Unity in diversity, last response to John Wilkins

John put up his last thoughts on race, and Evolgen chimed in with his ruminations.

First, nice exchange. Quick points....

1) I'm not hung up on a word. If you want to agree on another word that captures what I'm trying to say, I'm willing to go along with it.

2) One key point I want to make is that we should be cautious about relying on Lewontin's 85% intragroup vs. 15% intergroup data. Evolution and genetics are sciences of some subtly, and one can not draw a straight line between statistical data and pithy verbal conclusions. Context and framing matters a lot. What exactly does "sharing 99% of our genome with chimpanzees" mean? What does it mean to say that "most molecular evolution is neutral or nearly neutral." These are issues that require comment, and I think that commentary has been sorely lacking when it comes to conflations of partitioning of variation between and within groups on just one single locus. I find this somewhat ironic, because on a technical level I believe Lewontin and Elliott Sober coauthored a paper in the early 1980s urging caution (warranted) in regards to generalizing models of selection on one locus to multiple loci. In my mind this does not mean that one locus models are worthless, nor does Lewontin's insight in regards to variance on one locus need to be discarded, but it needs to be put into its proper context.

3) In terms of context, I think it is probably true that being at least one standard deviation above the median in intelligence (about the top 15%) is necessary to understand some of the issues here, so I am not averse to the likelihood of both John and Evolgen's contentions that the public will misunderstand human population substructure. The key I think is to mold the discourse among the elite, because I believe elites drive history. 40 years ago someone with my complexion would no doubt be a "nigger" in the eyes of the majority of Americans. I have seen memory research which suggests that many people who don't express racist opinions today don't recall expressing them in the past, even though researchers tracked these people between 1970 and 1983, and did note that their opinions tracked social changes. In other words, the human mind is a beautiful and manipulable object.

The key for me is that at least amongst those who the intuition of probability distributions is easy to grasp there is less need for caution. I will grant philosophers like John or other sorts who are one step closer to the nexus of public policy the task of mediating, interpreting and shaping how science maps onto the everyday life of a typical human being. My point is that the underlying science does show evidence of substructure that is not trivial. I would also contend that roughly speaking the substructure follows many common intuitions, though not all of them.

4) This last is a key point. There are two separate issues here which need to be decomposed. Ancestry and morphology. A few years ago anthropologist Henry Harpending sent me the following email, where he said:

Mitochondrial DNA paints an extreme picture of greater African diversity. It is just one locus and hence expected to be unreliable. Large collections of repeat polymorphisms show that there is a diversity cline away from Africa so that Japanese are about 15% less diverse than Africans, Amerindians maybe 25% less diverse. These are all probably neutral and hence passive indicators of population history.

Khoisan people are, according to their neutral genes, just generic Africans, but they sure don't look anything like [that] to their neighbors. They also have no 7R alleles at the dopamine D4 receptor locus while the frequency in their neighbors is 20%. There are a lot of indications in the genetics literature that a lot of selection is going on in our species, and I expect that the differences in appearance and in DRD4 and in other behaviorally significant loci are maintained by selection in the face of gene flow that has homogenized the neutral genome.

Henry's point, and one I am aware of, is that phylogeny and morphology need not always track. Melanesians are a classic example, though they differ from sub-Saharan Africans in many ways, they "resemble" them to the eyes of many outsiders. In fact, those populations which exhibit functional constraint on the MC1R locus for the ancestral consensus allele which confers dark skin, Africans, South Asians, Papuans, Melanesians, Australian Aboriginals and other dark skinned groups are not necessarily phylogenetically particularly close.

This is where the standard mtDNA narratives come in. And Evolgen points to this when he refers to Africans being more diverse than other groups. I would offer that one should be sensitive to loci here, as I noted, Europeans exhibit extreme polymorphism on MC1R, while Africans are not diverse here. The standard model is that mtDNA is "neutral," so the diversity Evolgen speaks of is one of is not functionally relevant, it is "junk." The reality is that I don't think it is all that neutral, but that is for a different day. On a preponderance of the genome I willing to agree that I suspect it will be found out that Africans are more diverse, in particular on the neutral loci (those that are not selectively beneficial). I have offered one idea why this might be so, larger long term effective population. Another, more common, idea is that Eurasians are a subset of Africans who emigrated about 50,000 years ago. But, I would offer that on the functionally relevant loci Eurasians exhibit their own unique polymorphisms and are not necessarily a subset of Africans. In fact, there is evidence for selective sweeps once humans left Africa, which makes sense since different environments (social and physical) would result in different pressures.

