April 21, 2010

More on geographical divide between Asian and Melanesian types in Indonesia (Cox et al. 2010)

I had previously posted about a paper showing a sharp divide in Indonesia between "Asian" and "Melanesian" Y chromosomes. A reader alerts me to another paper from this year, which discovers this divide using autosomal and X chromosome polymorphisms.

From the paper:
this transition is shifted eastward relative to Wallace’s line—a boundary that separates the biogeographic regions of Asia and Wallacea. At its southern limit, Wallace’s line falls between the islands of Bali and Lombok (figure 1), which are separated by a deep-water sea channel that marks the southern edge of the Sunda
Shelf. During ice-age glacial advances, the Sunda land mass included Borneo, Bali, Java and Sumatra, together with mainland Southeast Asia. However, even in periods
of low sea level, deep water in Wallacea separated the Sunda shelf from the eastern landmass of Sahul (connecting New Guinea and Australia). While the distribution of
many flora and fauna conforms to Wallace’s line, the seafaring capabilities of human settlers to this region undoubtedly overcame this barrier to dispersal. Indeed, Asian ancestry exceeds 50 per cent as far east as the island of Alor, which is well within Wallacea and approximately 1000 km east of Bali, as well as on the island of Sulawesi, which is located east of Wallace’s line in the north (figure 1). Curiously, Wallace himself noted this difference, positing a second line in eastern Indonesia corresponding to changes in human phenotype (Wallace 1869; Cox 2008). Wallace’s second ‘phenotypic’ line broadly parallels the rapid decline in Asian admixture identified here. It is refreshing to see (for once) a paper which acknowledges that modern genetics did not discover the wheel but has to a large extent confirmed what previous generations of scientists, working with their eyes (and later their calipers) could plainly see.
A visually interesting figure from the paper illustrates what a "cline" actually is.

We can see how west of 120 degrees longitude there is a uniform area of Asian ancestry, then a sharp transition zone and then a fairly uniform area of Melanesian ancestry.

The above figure illustrates one of the arguments of those (like me) who assert that racial variation in humans is real: the fact that it geographically punctuated (no smooth cline). The smooth areas of uniformity east/west of 120deg deserve to be recognized as real entities.

For visual illustration, three examples from Deniker's The races of man: a New Caledonian woman representing an "eastern" Melanesian type, a group of people from Flores (where, according to the current paper Asian admixture runs at 62%), and finally a Javan man representing a "western" Indonesian Mongoloid type.


Proc. R. Soc. B (2010) 277, 1589–1596
doi:10.1098/rspb.2009.2041

Autosomal and X-linked single nucleotide polymorphisms reveal a steep Asian–Melanesian ancestry cline in eastern Indonesia and a sex bias in admixture rates

Murray P. Cox

Abstract

The geographical region between mainland Asia and New Guinea is characterized by numerous small islands with isolated human populations. Phenotypically, groups in the west are similar to their neighbours in mainland Southeast Asia, eastern groups near New Guinea are similar to Melanesians, and intervening populations are intermediate in appearance. A long-standing question is whether this pattern primarily reflects mixing between groups with distinct origins or whether natural selection has shaped this range of variation by acting differentially on populations across the region. To address this question, we genotyped a set of 37 single nucleotide polymorphisms that are evolutionarily independent, putatively neutral and highly informative for Asian–Melanesian ancestry in 1430 individuals from 60 populations spanning mainland Asia to Melanesia. Admixture analysis reveals a sharp transition from Asian to Melanesian genetic variants over a narrow geographical region in eastern Indonesia. Interestingly, this admixture cline roughly corresponds to the human phenotypic boundary noted by Alfred Russell Wallace in 1869. We conclude that this phenotypic gradient probably reflects mixing of two long-separated ancestral source populations—one descended from the initial Melanesian-like inhabitants of the region, and the other related to Asian groups that immigrated during the Paleolithic and/or with the spread of agriculture. A higher frequency of Asian X-linked markers relative to autosomal markers throughout the transition zone suggests that the admixture process was sex-biased, either favouring a westward expansion of patrilocal Melanesian groups or an eastward expansion of matrilocal Asian immigrants. The matrilocal marriage practices that dominated early Austronesian societies may be one factor contributing to this observed sex bias in admixture rates.

Link

116 comments:

Vincent said...

A visually interesting figure from the paper illustrates what a "cline" actually is.

This is not what a "cline" actually is. This what people who don't know what a cline actually is CALL a "cline".

What you actually see here is a sample in which there is a huge hole: no sampling at all between longitude 125 and longitude 143.

It's hard to say what you might find in that 2,000 km gap, but it's a fair bet you'd find "Asian admixture" coefficients between Alor and highland PNG.

Also, this post implies that non-racists deny that genetic variation exists. Of course this isn't even remotely correct. Every geneticist knows that human exhibit genetic variation.

The fact that humans lack genetically-defined races has nothing to do with whether or not humans exhibit genetic variation.

VV

Onur Dincer said...

I think with the accumulation of knowledge about human genetic variation (especially with large-scale whole-genome analyses in all corners of the world in the near future), a consensus will emerge regarding the idea of human races.

My guess is that our familiar classifications of human races will gradually be abandoned and will eventually give way to a much more intricate and racially varied (= involving more races) classification of human races.

Marnie said...

Thank you, Vincent.

Dienekes said...

This is not what a "cline" actually is. This what people who don't know what a cline actually is CALL a "cline".

I guess you think the authors of the study don't know what a cline is.

After your last attempt to redefine what a "haplogroup" is for the entire population genetics community, I suppose you should start your own dictionary.

Also, this post implies that non-racists deny that genetic variation exists. Of course this isn't even remotely correct. Every geneticist knows that human exhibit genetic variation.

This post (or the paper it's about) has nothing to say about "racists", so you must be imagining things.

Onur Dincer said...

This post (or the paper it's about) has nothing to say about "racists"

For your whole post (including all the quoted parts of the paper in it) I concur. I don't what the non-quoted parts of the paper say, so I can't comment on them.

Maju said...

"What you actually see here is a sample in which there is a huge hole: no sampling at all between longitude 125 and longitude 143".

This is something that authors acknowledge in the text but refer to another paper (Mona 2007, I think) to justify that the cline is sharp, probably even sharper than shown in that graph, as most of the gap is occupied by West New Guinea, which should be quite similar to PNG.

I'm going to avoid the race/not race discussion, which for me is not the crux of the issue (after all "race" is just a word). For me the most important issue that seems to be the striking homogeneity of the West Indonesian and Filipino populations, which seem to contradict the theories of population replacement with "Australoid" (Melanesian) substrate admixture. If there was a population replacement, then either it was incredibly absolute or the substrate population(s) cannot be identified with Melanesians at all.

More informative in this sense might be the HUGO consortium paper on Asian autosomal DNA (that someone hang here). This identifies several pre-Austronesian clusters (Filipino Negrito, Orang Asli Negrito, Proto-Malay) but places again all archetypal Austronesian populations as mostly unrelated to these groups but with a marked "Austroasiatic" component (red) instead.

Questions remain in any case.

Onur Dincer said...

Correction:

Wrong:

I don't what the non-quoted parts of the paper say, so I can't comment on them.

It would be:

I don't know what the non-quoted parts of the paper say, so I can't comment on them.

Marnie said...

onur, we all know what you are saying. btw, as you've probably noticed, my blog compositions aren't perfect either, even though I am a native speaker of English. I appreciate the effort, but you don't need to worry so much about making a mistake. We'll fill in the blanks for you.

Vincent said...

I guess you think the authors of the study don't know what a cline is.
I was commenting on your blog, not on the paper. What I think is that YOU don't know what a cline is, or else you are willingly misrepresenting it in an attempt to advance your racial arguments.

My point is that it is impossible to characterize the figure you excerpted as a cline, since there is (coincidently?) a gap in sampling exactly where you are claiming a "sharp" boundary between the populations.

I call it out because it is a common tactic among race-believers: sampling two divergent populations (while omitting the intermediate populations) then yelling "cluster".

That human variation is essentially and globally clinal is true. The only way to make it look otherwise is to attempt some sort of shell-game with the data, either by goofy sampling or by focusing on the most extreme exception you can uncover.

VV

Andrew Oh-Willeke said...

"racial variation in humans is real"

I have to concur with Vincent here. Nobody denies that there is human phenotypic variation with genetic origins or that the phenotypes have some connection to geographic place of ancestry in a clustered way. Nobody thinks that people with black African origins have that skin color because it was culturally acquired, for example, or that people of Asian origins have epicantic folds because they had different diets as children. Likewise, nobody denies that society already assigns people in culturally meaningful ways to race categories based on appearances somewhat related to ancestral phenotypes.

Those opposed to race as a valid biological concept has more to do with how one draws the line which is cultural. Every "race" includes multiple genetic clusters, e.g. Berbers and Germans are both classified as "white" in the American system. Someone who might be classified one way, "mulatto" perhaps in Brazil, might be classified another way "black" perhaps in the Bronx.

Cultural identity isn't a perfect fit for biological identity (e.g., a Kenyan or Jamacian who has lived his whole life in London is not treated identically to an African-American). And, there are far more people than traditional race theory acknowledges who are admixtures including most African-Americans, mestizos, South Asians, West Wallaceans, a large share of Americans of East Asian origins, and all South African "Colored" individuals, all of completely different characteristic mixes.

One could try to rehabilitate the highly tainted concept of race to fit the perhaps dozen or score of phenotypic clusters (depending on the statistical test you used to distinguish them), but why bother with something that the culture wouldn't accept anyway. Biology is capable of creating many categories of race; the cultural context that is race is not(e.g. both mestizos and South Asians are both considered "brown" in the United States despite having extremely little genomic similarity. Trying to divide the world into three main net categories still ends up hopelessly muddled from a biological perspective with many, many people fitting in more than one category or none.

This may not be as obvious a problem in Greece as it in the U.S. but the problems with race are real.

Andrew Oh-Willeke said...

"most of the gap is occupied by West New Guinea, which should be quite similar to PNG."

West New Guinea is actually very racially and ethnically different from PNG right now, for quite complicated reasons only some of which are historically established. Some of the divide dates to Dutch colonial policy, some dates to a historical Indonesian policy of intentional resettlement of West Indonesians there, and some of that pre-dates Dutch colonialism or is a product of non-state sponsored migration (e.g. the large Chinese merchant class). Some pre-Dutch ethnic distinctions have roots in Highlands v. Wetlands v. Coastal ecology distinctions.

I'm also not at all clear on how much of the admixture if taking place on a population level in which there are non-admixed subpopulations, and how much involves actually admixed ancestry individuals. The pie charts can't distinguish the two. Admixed subpopulations are likely to be from historical era migrations (such as those that had already been well underway in Wallace's day), admixed individuals would be older in origins presumably but would have to be statistically teased from non-admixed subpopulations to be meaningfully described.

Onur Dincer said...

Although the cultural concept(s) of race and the biological concept(s) of race have many parallels (actually they were one and the same thing in the primitive stages of the science of biology) as Dieneke so beautifully illustrated in the beginning of this thread, nonetheless they are vastly different concepts today. So, for the sake of science let's put aside the cultural concept(s) of race and just focus on the biological concept(s) of race.

It is so characteristic of race deniers to give examples from hybrid populations, overwhelming majority of which are relatively recent (last a few thousand or even hundred years) products, to "prove" their case. Instead, we should focus on archetypal or prime races, which are still to be found in most lands of Earth in high numbers, and which are products of ten thousands of years of highly resolute inbreeding and isolation in specific conditions. That is why each prime race has specific characteristics and talents, which largely reflect the original conditions of their original homelands (urheimats - not to be confused with the linguistic concept of urheimat).

Marnie said...

Again, thank you Vincent and Andrew.

I'd add that perception weighs heavily in our classification of race. That's particularly obvious with Indians, for example. Their skin color is brown, but their features are quite "Caucasian". Our perception probably over estimates genetic difference in this case.

Our perception probably under estimates the genetic difference between a native of Moscow and a native of Cork.

terryt said...

"it is impossible to characterize the figure you excerpted as a cline"

I think it's a pretty good example of a cline. Even in the map you see completely white pie charts in the west and completely black ones in the east. In between the proportions vary in relation to how far the populations are from either extreme.

"It is so characteristic of race deniers to give examples from hybrid populations, overwhelming majority of which are relatively recent (last a few thousand or even hundred years) products"

I strongly suspect that hybrid populations have been forming since Homo erectus first left Africa, if not before.

"The fact that humans lack genetically-defined races has nothing to do with whether or not humans exhibit genetic variation".

But, like most species, the greatest genetic extremes tend to be found at the geographic extremities of the species' range. The 'East Asian' phenotype tends to be at its most extreme in Northeast Eurasia, the 'Australoid/Melanesian' offshore from Southeast Eurasia, and 'Caucasians' in Northwest Eurasia. Every other genetic variation could easily be the result of 'hybrids' between particular neighbouring physically discrete populations.

"For me the most important issue that seems to be the striking homogeneity of the West Indonesian and Filipino populations, which seem to contradict the theories of population replacement with 'Australoid' (Melanesian) substrate admixture".

But there's obviously some sort of substrate other than 'East Asian' though. The photograph of the Javan man would not be confused with any photograph of an East Asian from further north. And we'd have to assume that the original Melanesians came from somewhere. Presumably at some stage people to the east and west of Wallace's line looked much the same as each other.

Maju said...

"The photograph of the Javan man would not be confused with any photograph of an East Asian from further north".

Probably the turban. ;)

Seriously I don't see major differences between that guy and the random Cantonese. Dress him in Mandarin clothes and lengthen his mustache and he's ready to play the super-villain of a film of Fumanchu.

There must be some differences, the same that there are some differences between Germans and Britons but they are not so easy to discover and are very irregular.

"And we'd have to assume that the original Melanesians came from somewhere. Presumably at some stage people to the east and west of Wallace's line looked much the same as each other".

Sure. Once upon a time... in the heat of the Great Eurasian Expansion... people still looked all much alike. Once upon a time...

Vincent said...

So, for the sake of science let's put aside the cultural concept(s) of race and just focus on the biological concept(s) of race.

