Monday, July 20, 2009

172. Articles Now Available for Download

I'm pleased to announce that Amand Aglaster, director of VWB-Verlag, publisher of the journal The World of Music, has given permission for me to make my essay, Echoes of Our Forgotten Ancestors, freely available for download via this blog. Just click on the title and select "Save Page As . . ." If you'd like to order a beautifully printed and bound hard copy of the entire issue, you may do so from the following website: http://www.vwb-verlag.com/Katalog/m748.html.

A preprint version of my most recent paper, Some Notable Features of Pygmy and Bushmen Polyphonic Practice, with Special Reference to Survivals of Traditional Vocal Polyphony in Europe, is now available for download as well, by permission of the editors of the Proceedings of the Fourth Annual Symposium on Polyphonic Music, held last year in the Republic of Georgia, where it is scheduled for publication next year.

Finally, I'll remind you that my essay New Perspectives on the Kalahari Debate, published a couple years ago in the anthropology journal, Before Farming, can also be downloaded via the link provided above. Thanks again to the publisher WAS Press, for permission to freely share this work.

26 comments:

Maju said...

Fantastic!

I'm right now reading the first one and I can only warn you against taking Oppenheimer too seriously. He's not even a geneticist and certainly some of the theories he presents are at least controversial.

Nowadays, I'd say it's mainstream the understanding that there was "rapid coastal migration" and that South Asia is either the main or one of the main (together interactively with SE Asia) urheimats of Eurasian humankind. South Asia has highest diversity for 2/3 non-African macro-haplogroups (M and R) and for the main Y-DNA macro-lineage (F), the rest can confidently be thought as to have spread from SE Asia (South China included).

There is archaeological evidence of pre- and post-Toba continuity in Jawalpura (though we don't know for sure if the authors of that culture were AMHs or some other hominin) and anyhow such a catastrophe as Toba should have caused global effects. If a much smaller volcano could cause "the year without summer" and major starvation in Europe and elsewhere, imagine the chaos that Toba ensued.

In fact most geneticists tend to think of the OOA migration as post-Toba, though I'm undecided.

But what I totally disagree with is with claiming the North Caucasus as some sort of craddle of Europeans or West Eurasians. That goes against all evidence, which suggests migration from West Asia via Anatolia and the Balcans, with Central Europe (Moravia and surroundings) playing the main role in the genesis of finished Aurignacian and Gravettian cultures. No archaeological trail will ever bring you the North Caucasus.

This doesn't mean that the area cannot be refuge of ethno-cultures in other times more widespread. Personally I speculate with NE Caucasians (Chechens, Ingush, etc.) as being remnants of East Epi-Gravettian, which expanded into the Caucasus and Zagros area in the Epipaleolithic. That would explain their alleged linguistic affinity with Hurro-Urartean.

But you are talking of "Russians" anyhow, not Caucasian peoples. And it would seem that the pan pipe at least is not spread through the Caucasus but Ukraine, Rumania and some areas of North Italy (not sure if the Greek syrinx is still used). This could mean a Gravettian origin hypothetically or even a Balcanic Neolithic or even Indoeuropean one as well. But in any case it does not seem to originate in the Caucasus but one would have to look for a lost West Asian ultimate origin or, if it is Indoeuropean, maybe a Central Asian one.

DocG said...

I happen to like Oppenheimer, because 1. he's not afraid to think big and speculate, which is what I also like to do; 2. his speculations make a good deal of sense; 3. his book is very thoroughly researched, with a huge reference list not only from genetics but archaeology as well -- in fact most of what he has to say, even the most controversial things, can be traced to very knowledgeable and reliable sources.

What particularly appeals to me in Oppenheimer's book is how much sense so many of his arguments make in terms of what is known about the musical evidence. His speculations have tremendous explanatory power in that respect, so as far as I'm concerned, it's very likely he's on the right track.

As far as Toba is concerned, he is the only one to really take this idea and run with it. And what he has to say makes much more sense to me than what Ambrose has to say, since it seems highly unlikely to me that the Toba event could have had much impact as far away as Africa.

Toba, as Oppenheimer sees it, has too much explanatory power to dismiss. He could be wrong, but if he is, then we are left scratching our heads, looking for some other catastrophe (perhaps a Tsumani?) that could account for all the genetic diversity centered in India and other points west of that volcano. And why it is that we find so many "African signatures" east of Toba and so little of that sort of thing west of it. (The prevailing winds blew west.)

DocG said...

It's true that "There is archaeological evidence of pre- and post-Toba continuity . . ." but that by no means precludes the extreme genetic bottleneck effects such a violent event would have caused. The tools found in the upper strata would have been those of survivors in desperate straits, or else visitors from an unaffected region, not people carrying on with life as usual in the same place. As I recall, we are talking about an 8 ft. thick layer of volcanic ash. How could that not have produced extreme effects on every life form affected by it?

What's especially telling about that discovery is Petraglia's strong opinion regarding the nature of the tools he found, as most likely produced by AMH. If he's right, then AMH were in fact present in that region and most certainly affected by Toba, which is huge.

"anyhow such a catastrophe as Toba should have caused global effects. If a much smaller volcano could cause "the year without summer" and major starvation in Europe and elsewhere, imagine the chaos that Toba ensued."

There is no controversy whatsoever about the Toba explosion, it's tremendous power and the huge extent of its "fallout." Toba happened! Which means there most certainly would have been disastrous affects, if not global then certainly for thousands of miles to the northwest. If the "global effects" are not yet apparent, that doesn't mean they aren't there. There DOES seem to be evidence of a population bottleneck among tigers in India around the same time as Toba. Whether other studies involving other life forms have been done I don't know.

