April 18, 2012

Laos inhabited by H. sapiens since maybe 56,000 years ago?

Just a quick note because I really could not find too much.

I just read at a very succint note at Science Daily that:

Researchers have discovered the oldest known human remains in Southeast Asia, a partial human skull dating to at least 40,000 years ago. Excavations at Tam Pa Ling cave in northern Laos produced a dozen pieces from a Stone Age person’s skull, including a skullcap and a lower jaw, anthropologist Laura Shackelford of the University of Illinois at Urbana-Champaign reported April 14. Small front teeth, a rounded brain case and other traits identify the reassembled fossil as a modern Homo sapiens, Shackelford said. The find supports proposals that at least some human migrations out of Africa around 100,000 years ago followed a southern route that led to Southeast Asia.

Nothing more of relevance. I could find a paper (ppv) by Dr. Shackleford on Laos but it discusses another cave and much more recent dates (c. 15 Ka ago).

On the other hand I found a paper on another Laotian cave by different authors which mentions undefined human presence c.56 Ka ago:

Valéry Zeitoun et al, Multi-millennial occupation in northwestern Laos: Preliminary results of excavations at the Ngeubhinh Mouxeu rock-shelter. Comptes Rendus Palevol 2011. Pay per view.

Abstract

With over half a century of political instability, resulting from armed conflicts, decolonisation and the Cold War, archaeological investigations in Laos have been rare, leaving little more than a blank page in the chapter of Southeast Asia's prehistory. Recent research has shown that Laos holds a rich prehistoric heritage. In conjunction with the research initiated by J. White who conducted the first professional archaeological survey of northern Laos since decades, we have extended the investigations to the Luang Namtha province. This work allowed us to gather important data about Hoabinhian stone tool assemblages and former cultures. In particular, the archaeological remains and dating from the Ngeubhinh Mouxeu rock-shelter indicate that this mountainous region of Laos has been inhabited over a long period of time that possibly spans as far back as 56,000 ± 3000 BP.


While, by the moment, I ignore many details, I believe it is very interesting to mention as the early colonization of SE Asia by our species is still not well understood (and yet may be critical to understand all, or at the very least much, of Eurasian, plus aboriginal Australasian and American origins).

With due caution, I think we can infer from these and other less direct (Indian and Arabian archaeology) or more controversial materials (Liujiang skull, Luzon foot bone, speculations about very early colonization of Australia) that the colonization of SE Asia by our kind can be traced to at least that date of 56 Ka ago, probably even earlier.


Update (Aug 21): see this newer entry.

61 comments:

  1. "The find supports proposals that at least some human migrations out of Africa around 100,000 years ago followed a southern route that led to Southeast Asia".

    I have (finally) put up the list of mt-DNA M at Wikimedia. You will see from that that something like 20 mt-DNA haplogroups seem likely to have coalesced somewhere in the hills along the Norhteast India/Burma/Yunnan border region. Laos lies at the eastern end of these hills (well, actually the hills stretch into Vietnam). So obviously members of mt-DNA M have been in Laos, or to the immediate west of it, for as long as M has been around. But at the same time the presence in land-locked Laos argues eloquently against a 'coastal' element in the southern migration.

    ReplyDelete
    Replies
    1. "So obviously members of mt-DNA M have been in Laos, or to the immediate west of it, for as long as M has been around."

      Not quite obvious to me. Highlands tend to be refugia. They are where peoples migrate to when some new people arrives on the scene and crowds them out (or wipes out/assimilates beyond recognition those who don't migrate to the highlands). Older population strata are going to tend to end up in highland areas after enough waves of population migration have moved through neighboring areas regardless of where the old population strata were originally located.

      Horticulturalists whose demographic impact is more intense than non-horticulturalists, because horticulture supports so much higher population densities than the alternatives (with the possible exception of coastal fishing based food production), have a particularly strong preferences for valleys and plains over highlands due to the realities of farming technologies.

      I would not find it to be at all obvious that mtDNA M of the haplotypes found in Laos today are representative any further back than the onset of the Neolithic revolution in Southeast Asia, thousand of years after the Neolithic revolution arrived in either East Asia or South Asia.

      I would not be very comfortable drawing inferences about the population genetic makeup of Laos in the Upper Paleolithic from modern population genetics in the larger region.

      Delete
  2. I was wondering where you got that quote. I had not realized that the article says so clearly that the find supports the southern route.

    I'll check tomorrow what you have added to the wiki, thanks. It may be of use in the future.

    ReplyDelete
  3. Minor point, "land locked" Laos hosts the mighty Mekong river, and many tributaries. Laos lies almost entirely within the Lower Mekong Basin Vientiane is a mere 174 metres above sea level. It is not hard at all to imagine coastal people rather rapidly entering Laos by migrating upriver. True the Khone Phapheng Falls on the border with Cambodia are not navigable for shipping, but the walk around the white water would not have daunted primitive people at all.

    ReplyDelete
    Replies
    1. Terry's obsessions - hehe! Technically Laos is landlocked (it has no coast) but there was no "Laos" in the Upper Paleolithic so the concept is meaningless. Also, as you say well, the Mekong is the major river of the area and a navigable one (except for a few falls).

      In principle the findings, together with others in nearby regions and mtDNA structure, seem to support again the southern route and a very important role for SE Asia (including South China and Sundaland) in the early prehistory of Greater Eurasian peoples.

      It's much easier to make findings in highland areas where caves abound, silting is low and sea level changes do not matter than in the coasts. Also findings depend a lot on the focus of researchers and availability of access for research (for example right now Burma is a no-no, even if it should be a key location).

      Delete
  4. "Minor point, 'land locked' Laos hosts the mighty Mekong river, and many tributaries".

    But Most of Laos is mountainous and it looks very unlikely that humans moved into Laos from the south, via the Mekong. Vientiane is close to the border with Thailand and most agree that Thailand was settled from the north. And Most M haplogroups in Laos have close connections to the north. Besides which M haplogroups could hardly be said to string out along the SE Asian coastline. The majority of M haplogroups in the region are concentrated in the hill country of South China and NE India.

    "Technically Laos is landlocked (it has no coast) but there was no 'Laos' in the Upper Paleolithic so the concept is meaningless".

    Don't be stupid. Did you not read what I wrote? I stressed that Laos is geographically part of the Northeast India/Burma/Yunnan border region. It is therefore totally irrelevant that 'there was no Laos in the Upper Paleolithic'.

    "In principle the findings, together with others in nearby regions and mtDNA structure, seem to support again the southern route and a very important role for SE Asia (including South China and Sundaland) in the early prehistory of Greater Eurasian peoples".

    Rubbish. South China (especially Yunnan, which is what 'South China' usually means in these circumstances) is completely land-locked, as is Laos. As for Sundaland you will see there are very few M haplogroups that are specifically 'Sundaland' in origin. The paper shows merely that humans were in Laos a long time ago. It says nothing about how they got there.

    "right now Burma is a no-no, even if it should be a key location"

    Again the evidence, such as we have, indicates entry into Burma from the north or from Thailand. Not via the coast. But I agree that the evidence from Burma will prove to be most interesting.

    ReplyDelete
    Replies
    1. You mean Zomia. As I mentioned in that entry: "Essentially the Zomia highlands enclose the following rivers: Brahmaputra (by the West), Irrawady, Chao Praya, Mekong, Red River and Pearl River (by the NE). That Zomia can't be considered alone but in relation with those rivers and the coasts of that area as well.

      "South China (especially Yunnan, which is what 'South China' usually means"...

      Not really. I'm not even sure we should consider Yunnan as part of China proper, i.e. historically Han China - it's more like part of "Greater Tibet", so to say, or a region on its own merit.

      Anyhow normally South China means all china (except the West, which is not part of "China proper") south of the Huai River. This is the historical division (alternatively that area minus Jiangsu and Anhui (i.e. all China South of Shanghai).

      At the very least South China means most of the Yangtze and all Pearl River basins, roughly the provinces of Yunnan, Sichuan, Chonquing, Ghizhou, Guangxi, Guandong, Hunan, Jiagxi, Fujian and Hubei.

      So South China is not "landlocked" even if Yunnan is.

      "As for Sundaland you will see there are very few M haplogroups that are specifically 'Sundaland' in origin".

      Maybe but there is a high diversity of R, quite curiously, second only to South Asia.

      Delete
    2. Anyhow, the concept "landlocked" makes no sense without precise boundaries such as those of modern states. I could agree that you said that Altai is "landlocked" because it's very far from any sea in all directions but Laos?! The sea is just a few hundred kilometers away either downstream by the Mekong or, shorter but harder, across the mountains at what is now Vietnam.

