Chapter Seven
The Genetics of Ethnocentrism
1. Introduction
We have seen so far that ethnocentrism is more likely to occur, and more likely to be selected for, when two or more different ethnic groups are in conflict under conditions of natural selection or when the ecology is particularly harsh. We have also noted that the disposition to be ethnocentric is partly heritable and that it is a trait in itself, which cannot be explained simply in terms of personality dispositions. Accordingly, we would expect to be able to find genetically based mechanisms which would cause certain races to be more ethnocentric and, in addition, we would expect certain ethnic groups to be more ethnocentric than others even when controlling for environmental variables which might influence the extent of ethnocentric behaviour. We will explore these possibilities in the following chapters by taking factors which cause differences in ethnocentrism at the individual level and testing whether they are behind such differences at the group level.
However, in order to properly test whether or not certain factors are causing group differences in ethnocentrism we must find data on this subject. Having done this, this chapter, and those that follow it, will look at a number of candidate causes to explain population differences in ethnocentrism. Dutton et al. (2016a) conducted a literature review to find the key individual level causes of differences in ethnocentrism and their research will be drawn upon and expanded upon here. In this chapter, we will look at: (1) Direct genetic differences (2) Parasite Stress, an alternative model (3) Differences in Life History Strategy underpinned by genetic differences.
2. Measures of Ethnocentrism
For measures of ethnocentrism, Dutton et al. (2016a) drew upon the World Values Survey 6 (2010–2014) which surveyed 57 countries. For Positive Ethnocentrism they looked at percentage who ‘Would fight for your country’ (Fight for Country) and, in reverse, the percentage who were ‘Not at all proud of my nationality’ (No Pride) This was superior to looking at the positive responses because they were divided into subjective qualifications such as ‘very’ and ‘quite’. For negative ethnocentrism, they looked at the percentage who ‘Would not want as a neighbour’ ‘someone of a different race’ (No Other Race) or ‘an immigrant’ (No Foreigner). Obviously, it would be preferable to possess data on a far larger number of nations but, unfortunately, data from the World Values Survey was the best that they could find. As can be seen from Table 2, the measures of positive ethnocentrism strongly correlated in the expected direction, as did the measures of negative ethnocentrism, and they were statistically significant. However, the measures of the separate forms of ethnocentrism did not strongly correlate and were not statistically significant.
Table 2. Correlations between Positive and Negative Ethnocentrism (Dutton et al., 2016a).
Positive Ethnocentrism
Negative Ethnocentrism
No Foreigner
0.95
−0.193
N = 56
56
p = 0.001
p = 0.88
No Other Race
0.95
0.119
N = 56
56
p = 0.001
p = 0.41
Fight for Country
0.24
0.86
N = 56
56
p = 0.71
p = 0.001
No National Pride
0.13
−0.89
N = 56
56
p = 0.32
p = 0.001
Accordingly, they have what appears to be a sound measure of ethnocentrism with which they could attempt to test its causes. In the following, we will draw upon their research in order to test these causes.
3. Genes for Race Differences in Ethnocentrism
We will begin by looking at specific genes which might cause differences in ethnocentrism. And there is certainly good reason to suppose that there might be racial differences in ethnocentrism. Previous speculation on this has been evoked by supposed evidence of high ethnocentrism among Northeast Asians — and we test in the next chapter just how ethnocentric they are. There is, however, evidence from a variety of sources that Northeast Asian people are particularly ethnocentric, both positively and negatively. For example, Neuliep et al. (2001) conducted a survey with American and Japanese college students and found that on all measures the Japanese students were significantly more positively ethnocentric than were the American students and that, in both cases, the males were more ethnocentric than the females to roughly the same degree. Research with South Koreans and Chinese has garnered similar results. There is a large Chinese minority in Canada. A study of seventy-nine pupils in Toronto aged twelve to fourteen found that East Asian children were more positively ethnocentric than those of other races. There was a non-significant tendency for students to be more ethnocentric in their choice of best than in their choice of other friends. The East Asian participants rated their friendships with in-group members as being of higher quality than those with out-group members and this was a significant difference from other races. It was not true for the Anglo-European, West Indian or South Asian groups (Smith & Schneidner, 2000). Other studies with second generation immigrants have also pointed to a higher level of ethnocentrism among Northeast Asians. Stephan and Stephan (1989) found that Asian-Americans were significantly more negatively ethnocentric than Hispanic Americans and became significantly more anxious at the prospect of having to interact with white Americans than did Hispanic Americans. Li and Lui (1975) found that Taiwanese students in the USA were significantly higher scoring on all aspects of ethnocentrism than white American students. Accordingly, there would appear to be sound evidence that Northeast Asians are more ethnocentric than are Europeans, even when they are raised in multi-ethnic Western societies. The fact that this is noticeable even among Northeast Asian children in these societies would imply that it is likely to have a partially genetic nature, something substantiated by the more general evidence that ethnocentrism is partially genetic.