We have no way of knowing for sure whether modern humans were affected by Toba, true. But if not then we are faced with the problem of accounting for not only the musical evidence, which no one but me seems to care about anyhow, but also the sudden explosion of genotypic and phenotypic diversity that must have occurred at some point during the earliest phase of the OOA migrations.

DocG said...

"But what I totally disagree with is with claiming the North Caucasus as some sort of craddle of Europeans or West Eurasians."

This is not Oppenheimer's theory, though it clearly works for him. He cites research by "a team of Estonian geneticists" led by Villems and Kivisild, who argue simply that the Caucasus was a staging ground for "several important early European maternal clans," including those marked with the HV haplogroup. There will always be a certain disconnect between the genetic and archaeological evidence anyhow, because for one thing the latter is extremely fragmentary and clearly incomplete.

As far as I'm concerned, it makes no difference anyhow, as I have no interest in determining any of the specific paths or origin points of any European group, but only in attempting to speculate regarding the big picture. Georgia, in the Caucasus, happens to be an extremely important musical refuge area for practices that as I see it clearly have African roots, so it makes sense to me that the Caucasus could have played a major role in the earliest migrations. On the other hand, the Caucasus could simply, as you say, have become a refuge area, where people with conservative cultural values retreated at some point.

As far as panpipes are concerned, as I understand it, there is evidence that they were of importance in this region at some time in the recent past, though that tradition seems to have been lost. And we need to remember that there is a big difference between panpipes as diffused among tribal peoples during the Paleolithic and the same instrument as diffused via ancient Greece, Rome, India, etc., during a much more recent period.

Maju said...

I do not mind considering that humans were in Asia before the Toba event, of course (I am not sure but could make some sense), but you still have that most of Eurasian (incl. Oceanian and American) genetics belong to macro-lineages that have their highest diversity by far, and therefore their likely origin in South Asia. These are Y-DNA F and mtDNA M and R. Only mtDNA N(xR) and Y-DNA C and D can be argued to have another point of origin, most probably SE Asia. F-derived K is controversial but could also have a SE Asian origin maybe but its immediate ancestor IJK does seem to have also a more western origin.

What I argue is that, probably after Toba, the main source of Eurasian genetics was still in South Asia and that at least South Asian mtDNA lineages do not show the bottleneck that Oppenheimer argues for. My impression is instead that other lineages had a hard time expanding in South Asia because mtDNA M was already too widespread. They found more room to expand in the East and later also in the West instead.

I think you're relying too much on one interpretation that may not necesarily be the correct one and hence linking the fortune of your exploration to that of Oppenheimer.

But well...

This is not Oppenheimer's theory, though it clearly works for him. He cites research by "a team of Estonian geneticists" led by Villems and Kivisild, who argue simply that the Caucasus was a staging ground for "several important early European maternal clans," including those marked with the HV haplogroup. There will always be a certain disconnect between the genetic and archaeological evidence anyhow, because for one thing the latter is extremely fragmentary and clearly incomplete.

Not in Europe. In other areas there are way too many blanks to be filled but in Europe the overall picture is quite well known and it's clear that the Caucasus played no role (in fact it was one of the last Neanderthal refuges, I was just reading today). Also the position of the Caucasus in the genetic history of Europe has been pretty much downplayed as of late. More important at some moment may have been the area south of it, including Anatolia and maybe Kurdistan.

Georgia, in the Caucasus, happens to be an extremely important musical refuge area for practices that as I see it clearly have African roots, so it makes sense to me that the Caucasus could have played a major role in the earliest migrations.

Well, notice anyhow that Georgia is in the South Caucasus region, i.e. West Asia.

As far as panpipes are concerned, as I understand it, there is evidence that they were of importance in this region at some time in the recent past, though that tradition seems to have been lost. And we need to remember that there is a big difference between panpipes as diffused among tribal peoples during the Paleolithic and the same instrument as diffused via ancient Greece, Rome, India, etc., during a much more recent period.

I don't know how would you make the difference. Ukrainians may have got it from Greeks or Colquidans (West Georgians) within the Greek trade area of the Black Sea for example. Inversely, the transmission could have been in any other direction, for instance if it'd have been transmitted by Indoeuropeans (Ukranians to the rest) or maybe Neolithic diffusion (West Asians to the rest).

What we know of the panpipe in Greece is that it was considered a rural instrument and is most directly associated with archaic divinities such as Pan or with ancient cultural remnants like the so much idealized Arcadia. So guess it should be old in Greece too and therefore we could discard up to a point the IE possibility. But I don't see clear how Occam's razor could cut between the Paleolithic/Neolithic origin options.

And these imply different scenarios: Paleolithic would mean a European, probably Gravettian pervivence, while Neolithic instead would rather suggest a West Asian origin. Neither is attested so far.

Maju said...

Anyhow, just throwing food for thought. I really like your exploration, as you know, just that some hypothesis within it I don't find too convincing.

I made a link to this post from my blog, so my readers can download your research and also have an opportunity to dive in this interesting investigation of yours.

Enjoy.

DocG said...

Maju:
"I do not mind considering that humans were in Asia before the Toba event, of course (I am not sure but could make some sense), but you still have that most of Eurasian (incl. Oceanian and American) genetics belong to macro-lineages that have their highest diversity by far, and therefore their likely origin in South Asia."

As you know, SE Asia, Island SE Asia. New Guinea and Island Melanesia constitute a tremendously complex, mixed-up region that won't be sorted out genetically for some time, I would imagine. Many important groups have never been studied, or studied inadequately. If we assume Toba had the effect that Oppenheimer suggests it had, it looks as though the survivors most likely expanded not only north into northern India and other parts of mainland Asia, but also eastward, into parts of SE Asia, and the other areas I mentioned above. At the same time, however, we can assume that there would have already been, at the time of the explosion, a population east and south of Toba that remained either unaffected or less seriously affected, either genetically or culturally. While many of these groups have never been studied, there is some evidence to this effect:

1. Phenotypically, many of these groups strongly resemble Africans and some so-called Negrito groups in SE Asia strongly resemble African pygmies. Such African phenotypes are not found west of Toba (aside from the Andaman islanders, who could have migrated westward after the explosion).