      Your nagging is mere grasping for straws.

      Delete
    3. Actually there is less than 100 km between the Laoitian border and the sea via Vietnam at some points. 100 km across mountain and jungle can be journeyed by most Papuans (just an example of modern hardy mountain people) in a single day if need be. It is not really any meaningful distance from the sea.

      Delete
  5. The 56,000 years before present date confirms expectations driving by the timing of the colonization of Australia and Papua New Guinea that must have come later than a modern human presence in Southeast Asia.

    I'm not sure in this context that a "Southern route" and a "coastal route" are synonymous. I'd generally think of a "Southern route" as merely distinct from routes along the Silk Road, Central Asia, and Siberia, or perhaps a circumpolar route. Pretty much any route to East Asia via India rather than Russia or the Ukraine would qualify as a Southern route.

    In practice, I think route that is primarily coastal and makes its way inland via tracing rivers to their higher altitude sources and only tertiarily away from coasts or rivers, is a model that is a pretty good fit to almost all pre-Austronesian human migrations. But, the terminology doesn't naturally suggest that.

    Of course, the migration to inland and highland areas need not be particularly rapid given the evidence that in the New World, the trip from Beringia to Peru took only a few thousand years. With a 56,000 years before present date inland, one can take three thousand years to get there from the coast and still need modern humans to arrive on the coast by just 59,000 years before present. And, 3,000 years to travel a few hundred kilometers inland, perhaps 1-5 km inland per generation for a hunter-gather population, is almost imperceptible in the course of human affairs.

    ReplyDelete
    Replies
    1. I agree that a southern route does not need to be strictly coastal. I have since long ago sustained that, according to archaeological evidence, GIS modeling (Fields 2007) and common sense that crossing the South Asian subcontinent was probably done by three main routes:

      1. The Narmada-Son-Ganges corridor (between roughly Mumbay and Kolkatta). Here the coastal aspect of the route ends near Mumbay and probably began again at the Ganges Delta, however boating was surely useful for riverine exploitation and travels.

      2. The Krishna River route across the Ghates (roughly crossing through the middle of the Dravidian countries of South India). Here the coastal route continues (and is attested archaeologically) until Kerala, then continues through the interior. There's no archaeological evidence of it continuing beyond Andrah Pradesh towards Bengal but it's possible that the evidence is underwater.

      3. The purely coastal route around the whole Indian subcontinent. This one is modeled as "least resistance" route by the GIS paper but it's not really documented archaeologically. Again it's perfectly possible that the evidence is underwater or is yet to be found.

      In any case the shortcut through the Narmada route is quite obvious and apparent and precludes the necessity of a 100% coastal route towards SE Asia and beyond - but does not exclude it completely.

      The main argument for the "rapid coastal" route (emphasis in "rapid" rather than just "coastal") is genetic: there are several very old lineages that appear to be centered in SE Asia (notably Y-DNA C and D and mtDNA N) rather than farther West. However we could here also retake the old theory of the Toba bottleneck (affecting mostly South Asia and not so much SE Asia) to explain some of this (unsure) - it'd be IMO somewhat different of what Oppenheimer argued in the past but can still fit.

      Delete
  6. The Ngeubhinh Mouxeu rock-shelter at N 20° 42.98′, E 101° 00.96′ is just 80 km NE of the main channel of the Mekong. Must be a nice area because there are towns on both sides of the river now, Huay Xai on the Laotian side and Chiang Khong on the Thai side. So the shelter is up in the mountains, but not far from the wide watery superhighway to the sea, basically just a week of 10 km walks.

    ReplyDelete
  7. River valleys would have been extremely inviting to follow for migrating humans. I don't really see an either-or in close-to-coastline vs. inland, but that both happened, just in different places and at different times. Up to Pakistan, anything but a coastal route is IMO pretty much out of the question. From there on, there are several plausible inland routes. Also, the more tropical things got, humans probably stayed close to the coast or perhaps even stopped completely for a long period (dangerous, different environment/diseases, likely inhabited by erectus).

    The other point is that recent tree models of autosomal DNA all seem to show that NE Asians and Amerindians branched out first, e.g., from Indians. So there is some indication that one branch indeed moved inland, and I also would think Myanmar/Yunnan seems the most likely route. This could have been potentially very early on, in the pre-Toba time frame. SE Asians than would derive from the same source population (probably in the-then non-tropical portions of NE India), but perhaps were simply more successful in occupying SE Asia and Sundaland and later the Sahul after Toba. At least one of the Denisovan admixture papers makes something like this likely because they found a to-step admixture process explaining the results best.

    ReplyDelete
    Replies
    1. I am in general agreement but I don't think we should consider that people "even stopped completely". Of course there were surely people who remained at the point of arrival (say, Sindh) and whose descendants may be still today right there but the great migration was not a journey of leisure, not even one of survival... it's one of successive overpopulation.

      If conditions are good, people will grow fast, very fast (World population has multiplied times 7 in just a century and that's probably not a maximum). I think that in optimal conditions, assuming just x2 each generation is not weird at all. Hence some people have to move a bit further each generation or fight for resources with their cousins. That's how migration happened.

      You know the tale of the emperor, the chessboard,and the rice grains, right? A chessboard only has 64 spaces but that was more than enough to make the promise of the tale exponentially absurd. 64 generations is just a bit more than 1000 years, almost nothing in the time scales we are considering here. That explains that basal descendants of mtDNA M are spread in all the Asia east of Iran and even in Melanesia: the Great Eurasian expansion (or most of it) was a matter of one or two millennia most likely.

      Of course the exact subjective reasons could be of the kind of "John is such a jerk!" (hunter-gatherers often "vote with their feet") rather than a true famine but emotional and material worlds are not wholly detached from each other in fact. In any case, the people moved away when they felt constrained one way or another in their original group and lands and that happened inevitably with merely biological population growth, which was in turn fueled by the optimal conditions of a newly settled "virgin land" (more so in the warm Tropics).

      I doubt H. erectus would be a meaningful problem in most cases: they were too comparatively simple, dumb, and surely very sparse. They could win a skirmish, sure, just like a tiger can, but they were bound to lose every single protracted "war".

      If the expansion stopped at times it was only because they run short of places to go: surrounded by your own kin (as smart as yourself and also probably your relatives in some sense) and/or unsurmontable geographic or climatic barriers (or maybe the Neanderthal brawn&brains dangerous mix), they had to consider growth control one way or another eventually.

      Delete
    2. Regarding the second point: link?

      Amerindians are not 100% of East Asian stock but most of the paternal lineages (Q) and some of the maternal ones (X2) are from West Eurasia (both these lineages are most diverse in West Asia and therefore must have originated there and spread to Beringia via Altai and other less precise Siberian locations.

      I can only imagine that the branching produced in such trees is actually an artifact produced by admixture (mixed populations do behave that way, sometimes confusing the casual and not-so-casual observers).

      The East Asian component (which is surely dominant) of Amerindians is from the rather coastal zones of East Asia, mostly from NE Asia (but B is very rare so far North). I would associate the Yunnan route with Y-DNA D but there was probably never a strict distinction between these populations because a basal branch of D managed to make a strong founder effect in Japan anyhow, even if most of the diversity is concentrated in the Tibet-SE Asia arch. Even if they split at some moment, they later remixed, making the result quite blurry.

      Delete
    3. Maju,

      I strongly think by far the most likely time people moved ooA was during the warm&wet phase of the green Sahara - ~130,000-110,000ya. The adjacent Arabia and southern pathway to Pakistan would have similarly turned into a partially lush savanna - the ideal living conditions for early modern humans, and the pathway providing an ideal outlet for the ensuing population growth. After that, the climate deteriorated for almost 10,000 years - clearly long enough to make populations shrink to a tiny fraction, in comparison. Also, the climate never fully recovered, remaining much cooler and dryer, with severe fluctuations. That's not a time of driven population expansion and migration. So, let's say indeed these people moved into the subcontinent and still had a couple of thousands of years of preferable climate left, or else got stuck before cooling struck, and instead made it during the short recovery @100,000ya - not a huge difference.

      Which path would they have the best chance of being adapted to and multiply? Follow the savanna - not the dangerous, disease-infested, unfamiliar, and erectus-inhabited tropical rain forest. Also, during cooling and onset of a dryer climate, it was the savanna that conveniently expanded. In addition, the (sub-Himalayan) North offered additional great access to year-round fresh water and wildlife.

