World turned upside down!
A site with the URL Anthropology.net puts up a post titled Fighting the mantra, “People vary more within the groups than vary between groups”? What’s going on here! Since Lewontin’s Fallacy is one of the axioms at the heart of much of modern anthropology (that is, when anthropologists bother to accept the validity of linear-logo-centric analytic frames), I think someone might have their card revoked soon. Be afraid! Of course, we know what type of person runs Anthropology.net….
Readers who want to dig deeper into the nature of the pretty charts in the papers that the apostate anthropologist mentions above should check out Inference of Population Structure Using Multilocus Genotype Data: Linked Loci and Correlated Allele Frequencies, an early paper introducing STRUCTURE. K‘s in the hiz houze, as they say.
Update: Some concern in the comments that I’m painting with too broad a brush in regards to anthropologists. Obviously since I am avid follower of the work of people such as John Hawks, Henry Harpending and Heather Norton, who are all trained as anthropologists, I don’t think it’s all crap. That being said, physical anthropologists are anthropologists in the same way that a follower of Milton Friedman is a liberals. There are historical and definitional reasons to call oneself a liberal in such a case, but quite often it simply results in confusion because of other definitions in circulation (context matters here, since the older definition of liberal is still in currency in much of the rest of the world). In this case, there is Anthropology which deals in Theory and Ways of Knowing, and anthropology which is driven by data, models and analysis.
Of course, this isn’t a total dichotomy. There are cultural anthropologists who attempt to work within a scientific framework, general hypotheses, see how the data fits the inferences, and so forth. But they seem a very small minority at this point.
Labels: human biodiversity
The fallacy lies not so much in the data (which do show more variation within than between populations for gene loci taken individually) but in the way people interpret this finding.
blogspot.com/2008/01/85-truism.html
I addressed this point in one of my recent posts:
http://evoandproud.
I think you are a little hard on anthropologists writ large, Razib. After all, it is anthropologists that are responsible for much of what we know of human genetic and phenotypic diversity. You should specify “social anthropologist” in your critique.
jason, numerically i think biological anthropologists are so outnumbered. what’s your perception of the ratio?
Within anthropology I would guess biological anthropologists are outnumbered by cultural anthropologists probably 5 to 1. However, I think the biological anthropologists overwhelmingly outproduce social anthropologists in terms of research output.
I often feel like anthropology has been hijacked by the radical social anthropologists, and it really hurts the reputation of the field.
The whole debate on whether race exists is robbed of its middleground by cultural anthropologists. Their denial of significant genetic differences between groups is easy to debunk. What’s needed is not blank-slatism, but a taxonomic discussion about whether the differences between human populations are large enough to demonstrate the existence of subspecies, and if so, what specifically these subspecies are.
What’s needed is not blank-slatism, but a taxonomic discussion about whether the differences between human populations are large enough to demonstrate the existence of subspecies
Angels on pinheads.
What people really want to know (or make unknown) is the extent to which ethnic differences in socially valued traits have a genetic basis.
All the race fallacies that were forcefed into folk wisdom are predicated in trying to moot that such differences in socially valued traits could stem from genetics. (and more specifically, almost all the nurture-nature controversy (created entirely by the nurturists), exists because of the black-white IQ issue.)
Lewontin’s Fallacy is a prime example, and, indeed, Lewontin has asserted erroneously on many occasions that the 85/15 statistic demonstrates that socially valued traits can’t differ for genetic reasons. That was the implied conclusion of the original paper.
The subspecies thing is another misdirection, and often explicitly or implicitly paired with the same conclusion: differences in socially valued traits cannot plausibly exist unless humankind can be partitioned neatly into a mythical conception of ‘subspecies’.
The cline vs. cluster ‘debate’ is another misdirection. None of these issues have much if any implications for the grand question, even though they are not so subtly (and many times explicitly) argued as if they do.
What people really want to know (or make unknown) is the extent to which ethnic differences in socially valued traits have a genetic basis.
I think people are interested in that, but they are also very interested in what the major genetic divisions in the human race are.
All the race fallacies that were forcefed into folk wisdom are predicated in trying to moot that such differences in socially valued traits could stem from genetics.
I’d say that they’re more predicated on trying to oppose the older folk wisdom– racism. I agree with you, however, that fallacies here should be avoided, and that the blank-slatists are making way more fallacies than the race realists.