Candidate genes have been suggested for this, though we need to be relatively cautious of candidate gene literature as it is riddled with false positives and failure to replicate. De Dreu et al. (2010) hypothesized that ethnocentrism may be modulated by brain oxytocin, a peptide which has been shown to promote cooperation among in-group members. In double-blind, placebo-controlled designs, males self-administered oxytocin or placebo and privately performed computer-guided tasks to gauge different manifestations of ethnocentric in-group favouritism as well as out-group hostility. Results showed that oxytocin creates intergroup bias by motivating in-group favouritism and, to a lesser extent, out-group hostility. For example, in a computer war game subjects were more likely to self-sacrifice to save an in-group member and were more likely to permit the sacrifice of an out-group member if they had been injected with oxytocin. Oxytocin has previously been referred to as the ‘cuddle chemical’ as it makes people more affectionate. However, De Dreau et al.’s research demonstrates that it clearly has a darker side and that this relates to ethnocentrism.
Their research also means that there is at least one clear physical basis for differences in ethnocentrism: the degree of oxytocin that is transported and the strength of stimuli required to transport it. As such, we would expect Northeast Asians to disproportionately possess the short form of any gene relating to oxytocin, making them highly sensitive to it. In this regard, it has been reported that A118G (OPRM1) is a genetic basis of the fear of social exclusion. G and A polymorphisms in this gene regulate μ-opioid receptors. A study showed that subjects with the G allele showed stronger unpleasant feelings (based on fMRI) when they were excluded in ball-toss games (Way et al., 2009). Furthermore, Way and Lieberman (2010) found a positive correlation between the frequencies of G alleles in a population and the collectivism of the culture. They also reported that the G allele frequencies among East Asian populations are in fact much higher than those in
European populations. Also, the G allele in rhesus macaques has been reported to strengthen mother-infant attachment and to be associated with higher oxytocin levels when lactating (Barr et al., 2008; Higham, et al., 2011). Thus, there is indirect evidence that Northeast Asians are more sensitive to oxytocin and that this, indeed, causes their cultures to be more collectivist and more ethnocentric than European cultures.
Pursuing this line of research, Cheon et al. (2014) reported that the serotonin transporter gene polymorphism (5-HTTLPR) has been associated with individual variations in sensitivity to context, particularly with regard to stressful and threatening situations. The authors examined how 5-HTTLPR and environmental factors signalling potential out-group threat interacted to shape in-group bias. Across two studies, they provided evidence for a gene-environment interaction on the acquisition of intergroup bias and prejudice. Greater exposure to signals of out-group threat, such as negative prior contact with out-groups and perceived danger from the social environment, were more predictive of intergroup bias among participants possessing at least one short allele (vs. two long alleles) of 5-HTTLPR. Moreover, this gene-environment interaction was observed for biases directed at diverse ethnic and arbitrarily defined out-groups across measures reflecting intergroup biases in evaluation and discriminatory behaviour. Accordingly, their research presents us with a candidate genetic mechanism for ethnocentrism. There would appear to be a biological mechanism, which means that evidence of out-group threat is more likely to lead to ethnocentrism of both kinds; that is, in-group bias and out-group prejudice. Further research on this topic found that 70–80% of an East Asian sample carried the short form of this gene, that is to say the form that makes you more ethnocentric. Only 40–45% of Europeans in the sample carried the short form of the gene. Indeed, it was found that across twenty-nine nations, the more collectivist a culture was the more likely it was to have the short form as the prevalent allele in the population (Chiao & Blizinsky, 2009).
Dutton et al. (2016a) tested the genetics of ethnocentrism using their own measures. They tested the association between 5HTTLPR and the measures of ethnocentrism. In this instance, they had a sample of twenty-five countries. They found that the correlations, which were generally weak, were nowhere close to significant. This would seem to highlight the problem with using ‘collectivism’ as a proxy for ethnocentrism. It involves some of the same dimensions as ethnocentrism, but is conceptually highly distinct. So, group differences in ethnocentrism do not appear to be strongly explicable in terms of this gene, though it may be germane to some extent.
4. Parasite Stress
But why exactly is there racial variation in these ethnocentrism-causing genes? There are two main evolutionary possibilities for explaining high levels of ethnocentrism at the group level. The first we will examine is the so-called ‘Parasite Stress Model’, though we will highlight a number of serious problems with it.
There is evidence that parasite stress is a factor in differential levels of ethnocentrism, especially negative ethnocentrism. Parasites are generally highly detrimental to health, meaning that in an ecology with a high parasite load, people will tend to differentially select as mates those who evidence a high level of health and disease resistance. However, we would also expect different forms of behaviour to be selected for when comparing environments of high and low parasite stress. In this regard, American evolutionary biologist Randy Thornhill and colleagues (2009) looked at parasite stress prevalence across countries based on twenty-two significant human diseases. They hypothesized and found that collectivism, autocracy, women’s subordination relative to men’s status, and women’s sexual restrictiveness positively co-varied, and that they corresponded with a high prevalence of infectious disease. They further found that these values were linked to xenophobia and to ethnocentrism.
