the identification of the number of notes in a chord); and rhythm (two pieces of music are played and the problem is to identify whether the rhythms are the same or different). The association between intelligence and musical ability has been shown in two studies carried out by Lynn, Wilson, and Gault (1986). In the first, a sample of 217 10-year-olds were given a number of tests of reasoning, vocabulary, visualization, and perceptual speed abilities together with four musical ability tests (pitch, memory, chords, and rhythm). All the tests were positively intercorrelated and loaded on the first principal component as a measure of general intelligence (g). The loadings of the four musical tests lay between 0.45 (cords) and 0.59 (rhythm). This shows that the musical tests are measures of g. In the second study 93 9-11-year-olds were given three tests of musical ability (pitch change, chord analysis, and memory) together with the Standard Progressive Matrices, a measure of g. The three musical tests were significantly correlated with the Progressive Matrices at 0.27, 0.40, and 0.37. This confirms that musical ability is associated with intelligence. Further evidence for this associa­tion has been provided by Carroll (1993).

There has been some work on the musical ability of African Americans but this is little known because it has not been summarized in general textbooks on intelligence such as those of Brody (1992) and Mackintosh (1998) or in specialist textbooks on race differences in intelligence such as those by Loehlin, Lindzey, and Spuhler (1975) and Jensen (1980, 1998). The general outcome of these studies is that African Americans perform less well than Europeans on tests of musical abilities of pitch discrimination, tone discrimination, and memory, but they perform about the same as Europeans on tests of rhythm. To show this pattern of musical abilities, the results of these studies have been aggregated to give a Musical Quotient (MQ) derived from tests of musical ability other than rhythm, and a Rhythm Quotient (RQ). The results of these studies are summarized in Table 4.8. Row 1 gives results for a large sample of African Americans in North Carolina, South Carolina, and Virginia calculated from the Seashore Test and shows that they obtained an MQ (Musical Quotient) of 90 but a higher RQ (Rhythm Quotient) of 106. Row 2 gives results from a comparison of 272 European and 288 African American college students attending colleges in Tennessee, again showing that Europeans achieved higher scores on general musical ability (pitch, intensity, time, consonance, and total memory) but African Americans achieved a higher RQ (Rhythm Quotient) of 102. Row 3 gives results for a sample from a poor neighborhood in Washington, D.C. showing an MQ of 83 and an RQ of 96. Row 4 gives results for a sample of African-American 5-to-8-year-olds in Rochester in New York State with an MQ of 89 and an RQ of 104. Row 5 gives results for a sample of African-American 18-year-olds drawn from senior high school largely in Texas, with some from Illinois and Rochester in New York State, with an MQ of 86 based on tonal memory and pitch discrimination and an RQ of 100. The comparison group was 541 Europeans attending the same schools. All the studies show that African Americans have Rhythm IQs substantially greater than general Musical IQs by about 15 IQ points. There appears to be no change in the musical abilities of Africans over the period of approximately half a century from the 1920s to the late 1970s over which the studies have been conducted. The relatively high rhythm ability of Africans is expressed in their music in which a strong rhythmic element is frequently present. This is notably the case in the hymns sung by congregations in African and African-American churches. It also appears in jazz, which was first developed by African Americans in New Orleans in the early years of the twentieth century, and in its subsequent development in "swing," with its strong syncopated rhythms.

Several twin studies have shown that there is a genetic basis for musical abilities. For instance, a study by Vandenberg (1962) of the heritability of rhythm ability obtained from the correlations of 33 pairs of identical twins and 43 pairs of same-sex fraternal twins calculated a heritability of 0.52,

Table 4.8. Musical (MQ) and Rhythm (RQ) Quotients of African Americans

not corrected for measurement error. Heritabilities of this magnitude make it likely that the low general musical abilities and the high rhythm ability of Africans have some genetic basis.

The low musical abilities of Africans, except for their strong sense of rhythm, are consistent with their generally poor achievements in classical music. There are no African composers, conductors, or instrumentalists of the first rank and it is rare to see African players in the leading symphony orchestras.

10. Reaction Times

Reaction times consist of the speed of reaction to a simple stimulus such as the onset of a light. The task is to press a button when this occurs and the reaction time is the time taken to respond, which typically takes about a third of a second. Numerous studies reviewed by Jensen (1998) and Deary (2000) have shown that reaction times are positively related to intelligence at a magnitude of around 0.2 to 0.3. It has been persuasively argued by Jensen (1998) that reaction times are a measure, of the neurological efficiency of the brain in processing information. This makes it an interesting question whether the differences between Europeans and Africans in intelligence are also present in reaction times. If they are, it means that there are race differences in the efficiency of the brain. If there are not, it means that there are no race differences in the efficiency of the brain and the differences in intelligence must be due to some other factors, such as opportunities for learning the problems in the tests, educational experiences, or test bias.

The most complete studies of African-European differences in reaction times have been carried out by Jensen (1993) in the United States and Lynn and Holmshaw (1990) and Sonke (2000) in South Africa. Jensen's study compared 585 European and 235 African 10-year-olds whose IQs assessed by the Progressive Matrices differed by 11 IQ points. The Lynn and Holmshaw study compared 350 African and 239 British 9-year-olds whose IQs differed by 37 IQ points. Both studies used the same computer-controlled apparatus, so that no human error can affect the times registered. Both studies measured the twelve components of reaction times. Three different kinds of reaction time were measured. These were simple reaction time (SRT) consisting of reactions to a single light, choice reaction time (CRT) consisting of responses to one of eight lights, and odd-man reaction time (OMRT) consisting of reaction to the one of three lights that was farthest from the other two. Each of these three reaction times was measured for four components consisting of the reaction time proper (the decision time), the movement time (time taken to move the finger to the button), and the standard deviations of the reaction times and movement times. The results are shown in Table 4.9. Column 1 gives the different measures of reaction time. Columns 2 and 3 give the Jensen data for the correlation with the Progressive Matrices IQ and the d difference between Africans and Europeans with negative signs denoting faster times by Europeans. Columns 4 and 5 give the same data for the Lynn and Holmshaw data. Correlations between reaction times and IQs are consistently positive in all the data and 16 of the 24 correlations are statistically significant (designated by the asterisks), but the correlations are very low. Reaction times shown in rows 1, 5, and 9 are faster in Europeans than Africans except for CRT in the Jensen data. Simple movement times show no difference, but Africans are significantly faster than Europeans in both CMT and OMMT in the Lynn and Holmshaw data. The faster movement times of Africans may be a factor in the fast sprinting speed of Africans shown in Olympic records. The standard deviations are consistently greater in Africans in the Lynn and Holmshaw data and in four of the six differences in the Jensen data. In general the African-European differences are much greater and more consistent in the Lynn and Holmshaw data than in the Jensen data. This would be expected because the intelligence difference is some four times greater in the Lynn and Holmshaw data. However, in the Lynn and Holmshaw data the mean of differences of the six reaction times and standard deviations between the Africans and Europeans amounts to only 0.67J as compared with a 2.5d difference in IQ. The best interpretation of the results is that approximately a quarter of the African-European difference in intelligence may be explicable by the speed of neurological processing, while the remainder must be attributed to other processes.

Although reaction times have a significant heritability of around 50 percent (Deary, 2000), they are also affected by nutrition. An Italian study found that children aged 6-10 in iodine deficient villages had slower reaction times as well as lower IQs (Vitti et al., 1992). Slow reaction times in children from iodine deficient villages have also been reported by Bleichrodt et al. (1987).

Table 4.9. Correlations between reaction times and IQ and differences between Africans and Europeans in reaction times.

Sonke (2000) has reported a study of three groups consisting of 26 illiterate Africans in South Africa aged 16, with "only a few years of schooling," 29 African university students at Venda University in the Northern Transvaal, and 30 European Dutch university students at Tilberg university. The three groups were given an intelligence test (Raven's Progressive Matrices) and simple and complex reaction time tasks, and an EEG measure was taken of the latency of the evoked potential (P3) to the presentation of the reaction time stimuli, a measure of the speed with which the stimulus is registered in the brain. There were equal numbers of males and females in all three groups.

The results are shown in Table 4.10. Row 1 gives the IQs of the three groups. Row 2 gives the mean simple reaction times showing slowest reaction times in the African illiterates and fastest in the European students. Row 3 gives complex reaction times showing the same group differences. Row 4 gives the evoked potential latencies for task Bl, showing longest latencies in the African illiterates, and the shortest latencies in the European university students. All the group differences are statistically significant.

Table 4.10. Reaction times and EEGs of Africans and Europeans

There are six points of interest in this study. First, the South African illiterate sample's Progressive Matrices IQ of 68 is closely similar to that of a large number of samples of Africans in South Africa and in other countries of sub-Saharan Africa. Second, the African university students have a somewhat higher IQ of 82, again similar to that of other African South African university student samples. Third, there are significant differences between the three groups in reaction times, confirming other studies summarized in this chapter. Fourth, there are significant African-European differences in the EEG evoked potential showing that in European students the brain reacts more rapidly to a stimulus than in African students. Fifth, there is a statistically significant correlation of 0.213 between the complex reaction times and the Progressive Matrices, confirming many other studies of this association. Six, the correlation between the Progressive Matrices and the EEG evoked potential is not statistically significant. The differences between the African illiterates and the African students on reaction times and evoked potentials are probably attributable to the students having higher IQs.