As you point out, Lewontin is a good example of this. He lets his Marxist views get in the way of his science.
The subspecies thing is another misdirection, and often explicitly or implicitly paired with the same conclusion: differences in socially valued traits cannot plausibly exist unless humankind can be partitioned neatly into a mythical conception of ‘subspecies’.
The subspecies thing is not a misdirection unless we’re single-mindedly focused on group differences in social traits. it’s a notable question, just as the nature of group genetic differences is a notable question. Also, subspecies is not a mythical concept if handled correctly.
OK, I’m not getting this.
IIUC, Lewontin says that the additional variance induced by continental divisions is small with regard to inter-personal variance: between-groups variance is “only just” larger than within-groups variance.
OTOH, the guys at anthropology.net point out that if you look at enough loci, you can build a reliable statistical estimator of someone’s origins.
I fail to see where the contradiction is.
The question is how big the within-group and between-group heritable _phenotypic_ differences are, and these genetic stats don’t tell you. Much of the between-group genetic differences are the product of recent selection, which means that they changed rapidly, and they couldn’t have changed rapidly unless their selective advantages were in the few-percent range. Recent sweeping alleles have large effects.
And the importance of recent selection varies from trait to trait. So, if you look at skin color, between-group differences are _bigger_ than within-group differences. In principle, and sometimes in practice, differences in a single gene that’s gone to fixation locally (and a number have) will dominate trait variation.
The only way to know the relative effects of the within-group and between-group genetic variation on a particular trait is to look at that trait: overall genetic statistics tell you nothing. Within-group genetic variation in dog breeds is more than twice as big as between-group variation – so does this mean that the differences in size between individual dachshunds are bigger than the average difference in size between dachshunds and Great Dane? No.
In that case most of the difference in size boils down to a single mutation, one with a large effect.
Or you could look at the cichlid fishes of lake Victoria: they’ve evolved into many different species in a short time and you can hardly detect _any_ genetic variation. Presumably the variation that counts is in a few developmental/regulatory genes; but those cichlids are just as different as they look, just as different as they act.
The subspecies thing is not a misdirection unless we’re single-mindedly focused on group differences in social traits.
I didn’t mean to imply you were using the question this way (and I didn’t think so), just that it often was, especially by most scientists arguing for and against it. The reason for this is apparently because scientists themselves are unclear about what significance to take from or make out of ‘sub-species’.
By ‘angels on pinheads’ I did mean to say that ‘sub-species’ is not a concept or classification with great importance or great theoretical standardization or rigor.
Scientists have never stopped piecing together phylogenetic relationships, which is important, but not predicated on such classification.
Think about it like this: when you ask if Europeans are a human ‘sub-species’, what information, more precisely, are you trying to gain?
More would be gained by asking about that information directly.
The difference between ‘race’ and ‘subspecies’ is that the latter contains a fallacious judgment about ‘differentiatedness’ not standardized or justified by scientists. Perhaps it should be applied to obviously incipient species headed towards allopatric speciation, but it isn’t in practice. (and by that definition, by the way, there are no human “sub-species,” AFAICT) Many times it is simply based on phenotypic dissimilarity or marked regional differences, but that isn’t necessarily (or even most commonly) the result of speciation or extreme isolation.
A race, on the other hand, is a convenience grouping of phylogenetically related organisms below the species level that may or may not share specific traits.
The subspecies thing is another misdirection, and often explicitly or implicitly paired with the same conclusion: differences in socially valued traits cannot plausibly exist unless humankind can be partitioned neatly into a mythical conception of ‘subspecies’.
Because, after all, it’s not as if there is no variance in IQ among caucasians, is there?
“Them there ‘tards must be a different subspecies!”
Jason, you link to Goodrum’s race faq, but if you re-read that faq you’d see that “race” and “subspecies” are essentially the same thing. The only reason we say “race” is because of human exceptionalism– thinking that we can be grouped in a different way than other species:
The terms ?race? and ?subspecies? are most often used synonymously [1,2] although the former is normally used when talking about human populations. When a distinction is made, ?race? generally implies a lower level of differentiation, but because this term is not commonly used in the recent non-human literature, ?race? and ?subspecies? are used interchangeably throughout this paper.