2. Certain groups, notably certain highland N. Guinea groups, are considered by most researchers to represent some of the oldest populations outside of Africa -- and as I understand it, this conclusion has been borne out by whatever genetic studies of them have yet been done. Unfortunately not much genetic research appears to have been done yet with these groups, so we can't be sure. (A good example of the problem is a study done by Ingman and Gyllensten in 2003 on the "Evolutionary History of Australians and New Guineans" based on DNA from 10 individuals from the New Guinea coast and 11 from the Highlands. Not nearly enough of a sample, obviously.)

3. The musical evidence points to various instances of an "African signature," which, in New Guinea, is found almost exclusively among Highland groups. Very similar instances of P/B related practices can be found in a wide area to the east and south of Toba, but none to the west.

My point is that there IS evidence, albeit circumstantial evidence, of a "Toba effect," but that it is too soon to say whether or not the genetic research will reinforce that picture or not.

DocG said...

Maju: "What I argue is that, probably after Toba, the main source of Eurasian genetics was still in South Asia and that at least South Asian mtDNA lineages do not show the bottleneck that Oppenheimer argues for."

Oppenheimer writes (p. 193):
"Now, if Toba really did blow its top after India was first colonized, we would expect a mass extinction event on the Indian peninsula which would have effected the eastern side more than the west. This is certainly one interpretation of the paradox of the Indian genetic picture, in which the genetic trail of the beachcombers can be detected, but the bulk of Indian subgroups...are unique to the subcontinent, especially among the tribes of the south-east. This is what we would expect for a recovery from a great disaster.” If indeed "the bulk of Indian subgroups are unique to the subcontinent" that would seem consistent with at least one major bottleneck, if not more -- no? Have you read Oppenheimer's book? If not, I suggest you get it and go immediately to Chapter 4, where all these possibilities are discussed in some detail.

Maju: "My impression is instead that other lineages had a hard time expanding in South Asia because mtDNA M was already too widespread. They found more room to expand in the East and later also in the West instead."

That makes perfect sense to me. Which provides an excellent explanation of why Asia to the south and east of Toba is such a mishmash. It would seem as though Toba survivors moving into that region from India could have encountered OOA migrants who were already established there. These would have been less agressive and consequently retreated to the hills, or else to more remote islands. This might not be reflected (yet) in the genetic evidence because the samples and the number of groups studied are too small and also because the more remote groups have been less accessible to researchers.

"think you're relying too much on one interpretation that may not necesarily be the correct one and hence linking the fortune of your exploration to that of Oppenheimer."

I've focused on Oppenheimer because his theories enable me to more easily account for certain patterns that otherwise would be very hard to explain. But the fortune of my research will not, I hope, be linked to any one theory of human migration and evolution -- because my main concern is doing the research itself, and seeing where it leads, rather than using it to prove any particular theory.

DocG said...

As far as the Caucasus is concerned, Oppenheimer refers to the Trans-Caucasus, not the northern Caucasus. And for me it is definitely Georgia, not the Caucasus in general, that is of the greatest interest. It is here as well that we find the only language family in Europe, other than Basque, or Lapp, that is not Indo-European. And I find it difficult to accept that the archaeology of Europe gives us a complete picture, I think that could be an illusion, especially when we are considering the Paleolithic.

And as far as panpipes are concerned, it is possible to distinguish between two very different traditions, the oldest being group oriented and employing the well known "hocket" technique, a practice which appears to have been lost when the panpipe was adopted as the Greek "syrinx" and was usually performed solo. The instrument appears to have spread rather widely from Greece and Anatolia in ancient times, but not necessarily the original group performance style, which seems to have survived only in refuge areas. So the panpie of the archaeologists must be distinguished from the panpipe of the comparative musicologists.

DocG said...

Maju: "Anyhow, just throwing food for thought. I really like your exploration, as you know, just that some hypothesis within it I don't find too convincing."

Thanks. Hashing through various aspects of this research with interested people like yourself is far more important to me than the question of whether any of my speculations are "right" or "wrong." I love to speculate and theorize but what really interests me is the research itself -- and its implications.

Maju said...

As you know, SE Asia, Island SE Asia. New Guinea and Island Melanesia constitute a tremendously complex, mixed-up region that won't be sorted out genetically for some time, I would imagine. Many important groups have never been studied, or studied inadequately.

Agreed but India's genetic diversity has also bee just scratched, so to say. Continuously we learn of new small haplogroups or new presence of already known ones. But the overall picture seems to have been drafted pretty much already (and I mean now in 2009, probably not so much yet in 2004, when Oppenheimer wrote).

Phenotypically, many of these groups strongly resemble Africans and some so-called Negrito groups in SE Asia strongly resemble African pygmies. Such African phenotypes are not found west of Toba (aside from the Andaman islanders, who could have migrated westward after the explosion).

I can detect some affinities but there are also many differences. One important element to consider is that similar phenotypes would also be similarly adaptative to similar climates and environments.

I do not doubt anyhow that Negritos have been in most cases isolated or almost so since the early colonization, while other peoples, in India for example, had much more complex interactions.

Certain groups, notably certain highland N. Guinea groups, are considered by most researchers to represent some of the oldest populations outside of Africa...

Papuans, like Australian Aboriginals, are also surely product of the early colonization and have remained isolated (or just semi in the case of Papuans, who are Neolithic since long ago) since then. But all Papuan mtDNA lineages are derived from M and R, which have highest diversity by far in South Asia.

