Race Differences in Ethnocentrism

by Edward Dutton

  Equally, we could divide races, or even nations, in accordance with the percentage of genes the members share in comparison to outsiders. Populations that look physically different are evolved to different environments. As such, they are separate breeding populations, and thus have more genes in common with each other than with outsiders. In this sense, they are an extended family and a different race is a different extended family. Salter’s (2007) analysis showed that if the world population were just English, then the kinship between any random pair of Englishmen would be zero. But if the world population consisted of both English people and Danes, then two random English people would have a kinship of 0.0021. This would make them sixth cousins when compared to a Dane. As genetic distances between populations become larger, the kinship coefficient between random co-ethnics within a population gets larger. But this again shows that the racial division is meaningful and has a clear statistical basis: members of a race have more genetically in common with co-ethnics than with members of any other race, as we have seen from research on polymorphism clusters. What it also means, and this should be emphasized, is that races are constantly evolving as different groups within the broader category breed according to different patterns.

  The third supposed problem with race is that deploying it leads to bad consequences. It legitimizes ‘racist groups’ and so forth. That it does this is clearly of no relevance to whether or not it is a philosophically justifiable and predictive category. This argument commits the fallacy of ‘appeal to consequences’ and, depending on how the consequences are described, ‘appeal to emotion’.

  The fourth criticism is that there are more differences within races than there are between them. Likewise, you could argue that there are more differences within humanity than there are between humans and chimpanzees. There is, after all, only a 1.5% difference between humans and chimpanzees (Caccone & Powell, 1989). I do not think many people would argue that the distinction between humans and chimpanzees is meaningless. We are talking about comparative differences. Dividing between two racial categories, for example, permits accurate predictions to be made about each, even if the differences are very small (e.g. Hoffman, 1994). The genetic differences (in terms of heritable musical ability) between a standard musician and Mozart are probably rather small but these differences have clear and important consequences. Tiny genetic differences (humans only differ by 0.0012%) can have significant consequences. As we will see, it is possible to extend this understanding of within-group differences to between-group differences.

  In addition, as Cochran and Harpending (2009, p. 15) have noted, there are more genetic differences within breeds of dog than between breeds of dog, but nobody would dismiss as insignificant the differences between a Great Dane and Chihuahua. In addition, they note that ‘information about the distribution of genetic variation tells you essentially nothing about the size or significance of trait differences … If between-group genetic differences tend to push in a particular direction — tend to favour a certain trend — they can add up and have large effects’. Thus, we can conclude that the criticisms of the race concept can be successfully refuted.

  6. Conclusion

  In this chapter we have examined race taxonomies and the concept of race. We have argued that ‘race’ is nothing more than what would be termed ‘subspecies’ or ‘breed’ in the animal kingdom of which the human species is a member. We have demonstrated that, following our previous discussion of what constitutes a meaningful and scientific category, race is indeed a meaningful category, because it permits correct predictions to be made. We have seen that the traditional races of classical anthropology have a clear quantitative basis and that gene and disease frequencies cluster along racial lines. Finally, we have examined the criticisms of the concept of race and shown that they can be refuted.

  Chapter Three

  What Is Intelligence?

  1. Introduction

  In this study of ethnocentrism, the issue of intelligence will be examined as a possible factor in explaining differences in group levels of ethnocentrism. Accordingly, in this chapter we will examine the issue of intelligence, as well as the evidence for racial differences in average intelligence.8

  2. What Is Intelligence?

  For the purposes of this book, it suffices to say that we define ‘intelligence’ as the ability to solve complex problems and to solve them quickly. The more quickly you can solve a given problem the more intelligent you are and the more intelligent you are the more complex the problem has to be before you are stumped.

  Intelligence is a single entity that can be measured by IQ (intelligence quotient) tests. These tests are divided into three components testing linguistic, mathematical, and spatial intelligence. People vary in their performance on the components, but performance is positively correlated, proving that there is a single entity known as g (general intelligence) which underpins these intelligences. This model of intelligence is widely accepted by experts in the field such as Lynn (2006), Jensen (1998), and Mackintosh (1998). There are three fundamental forms of intelligence: mathematical reasoning, verbal reasoning, and spatial reasoning, with these three understood to be underpinned by g.

  In statistics, a correlation refers to a relationship between two variables and the degree of its strength. So, if the correlation was 1, the two things always go together and if it’s −1 then they never do. Usually, correlations are between 0 and 1. So, a 0.7 correlation is strong and means that the two variables often go together. Correlations tell us what percentage of the variance is explained by a particular variable. So, if the correlation between IQ and how well people perform in school leaving exams in 0.7 (which it is, see Jensen, 1998) then IQ explains 49% of the variation in how well people do in school leaving exams. The percentage of the variance is always the correlation squared.

  The criticisms of the concept of intelligence, as defined above, are highly problematic. As we will draw upon the concept of intelligence and upon IQ tests, we will now look at these.