But race is a cultural concept, not a biological one. There's no ignoring that this is the essential point of contention between racists and race-deniers.

Dienekes said...

I was commenting on your blog, not on the paper. What I think is that YOU don't know what a cline is, or else you are willingly misrepresenting it in an attempt to advance your racial arguments.

Doesn't fly. It is the authors who see a "steep Asian-Melanesian ancestry cline" not me.

My point is that it is impossible to characterize the figure you excerpted as a cline, since there is (coincidently?) a gap in sampling exactly where you are claiming a "sharp" boundary between the populations.

First of all, as the authors argue, the gap is probably much smaller, as the Melanesian element is dominant east of 125 degrees.
Second, there is of course nothing "wrong" with characterizing a figure with sampling gaps as a cline.

I call it out because it is a common tactic among race-believers: sampling two divergent populations (while omitting the intermediate populations) then yelling "cluster".

First, "race-believers" are not "racists" as you misleadingly claimed. It is a common tactic among race deniers to conflate the two.
Second, the authors did not sample "two divergent populations" but many populations, and they made a reasonable inference about the parts of New Guinea which they didn't sample.

Onur Dincer said...

But race is a cultural concept, not a biological one.

How can you be so sure of that?

First, "race-believers" are not "racists"

Yes of course. But I wouldn't call myself a "race-believer" either, as believing is a characteristic of religious faiths. If I were to classify my position on human races, I would define myself as a supporter of the biological concept of human races (subspecies), or shortly, "race-supporter". I should make it clear that there is no fixed biological concept or classification of human races that I support and my opinions on biological human races are open to change depending on developments in science (especially genetics).

Vincent said...

Second, there is of course nothing "wrong" with characterizing a figure with sampling gaps as a cline.

YOU are trying to advance the argument that this data shows two races: one west of 120 degrees latitude and one east of it.

YOU make this argument based on what YOU call a "sharp transition zone". You characterized the data as "geographically punctuated (no smooth cline)".

However, as you appear to concede, there isn't any data AT ALL for the so-called "transition zone". There is a 2,000 km hole, where no samples were taken, in exactly the spot in the "cline" which you are trying to characterize.

So there are two things wrong with what you wrote. One is that there is no cline, because there is no data. Two is that by drawing a conclusion from data that doesn't exist, you perpetuating a fantasy instead of science.

I have no expectation that you care enough about science to overcome your racist approach, but leaps of logic this great deserve to be called out.

VV

Onur Dincer said...

I strongly suspect that hybrid populations have been forming since Homo erectus first left Africa, if not before.

Hybridization (admixture) between human subspecies is a very old phenomenon. But with the passage of time most of the former human subspecies disappeared as a result of extreme admixture with other subspecies or extinctions. Today's archetypal human subspecies (races) are remnants of a very long process of dispersals, hybridizations, extinctions and isolations, so their archetypal character is of course a relative one, not an absolute (that would be completely unscientific and ridiculous, very reminiscent of religious thinking) one.

Dienekes said...
This comment has been removed by the author.
Dienekes said...

However, as you appear to concede, there isn't any data AT ALL for the so-called "transition zone". There is a 2,000 km hole, where no samples were taken, in exactly the spot in the "cline" which you are trying to characterize.

You must be blind, because the transition occurs clearly right in the middle of the figure where there are plenty of data points.

As for the alleged "gap" it coincides with a region where 97.5% of the Y-chromosomes are of Melanesian origin and thus there are excellent chances that in the so-called gap there is a preponderance of Melanesian ancestry. It's not a "fantasy", but a reasonable estimate, which is also shared by the authors of the study.

I have no expectation that you care enough about science to overcome your racist approach, but leaps of logic this great deserve to be called out.

Insult away, it doesn't change the fact that there is a zone of ancestral homogeneity from southern China and Indochina through most of Indonesia, then a sharp transition zone east of the Wallace line and then a zone of ancestral homogeneity in New Guinea and islands to the east.

Maju said...

I understand that there is a sharp cline already in East Indonesia and that it's surely sharper than suggested by the graph further east (in West Papua).

I don't understand that means that races exist because races are a concept, not any objective fact.

In every cline there are clusters and for every cluster there is a cline. Racialists emphasize clusters but that's only product of their subjectivity. Racialists would like clusters to be even neater and more distinct (and sure: sometimes clines are pretty sharp) and they probably have everything in their homes compulsively sorted in well tagged boxes, carefully closed so their contents don't ever mix accidentally.

These people do not understand that all orders are not but subsets of Chaos. They still believe in a Newtonian punctuated and machine-like Universe with preference to Quantum uncertainty and (even worse) Relativity!

Sadly, it's a psychological problem and I doubt we can fix them because it's part of the conservative syndrome, defined by fear to disorder and difference (a phobia after all). They'd like to be like ants or better even: robots with perfectly logical algorithms inside them and not just the, often even more "accidentally perfect", unstable products of Chaos, which is the matrix of all orders.

That's why I think that the discussion on race is to a large extent pointless: because we are discussing how people want to see reality: as neatly packed boxes (even if forcibly so) or as a rather unorganized flow of multicolor difference. In fact both visions have some reason (as usually happens) but as Chaos predominates over all orders, racialism (with its compulsive obsession with sorting people) is necesarily more limited and short-sighted.

And, of course, it is largely subjective.

Maju said...

PS- Also, it's easier to forge two races when you make two clusters. This is a potential weakness of the paper because maybe there are three, four, five, six or two hundred million different clusters.

That's also a reason why I pointed to the HUGO paper, which shows a larger variability of clusters, not so neatly packed in most cases, for the region (but that fails to sample New Guinea mostly anyhow).

Marnie said...

Dienekes,

I'm going to avoid for the moment, the issue of whether or not there is missing data in the cline, or racial gradient of this paper. At some point, I will look myself.

Also, I think it is not helpful to spend too much time trying to draw hard boundaries between people who look racially different.

I do enjoy your blog very much.

I'm concerned about a comment that you made in your post on "Hard Problems in Social Science." I'll quote:

"By "hierarchy", I refer to the fact that some things that are unequal may in fact be organized in terms of superiority or inferiority. Different foods may all be pleasant to have as options, but they are not all equally good for one's health. Sickle-cell trait may harden the population against outbreaks of malaria, but it is not exactly pleasant at the level of individual. Monogamy or polygamy may be different social practices, but they have important adverse consequences for societal functioning (e.g., cheating by polygynous-inclined males in monogamous societies vs. an excess of unattached, potentially troublesome, males in polygynous ones)."

I will state simply that I think there is great plasticity in human behavior. There are also many genetic diseases that people of different genetic backgrounds are resistant or susceptible to.

I'm very troubled with the concept of your suggestion that there is a racial hierarchy.

It would be helpful if you further illustrate your thoughts on this.

I was following up on that "Hard Problems in Social Science." The Wall Street journal wrote an excellent article on it:

"Hard Problems from 'Soft' Sciences" by Christopher Chabris. One of his comments:

"There was at least one problem that went unmentioned at Harvard: the problem of hubris."

I'd appreciate your thoughts on what you mean by the "problem of racial hierarchy."

Onur Dincer said...

Maju, you are talking just like a Turkish friend of mine who is likewise unsympathetic to the idea of biolological human races. Like you, he mentions things like fear or phobia to disorder, difference and adulteration of the assumed pure (relatively speaking) form while enumerating the characteristics of race-supporters. Not surprisingly, he is a leftist too. Based on my personal experiences I think that political and ideological views have a strong correlation with (like many other things) views about human diversity (genetic and/or cultural).

Onur Dincer said...
This comment has been removed by the author.
Onur Dincer said...

Btw, Maju, I don't think Newtonian mechanics, chaos theory or any other concept of physics have anything to do with our discussion about biological human races (subspecies). Human races, like most of the other subspecies of living beings, in general weren't separated from each other in a perfect manner even before the increasing contacts of the last a few thousand years. There were periods of genetic leakage from one race to the other, but most often these were so small that they were rapidly being absorbed by the dominant race of a territory without leaving any significant trace or even any detectable trace at all.

Maju said...

There may be also (or "rather") an effect of personality (individual psychology) on ideological preferences.

In any case, this idea of conservative people having fear of disorder is not taken from any Marxist or anarchist manual but rather something I have read about on scientific media (just too lazy to search for the source right now).

If correct (and I think it's essentially correct), the variance of human psychologies in this aspect: maniacally orderly (fearful) and carelessly disordered (fearless) at the extremes, may be in dynamic equilibrium, as both trends may have their pros and cons.

In any case, I think that Dienekes' (and others') obsession to "demonstrate" that races do exist has such psychological element to it and also that it is not pure science but a pseudoscientific endeavor to justify an aprioristic belief. Not too different from trying to justify Noah's deluge legend on whatever scientific evidence of any prehistorical flood.

The opposite endeavor, i.e. claiming in absolutist terms that "races do not exist" may also be equally misguided. In my opinion we should all take a relaxed attitude on the issue, the same we normally do about traits such as blond and brunette hair, blue or brown eyes, etc.

However there are serious political and social implications to racism and ethnocentrism and therefore we have to take sides sometimes. In these cases I will always take side for Humanism.

Also I happen to think that racism and ethnocentrism are the consolation prize of the mediocre: I am just petty and insignificant (though I won't admit) but my race/nation/sect is the best. It's not a worthy choice, IMO.

Vincent said...

You must be blind, because the transition occurs clearly right in the middle of the figure where there are plenty of data points.
Sorry, but you are wrong. This paper has no samples between 125 and 143 degrees longitude. That's the gap, in which there is no data. How can you pretend to know WHAT the transition looks like when you can't observe the transition?

Put that in context: it'd be like sampling Europe in London and Palermo with nothing in between. You could extrapolate between those two points for, say, Strasbourg. But you'd be on thin ice to claim there was a "sharp boundary" between London and Palermo - because you have not EXAMINED anything between the two locations.

If you want to know what the cline between Alor Island and PNG looks like, then you have to sample in between. The authors haven't done that, which may be fine for the aims of their research.

But twisting their research in an effort to support your own agenda is something. But it's not good science.

Dienekes said...

Sorry, but you are wrong. This paper has no samples between 125 and 143 degrees longitude. That's the gap, in which there is no data. How can you pretend to know WHAT the transition looks like when you can't observe the transition?

I'll say it one more time for the slow-witted: the sharp transition zone which the authors (and myself) observe occurs around 120 degrees longitude where you go from almost complete Asian ancestry to about half.

So, there is no problem with the fact that the sample has a gap. As to what that gap contains the authors make a very reasonable inference that it's predominantly Melanesian rather than Asian, and I see no reason to think otherwise.

But twisting their research in an effort to support your own agenda is something. But it's not good science.

Their research indicates mixing of two long-separated ancestral source populations in a sharp transition ... over a narrow geographical region.

So, whether you like it or not it is the authors that yell "cluster", and not me.

Vincent said...

I'll say it one more time for the slow-witted: the sharp transition zone which the authors (and myself) observe occurs around 120 degrees longitude where you go from almost complete Asian ancestry to about half.
You're changing your story. In your blog post, you write "We can see how west of 120 degrees longitude there is a uniform area of Asian ancestry, then a sharp transition zone and then a fairly uniform area of Melanesian ancestry."

The only pace there is "uniform area of Melanesian ancestry" is east of 143 longitude. So if you want to change the argument, fine. It still takes over 1000 km to go from 100% Asian to 50% Asian: that's the distance between Greece and Syria. Sharp, indeed.

That's your argument for distinguishing Melanesians from Asians as discreet races? Really?

VV

aargiedude said...

"we genotyped a set of 37 single nucleotide polymorphisms"

That's very bad, the decent studies use 500,000 SNPs. Their argument is that they're informative for Asian/Melanesian ancestry, which is roughly equivalent to using 500 randomly selected SNPs, and which is still very lousy (1,000 times worse than the 500K SNP studies). Also, the "Caribbean" region between west Indonesia and New Guinea is a special case, because it's never been physically connected to either Asia or Australasia, so we can easily expect these people to have developed their own distinctive genetic profiles, kind of like Sardinians for the same reason, and thus not be exactly compatible with a test that looks at their Asian and Melanesian ancestry. In other words, they're missing a potential 3rd cluster. Or put yet another way, if they had done a 500K SNP test, with no preconceived ancestry markers, they might have discovered that the "Caribbean" islands between west and east form a 3rd genetic cluster, and that the Asian and Melanesian components in which they're trying to pigeon-hole these people are 5% or less of their ancestry.

Imagine the situation like if we were trying to determine the ancestry of West Asia and we use SNPs that are informative for sub-Saharan/East Asian ancestry. We discover that Egyptians, North Africans, and south Europeans are completely sub-Saharan, Russians and Indians are completely East Asian, and everybody else has a combination of sub-Saharan and East Asian ancestry that adds up to 100%. So this study has erred. It should have first determined if Wallacea formed its own genetic cluster/s, which given the circumstances (they're islands that have never been physically connected to Asia/Australasia even at the height of the Ice Age) is a highly likely YES. After that, determine any ancestry informative SNPs for the THREE genetic clusters and see how they stack up. But that wouldn't be necessary because to determine the existence of potential 3rd or 4th clusters they would have had to do a large scale 500K SNP scan, from which they could then produce a more informative chart with STRUCTURE. The PC analysis they did is usually just a visual simplification in most autosomal studies. They took a shortcut and arbitrarily decided that Wallaceans shall be pigeon-holed into either East Asian or New Guinean ancestry.

"A higher frequency of Asian X-linked markers relative to autosomal markers throughout the transition zone suggests that the admixture process was sex-biased, either favouring a westward expansion of patrilocal Melanesian groups or an eastward expansion of matrilocal Asian immigrants."