Much of the debate over the existence of human races stems from how one chooses to define ?race? (or ?subspecies?).
Also, I think it’s worth noting that the race/subspecies debate is quite relevant to the group genetic differences debate. If we discover that human variation is best summarized by dividing the human race into groups A, B, and C, then our hypotheses of group genetic differences should make claims about A, B, and C, not everyday racial categories.
also, i agree with gcochran that phenotypic differences are much more important to the group differences debate than genotype (which will tell us more about our natural history than anything else). so it would probably make sense to see how the variance in phenotype is distributed throughout the globe and consider what the major groupings are when we make our hypotheses about group differences.
Jason, you link to Goodrum’s race faq, but if you re-read that faq you’d see that “race” and “subspecies” are essentially the same thing. The only reason we say “race” is because of human exceptionalism– thinking that we can be grouped in a different way than other species:
I don’t know why you are asking me to “re-read” my own link, when my purpose for linking was precisely to show that I believe some of its ideas are flawed!
No, animals are, of course, commonly grouped by ‘race’. And, no, ‘subspecies’ and ‘race’ are not used interchangeably. ‘Race’ is used interchangeably in the literature to mean ‘subspecies’ and ‘population’, but ‘subspecies’ is not used interchangeably to mean ‘population’. ‘Population’ is used to designate any group of phylogentically related organisms below the species level, while ‘subspecies’ is used to designate some grouping below the species level arbitrarily (and often problematically) deemed to be so phenotypically or ecologically differentiated that it deserves a special classification and trinominal name.
If ‘subspecies’ and ‘population’ were used interchangeably, then there would be nearly as many trinominal names per species as there are individuals within that species.
I don’t see much use in ‘subspecies’, unless the concept is corrected to deliver some specific, agreed upon, and relevant information about a population within a species.
‘Race’ is useful in the neo-Darwinian era only as a synonym for ‘population’. Dropping it would be fine, if so many weren’t trying to use it as a trojan horse to deny the latter. (Note Watson never used the word ‘race’!)
Also, I think it’s worth noting that the race/subspecies debate is quite relevant to the group genetic differences debate. If we discover that human variation is best summarized by dividing the human race into groups A, B, and C, then our hypotheses of group genetic differences should make claims about A, B, and C, not everyday racial categories.
See above. This has nothing to do with “subspecies”, a category based on arbitrary differentiation. Note the New York Times article linked above. They didn’t need to decide if Polynesians were a “subspecies” to discover if they were more closely related to E Asians than Melanesians. But implicit to the research is that “Polynesians” and “Melanesians” – groupings of individuals by racial ancestry – are scientifically justifiable categories for study.
I don’t know why you are asking me to “re-read” my own link, when my purpose for linking was precisely to show that I believe some of its ideas are flawed!
Sorry, I didn’t read it that way. I thought the link text was “fallicious judgement” because you thought the link proved that it was.
‘Race’ is used interchangeably in the literature to mean ‘subspecies’ and ‘population’, but ‘subspecies’ is not used interchangeably to mean ‘population’
Then subspecies is the more specific term, no? Shouldn’t we stick with it, then?
‘Race’ is useful in the neo-Darwinian era only as a synonym for ‘population’
Sailer’s definition is a great example of this. He basically wrote an entire opinion article/FAQ defining the word “population” and replacing it with “race”… from a scientific POV “race” should be thrown away because of its superfluity as well as the fact that reporting research in terms of “race” will lead people to confused interpretations, as Sailer’s definition is not the one that most people have in their head. (As a sidenote, Sailer thinks that “population” is lefty PC talk for race, when in fact its a much more specific term and has been used for nearly a century.)
Note the New York Times article linked above. They didn’t need to decide if Polynesians were a “subspecies” to discover if they were more closely related to E Asians than Melanesians.
I never claimed that subspecies should be required to look at the differences between two populations, so this is a straw man. What I’m saying is that on a global level, in the context of the “does human race/subspecies exist?” debate, knowing whether there are human subspecies– and if so, what they are– would appear to answer our question.
… from a scientific POV “race” should be thrown away because of its superfluity as well as the fact that reporting research in terms of “race” will lead people to confused interpretations, as Sailer’s definition is not the one that most people have in their head.