      At this point, AMHs still had another 50,000 years before they started to (almost) completely wipe out the native early homos and generally learned how to adapt to different climates and niches. They probably were half way there just before Toba, and then, another 10,000-20,000 years later, were ready to go almost anywhere.

      The thing is, we have to explain this almost 50,000 year delay. It's huge. I think the best explanation is that AMHs:

      - were initially not culturally/mentally ready
      - had initially a clear disadvantage (unfamiliarity of local conditions and disease-wise) compared to the "natives."

      Just imagine exploring groups trying to head into the rain forest, never to come back, or if, deliriously diseased, or badly wounded - speaking of some wild and ferocious forest people. That would scare any pioneer today - even more so then, when before further adaptations the difference was small, and the advantages of the natives plenty. AMHs would have concluded to stick with what the knew: the savanna and cooler/dryer locations - which were then expanding, thus offering a much more fitting outlet.

      Then, again, starting ~70,000ya climate worked AMHs' way with ever expanding zones of dry and cooler savanna.

      As to the West Asian portion of Amerindians' origin, of course one needs to explain how and when their progenitors got to the Altai (or even further west), in the first place. AFAIK, there is no archaeological evidence from before ~40,000 ya - that's getting pretty late on the way to Beringia. And, those haplogroups don't seem to be related to the ones moving into Europe at the same exact time. Perhaps they came from the East, instead, initially, expanded and diversified during good times, to then move back East.

      Delete
    4. One of the trees I was referring to was Dienekes', which actually includes branching. There are several others, one is "Scientific Reports 2, Article number: 355 doi:10.1038/srep00355l, Paleolithic Contingent in Modern Japanese: Estimation and Inference using Genome-wide Data, Yungang He et al." also recently discussed by Dienekes. However, every single tree I have seen behaves like this. If true, it would support a "holding population" west of SE Asia, and a split (inland migration north) from this population before SE Asia and coastal E Asia was populated in earnest.

      There are a couple of Denisovan admixture papers supporting a similar line-of-thought based on admixture - especially two-step admixture; the most important one is: "Reich et al., Denisova Admixture and the First Modern Human Dispersals into Southeast Asia and Oceania, The American Journal of Human Genetics (2011), doi:10.1016/j.ajhg.2011.09.005."

      Delete
    5. I also consider the early Abbassian window as possible but the evidence in South Asia (or elsewhere in non-Arabian Asia) is lacking for that period. Hence I imagine that there were flows towards the now desertic areas of Sahara, Arabia, etc. but that these did not feel the need to migrate further or found the route to a better place in South Asia until c. 80 Ka., at the very end of the Abbassia Pluvial.

      Using the "Saharan pump" theory in a different context, we must understand that the pump, like the heart or a sponge, first expands and absorbs and then contracts and expels. It may well have been only in the contracting phase when our ancestors were "climatically expelled" towards South Asia.

      Not sure what to do with all the savanna/jungle/erectus speculations. In general I'd bet for the riverine and coastal routes because these ecosystems are naturally the richest ones. I'd also imagine that H. erectus was not ever a big problem, i.e. a real challenge. It could steal children or maybe cause punctual conflicts but not really stop our species from doing what they wanted in general. Not much more problematic than, say, elephants or tigers or crocodiles.

      I mean those people were used to respect but also to kill such menacing beasts. They also were probably used to brawling and the occasional deadly conflict with other Homo sapiens. I think it's more likely that H. erectus just retreated once and again, with the occasional skirmish, until they were effectively extinct.

      It's more likely that we killed them than the opposite: our species can be very brutal and merciless: we are no "angels".

      A quite different case was the big brained and extremely muscular H. neanderthalensis. That one surely was a clear barrier for the expansion of our species.

      On the Altai-Amerindian issue, I see no problem with the dates: c. 40-30 Ka. in Altai, at the latest c. 20 Ka. push towards the East (attested in Mongolia, maybe earlier further North in Siberia?) and c. 17 Ka. colonization of America. Q would have first coalesced maybe c. 55 Ka. in Iran, Kurdistan or the the Persian Gulf Oasis (but its ancestor P seems from the area of Bengal-Bihar, while its "brother" R seems to be from NW South Asia).

      Re. the autosomal ML trees:

      Amerindians (or any other admixed population) must behave like that in ML trees of autosomal genetic pools: admixed populations always end up very high within one of the components' branches, in this case the East Asian one. But the high root does not mean that they branched earlier but that they have West Eurasian admixture.

      In order to understand the actual order of branching you need to look at the haploid genetics, which effectively belong to two different groups: East Asian and South/West Eurasian ones. Autosomal trees alone can't explain this, not when admixture is present (and it is way too often).

      Delete
    6. Maju,

      The main point I was trying to make is, if you accept that AMHs entered the subcontinent 100,000 or even 80,000 ya, you need to explain that 50,000 - 30,000 years delay before further, world-wide spread, and before large-scale wipe-out if competing homo lineages. To me, it seems most likely that the combined difficulties of non-adaptation to out-of-comfort climate zones (tropical rain forest or coldish areas, including their pre-existing, well-adapted homo lineages) and inferior mental/cultural capability were the main culprit.

      As to Amerindians, the earliest archeaological dates in the trail from northern Siberia to Beringia are just that: very early - as are some studies about the Beringian hiatus.

      ~50,000- 60,000 ya, most of Iran and surrounding regions were still in the hands of Neanderthals. What exact path did these Q-people take? I see different haplogroups going through the northern Pakistani and Afghani passes moving towards NE Europe. I'd better believe marching west north of the Himalayas ~45,000 - 50,000 ya, while multiplying, and then partially moving east, again.

      As to the trees, Dienekes' and that of the origuinal authors actually include admixture (side branches meeting again), although I agree we have to be careful about the interpretation, at this point.

      Delete
    7. I see. I totally discard "intellectual inferiority", more so when animals and other Homo species could do the same thing with less IQ.

      One reason may be that the actual (and not subjective) environmental pressure of the intermittent Arabian deserts (whose exact patterns are not well known but could have been a burden in the less humid periods was enough to keep populations small and very localized:

      "Uncertainties remain concerning the extent to which the climate deteriorated in the intervening sub-stages 5d (120–110 ka) and 5b (100–90 ka)". (from Rose 2012, on the Nubian complex in Dhofar).

      So there is a very substantial wet phase at the earliest MI5 (5e: roughly 128-120 Ka ago) and later two "less substantial" wet episodes at 110-100 (5c) and 90-74 Ka ago (5a).

      So most likely the people who explored Arabia and established themselves there in the wet phases, suffered in the dry ones (but we have no evidence that they emigrated in these) and re-expanded in the successive less plentiful wet phases, the last of which is the most certain date of emigration to South Asia.

      Why did not they arrive to South Asia earlier? Maybe:

      (a) Geographic ignorance + desertic barriers
      (b) Luck (among such small groups luck may have played certain roles), randomness in decision making or survival chances.
      (c) Lack of boating tech??? (Does not make sense because the Nubian culture would have needed it to reach out to Dhofar and the Persian Gulf group reported by Armitage at about the same dates looks totally coastal (there's people who denies it but they are forcing the evidence to adapt to their ideas instead of adapting their ideas to the evidence - wishful thinking).

      There's also the possibility that they actually arrived to South Asia but then why we don't find any evidence of expansion? I'd discard this one unless new evidence changes the conceptual landscape.

      So basically a mix of luck and deserts. They would have never been too many anyhow so maybe they never really felt the need to explore.

      (continues)

      Delete
    8. (comes from above)

      My best guess is that the (mostly L3) tribe that actually performed the migration to South Asia was maybe a late arrival to Arabia and was more used to moving around.

      That may be a cultural issue: how much nomadic is each group. Because a difference on when to move camp (every other day or just seasonally, for instance) may make a difference on the ability to cross the Baloch deserts and reach the much more promising lands of Sindh and the rest of the subcontinent.

      I wouldn't say that either option is "inferior" or "superior" because if moving often increases chances of finding empty non-exploited lands, moving little increases comfort and certainty and therefore also adds to survival. Probably, specially as population increased, peoples tended to the latter model (eventually culminating in the Neolithic Revolution - but I'm going too fast).

      For example, a highly nomadic behavior like that of the Roma People would probably have been rewarded in the context of an "empty" Eurasia but not so much in the context of a full one (which is the one they actually found). That's the kind of variables I imagine could have played into this issue.

      In any case the surrounding deserts (and Neanderthals) were enough to contain the expansion until a critical breach was opened.