Unfortunately not much genetic research appears to have been done yet with these groups, so we can't be sure...

The picture of Papuan genetics has improved since 2003. Have no doubt. Their Y-DNA seems C2 and K-derived (M and others), which could be of SE Asian origin but their mtDNA is of South Asian derivation.

The complexity of such criss-cross of lineages across the South and Eastern Asian and Australasian landscape only tells me that the urheimat of Eurasians, including Negritos, Papuans, etc., was a dynamical double region: South and SE Asia, through complex interaction. But you cannot dismiss South Asia as Oppenheimer apparently does: it just clashes with the huge diversity it has.

Maju said...

My point is that there IS evidence, albeit circumstantial evidence, of a "Toba effect," but that it is too soon to say whether or not the genetic research will reinforce that picture or not.

I don't think it does, at least not the mtDNA. This may be because the Eurasian expansion happened after Toba. I don't detect any signature of a Toba-caused bottleneck in mtDNA (like a very long undivided branch, of the kind we do see in Y-DNA) and that has lead me as of late to think of the OOA or at least the Eurasian expansion from Tropical Asia as being post-Toba.

My opinion anyhow.

Oppenheimer writes (p. 193):
"Now, if Toba really did blow its top after India was first colonized, we would expect a mass extinction event on the Indian peninsula which would have effected the eastern side more than the west. This is certainly one interpretation of the paradox of the Indian genetic picture, in which the genetic trail of the beachcombers can be detected, but the bulk of Indian subgroups...are unique to the subcontinent, especially among the tribes of the south-east. This is what we would expect for a recovery from a great disaster.” If indeed "the bulk of Indian subgroups are unique to the subcontinent" that would seem consistent with at least one major bottleneck, if not more -- no? Have you read Oppenheimer's book? If not, I suggest you get it and go immediately to Chapter 4, where all these possibilities are discussed in some detail
.

No, I have not. I may go to the library one of this days for it though.

I agree that the East coast looks kind of late populated and less diverse but this is consistent with archaeological data and migration modelling that suggest the Narmada-Son-Ganges as the main route east.

The only thing that SA lacks somewhat in mtDNA is macrohaplogroup N(xR) which I speculate may have coalesced in SE Asia (though even that is not fully clear either). In Y-DNA lacks D, though C may have also coalesced in SE Asia as well. But still the bulk of Eurasian lineages (both genders) seems to originate in South Asia.

I could accept Oppeheimer's hypothesis if I'd look only at Y-DNA but Y-DNA is much more easily subject to random drift it seems and has overall a much more obscure structure. MtDNA is easier to read (once you overcome the overwhelming list of haplogroups) and shows more consistent patterns. Occam's razor in hand I always choose mtDNA over Y-DNA if in doubt, as usually shows more clearly the old layers and generally correlates better with autosomal DNA and even phenotype (like Finns or North Africans, who have mostly "exotic" Y-DNA but rather West Eurasian mtDNA, autsomomal structure and phenotype).

Maju: "My impression is instead that other lineages had a hard time expanding in South Asia because mtDNA M was already too widespread. They found more room to expand in the East and later also in the West instead."

That makes perfect sense to me. Which provides an excellent explanation of why Asia to the south and east of Toba is such a mishmash. It would seem as though Toba survivors moving into that region from India could have encountered OOA migrants who were already established there
.

This could apply again for Y-DNA (D could be oldest and largely replaced by C and then "Indian" F derivates) but it is not what we see in mtDNA - unless you focus in oddballs like Australian Aboriginals, mostly mtDNA N(xR). It is not the case of Andamanese, Papuans or Melanesians nor, for what I know, of Negritos, all showing mtDNA M (or in the case of Papuans also R), apparently original from South Asia.

Also if Toba caused such devastation in South Asia, we should expect a recolonization from the East, at least in part. This could be tentatively attributed to Y-DNA K but hardly to any mtDNA lineage. There is also no signature in India of secondary demic explosion after the early M one. R did some of that but, as discussed before found only limited room for expansion and had to find venues outside South Asia where to really leave its mark.

Maju said...

It is here as well that we find the only language family in Europe, other than Basque, or Lapp, that is not Indo-European.

Some theories consider Georgian to be remotely related with Indoeuropean. There are other rare linguistic families in the area and they are in the North Caucasus. One is possibly associated with Hattic and the other with Hurro-Urartean. In total there are four linguistic families in Europe that are not Indo-Uralic (Indo-European plus Uralic, related at least by ancient sprachbund at their prehistorical coalescence are at the border with North Asia): three Caucasian families and Basque.

Hashing through various aspects of this research with interested people like yourself is far more important to me than the question of whether any of my speculations are "right" or "wrong." I love to speculate and theorize but what really interests me is the research itself -- and its implications.

That's what you're doing that is really important, IMO. I may have contentions with some conclusions but I do think this exploration is really interesting in any case.

I'd just wish sometimes we had a time machine, to make some time turism... or rather time anthropology and come back with some photos and, maybe more importantly, some sound recordings.

Just make sure you do not land near Toba caldera when it exploded. ;)

DocG said...

Maju: "But the overall picture [for India] seems to have been drafted pretty much already (and I mean now in 2009, probably not so much yet in 2004, when Oppenheimer wrote)."

I've seen several papers dealing with specific groups in specific parts of India, but I don't recall anything that provides a general overview, nor do I recall anything about bottlenecks or their absence. But you seem to be following the literature more carefully than I am. What do you recommend?

"But all Papuan mtDNA lineages are derived from M and R, which have highest diversity by far in South Asia."