Throwing the word ‘race’ away would confuse people even more, and in the entirely wrong direction. The main reason for the “confused interpretations” is because certain scientists have actively promoted that confusion to advance their political agendas at the expense of science.
The architects of the Neo-Darwinian synthesis – all of them – accommodated ‘race’ to the paradigm shift of genetics, as they did evolution itself. This is how they used race and how it is most commonly used in the literature (Sailer did not invent this understanding, but he came to it and communicated it in his own way). I will gladly side with them over Lewontin et al.
What I’m saying is that on a global level, in the context of the “does human race/subspecies exist?” debate, knowing whether there are human subspecies– and if so, what they are– would appear to answer our question.
Yes, knowing whether there are subspecies would appear to answer our question if there are subspecies….
(??)
the semantic argument has utility in regards to communicating population level concepts to the general public. on the other hand, in a within-GNXP context the race or not race debate (which has been had by contributors to this blog from the beginning of its inception) is similar to libertarians and left-liberals arguing over the world “liberal.” all these terms are basically capturing and reducing (and simplifying) the structure of gene frequencies across human populations, which we can agree upon (at the least, that it is an empirical, not a definitional, issue), no matter what our position on more controversial topics such as the genetic component of between group IQ differences.
if i’m doing a post where the term “haplotype” shows up, i use the word “subpopulation.” otherwise, i am liable to use the term “race.” the primary determinant is the audience. this isn’t a hard & fast rule i adhere to, it just seems natural to use different terminology in different contexts because of their inferential potential. you know, like copulation vs. fucking.
Throwing the word ‘race’ away would confuse people even more, and in the entirely wrong direction. The main reason for the “confused interpretations” is because scientists have actively promoted that confusion to advance their political agendas at the expense of science.
Two wrongs don’t make a right. Lay people have very muddled views about human variation, ranging from racial essentialism to denial of genetically meaningful human population groups. I don’t see how talking about human population groups instead of races encourages either pitfall.
The architects of the Neo-Darwinian synthesis – all of them – accommodated ‘race’ to the paradigm shift of genetics, as they did evolution itself. This is how they used race and how it is most commonly used in the literature (Sailer did not invent this understanding, but he came to it and communicated it in his own way). I will gladly side with them over Lewontin et al.
I would side with them over Lewontin as well. But this is what I was talking about when I said blank-slatists rob the discussion of its middle-ground.
Ernst Mayr (an architect of the Neo-Darwinian synthesis extraordinaire), defines “race” as follows:
“an aggregate of phenotypically similar populations of a species inhabiting a geographic subdivision of the range of that species and differing taxonomically from other populations of that species.”
is that not identical to the definition for “population”?:
“a stable group of randomly interbreeding individuals.”
You can’t interbreed unless you share a geographical locale, so Mayr’s concept of “race” is an unnecessary synonym. Who would think that scientists would actually want to introduce an unnecessary term into their work? Especially one that causes so much confusion and emotion among lay people…?
Yes, knowing whether there are subspecies would appear to answer our question if there are subspecies….
1) you clipped out the first part of that paragraph. the paragraph was in response to your claim that was these questions about subspecies would interfere with examining group differences like the ones in that ny times article.
2) more importantly it would answer the question of whether there are human races (assuming we’re not using the stretched, unnecessary definition which is synonymous with “population”)
…
razib’s comments seem spot on by the way.
is that not identical to the definition for “population”?:
No, not entirely, but identical in the neo-Darwinian fundamentals. Mayr was technically defining ‘race’ as a group of populations. But technically a group of populations is conceptually indistinguishable from a population. Virtually all populations, after all, would be groups of populations. The English comprise people from Northern England and Western England, etc, and then people from Northern England, etc, from various towns, etc, on down to local families. As a technical (quantifiable) definition, I don’t think Mayr’s definition works, but on the less ambitious (but still justifiable) level of semantics, the definition makes sense, and reflects common usage among the appropriate working scientists.
Two wrongs don’t make a right… Who would think that scientists would actually want to introduce an unnecessary term into their work?