      Also, per Petraglia, the wave of c. 90 Ka ago into Arabia, actually occupied more arid regions inland, essentially taking over the peninsula. This is something we do not see in the previous waves, limited to certain locations, often not far from the benevolent influence of the sea, which almost invariably (exception: coastal deserts) improves the climate by bringing rain and dew and reducing the extremism in temperatures.

      It's possible, I guess that this last wave, would have been readier to live in desert conditions, allowing some of them to overcome the barrier Baloch desert after due interaction with the peoples of the Persian Gulf marshlands (which may have been more inclined to use boats and such).

      Just speculating but, as you can see, I go through quite different paths than yours. I do not think that anything other than deserts, seas or Neanderthals acted as pressure. Cold might but that was only a serious problem quite further North, so nope.

      Delete
    9. I agree that much of this is speculation, today - but a better word for this is Gedankenexperiment - or simply thinking about different, most likely scenarios. There's nothing wrong with that; it's a very creative process that can lead to much insight - especially once further data arrive and as long as no one thinks things are set in stone.

      If you look at Google maps, you can see that there are plenty of wadis from the end of the Persian Gulf to about opposite of Ras-Al-Khamah. I really don't understand why there wouldn't be AMHs present here contemporary to the Arabian population boom(s) - unless Neanderthals dominated the slightly more northern regions and/or the Iranian/Iraqi marshlands (perhaps a hot spot for interbreeding, BTW).

      Now, there clearly is a gap (paucity of wadis) until about Pakistan, if moving along the shoreline. But all it would take is a few years of exceptional rainfall (within a wet period) for this gap to become easily traversable by lots of motivated people fleeing over-population.

      My motto is, if in doubt - it's the climate, stupid! (derived from ...economy... - in politics).

      So there is a very substantial wet phase at the earliest MI5 (5e: roughly 128-120 Ka ago) and later two "less substantial" wet episodes at 110-100 (5c) and 90-74 Ka ago (5a).

      On a more global level, the advantageous stable, wet (and warm) climate of ~130,000 - 110,000ya was never seen again until recent times. Yes, there are a couple of other, extremely short intervals - but I believe most emigrations happen under population pressure when conditions are good - not long after they have turned bad - especially if the adjacent regions are climatically more challenging.

      http://upload.wikimedia.org/wikipedia/commons/b/b8/Vostok_Petit_data.svg

      Delete
  8. "You mean Zomia".

    OK. The present paper concerns the 'Laotian region of Zomia'.

    "The Ngeubhinh Mouxeu rock-shelter at N 20° 42.98′, E 101° 00.96′ is just 80 km NE of the main channel of the Mekong".

    Thanks for that information. That is well within Maju's 'Zomia'. The Mekong (or more likely a tributary) probably provided fresh water but not necessarily a migration route.

    "I'm not even sure we should consider Yunnan as part of China proper"

    It is considered 'China proper' in all the relevant papers.

    "At the very least South China means most of the Yangtze and all Pearl River basins"

    Which makes the distinction between 'north' and 'south' meaningless. 'South China' includes virtually all of the heavily populated region of China.

    "So South China is not 'landlocked' even if Yunnan is".

    Yunnan is part of Zomia, and Zomia is landlocked.

    "Maybe but there is a high diversity of R, quite curiously, second only to South Asia".

    The 'high diversity' of R in Sundaland (which is doubtful anyway, unless you're counting R11'B6/B4'5/R21 as three haplogroups) is of no relevance whatsoever in regard to M's diversity there.

    "concept "landlocked" makes no sense without precise boundaries such as those of modern states".

    Rubbish. It is obviously a concept foreign to your way of thinking but even today humans occupy ecological regions.

    "but Laos?! The sea is just a few hundred kilometers away either downstream by the Mekong or, shorter but harder, across the mountains at what is now Vietnam".

    An even today no ethnic groups occupy both the coastal regions and the mountains. They are separate ecological regions.

    "100 km across mountain and jungle can be journeyed by most Papuans"

    Possibly today, but can you provide any historical examples?

    "I'm not sure in this context that a 'Southern route' and a 'coastal route' are synonymous".

    I doubt that they a synonymous in any realistic context.

    "In practice, I think route that is primarily coastal and makes its way inland via tracing rivers to their higher altitude sources and only tertiarily away from coasts or rivers, is a model that is a pretty good fit to almost all pre-Austronesian human migrations".

    The distribution of mt-DNA M haplogroups argues against such a scenario.

    ReplyDelete
    Replies
    1. (1) There's no fixed frontier for "China Proper". My definition is that where Han/Hui Chinese are nearly 100% of the population, and, by that definition Yunnan is out, as is Guangxi. But up to you.

      (2) The distinction between North and South China is still important and there are very important areas in population numbers in North China. In fact most of the most densely populated provinces of China are North of Shanghai.

      (3) "Zomia is landlocked"? Natural landscapes can be described that way, much less when the sea is just 50 km away. Why do I have to even discuss this again?

      (4) Of all mtDNA R (18 basal sublineages), 8 lineages are from South Asia, 7 from SE Asia (incl. China and Melanesia) and 3 from West Eurasia. All those seven R basal sublineages are present in Sundaland or beyond (Philippines, Wallacea, Melanesia). In Sundaland itself we find several clearly Sundalander lineages: R22, R23, R21'12 (actually the R21 part because the R12 part is from Australia), which is 3/7. Some other lineages are from beyond Sundaland: R14 (from Nicobar to Papua, may still have originated in Sundaland), P (Papua, Philippines and Australia).

      While P seems the only clear beyond-Sundaland R lineage, R9 (pre-F) and R11'22'B seem the only plausible non-Sundalander lineages among all the Eastern set of R. So 4/7 (or 3/7 if you want to be nitty-picky about R14) are from Sundland most likely. That's pretty diverse in its context and, as I said, second only to South Asia.

      Ref. http://ourorigins.wikia.com/wiki/MtDNA_haplogroup_R

      The high diversity of R in Sundaland is interesting re. the secenarios of M and N, if we understand that the "R clan" may have pushed around some predecessors in SE Asia and Melanesia bearing M and N lineages, very possibly in association with Y-DNA MNOPS (MP in Melanesia, NO in Sundaland and mainland East Asia).

      (4) While is possible that today demographic and socio-politcal-economic pressure keeps ethnic groups contained in their specific valleys (when they go out, they are assimilated), that was no constraint at the time of the earliest colonization of the region: people very probably exploited mountain, lowlands and sea seasonally, as they did here in the Catabrian strip and other areas where mountains and sea are close. In the inland basin, they would have done the same between the river and the mountains. They were seminomads, remember?

      Your "separate ecological regions" are only complementary ecological niches when they are close enough in geography and make for a more diversified and sustainable economy (and Paleolithic peoples are documented to have used all the niches through seasonal or even circumstantial rotation).

      Delete
  9. "River valleys would have been extremely inviting to follow for migrating humans".

    Except in regions of tropical jungle where the vegetation is particularly dense and the ground underfoot particularly swampy.

    "There's no archaeological evidence of it continuing beyond Andrah Pradesh towards Bengal but it's possible that the evidence is underwater".

    Possible. But it's more likely that the Ganges Delta was impassable until effective boating had been developed.

    "In any case the shortcut through the Narmada route is quite obvious and apparent and precludes the necessity of a 100% coastal route towards SE Asia and beyond"

    And it is possible that the mt-DNA M haplogroup distribution supports such a route. However the 'coastal' element prior to arrival there is not supported by the mt-DNA evidence.

    "The main argument for the 'rapid coastal' route (emphasis in 'rapid' rather than just 'coastal') is genetic: there are several very old lineages that appear to be centered in SE Asia (notably Y-DNA C and D and mtDNA N) rather than farther West."

    I agree those haplogroups are particularly ancient in the east, but one basal M haplogroup in the Melanesian islands cannot possibly be older than about 30,000 years, at the most. Probably much younger.

    "The Ngeubhinh Mouxeu rock-shelter at N 20° 42.98′, E 101° 00.96′ is just 80 km NE of the main channel of the Mekong".

    And you'd be lucky to make even 10 kms a day through the jungle.

    "Also, the more tropical things got, humans probably stayed close to the coast or perhaps even stopped completely for a long period (dangerous, different environment/diseases, likely inhabited by erectus)".

    Above a particular altitude the vegetation thins out considerably. That is why the New Guinea Highlands were so heavily populated. They were far more inviting than the lowlands. In times of climate cooling the timber line drops opening a larger region of more open habitat.

    "most of the paternal lineages (Q) and some of the maternal ones (X2) are from West Eurasia (both these lineages are most diverse in West Asia and therefore must have originated there and spread to Beringia via Altai and other less precise Siberian locations".

    I Agree.