11. Brain Size of Africans and Europeans

Studies showing that Africans have smaller average brain size than Europeans are summarized in Table 4.11. The figures given in the table are in cubic centimeters (these figures have been converted from cubic inches given by Morton and Gould, and from grams given by Ho et al., 1980). It should be noted that estimates of cranial capacities are to some degree affected by the method of measurement. The cranial volume of skulls is measured by filling them with lead shot or mustard seed and measuring the volume of the shot or seed. Lead shot gives slightly larger volumes than mustard seed because it cannot be compressed so tightly. For living humans brain size is calculated from the length, breadth, and height of the head, or from the circumference. These different methods of measurement explain some of the differences obtained in the studies. Despite different methods of measurement, there is considerable consistency in the various studies.

In Table 4.11 the results are given of eight studies of the brain size of samples of Europeans and Africans and the difference between the two means. All the studies show that Europeans have a larger average brain size than Africans. Row 1 gives the results of the first study showing that there are African-European differences in brain size published in the nineteenth century by the American physician Samuel Morton (1849), who assembled a collection of skulls, categorized them by race, and calculated their average cranial capacities. His results were criticized by Gould (1996), who accused him of massaging the figures to demonstrate that Europeans have larger brains than Africans. Gould recalculated Morton's skull sizes and his results were closely similar. It is Gould's figures that are given in the Table. He dismissed the 41cc difference as of no consequence. Gould characteristically failed to mention any of the other studies that all confirmed Morton's conclusions and found larger differences. Notice that the numbers of skulls in Morton's collection are quite low, consisting of 52 Europeans and 29 Africans, as compared with the other studies.

Row 2 gives the results of an analysis of a much larger collection of skulls held at Western Reserve University in Ohio, showing 50cc African-European differences in brain size. Row 3 gives results presented by Tobias, a committed equalitarian, who asserted that there is no race difference in brain size but whose results actually show a rather larger African-European brain size difference than those of Morton. Row 4 gives the results from autopsies in the United States showing a larger African-European difference, of 103cc, than in the other studies. Row 5 gives results from the largest collection of approximately 20,000 skulls from all over the world analyzed by the American anthropologist Kenneth Beals. Row 6 gives results calculated by Groves (1991) by combining estimates of cranial capacities of 36 samples of males from figures given by Coon, Molnar, and Martin and Sailer and again showing that Europeans have larger average brain size than Africans. Row 7 gives results for the United States for military personnel. These figures are adjusted for height and weight. The brain sizes of the Europeans are virtually identical to those found by Ho et al. given in row 4, but the brain size of the Africans is much greater, at 1,359 as compared with 1,267. The explanation for this is that the U.S. military screens applicants for intelligence and rejects those with IQs below 81 (Nyborg and Jensen, 2000). Flynn (1980) has estimated that military rejection rates for low IQ are 3.4 percent for Europeans and 30 percent for Africans, and that the result of this is that Africans in the military have an average IQ of 91.5. The effect of not accepting Africans with low IQs is to screen out many of those with low intelligence and small brains, making the African-European brain size difference much smaller than in other samples. Row 8 gives average brain size of six samples of male Europeans from North America and Europe and two samples from sub-Saharan Africa from data compiled by Jurgens, Aune, and Pieper (1990) and analyzed by Rushton (2000, p.124) showing a European advantage of 109cc. The results in the eight data sets all show greater brain size of Europeans than of Africans and are reasonably consistent considering that they were compiled using different methods and different kinds of samples, including autopsies (Ho et al., 1980), skull volumes (Reals et al., 1984), and external head measurements of living individuals.

These results are corroborated by a further large-scale study of children carried out by Broman, Nichols, Shaughnessy, and Kennedy (1987). They examined and followed up approximately 17,000 European and 19,000 African children in the United States from conception to the age of 7 years. At the age of 7 there was the typical gap of approximately 15 IQ points between the two groups. The head circumference of the two groups calculated from the published data are 50.9cm (sd 1.6) for Africans and 51.7cm (sd 1.6) for Europeans. This difference is statistically highly significant and provides an approximate measure of differences in brain size, since head circumference and brain size are correlated at about 0.8 (Brandt, 1978). The brain volumes have been estimated by Rushton (1997) at 1,134 for Africans and 1,150 for Europeans. The difference is much smaller than in the other samples, possibly because Europeans mature later than Africans. In this study the African children were slightly taller than the Europeans, suggesting that possible differences in nutrition are not likely to be responsible for the differences in head size.

Table 4.11. Brain size (cc) of Europeans and Africans

12. IQs of African-European Hybrids

We now consider studies of the IQs of African-European hybrids. The prediction from the genetic theory of race differences is that the IQs of racial hybrids should fall approximately midway between those of Europeans and Africans. To examine this prediction, studies of African-European racial hybrids are summarized in Table 4.12. Row 1 gives results for Brazil showing that hybrids known as "browns" score intermediate between Europeans and Africans. Row-2 gives results from Germany from the Eyferth (1961) study showing the IQ of African-European hybrid children was 94 in relation to 100 for European children. The mean IQ of the African-European hybrids was 96.5 but is reduced in the table to 94 to allow for the secular increase of the IQ from the date of the standardization. Row 3 gives results from South Africa for Europeans, Africans, and Coloreds, who are largely African-European, and shows that the IQ of 83 of the Coloreds falls exactly half way between that of Europeans (100) and that of Africans (65). Row 4 gives results from a more recent study in South Africa collected approximately sixty years later and showing a sample of Coloreds with an IQ of 86 compared with an IQ of 100 for Europeans. Africans were not included in this study but the IQ of 86 is much higher than that of pure Africans in South Africa. Row 5 gives results from a further South African study showing an IQ of 80 for Coloreds.

Rows 6 through 9 give four results for hybrids in the United States.

Table 4.12. IQs of Europeans, African-European hybrids, and Africans

Row 6 gives results from the Minnesota Transracial Adoption Study showing that hybrids score halfway between African Americans and Europeans. The numbers are very low, but the results are informative because all three groups were reared by European adoptive parents and this rules out any reasonable environmental interpretation of the differences. Row 7 gives a further result from the United States showing once again that hybrids score intermediate between Europeans and African Americans (results of this study are given by Loehlin, Lindzey, and Spuhler, 1975). Row 8 gives another result from the United States that divided African Americans into dark skinned and lighter skinned and showed that the lighter skinned African Americans, taken as an index of hybridization with Europeans, have an IQ of 92, halfway between the Europeans and Africans. Row 9 gives the last result, showing that African-American-European hybrids have an IQ of 97 and again score intermediate between African Americans and Europeans.

13. Heritability of Intelligence in Africans Americans

There have been three studies of the heritability of African Americans in the United States. They are all obtained from a comparison of identical (Mz) and non-identical twins (Dz). The results are given in Table 4.13.

This gives the age of the sample, the numbers of identical (Mz) and non-identical (Dz) twins, the correlations between the twin pairs, the heritability obtained by doubling the difference between the Mz and Dz correlations, and the corrected heritability, corrected for attenuation assuming a test reliability of 0.9. Row 1 gives data from Loehlin, Lindzey, and Spuhler (1975) obtained from a doctoral dissertation by P. L. Nichols for 4-year-olds tested with the Stanford-Binet. This shows a corrected heritability of 0.56, a little higher than that of Europeans of this age. Row 2 gives results of a further data set from Osborne (1980, p. 72) for general intelligence calculated as the average of twelve tests and showing a corrected heritability of 1.00. Row 3 shows data for the Progressive Matrices (Scarr, 1981, p. 282) giving a corrected heritability of 0.60. Taken together the three results show a heritability of 0.72 in African Americans and higher heritability in the two studies of adolescents than in the 4-year-olds. The heritabilities of the African Americans are virtually the same as those in Europeans given in Chapter 3.

14.  Genetic and Environmental Explanations of the Low African IQ

The problem of the genetic and environmental contributions to the low IQ of Africans has been debated since the early decades of the twentieth century, particularly in regard to the problem of the low IQs obtained by African Americans in the United States. Many hundreds of papers and a number of books have been devoted to this problem and it is not possible to deal with it adequately. Three positions have been taken on this question:

1. The IQ difference between blacks and whites is wholly environmentally determined or at least there is no compelling evidence for any genetic contribution to the low black IQ. This position has been taken by Flynn (1980), Mackintosh (1998), Nisbett (1998), Fish (2002), Brody (2003), and many others.

2. The IQ difference is determined by some mix of genetic and environmental factors. This position has been taken by Loehlin, Lindzey, and Spuhler (1975), Vernon (1979), and Waldman, Weinberg, and Scarr (1994, p. 31), who conducted one of the most important studies of this question involving the IQs of black children adopted by white couples.

3. The IQ difference is largely genetically determined. This position has been taken by Garrett (1945, 1961); McGurk (1953a, 1953b), who showed that when blacks and whites were matched for socio-economic status, blacks scored 7.5 IQ points below whites; Kuttner (1962), who argued that black-white differences in intelligence were reflected in the differences in the building of early civilizations; Shuey (1966), who made the first compilation of black-white IQ differences, from 1916 up to 1965; Osborne and McGurk (1982), who made an updated compilation of Shuey's work covering the years 1966-1980; and Jensen (1969, 1974, 1980, 1998), who has made numerous contributions to this issue and concluded that about two thirds of the American black-white IQ difference is attributable to genetic factors. Others who have taken the largely genetic position are Shockley (1969), Eysenck (1971), Baker (1974), Levin (1997), Rushton (2003), and the writer (Lynn, 1994c, 2001).