This explanation about patrilocality/matrilocality doesn't really hold up. If a group of men and women move into a region they're going to displace the previous group without distinctions, regardless of wether they have a patrilocal or matrilocal social system. The explanation would make sense if the parties involved had been living next to each other for thousands of years. In such a case, the lopsided interactions of patrilocal tribes would gradually result in the female-linked chromosomes dispersing wider than the male-linked chromosomes, leading to the current anomaly of their distribution. And this is a paleolithic continuity explanation, the opposite of what they propose. This is also what I think is happening in West Asia, with East Asian and sub-Saharan mtdna, which in both cases is clearly vastly more common than their y-dna counterparts. The sub-Saharan lopsided distribution is usually explained as a result of slavery, but when it comes to the same phenomenon regarding East Asian mtdna there's just no reasonable explanation that can be offered... from a historical angle.

aargiedude said...

Another attempt to explain the problem with their Wallacean results. Iberians probably have 10 times as much sub-Saharan ancestry as they do East Asian ancestry, so if we tried to pigeon-hole them into either sub-Saharan or East Asian groups, using ancestry informative markers for African and East Asian ancestry, they would probably come out as 90% sub-Saharan and 10% East Asian. But in fact, Iberians have a genetic distance of 0,15 with Nigeria and 0,11 with Japan, so they're CLOSER to East Asia, but their results using wrongfully preselected informative markers would label them as being overwhelmingly of sub-Saharan ancestry.

terryt said...

"Dress him in Mandarin clothes and lengthen his mustache and he's ready to play the super-villain of a film of Fumanchu".

No he's not. He has a quite subdued eyefold compared to Mongolians, his moustache is quite luxurious compared to Northern Chinese and his hair (what we can see of it) appears to be quite wavy, not straight. Many studies have emphasised the fact that Northern and Southern Chinese are different.

"But you'd be on thin ice to claim there was a 'sharp boundary' between London and Palermo - because you have not EXAMINED anything between the two locations".

In the present case it's obvious that there is a transition zone of some sort across Wallacea. Wallacea is undoubtedly a 'boundary' between two phenotypes, even though you might prefer to claim it is not a 'sharp boundary'.

"This paper has no samples between 125 and 143 degrees longitude. That's the gap, in which there is no data. How can you pretend to know WHAT the transition looks like when you can't observe the transition?".

By 125 degrees we have already reached a 50% mix of each type, and we can be pretty sure the transition is associated with the Austronesian expansion. Half of the transition zone, or cline, occurs to the west of the region for which there's no data. I think it's safe to assume the other half of the transition zone follows the pattern of the Austronesian settlement pattern, so it too contains a mix of the two phenotypes. And you will note that Dienekes has now put up a blog on New Guinea, which amplifies the pattern

"That's your argument for distinguishing Melanesians from Asians as discreet races?"

I would think that your own eyes should tell you that.

"Today's archetypal human subspecies (races) are remnants of a very long process of dispersals, hybridizations, extinctions and isolations"

Quite. So we've finished up with ill-defined subgroups with phenotypic extremes found in the geographic extremities, exactly the same as for most other species. Although most other species are not as mobile as humans have been. As a result sub-species are more clearly defined for non-human species.

"These people do not understand that all orders are not but subsets of Chaos".

But the different human phenotypes are largely geographically distributed, so they are not a result of 'chaos'. Human genes have expanded at times, but limited or diverted by geographical features.

terryt said...

"it's easier to forge two races when you make two clusters. This is a potential weakness of the paper"

Quite. They lump all East Asians for a start.

"which is roughly equivalent to using 500 randomly selected SNPs"

The authors are careful to point out that the SNPs were not 'randomly selected'.

"they might have discovered that the 'Caribbean' islands between west and east form a 3rd genetic cluster"

Isn't that exactly what they did find? The people of Wallacea are a different population but this is a result of a particular mixture of mainland populations.

"and that the Asian and Melanesian components in which they're trying to pigeon-hole these people are 5% or less of their ancestry".

If these two populations made up only 5% of their ancestry where on earth could the other 95% have come from?

"If a group of men and women move into a region they're going to displace the previous group without distinctions, regardless of wether they have a patrilocal or matrilocal social system".

Not always so, and I suspect that mixing with pre-existing populations has been much more common than usually accepted.

"but when it comes to the same phenomenon regarding East Asian mtdna there's just no reasonable explanation that can be offered... from a historical angle".

The long-accepted explanation is that the population of Wallacea became a mixed population and then expanded by sea to both the east (into the Pacific) and west (to Madagascar). By the time Polynesia was settled the male side was mainly Melanesian (Y-hap C2) and the female Asian (mtDNA B).

"I'm very troubled with the concept of your suggestion that there is a racial hierarchy".

I'd hope we'd all be uncomfortable with such an idea.

Maju said...

@argiedude

"This explanation about patrilocality/matrilocality doesn't really hold up."

Yeap. I'm not really convinced either, specially as Papuans are not just patrilocal but extremely Patriarchal in general.

However I don't agree necesarily with your claim of: "If a group of men and women move into a region they're going to displace the previous group without distinctions, regardless of wether they have a patrilocal or matrilocal social system".

They do not need to displace. They can perfectly mix and/or live side by side. It happens often in fact and the overall byproduct is admixture.

Surely we are in front of complex processes for which the hypothesized Austronesian matrilocality may be a factor but not the only one. Founder effects (i.e. prehistorical incidents we can't reconstruct) must be at play too.

"Iberians probably have 10 times as much sub-Saharan ancestry as they do East Asian ancestry"...

Source? I don't know if it's correct or not but in principle I don't have any reason to believe this claim at all. As a hunch it might be correct for Portuguese (because of where they expanded in the colonial period and also because they generally have more North African ancestry for some founder effect) but not for all the region.

"But in fact, Iberians have a genetic distance of 0,15 with Nigeria and 0,11 with Japan, so they're CLOSER to East Asia"...

This in fact contradicts your previous claim.

Maju said...

@Terry:

"He has a quite subdued eyefold compared to Mongolians"...

I thought we were comparing him with Cantonese.

"his moustache is quite luxurious"

Really? It's a tiny moustache. Fumanchu had it much bigger.

"his hair (what we can see of it) appears to be quite wavy, not straight"

Can't really be said but in that case he would not be so archetypal from Javanese, who overwhelmingly have flat hair.

"Many studies have emphasised the fact that Northern and Southern Chinese are different".

Somewhat different.

You live in an imaginary Coonian universe where there was once upon a time a pure Mongoloid race in Mongolia or Northern China that expanded southwards miraculously and without any clear archaeological nor historical backing. Not even a genetic one (see the HUGO paper, for example).

The reality seems to be much more complex. Even Dienekes (the paladin of racialism) came up with anthropometric results of his own that clearly split Mongoloids in various clusters and make North Chinese and Mongolians markedly different anthropometrically. This while Egyptians and Norwegians remained the same (for comparison).

So both genetics and anthropometry seems to depict a much more complex landscape for Eastern Eurasia than your oversimplified Coonism.

"In the present case it's obvious that there is a transition zone of some sort across Wallacea".

But what if Wallacea had one or several distinctive clusters of its own? They'd be hidden by the bipolarity.

And, considering the highly diverse insular nature of the region, I'd suspect there could well be a distinctive cluster (or several, specially if you insist in including Philippines) for those islands.

So far nobody seems to have located it but a more focused study (and less B&W than this one) should IMO.

The HUGO paper seems to detect a Melanesian cluster that is the same as the distinctive Wallacean one. But it is clearly distinct from the Filipino Negrito cluster, as well as from any other minority cluster such as Malaysian Negrito or Proto-Malay.

So it's not just a B&W bipolar transition but we have a whole array of several uniquely distinct populations, both in the insular area and in mainland East Asia.

So should we speak of the Filipino Negrito race as a distinct one (I think it's the case if we're going to use racespeak)? Should we maybe speak of Ryukyuan, Sinoid (general mainland EA cluster, not found anywhere in pure form), Austroasiatic and Austronesian races? (And let's not forget the Hmong one, nowadays found only diluted).

You said that Northern and Southern Chinese are different and that is somewhat correct: Northern Han are essentially Blue-Yellow ("Sinoid-Ryukyuan"), while Southern Han are essentially Blue-Green ("Sinoid-Austronesian"). But neither shows any Melanesian or Negrito element, as do not Austronesians outside of Wallacea (instead they have the "Austroasiatic" component often and in large amounts).

Maju said...

"I would think that your own eyes should tell you that".

My eyes, which watched a documentary on Wallacean fishermen the other day, tell me that they look like any other people in the world. In fact they look like all them together: one was more Papuan-like (but not quite), the other very much Caucasoid-looking, a woman looked Khoisanid and yet one or two looked close to the Mongolid archetype. My eyes say that phenotype Chaos is boiling in their genes.

"But the different human phenotypes are largely geographically distributed, so they are not a result of 'chaos'".

You don't understand Chaos. Even if some polymorphic alleles (and remember that some 99% are simply not polymorphic at all) are more or less geographically distributed, they are seldom in fixed manner but as a matter of degree in fact. Also expression is altered by epigenetics, random recombination, etc. Recombination in fact is pretty much chaotic in nature (I say that every person is mixed for a reason) but if you don't like it you have only one choice: clonation. And even clonation will not allow you to escape the indetermination and unpredictability of Chaos.

The fact that a bunch of Wallaceans could display such a diversity of phenotypes as I described above shows how rich is the potential hidden in your apparent but fallacious clusters.

"I'd hope we'd all be uncomfortable with such an idea".

I am but I don't think the author of the blog nor some of his readers are. You know: a lot of people like to think that their (often very poor) individual status improves if their groups status is higher, even if it's just an imaginary thing altogether. It's a psychological consolation prize: I can barely meet ends and my life is a hell but I am "Aryan" (or whatever your favorite race/ethnos/sect). It's also used to keep the working class divided along ethnic lines and to keep some "privileged" (???) groups more loyal to the oligarchies: "my boss exploits me badly but he's white so we are in the same ship. We can go together to the KKK/neonazi demo in the weekend and focus our repressed anger against blacks/gypsies/immigrants, who surely are to blame for all (add absurd pseudo-reasoning pre-designed by your local racist leader). Maybe even I get a promotion out of all that".

Marnie said...

Maju,

Many interesting comments.

On your last, about groupthink and status: 'White' people are by no means the only people who demonstrate hubris regarding their perceived racial status.

I can describe many laughable examples of non-white ethno-centrism.

It's a human tendancy.

Here, Dienekes really does have a point about racial friction. There's no escaping the friction, because we all, all human beings, have a tendancy toward racial hubris. Call it tribalism.

I keep an article on my fridge about the experiences of an African American (brown) woman in Ghana (black). I was going to put it up to illustrate that it is not just white people who have a tendency to think they are racially superior. I haven't put it up, because I think it is somewhat OT. But it is quite funny, and demonstrative of our human foibles.

I do not live in a racially homogeneous place. I'm everyday reminded of all the possible variations of us and them.

I take it as a warning to myself.

Maju said...

"On your last, about groupthink and status: 'White' people are by no means the only people who demonstrate hubris regarding their perceived racial status".

Of course, you are right. It was just an example. Replace the term as needed. I was in fact thinking that I should have included an example of Jewish ethnocentrism too... but it'd get too cumbersome, right?

"It's a human tendancy".

It's a pitiful human weakness.

Marnie said...

"Jewish ethnocentrism too... but it'd get too cumbersome, right?"

Maju, you're no shrinking violet.

I wasn't intending my point about tribalism to descend into a finger pointing session, except to point the finger at myself.

Tribalism may be a problem in modern societies, but as we are on an anthropology blog, we should all be quite aware of human history. Tribalism is a necessary fact of human nature, without judgement.

The question is not whether or not tribalism exists, or should exist. The question is how we mold tribalism in the modern world, in order to have a functioning society and world.

Unknown said...

It is important to note that the cline is longitudinal. If you go due east from Alor it is mostly open water (or sparsely populated islands) until you reach mainland New Guinea, where there are samples. Furthermore, there are many reasons to believe that other PNG samples would be mostly if not entirely Melanesian (partially by design!), barring coastal samples that had some Austronesian history. The gap in sampling is a gap in LAND. It is not equivalent to leaving out half of Europe.

The SNPs are AIMs, chosen with high Fst between Han and PNG, so the fact that there are only 37 is the point, not the weakness. In general, many SNPs on a chip will not be ancestry informative. It is true that one weakness is this lumps "Asians" together. However, this approach works better than you might naively think. To verify this, look closely at the HUGO data (50K), and you will see the same cline from light green (Asian/Austronesian?) to dark green (Melanesian) in their STRUCTURE plot (with a little red on Flores).

We found the transition/cline in the Y data right about the same time Murray found it in the autosomes/X (I work with Murray and Tanya). What's amazing is that nobody noticed either before, well, unless you count Wallace. ;-)

Onur Dincer said...

Marnie: "I'm very troubled with the concept of your suggestion that there is a racial hierarchy."

Terry: "I'd hope we'd all be uncomfortable with such an idea."

Maju: "I am but I don't think the author of the blog nor some of his readers are."


As a reader of this blog, I am neither comfortable nor uncomfortable with such an idea; I am neither racist nor anti-racist. Besides, racial hierarchy is a subjective concept and I am not interested in subjective concepts, I only care about objective facts and their scientific and rational interpretations (at least in scientific forums and blogs like this). The concept of subspecies of living beings is a more or less objective concept (even if not as objective as the concept of species). Moreover, in parallel with the developments in genetics, subspecies (including human subspecies) are being classified more and more precisely and objectively. Modern human subspecies (races) are as real as the subsepecies of any other species in the biosphere of Earth. Here is the conclusions of a new study (Formulating a Historical and Demographic Model of Recent Human Evolution Based on Resequencing Data from Noncoding Regions) about human genetic variation, which is just one of the countless proofs of the reality of human races:

"Our results support a model in which modern humans left Africa through a single major dispersal event occurring ~60,000 years ago (i.e., the time of divergence of Africans and Eurasians), corresponding to a drastic reduction of ~5 times the effective population size of the ancestral African population of ~13,800 individuals. Subsequently, the ancestors of modern Europeans and East Asians diverged much later, ~22,500 years ago, from the population of ancestral migrants. This late diversification of Eurasians after the African exodus points to the occurrence of a long maturation phase in which the ancestral Eurasian population was not yet diversified." (emphases and parenthesis mine)

http://www.plosone.org/article/info:doi/10.1371/journal.pone.0010284#pone.0010284-Akey1

Even if the study's results points to a relatively late time of divergence between Europeans and East Asians (~22,500 years ago), it is still early enough for subspeciation.