I don’t think it is wrong. I certainly don’t believe it is ‘wrong’ to have synonyms in science jargon, and like with the English language as a whole, synonyms have slightly different connotations which may be useful in communicating ideas. ‘Demes’ and ‘subpopulations’ also technically add no more information than ‘population’, but both are used widely and imply populations of smaller levels of differentiation within populations of greater differentiation. ‘Race’ and ‘subspecies’, OTOH, imply larger populations of larger differentiation. (e.g. continental divisions. Something akin to ‘factors’ in psychometrics)
And if it weren’t for the claimed technical distinctions of ‘sub-species’, I would have less problems with it. (And I suppose the species concept is not strictly technical or quantifiable either, though ‘sub-species’ compounds what is already mucky.) One problem with the ‘race’ debate, has been the unhelpful denigration of fuzziness and practicality in biological taxonomy.
I think ‘race’ is a fine word of deep organic pedigree that communicates some ideas best in the current semantic climate. Kudos to the many scientists who continue to use it correctly in the literature. What I say has little bearing on the daily operations of ‘normal science’. Scientists decide in their everyday work which terms and concepts are helpful and which are not. And biologists have persisted using ‘race’, despite the infinite pointless agitation.
the paragraph was in response to your claim that was these questions about subspecies would interfere with examining group differences like the ones in that ny times article.
I did not claim this. I only said ‘subspecies’ was irrelevant to such research, in response to your:
Also, I think it’s worth noting that the race/subspecies debate is quite relevant to the group genetic differences debate. If we discover that human variation is best summarized by dividing the human race into groups A, B, and C, then our hypotheses of group genetic differences should make claims about A, B, and C, not everyday racial categories.
Correct categories for A, B, and C is exactly the kind of issue the research in the NYT was resolving. And the ‘subspecies debate’ is entirely irrelevant to this research.
As a technical (quantifiable) definition, I don’t think Mayr’s definition works
Then we’re in agreement on my main point about it.
the definition makes sense, and reflects common usage among the appropriate working scientists.
1) you’ll have to tell me what you mean by “makes sense.” do you mean it has an intelligible meaning, or do you mean it’s somehow more reasonable than the alternative (“population”)? if it’s the former, then that’s not really an argument for using it in science. if it’s the latter, i’d like to know why.
2) The ad popularum “common usage” argument, aside from being a fallacy, favors using “population” over “race.”
I certainly don’t believe it is ‘wrong’ to have synonyms in science jargon, and like with the English language as a whole, synonyms have slightly different connotations which may be useful in communicating ideas.
I agree. But not all synonyms are created equal. Some are needlessly vague and confuse scientists and laymen alike.
And if it weren’t for the claimed technical distinctions of ‘sub-species’, I would have less problems with it.
I currently think the phylogenetic sub-species definition might be useful. It defines sub-species as populations *existing* (present tense) in the same geographical locale with a certain degree of phylogenetic similarity as a result of interbreeding. Goodrum’s FAQ says that it was useful in conservation work.
…
I said: “Also, I think it’s worth noting that the race/subspecies debate is quite relevant…”.. that doesn’t mean I think it’s relevant in all circumstances. Just to clear that up.
Correction to what I said:
“Then we’re in agreement on my main point about it.”
This wasn’t my main point… what I should have said is, “that seems to be the most important point.”
I currently think the phylogenetic sub-species definition might be useful.
different species definitions are useful in different concepts. similarly, i think different ‘race concepts’ have different uses for different contexts. e.g., sailer’s ‘race as an extended family’ has more utility for the united states than it does for much of latin america, where within family phenotypic variation can be high. on the other hand, a phenotypic race can totally mislead in terms of overall genome phylogeny and non-salient phenotypic characteristics (e.g., immune resistance, etc.).
1) you’ll have to tell me what you mean by “makes sense.” do you mean it has an intelligible meaning, or do you mean it’s somehow more reasonable than the alternative (“population”)? if it’s the former, then that’s not really an argument for using it in science. if it’s the latter, i’d like to know why.
2) The ad popularum “common usage” argument, aside from being a fallacy, favors using “population” over “race.”
I’m not really going to continue quibbling, this is my last comment. The disagreements are minor here and I’ve already said what is important.
‘Race’ is an old and widely used synonym for ‘population’ in the biological sciences with it’s own subtle connotations, that evokes biological (and Galtonian) notions and differentiation among larger population units. I approve of it, but more importantly working scientists approve of it in their work. There is no more need to “get rid of it” than the words ‘deme’ or ‘subpopulation’. Public misunderstandings of conceptual race will not be cleared with a verbal shell game, and are exacerbated not by terms, but by systematic and deliberate antiracist obfuscation. If anything, loss of the term would be a victory for such obfuscations. But I promise you the term isn’t disappearing anytime soon.