There are six major arguments for the presence of some genetic determination of the intelligence difference between Africans and Europeans.

First, the two races have evolved independently in different environments over a period of approximately 100,000 years (Mellars and Stringer, 1989; Cavalli-Sforza, 2000). When two populations evolve largely in isolation from each other for this period of time genetic differences between them

Table 4.13. Heritability of intelligence of African Americans

inevitably evolve for all characteristics for which there is genetic variability. These differences evolve as a result of genetic drift, mutations, founder effects, and most important, adaptation to different environments. The extreme environmentalist position that there is no genetic difference between the two races for intelligence defies this general principle of evolutionary biology and should be ruled out as impossible.

Second, the consistency with which Africans obtain low IQs in so many different locations can only be explained by the operation of a strong genetic factor. If only environmental factors were responsible for the different IQs of different populations, we should expect to find some countries where Africans had higher IQs than Europeans. The failure to find a single country where this is the case points to the presence of a strong genetic factor.

Third, the high heritability of intelligence found in twin studies of blacks and whites in the United States, in Europe, Japan, and India shows that intelligence is powerfully affected by genetic factors and makes it improbable that the differences between Africans and Europeans, or between any other pairs of races, can be solely environmentally determined.

Fourth, the brain size difference between blacks and whites points to a genetic difference, considering the high heritability of about 0.9 of brain size and the correlation of approximately 0.4 between brain size and intelligence.

Fifth, several egalitarians have proposed that white racism may be responsible for impairing the IQs of the blacks. Thus, Weinberg, Scarr, and Waldman write that their result that black children adopted by whites have low IQs "could indicate the results of environmental influences such as the pervasive effect of racism in American life" (1992, p. 41) and "the IQ results are consistent with racially based environmental effects in the order of group means" (p. 40). Mackintosh (1998, p. 152) also falls back on white racism in a final attempt to argue that the low IQ of the black adoptees can be explained environmentally and suggests that perhaps "it is precisely the experience of being black in a society permeated by white racism that is responsible for lowering black children's IQ scores." These egalitarians do not explain how hypothetical white racism could impair the IQs of black children reared by middle class white parents. There is no known or plausible mechanism by which supposed white racism could impair the IQs of blacks. Nor do they attempt to explain how it is that Africans throughout sub-Saharan Africa, who are not exposed to white racism, except in South Africa, have IQs of approximately 67.

Furthermore, if racism lowers intelligence, it is remarkable that Jews in the United States and Britain should have IQs of around 108, as shown in Lynn (2004), since Jews have been exposed to some degree of racism for many centuries. The high IQ of American Jews has been well known since the 1930s and has been extensively documented by Storfer (1990), MacDonald (1994), and Herrnstein and Murray (1994), yet it goes curiously unmentioned by environmentalists like Flynn (1980), Brody (1992, 2003), Neisser (1996), Mackintosh (1998), Jencks and Phillips (1998), Nisbett (1998), Montagu (1999), and Fish (2002).

Sixth, the Minnesota Transracial Adoption Study carried out by Waldman, Scarr, and Weinberg (1994) was designed to show that when black infants are adopted by white parents they would have the same IQs as whites. The authors of this study examined groups of black, white, and interracial babies all adopted by white middle class couples. In the event it turned out that at the age of 17 the IQs were 89 for the blacks, 98 for the interracial, and 106 for the white. Thus, a 17 IQ point difference between blacks and whites remains when they are reared in the same conditions. Being raised by white adoptive parents had no beneficial ef fects on the intelligence of the black children because their IQ of 89 is thesame as that of blacks in the north central states from which the infants came. The interracial group with its IQ of 98 falls midway between the black and the white, as would be predicted from the genetic cause of the difference. A full analysis and discussion of this study has been given by Levin (1994) and Lynn (1994c), together with an unconvincing reply by Waldman, Weinberg, and Scarr (1994, p. 43) in which they assert "we feel that the balance of the evidence, although not conclusive, favors a predominantly environmental etiology underlying racial differences in intelligence and that the burden of proof is on researchers who argue for the predominance of genetic racial differences". Notice that their use of the term ''predominantly environmental etiology" concedes that accept that genetic factors are also present. While the results of this study show that differences in family environment cannot explain the low black IQ, it remains possible that blacks provide an inferior prenatal environment as a result of poorer nutrition of pregnant black women or possibly of the greater use of cigarettes that might impair the growth of the fetal brain. These possibilities are rendered improbable by studies showing that the nutrition of American blacks throughout the twentieth century was not inferior to that of whites (see Chapter 13, Section 7). Another possibility is that black babies might suffer greater impairment of the brain because pregnant black women might smoke cigarettes more, since there is some evidence that smoking retards fetal growth, but this is rendered improbable by numerous studies showing that blacks smoke cigarettes less than whites.

Despite their commitment to the egalitarian position, it is interesting to note that Waldman, Scarr, and Weinberg conclude that their evidence shows that both genetic and environmental differences contribute to the black-white IQ difference: "We think it is exceedingly implausible that these differences are either entirely genetically based or entirely environmentally based" (p. 31). Thus, while there is nothing in their data that can justify this conclusion, because they provide no evidence for any environmental contribution to the low black IQ, their final position is not greatly different from that advanced by Jensen (1969), that both genetic and environmental factors are responsible for the low black IQ, but where Jensen proposed that the relative contributions are about two thirds genetic and one third environmental, Waldman, Scarr, and Weinberg have concluded that both factors are involved, although they do suggest a quantification of the magnitude of the respective contributions.

In fact, the results of the Minnesota Interracial Adoption Study show that both conclusions are incorrect. The conclusion to be drawn from this study is that rearing black children in a white middle class environment has no effect at all on their IQs at age 17.

15. Estimation of the Genotypic African IQ

The IQs of approximately 67 of the African populations of sub-Saharan Africa shown in Table 4.1 are a function of both genetic and environmental factors. We now undertake the task of estimating the genotypic African IQ. This is the IQ that Africans would have if they were raised in the same environment as Europeans. The starting point of this analysis is the Minnesota Transracial Adoption Study, the results of which are summarized in section 14 and which showed that a 17 IQ point difference between African Americans and Europeans is still present when they are reared in the same family environments. The conclusion to be drawn from this is that the African-American-European IQ difference in the United States is wholly genetically determined. Although this study showed a 17 IQ point African-European IQ difference, it is reasonable to assume that the true African-American-European difference is 15 IQ points, as shown by the numerous studies summarized in Table 4.4, and that the 17 IQ point difference obtained in this study is a sampling error. We conclude therefore that the genotypic IQ of African Americans is 15 IQ points below that of American Europeans. A further argument for believing that the IQ of African Americans is wholly genetically determined is that it has remained constant over a period of approximately 80 years despite the great improvements in the environment of African Americans relative to that of Europeans.

The conclusion that African Americans have a genotypic IQ of 85 does not mean that Africans in sub-Saharan Africa also have a genotypic IQ of 85. African Americans are not pure Africans but are a hybrid population with a significant amount of European ancestry. This has been estimated at 25 percent by Reed (1971) and by Chakraborty, Kamboh, Nwamko, and Ferrell (1992). We can estimate that pure Africans in Africa and in the United States have a genotypic IQ of 80 and that this IQ increases by 0.2 IQ points for every 1 percent of Caucasoid genes. Thus, the average African American will have an IQ of 85 (80 + 25 X 0.2 = 85), a figure confirmed by numerous studies summarized in Table 4.4. In the Southeastern states the percentage of European genes among African Americans is quite low. For instance in South Carolina it has been estimated at 6 percent (Workman, 1968) and in Georgia at 11 percent (Reed, 1969). These admixtures of European genes should raise their IQ by 1.2 and 2.2 IQ points, respectively, giving them an IQ of 81.2 and 82.2. This prediction has been confirmed by the study of 1,800 African Americans in five Southeastern states by Kennedy, Van der Reit, and White (1963), which found their IQ on the 1960 Stanford-Binet was 80.7.

African Americans with 50 percent European genes will have an IQ of 90 (80 + {50 by 0.2 = 10} = 90). This is about the mean IQ of African Americans in the Northern states, where the proportion of European ancestry approaches 50 percent. African Americans with 75 percent European genes will have an IQ 15 points higher at 95 (80 + {75 x 0.2 = 15} = 95), which is very close to the IQ of 94 of the interracial children in the Minnesota Transracial Adoption Study. Europeans with 100 percent European genes will have an IQ at 100.

This estimate of the genotypic African IQ as 80 means that the average IQ that Africans would obtain if the environments in which they were raised were the same as those of Europeans would be 80. Throughout sub-Saharan Africa the mean IQ of Africans is approximately 67, so it can be inferred that adverse environmental conditions in sub-Saharan Africa impair the African IQ by around 13 IQ points.

Chapter 5. Bushmen and Pygmies

1. Intelligence of Bushmen

2. Brain Size of Bushmen

3. Pygmies

The bushmen, also sometimes called Khoisans, Sanids, or Capoids, and the Pygmies are two of the minor races of sub-Saharan Africa in the taxonomies of classical anthropology such as that of Coon, Garn, and Birdsell (1950). Cavalli-Sforza, Menozzi, and Piazza (1994) in their genetic analysis of human populations have confirmed that these two peoples have distinctive but closely related genetic characteristics and form two related "clusters." The Bushmen together with the Pygmies and Africans evolved from the original Homo sapiens peoples of equatorial East Africa. The ancestors of the Bushmen migrated south and by about 100,000 years ago occupied most of southern Africa. Extensive human bones and artifacts have been found in the Border Cave in present day Swaziland and have been dated at about 100,000 years old. The morphology of the bones indicates that these peoples were a mix of Africans and Bushmen (Beaumont, deVilliers, and Vogel, 1978).