The HUGO paper seems to detect a Melanesian cluster that is the same as the distinctive Wallacean one. But it is clearly distinct from the Filipino Negrito cluster, as well as from any other minority cluster such as Malaysian Negrito or Proto-Malay.

So it's not just a B&W bipolar transition but we have a whole array of several uniquely distinct populations, both in the insular area and in mainland East Asia.

So should we speak of the Filipino Negrito race as a distinct one (I think it's the case if we're going to use racespeak)? Should we maybe speak of Ryukyuan, Sinoid (general mainland EA cluster, not found anywhere in pure form), Austroasiatic and Austronesian races? (And let's not forget the Hmong one, nowadays found only diluted).

You said that Northern and Southern Chinese are different and that is somewhat correct: Northern Han are essentially Blue-Yellow ("Sinoid-Ryukyuan"), while Southern Han are essentially Blue-Green ("Sinoid-Austronesian").


Most of such minor races are actually sub-races (which are generally much less rigid and homogeneous entities than the archetypal races, so it is very difficult and problematic to classify them as races proper) of the archetypal (major/prime) races. The rest of the minor races are hybrids (mostly recent, as I stated in my above posts) of the archetypal races. I think the real human races are the archetypal races, the minor races aren't really races. Still, in a relaxed sense they can be classified as sub-races (itself a relaxed category), or in the case of the archetypal hybrids, hybrid races or racial hybrids.

Onur Dincer said...

Hey Marnie,

In a thread you asked me about sacred springs in Turkey. Yesterday I finished reading Speros Vryonis' classic book "The Decline of Medieval Hellenism in Asia Minor and the Process of Islamization from the Eleventh through the Fifteenth Century". In a part of the book he says that the hot springs of Yalova, Turkey, are sacred for the Turks and that their sacredness goes back to the pagan times through Byzantine. Here is the relevant part:

"in many instances the Muslims appropriated characteristics and peculiarities attendant upon a particular Christian sanctuary along with the building itself. This resulted, not infrequently, in a syncretization of the old and new elements, particularly at the lower levels of society. This transferral of Christian practice, associated with a church converted to a mosque, was usually intensified or facilitated whenever conversion and intermarriage occurred. This is apparent in the Muslim and Christian practice of equating saints of one religion with those of the other religion. The best known equation of a Muslim with a Christian saint revolves about the figure of Khidir. As early as the fourteenth century, John Cantacuzene noted that the Turks worshiped St. George in the figure of Khidir Elias, and at Elvan Chelebi, east of Chorum, he was associated with St. Theodore. In both cases he bas been identified with equestrian, military, dragon slayers. In Konya the Muslims reverenced St. Amphilochius under the guise of Plato or Eflatun, and elsewhere they identified Sari Saltik with St. Nicholas, and St. Charalambos with Hadji Bektash. A religious ceremony held in honor of Hadji Bektash by the Greek Christians of Sinassus is recorded in the fifteenth century, and in the nineteenth century local Christians believed that the tomb of Hadji Bektash near Kirshehir was the tomb of St. Charalambos. Often the sanctuaries that were relinquished to Islam retained a certain Christian holiness. The Muslims of fifteenth-century Konya attributed miracles to the tomb of St. Amphilochius. At the extremes of Anatolia-at Edessa in the east and at Yalova in the west-the survival of the holy went back through the Byzantine era to pagan times; the famous hot springs of Yalova were in antiquity dedicated to Apollo, in the Byzantine days to St. Michael, and under the Turks they were resanctified by the burial of a dervish; at Edessa the sacred fish in the fountain of the mosque of Abraham continue a pre-Turkish tradition." (p. 485, emphasis mine)

Dieneke, sorry for the off the topic.

terryt said...

"It's also used to keep the working class divided along ethnic lines"

I'm sorry to say that I agree with that observation.

"I was going to put it up to illustrate that it is not just white people who have a tendency to think they are racially superior".

One of my friends was born in Hong Kong, and is half Chinese. He claims the Chinese as a group are far more racist than any white people he's ever met.

"Really? It's a tiny moustache. Fumanchu had it much bigger".

What film was that in? This one:

http://www.toptenz.net/wp-content/uploads/2008/06/fu-manchu-mustache.jpg

Does the Indonesian really look like this Mongolian:

http://www.lax-a.is/media/hausmynd/M60.jpga.is/media/hausmynd/M60.jpg

"your apparent but fallacious clusters".

But even when we look at Africa we find the phenotypes are not chaotically distributed. Khoi-San in the south look quite different from West African people. The situation of course has become complicated by the expansion of people originating in West Africa through much of the rest of Sub-Saharan Africa. And Pygmies form another cluster, although how ancient is debatable. In East Africa the people form another cluster because they have varying levels of Eurasian admixture, as do those from the Mediterranean coast. So: Chaos?

Dienekes said...

It is important to note that the cline is longitudinal. If you go due east from Alor it is mostly open water (or sparsely populated islands) until you reach mainland New Guinea, where there are samples.

That's a good point, it makes sense to sample geographically trying to get as much coverage as possible, but sample points should also reflect the underlying population densities. The Sahara for example is about the same area as China, but probably has several orders of magnitude fewer people in it.

Onur Dincer said...
This comment has been removed by the author.
Onur Dincer said...

For those who don't know, Speros Vryonis is a famous American historian of Greek descent specialist in Greek and Byzantine history:

http://en.wikipedia.org/wiki/Speros_Vryonis

Again sorry Dieneke, this will be my last post on this subject in this thread (at least that is what I hope! :D).

Marnie said...

onur,

Speros Vryonis' book: I'll read it. Thanks for telling me about it!

Rather smart, to sanctify a spring, don't you think? Also, its impressive that people manage to hold onto meaning in their lives (such as a spring) as the centuries wash over them.

Here's a link I've been meaning to put up for you, onur:

http://www.facebook.com/topic.php?uid=2209228234&topic=4459

It's from the public facebook page for the prefecture of Kozani, in Northern Greece. It's a public page. The thread is interesting, but you want to go down to Christos' comment "the map shows where the Vallahdes lived, greek muslims but not really muslims, long story, but worth reading" and then push into the google link. You'll have to read the google translated version. Like your post above, it discusses synchretization of Christian Orthodox and Muslim beliefs and practices.

The information happens to be on a promacedonia link. (We've had that discussion.)

Onur, if you're interested in reading this, and you can't find the link I am talking about, let me know.

I appreciate the information.

Marnie said...

Onur,

As to your comments on race and objectivity, I will say that it is very difficult to be completely objective about race.

Genetic anthropology is about as close as it gets. However, like any science, hypotheses about race are subject to uncertainty, outright error, and complexity.

Here's an idea: Come up with an on the spot racial calculator. Say, you were in a cafe, checking out some hot babe, and you decide you are more than a little interested. But you were not sure of her "race". She's checking you out to. So you pull out your racial calculators to see if you have a match. "Sorry, wrong subrace." Bummer.

Do we really want to live in such a world? (If we're not careful, Apple will tell us that we do.)

My interests in genetic anthropology lie with understanding the human journey. It's also likely to have medical uses that could provide more effective drugs, etc.

But as to race, as far as I'm concerned, there is only one. A human one.

Marnie said...

Terry,

"But even when we look at Africa we find the phenotypes are not chaotically distributed. Khoi-San in the south look quite different from West African people. The situation of course has become complicated by the expansion of people originating in West Africa through much of the rest of Sub-Saharan Africa. And Pygmies form another cluster, although how ancient is debatable. In East Africa the people form another cluster because they have varying levels of Eurasian admixture, as do those from the Mediterranean coast."

Good observation.

In so far as we want to understand the peopling of SE Asia, this is an interesting discussion.

Also, an intersting discussion about sampling. I never knew sampling could get so heated.

Maju said...

"Most of such minor races are actually sub-races"...

Because you say so!

There's no scientific way of saying this is a race, this is a subrace, this is individual variation...

Anyhow, my point was to illustrate that there is no genetic evidence of "Melanesian" or "Australoid" admixture among mainline SE Asians. The admixture they have (on genetic data) is mostly between "Mongoloid" groups.

Some of the "purest Austronesians" (per the HUGO paper), the Mentwai, have loads of curled hair, what is in itself a rejection of Terry's Coonist hypothesis of such traits being derived from Melanesian/Asutraloid admixture.

"The rest of the minor races are hybrids"...

Again I say that this is false. It's aprioristic and baseless.

There were never just three or five or eleven isolated populations that gave birth to the extant ones but a large continuity with some interspersed areas of discontinuity (low density buffers, natural barriers). It was homogenization by "admixture" (gene flow) what generated the archetypes or more or less homogeneous phenotypes in these regions and it was the lack of such flow (or its drastic limitation) at the natural barriers what created the differences.

There is no urheimat for the "Mongoloid" or "Caucasoid" "races": all the East Asian and West Eurasian world regions are as a whole. They exist only because these regions are geographically separated by deserts and other such buffers, preventing to a large extent genetic flow between the regions but clearly allowing it inside them. The "races" are product of repeated admixture, the same that a puree is the product of disintegrating and stirring its ingredients in a single pot.

Onur Dincer said...

Onur Dincer: "Most of such minor races are actually sub-races (which are generally much less rigid and homogeneous entities than the archetypal races, so it is very difficult and problematic to classify them as races proper) of the archetypal (major/prime) races."

Luis Aldamiz: "Because you say so! There's no scientific way of saying this is a race, this is a subrace, this is individual variation."


Maju, the essential difference between our views is in terminology. I prefer much more clearly defined terms like "race", "subspecies", "subrace", "hybrid race", "racial hybrid" for various human groups, while you prefer vague words like "human group", "population", etc. for the same groups. Apart from these differences in terminology, we are actually saying mostly the same things.

As a clear example, you say: "Anyhow, my point was to illustrate that there is no genetic evidence of "Melanesian" or "Australoid" admixture among mainline SE Asians. The admixture they have (on genetic data) is mostly between "Mongoloid" groups."

I could make the same statement with the exact same words without hesitation. But alternatively I could also make a small change by saying "Mongoloid subraces" instead of "Mongoloid groups", while you would never do such a thing. Also I wouldn't normally (never for archetypal races) use quotation marks in the sense of emphasizing the artificialness of a category while mentioning race/subrace names like "Caucasoid", "Mongoloid", "Melanesian", etc., while you usually do that.

Onur: "The rest of the minor races are hybrids (mostly recent, as I stated in my above posts) of the archetypal races."

Luis: "Again I say that this is false. It's aprioristic and baseless."


Again we are faced with an essentially terminological clash, it seems.

Luis: "There were never just three or five or eleven isolated populations that gave birth to the extant ones but a large continuity with some interspersed areas of discontinuity (low density buffers, natural barriers). It was homogenization by "admixture" (gene flow) what generated the archetypes or more or less homogeneous phenotypes in these regions and it was the lack of such flow (or its drastic limitation) at the natural barriers what created the differences."

This isn't different from what I have been saying all along. As a clear example, I say: "Today's archetypal human subspecies (races) are remnants of a very long process of dispersals, hybridizations, extinctions and isolations".

If your objection is to my this statement:

"Human races, like most of the other subspecies of living beings, in general weren't separated from each other in a perfect manner even before the increasing contacts of the last a few thousand years. There were periods of genetic leakage from one race to the other, but most often these were so small that they were rapidly being absorbed by the dominant race of a territory without leaving any significant trace or even any detectable trace at all."

Then I should clarify that in this statement I used the term "race" to designate only today's archetypal human races. Wish I had been more explicit.

Onur Dincer said...

Luis: "There is no urheimat for the "Mongoloid" or "Caucasoid" "races""

By urheimat I mean the territory where a certain race or subrace formed as a distinct entity. Surely all races and subraces, or with your terminology, human groups formed somewhere, didn't they?

Luis: "They exist only because these regions are geographically separated by deserts and other such buffers, preventing to a large extent genetic flow between the regions but clearly allowing it inside them. The "races" are product of repeated admixture, the same that a puree is the product of disintegrating and stirring its ingredients in a single pot."

Again, I could use the exact same words with the sole difference that I wouldn't enclose the word "races" in quotation marks.

In conclusion, our main point of contention is our different terminological preferences. But this is no less a significant issue, as the biological human races (subspecies) are scientifically valid entities. Here you err by denying the validity of scientifically valid entities. Worse still, it seems you are making this error (consciously or unconsciously) because of your ideological conditioning.

Onur Dincer said...

Marnie, thanks for the links. The stories of Vallahdes and the other converting (to Islam and in many, maybe even most, groups to also Turkishness) groups of Greece during the Ottoman rule are really interesting. Eventually most of the Muslim population of Greece were settled in Turkey with the 1923 population exchange between Greece and Turkey, and the remaining still non-Turkified Muslims in this newly exchanged Muslim population were too eventually Turkified (became Turkish-speaking) in the newly founded highly nationalist (less so today) Turkish Republic that they were settled.

As to Vryonis' book, it is freely available in the Internet in PDF form. You can easily find it with a little googling. Also I have it in my harddisk so I can send it to you through email if you give me your email address. The only noteworthy shortcoming of the PDF version (in both of the two files I downloaded from different websites) is that it has some bugs and because of that about ten pages of the book are missing.

Maju said...

I prefer much more clearly defined terms like "race", "subspecies", "subrace", "hybrid race", "racial hybrid"...

And how are these terms "clearly defined"? They are not: they are very diffuse concepts. So it's not an issue of words only but of their definitions, which I say don't exist. That's what makes race an almost totally subjective matter.

Also I wouldn't normally (never for archetypal races) use quotation marks in the sense of emphasizing the artificialness of a category while mentioning race/subrace names like "Caucasoid", "Mongoloid", "Melanesian", etc., while you usually do that.

That's because I have no reason to think they really mean anything beyond pretty much diffuse clusters of genetic/phenotype variance. So it's not like they are absolute categories but rather common places.