Some are needlessly vague and confuse scientists and laymen alike.
It’s really not that confusing. It wasn’t “confusing” until important people with an agenda intentionally confused it.
I currently think the phylogenetic sub-species definition might be useful. It defines sub-species as populations *existing* (present tense) in the same geographical locale with a certain degree of phylogenetic similarity as a result of interbreeding. Goodrum’s FAQ says that it was useful in conservation work.
The criteria are so vague and easy to satisfy it should involve naming many millions of new subspecies. As I said: “Perhaps [subspecies] should be applied to obviously incipient species headed towards allopatric speciation, but it isn’t in practice… Many times it is simply based on phenotypic dissimilarity or marked regional differences, but that isn’t necessarily (or even most commonly) the result of speciation or extreme isolation.” The link does not support Goodrum.
sailer’s ‘race as an extended family’
is there anything original about his conception of race? It seems like he took the meaning of “population” which has been around at least since Hardy-Weisenberg and called it ‘race’..
to sailer’s credit, he seems to have come up with a really interesting way of teaching some basic concepts in popgen.
…
jason, you’re right that at this point we’d be quibbling if we continued to argue. i think the source of our remaining disagreement, which wouldn’t go away upon further discussion– is that you see “race” as comparable to “deme” or “subpopulation.” i see “race” as something that began as an unscientific concept and was later given a scientific face lift when people wanted to continue talking about “races”, and I think this is why it confuses people.
to sailer’s credit, he seems to have come up with a really interesting way of teaching some basic concepts in popgen.
well, i think that’s the point. i doubt steve was assuming he’d made a novel population genetic insight, just that explaining in the way he does makes more sense than going into Fst or coefficients of relatedness.
i see “race” as something that began as an unscientific concept and was later given a scientific face lift when people wanted to continue talking about “races”, and I think this is why it confuses people.
but “race” was never arbitrary. the scientific method didn’t exist before a particular period, so race classification was pre-scientific. but that doesn’t mean that it was without biological content; folk taxonomy is imperfect, but it often captures a part of reality (this is alluded to in coyne & orr’s speciation). it’s like saying that taxonomy before cladistics was useless; it was certainly far less rigorous and not very hypothetico-deductive, but cladistics nevertheless confirmed that most of the assumed lineages did have some grounding in reality.
i see “race” as something that began as an unscientific concept and was later given a scientific face lift when people wanted to continue talking about “races”, and I think this is why it confuses people.
This is ridiculous. ‘Race’ began with all the other taxa, and progressed like all the other taxa. Linnaeus applied his order to humans like the other animals, naming four races: European, Amerindian, African, and Asian, and one mythical race that was dropped. This is scarcely different than the primary genetic clusters found by modern geneticists.
All evolution was given “a scientific face lift” by the Modern Synthesis architects. They kept what they did because it still fit. They kept Linnaean taxonomy because it still fit. They kept ‘race’ not because they “wanted to continue talking about “races”", but because they felt it still helpfully described the order of nature within the improved models.
Linnaeus’s races seem like subspecies (they don’t go from the bottom up like populations). Why redefine “race” to mean “population” when we already have a word for it?
Razib,
Re your update:
>In this case, there is Anthropology which deals in Theory and Ways of Knowing, and anthropology which is driven by data, models and analysis.
I managed to really annoy a poster here a while back by trying to make the same point.
I think it may be worth adding that there is nothing per se wrong with folks? meditating upon ?Theory and Ways of Knowing,? at least if they refrain from lying and consider the possibility that they might write in clear English (and some of them actually do). There is also nothing at all wrong with anthropologists? who simply clearly and honestly present ethnographic data for the use of anyone who finds it interesting or enlightening.
Whether or not that is ?science? (most of its practitioners seem to agree that it is not ?science?) is moot.
But some people (most of the posters here) are also curious as to know how far you can go with ?anthropology which is driven by data, models and analysis.? Such efforts seem to be largely (not exclusively) driven by physical anthropologists nowadays.
No need for a fight here ? just different interests for different people.
All the best,
Dave
We need some more Indiana Jones-type anthropologists!