    "The East Asian component (which is surely dominant) of Amerindians is from the rather coastal zones of East Asia, mostly from NE Asia (but B is very rare so far North)".

    Which suggests that B entered America via the coast. It is common all along the East Eurasian coast north to Japan.

    ReplyDelete
    Replies
    1. (1) Between Andrah Pradesh and Bengal there is no "Ganges Delta". Check the map again, please.

      (2) Regardless of whether you might be right or not about how long it takes to cross jungle (it surely varies), 50-80 km is not any real distance: only a magical wall (or maybe the sea for some time) could stop people from going from A to B if the distance is so small. They are people, not trees.

      (3) We seem to agree re. Native American origins.

      Delete
  10. "(2) Regardless of whether you might be right or not about how long it takes to cross jungle (it surely varies), 50-80 km is not any real distance: only a magical wall (or maybe the sea for some time) could stop people from going from A to B if the distance is so small".

    You obviously have never had anything to do with a tropical jungle.

    "(3) 'Zomia is landlocked'? Natural landscapes can be described that way, much less when the sea is just 50 km away. Why do I have to even discuss this again?"

    Because you still do not understand that, although humans may be as extremely adaptable as you claim, human groups would first occupy the full extent of their preferred ecology before venturing into unknown habitats. For example the sea-going Polynesians occuppied coastal regions on even the largest islands. It was only with increasing population that they were forced inland. Zomia not landlocked? Have a look at this:

    http://en.wikipedia.org/wiki/Zomia_(geography)

    You can see that Zomia extends east of the Ganges Delta from the Khasi Hills through Burma and Yunnan to Laos and then south through Vietnam along the Annam Mountain range. As a result we should probably confine the 'Sunda ecological zone' to the region between Zomia and Wallace's Line. The coastal strip of Vietnam is not part of Zomia nor of Sunda. It has much in common with Southeast China.

    "(4) Of all mtDNA R (18 basal sublineages), 8 lineages are from South Asia, 7 from SE Asia (incl. China and Melanesia) and 3 from West Eurasia. All those seven R basal sublineages are present in Sundaland or beyond (Philippines, Wallacea, Melanesia)".

    Wrong. Six are from South Asia: R5, R6, R7, R8, R30 and R31. What do you claim as the seventh? I agree that R1, R2'JT and U are all present in South Asia but in the form of derived haplogroups. That would leave just two in West Eurasia: R0 and R3.

    "In Sundaland itself we find several clearly Sundalander lineages: R22, R23, R21'12 (actually the R21 part because the R12 part is from Australia), which is 3/7".

    I realise that the geography of SE Asia is not your main interest so it may come as a surprise to you to learn that the Lesser Sunda Islands are not part of Sundaland. They are part of Wallacea, the region that is neither Sunda nor Sahul. Sunda and Wallacea are two separate ecological regions. Of the eight remaining R haplogroups four are Wallacean (R12'21, R14, R22 and R23) and one is Sahul (P). That leaves just two as possible Sunda haplogroups: R9 and R11'B6/B4'5/R24. And they are much more likely to have originated in South China/Vietnam, not in Sunda.

    "R14 (from Nicobar to Papua, may still have originated in Sundaland)"

    More likely to be yet another Wallacean haplogroup.

    "The high diversity of R in Sundaland is interesting"

    There is no 'high diversity of R in Sundaland', except in your imagination.

    "the 'R clan' may have pushed around some predecessors in SE Asia and Melanesia bearing M and N lineages"

    Yet another of your postulated genocides?

    ReplyDelete
    Replies
    1. (3) Humans will explore everything, regardless if they prefer mountain or beach. When you are hunter-gatherer tribe, you must know your surroundings. We are not trees: we are curious intelligent adaptive beings who like to know what we're doing - and that implies knowing what goes on maybe 1000 km around (variable of course but for sure quite a bit more than 100km, which is a journey or few journeys walk). Your "boundaries" are arbitrary and unreal.

      (4) The South Asian basal sublineages of R are: R1, R3, R5, R6, R7, R8, R30 and R31 AFAIK. R1 (which has nothing to do with R0 nor N1 nor M1, which you may have in mind I guess) is only reported in South Asia, same for R3. Western Eurasian R subclades are R0, U and JT.

      I know well that Lesser Sunda is not Sundaland - but Bali is!

      Also:

      R21: the Orang Asli from Malaysia (R21) have nothing to do with Wallacea. The remnants of the journey to Australia (of its sister clade R12) are not found anymore in the island chain. Malaysia (both halves) was part of Sundaland in the Pleistocene.

      R22 is found in Indonesia (mostly), West Malaysia, Cambodia/Thailand, Vietnam, etc. That's "Wallacean"?! You should check (Min-Sheng Peng 2010) before you invent stories (wasting my time, angering me).

      R23 Bali!!! (and Sumba). Bali was part of Sundaland and they do not even navigate at all (they have beliefs against navigation, which they leave gladly to their Lombokese neighbors, which may well come from the time Bali was joined to Java and Asia).

      The only case you can argue maybe is R14 but the fact that it's also found in Nicobar (Austroasiatic speakers) suggests a Sundaland or otherwise mainland origin. Arguable but I do not think it is "Wallacean": it'd be the only lineage to expand from Wallacea in Western direction - very anomalous.

      "Yet another of your postulated genocides?"

      Maybe (if you wish to use that word, which I do not) - it'd be my only one, because I only see one such plausible "second wave" in the Eurasian expansion process, which is the one of mtDNA R and Y-DNA MNOPS (coupled). Or maybe they expanded all more or less together and there was no "genocide" but just randomness.

      But in this case I do see a second demographic pulse which partly displaced or reduced elements of the first one. However the displacement is only apparent in SE Asia, parts of mid-East Asia and Melanesia, because in South Asia or NE Asia there was no such displacement and the impact in the Far West was "genocidal" mostly towards Neanderthals, not the first wave people (and in fact incorporated male lineages from the first wave such as IJ and G).

      Delete
  11. "(1) Between Andrah Pradesh and Bengal there is no 'Ganges Delta'. Check the map again, please".

    I was not in fact claiming that there was. I was actually trying to draw attention to the fact that the Ganges Delta appears to separate South Asian haplogroups into Zomian and Indian, with few being common to both east and west of the Delta.

    "The high diversity of R in Sundaland is interesting re. the secenarios of M and N"

    It is actually the low diversity of R in Sundaland that is interesting re. M and N. Both M and N are reasonably well represented there. N21 and N22 qualify as Sunda as do M17, M23'75, M26, M31, M32'56, M47, M73'79 and, possibly, M21, M77 and M22. Those last three may be originally from Zomia though. Which brings us to the huge number of M haplogroups that look to have coalesced in Zomia: M7, M8, M9, M10, M11, M12'G, M13'46'61, M19'53, M24'41, M33, M39'70, M40'62, M48, M49, M50, M52'58, M60, M71, M72, M76 and M80'D, as well as the previous three. That's a total of 24 haplogroups. In contrast R looks to have no Zomian haplogroups and N has just two possible candidates: N8 and N11. In contrast neither M nor N have what might be called Wallacean haplogroups, unless we consider the possibility that R coalesced in Wallacea. Yet M, N and R all have Sahul representatives. Very interesting.

    ReplyDelete
  12. Regarding the 'great southern coastal migration route'. I'm sure you'll find this interesting:

    http://www.biomedcentral.com/1471-2148/8/45

    Quotes:

    "Although there is evidence of Neolithic and more recent expansions in the Arabian Peninsula, mainly detected by (preHV)1 and J1b lineages, the lack of primitive autochthonous M and N sequences, suggests that this area has been more a receptor of human migrations, including historic ones, from Africa, India, Indonesia and even Australia, than a demographic expansion center along the proposed southern coastal route".

    "However, the link found between the M Saudi 201 sequence and an M14 Australian sequence is puzzling. Although at first sight it could be taken as a signal of the connection between the two utmost ends of the southern route, it seems not to be the case. First, both lineages share three basal positions and this hypothetical link would considerably delay the arrival age of M in comparison to that of East Asia. It would be improbable that similar Australian links with other M lineages mainly from India were not found. Third, if the Arab lineage had such an old implantation in the Arabian Peninsula some detectable autochthonous radiation should be expected. Most probably, the M42 sequence belongs to an Australian clade and its related lineage found in Saudi Arabia is also of Australian origin. Historical links as those invoked to explain the presence of Indian and Indonesian sequences in the Arabian Peninsula pool should also be valid for this case".

    To your comments:

    "Humans will explore everything, regardless if they prefer mountain or beach. When you are hunter-gatherer tribe, you must know your surroundings".