Until around 1,500 years ago the Bushmen occupied most of southern Africa and the Pygmies occupied the rain forests of west and central Africa. From about 500 AD Africans (Negroids) from the north began to encroach on their lands, killed large numbers of them, and drove most of the surviving Bushmen into the Kalahari desert and the Pygmies into the dense rain forests of central Africa. Related to the Bushmen are the Hottentots, small groups of whom survive in a few locations in southern Africa. Although the two groups are genetically closely similar there are some genetic differences, such as the low incidence of the B blood group in the Bushmen and the high incidence in the Hottentots.

Many of the Hottentots are racial hybrids with Bushmen and European ancestry, which has given them lighter skin color and taller stature than the Bushmen (Cole, 1965). The Bushmen survive principally in the Kalahari Desert, where they number about 50,000. There are about the same number of Hottentots. The largest surviving group is the Nama in Southwest Africa, where they are around 24,000 (Cole, 1965), and there are a few other smaller groups north of the Orange River.

The Bushmen have a number of physical characteristics that distinguish them from Negroid Africans. They have peppercorn hair that grows in spirals with open spaces between tufts, whereas most Africans have helical woolly hair that forms a tight mat. It is believed that the peppercorn hair of the Bushmen evolved as an adaptation in hot and damp forests in which they lived for many millennia because it affords protection from strong sunlight but at the same time the open spaces between the tufts allow sweat to evaporate. Pygmies who have remained in tropical rain forests have the same peppercorn hair. The mat woolly hair of Negroid Africans is a more advantageous adaptation in dry hot environments because it gives greater protection from strong sunlight and reduces sweating. The skin color of the Bushmen is yellowish brown, while that of Negroid Africans is black or dark. Some of the Bushmen have an epicanthic fold on the upper eyelid, similar to but less pronounced than that of East Asians and Arctic Peoples. The advantage of the epicanthic fold for Bushmen is probably that it reduces the dazzling effect of glare from strong sunlight reflected from the desert, as it does the glare from snow for the East Asians and Arctic Peoples. This characteristic must have arisen independently through convergent evolution.

A distinctive characteristic of Bushmen is the very large buttocks of the women, known as steatopygia. The adaptive advantage of these may have been to store food and water in times of famine and shortage. The genitalia of the Bushmen are unique among the human races. Bushmen have penises that stick out horizontally, while Bushwomen have prominent minor labia that descend about 3 inches below the vagina. The adaptive advantages of these characteristics are unknown.

1. Intelligence of Bushmen

There have been only three studies of the intelligence of the Bushmen. In the 1930s a sample of 25 of them were intelligence tested by Porteus (1937) with his maze test, which involves tracing the correct route with  pencil through a series of mazes of increasing difficulty. The test has norms for European children for each age, in relation to which the Bushmen obtained a mental age of seven and a half years, representing an IQ of approximately 48. In the second study, Porteus gave the Leiter International Performance Scale to 197 adult Bushmen and concluded that their mental age was approximately 10 years, giving them an IQ of 62. In the third study, Reuning (1972), a South African psychologist, tested 108 Bushmen and 159 Africans with a pattern completion test involving the selection of an item to complete a pattern. In the light of his experience of the test, Reuning concluded that it "can be used as a reliable instrument for the assessment of intelligence at the lower levels of cognitive development and among preliterate peoples" (1968, p. 469). On this test the Bushmen scored approximately 15 IQ points below the Africans, and since it is known that Africans have a mean IQ of approximately 67 (see Chapter 4), this would give the Bushmen an IQ of approximately 52.

Reuning also gave a figure drawing test involving the drawing of a man to the Bushmen and the Africans. This test is the same as the Goodenough Draw-a-Man test (DAM), which is a reasonably good measure of intelligence. The drawings produced in the Goodenough Test are scored for detail and sophistication, which improve as children grow older. Young children typically draw stick men with little detail, while older children draw full-bodied men with many details such as eyebrows, thumbs, and so on. Reuning (1968, p. 476) recorded that the Bushmen's drawings were significantly less advanced than those of Negroid mineworkers, whom he also tested and 76 percent of whom were illiterate. He described the Bushmen's drawings as characterized by "extreme simplicity...the majority were stick figures...no details of fingers, toes, hair, eyes, etc...." The simplicity of the Bushmen's drawings "contrasts with the tendency of the blacks to include much small detail in their drawings (buttons, hair, fingers, toes, a pipe, etc.)." The difference between Bushmen and Africans in the sophistication of drawings provides further corroboration of the lower intelligence of Bushmen.

Reuning noted that there was considerable variability in intelligence between individuals among the Bushmen, just as there is among other peoples, and that they themselves recognize that some individuals are intelligent while others are dull. Their languages have the word "clever" to j describe this attribute. He records that "When the tester at the end of a test ) had praised a good performance, they let us know, through the interpreter: : 'We could have told you so, he (or she) is clever' (1968, p. 479). There is, furthermore, a general factor (g) among the Kalahari Bushmen shown by the positive intercorrelation of a number of tests and the correlation between test performance and the general consensus of who is intelligent.

In addition to administering a test of intelligence, Reuning tested Bushmen and comparison groups of Africans and whites for size constancy. This is the ability to estimate the size of an object at a distance. He found that Bushmen had more accurate size constancy than Africans and Europeans and attributed this to the need for this ability for using the bow and arrow for hitting an animal at a distance. He found that Bushwomen also have good size constancy although they do not hunt. He suggests that the development of this ability is probably attributable to its advantage for efficient hunting. If this is correct, it implies that the ability may have deteriorated in European and East Asian peoples who gave up hunting about 8,000 years ago and adopted agriculture.

The three studies of Bushmen by Porteus and Reuning give IQs of 48, 62, and 52 and can be averaged to give an IQ of 54. It may be questioned whether a people with an average IQ of 54 could survive as hunter-gatherers in the Kalahari desert, and therefore whether this can be a valid estimate of their intelligence. An IQ of 54 is at the low end of the range of mild mental retardation in economically developed nations. This is less of a problem than might be thought. The great majority of the mildly mentally retarded in economically developed societies do not reside in hospitals or institutions but live normal lives in the community. Many of them have children and work either in the home or doing cognitively undemanding -jobs. An IQ of 54 represents the mental age of the average European 8-year-old child, and the average European 8-year-old can read, write, and do arithmetic and would have no difficulty in learning and performing the activities of gathering foods and hunting carried out by the San Bushmen. An average European 8-year-old can easily be taught to pick berries, put them in a container and carry them home, collect ostrich eggs and use the shells for storing water, and learn to use a bow and arrow and hit a target at some distance. Before the introduction of universal education for children throughout North America and Europe in the second half of the nineteenth century, the great majority of 8-year-old children worked productively on farms and sometimes as chimney sweeps and in factories and mines. Today, many children of this age in Africa, India, Pakistan, Bangladesh, through out much of Latin America, and in other economically developing count tries work on farms and some of them do semi-skilled work such as carpet weaving and operating sewing machines. There is a range of intelligence among the Bushmen and most of them will have IQs in the range of 35 to 75. An IQ of 35 represents approximately the mental age of the average European five-and-a-half-year-old and an IQ of 75 represents approximately the mental age of the average European eleven-and-a-half-year-old. The aver­age five-and-a-half-year-old European child is verbally fluent and is capable of doing unskilled jobs and the same should be true for even the least intelligent Bushmen.

Furthermore, apes with mental abilities about the same as those of human 4-year-olds survive quite well as gatherers and occasional hunters and so also did early hominids with IQs of around 40 and brain sizes much smaller than those of modern Bushmen. For these reasons there is nothing puzzling about contemporary Bushmen with average IQs of about 54 and a range of IQs mainly between 35 and 75 being able to survive as hunter-gatherers and doing the unskilled and semi-skilled farm work that a number of them took up in the closing decades of the twentieth century.

2.  Brain Size

The brain size of the Bushmen was estimated at l,250cc by Drennan (1937) and a little higher at l,270cc by Smith and Beals (1990). The Smith and Beals data set also includes Negroid Africans whose brain size is l,282cc and therefore a little larger than that of Bushmen. This is consistent with the higher average IQ of Africans at 67, as compared with the 54 of Bushmen, although the brain size difference of this magnitude can only explain a small fraction of the intelligence difference. The smaller brain size and lower intelligence of the Bushmen compared with the Africans implies that the brain size of the Africans increased over the last 100,000 years or so, since contemporary Africans and Bushmen came from the same ancestral stock. Their brain sizes must have originally been the same and that of the Africans must have increased either as a result of stronger selection pressure or advantageous mutations.

3.  Pygmies

The Pygmies inhabit the equatorial rain forests of Zaire, now called the Congo, and the Central African Republic. At the close of the twentieth century they were thought to number around 100,000 to 200,000. The purest Pygmies are the Mbuti who live in the Ituri forest of northeastern Congo and are thought to number somewhere between 30,000 and 60,000. The other Pygmies are more interbred with Africans. Mbuti Pygmies average around 4' 7" in height. Pygmy children up to the age of puberty have normal height, but when they become adolescents they do not have the growth spurt of other peoples because of their low output of the insulin-like growth factor 1. Most of the Pygmies have remained hunter-gatherers. Typically they live in small groups of around 30 and move from place to place. They have made no progress in the domestication of either animals or plants. In the early twenty-first century the Pygmies in the Congo were described by Cheung (2003) as living "deep in the forests, eking out an existence by hunting and gathering food."