It's great discussing this sort of thing with someone so knowledgeable and well read in the literature as you, Maju. I read a lot too but have to admit, a lot goes over my head. Right now I'm wondering about the origin of M. As I understand it there are no instances of L3 or any other L haplogroup outside of Africa. So M is usually discussed as a mutation (or set of mutations) that must have occurred somewhere in Asia. But where?

Can it be regarded as the mark of the original OOA band? Or would they have carried L3? And if they carried L3, then what happened to L3 when they were in Asia? If M originated with the OOA migrants, either in Africa or shortly after they entered Asia, then wouldn't we expect to find it, or its descendents, among all Asians, regardless of whether they experienced a bottleneck in India or not? If so, then the presence of M among New Guinea highlanders wouldn't matter one way or another with respect to Oppenheimer's theory, no?

As for the diversity of M in India, here again I'd appreciate it if you could provide a reference. There seems to be an enormous diversity in India from one group to another, but I'm wondering about the internal diversity of specific groups.

"The picture of Papuan genetics has improved since 2003."

Here again, I could use a reference. I appreciate what you are doing on your blog, by the way, and definitely need to spend some time there catching up. But I'd appreciate references to specific articles as well as phylogenetic trees. To my knowledge most of the research has been done by Friedlaender, who has concentrated mostly on Island Melanesia.

"Also if Toba caused such devastation in South Asia, we should expect a recolonization from the East, at least in part. This could be tentatively attributed to Y-DNA K but hardly to any mtDNA lineage."

Again a reference would be useful with regard to the above. You seem very sure of your interpretations of the genetic data and the phylogenetic trees and I'm wondering what the source of your certainty might be. Don't get me wrong, I'm very impressed with your mastery of all this material, and grateful for your interpretations, but I'm wondering what it's based on, and whether other researchers would concur.

It seems to me that, Toba or not, there must have been some sort of major bottleneck producing event or events in order to have produced the phenotypical and also cultural diversity we see in so many parts of Asia, SE Asia and Oceania. Otherwise Asia would look, sound and feel very much like Africa, no?

Maju said...

I've seen several papers dealing with specific groups in specific parts of India, but I don't recall anything that provides a general overview, nor do I recall anything about bottlenecks or their absence. But you seem to be following the literature more carefully than I am. What do you recommend?

Absoutely true. Most papers are focused on this or that lineage or population. Few comprehensive reviews exist, that's why I've worked my own largely (and I am so glad that the phylotree site has finally set an updated online reference for mtDNA phylogeny, much like ISOGG has been for years for Y-DNA one).

Sadly my old computer collapsed some weeks ago and I'm still recopilating my bookmarks as I go. I am making an effort to find some of the papers I recall important to define the coastal migration process as most likely scenario but it'll take me some time.

By the moment I found Metspalu 2004, but it's anyhow an old paper. 2004 is beginning to be really old in genetics, as the field moves so fast...

I'll try to locate more and better stuff for you.

Right now I'm wondering about the origin of M. As I understand it there are no instances of L3 or any other L haplogroup outside of Africa...

Basically correct, though L6 has been found to be more common (and quite diverse) in Yemen, while it's very rare in Africa (and concentrated among Ethiopian Semitic speakers there). There's also quite an ammount of L(xM,N) in West Asia and even some in Europe but it's believed to be of modern origin or mediated by North Africa also in more recent times.

M and N are for our purpose the only mtDNA "grand" lineages of Europe: all the rest are derived. And M and N are both sublineages of L3, which has also several branches in Africa (5 in fact, plus some L3*) and is generally thought to be of East African (Horn, Sudan) origin.

So basically you connect the dots and you have a line going from Addis Ababa to Karachi or Mumbay, so to say. N is in fact more problematic because its sublineages are scattered though all Eurasia-plus but is rare in South Asia, excepting R and other lineages related to Western mtDNA. That's why I think that N had a SE Asia core, like Y-DNA D and probably also C.

So M is usually discussed as a mutation (or set of mutations) that must have occurred somewhere in Asia. But where? -

In the way to India maybe or even in pre-Toba India if you think that AMHs were already there when Toba happened. But what is clear is that M has the largest star-like structure of all haplogroups ever: more than 30 lineages hang directly from the N node (in comparison L3 has 7, N and R are in the dozen area, only Western H can compare somewhat, and still seems smaller). This is a clear signal of large sudden expansion. As I say above, this may mean post-Toba expansion or maybe expansion after the journey of the desert, so to say, between Africa and South Asia. Whatever the case it does appear as the signature of a people that had no barriers to rapid demographic growth.

And the main place where all these "daughters of M" went to live (diversity) was South Asia, what means that the explosion happened there with all likelihood.

This is incompatible with Oppenheimer's interpretation.

(will continue)

Maju said...

(continues from above)

Actually I've been thinking a bit about that because, I just came to find out that the Aurignacian expansion in Europe is coincident with another huge volcanic explosion: the Campanian Ignimbrite eruption. Not as large as Toba but certainly enough to send pyroclastic bombs as far as Ukraine.

We can't be sure of the details but probably the eruption and derived climate alteration caused serious troubles to all but certain group (Aurignacians) was better placed to take advantage of the post-catastrophe scenario.

A similar case is that of the recent news on the meteorite that apparently stroke North America c. 13,000 years ago and whose possible effects you discuss in your paper. Again in this case, certain groups (Clovis people for instance) benefited from the desolation.

Whatever the case it is clear that mtDNA M people benefited from an empty India (and beyond) that allowed them to expand very rapidly. That is what M's huge starlike structure says: that a lot of different sublineages had the chance to succeed. But not only in South Asia: also in Melanesia, SE and East Asia (even some M lineages arrived to Australia - though AAs are mostly N).

Of course, you could imagine, I guess, that M coalesced in SE Asia and only expanded in South Asia because it was empty but this is not the most parsimonious model because the highest diversity of M (by far) is in South Asia. So in general geneticists lean to SA as the core of Eurasian expansion, along with SE Asia.