A much more solid category is something like "people with black curly hair". This is an specific clearly measurable phenotype (and possibly a single genetic clade too) but it goes across "races" and defines none of them. To make a "race" you have to add and subtract and average and reduce... it's an statistic simplification. I see pointless to even discuss its relevance because it has none. It may be convenient to talk of Caucasoids or Melanesians now and then but we must not forget that they don't really exist as absolute categories but only as statistic simplifications, maybe useful but always artificial constructs.

"This isn't different from what I have been saying all along".

Maybe (unsure). But it's different from what others say.

"Surely all races and subraces, or with your terminology, human groups formed somewhere, didn't they?"

What I say is that the formed and are still forming as we speak (they don't remain static but change) within pretty much separated world regions by means of "stirring up" these pots once and again. Each "race" is a homogenized puree of as many ingredients as you can imagine, not the product of the expansion of a single imaginary group living once in some mythical Land of the Hyperboreans or whatever.

"... the biological human races (subspecies) are scientifically valid entities".

They have only statistical value. You can easily find a zillion of those averaged components that cross all "racial borders" or at least many of them. And this is not because of admixture between "pure races" (at least not in most cases) but because those phenotypes pre-date the process of homogenization and remain within them at higher or lower frequencies, the same that the onion in the lentils puree is the same as the onion in the pumpkin one, regardless that the overall purees are different.

terryt said...

"Some of the 'purest Austronesians' (per the HUGO paper), the Mentwai, have loads of curled hair, what is in itself a rejection of Terry's Coonist hypothesis of such traits being derived from Melanesian/Asutraloid admixture".

How so? We know that after the Austronesian expansion east (Lapita pottery) many Melanesian people adopted the improved boating and followed on behind. This has given rise to a cline across the Pacific, from Melanesian-looking people in New Britain to Polynesian-looking people in the Central/East Pacific. Even where Polynesian-speaking people have subsequently moved back to the west and formed the 'Polynesian outliers' the people have become mixed, to at least some extent, and the cline is maintained. As a result it's hardly surprising that 'the Mentwai, have loads of curled hair'. They are a product of admixture with the later eastward-moving Melanesian groups.

"There were never just three or five or eleven isolated populations that gave birth to the extant ones"

Of course not. But people at the geographic extremes have presumably always looked different from people in the centre of the human distribution. And they have presumably changed over time. Anyway you say:

"It was homogenization by 'admixture' (gene flow) what generated the archetypes or more or less homogeneous phenotypes in these regions and it was the lack of such flow (or its drastic limitation) at the natural barriers what created the differences" ... "The 'races' are product of repeated admixture".

Yes. Even the extreme phenotypes at the geographic extremes are themselves a product of immigration and hybrid formation. The movement and mixing is ancient.

"There is no urheimat for the 'Mongoloid' or 'Caucasoid' 'races'"

There probably is, even though you go on to claim:

"all the East Asian and West Eurasian world regions are as a whole".

So how is it that it's possible to say instantly whether a particular individual comes from east or west Eurasia?

"it seems you are making this error (consciously or unconsciously) because of your ideological conditioning".

My impression too.

terryt said...

"And how are these terms 'clearly defined'? They are not: they are very diffuse concepts".

And they're nomore 'clearly defined' in the rest of biology either. So what's your problem?

"diffuse clusters of genetic/phenotype variance".

The same as subspecies then. And once a cluster diverges sufficiently from other clusters it become a separate species. But, again, the word 'species' is ill-defined.

"It may be convenient to talk of Caucasoids or Melanesians now and then but we must not forget that they don't really exist as absolute categories"

But in most cases they're confined to specific regions too, so although they may not exist as 'absolute categories' they do exist as 'valid categories'.

"Maybe (unsure). But it's different from what others say".

It's pretty much what I say.

"And this is not because of admixture between 'pure races'"

But it is the product of admixture between populations with different evolutionary histories. We can call these different populations 'races' or more long-winded terms such as 'geographically separated by deserts and other such buffers, preventing to a large extent genetic flow between the regions but clearly allowing it inside them'.

Maju said...

Terry: have you even bothered looking at the HUGO paper (link in my first comment, above), specifically on Mentawai genetics? They have the purest "Austronesian" (light green) component of all sampled peoples, along with the Taiwanese Ami. If there is one Austronesian people that is not admixed, these are, it seems, the Mentawai. So their rather common curled hair should be part of the Austronesian package. This does not surprise me the least, after all there are many "Mongoloid" peoples displaying curly hair, even in America.

But for you, with your many preconceptions about how things "should" be and your blindness towards the actual facts when they contradict these, this makes no sense, so you repeat your preconceived argumentation instead of looking at the problem with the attention it deserves.

"So how is it that it's possible to say instantly whether a particular individual comes from east or west Eurasia?"

Because you're looking at this from an statistical viewpoint: you can say and be right in maybe 90% or 95% of cases but there are some which look totally out of place, which you cannot really tell.

And also because you are looking at tablespoons of the same puree, so all pumpkin puree ones are orange, while the lentils puree ones look brown, regardless on whether pumpkin was also an ingredient in the puree.

Enfin.

terryt said...

"If there is one Austronesian people that is not admixed, these are, it seems, the Mentawai. So their rather common curled hair should be part of the Austronesian package".

And probably was. So what's your problem? A mixed population will surely give rise to variation in its offspring. Everyone knows that the F2 generation is extremely varied. It's the F1 where all individuals tend to look much the same. The F2 generation is where animal breeders start their serious selection.

"there are many 'Mongoloid' peoples displaying curly hair, even in America".

Many? You have to search very hard to find them, and even then their hair is nowhere near as curly as Melanesians'. Besides which they are more likely to come from the south or have some admixture in their ancestry.

Marnie said...

Onur,

Vryonis' book is for sale on Amazon. I'll get it there. Again, thanks for telling me about it.

terryt said...

Maju. I realised as I was potting up some plants that you have no grasp at all of the Wallacean/Austronesian situation. This is a product of 'your many preconceptions about how things should be and your blindness towards the actual facts when they contradict these'. I'll try to explain again: the Austronesians are a MIXED population. They originated in Wallacea. From the article:

"We conclude that this phenotypic gradient probably reflects mixing of two long-separated ancestral source populations—one descended from the initial Melanesian-like inhabitants of the region, and the other related to Asian groups that immigrated during the Paleolithic and/or with the spread of agriculture".

I'd prefer to confine the word 'Melanesian' to people beyond New Guinea, and therefore to say that the Austronesians were 'descended from the initial [Papuan]-like inhabitants of the region, and the other related to Asian groups that immigrated during the Paleolithic and/or with the spread of agriculture'. The Melanesians are at least partly Austronesian genetically, and mostly Austronesian-speaking.

It's therefore not at all surprising that members of this mixed population might have curly hair. In fact it's rare for Polynesians to have straight hair. It's usually wavy rather than tightly curled though.

People from New Guinea:

http://cache.daylife.com/imageserve/08Xl9rD5tNfrs/610x.jpg

Young Melanesians:

http://www.friendly-bungalows-tanna-vanuatu.com/images/girls-custom.jpg

And young Polynesians doing kapa haka:

http://productsfromnz.com/pics/big_1793.jpg

terryt said...

And I've just gone back to your link. Seems that, contrary to what you've maintained, there is quite a 'Papuan' influence in Indonesia and even the Han Chinese are basically a mixture between a southern 'Thai/Kradai' population and a northern 'Altaic' one.

Maju said...

"Many? You have to search very hard to find them, and even then their hair is nowhere near as curly as Melanesians'"

I mean curly in the regular sense of Europeans, not in the thinly curly of Africans, which is rare in Eurasia, except maybe among some Mediterraneans and Negrito/Melanesians. It's enough anyhow to make my example perfectly valid.

My whole point is that such trait is not generally present among "Mongoloids" because of admixture but because it's "always" been there, even if as minority trait, exactly as blond hair among Europeans... or Melanesians.

"And probably was. So what's your problem?"

My problem is that this is not what you were saying before but "They are a product of admixture with the later eastward-moving Melanesian groups".

So my problem is that such claim was a nonsense. Are you reconsidering it or not?

Maju said...

"Maju. I realised as I was potting up some plants that you have no grasp at all of the Wallacean/Austronesian situation".

It's maybe better that way, considering how you manage your "grasp" (preconceived ideas) of it.

"I'll try to explain again: the Austronesians are a MIXED population. They originated in Wallacea".

I don't think that's likely from what we know: they might have originated in West Indonesia (Mentawai), Philippines or Taiwan. These groups are the only ones with "pure Austronesian" autosomal genetics (per the HUGO paper)... unless you postulate some post-Austronesian major migration onto their homeland by another group (Austroasiatics, Melanesians?), which I don't make much sense of.

I suppose that the Mentawai are some sort of a colony in a previously desert island group, so the HUGO K=14 map seems to support the generally accepted Filipino-Taiwanese origin of this group. But it also supports the Mentawai as a fossil of the earliest colonists probably, before admixture with pre-existent Austroasiatic populations (red cluster).

"It's therefore not at all surprising that members of this mixed population might have curly hair. In fact it's rare for Polynesians to have straight hair. It's usually wavy rather than tightly curled though".

But the Mentawai show no sign of admixture per the HUGO paper: they have one of the "purest" gene pools in all East Asia. The only Melanesian sample (insular Melanesian, Nasioi language) is also among those "purebred" groups but with a totally different component (dark green) that cannot be associated to early Austronesians. And the only other three such "pure" groups (excepting Austronesians) are Ryukyuans, the Mlabri (an Austronesian spekaking hunter-gatherer tiny group) and the control HapMap CEU sample.

So the Austronesian original stock is best represented by Filipinos, Taiwan Aborigines and Mentawai, yet these have curly hair quite often. So this trait was part of the original Austronesian genetic package. Yes or yes?

"And I've just gone back to your link. Seems that, contrary to what you've maintained, there is quite a 'Papuan' influence in Indonesia"...

There is a Melanesian component in Wallacea (dark green), of course. I never said otherwise.

"and even the Han Chinese are basically a mixture between a southern 'Thai/Kradai' population and a northern 'Altaic' one".

Well, Northern Han have the Mainland EA component (blue) and the Ryukyuan (or NE Asian) one (yellow) as main ones. Also some Hmong (light blue) and Austronesian (light green) components. I see no reason to use the adjectives you chose. There's no Kradai component (all Kradai speakers are blue-green mixture, with Austronesian red sometimes) and the blue component is so evenly spread among mainland East Asians that the only name it can be given is "Mainland(-er)".

aargiedude said...

[i]"Iberians probably have 10 times as much sub-Saharan ancestry as they do East Asian ancestry"...

Source? I don't know if it's correct or not but in principle I don't have any reason to believe this claim at all. As a hunch it might be correct for Portuguese (because of where they expanded in the colonial period and also because they generally have more North African ancestry for some founder effect) but not for all the region.[/i]

Iberians have 1% to 2% mtdna L, excepting Portuguese, but less than 1% East Asian mtdna. They also have 0,3% to 0,5% E1b1a, but only 0,1% East Asian y-dna.

[i]"But in fact, Iberians have a genetic distance of 0,15 with Nigeria and 0,11 with Japan, so they're CLOSER to East Asia"...

This in fact contradicts your previous claim.[/i]

If originally Iberians had a genetic distance of 0,10 to East Asia and 0,15 to sub-Saharan Africa, receiving 1% African ancestry and 0,3% East Asian ancestry would tilt them a little towards sub-Saharan Africa, but their genetic distances would still be generally the same.

.........................

What this study did is pretty much a PC plot, using a couple dozen ancestry informative markers for separating 2 genetic clusters: East Asia and Australasia. If they had included Europeans in their analysis they would probably have come out 100% East Asian. This is not an in depth analysis, a la STRUCTURE and its cluster bars. The people that live in the islands east of Java and west of New Guinea (the Wallacean area) could well be descended overwhelmingly from the people that have lived there for tens of thousands of years, and would form their own genetic cluster in a 500K SNP STRUCTURE analysis, but what this study is picking up are the signals from their small amount of ancestry from the genetic clusters of East Asia and Australasia. That's what I meant with my example about Iberians; if the same analysis had been done with Iberians, using ancestry informative markers for African and East Asian but not Caucasian, then the composition of Iberians would be determined by whatever small shreds of non-Caucasian ancestry they have, and ironically they could thus be found to belong overwhelmingly to the African genetic cluster, despite in fact being closer to East Asians. Because their overall genetic make-up is being determined by looking at their 1% ancestry from those 2 genetic blocs; the whole picture is being missed.

aargiedude said...

To onur, I think, who responded to me that the people of Wallacea can't be anything other than a mix between Australasians and East Asians. A people who live in a region for a long time will develop their own regional genetic quirks that can later be picked up by a program such as STRUCTURE and identified as a genetic cluster. If we perform a true in depth analysis of the people of the Wallacean islands, we will find out if they belong or not to their own genetic cluster, and thus if their population history is mostly a recent phenomenon of trans-Wallacean population movements, or if they are mostly descended from the people that lived there for tens of thousands of years ago. As a comparison, say nobody knows the real ancestry of Mexicans. When we do an in depth analysis of their dna, they could form their own genetic cluster, different from Indians and Europeans, in which case we would conclude they are overwhelmingly descended from the autochtonous people of Mexico. Or, as effectively is the case, they could be split into different genetic blocs, some or all of which are clearly dominant in other parts of the world, and thus we would conclude that they are descended at least partly from some more recent population movements. If left alone for enough time, they would once again form into a genetic bloc that could be separated from others, even though their dna originally comes from people that belonged to those genetic clusters. If we found out that Wallaceans forms their own genetic cluster/s, the next step would be to validate if this phenomenon is due to something concrete (a long established population) or if it's an artifact of inbreeding in an isolated small population or some other other oddity, but those kinds of odd results are always limited to very small populations, as in last year's autosomal study of Sardinians, which was able to separate 8 isolated mountain villages into 8 genetic clusters.

Dienekes said...

To onur, I think, who responded to me that the people of Wallacea can't be anything other than a mix between Australasians and East Asians. A people who live in a region for a long time will develop their own regional genetic quirks that can later be picked up by a program such as STRUCTURE and identified as a genetic cluster. If we perform a true in depth analysis of the people of the Wallacean islands, we will find out if they belong or not to their own genetic cluster, and thus if their population history is mostly a recent phenomenon of trans-Wallacean population movements, or if they are mostly descended from the people that lived there for tens of thousands of years ago.