    If it is relatively easy to get there.

    "R1 (which has nothing to do with R0 nor N1 nor M1, which you may have in mind I guess) is only reported in South Asia"

    Not according to that very reliable source Wikipedia:

    http://en.wikipedia.org/wiki/Haplogroup_R_(mtDNA)

    It is the derived form R1a that is found amoung Brahmins of Uttar Pradesh. It is probably an immigrant there as R1a1 is found in Eastern Europe and the northwest Caucasus. The same source gives R3 as Armenian. Neither therefore can be described as 'South Asian'.

    ReplyDelete
    Replies
    1. I know of Abu-Amero 2008 since its publication date, thanks. It was a pioneer study on Arabian mtDNA genetics, at least in the open access section of knowledge (there was some other paper on Yemen previously, I think, but was PPV and therefore largely ignored).

      Even in Abu-Amero 2008, which is not focused on L(xM,N) as its near-contemporary Behar 2008, we see the following observations:

      Particularly, Yemen has the largest contribution of L lineages [30]. So, most probably, this area was the entrance gate of a portion of these lineages in prehistoric times, which participated in the building of the primitive Arabian population. Later, received gene flows from North Africa and the Near East, and suffered expansions and retractions in humid or arid climatic periods. These fluctuations are also reflected in the frequent loss of diversity for several African clades as the L6 in Yemen [30] or the L5 in Saudi Arabia. However, the presence of western Africa L lineages and the different composition of L subclades in the African pool of different countries might reflect unequal participation of the primitive and the recent slave trade substrates in their respective African components.

      Note [30]: Kivisild 2004: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1182106/

      ... which is also interesting. I have used mostly Behar 2008 for my reconstruction because it lists many lineages in Arabia that are most unlikely to have arrived recently and should belong to the earliest flows into Eurasia from Africa, i.e. those related to the Out of Africa migration (if not exactly the same ones).

      Delete
    2. Re. R1 and R3 I must admit you are right. I was trusting my memory and seems not. Palanychami 2004 (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1182158/) does not mention R3 as I though, and, although he does mention R1, he classifies it as West Eurasian, not South Asian. That makes the landscape of R more interesting and somewhat more westerly in origin (maybe more Madhya instead of Bengal? It certainly weights a bit to exclude a SE Asian origin).

      But all the rest stands, it seems to me.

      Delete
    3. Caution, I've tracked the reference for R3 and "Armenia" wrong: the sequence now known as R3 was described as R1 by Maca-Mayer in 2001 and is from an Homo sapiens (in case anyone had doubts) from Jordan.

      http://www.ncbi.nlm.nih.gov/nuccore/AF381997

      Via: http://www.mtdnacommunity.org/human-mtdna-phylogeny.aspx (where AF381997 is described as the only R3 known ever).

      Delete
  13. "I know well that Lesser Sunda is not Sundaland - but Bali is!"

    I presume you're referring here to R23 which is also found in Sumba, which is east of Wallace's Line. If R23 is not actually Wallacean it is at least partly so. The haplogroup could very easily have coalesced in Wallacea and moved more recently to Bali. Bali has been the destination for much movement in Indonesia for many centuries, if not longer.

    "the Orang Asli from Malaysia (R21) have nothing to do with Wallacea. The remnants of the journey to Australia (of its sister clade R12) are not found anymore in the island chain".

    But the ancestor of R12'21 must have moved through the island chain at some stage, either as R12'21 or one or other already formed haplogroups. My bet is that haplogroup R12'21 coalesced in the lesser Sunda Islands and R12 went east to Australia and R21 went west to Sundaland. The 'original' haplogroup has been replaced in Wallacea.

    "R22 is found in Indonesia (mostly), West Malaysia, Cambodia/Thailand, Vietnam, etc."

    R22 is especially common in the Lesser Sunda Islands. Why don't you bother checking the references I've provided for it? Eastern Indonesia is across Wallace's Line anyway. The haplogroup is quite likely to have expanded to the mainland after it had coalesced.

    "R23 Bali!!! (and Sumba). Bali was part of Sundaland and they do not even navigate at all (they have beliefs against navigation, which they leave gladly to their Lombokese neighbors, which may well come from the time Bali was joined to Java and Asia)".

    Check your geography. As I said above Sumba is across Wallace's Line. And so is Lombok of course. Representatives of the haplogroup somehow managed to cross Wallace's Line.

    "The only case you can argue maybe is R14 but the fact that it's also found in Nicobar (Austroasiatic speakers) suggests a Sundaland or otherwise mainland origin".

    Again why don't you check those references you demanded I provide? Why did I bother? R14 is found in New Guinea, the Nicobar Islands (which were settled far more recently than was New Guinea) and in the Lesser Sunda Islands, in Wallacea. None of your arguments for a Sunda origin for any of the above haplogroups is very convincing at all.

    "Arguable but I do not think it is 'Wallacean': it'd be the only lineage to expand from Wallacea in Western direction - very anomalous".

    That's really rich coming from someone who has just written:

    "Humans will explore everything, regardless if they prefer mountain or beach. When you are hunter-gatherer tribe, you must know your surroundings. We are not trees: we are curious intelligent adaptive beings who like to know what we're doing - and that implies knowing what goes on maybe 1000 km around (variable of course but for sure quite a bit more than 100km, which is a journey or few journeys walk). Your 'boundaries' are arbitrary and unreal".

    Why on earth would people not move west as well as east once they had the technology? Anyway R14 must have moved in a westerly direction to reach the Nicobar Islands. And the other haplogroups you've mentioned all demonstrate at least some westward movement.

    "But in this case I do see a second demographic pulse which partly displaced or reduced elements of the first one".

    Why just two? Surely humans have been moving in many directions since they first emerged from Africa. And the link I provided re. the southern migration claims considerable westward movement from as far east as Australia, Why are you still looking for some elusive single 'Garden of Eden'?

    ReplyDelete
  14. "But the ancestor of R12'21 must have moved through the island chain at some stage"...

    Not really. Only R21 (or pre-R21) must. R12 is clearly the remnant of the ones who stayed behind, while R21 is the founder effect in Australia.

    Of course if you use decontextualized geometry the centroid falls somewhere between Malaysia and Australia but then the origin of R as a whole must be to the NW, well into India, so the corrected centroid is close to the location of R12, in Sundaland.

    "R22 is especially common in the Lesser Sunda Islands".

    I could not find that precision of detail, sorry: just "Indonesia" (where it's also basally most diverse). And considering that the rest of R22 is found North of Indonesia, in Indochina (and not Australasia), it's logical to think that the origin was in Sundaland rather than Lesser Sunda.

    Prove me wrong.

    "Sumba is across Wallace's Line"...

    I know. But it's 1 vs 1.

    "Representatives of the haplogroup somehow managed to cross Wallace's Line".

    Naturally. Nobody said the opposite but we can reconstruct a number of scenarios in which populations (Austronesians for example but also early Neolithic peoples, first settlers who eventually, some, arrived to Sahul...) migrated from Sundaland to Wallacea, while the opposite is not really apparent in anything.

    "Why on earth would people not move west as well as east once they had the technology?"

    I don't say it's impossible but it's not evident in anything. If you'd be right it'd be the first such reported Wallacea->Sundaland migration and that is a major novel claim that requires strong support from evidence (extraordinary claims need extraordinary evidence).

    The main barrier would be demographic anyhow: any flow would be absorbed ('drifted out' leaving no trace) unless very numerous (a full fledged migration). Also because of its more centric location Sundalanders were likely to have the technological edge over Wallaceans. Communication is important and Wallacea is a very remote region.

    But, again, prove me wrong. But prove it not just speculate, rant, confront, antagonize, cry... Proving things require less overreaction and more thoughtful dedication: it's not a mere job of responding to correspondence with the first thing that comes to mind (in fact that may be at times a distraction) but one of thinking, drafting, checking evidence, thinking again... and being a bit more humble than you are about what you know and about what we as Humankind-in-2012 know.

    "And the link I provided re. the southern migration claims considerable westward movement from as far east as Australia"...

    What link? Abu-Amero 2008. The lineage has been redescribed since then and is not anymore considered to have any relation to Australia at all. I'm not sure (feel free to check the sequences) but it seems to be a South Asian related lineage, possibly now known as M36-something (???)

    Please do not build a theory on such a feeble "finding" without first checking the facts.

    In any case, there is not a single piece of evidence supporting any westward migration from Australia. Your "evidence" is very elusive (and illusory) but even if it was true, it'd just mean most likely that an ancient haplogroup had split in two branches one heading to Australia and the other to Arabia - nothing unusual.

