Sunday, August 14, 2005

Interracial Marriage   posted by DavidB @ 8/14/2005 12:50:00 PM

I see that one of my posts from earlier this year is being quoted on White Nationalist sites as ‘sophistry’ aimed at encouraging interracial marriage, or ‘miscegenation’, as they charmingly call it.

I suppose it is an honour to be criticised in such quarters, but I don’t think I deserve it. My post (here) expressed no opinion either for or against interracial marriage. I merely pointed out a fallacy in an argument sometimes used against it. I don’t think anyone has answered my criticism, other than by the usual tactic of changing the subject.

But for what it is worth, I don’t see any strong biological reasons either for or against interracial marriage in general.

To expand on that, we can consider the biological implications at both individual and population level…

Individual level

At individual level the key consideration is the effect on the offspring. There are theoretical arguments on both sides.

Against interracial marriage (or other interracial mating), it may be said that:

a. the main races [see note 1] have evolved over thousands of years to be adapted to a specific geographical environment, whereas their offspring will not be well-adapted to any environment, and

b. the genes at different loci in each major race have evolved to be adapted to each other, forming ‘co-adapted gene-complexes’. Interracial mating would break up these gene-complexes.

For interracial mating it may be said:

a. Mating between different races of animals or plants within the same species often results in ‘hybrid vigour’: the offspring are superior to either of the parents (or at least to their mean). The main reason for this is probably that it reduces the harmful effects of inbreeding. As human racial and ethnic groups are to some extent inbred, then some degree of hybrid vigour might be expected when different groups interbreed.

b. As a more specific benefit, the offspring of interracial mating are likely to avoid genetic diseases arising from homozygous harmful recessive genes found mainly in particular groups, such as cystic fibrosis in Europeans and sickle-cell anaemia in West Africans.

The last of these points is the only one that can be asserted with any confidence. The others are more speculative, and it is a matter for research whether they are significant or not. The subject has not been studied very fully. On searching I found only three areas where mixed-race offspring are said to have some health problems:

a. it is more difficult for mixed-race individuals to find genetic matches if they need tissue or organ donors.

b. in a study of American students, individuals who identified themselves as mixed-race were reported to have higher than average levels of stress and stress-related health problems. Of course, there is no evidence that this is a genetic effect.

c. according to one study (link here), mixed-race offspring may have a moderately higher risk of certain birth defects, but a moderately lower risk of others.

On the positive side, it is sometimes suggested (e.g. by Armand Marie Leroi) that mixed-race offspring are physically fitter or better-looking than the average of their parents. From personal observation I find this plausible, but I don’t know of any statistically valid evidence: it is easy to point to individual cases like Halle Berry or Thandie Newton, but this doesn’t prove much. (Razib recently commented on a study which provides some of the kind of evidence needed.)

It is possible that with more research further drawbacks or advantages to interracial mating will emerge. Just as each ‘pure’ group has its own distinctive profile of genetic risks, we might expect each ‘mixed’ group to have its own special features. But on common-sense grounds these are unlikely to be very important. The reason is simply that the existing races are already genetically variable. If there were any ‘co-adapted gene complexes’, they would be broken up by mating within races almost as frequently as in mating between races. As for being ‘adapted to the environment’, the differences between human races are comparatively recent and superficial in evolutionary terms. This is true whether we adopt a pure Out-of-Africa model or some mixture of Out-of-Africa and multiregionalism. In the latter case it would still be true that races as they exist now are comparatively recent. It is also likely that some of the differences between races are not adaptive at all, but the result of genetic drift at a time when populations were much smaller and more isolated from each other than today. And some of the differences that were adaptive in Paleolithic and Ice Age conditions will no longer be adaptive now.

Population level

Turning to the pros and cons at population level, the first key point is that interracial mating in itself makes no difference to gene frequencies in the total population. Any trait that depends on the additive effect of genes will therefore not be affected so far as its mean value in the combined population is concerned. Interracial mating does however affect the amount of diversity within and between populations, as measured by average heterozygosity (though, as I pointed out here, this is not always a very useful measure). If we consider the fanciful case of global panmixia, where the entire human population mates entirely at random, then races within the human population would cease to exist. The resulting level of diversity in the global population would be lower than at present, because between-group variance would be eliminated, but the remaining diversity would be greater than in any of the present races. For quantitative characters such as height or IQ (assuming additive inheritance), the genotypic mean would be that of the present world population, while the standard deviation would be somewhat greater than in existing populations. Effects due to dominance and epistasis are of course more difficult to predict.