Can it be regarded as the mark of the original OOA band? Or would they have carried L3? And if they carried L3, then what happened to L3 when they were in Asia? -

It became M and N. There's no such thing as an immutable lineage, you know: everything evolves. M and N are the Eurasian L3 and are sometimes written as L3(M) and L3(N) or something like that.

So yes, M and N are the mtDNA signature of the original OOA band or bands.

If M originated with the OOA migrants, either in Africa or shortly after they entered Asia, then wouldn't we expect to find it, or its descendents, among all Asians, regardless of whether they experienced a bottleneck in India or not? If so, then the presence of M among New Guinea highlanders wouldn't matter one way or another with respect to Oppenheimer's theory, no? -

Well, if M (and R) "exploded" in South Asia and Papuans carry some of the derived lineages, as they do, this means, almost necesarily, that proto-Papuans arrived there from South Asia and not the other way around.

As for the diversity of M in India, here again I'd appreciate it if you could provide a reference. There seems to be an enormous diversity in India from one group to another, but I'm wondering about the internal diversity of specific groups.

I can't provide a single source. Researching where each lineage was found at took me some time, checking different sources listed at the Phylotree site (some I could not because they are behind paywall) and others and occasionally being informed directly from my readers.

I just can tell you that, according to my own research, based on very diverse primary and secondary sources, South Asia has highest mtDNA M diversity, though followed closely by Eastern Asia. I concluded this must mean a SA>SEA fast migration, in line with mainstream ideas, but maybe I am wrong - who knows.

A good list of papers on South Asian DNA can be found here anyhow.

Maju said...

Here again, I could use a reference. I appreciate what you are doing on your blog, by the way, and definitely need to spend some time there catching up. But I'd appreciate references to specific articles as well as phylogenetic trees. To my knowledge most of the research has been done by Friedlaender, who has concentrated mostly on Island Melanesia.

Aso Hudjashov 07, which also has some nice illustrative graphs on mtDNA distribution through regions. Not sure if it'll be enough.

Again a reference would be useful with regard to the above. You seem very sure of your interpretations of the genetic data and the phylogenetic trees and I'm wondering what the source of your certainty might be. Don't get me wrong, I'm very impressed with your mastery of all this material, and grateful for your interpretations, but I'm wondering what it's based on, and whether other researchers would concur.

I'll see what I can find. I'm already quite tired at this stage of writing back and searching for papers at the same time. Check anyhow if some of what I have mentioned so far may help. I'll try to look for more and hopefully better stuff tomorrow or whenever I can.

(continues)

Maju said...

It seems to me that, Toba or not, there must have been some sort of major bottleneck producing event or events in order to have produced the phenotypical and also cultural diversity we see in so many parts of Asia, SE Asia and Oceania. Otherwise Asia would look, sound and feel very much like Africa, no? -

Whoa! Phenotype formation, "racialization" is a hot issue. Anyhow my opinion follows: Africa is anything but a single bloc. But there seems to have been a greater deal of homogenization there than in Eurasia. In Eurasia we can be reasonably safe that the modern phenotypes are product of regional homogenization after the "great Eurasian expansion". As I see it, after the expansion, there was much less gene flow between regions.

Also we do know that West and East Eurasians have depigmented following distinct genetic pathways, at least in part, what means that there was not meaningful genetic exchange between these two populations as they moved north (otherwise introgression would have worked and the adaptative genes of ones would have been rapidly incorporated into the others by positive selection).

This basically supports the idea that Eurasians expanded following a U pattern from Tropical Asia.

Broadly we can divide Eurasians into Cauasoids (including largely South Asians, who are at the origin of the colonization of West Eurasia, at least in part), Mongoloids and an array of conservative or archaic-looking groups in Australasia and some lost corners of tropical Asia (and the Ainu too would fall in this array of minor groups probably). This is totally consistent with the rapid spread model followed by a longer period of regional homogenization (after corresponding founder effects or whatever).

Not everybody agrees with this. Some argue that modern phenotypes only formed in late Paleolithic and Neolithic periods, implying more or less mysterious expansions that wiped everyone else off the map more or less. They support that view on the relative aboundance of archaic features in known archaic human specimens. But I find this view largely inconsistent with the large continuum areas we find now that seem much more consistent with my model.

Up to you. What is clear is that nothing of all this is still written on stone.

Anyhow, after all this discussion I'm starting to see Oppenheimer's theory on a different light: what if the high South Asian diversity is not the product of a local origin but that of a SE Asian one, favored by the emptiness of the territory?

One problem is that I would expect not a starlike structure but a more complex one, with specific founder lineages. And that's not what we see in South Asia but an vast array of locally relevant or just widespread lineages, spawning from exactly the same origin as Australasian Q or Sibero-American D.

Well, can't say much more. Will try to search for more and better papers, if available.

Maju said...

Another important paper on OOA and coastal migration is Pope et al., 2007. But, while I used to have free access to it, it's now behind a paywall apparently.

DocG said...

Hi Maju: I just downloaded Metspalu 2004, which I hadn't seen before. I'll read it over this evening. Thanks.

Meanwhile, what do you think of these quotations from Cordeaux et al, "Mitochondrial DNA analysis reveals diverse histories
of tribal populations from India":
"Southern Indian tribes showed reduced diversity and large genetic distances, both among themselves and when compared with other groups, and no signal of prehistoric demographic expansions. These results probably reflect enhanced genetic drift because of small population sizes and/or bottlenecks in these groups. By contrast, northern groups exhibited more diversity and signals of prehistoric demographic expansions. . . Phylogenetic
analyses revealed that southern and northern groups (except northeastern ones) have related mtDNA
sequences albeit at different frequencies, further supporting the larger impact of drift on the genetic
structure of southern groups."