    Demonstrating a westward migration from Australia would require that there was (hypothetical not simple past, do you say "there were"?) clear clearly greater basal diversity in Australia, much like there is more basal diversity (without any kind of doubt) for R22 in Indonesia when compared relation to Indochina, or for the human mtDNA (or L3 if you wish) in Africa and not Eurasia (or anywhere else).

    ReplyDelete
  15. "I know of Abu-Amero 2008 since its publication date, thanks. It was a pioneer study on Arabian mtDNA genetics"

    Thanks for the following:

    "Abu-Amero 2008. The lineage has been redescribed since then and is not anymore considered to have any relation to Australia at all. I'm not sure (feel free to check the sequences) but it seems to be a South Asian related lineage, possibly now known as M36-something"

    And the discussion in the paper regarding this quote is particularly interesting:

    "Particularly, Yemen has the largest contribution of L lineages [30]. So, most probably, this area was the entrance gate of a portion of these lineages in prehistoric times, which participated in the building of the primitive Arabian population".

    Yes. But the author stresses that was after any realistically dated OoA. And those lineages contributed only to the 'building of the primitive Arabian population'. Nowhere else. Read all the paper. You even quote:

    "the presence of western Africa L lineages and the different composition of L subclades in the African pool of different countries might reflect unequal participation of the primitive and the recent slave trade substrates in their respective African components".

    I'm prepared to accept that much of the flow predates the 'recent slave trade' though.

    "That makes the landscape of R more interesting and somewhat more westerly in origin (maybe more Madhya instead of Bengal?"

    Only in your imagination. How does it do so? It certainly seems as though R became a very widespread haplogroup, and it makes sense that it had some supreme advantage over the earlier haplogroups. I agree that some already coalesced M and N haplogroups may have been carried along to some extent though.

    "It certainly weights a bit to exclude a SE Asian origin)".

    How does it do that? The greatest diversity in the haplogroup is still in the Far East.

    "Caution, I've tracked the reference for R3 and "Armenia" wrong: the sequence now known as R3 was described as R1 by Maca-Mayer in 2001 and is from an Homo sapiens (in case anyone had doubts) from Jordan".

    Thanks for following that up. So it's from Jordan, not Armenia. And certainly nopt from South Asia. Feel free to change the entry on Wikimedia.

    "Not really. Only R21 (or pre-R21) must. R12 is clearly the remnant of the ones who stayed behind, while R21 is the founder effect in Australia".

    On what grounds do you claim that to be so? I actually agree with you though, but not for the reason you would wish. You have the haplogroups round the wrong way. R12 is Australian. So 'R12 is clearly the remnant of the ones who stayed behind, while R21 is the founder effect in' the Malaya Negritos. Either way, one of the haplogroups must have passed through the Lesser Sunda Islands.

    ReplyDelete
  16. "Of course if you use decontextualized geometry the centroid falls somewhere between Malaysia and Australia but then the origin of R as a whole must be to the NW, well into India"

    Why are you so obsessed with India? Surely R could have originated almost anywhere its basal haplogroups are found today. You have consistently been quite happy to use the centroid as indication of origin except when such use fails to fit the belief you already hold.

    "I could not find that precision of detail, sorry: just 'Indonesia' (where it's also basally most diverse). And considering that the rest of R22 is found North of Indonesia, in Indochina (and not Australasia), it's logical to think that the origin was in Sundaland rather than Lesser Sunda".

    I can't actually remember where I got the 'especially in the lesser Sunda Islands' remark from but in the study referenced it is found amoung the Cham of South Vietnanm. The Cham are Austronesian-speaking and most accept that the Austronesian expansion began somewhere in the islands of Indonesia. The only Indonesian islands that were part of Sunda are Sumatra, Borneo and Java. The others are across Wallace's Line. So it's logical to assume that the origin was in Eastern Indonesia rather than in Sunda.

    "I know. But it's 1 vs 1".

    So we'll agree that R23 stradles Wallace's Line. I'm quite happy to accept that.

    "But prove it not just speculate, rant, confront, antagonize, cry..."

    That is what you have been doing the whole of this comments section.

    ReplyDelete
  17. Sorry Eurologist. I missed your comments:

    "Which path would they have the best chance of being adapted to and multiply? Follow the savanna - not the dangerous, disease-infested, unfamiliar, and erectus-inhabited tropical rain forest".

    Im totally agree.

    "Just imagine exploring groups trying to head into the rain forest, never to come back, or if, deliriously diseased, or badly wounded - speaking of some wild and ferocious forest people".

    I agree with that too. Even African Pygmies are not spread right through all the Central African rainforest. Much of it was inaccesible even to them until fire was used to open the jungle to slash and burn agriculture.

    "AMHs would have concluded to stick with what the knew: the savanna and cooler/dryer locations - which were then expanding, thus offering a much more fitting outlet".

    Yes.

    "you need to look at the haploid genetics, which effectively belong to two different groups: East Asian and South/West Eurasian ones".

    Maju and I actually agree on that.

    ReplyDelete
  18. The greatest diversity of R, after the R1/R3 correction is still NOT in SE Asia (the number of basal lineages is identical (6) to that of South Asia) and the geographical centroid is moved westward - because West Asia (where R1 and R3 belong) is West of South Asia (where I imagined them in my previous calculations), so the resulting centroid, which is the putative origin, goes westward somewhat (approx 1/9 of a subcontinental region).

    (Now, you may want to correct it towards the putative origin of its direct ancestor, N, which is probably a bit more to the East but not too much).

    "Feel free to change the entry on Wikimedia".

    I do not edit Wikipedia anymore: not while they do not include the alternative theories in the 9/11 attacks page. I'm not cooperating with manipulating Zionist-Imperialist entities that allow the secret services to shape their contents and control their internal power structures.

    "You have the haplogroups round the wrong way. R12 is Australian".

    Alright, then the other way around.

    Whatever the case the direction of migration within R is from mainland Asia (wherever) towards the periphery. There's no doubt about that, and Australia is clearly in the periphery and very specially in relation with R. I fail to understand what's the difficulty here.

    "Why are you so obsessed with India?"

    I'm not: the geography of Eurasia (at least all the non-frozen parts of Eurasia) is centered on South Asia. It's as simple as that. I also claim an important role for SE Asia (but not for Wallacea particularly). It's clear that our ancestors, the ancestors of all non-Africans, scattered first through Tropical Asia and from there to everywhere else.

    "You have consistently been quite happy to use the centroid as indication of origin except when such use fails to fit the belief you already hold".

    I use the same method all the time: basic centroid and corrected centroid, both. And they say all the time the same (with minor variations). Can you understand that no matter if you "move" a sublineage from here to there, it has a very little weight on the overall result of 18 lineages?

    You should get the map and calculate the centroids yourself instead of forcing me to waste my time discussing this with you.

    Because each half an hour I waste in the comment sections is, maybe, half an hour I'm not doing something more useful.

    "the study referenced it is found amoung the Cham of South Vietnanm"

    It is not only found among them: it's elsewhere in Indochina. You can check that yourself and save me time.

    "The only Indonesian islands that were part of Sunda are Sumatra, Borneo and Java".

    Wrong: Bali (and several other small islands) were part of Sundaland: Bali is this side of Wallace Line, which goes through the Lombok Strait.

    Wikipedia: "When sea levels dropped during the Pleistocene ice age, Bali was connected to Java and Sumatra and to the mainland of Asia and shared the Asian fauna, but the deep water of the Lombok Strait continued to keep Lombok and the Lesser Sunda archipelago isolated".

    Mr. Wallace Jr., you, should know that.

    ReplyDelete
  19. "you need to look at the haploid genetics, which effectively belong to two different groups: East Asian and South/West Eurasian ones".

    I disagree. We now are starting to have tools that allow us to explore the full autosomal genome. You can't tell me that the information contained in there (albeit, admittedly, still difficult to extract) is not orders of magnitude more meaningful than that of y-DNA or mtDNA - both because it doesn't get lost as easily, and that because of sheer size it has a greater phenotype impact.

    ReplyDelete
  20. "The greatest diversity of R, after the R1/R3 correction is still NOT in SE Asia (the number of basal lineages is identical (6) to that of South Asia) and the geographical centroid is moved westward - because West Asia (where R1 and R3 belong) is West of South Asia (where I imagined them in my previous calculations), so the resulting centroid, which is the putative origin, goes westward somewhat (approx 1/9 of a subcontinental region)".