  1. But what do you mean by ‘Intelligence’? Critics argue that ‘intelligence’ is difficult to precisely define. Where do you draw the border between ‘intelligent’ and ‘not intelligent’? This point could be made about any concept to varying degrees. It could be argued that ‘tall’ is difficult to define, but that does not mean we cannot talk about ‘tall people’. The world is a mass of information which we make sense of through categories which allow seemingly correct predictions to be made. If we can’t do that, then we wouldn’t be able to do anything because we could never make any predictions. Insisting categories be perfectly defined fails the test of pragmatism.

  2. There are different kinds of intelligence, such as ‘emotional intelligence’. It has been shown that emotional intelligence correlates with g at 0.3 (Kaufman et al., 2011). Also, since in general (in dictionaries, for example), intelligence is defined as we have defined it, it is confusing to use the concept differently. Additionally, if everyone has some kind of ‘high intelligence’ then the concept becomes meaningless in terms of making predictions.

  3. Intelligence means different things in different cultures. We have stated what we mean by intelligence. If a different culture talks about something different, then they’re not talking about intelligence.

  4. Intelligence is simply what IQ tests test. As we will see below, IQ test results are statistically significantly and strongly correlated with other differences. These differences include health, law-abidingness, access to resources (Lynn & Vanhanen, 2012), and general knowledge (Spearman, 1904) so intelligence is germane in all cultures.

  Correlations, it should be noted, are tested for ‘statistical significance’. This is a way of proving that the correlation is not a fluke, based on the sample size. For example, with a sample of twelve people you might find a strong negative correlation between education level and wearing blue clothes but this could just be down to chance. So, a p test, a test of significance, allows us to establish whether
or not the relationship is a fluke. If we can be at least 95% confident that it is not a fluke (p = < 0.05) then we accept that the relationship is genuine. This, of course, involves drawing a random border. If something has a 94% probability of not being a fluke then we might say that ‘near significance’ has been attained. But the border of acceptability has to be drawn somewhere and it is accepted among scientists that this is at 95% certainty.

  5. We do not fully understand intelligence, so intelligence research is speculative. It is true that we do not yet understand the precise brain architecture of intelligence, but this does not mean that we cannot talk about intelligence. We could talk about stars before we understood their architecture (Levin, 2005).

  Criticisms have also been leveled against IQ tests. We respond to the most common of these below.

  1. A few dozen questions are insufficient to test mental ability. It is quite true that, in a minority of cases, an IQ test score may be skewed by illness, stress, or even developing slightly later than one’s peers, but there is a significant correlation between adolescent IQ score and later achievement in various fields which relate to mental ability. For example, IQ test scores predict educational success at 0.5 overall, 0.7 at school, 0.5 at undergraduate level, and 0.3 at postgraduate level (Jensen, 1979, p. 319). They also predict income, health, criminality (negatively), and many other factors (Lynn & Vanhanen, 2012).

  2. IQ tests are unable to measure intelligence. To argue that intelligence is real yet IQ tests do not measure it is like claiming that weight is real, and some people are heavier than others, but bathroom scales do not accurately measure it. A pair of scales is reliable if its estimation of the heaviness of different people positively correlates with our own estimation when trying to lift the same people. Likewise, an IQ test is reliable if its estimation of the intelligence of different people positively correlates with differences in their intelligence as measured by more intuitive measures of intelligence, such as academic performance. The instrument, in both cases, may be imperfect, but it is the best instrument we have. In that IQ scores positively correlate with evidence of intelligence (such as educational attainment), they are the best (if imperfect) means we have of measuring intelligence, just as bathroom scales are the best (if imperfect) means we have of measuring weight. Different scales will give people slightly different weights just as different IQ tests will give different people slightly different IQs.

  3. Intelligence and IQ are not the same thing. We have defined intelligence as ability in cognitive tasks. Academic exams involve cognitive tasks, and successful performance in school exams is predicted by IQ at 0.7 (Jensen, 1979, p. 319).

  4. IQ tests are unreliable. No test instrument is perfectly reliable. Modern IQ tests, in particular the Raven Progressive Matrices (first developed in 1938), have been argued to have a reliability of at least 0.9 (Jensen, 1998, pp. 49–50), so it is simply inaccurate to brand them unreliable.