Only one study has been made of the intelligence of Pygmies. This was carried out by Woodworth (1910) using the Sequin Form Board test, which consists of a set of blocks of various shapes that have to be fitted into the appropriate holes. He found that Pygmies performed much worse than other peoples including Eskimos, Native Americans, and Filipinos but he did not quantify their abilities. Judging from their life-style, their intelligence appears to be lower than that of Negroid Africans. Most of them still retain a primitive hunter-gatherer existence while many of the Negroid Africans became farmers during the last few hundred years. In the twentieth century a number of Pygmies worked for Negroid African farmers and these "are always the lower caste, being the farmers' hereditary servants," according to Cavalli-Sforza, Menozzi, and Piazza (1994, p. 178). The term "hereditary servants" appears to be a euphemism for slaves. The enslavement of Pygmies by Negroid Africans is consistent with the general principle that the more intelligent races typically defeat and enslave the less intelligent, just as Europeans and South Asians have frequently enslaved Africans but not vice versa.

Chapter 6. South Asians and North Africans

1. Intelligence of Indigenous South Asians and North Africans

2. South Asians and North Africans in Britain and Australia

3. South Asians and North Africans in Continental Europe

4. Indians in Africa, Fiji, Malaysia, and Mauritius

5. High School and University Students

6. Brain Size

7. Heritability of Intelligence in South Asians and North Africans

8. Genetic and Environmental Determinants of the Intelligence of South Asians and North Africans

9. Intelligence in Israel

The south asians and North Africans are the indigenous peoples of southern Asia from Bangladesh in the east through India, Pakistan, Iraq, Iran, the Gulf states, the Near East, and Turkey, and of North Africa, north of the Sahara desert. They are closely related to the Europeans and in some of the taxonomies of classical anthropology, such as that of Coon, Garn, and Birdsell (1950), the two peoples have been regarded as a single race designated the Caucasoids. But Cavalli-Sforza, Menozzi, and Piazza (1994) in their genetic analysis of human differences have shown that the South Asians and North Africans form a distinctive genetic "cluster" that differentiates them from the Europeans. They are therefore treated here as a separate race.

1. Intelligence of Indigenous South Asians and North Africans

Studies of the intelligence of indigenous South Asians and North Africans are summarized in Table 6.1. The figures are only for general intelligence (g) because there are virtually no data for verbal or visualization abilities. Rows 1 through 13 give twelve results for various locations in India lying in the range between 78 and 88 and with a median of 82. Rows 14 through 17 give IQs of 84, 83, 89, and 80 for four samples of school children in Iran, all taken from the city of Shiraz. Rows 18 and 19 give IQs of 87 for children and adults in Iraq. Row 20 gives an IQ of 86 for Arabs in Israel obtained in the standardization sample of the WISC-R as compared with Jews (IQs of Jews in Israel are in the range of 90-97 and are discussed in section 9). Row 21 gives an IQ of 84 for Jordan calculated from the standardization sample of the KABC. Row 22 gives an IQ of 86 for Kuwait obtained from a standardization of the Progressive Matrices on school children. Row 23 gives an IQ of 82 for school children in various locations in Lebanon. Row 24 gives an IQ of 78 for Nepal obtained from children in 34 schools in towns, villages, and the jungle area. Row 25 gives an IQ of 84 for adolescents at schools in Pakistan in the region of Islamabad. The comparison group is 707 European children in Canada reported in the same study. Row 26 gives an IQ of 84 for children in Pakistan obtained from nine schools around Karachi representing poor and affluent areas. Row 27 gives an IQ of 78 for school children in Qatar. Row 28 gives an IQ of 79 for young children attending a village school in Sri Lanka. Row 29 gives an IQ of 83 for young children in first grade in schools in Syria compared with a sample of 200 of the same age in Germany. Rows 30 through 32 give IQs of 84, 90, and 96 for three samples of school students in Turkey. Row 33 gives an IQ of 85 for Yemen derived from a standardization of the CPM on a sample of 6-11-year-olds.

Rows 34 through 38 give IQs for North African samples. Row 34 gives an IQ of 84 obtained from a mixed sample of North Africans from Algeria, Morocco, and Tunisia. Rows 35, 36, and 37 give IQs of 77, 81, and 83 for school children in Egypt. The thirty-eight IQs of the South Asians and North Africans show reasonable consistency. With the exception of Turkey, all the IQs lie in the range between 77 and 89. The median IQ of the entire set of results is 84 and is considered the best estimate for the IQ of South Asians and North Africans. The median IQ of 90 derived from the three studies for Turkey is higher than that in the remainder of South Asia and North Africa. The most likely explanation for this is that Turkey has straddled Europe and Asia for many centuries and the geographical proximity of Turkey and southeast Europe will have brought about a mixing of Turkish and European genes to produce a European-South Asian cline or genetically hybrid mixed population, with the result that contemporary Turks and Greeks are genetically quite similar, as shown by Cavalli-Sforza, Menozzi, and Piazza (1994) and noted in Chapter 3.

Table 6.1. IQs of South Asians and North Africans

2. South Asians and North Africans in Britain and Australia

IQs of South Asians in Europe and Australia are given in Table 6.2. Row 1 gives an IQ of 87 for Indian children in London collected in the mid-1960s by the Inner London Education Authority (ILEA) and calculated by Vernon (1969, p. 169). Row 2 gives an IQ of 93 for a sample of Pakistani children in London. Row 3 gives an IQ of 91 for Indian children at a comprehensive school in Essex in a study in which Afro-Caribbean children at the same school obtained an IQ of 85. Rows 4 and 5 give IQs of 94 and 89 for Indian and Pakistani children in a town in the British Midlands. Afro-Caribbean children at the same schools obtained an IQ of 86, confirming the result in row 3 finding higher IQs of South Asians than of Africans. Rows 6 and 7 give IQs of 83 and 97 for Indians nation wide and rows 8 and 9 give IQs of 93 and 96 of Pakistani children nationwide..

Rows 10 through 12 give results from a study in which Pakistani, Indian, and Bangladeshi children obtained IQs of 93, 92, and 92, and therefore there were no IQ differences between these three groups from the Indian sub-continent. These IQs are relative to 100 for white children attending the same schools and are likely to be somewhat inflated because 7 percent of white children, mainly middle class with higher IQs, attend private schools and white middle-class par-ents who send their children to state schools typically tend to avoid sending them to schools with large numbers of immigrants. The effect of this will have been that the IQs of the South Asians will be inflated relative to national norms. There are no national norms for the tests used, so the amount by which the IQs are

Table 6.2. IQs of South Asians in Britain and Australia

inflated cannot be determined but is probably around 5 IQ points. Row 13 gives an IQ of 89 for South Asian immigrants in Australia.

The range of IQs of South Asian Europeans in Britain is quite large, from 83 to 97. One reason for this considerable range is that the IQs increase with length of residence in Britain. This is shown in two of the studies. First, rows 6 and 7 give non-verbal reasoning IQs of 83 for Indian children resident for fewer than four years in Britain and 97 for those resident in Britain for four or more years, indicating a gain of 14 IQ points arising from residence in Britain. It is interesting to note that the IQ of 83 of Indian children resident for fewer than four years in Britain is almost the same as the IQ of 82 for Indians in India given in Table 6.1. Second, rows 8 and 9 give non-verbal reasoning IQs of 93 for Pakistani children resident for fewer than four years in Britain and 96 for Pakistani children resident for four or more years in Britain, indicating a gain of 3 IQ points with longer residence in Britain.

The median IQ of the studies of South Asians in Britain is 89 and the IQ of South Asian immigrants in Australia given in the last row is the same. This is a little higher than the IQ of 84 of indigenous South Asians, con­sistent with the results showing that IQs improve with length of residence in Britain and Australia. These IQ gains may be due to a variety of factors. Recent immigrants will have had difficulty in speaking and understanding English and this will have impaired their performance even on non-verbal tests because of difficulty in understanding the instructions given in English. In addition, those who had been born in Britain may have benefited from better nutrition and education than comparable children received in their own countries.

3. South Asians and North Africans in Continental Europe

IQs of South Asians and North Africans in Continental Europe are given in Table 6.3. Row 1 gives an IQ of 86 for Turkish immigrants in Germany. Rows 2 though through 17 give results of 16 studies of the IQs obtained by South Asians and North Africans immigrants in the Netherterrds. A useful review, of a number of these studies has been given by Te Nijenhuis and van der Flier (2001). Row 2 gives an IQ of 78 for a sample of the children of first-generation immigrants from Turkey and Row 3 an IQ of 79 for a sample of the children of second-generation immigrants from Turkey. Both IQs are low and indicate no significant improvement in the intelligence of second-generation immigrants. Row 4 gives an IQ of 75 for a sample of children of first-generation immigrants from Morocco and row 5 an IQ of 79 for a sample of children of second-generation immigrants from Morocco.

Table 6.3. IQs of Solith Asians and North Africans in Continental Europe

Again, both IQs are low but there appears to be some improvement in the intelligence of second-generation immigrants, as has been found in the studies of immigrants in Britain.