What would undeniably be lost would be the clustering of genes and phenotypic traits which at present makes it possible to identify individuals as belonging to a particular race, despite the overlap in traits between races. I cannot see any biological reasons for objecting to this, other than the dubious ‘co-adapted gene complex’ idea discussed above. From a purely aesthetic point of view, I would regret the disappearance of such fine types as the Nuer or the Icelander. From a scientific point of view, I would be sorry to lose such curiosities as the Andaman Islanders. More generally, I think it would be undesirable to make any large and sudden change in the genetic makeup of the human species, on the prudential ground that we can’t predict all the consequences [see note 2]. ‘Global panmixia’ would be difficult to reverse [note 3]. But it would also be unreasonable to resist all change.

In any event, global panmixia is hardly an imminent prospect. In much of the world (e.g. Asia, Africa, Eastern Europe), there is little interracial or even inter-ethnic mating. Significant rates of interracial mating occur only in rather special circumstances, where different racial groups are present in large numbers and have few barriers of language, religion or custom between them. Like it or not, in most localities the distinctions between racial groups will not disappear in the foreseeable future.

In the longer term, it is not clear whether interbreeding will eventually lead to the disappearance of geographical races. In so far as the existing racial distinctions are due to genetic drift in small Paleolithic populations, or to obsolete selective factors, then they cannot be expected to persist in the face of even modest levels of interbreeding. If on the other hand differences are due to significant differences in selective pressures (including sexual selection) in geographical regions, then they may persist indefinitely in the form of clines. The existing pattern of geographical variation is already clinal, except where there has been recent inter-continental migration. The amount of gene flow between major races at global level (e.g. between East Asians and Africans) due to interbreeding must be less than 1 per cent per generation, so even quite modest levels of counter-selection would suffice to maintain some genetic differentiation. In any event, even if gene flow is not resisted by selection, at present levels of interbreeding it would take several millennia for all distinctions to be erased. During this time scale other events affecting the human gene pool are likely to supervene, so speculating about the distant future of racial differences seems pointless.

Finally, I should stress that I have only been discussing biological, and not social or political, implications.

Note 1: I use the term ‘race’ for convenience, though it has no generally accepted meaning. It should not be taken as implying that human ‘races’ are or ever have been sharply defined and distinct entities. To illustrate the difficulty, consider the case of South Asia, which contains over one fifth of the world’s population - about the same as that of Africa, Australasia, and the Americas combined. Do the inhabitants of South Asia belong to a single ‘race’, or several? Do South Indians belong to the same ‘race’ as North Indians and Pakistanis? Do South Asians (or any of them) belong to the same ‘race’ as major populations elsewhere? Genetically they may be closer to Europeans and other ‘Caucasoids’ than to East Asians (but see this paper by Jorde and Wooding for some more complex evidence), but should they be grouped in one ‘Caucasoid’ race with Europeans (and North Africans) or are they sufficiently different to be regarded as a ‘race’ in their own right?

Note 2: It is conceivable, for example, that the distinctive achievements of European culture have depended on the prevalence of certain combinations of intelligence and personality which in turn depended on gene-combinations that are more common among Europeans than elsewhere, and which would become rarer if the European gene-pool were to merge with, say, the Chinese gene-pool. I don’t know of any serious evidence for this, but I don’t think it can be dismissed as absurd.

Note 3: I was going to say impossible to reverse, but that cannot be strictly true. Heredity is particulate, not blended. Reversing the effects of panmixia would be like unshuffling a pack of cards, not like unscrambling an egg. Assuming that the differences between races that are of practical interest (such as appearance, personality, and cognitive ability) involve differences of allele frequency at no more than a few hundred loci, I guesstimate that it would take between 10 and 20 generations of careful breeding to recreate populations with allele frequencies similar to the original major races with respect to these loci. The more loci, the longer it would take.