  5. The tests are not predictive of life outcomes because some successful people, such as Einstein, are brilliant at mathematics but less good at linguistic tasks. This criticism fails to appreciate that this kind of contrast is relatively rare. In general, those who perform above average on linguistic tasks also perform well on spatial and mathematical tasks, and this implies the presence of g (general intelligence). The correlation at age sixteen between verbal and mathematical intelligence on the NCDS (the UK-based National Child Development Study, N. 17,000) is 0.65 (Kanazawa, 2012, p. 42). The subjects generally perform better on one kind of task than on another, but the crucial point is that there is a strong positive correlation. Spearman’s (1904) own research found a correlation of 0.64 between performance in English (mainly linguistic intelligence) and performance in Math. This demonstrates that many of the subjects were better at English than Math or vice versa. But it also evidences our ability to posit g and shows a strong positive correlation. With this in mind, IQ tests can be ‘g-loaded’ such that they more accurately test g, steadily eliminating aspects of the test which have been shown not to relate to g. This has led some IQ tests to have g-loadings of around 0.9, which means that the argument that they are unfair is very difficult to sustain.

  6. IQ Tests are culturally biased. The tests are argued to be culturally biased and unfair on certain races and classes (e.g. Ryle, 1974, p. 54). This is simply inaccurate. With regard to race, East Asians score better on IQ tests than the Europeans who developed them, and East Asian Americans score better than white Americans (Jensen, 1981, p. 205). In addition, reaction times correlate with IQ at about 0.4 (Hunt, 2011, p. 151), and racial differences in reaction times are in the same direction as racial differences in IQ scores (Jensen, 1998, Ch. 11). Also, relative brain size correlates with IQ at around 0.3, and racial differences in relative brain size are in the same direction as the differences in IQ scores (Jensen, 1998, p. 437). As such, IQ tests significantly correlate with objective measures so that it is very difficult to argue they are culturally biased.

  In addition, the specific theory of bias known as ‘stereotype threat’ is not persuasive. The argument runs that blacks, for example, are stereotyped to do worse than whites on IQ tests, so they do worse solely because this expectation creates stress. However, in the case of blacks, this is a misreading of Steele and Aronson (1995) who actually found a 1 SD (standard deviation) difference between black and white IQ scores even when controlling for stereotype threat (see Ganley et al., 2013 for meta-analysis). Also, large-scale strongly controlled attempts to replicate stereotype threat, for example in relation to females and mathematics, have consistently failed (Ganley et al., 2013). As such, it seems clear that our definition of intelligence is the most useful one and that IQ tests are reliable.

  3. Race Differences in Intelligence

  The different races have adapted to their different environments through different modal personalities and average levels of intelligence. Beginning with intelligence, we will set out below the average IQs for different races. However, before doing this we must respond to the argument that though there are racial differences in IQ these are for environmental rather than genetic reasons.

  Environment is a significant factor in understanding intelligence. Intelligence is less genetically predicted in a poor environment in which people are prone to serious childhood illnesses which might reduce IQ. British psychologist Richard Lynn (2006) has set out the following average racial IQs based on a detailed meta-analysis of average national IQs.

  1. Ashkenazi Jews: 112.

  2. Northeast Asians: 105.

  3. Europeans: 100.

  4. African Americans: 85.

  5. South Asians/Middle Easterners: 80.

  6. Sub-Saharan Africans: 70.

  The average IQ of African Americans is especially interesting because it sits between that of Africans and Europeans. This implies that much of the difference between Europeans and Africans is genetic, since the environment of African Americans is much closer to the European one than it is to the African one. Thus, the African American IQ probably reflects an element of environment and some admixture with Europeans. African Americans are, as we have noted, up to about 25% European genetically. Interestingly, the IQ in Jamaica, where European admixture is much lower and living standards lower, is around 75 (Lynn & Meisenberg, 2015). These differences make evolutionary sense. A cold environment, such as that to which Northeast Asians are adapted, would have strongly selected in favor of intelligence because it would require the ability to plan, conceive of survival strategies (such as building shelters and making clothes) in extremely harsh conditions, and it would select in favor of very low time preference, of those able to think far into the future to plan for the colder times. This is known as the Cold Winters Theory. It is also possible that a trade-off had to be made, in terms of sexual selection, in selecting for parasite resistance or for intelligence. Intelligence would be more important in colder climates while parasite resistance (advertised through physical prowess) would be more important in tropica
l climates (Miller, 2000).

  Lynn (2006) has found important differences in IQ within racial categories. Thus, for example, in Europe, though the average IQ is 100, there are clear regional differences. The average IQ of white British people is 100, but it is only 93 in the Republic of Ireland, possibly reflecting a long history of outward migration, with intelligence predicting migration (Lynn, 1988 or Lynn, 1980). The average IQs in Greece (92), Portugal (94), and in certain Balkan countries are around this level, reflecting the fact that they are clines between Europeans and South Asians in Turkey in the case of Greece and the Balkans or, in the case of Portugal, between Europeans and Sub-Saharan Africans (Lynn, 2006, p. 15). In Eastern Europe, some countries, such as Latvia, have IQs below 100. In the case of Latvia, for example, this is 97. If this is not down to sampling errors, then it is also likely to reflect emigration, especially during the turbulence of World War II. Northeast Asian IQs also vary, with the Mongolian IQ being only 100. Significant variation occurs within racial groups.9