Your logic is flawed. A population can be BOTH a mix of two distinct populations (Mongoloids and Australo-Melanesians in this case) AND be genetically distinct forming their own cluster.

Indeed, if a population forms its own distinct genetic cluster, we cannot at all infer that they are descended from "people who lived there tens of thousands of years ago".

If that was the case, then we would infer that Ashkenazi Jews who form their own distinct cluster are an ancient population, rather a recent one (Middle-Eastern+European).

The flaw in your logic is in assuming that a population can become genetically distinct and form its own cluster over long evolutionary time spans, but that is simply not the case.

aargiedude said...

The pie charts produced by this study show the people of Sumba, Flores, and Alor to be mostly of East Asian ancestry. We know their y-dna, thanks to the Lansing study, is 90% Australasian. To reconcile this picture, we need these people to basically consist of 100% male Australasians and 100% female East Asians. That's stretching the bounds of matrilocal/patrilocal effects on geneflow to the surreal absurd.

And even then, that would result in Wallaceans being 50% East Asian, not 75% as this study claims.

They made a PC plot using ancestry informative markers. This study does not tell the real story of Wallaceans' genetic make-up.

aargiedude said...

Dienekes, you didn't catch my last sentence, where I said that if a genetic cluster is determined to exist, the next step is to

"validate if this phenomenon is due to something concrete (a long established population) or if it's an artifact of inbreeding in an isolated small population or some other other oddity, but those kinds of odd results are always limited to very small populations, as in last year's autosomal study of Sardinians, which was able to separate 8 isolated mountain villages into 8 genetic clusters"

Your example of Ashkenazis would qualify as an "isolated small population" (1,000 years ago when they went through a serious genetic bottleneck).

aargiedude said...

they could form their own genetic cluster, different from Indians and Europeans, in which case we would conclude they are overwhelmingly descended from the autochtonous people of Mexico.

A slight lapse on my part, the single genetic cluster could well (and probably will) belong to the American Indian genetic cluster. But the general idea stands. Do a 500K genome analysis of Mexicans, or Wallaceans, and see how their genetic structure results come out. Don't pre-determine they WILL be a mix of East Asian and Australasian and try to determine their composition by pigeon holing them into these 2 options with a couple dozen ancestry informative markers for those 2 genetic clusters, which may well constitute an insignificant fraction of their ancestry, thus leading to results that are just noise, as I explained in my theoretical exmaple of what would happen if we tried to determine Iberians' ancestry by using ancestry informative markers for East Asians and sub-Saharan Africans, only.

Do they form their own group, distinct from East Asians and Australasians? Then they're almost certainly descended from the aboriginal people of Wallacea, barring an unlikely case of genetic drift or inbreeding affecting them after becoming mixed (producing a "fake" cluster, as in the case of Sardinians I noted previously).

Dienekes said...

Would you say that the Japanese are a small isolated populationt too? THey are certainly distinct from other Mongoloids, but are also certainly the product of admixture which hasn't occurred "tens of thousands of years ago".

aargiedude said...

Would you say that the Japanese are a small isolated populationt too? THey are certainly distinct from other Mongoloids, but are also certainly the product of admixture which hasn't occurred "tens of thousands of years ago".

No way is it a certainty that the Japanese are a result of a mix between Ainu and people from the mainland. That's a theory, like the Aryan Invasion Theory.

Maju said...

"Iberians have 1% to 2% mtdna L"...

MtDNA L does not automatically mean "sub-Saharan": there are many L lineages that look more typically North African or West Asian than anything else and must have been there since about the OoA time or just a little later.

"If they had included Europeans in their analysis they would probably have come out 100% East Asian".

Probably.

"The flaw in your logic is in assuming that a population can become genetically distinct and form its own cluster over long evolutionary time spans, but that is simply not the case".

He is right on that. Otherwise populations could not form clusters but would be just a mish-mash of diverse ancestral clusters. But that's not how statistical genetics work: instead they take the most statistically significant differences (methods may vary and be questioned but that's their intent anyhow) and group the real peoples' genetic pools on such basis, forming clusters (K-means) or similarly identifiable patterns (PC, NJ, ML, etc.) Admittedly it's the only way to make some sense with the huge complexity of the recombinant part of the genome but has its limitations too.

When you look at specific genes and their alleles instead you get highly complex patterns even where founder effect and drift have clearly left their mark. A good example is fig. 4 (right hand panels) of Coops 2009.

Marnie said...

Why does a distinct cluster necessarily indicate descent from an aboriginal group?

Aren't their standard ways to estimate the timescale and population size needed to get a distinct cluster from two or more distinct clusters?

Marnie said...

Also, aren't there others ways to examine the length of time that a group has been distinct (besides looking at clustering behavior)?

I'm thinking of some of the techniques discussed in the "Khoisan and Bantu genomes from southern Africa" paper,
Stephan C. Schuster et al.

terryt said...

"So my problem is that such claim was a nonsense. Are you reconsidering it or not?"

No. Of course I'm not reconsidering, because you are completely wrong. After the Lapita people (Austronesian-speaking) moved east into the Pacific beyond the Northern Solomons people from New Britain and New Ireland, who had been bypassed during the initial movement, then followed on. This situation has been widely accepted by anthropologists for years, supported by both archeological and genetic evidence. Ebizur even pointed out somewhere that Y-hap C2 is virtually unknown north of New Guinea but is common both east and west of it. The reason is that Melanesian haplogroups such as K and M have expanded and replaced C through the region. These Melanesian haplogroups reach as far as Fiji.

"These groups are the only ones with 'pure Austronesian' autosomal genetics (per the HUGO paper)... "

How on earth can you have 'pure Austronesian' genes if the Austronesians are a mixed people? Talk sense man. That the Austronesians are not a single population is shown simply by the fact that in the east the Y-haps are O3 and C2 and in Madagascar O2 with no C, along with some other haplogroups not found in other Austronesians. Even though you absolutely refuse to accept it the posted study shows a mixture of two populations in Wallacea. So how can you talk of 'pure Austronesians'?

"The people that live in the islands east of Java and west of New Guinea (the Wallacean area) could well be descended overwhelmingly from the people that have lived there for tens of thousands of years"

Probably not. There's a good chance many of the islands had become uninhabited by the time of the Austronesian expansion, a little more than 5-6000 years ago. This paper shows people did survive in parts of the Philippines and some Indonesian islands though.

"what this study is picking up are the signals from their small amount of ancestry from the genetic clusters of East Asia and Australasia".

Not a 'small amount of ancestry'. Virtually totally East Asian and Papuan. Not Australasian. There were no people in Australasia at the time, except for the Australian part, and they didn't contribute to the Wallacean population at all. I'll excuse 'Melanesian', although even that is not an accurate term.

"To reconcile this picture, we need these people to basically consist of 100% male Australasians and 100% female East Asians. That's stretching the bounds of matrilocal/patrilocal effects on geneflow to the surreal absurd".

Surprisingly that's exactly what we find in Eastern Polynesia. The mtDNA is almost completely B (East Asian) while the Y-hap is almost completely C2 (Australo-Melanesian). I admit that's likely to be the result of a series of bottlenecks during the migration through the Pacific, but that's what they've finished up with.

terryt said...

Regarding the terminolgy of Pacific people. Here's an explanation I offered earlier:

'Papuans' is shorter to write than 'New Guinea natives'. Australian Aborigines and New Guinea natives are fairly different, so I tend to use 'Papuans' just for New Guinea/Northern Melanesians. And the word 'Melanesian' includes a wider group than just the Papuan-type people of Northern Melanesia. Some Melanesians, especially those from the Southern Solomons and beyond, speak Austronesian languages and so are probably at least partly a product of the Austronesian expansion. So I would tend to separate them from 'Papuans' too. So we finish up with a number of different-looking Pacific people: Papuans, Australian Aborigines, Melanesians, Polynesians and Micronesians.

From that it seems that this paper is using 'Melanesian' where I would use 'Papuan'. 'Australasia' is a term used to combine Australia and New Zealand. Their indigenous people are quite different from each other.

Maju said...

"Why does a distinct cluster necessarily indicate descent from an aboriginal group?"

It does not automatically. The only thing it says is that the group has been very much isolated from other groups for quite long. Otherwise they would have various components. How long exactly? Could not say - I would analyze it depending on context (samples, geography, known prehistory and history) and depth/quality of the analysis.

Maju said...

"No. Of course I'm not reconsidering, because you are completely wrong. After the Lapita people (Austronesian-speaking) moved east into the Pacific beyond the Northern Solomons people from New Britain and New Ireland, who had been bypassed during the initial movement, then followed on".

How is this related to curly hair among Mentawai?

"How on earth can you have 'pure Austronesian' genes if the Austronesians are a mixed people?"

That's what seems apparent from the HUGO K=14 graph. I'm stating the obvious when you look at it.

"There's a good chance many of the islands had become uninhabited by the time of the Austronesian expansion"...

This is not a matter of "chance" (your way to say wishful thinking, it seems) but of whether there's data to support such claim or not. In the case of the Mentawai archipelago it is plausible, but I know of no other documented cases. Also the fact that what appears most among Austronesians (Malays specifically) is the red "Austroasiatic" component seems to suggest that Austronesian mostly advanced on inhabited land and most probably Neolithic areas with some density.

terryt said...

"Also the fact that what appears most among Austronesians (Malays specifically) is the red 'Austroasiatic' component seems to suggest that Austronesian mostly advanced on inhabited land and most probably Neolithic areas with some density".

Yes. Because the Austro-Asiatic component is mainland SE Asia (Sunda in fact) we can assume that the Austronesians 'advanced on inhabited land' on the mainland. This suggests they did so already in some numbers. But that tells us nothing about whether the Wallacea the Austronesians developed in was populated before the Neolithic. In fact your much-prized HUGO diagram actually has very few examples from Wallacea, and the red doesn't appear in them. From what I can make out most of the Austronesian samples are actually from what had been Sunda. What's more the Papuan genes are by no means confined to Eastern Indonesia. They seem to spread as far as Sumatra and Southern Borneo.

Just a reminder from the original post:

"Phenotypically, groups in the west are similar to their neighbours in mainland Southeast Asia, eastern groups near New Guinea are similar to Melanesians, and intervening populations are intermediate in appearance. A long-standing question is whether this pattern primarily reflects mixing between groups with distinct origins or whether natural selection has shaped this range of variation by acting differentially on populations across the region".

The pattern is hardly likely to reflect natural selection. If that was the case we wouldn't see a cline across the region. The variation would be random.

Further on in the article the authors state, 'Certainly, the spread of some Asian groups was relativley recent, coinciding with the first agricultural settlements in ISEA (Bellwood 2005). This process is often termed the Austronesian expansion, and is putatively linked to demic dispersals from mainland China during the Mid-Holocene'.

Of course I realise you know much more about the subject than Peter Bellwood, Michael Hammer or Tatiana Karafet. I'm not familiar with Murray Cox's work but I'm sure you know more about the settlement of SE Asia than he does.

"How is this related to curly hair among Mentawai?"

The Mentawai are probably mainly descended from the pre-Austronesian inhabitants. Just because they're mainly green in the HUGO diagram doesn't mean they're 'pure' Austronesian. The green could mean almost anything. Presumably they adopted the Austronesian language, as did many other people through the region. Looking for 'pure' Austronesians is like looking for 'pure' Indo-Europeans, or 'pure' Baques.

terryt said...

Some may find this extract from a Bellwood book interesting. He looks at the languages of SE Asia and New Guinea. His comments on occupation of the deep forest are also very interesting:

http://books.google.co.nz/books?id=4obAfGBGKY0C&pg=PA123&lpg=PA123&dq=halmahera+prehistory&source=bl&ots=qPhsUi1tib&sig=UC5kV3y2cOpRj_5LIGbac3TXyXk&hl=en&ei=eFrVS_HkB4jasQO5wdn0CQ&sa=X&oi=book_result&ct=result&resnum=2&ved=0CAoQ6AEwAQ#v=onepage&q=halmahera%20prehistory&f=false

Marnie said...

aargiedude (and anyone else who wants to take this on):

what do you mean by this statement:

A people who live in a region for a long time will develop their own regional genetic quirks that can later be picked up by a program such as STRUCTURE and identified as a genetic cluster. If we perform a true in depth analysis of the people of the Wallacean islands, we will find out if they belong or not to their own genetic cluster, and thus if their population history is mostly a recent phenomenon of trans-Wallacean population movements, or if they are mostly descended from the people that lived there for tens of thousands of years ago."

What do you mean by a true, in depth analysis?

By what techniques and methods do you perform a "true, in depth analyis" on a population that presents itself as a cluster.

Also, what are the assumptions that a program like STRUCTURE uses to locate a cluster in the first place? It's looking at a subset of locations on the whole genome. How are these points chosen?

I asking because this seems like a cutting edge and very interesting area of genetic anthropology. For those of us who are not in the field, it would be nice if some of you could shed some light on this.

Maju said...

"In fact your much-prized HUGO diagram actually has very few examples from Wallacea, and the red doesn't appear in them".

First of all I was not talking of Wallacea. Second the red component does appear among the Manggarai of Flores (though it's likely an import from Sundaland). Third, there are five samples from Wallacea, all high in the Melanesian component and clustering with Melanesians in the ML tree.

"From what I can make out most of the Austronesian samples are actually from what had been Sunda".

Sundaland (Sunda is a chain of islands in both Sundaland and Wallacea). There are many West Indonesian, Malaysian and Filipino samples and "only" five Wallacean samples but it's a lot for an All-Asia study.

I don't understand what's the problem here. SE Asia is very well sampled overall and Wallacea is no exception.

"What's more the Papuan genes are by no means confined to Eastern Indonesia".

Practically so. The Toraja are from Sulawesi. Only very tiny fractions (practically "noise") appear in two Filipino samples: one Negrito and the other Austronesian "urban". Not anything meaningful in any case.