Row 6 gives an IQ of 83 for children of immigrants from Morocco and Turkey. Rows 7 and 8 give IQs of 85 and 84 for further samples of Moroccan and Turkish immigrant children. Row 9 gives an IQ of 88 for a sample of Indians, 6 IQ points higher than the median IQ of 82 in India. Row 10 gives an IQ of 93 for Moroccan and Turkish children, the average of 92 obtained on the Otis and 94 on the Cito, both of which are largely verbal tests. Row 11 gives an IQ of 84 for Turkish (n=24) and Moroccan (n=9) children obtained on the standardization sample of the Snijders-Oomen Non-Verbal Test; IQs of those born in the Netherlands were the same as those who had only been in the country from 1 to 6 years. Rows 12 and 13 give IQs of 85 and 84 for samples of Moroccan and Turkish children. Rows 14 and 15 give IQs of 84 and 88 for Moroccan and Turkish adults on the General Ability Test Battery (GATE); this is a Dutch test with eight subtests measuring vocabulary, arithmetical ability, perceptual speed, etc. The Turkish and Moroccan immigrants performed poorly on vocabulary because they had not learned Dutch well and this test has therefore been omitted in the calculation of the IQs. The figures for g are the average of the remaining seven subtests. Row 16 gives an IQ of 92 for Muslims in the Netherlands from Turkey and Morocco compared with approximately 69,000 Dutch Europeans; this figure is obtained from a test of arithmetic entered as verbal IQ. The mean vocabulary IQ of this sample was 85 but this is not entered because most of these children did not speak Dutch as their first language. Row 17 gives an IQ of 94 for second-generation immigrant children of whom 72 percent were from Turkey and Morocco and 10 percent from Surinam and the Netherlands Antilles. Their verbal IQ was 80 but this has been omitted on the grounds that most of them did not speak Dutch as their first language. The results of the studies from the Netherlands are closely similar to those from Britain. The median IQ of the first eight studies of first generation immigrants is 84, the same as that of indigenous South Asians and North Africans. Row 18 gives an IQ of 83 for gypsies in Slovakia. The result is given here because gypsies, or Roma as they are coming to be called, are of South Asian stock who migrated from northwest India between the ninth and fourteenth centuries. This has been shown by linguistic analysis of their Romani language, which has been found to have an Indian origin, and by genetic analysis (Pearson, 1985; Fraser, 1995). They have remained largely isolated from Europeans and their IQ is typical of South Asians.

4. Indians in Africa, Fiji, Malaysia, and Mauritius

There are Indian populations in several countries in Africa. In South Africa they number about one million, of whom approximately 84 percent are in Natal and 14 percent are in the Transvaal. There are also Indians in Kenya and Tanzania, whose ancestors were brought in by the British and Germans under colonial rule to do work of various kinds including building railways. Studies of the IQs of Indians in Africa are summarized in Table 6.4. Row 1 gives an IQ of 77 for the first study of the IQ of Indians in South Africa, compared with 65 for Africans. Row 2 gives an IQ of 88 for Indian computer programming students compared with an IQ of 100 for a comparable sample of 243 whites. Row 3 gives an IQ of 86 calculated from the standardization samples of the Junior South African Individual Scales. This test resembles the Wechsler. The norms for Indians have been calculated in relation to the South African white standardization sample. The test contains a scale for numerical ability, on which the Indians obtained an IQ of 86, which contributes to the overall IQ. Row 4 gives an IQ of 91 in relation to British 1979 norms. White South Africans obtained an IQ of 98; hence Indians scored 7 IQ points below South African whites. Row 5 gives an IQ of 83 in relation to South African whites on the South African Junior Aptitude Test. This test also has two memory subtests on which the Indians obtained an IQ of 89; in the same study Africans in South Africa obtained an IQ of 63, showing again that Indians in South Africa obtain much higher IQs than Africans.

The median IQ of Indians in South Africa derived from the five studies is 86. This is a little higher than the median IQ of 82 of Indians in India and a little lower than the IQ of approximately 89 of Indians born in Britain. Possibly a reason for these differences is that standards of living are lowest in India, higher in South Africa, and highest in Britain, and these have had some effect on intelligence levels. There may also have been differences in the intelligence of the migrants from whom the Indians in South Africa and Britain are descended. The ancestors of the Indians in South Africa were largely recruited to work in the sugar and tobacco plantations and may not have had such high IQs as those who migrated to Britain in the second half of the twentieth century.

Row 6 gives an IQ of 91 for Indians in Tanzania. The sample consisted of secondary school students who had to pass an entrance examination for entry to their schools and the IQ is therefore somewhat inflated. The IQ of this sample is probably about 8 IQ points higher than that of the general population of Indians in Tanzania, which can therefore be estimated at approximately 83,

Table 6.4. IQs of Indians in Africa, Fiji, Malaysia, and Mauritius

closely similar to the IQ of 82 of Indians in India given in Table 6.1. In the same study Africans at the same selective schools obtained an IQ of 78. This difference confirms a number of studies in South Africa and Britain showing that when Indians and Africans are in the same environment, Indians obtain substantially higher IQs than Africans.

Row 7 gives an IQ of 82 for Indians in Fiji, in which there are approximately the same number of Indians and indigenous Fijians. The Fijians obtained a mean IQ of 84 in the same study. Row 8 gives an IQ of 88 for Indians in Malaysia obtained from a standardization of Raven's Standard Progressive Matrices. Row 9 gives an IQ of 89 for 11-year-olds in Mauritius described as "a community sample," of which 69 percent were Indians and the remainder Creoles of mixed European and sub-Saharan African descent, whose IQ is 2.5 points lower than that of the Indians (Raine, Reynolds, Venables, and Mednick 2002). The studies summarized in Table 6.4 lie in the range between 77 and 91 and have a median IQ of 88, a little higher than that in India and about the same as that of Indians in Europe. This is probably because Indians outside India generally enjoy higher living standards and possibly because those who have emigrated from India have had above average intelligence.

5. High School and University Students

Studies of the intelligence of South Asian and North African students in high school, colleges, and universities are summarized in Table 6.5. It would be expected that these would be somewhat higher than the intelligence of general population samples and it will be seen that this is the case. Row 1 gives an IQ of 85 for students at the University of Alexandria and is only fractionally higher than the IQ of 83 on the same test of a general population sample given in Table 6.1. Row 2 gives an early study of1926 in which second year students at the University of Calcutta obtained an IQ of 95. The test used was the Stanford, on which American students at the University of Stanford obtained a mean IQ of 113. Row 3 gives an IQ of 93 for 14-year-old students at St. Xavier's School in Delhi described as coming from upper-class families. Rows 4 and 5 give IQs of 90 and 88 for students at the Punjab University. Row 6 gives an IQ of 88 for women students at various colleges in the Indian city of Amritsar. Row 7 gives an IQ of 90 for university students in engineering, economics, and the liberal arts in Tehran. Row 8 gives an IQ of 92 for high school students in Baghdad described by the author of the study as "a highly selected group, since education is not compulsory at the high school level and students who do reach this level have to pass rigid examinations" (Alzohaie, 1966, p. 476).

Rows 9 and 10 give IQs of 98 and 102 for engineering students at the University of the Witwatersrand in South Africa. In this study European students in the same faculty obtained IQs of 106 and 113, and black African students IQs of 93 and 99. Row 11 gives an IQ of 106 for a further sample of engineering students at the University of the Witwatersrand in South Africa. In this study European students in the same faculty obtained an IQ of 116, and black African students an IQ of 101. Thus in these three studies of students the IQs of the Indians fall midway between those of whites and blacks, as they do in general population samples. Also, in these studies the IQs of the Indians are somewhat higher than those in South Asia and North Africa. This is probably attributable to the IQs of Indians in South Africa being higher and because the engineering department of the University of the Witwatersrand takes relatively talented students.

Rows 12 and 13 give IQs of 92 and 101 for university students in Turkey.

Table 6.5. IQs of South Asian and North African high school and university students

Row 12 gives an IQ of 92 for 15 women and 24 men students of dentistry at Attaturk University in the city of Ezurum. Row 13 gives an IQ of 101 for medical students at the same university.

The median of the studies is an IQ of 92, eight points higher than that of general population samples of South Asians and North Africans. The interest of these studies is that they show that South Asian and North African university students with extensive education and from upper and middle class families have lower IQs than average Europeans. This indicates that lack of education is unlikely to be a major factor responsible for the low IQs of general population samples. The IQs of South Asian and North African students are also lower than the median of 105 for European college stu­dents (Table 3.3). Thus the 15 IQ point difference between Europeans and South Asians and North Africans in general population samples is closely similar to the 14 IQ point difference between college students.

6. Brain Size of South Asians

Four sets of data on the brain size of South Asians compared with that of Europeans are shown in Table 6.6. Row 1 gives data assembled by Smith and Beals (1990) from approximately 20,000 crania collected worldwide

Table 6.6. Brain size (cc) of Europeans and South Asians

and shows a European advantage of 84cc. Row 2 gives data assembled from various sources by Groves, who contends that there are no racial differences in brain size, but which nevertheless show a European advantage of 63cc. Row 3 gives average brain size of six samples of Europeans from North America and Europe and two samples from India from data compiled by Jurgens et al. (1990) and analyzed by Rushton (2000, p. 124) showing a European advantage of 126cc. Row 4 gives data compiled by the U.S. National Aeronautics and Space Administration (NASA) for the average of 19 European male military samples and for a male Iranian military sample showing a European advantage of 114cc. The figures in the four data sets all show greater brain size of Europeans and are reasonably consistent, considering that they were compiled using different methods. The Smith and Beals data are derived from measurements of the volume of skulls, the Groves data come from various sources, while the data sets in rows 3 and 4 have been calculated from external measurements of the heads of living individuals. The average of the four data sets is a European Caucasoid advantage of 97cc.