Sum: The key, for me, is that people tend to view populations as unitary lumps. In fact, they tend to view people as unitary lumps. Fundamentally we are all packets of quanta, but on the genetic level, we are collections of alleles which are trapped within bodies. Populations, clusters of individuals bodies, are a higher order structure which is riven with the flux of these alleles as they percolate through the system. If we abandon our pre-population thinking biases the "race" questions is far more tractable on the scientific level. As long as we point to the same entity, I don't care if you call it red or blue :)

On a social level, we need to prep the public for the coming postgenomic revolution. There are many reasons for this, I suspect that genomic data will be relevant to insurance companies, for example. But on a more idealistic and less concrete plane ancestry tests are already playing havoc with public intuition. The HapMap, which is plumbing the depths of variation amongst hundreds of individuals from 3 large population clusters is only the prologue when it comes to creating a detailed map of our genomic diversity. John and Evolgen both expressed the worry that these data, or at least entertaining the idea of race, is dangerous. But it doesn't matter, the data is out there in the public domain, and people are already flying off with it, we need to control the discourse now. On my other weblog I have noticed a trend where referrals come in from white racialist websites when it comes to topics like the HapMap. Unless you want a blanket moratorium on this data the usual suspects will mine it for what they want. How data maps onto public policy is still up in the air, but I think these skirmishes about "is there a race" are in the end irrelevant, we need to start figuring out how we are going to the ride the tigers into our species' future. Lewontin's 85/15 figure is correct, for one locus, but you can't pretend that substructure is irrelevant and trivial when people can go to PUBMED and find a paper which suggests that genetic interactions between European modal and African modal alleles in African Americans (who are admixed to some extent) might increase the risk of death.

Addendum by analogy: In the early 7th century the Persians and Byzantines engaged in what were by the standards of the day "World Wars." The Persians expanded and drove the Byzantines back to the point where they had recreated the empire of the Achaemenids. Under the emperor Heraclius the Byzantines fought back by engaging in a daring end around by attacking the Persian heartland while the main armies of the enemy were engaged to the west at the gates of Constantinople. To the participants in these events they were of great import. But the realty was that in the mid-7th century Islam swept Persia aside and rendered Byzantium and empire in name only. The data is going to be our Islam, it will conquer not with the sword but by the force of its own reality.

Tags

More like this

Voight et al addresses the question of comparing the number of loci showing evidence of selective sweeps in the three HapMap populations

In this paper we provide the first genome-wide map of incomplete selective sweeps in humans. We find widespread signals of selection in all three populations. These selective events are generally very recent, falling mainly within the Holocene era, and substantially postdating the separation of the three populations. Selective sweeps in Yoruba tend to be narrower and apparently older than in the non-African populations, perhaps explaining why previous low-resolution scans for selection have reported a deficit of selective events in African populations [13-16]. (Two of those studies also used African-American samples, in which it is possible that European admixture may dilute evidence for selection in Africans.) Though most selected regions are not shared across populations, there is still a clear excess of shared selective events. Indeed, since we have incomplete power to detect selection, it is likely that we tend to underestimate the degree of sharing across populations.

http://biology.plosjournals.org/perlserv/?request=get-document&doi=10.1…

By Rikurzhen (not verified) on 24 Apr 2006 #permalink

Your second-last paragraph is a really important point that I've grumbled about whenever I see people rasing the spectre of racism every time the subject of genetically mediated human differences comes up. They don't seem to understand that if they actually succeed in their goal of stigmatizing anyone who broaches the subject, the only people who are going to discuss it openly are people who don't give a damn about stigmatization. Much like the the NRA's favourite saying about guns, if talk of "race" is racist, then the only people who will talk about it will be racists. The hardline "there is no such thing as race" folk think they're fighting the good fight, but in actuality are practicing unilateral intellectual disarmament.

I think this is a rather more urgent problem in American discourse than elsewhere, although it is a problem in many countries.

But say rather "Race doesn't mean what you think it means" than "Race is a biological fact". That is too dangerous.

I think what got lost in your discussion was my attempt to distinguish between neutral loci (good for elucidating population structure) and non-neutral loci (good for finding out what makes those populations biological different). Yes, I am squeamish about saying one of those populations is smarter than another due to some set of alleles (although I also don't think intelligence can be measure by a single quotient -- IQ measures something, it doesn't measure everything, and I'm just not sure what it measures exactly). But I do admit that we can identify functional differences between our populations.

Because the races currently tossed around in public discourse are based on a few morphological characteristics (skin color, facial structure, hair, etc), it would be most interesting to first reconstruct the evolutionary relationships of the different populations (using multiple neutral loci), and then overlay the changes that occurred within and between populations on top of that genealogy. Jumping right in using loci under selection (eg, MHC/HLA) can tell you that some populations are functionally different from others, but it doesn't tell you how those differences evolved.

Insurance companies, heh:) We can safely bet that it'll be rich men who own stock in genomic-technology corporations who'll benefit most in the postgenomic era. What it will mean for the average joe, not sure. Science becomes money; Science IS money.