Could this be consistent with multiple bottlenecks in the south, due either to Toba, or perhaps a Tsunami, with a south-north gradient of population impact?

"In addition, Forster et al have proposed an mtDNA
control region motif (16223C and 16357C) which could
represent a signature of an early migration from Africa to Sahul through the southern route. This motif was not found in any Indian tribal mtDNA; 16357C had a
frequency of only 2.1% and was always associated with 16223T, while 16223C had a frequency of 27.9%. Furthermore,
Indians do not show particular affinities to Africans."

Finally, "we could not find in Indian tribals any unquestionable genetic signature of the B60 000 year old
migration from Africa to Sahul following the postulated
southern route. A possible explanation would be that such migration never occurred along that route. Alternatively, the early migrants from Africa may have made their way to Sahul following the southern route without settling
in India." Would this latter picture also be compatible with the effects of Toba in India, which could have erased the African markers?

I realize this picture is very different from the picture you see, and I'm wondering why.

Maju: "Whoa! Phenotype formation, "racialization" is a hot issue."

I'm puzzled as to why you associate phenotype with race. They are two completely different things, although it's true that those arguing for a recognition of race as science often get them confused. Race is certainly a social construct. But phenotype differences are certainly a legitimate subject for scientific study, why not? The notion of race implies the notion that people can be classified according to so-called "racial" differences, but no one has ever been able to pinpoint those differences and no one has ever been able to produce a truly scientific classification based on that.

The notion of phenotype, on the other hand, is simply technical, with no implications as to whether or not classifications would be possible or what they might mean. "Race" has huge social baggage, but phenotype doesn't -- or shouldn't.

Meanwhile thanks for the references. I hope you can locate some more.

Maju said...

I'm puzzled as to why you associate phenotype with race.

They mean more or less the same (at least in anthropological contexts) but I was not making an issue about debating races/phenotypes but what I find difficult to deal with is how the different types formed. I tend to favor the model of regional homogeneization but for others (normally for racialists) it's a matter of groups carrying the "essence" of a phenotype and expanding around (at the expense of other groups).

All or most remains prior to Late UP appear somewhat different from modern populations, so this kind of favors the essentialist position but the archaeological record is inconclusive anyhow, and there are many interpretations as well.

Slippery terrain, not so much for its socio-political implications but because it's hard to get an easy really scientific picture.

Maju said...

Re. Cordeaux' paper.

It seems interesting for the sampling of so many different tribal groups but:

1. Tribals are a tiny minority of the population of South Asia and it is most likely that most of caste populations ancestry (specially mtDNA) is also native (tribals became something else as they integrated in Neolithic and later Hindu society).

2. Their sample of other Indians is minimalistic and necesarily not representative (there are many more mtDNA lineages in India than all the non-tribals sampled in that study).

Could this be consistent with multiple bottlenecks in the south, due either to Toba, or perhaps a Tsunami, with a south-north gradient of population impact?-

I cannot discard a bottleneck because of Toba but we can discard it because a tsunami. All people would have to live at the coasts affected and the data from the 2006 tsunami shows that tribals like the Andamanese react very well to such catastrophes (they survived by taking high ground before the wave hit).

I think that what Cordeaux means is that the small tribal populations have always remained tiny and hence suffered strong drift. Drift, as you may know, is the process by which an allele or, in this case a haploid lineage, tends to fixation by mere means of randomness. This is strongly favored by small population sizes and disfavored by large populations, where the multiple random accidents tend to cancel each other.

If a small isolated tribe begins with four lineages it is very likely that after some centuries they end up with one or two only and that with more time they end up with just one (fixation). This happens way too easily with haploid lineages because a man can have all daughters or vice versa, with enough time these accidents accumulate in one direction and fixation happens.

More interestingly, if the tribe expands after fixation and splits into two or more groups, these new groups will eventually tend to fixation in one of the derived lineages but each one in a different one, so (up to a point and with due caution) you can imagine the branching out of lineages as that of "tribes" or clans that expand or not.

So when you see a lineage that splits in two branches it is a sign of a modest expansion by the group that carried it, and when you see a lineage like M that splits in more than 30 (known, surviving) lineages, it is sign of a huge expansion. When a lineage is only found in one variant, no matter how old it is since divergence, then it means it has not expanded, at least successfully.

Not sure if all this is known to you. But well, Cordeaux only means that tribes have remained in great reproductive isolation as far as he can tell.

Maju said...

Re. the mention to Foster:

I have not read Foster's paper (do you have a link?) but it seems that the two loci belong to the hyper-variable region (aka control region) and mutations at either locus are found in many different lineages. While the control region is used in research (seems cheaper to use) it is not a reliable marker alone.

For example, remember when they discovered that the Cromagnon of Paglicci had "the CRS haplotype"? That was detected using the control region precisely but leaves us with too many doubts because modernly such haplotype is found in at least two haplogroups: U and H, even if in the area studied is more common in H1 (and in this case we can also support this conclusion because the analysis of a series of Cromagnons at Taforalt with similar methods produced a most likely haplogroup "pie" almost identical to the modern one at a neighbour village).

Sounds far-fetched, unless he can gather more evidence. In fact, if this hypothesis has not been accepted since 2003, probably means that he was speculating too much.

Finally, "we could not find in Indian tribals any unquestionable genetic signature of the B60 000 year old migration from Africa to Sahul following the postulated southern route. A possible explanation would be that such migration never occurred along that route. Alternatively, the early migrants from Africa may have made their way to Sahul following the southern route without settling
in India." Would this latter picture also be compatible with the effects of Toba in India, which could have erased the African markers?
-

This part (this is again from Cordeaux, right?) could make some sense but what lineages are they studying? The model says that macrohaplogroup M expanded and people settled in different areas, where the different local lineages got more or less fixated. I would not expect to find too many direct connections between the various regions and the ones that do exist may mean "late" (secondary) migrations and not the original M expansion (logically).