In effect, they argued that in a context of high parasite stress it would be sensible to avoid those whom you do not know — as they may carry new parasites to which you are not immune — and associate only with those whom you do know, as doing so has allowed you successfully to avoid parasites in the past. Clearly, even if the mechanism for the spread of disease — via pathogens — is not then known, nevertheless, genes which cause you to shun outsiders and focus strongly on insiders will become highly advantageous to survival in a context of high parasite stress. Accordingly, they will gradually spread throughout the population, rendering areas that are high (or have until recently been high) in parasite stress more ethnocentric than those which are lower in it, at least in general. Thornhill et al. found that, overall, less collectivist cultures tend to be at higher latitudes, meaning less parasite stress, and also that industrialism, and the consequent inoculation of the population against parasites, tended to reduce collectivist and thus ethnocentric behaviour. This being the case, we would expect that the length of time since a country industrialized would be negatively associated with ethnocentrism when other factors are controlled for and this is in fact what they find.
In another study, Fincher and Thornbill (2012) argued that strong family ties and heightened religiosity were both a reflection of parasite stress. They found that both within the USA and also between nations there was a positive association between parasite stress and heightened in-group assortative sociality and heightened religious commitment. The simplest explanation, they argued, is that in areas of very high parasite stress it pays to avoid strangers who may carry novel parasites into the community. These findings may even help to explain the relatively high levels of consanguineous marriage prevalent in Middle Eastern countries. Indeed, Thornhill and Fincher (2014, p. 334) note that ethnocentrism is comprised, partly, of ‘nuclear and extended family nepotism’ and ‘cooperation with in-group non-family members with the same values and immunity’. This propensity for cousin marriage may even then be passed on as both a cultural and genetic legacy to future generations, though whether this is indeed the case remains speculative. As for the finding in relation to religiousness, in 137 countries religious belief and religious participation were positively correlated at 0.67 and these correlated with parasite stress at 0.4–0.64 (Fincher & Thornhill, 2012). As such, parasite stress is seemingly relevant to understanding religious differences worldwide but it is far from the only factor.
5. Problems with the Parasite Stress Model
However, there are a number of difficulties with the parasite stress model as a total explanation and, accordingly, it is not one I would be inclined at accept in this study.
Firstly, Thornhill and Fincher’s theory may be criticized for overemphasizing the importance of parasite stress to understanding human behaviour and especially to ethnocentrism. For example, we have already observed that in environments of relatively low parasite stress high levels of ethnocentrism can seemingly evolve for quite different reasons. Thornhill and Fincher specifically argue that low ethnocentrism involves prioritizing the nuclear family but otherwise caring relatively little about the broader, extended family; instead concentrating on more general alliances with non-relatives. They predict, and find, that areas that are high in parasite stress are, therefore, less likely to be democratic — as this involves trust of strangers — and are more likely to be autocratic (p. 334). In addition, the strong boundaries that will be produced by highly ethnocentric cultures will lead to high levels of civil war, as well as high levels of cultural and linguistic diversity within a given country (p. 335). As we have already discussed, genetic homogeneity is also a factor behind low democracy levels and ethnic conflict (Vanhanen, 2012). Further, low national intelligence has also been shown to be associated with low levels of democracy and high levels of political instability (e.g. Lynn & Vanhanen, 2012). This may be because less intelligent societies are less trusting, less cooperative (these both being associated with intelligence), less organized, have less self-control (necessary for democracy to function) and lack the necessary foresight to notice any decline into dictatorship before it is too late (Vanhanen,
2012). In drawing upon the parasite stress model, it is important not to reduce everything down to it. We need to appreciate that ethnocentrism can be explained by a variety of non-exclusive models of which parasite stress is only one. A superior model would be able to explain ethnocentrism variation in all cases.
Secondly, further difficulties with the parasite stress model also arise when attempts are made to extend it into national differences in modal personality. Fincher et al. (2010) distinguish between ‘non-zoonotic’ and ‘zoonotic’ parasites. Non-zoonotic parasites have the capacity for human-to-human transmission while zoonotic parasites do not. The research group found that it was specifically non-zoonotic parasite stress that was associated with collectivism and thus, by implication, ethnocentrism. However, they also averred that parasite stress was negatively associated with certain personality characteristics, especially extraversion. Likewise, Murray et al. (2013) have argued that high national levels of ‘authoritarian personality’ are associated with higher levels of parasite stress. The fundamental problem with these findings is that the quality of data comparing different countries on the Big 5 personality traits is very poor. It often involves small and incomparable samples and leads to extremely counter-intuitive findings, such as that the Japanese are lower in Conscientiousness than Sub-Saharan African countries (e.g. Schmitt et al., 2007). Moreover, serious questions must be raised over whether it is even possible to compare different nations on these measures (see Meisenberg, 2015). The personality surveys involve people subjectively evaluating the degree to which they are ‘tidy’, for example. But they will be making this evaluation according to different cultural norms of how tidy ‘tidy’ really is, meaning it is extremely problematic to compare different cultures in this way. Equally, as German psychologist Gerhard Meisenberg (2015) has also noted, national differences exist in how likely people are to opt for the most extreme of the numbered possible options in a survey. Also, many of the surveys rely on psychology students. Not only are these not representative of the population but they cannot be assumed to be approximately comparable across populations in terms of the way in which their personality differs from the norm.
Race Differences in Ethnocentrism Page 13