"The Mentawai are probably mainly descended from the pre-Austronesian inhabitants. Just because they're mainly green in the HUGO diagram doesn't mean they're 'pure' Austronesian. The green could mean almost anything".

I don't think you can make that claim:

- Almost 100% 'bright green' component: Taiwan Aborigines, Filipino Austronesian, Mentawai.
- Strong 'bright green' component: most other Austronesians (often with red component), Kradai and Southern Han (with 'dark blue' and often also some 'red' or 'yellow' components).

If not strictly Austronesian it looks at least Austronesian-Kradai. The purest examples are all Austronesian anyhow and, excepting the Mentawai, from the widely acknowledged Austronesian homeland at Taiwan-Philippines.

So you are just repeating a preconception without looking at the data. Oddly enough, you are doing it even if what I say would seem to favor some of your thesis: a large demic Austronesian migration from somewhere in your "Greater Wallacea" geographic concept (Philippines to be precise) and even to the presence of some desert islands settled ex novo by the original Austronesian immigrants (Mentawai archipelago). So I am starting to suspect that you are just arguing for the sake of it and because you don't want to take 30 mins looking at what the heck the HUGO K=14 graph may mean.

"Looking for 'pure' Austronesians is like looking for 'pure' Indo-Europeans, or 'pure' Baques".

That's a funny blanket statement but nothing else. So Taiwan Aboriginals are then? What about Austronesian Filipinos? Where's the admixture in them? And in the Mentawai?

K=14 is a good depth (I only recall one even deeper on Native Americans). All TA, some Filipino and the Mentawai are >90% that component. The other Austronesian Filipinos have some "Chinese" (probably) component (blue-yellow) but are otherwise in the same situation. It surely represents well to the ancestral Austronesians, even if its presence in South China indicates some other connections as well (Kradai).

It's not any "pet theory" (I might have preferred other results), it's what I see on the actual data.

Maju said...

Sparked by this discussion, further analysis of my own on the HUGO data at Leherensuge now.

terryt said...

"It's not any 'pet theory' (I might have preferred other results), it's what I see on the actual data".

Again, it's taken a long time.

"a large demic Austronesian migration from somewhere in your 'Greater Wallacea' geographic concept (Philippines to be precise) and even to the presence of some desert islands settled ex novo by the original Austronesian immigrants (Mentawai archipelago)".

Sorry. I didn't realise you'd finally agreed with me. I wouldn't claim precisely the Pilippines. The weight of evidence actually suggests the Austronesians developed from a complex interplay of humans within and around Wallacea, including Taiwan to the north. And I'd be fairly sure of other 'desert islands settled ex novo' than just Mentawai. From the paper:

"A remaining question is why Asian-Melanesian ancestry changes over such a small area in eastern Indonesia. It may mark the region where indigenous Papuan groups were large enough to resist incursive Austronesian populations during the Mid-Holocene".

Almost certainly so. But that in turn begs the question as to how they could occupy Wallacea so completely. Was it mostly unoccupied?

As for the Mentawai and their hair: I don't know here you got the idea they all have curly hair. I've looked at many photographs on the Net and see none with curly hair. They all have relatively straight or wavy hair. And this link is interesting:

http://www.mentawai.org/histbackgr.htm

The author quotes from a 'John Crisp a civilian employee in the service of the English East India Company visited the Pagai ("Poggy") islands in August 1792'. He wrote:

"we should naturally expect to find their inhabitants to be a set of people originally derived from the Sumatra stock, and look for some affinity in their language and manners; but to our no small surprise, we find a race of men, whose language is totally different, and whose customs and habits of life indicate a very distinct origin and who resemble more the inhabitants of the late discovered islands in the great Pacific Ocean".

So he saw them as basically Polynesians. I'd guess the Polynesians would show up green as well. The 'original' Austronesian genes in the region between Mentawai and Polynesia have been diluted by the addition of Melanesians and yet more mainland Asians.

"So Taiwan Aboriginals are then? What about Austronesian Filipinos? Where's the admixture in them? And in the Mentawai?"

Surely you're not claiming all these groups share the same haplogroups. They're actually a diverse bunch genetically.

"The purest examples are all Austronesian anyhow and, excepting the Mentawai, from the widely acknowledged Austronesian homeland at Taiwan-Philippines".

On what grounds would you except the Mentawai as coming from the Austronesian homeland?

aargiedude said...

"Iberians have 1% to 2% mtdna L"...

MtDNA L does not automatically mean "sub-Saharan": there are many L lineages that look more typically North African or West Asian than anything else and must have been there since about the OoA time or just a little later.


Maju, you're preaching to the choir. Maybe I should have used a made-up example.

Why does a distinct cluster necessarily indicate descent from an aboriginal group?

EDIT: Maju said it better, but ok, here's my take:

It doesn't necessarily mean such a thing (*), but my point was that we should first fully analyze the Wallacean people to determine if their population history is indeed, as the authors simply assumed, a mix of 2 Neolithic population movements from neighboring west Indonesia and New Guinea, or wether they have a genetic profile that indicates the existence of a 3rd component (aka cluster), which would be aboriginal ancestry, and could in fact be nothing less than the vast majority of their ancestry. Till we look, we won't know, and this study didn't look.

(*) For example, Ashkenazis would very possibly form their own genetic cluster (or they have, already, in some autosomal study that's slipping my mind, right now?), because of their well known population bottleneck 1,000 years ago. Obviously, this doesn't mean there used to be an ancestral Ashkenazi population existing thousands of years previously. And there are many cases in which "fake" clusters have popped up which do not indicate an ancestral population component, such as in Tishkoff's study of Africa, in which 2 clusters were very likely the result of recent genetic drift (a small isolated city in the Sahara and a small hunter-gatherer population in Tanzania), or in the earliest of Rosenberg's autosomal studies, in which the Kalash of Pakistan, an isolated, small mountain population, formed their own cluster (later disappeared completetely when Rosenberg used 600K SNPs on the same data). If we studied every little isolated village in Italy (or Albania, or Norway, whatever) we might come up with thousands of clusters! In fact, that's what happened in a 2008 autosomal study of Sardinia. There were 8 isolated mountain villages, and they formed 8 clusters.

There's a good chance many of the islands had become uninhabited by the time of the Austronesian expansion

Wow, please tell me what is indicating this. [about the Wallacean islands]

Not a 'small amount of ancestry'. Virtually totally East Asian and Papuan. Not Australasian.

Australasia includes Australia and New Guinea. The point of my posts was to explain why I think they erred and have missed the potential existence of a 3rd genetic cluster, which would be the aboriginal component, by assuming there is none and separating them into East Asian and Australasian. Your counter-argument is to tell me the study found them to be a mix of East Asian and Australasian. You don't get it.

There were no people in Australasia at the time, except for the Australian part, and they didn't contribute to the Wallacean population at all.

You think Australasia is the Wallacean islands, so you're saying there were no people in the Wallacean islands. And that the Australian part didn't have anything to do in this, so their Melanesian ancestry is from New Guinea. Fine. So the Wallacean islands became completely uninhabited, but New Guinea didn't. >>> Something terrible affected the Wallacean islands but not New Guinea. And you're going to tell us what that was, cause I can't imagine.

terryt said...

"Wow, please tell me what is indicating this. [about the Wallacean islands]"

Several thing suggest it. But I've just provided an example.

"Australasia includes Australia and New Guinea".

The term 'Australasia' normally refers to New Zealand and Australia. It is sometimes used to include New Guinea, but not commonly, and not usefully. There were no people in New Zealand until about 800 years ago. So, yes, it was unpopulated until then.

"You think Australasia is the Wallacean islands"

They're miles apart and unrelated. Don't you know your pacific geography?

"Your counter-argument is to tell me the study found them to be a mix of East Asian and Australasian. You don't get it".

There's no evidence at all of a 'third' population in the region. Presumably that's why the authors ignored the possibility.

"Something terrible affected the Wallacean islands but not New Guinea. And you're going to tell us what that was, cause I can't imagine".

You are obviously unaware that sea level rose considerably around 10,000 years ago and isolated islands, and reduced their area, limiting possible population numbers. People survived on New Guinea because it's rather larger than most Wallacean islands.

Maju said...

"I don't know here you got the idea they all have curly hair".

I happened to watch a documentary on the Mentawai last Saturday. At first I was not aware of who were this isolated people but then I realized they were the odd western "pure" Austronesian offshot. Soon I realized that their often curly or wavy hair, often not different from my own in color and texture, did not match with some typical beliefs, including your own.

Your link does not seem to contradict this (no images).

"Surely you're not claiming all these groups share the same haplogroups".

I'm not talking haplogroups here but autosomal genetics. These two things can be quite different (in Asia, in Europe or in Africa).

"On what grounds would you except the Mentawai as coming from the Austronesian homeland?"

I'm not doing that. I'm saying that, per all accounts, the ancestral Austronesian urheimat was not at Sumatra but in the Taiwan-Philippines area. This makes sense to me on light of the genetic data. I could also think of a wider Austronesian homeland in all Sundaland-Philippines-Taiwan but then there was no Austronesian expansion in ISEA and would be hard to explain some stuff.

...

Australasia (Wikipedia): "Australasia is a region of Oceania: Australia, New Zealand, the island of New Guinea, and neighbouring islands in the Pacific Ocean.The term was coined by Charles de Brosses in Histoire des navigations aux terres australes (1756). He derived it from the Latin for "south of Asia" and differentiated the area from Polynesia (to the east) and the southeast Pacific (Magellanica). It is also distinct from Micronesia (to the northeast)".

In other words, it's the main landmass area of Oceania: Australia and surroundings. For our purposes it can be considered roughly the same as Sahul (and you can save us your misplaced Kiwi pride).

...

I cannot agree with your theory of the deserted islands either, Terry. Maybe there were some of those but it's clear that people can survive and surely survived in small archipelagos too. We would have to evaluate each case on their own merits but there are thousands of such small islands in the area... too much. All I know is that in Wallacea the native Melanesian blood survived well.

Onur Dincer said...

Sidenote: From now on I will always use the word "race" only for biological archetypical races (human or not), i.e., subspecies; for lesser biological races (again human or not) I will always use the word "subrace", or in the case of biological archetypical hybrids, "hybrid race" or "racial hybrid".

Onur Dincer: "I prefer much more clearly defined terms like "race", "subspecies", "subrace", "hybrid race", "racial hybrid" for various human groups, while you prefer vague words like "human group", "population", etc. for the same groups. Apart from these differences in terminology, we are actually saying mostly the same things."

Luis Aldamiz: "And how are these terms "clearly defined"? They are not: they are very diffuse concepts. So it's not an issue of words only but of their definitions, which I say don't exist. That's what makes race an almost totally subjective matter."


The definition of subspecies (race) can vary across different types of species (sexual vs. sporodic, unicellular vs. multicellular, etc.), but it is more or less fixed and consistent among the same or similar types of species. As humans are mammals, the most appropriate thing to do would be to compare human subspecies (races) with subspecies of other mammals. Many biologists have done just that and have come to the conclusion that there is no qualitative difference between modern human subspecies and subspecies of other mammals. I don't think you have objection to the existence of subspecies in other mammals, do you?

Onur: "Also I wouldn't normally (never for archetypal races) use quotation marks in the sense of emphasizing the artificialness of a category while mentioning race/subrace names like "Caucasoid", "Mongoloid", "Melanesian", etc., while you usually do that."

Luis: "That's because I have no reason to think they really mean anything beyond pretty much diffuse clusters of genetic/phenotype variance. So it's not like they are absolute categories but rather common places.

A much more solid category is something like "people with black curly hair". This is an specific clearly measurable phenotype (and possibly a single genetic clade too) but it goes across "races" and defines none of them. To make a "race" you have to add and subtract and average and reduce... it's an statistic simplification. I see pointless to even discuss its relevance because it has none. It may be convenient to talk of Caucasoids or Melanesians now and then but we must not forget that they don't really exist as absolute categories but only as statistic simplifications, maybe useful but always artificial constructs."


Why the obsession with absoluteness?

Onur Dincer said...

Onur: "Surely all races and subraces, or with your terminology, human groups formed somewhere, didn't they?"

Luis: "What I say is that the formed and are still forming as we speak (they don't remain static but change) within pretty much separated world regions by means of "stirring up" these pots once and again. Each "race" is a homogenized puree of as many ingredients as you can imagine, not the product of the expansion of a single imaginary group living once in some mythical Land of the Hyperboreans or whatever."


Even you yourself admit that the human races are pretty much confined (leaving aside the long distance migrations within the last few thousand years with the advances in mobility and transportation) to specific regions. So what is all the fuss?

Onur: "In conclusion, our main point of contention is our different terminological preferences. But this is no less a significant issue, as the biological human races (subspecies) are scientifically valid entities."

Luis: "They have only statistical value. You can easily find a zillion of those averaged components that cross all "racial borders" or at least many of them. And this is not because of admixture between "pure races" (at least not in most cases) but because those phenotypes pre-date the process of homogenization and remain within them at higher or lower frequencies, the same that the onion in the lentils puree is the same as the onion in the pumpkin one, regardless that the overall purees are different."


It is true that most human phenotypes pre-date modern human races (that is very normal when we compare the ages of modern human races with the age of our species, also many phenotypical differences you speak of pre-date our species). Even these common phenotypes in a great many cases have significant frequency and even genotypical differences between races (sure, also within races, but intraracial differences are most of the time (in genotypical differences, in the overwhelming majority of the cases) less (in a great many cases, much less) than interracial differences).

Onur Dincer said...
This comment has been removed by the author.
Onur Dincer said...
This comment has been removed by the author.
Onur Dincer said...

Correction: "From now on I will always use the word "race" only for biological archetypical races (human or not), i.e., subspecies; for lesser biological races (again human or not) I will always use the word "subrace", or in the case of biological archetypical hybrids, "hybrid race" or "racial hybrid"."

I should have written "archetypal", not "archetypical" (I am sure Marnie will now get angry with me :-D).

Onur Dincer said...

To onur, I think, who responded to me that the people of Wallacea can't be anything other than a mix between Australasians and East Asians.