I realize this picture is very different from the picture you see, and I'm wondering why.

I think I have explained why above.

I also think that different hypothesis, models, "clash" in the scientific debate and, normally, only the best survive. The coastal migration model so far has fared quite well, while Cordaux'03 right now will most likely only be mentioned for the raw data on tribal Indian mtDNA.

As of now, neither Indians nor other Eurasians/Sahullians are "closer" to Africans: all are derived from the same early Eurasian population(s) and this population(s) is believed to have lived in tropical Asia, including India, where most of the top level diversity is now.

Could we make it fit with a recolonization from SE Asia? I'm thinking we could but we would have to persuade too many people that the highest diversity of South Asia in mtDNA M and Y-DNA F is product of the rapid migration from SE Asia soon after the Toba disaster, more or less simultaneous with the same expansion in SE Asia and neighbouring areas (mid-East Asia, Sahul).

It could fit, I guess, but it's kind of forced.

I am anyhow starting to think on it as a possibility and trying to figure out how it could have worked (and if it could at all) even if only as an interesting exercise.

DocG said...

Hi Maju. Just a few comments here and there, as I don't want to turn our dialogue into a ten volume set. :-)

As I see it, a Tsunami as a possible multi-bottleneck event remains a possibility, yes. Because at the earliest OOA stage, we do seem to be dealing with an exclusively coastal culture. Yes, the Andaman Islanders saw it coming, but only because of certain traditions they had, which must have originated with earlier Tsunami experiences. But the OOA groups might well have been Tsunami virgins, with no such traditions to protect them.

Maju: "Drift, as you may know, is the process by which an allele or, in this case a haploid lineage, tends to fixation by mere means of randomness."

Yes, I know what genetic drift is. I also know that this is an ambiguous term because it can actually stand for two very different things: 1. random change over time in a small population, as you say; but also 2. the development of founder effects due to a population bottleneck. One is gradual the other sudden. Very different, but both characterized by the same term. NOT the clearest way to explain anything I'm afraid. Not your fault, of course, but someone's.

With respect to Cordaux et al., they clearly embrace both possibilities. Here is the link, by the way: http://www.nature.com/ejhg/journal/v11/n3/pdf/5200949a.pdf

And here is what they say: "These results probably reflect enhanced genetic drift because of small population sizes and/or bottlenecks in these groups." So for them bottlenecks are also a possibility, not just random drift over time. For me both make sense, i.e., a sudden bottleneck followed by isolation, followed by random drift over time, which would explain the lack of diversity even over a period of 72,000 years (since Toba). It seems to me that one or the other couldn't account for the lack of diversity, it would almost have to be both. And let me remind you of their conclusion:
"the early migrants from Africa may have made their way to Sahul following the southern route without settling in India." This for me is the interpretation most consistent with the musical evidence. And not only the musical evidence, because there's also a lot in the cultural evidence generally that supports this. As I see it what makes the most sense is that the OOA group quickly moved across the Indian Ocean coast all the way to Sahul and also Sundaland, leaving colonies along the Indian coast. At some point there must have been some disaster, either Toba or a Tsunami or perhaps a drought, or major cyclone, etc., that would have almost wiped out the colonies in India, producing multiple bottlenecks and founder effects, genetically, phenotypically and culturally. Whatever the disaster, the groups that had already made it to the Sahul and Sundaland would have been spared the worst effects and thus not suffered bottlenecks, nor serious genetic, phlogenetic or cultural changes (though very possibly some). After the disaster this area would have been populated by surviving groups from India, so that today it is a mix of both types, with the original settlers now living in refuge areas.

This makes sense to me for many reasons. And, reading not only Oppenheimer, but also Cordaux et al., plus some others it does seem at least consistent with the genetic evidence. You are correct in pointing out certain difficulties with such a theory and I'm grateful for that. The picture in my mind is far from being proven. But it's the only picture that makes sense for me to date, because it explains so much.

As far as the survival of only the "best" theories, I think it's far too early to tell which has survived and which has not.

DocG said...

As for the reference to Forster:
Forster P, Torroni A, Renfrew C, Ro¨hl A: Phylogenetic star
contraction applied to Asian and Papuan mtDNA evolution.
Mol Biol Evol 2001; 18: 1864–1881.

http://mbe.oxfordjournals.org/cgi/reprint/18/10/1864

It's a free download. Enjoy!

Maju said...

Thanks for the links, the first one I had already read but Forster's paper is truly "arcaheological" so to say: brings me back to a time when the phylogenetics were still being deciphered (for exapmple there's no mention of macro-haplogroup R yet). Fascinating - but not very helpful, really (too old).

AFAIK a bottleneck is a quasi-extinction event in which 10% or less survive. IMO there is no clear indication of this ever happening in human history.

Founder effect has similar effects to a bottleneck but it's caused by a migration, so it's normally easier to dientify. Drift would have similar effects too but requires the population to remain low for long periods.

I'd say that what we see in mtDNA can perfectly be explained by founder effect at the OOA period, in the arid lands between the Horn and Pakistan.

I'd discard the tsunami because archaeology and also geographic modelling show that the pure coastal route would not be real except maybe for some groups. "Coastal" here means along Tropical Asia, as opposed by continental through the Central Asian steppes. Though yes, some have argued for a strictly coastal migration and this may apply to the lineages that reached fastest to the East, mtDNA N and Y-DNA D (and maybe C too) but not to the "slow" ones that appear to expand from India (mtDNA M and Y-DNA F), which make the bulk of non-African haploid genetics.