7.  Heritability of Intelligence in South Asians

There have been two studies of the heritability of intelligence in India, both of which have used the method of comparing the IQs of identical (MZ) and non-identical (DZ) twins. Pal, Shyam, and Singh (1997) have reported a study of 30 MZ and 30 same-sex DZ adult twins and calculated the heritability at 0.81. If this is corrected for attenuation, as­suming a test reliability of 0.9, the heritability becomes 0.90. In a second study, Nathawat and Puri (1995) obtained a heritability of 0.90, which corrected for attenuation assuming a test reliability of 0.9 becomes 1.0. Thus, the heritability of intelligence in India is marginally higher than that of 0.83 in Europeans.

8.  Genetic and Environmental Determinants of the Intelligence of South Asians and North Africans

We saw in Table 6.1 that the median IQ of the studies of indigenous South Asians and North Africans is 84. This IQ is depressed environmentally because of the low standard of living of these peoples. A report on malnutrition in South Asia and North Africa in the early 1990s published by UNICEF (1996) estimated prevalence rates of stunting of 24 percent of children in the Middle East and North Africa and 60 percent of children in South Asia, indicating the effect of sub-optimal nutrition. There is little doubt that this has an adverse effect on intelligence. Nevertheless, it seems likely that genetic factors are also involved. First, the very high heritabilities of intelligence in both South Asians and Europeans show that genetic factors are largely responsible for differences in intelligence within the two populations, and this makes it likely that these contribute to the differences between the two populations. Second, it has been shown that South Asians and North Africans living in the affluent European environments of Britain, Australia, and the Netherlands have median IQs of 89, 89, and 94. All of these are higher than the average IQ of 84 of those in their indigenous homelands and poorer environments. These figures show that when South Asians and North Africans are reared in European environments their IQs increase but they do not increase to the same level as those of Europeans. This suggests the presence of genetic factors. Third, the IQ of Indians in South Africa is 86. This is higher than the IQ of 82 in India and is attributable to the better living standards, but it is substantially below the IQ of Europeans. The Indians were brought to Natal in the 1850s to work on the sugar plantations (Johnston, 1930). They have had some four to six generations to adapt to life in South Africa, live in the same country and in approximately the same environment as Europeans, yet a large IQ difference remains and suggests a genetic difference between the two populations. Fourth, the average brain size of South Asians is about 8 percent smaller than that of Europeans and may be partly due to sub-optimal nutrition but is likely also to have some genetic basis and contribute to the intelligence difference.

9. Intelligence in Israel

Intelligence in Israel is higher than in the other countries of South Asia and North Africa. Eight studies of intelligence in Israel are summarized in Table 6.7. The IQs lie in the range of 89-97 with a median of 95. This is substantially higher than the median of 84 for the remainder of South Asia, showing that Jews have higher IQs than other South Asians. In Israel approximately 20 percent of the population are Arabs, whose IQ of 86 (see Table 6.1, row 20) is virtually the same as that of other South Asians in the Near East. Forty percent of the population are European Jews (mainly Ashkenazim from Russia and Eastern Europe) and 40 percent are Oriental Jews (Mizrahim) from Asia and North Africa. Three studies carried out in Israel have found that the Ashkenazim have a mean IQ approximately 12 IQ

Table 6.7. Intelligence in Israel

points higher than the Oriental Jews (Zeidner, 1987a; Burg and Belmont, 1990; Lieblich, Ninio, and Kugelmass, 1972). The IQ of 95 for Israel is the weighted mean of the IQs of 103 of the Ashkenazim Jews, 91 of the Oriental Jews (12 IQ points lower), and 86 of the Arabs. The lower IQ of Arabs in Israel compared with Jews is confirmed by Zeidner (1987a), who has reported that Arab applicants for admission to university obtain an IQ 15 IQ points lower than that of Jewish applicants.

There are two questions concerning the Jewish IQ that require explanation. The first is why the Ashkenazim Jews in Israel have an IQ of 103. This is not particularly surprising because there is considerable evidence that Ashkenazim Jews in the United States and Britain have substantially higher IQs than Gentiles. In the United States, a study published in the 1920s reported that Jewish 10-year-olds had an IQ 13 points higher on the Stanford-Binet test than European Gentiles (Ns=110 and 689, respectively) (Bere, 1924). In the 1940s Nardi (1948) reported an IQ of 110 on the Stanford-Binet test for Jewish 12-year-olds (N=l,210), and in the 1950s Levinson (1957) found an IQ of 109 for Jewish 12-year-olds (N=2,083), also on the Stanford-Binet test. Herrnstein and Murray (1994) reported an IQ of 112.6 for Jewish adolescents in their study of the National Longitudinal Study of Youth, and the latest study has found an IQ of 107.5 in a nationally representative sample (N=150) of adults (Lynn, 2004). Similarly high IQs for Jewish children have been reported in Britain. In the 1920s Davies and Hughes (1927) found that Jewish 8-14-year-olds in London had an IQ of 110 (N=l,081), compared with 100 for British children. In the 1960s Jewish 10-year-olds in Glasgow had an IQ of 117.8 (N=907) compared with Scottish children in the same city (Vincent, 1966). However, this figure for the Jewish IQ is too high for a comparison with British children as a whole because the IQ of children in Glasgow is 93.7 in relation to 100 for the national average (Lynn, 1979). To compare the mean IQ of Jewish children in Glasgow with that of British non-Jewish whites we have therefore to subtract 6.3 IQ points from their score, giving them a mean IQ of 111.5. Thus, the IQs of Jews in the United States and Britain average between around 107 to 115 and are therefore higher than the 103 estimated for Ashkenazim Jews in Israel. Some possible explanations for this are that few American and British Jews have emigrated to Israel. Most of the Ashkenazim Jews in the United States and Britain fled persecution in Russia and Eastern Europe between 1880 and 1914 and in Germany between 1933 and 1939. It seems likely that these would have been the more intelligent who foresaw the dangers of staying and were able to organize emigration. Those who remained in Russia and Eastern Europe would likely have been a little less intelligent. These are the ones who emigrated to Israel after World War II ro escape persecution and poverty and whose IQs are a little lower than those of Ashkenazim Jews in the United States and Britain. A further factor is that many of these supposedly European Jews are not Jews at all but pretended to be Jews in order to get permission to leave the Soviet Union (Abbink, 2002).

A second problem concerning the intelligence of Jews is that all Jews were originally from the same stock, so why is the intelligence of Ashkenazim Jews approximately 12 IQ points higher than that of Oriental Jews? There are probably two answers to this question. The first is that despite strict Jewish prohibitions on exogamy, there has always been some inter-marriage and inter-mating between Jews and non-Jews living in the same localities. Even a small amount of exogamy over many generations is sufficient to introduce significant proportions of non-Jewish genes into the Jewish gene pool. The effects of this are visible in European Jews, a number of whom have fair hair and blue eyes. The result of this will have been that Ashkenazim Jews in Europe will have absorbed a significant proportion of the genes for higher intelligence possessed by the Europeans, while the Oriental Jews in the Near East and North Africa will have absorbed a significant proportion of the genes for lower intelligence from the South Asians and North Africans. The second factor that has probably operated to increase the intelligence of Ashkenazim Jews in Europe and the United States as compared with Oriental Jews is that the Ashkenazim Jews have been more subject to persecution. Jews were less persecuted over the course of many centuries in Southwest Asia and North Africa. Oriental Jews experienced some persecution sufficient to raise their IQ of 91, as compared with 84 among other South Asians and North Africans, but not so much as that experienced by Ashkenazim Jews in Europe.

The 12 IQ point difference between Ashkenazim Jews and Oriental Jews in Israel is almost certainly to some degree a genetic difference. Genetic analysis by Hammer, Redd, and Wood (2000) has shown that all Jews have some genetic affinity (except for the Ethiopian Jews) arising from their common original stock in the Near East but that European and Oriental Jews form two genetic families, the European Jews with some genetic affinity with gentile Europeans and the Oriental Jews with some genetic affinity with Southwest Asians and North Africans.

Chapter 7. Southeast Asians

1. Intelligence of Indigenous Southeast Asians

2. Southeast Asians in the United States and the Netherlands

3. Brain Size of Southeast Asians

4. Genetic and Environmental Determinants of the IQ of Southeast Asians

The southeast asians are the indigenous peoples of Burma, Thailand, Cambodia, Vietnam, Malaysia, Indonesia, the Philippines, and Borneo. In classical anthropology they were designated the Malays (Morton, 1849; Coon, Garn, and Birdsell, 1950) or the Indonesian-Malays (Cole, 1965). Their distinctive racial identity has been confirmed by the genetic analysis made by Cavalli-Sforza, Menozzi, and Piazza (1994) in which these peoples constitute a genetic "cluster." They have some genetic affinity with the East Asians with whom they are to some degree interbred, but the flattened nose and epicanthic eye-fold are less prominent.