I don't remember to have said such a thing (but of course that doesn't mean I necessarily disagree with that statement), so it should be someone else.

Maju said...

"I don't think you have objection to the existence of subspecies in other mammals, do you?"

No and yes. It's not a consistent, clearly defined, concept.

According to your criteria there is something like "Homo sapiens caucasicus" or "Homo sapiens mongolicus" but I've never seen such taxonomy anywhere. At most the, now abandoned, "H. sapiens sapiens" and "H. sapiens neanderthalensis" dichotomy.

In any case, the use of the subspecies category is arbitrary and not really regulated nor systematic, unlike higher taxons.

"It is true that most human phenotypes pre-date modern human races"...

That's a good reason to doubt the validity of the subspecies category.

"I will always use the word "subrace", or in the case of biological archetypical hybrids, "hybrid race" or "racial hybrid"".

There's no biological concept of "subrace" or "sub-subspecies". The scale ends at the subspecies level, which in itself is not a rigid category but one of mere convenience and subjective usage.

All this issue is discussed at length at Wikipedia-Race-Subspecies, where there are two contention points:

1. Fst index, which is 0.15 for interpopulation variation and only 0.1 for intercontinental variation among humans, much lower than among most mammal species.

2. The Russian doll structure of the human genome, with Africans holding all or nearly all the diversity (not a subspecies) and extremely peripheral populations having as little as 70% of it maybe.

Considering only point 2, we would have only two subspecies: H. sapiens africanus and H. sapiens asiaticus (if anything). But the latter is a subset of the former and not really a bifurcation in equal terms and, in turn, it's divided in several commonly accepted "races".

The weight on whether one accepts or rejects "races" as "real" is not an objective measure, that is clearly most elusive, but subjective factors. For example, people raised in a racially stratified society such as the USA are more likely to believe in the reality of race. It is therefore largely a social, ideological and political construct.

Onur Dincer said...

Luis, your surname sounds Arabic (starting with 'al-' like "Alhambra"), is it because your ancestors were Arabic-speaking (and very probably also Muslim, but most probably from the native genetic stock) during the Muslim rule of Iberia? Do you have any info about your medieval roots? Maybe you have Banu Qasi (a Basque Muslim emirate from the early Muslim era of Iberia) roots. Alternatively you may have medieval non-Basque roots from the much more Islamized (both in numbers and duration) central and/or southern part of medieval Iberia.

If you ask my roots, I can't go beyond even the 19th century, as, unlike Catholic/Protestant Europeans, who have had ecclesiastical birth, baptismal, marriage and death/burial registers for centuries (going to the medieval times in some places), Muslims (as far as I know also Orthodox Christians and Oriental Orthodox, though Dienekes and Marnie should know them better) never had any similar (even remotely) personal registers (whether secular or religious, or whether for taxation purposes or not) until the beginning of the eastern and southern Mediterranean modernization during the 19th century.

Onur Dincer said...

Btw, we have Ottoman tax registers beginning from the late 15th century, but they don't contain any personal or familial information, not even the number of people in a given region, they only contain information about the number of taxable "households" (unfortunately not a household strictly speaking, but a unit of taxable community somewhat vague and difficult to explain), their religion and information about whether they are sedentary (most of the time) or nomad (much less and limited to certain territories).

As I told in my previous post, these all changed with the adoption of Western style registers, which included personal, familial and even occupational information, by the Ottomans sometime during the first decades of the 19th century.

Onur Dincer said...

Luis, I already read that Wikipedia article. I suggest you to read (if you haven't already) "The Race FAQ" article of John Goodrum, "Human Genetic Diversity: Lewontin's Fallacy" paper of A.W.F. Edwards and "Race: a social destruction of a biological concept" paper of Neven Sesardic (all of them freely available in the Internet).

Onur Dincer said...
This comment has been removed by the author.
Onur Dincer said...

Correction: "is it because your ancestors were Arabic-speaking"

I should have written "some of your ancestors", especially as most European surnames, because of their ancientness, represent only an extremely minuscule fraction of ancestry.

Maju said...

There are many surnames and toponyms that begin or end with alde/alda, which means 'zone, area, part' or, when suffixed, also 'near, surrounding area' in Basque: Aldaiturriaga, Aldekoa, Aldonza, Aldama, Aldaitz... Errekalde (often hispanized as Recalde), Olalde, Elizalde or Elexalde...

My surname is as Basque as can be. I have diverse ancestry but not by my father's line.

Anyhow, I don't know why my private data should be subject to scrutiny here. I'm rather wary of spreading personal information through the net, and that's why I use a pseudonym.

Onur Dincer said...

I have diverse ancestry but not by my father's line.

So if I understand you correctly, you aren't "pure" Basque in your known family history, right?

Anyhow, I don't know why my private data should be subject to scrutiny here.

Like Gioiello, I don't think there is anything to hide in the Internet unless of course it might really harm you or someone else. I see no harm for you or anyone else in your revealing of your family history and ethnic and racial origins.

Maju said...

Well, there are very serious concerns about privacy on the net. Not just government agencies can spy on you but a host of unscrupulous private organizations can do as well in order to steal your money and other dangerous stuff.

You are very free to broadcast your own private data to the universe but I wish to keep some privacy and I also think it's something everybody should at least consider.

And sure: I'm not "purebreed Basque", though I doubt it matters at all.

Onur Dincer said...

And sure: I'm not "purebreed Basque", though I doubt it matters at all.

In a European or any White context, it might matter if all or part of your "non-Basque admixture" is from a non-White (cultural and/or biological) population. That is why I am especially curious about your non-Basque ethnic and racial origins.

terryt said...

"Your link does not seem to contradict this (no images)".

Just google 'Mentawai people'- images, and have a look. Pages of them. They're as close to Polynesians as any people I've seen outside Polynesia.

"the ancestral Austronesian urheimat was not at Sumatra but in the Taiwan-Philippines area".

I don't think anyone has ever claimed Sumatra as a place of origin. Sumatra is part of ancient Sunda so it has a different history. I think you'll find the following link to be an excellent summary of Pacific settlement:

http://www.plosgenetics.org/article/info:doi%2F10.1371%2Fjournal.pgen.0040019#pgen-0040019-sg001

"For our purposes it can be considered roughly the same as Sahul (and you can save us your misplaced Kiwi pride)".

Surely 'Sahul' is a far better word for what we're talking about here. Australasia includes far too big a region to be useful in this context. And 'Australasia' is in fact much more often used just for NZ and Australia. Nothing to do with 'pride'. Note, further on in the Wiki article: 'Geopolitically, Australasia is sometimes used as a term for Australia and New Zealand together, in the absence of another word limited to those two countries. Sometimes the Island of New Guinea (including Papua New Guinea and the Indonesian part of the island) is encompassed by the term'.

"All I know is that in Wallacea the native Melanesian blood survived well".

Not so, evidently. I'm sure I put this link up somewhere:

http://books.google.co.nz/books?id=4obAfGBGKY0C&pg=PA8&lpg=PA8&dq=bellwood+wallacea&source=bl&ots=qPhsWd1pja&sig=CWtI_oGbrMz3u0AyhykHQzP8efY&hl=en&ei=4aLXS6SSDoTCsgOTuK3rBQ&sa=X&oi=book_result&ct=result&resnum=2&ved=0CAgQ6AEwAQ#v=onepage&q=bellwood%20wallacea&f=false

It's a long extract but I'm sure you'll find most of it informative. Several comments he makes in the book suggest a very sparsely inhabited pre-Austronesian Wallacea (which suggestion I've also seen made over the years). Bellwood claims that even the large island of Borneo was sparsely inhabited (if inhabited at all) before the Austronesians arrived. He attributes this to the heavy jungle cover, and has some interesting comments to make on the subject. He also mentions the fact that, apart from parts of the Philippines, no Negrito groups are found in Wallacea (or in Borneo for that matter). Further, he shows that linguistic evidence indicates that the Papuan languages scattered through Eastern Wallacea diverged from the New Guinea languages relatively recently, possibly even post-Austronesian expansion. So the 'native Melanesian blood' in Wallacea hardly represents ancient survival.

The fact that the Austronesian languages are spread widely through previously uninhabited islands, from the Central Pacific to Madagascar, indicates they had developed a superior boating technology enabling them to finally reach such islands. The suddenness of their expansion suggests to me they had already discovered the benefits of exploiting such uninhabited islands.

As to the possibility of human extinction on islands. The phenomenon is certainly accepted for many Pacific islands, notably Norfolk, Pitcairn and many of the Phoenix and Line islands (along with many others). I've also seen it claimed for Indonesian islands but I don't know which particular ones.

terryt said...

"In any case, the use of the subspecies category is arbitrary and not really regulated nor systematic, unlike higher taxons".

Yes. I've actually read a paper on dabbling ducks that divides then into 'species', 'subspecies', 'superspecies', 'allospecies', 'subgenus', 'supergenus' and 'infragenus'. Wow.

"That's a good reason to doubt the validity of the subspecies category".

Not really, because that's probably the case with other subspecies as well.

"Fst index, which is 0.15 for interpopulation variation and only 0.1 for intercontinental variation among humans, much lower than among most mammal species".

But what about mammal subspecies?

"Considering only point 2, we would have only two subspecies: H. sapiens africanus and H. sapiens asiaticus (if anything)".

African vary geographically a great deal over the continent. The variation has been diminished by the recent migration of Bantu people of course. And I agree that this sort of migration is unusual in other species.

Maju said...

"In a European or any White context, it might matter if all or part of your "non-Basque admixture" is from a non-White (cultural and/or biological) population".

In the USA they wanted to classify me as "Hispanic" just because my passport reads "Spain". I was so outraged at the very fact that they classified people by race that I would not really dispute the matter. Eventually the bureaucrat (a High School teacher who should know better - but, well, their level of HS is like our levels of primary school or almost) realized that Spain is not some place south of Mexico.

Anyhow, among Basques, it's primarily language what matters for identification purposes. The world is divided between euskaldunak (Basque speakers or "those who have the Basque way" - 'euskal-' possibly 'holding or supporting way/language') and erdeldunak (half speakers or "those who have the half way"). We and them, like everywhere, but it is not a matter of mere ancestry but of language and identity.

"That is why I am especially curious about your non-Basque ethnic and racial origins".

From northern Spain and northern Italy. You're Turkish right?

Maju said...

"But what about mammal subspecies?"

For what I could understand it seems to be higher in most cases (0.25 or higher). 0.15 was mentioned as pretty low, 0.10 as clearly substandard, with no comparable case.

But you tell me. I'm not the one so interested in subspecies.

"I've actually read a paper on dabbling ducks that divides then into 'species', 'subspecies', 'superspecies', 'allospecies', 'subgenus', 'supergenus' and 'infragenus'".

You will notice that all those categories are higher than species, excepting 'subspecies' of course.

"African vary geographically a great deal over the continent".

Sure. But when you make a ML tree they cluster together vs. Eurasians (excepting North Africans, of course). So the high level division within Humankind, if any, should be that one. Fst is still too low though.

African Hunter-Gatherers vs. Bantus have one of the highest inter-group Fst (0.11 if I'm correct). But it's still very low and anyhow we are always talking of present populations, as it'd be a lot harder to judge fossil ones.

"And I agree that this sort of migration is unusual in other species".

Maybe not so unusual. In the latest PLoS ONE there's a paper on European wolf genetics and how the North American haplotype has apparently spread over the native European one. But gray wolves are a more mobile species than humans and anyhow are a subspecies themselves (of Canis lupus).

Btw, domestic dogs, with all their colorful diversity, are just one subspecies by all standards (Canis lupus familiaris). So visible phenotype and classification as subspecies are not so simplistically linked as some have suggested.

Onur Dincer said...
This comment has been removed by the author.
Onur Dincer said...

You're Turkish right?

Yes. As far as I know, during my known family history (as I told you, it doesn't go back even 200 years), all of my ancestors were Turks (i.e., Turkish-speaking Muslim). I say "as far as I know" because even my 19th century and early 20th century family past is somewhat blurry in both of my parents' lines (especially my mother's line as they came to the territory of modern Turkey as refugees from the Balkans in the early 20th century) because of the turbulent events and migrations (mostly forced) within and in the environs of the Ottoman Empire in those times.

terryt said...

"But when you make a ML tree they cluster together vs. Eurasians (excepting North Africans, of course). So the high level division within Humankind, if any, should be that one".

Now that's very convincing evidence in favour of the concept that non-African humans are just a small subset of Africans, not a representative cross-section. So the original OoA is unlikely to have carried the whole human variation we see today.

"But gray wolves are a more mobile species than humans and anyhow are a subspecies themselves (of Canis lupus)".

A race? Thanks for the link.

Maju said...

No, because ML trees don't describe (approach) variation but closeness. Get your facts straight.

terryt said...

"You will notice that all those categories are higher than species, excepting 'subspecies' of course".

True. But there's a problem. Many of the species are easily capable of forming fertile hybrids. And, in some cases even members of different superspecies can do so.

"No, because ML trees don't describe (approach) variation but closeness. Get your facts straight".

So obviously Eurasians are in no way a subset of Africans. The two groups are completely separate.

By the way. I remember you and Ebizur commenting on the separate genetic nature between northwestern and southeastern Borneo. Can you remeber the main points of differentiation?

Maju said...

"So obviously Eurasians are in no way a subset of Africans. The two groups are completely separate".

No. That's an illusion of statistical approaches. Simplified:

Stage 1: Africans > Eurasians (Eurasians are a subset with reduced genetic diversity).

Stage 2: Africans homogenize, Eurasians diversify as they expand (various subsets)

Result: Eurasians are still a subset of Africans and have several subsets within them. Africans appear as more homgoeneous than they used to be because of persistent intra-continental gene flow.

So you get various clusters but that doesn't mean that one cluster is not a subset of the other. It is.

Just that if Eurasians have allele A in apportion of 90% and Africans have it in 10%, then Statistics will consider it an "Eurasian" gene but it's ancestrally African anyhow. Same for all other bottlenecked alleles.

"By the way. I remember you and Ebizur commenting on the separate genetic nature between northwestern and southeastern Borneo. Can you remeber the main points of differentiation?"

No, sorry.