1. Intelligence of Indigenous Southeast Asians

IQs for samples of Southeast Asians from five countries are given in Table 7.1. Rows 1 through 4 give IQs for Indonesia. Row 1 gives an IQ of 86 for children in the city of Bandung in Java. Row 2 gives an IQ of 87 for children and adolescents in two villages in central Java. Row 3 gives an IQ of 87 for children of families working on a tea plantation. Row 4 gives an IQ of 87 for children in northern Jakarta. Row 5 gives an IQ of 90 for Lao children living in a village and "not from families living in abject poverty." Row 6 gives an IQ of 88 for mothers of the children given in row 5. Row 7 gives an IQ of 89 for Malays in Malaysia obtained in the standardization of the Standard Progressive Matrices. Row 8 gives an IQ of 85 for Malay college students at the International Islamic University in Kuala Lumpur in relation to college students at universities in Germany, Russia, and the United States. Row 9 gives an IQ of 86 for the Philippines obtained from school children in Manila. Row 10 gives an IQ of 93 for 13-year-old Malays at school in Singapore. Row 11 gives an IQ of 91 for school children in Thailand obtained from Chon Buri province, an agricultural area on the east coast. The IQs lie in the range between 86 and 93 and the median is 87.

Table 7.1. IQs of Southeast Asians

2. Southeast Asians in the United States and the Netherlands

IQs of Southeast Asians in the United States and the Netherlands are summarized in Table 7.2. Row 1 gives an IQ of 96 for an early study of a sample of Filipino children in Hawaii tested with the Porteus Mazes. Row 2 gives an IQ of 89 for a sample of Filipinos in Honolulu collected by Smith (1942) in 1924 and 1938. Row 3 gives an IQ of 91 for Filipino children on the Hawaiian island of Kauai. Row 4 gives an IQ of 93 for a sample of Filipinos in Hawaii obtained from the mathematics subtest of the STAS. Row 5 gives an IQ of 87 for a sample of Filipinos in the United States calculated by Flynn (1991). Row 6 gives an IQ of 92 for a sample of second-generation Indonesian immigrants in the Netherlands. Row 7 gives an IQ of 94 for a sample of mainly Vietnamese high school students in an American city calculated by Flynn (1991). This sample obtained a verbal IQ of 87 measured by the Mill Hill Vocabulary Scale. This is probably slightly depressed in relation to their non-verbal reasoning IQ because many  of them had not acquired fluency in English. The median of the seven studies is an IQ of 93 and is a little higher than the IQ of 87 of indigenous Southeast Asians. It is possible that a selective element in migration to the United States and the Netherlands may be part of the explanation for this, and a further possible factor is that Southeast Asians in the United States and the Netherlands enjoy a higher standard of living and of nutrition than indigenous Southeast Asians.

Table 7.2. IQs of Southeast Asians in the United States and the Netherlands

3. Brain Size of Southeast Asians

Studies of differences in brain size between Europeans and Southeast Asians are summarized in Table 7.3. Row 1 gives the results calculated by Gould (1981) from the collection of skulls assembled in the nineteenth century by the American physician Samuel Morton (1849). The number of skulls was quite low, consisting of 18 Southeast Asians and 52 Europeans, and not a great deal of weight can be attached to the results. They are given here largely for historical interest. Row 2 gives results from six populations of Southeast Asians compared with nine populations of Europeans showing a difference of 37cc. The standard deviations are given by Beals et al. (1984). The numbers of individuals are not given but are part of a total collection of approximately 20,000 and can be assumed to be several thousand. Despite the small size of Morton's sample and Gould's accusation that Morton massaged his results to give a larger brain size for Europeans, the results agree closely with the later study of Smith and Beals. Row 3 gives a much larger difference based on average brain sizes for 190 samples of Europeans and 20 samples of Southeast Asians. Thus, all three data sets show smaller brain size in Southeast Asians than in Europeans, consistent with their lower IQs.

Table 7.3. Brain size (cc) differences of Europeans and Southeast Asians

4. Genetic and Environmental Determinants of the IQ of Southeast Asians

The IQ of Southeast Asians in the United States is higher at 93 than that of indigenous Southeast Asians, 87. This difference is attributable to the better environment with higher living standards in the United States, with better nutrition, education, and welfare. The effect of these is that the IQ gap between Southeast Asians and Europeans is approximately halved. Nevertheless, a 7 IQ point difference remains when Southeast Asians and Europeans are raised and live in approximately the same environments. This suggests that genetic factors contribute to the difference in intelligence between the two races. The smaller average brain size of Southeast Asians compared with Europeans also suggests a genetic difference.

Chapter 8. Australian Aborigines

1. Intelligence of Australian Aborigines

2. Aboriginal-European Hybrids

3. Piagetian Intelligence

4. Spatial Memory

5. Brain Size

6. Genotypic Intelligence

7. Intelligence of New Guineans

8. Conclusions

The australian "aborigines are the indigenous people of Australia. They are also known as the Australids, have long been recognized as a race in classical anthropology, and are one of the seven major races in the taxonomy proposed by Coon, Garn, and Birdsell (1950). They have a distinctive profile of blood groups, about 73 percent of them having O group as compared with a little fewer than 50 percent among Europeans; the remaining 27 percent are A, and there are virtually none with the B group. Their distinctive racial identity has been confirmed by the genetic analysis made by Cavalli-Sforza, Menozzi, and Piazza (1994) in which the Australian Aborigines together with the original New Guineans constitute a genetic "cluster." The reason that the Australian Aborigines and the original New Guineans are closely related genetically is that the ancestors of the Australian Aborigines migrated from New Guinea to Australia about 60,000 years ago (Bradshaw, 1997). Those who migrated split from those who remained in New Guinea and today inhabit the interior highlands. Also closely related to the Australian Aborigines are the now extinct Tasmanians. The last pure Tasmanian died in 1876, but there are still a few mixed-race Tasmanians.

It has been estimated that before the Europeans arrived there were around 300,000 Aborigines in Australia. Their numbers were considerably reduced following the colonization of Australia by Europeans, partly as a result of diseases contracted from Europeans from which they lacked immunities, and partly as a result of Europeans killing them. In the second half of the twentieth century, the numbers of Aborigines in the censuses of 1961, 1971, and 1981 were recorded as approximately 106,000, 139,000, and 171,000. The rapid increase in numbers has been a result of high birth rates and a reduction of infant and child mortality.

In the second half of the twentieth century there were three groups of Australian Aborigines. The first lived on government reserves principally in the north and center of Australia. The second group lived on the outskirts of country towns and stations. The third lived in larger towns and cities. Both the second and third groups typically attended schools with Europeans. Many of the second and third groups have some European ancestry while those on the reservations are largely pure Aborigines.

1. Intelligence of Australian Aborigines

The first attempt to estimate the intelligence of the Australian Aborigines was made by Galton (1869). On the basis of travelers' accounts of their accomplishments he estimated their intelligence was approximately three "grades" below that of the English. In Galton's metric, a grade was equivalent to 10.4 IQ points. Hence in terms of the IQ scale, he estimated the Australian Aborigine IQ at 68.8. Subsequent studies of the intelligence of Australian Aborigines assessed by intelligence tests have shown that this was a fairly accurate assessment. These studies are summarized in Table 8.1. Row 1 shows the results of the first study, giving an IQ of 66 obtained by Porteus with his Maze Test, a series of paper and pencil mazes of increasing complexity from which mental age is measured as the success rate of the average child of the corresponding chronological age. The Maze Test was later incorporated into the Wechsler tests and provides a measure of g and of visualization. The mean mental age of his sample adults was 10.5, the approximate equivalent of an IQ of 66. Row 2 gives results for the next study that used the Porteus Mazes on a sample of Aborigines at La Grange Bay in northwest Australia. The men obtained a mental age of 10.5 and the women of 8.6. The average mental age of the two sexes was 9.55, equivalent to an IQ of 59. Row 3 gives a closely similar result obtained by Porteus for adults at the Beagle Bay Mission in the Kimberley region; the Aborigines obtained a mental age of 9.35, equivalent to an IQ of 58. Row 4 gives an IQ of 69 obtained from two visualization tests (Alexander passalong and Fergusson Form Boards). Row 5 gives an IQ of 70 from a study of the Wailbiri Aborigines of Central Australia carried out by Porteus and Gregor in the 1960s. Row 6 gives an IQ of 58 for a sample at a primary

Table 8.1. IQs of Australian Aborigines

school in Maningrida in the Northern Territories. Row 7 gives an IQ of 74 for a sample of adults who obtained a mental age of 11.8. Rows 8 and 9 give IQs of 62 and 64 for two samples of Aboriginal children attending schools with white children in a town in New South Wales.

Row 10 gives an IQ of 70 for a sample of Aboriginal adults tested with the Colored Progressive Matrices. Row 11 gives a verbal IQ of 67 for 3-and 4-year-old Aboriginal children attending pre-school with whites in Bourke. Row 12 gives a verbal IQ of 52 for children attending schools at the Hermannsberg Mission in central Australia. Row 13 gives an IQ of 58 for Aboriginals calculated in relation to the norms for European children in New Zealand. Row 14 gives an IQ of 62 on the Spiral Omnibus Reasoning Test for a sample of 13-year-old Aboriginal children attending school on an Aboriginal reserve in Queensland. Row 15 gives a verbal IQ of 59 for a sample of 6-10-year-old Aboriginal children in Alice Springs in central Australia. Row 16 gives a reasoning IQ of 60 for a sample of adults with an average age of 25. Row 17 gives a vocabulary IQ of 61 for a sample of 4-year-olds.

The IQs range between 52 and 74. The median IQ of the seventeen studies is 62 and represents the best estimate of the average intelligence of Australian Aborigines. Verbal ability is a little weaker than visualization abil