    OK. South Asia is a huge region so let's take a look at these Indian haplogroups, using the information as we have it at present:

    Two of them (R7 and R8) are considered to be associated with the early Austro-Asiatic speakers: the Munda. My guess is that they didn't come in with the Munda. They were just the first Indian haplogroups the Munda picked up. Both are at least eastern India anyway.

    R7 is centred on NE India and probably coalesced there. It spread south as far as Orissa and Andhra Pradesh. That it is an eastern Indian haplogroup is supportyed by the fact that it was not found in the Bhil, the most westerly of the Madhya Pradesh tribals tested.

    R8, in contrast, was not found in the most northerly of those tribals, the Sahariya. It looks to have coalesced south of R7, in Orissa, and spread up the Godavari River. It especially spread further south as far as Sri Lanka, and west as far as Maharashtra.

    R6 too is found along the southeast coast of India. But also through Uttar Pradesh, the Ganges River valley, and northwest to Kashmir and Pakistan. Presumably it coalesced somewhere in the lower Ganges, between R7 and R8.

    R30, surprisingly, was not found in the Madhya Pradesh tribals at all. It is widespread in India though, found along the Ganges valley in Uttar Pradesh (my guess is that it coalesced there) and out into northwest India, presumably north and west of Madhya Pradesh: Punjab, Rajasthan, Gujarat and Maharashtra.

    The remaining two Indian R haplogroups, R5 and R31, were found only in the most westerly of the Madhya Pradesh tribals, the Bhil, but not in the other two. Presumably the haplogroups' spread was mainly to the west of Madhya Pradesh. R5 is found through much of India, except the southeast coast. It is especially common along the southwest coast and beyond Cape Comorin in Sri Lanka. R31 is common in Rajasthan, Southern India and Sri Lanka.

    So there you have the pattern of R's expansion through India revealed by the geographic distribution of the basal haplogroups:

    Basal R entered Northeast India (where R7 coalesced) from somewhere further east, and then spread into the Ganges valley (where R6 coalesced) and down the east coast, where R8 coalesced in Orissa and moved up the Godavari River and as far south as Sri Lanka. These three haplogroups are eastern Indian versions of R.

    From the upper Ganges basal R spread south (possibly originally via the Indus) along the west coast of India (R5, R31), through southern India to Sri Lanka. From Northwest India basal R emerged into SW Asia, where U, R2'JT, R1, R0 and R3 coalesced.

    So R's spread is coastal and along the major river systems, and rapid. Such a spread could only have been possible if it was through an unexploited habitat. That is an eloquent argument against an original coastal OoA expansion.

    ReplyDelete
    Replies
    1. R7b is mostly Dravidian (more common among tribal than caste) and located further south in Andrah Pradesh. Only R7a is as you say (mostly Austroasiatic tribal but also present at low frequencies in IE and Dravidian speaking groups): http://www.biomedcentral.com/1471-2148/8/227 (Chaubey 2008). Please do not just trust Wikipedia at face value because sometimes it is wrong or incomplete (in this case it only showed half the picture).

      On the rest I can't follow your logic.

      Delete
  21. "Now, you may want to correct it towards the putative origin of its direct ancestor, N, which is probably a bit more to the East but not too much"

    Why would the region in which N coalesced have anything to do with where R coalesced? I agree that R must have coalesced somewhere within the region thet N had reached, but R coalesced AFTER N had expanded.

    "Alright, then the other way around. Whatever the case the direction of migration within R is from mainland Asia (wherever) towards the periphery. There's no doubt about that, and Australia is clearly in the periphery and very specially in relation with R. I fail to understand what's the difficulty here".

    If one haplogroup of the pair had left the region where the ancestor lived, while the other was a stay-at-home, we would expect the parting haplogroup to be a subclade of the resident one. That is exactly what we see with M31, but we don't see the phenomenon here. The two haplogroups split two mutations from basal M and went their separate ways. I agree that is not a deciding argument in favour of them both having moved, but it is perhaps indicative of such.

    "I use the same method all the time: basic centroid and corrected centroid, both".

    And you 'correct' the centroid to wherever it suits your pre-existing belief.

    "It is not only found among them: it's elsewhere in Indochina. You can check that yourself and save me time".

    And everywhere it is associated with Austronesian-speaking people.

    "Wrong: Bali (and several other small islands) were part of Sundaland: Bali is this side of Wallace Line, which goes through the Lombok Strait".

    Sorry. Yes I knew that because I wrote yesterday:

    "So we'll agree that R23 stradles Wallace's Line. I'm quite happy to accept that".

    However it remains a fact that most of the smaller Indonesian islands are east of Wallace's line. They are the regions most reliant on boating for survival.

    "I disagree. We now are starting to have tools that allow us to explore the full autosomal genome. You can't tell me that the information contained in there (albeit, admittedly, still difficult to extract) is not orders of magnitude more meaningful than that of y-DNA or mtDNA - both because it doesn't get lost as easily, and that because of sheer size it has a greater phenotype impact".

    I agree totally with your comments here. However the haplogroups show a more complicated scenario than does the autosomal DNA. Y-DNA Q almost certainly entered Central Asia from NW South Asia. It dominated the male line but the autosomal DNA kept being diluted as the migrating population moved towards America. Likewise mt-DNA X is not an East Asian haplogroup but again its associated autosomal DNA was diluted as East Asians were accepted into the migrating population. Interestingly many claim that the earliest Americans were surprisingly un-East-Asian-looking. Could be significant.

    ReplyDelete
    Replies
    1. "Why would the region in which N coalesced have anything to do with where R coalesced?"

      We have discussed this before and you agreed. In fact you seem to forget about some stuff now and then and then you even proposed it as if it was your own idea in one of our endless discussions.

      The reason is, I will refresh your mind, that if we assume that there was a process of vectorial expansion from an origin outwards, usually in a predominant direction, the purely geometric centroid will not show that vectoriality, so a correction towards the ancestors, after calculating the geometric centroid makes sense. For example L3 in Eurasia+ shows a pattern of expansion Eastwards, so, after estimating the geometric centroids of M and N we may wish to correct them towards East Africa.

      Inversely the centroid of R would need to be slightly corrected towards Indochina or Bengal instead.

      If you don't like, up to you. I do.

      "And you 'correct' the centroid to wherever it suits your pre-existing belief".

      No. To where the ancestor is.

      And you agreed with it just weeks ago.

      And you even proposed such a correction as if it was your own idea.

      And I see myself discussing all this again. WTF!

      "And everywhere it is associated with Austronesian-speaking people".

      No. There's no Austronesian-speaking people in Thailand and Cambodia.

      Think twice, please!

      "They are the regions most reliant on boating for survival".

      Blah-blah: no data: pointless preconception underlining rant.

      They are also the smallest areas holding the smallest populations and the most isolated ones often.

      Even in the frame of Austronesian migrations, Wallacea is not very important (Philippines and Borneo are instead).

      "Y-DNA Q almost certainly entered Central Asia from NW South Asia".

      There's very little Q in South Asia, however West and Central Asia... that's another story altogether!

      So almost certainly nope.

      Delete
  22. Y-DNA Q almost certainly entered Central Asia from NW South Asia

    I don't know. It just looks Siberian and Beringian, to me. And the parent, P, seems to be found mostly quite northerly in India (and perhaps with a mostly NE emphasis, data awaiting).

    ReplyDelete
    Replies
    1. Let's clarify carefully this matter for the time being, as you guys seem so interested.

      Following Wikipedia:

      Q* (unclassified) is found at low frequencies in South Asia but the source and markers used for testing are not really clear (just a cite from http://ycc.biosci.arizona.edu/new_binary_polymorphism/resultsanddiscussions.html)

      The next phylogenetic level is Q1/Q1a no mention of "asterisk" presence anywhere, which is split in two:

      > Q1b > Q1b1 (Europe, Western Jews, Central Asia, Sindh) - surely from West Asia therefore (it's hard to think that from Central Asia or Europe it would have penetrated into non-European Jews).

      > Q1a, which splits in four clades (mono or paraphyletic):
      >> Q1a*: Koryaks and a pre-Dorset individual from Greenland (very low diversity probably a single subhaplogroup)
      >> Q1a1: East Asia and Hazaras
      >> Q1a2: Iran, Lebanon, Turkey
      >> Q1a3: Europe, West Asia, South Asia and Native Americans - two main identified subclades:
      >>> Q1a3a > Q1a3a1: Native Americans
      >>> Q1a3b > Yemenite Jews

      So IMO Q1 is from West and Central Asia, whether Q as a whole is from South Asia is less clear - but it's plausible considering where P(xR,Q) is found: East India (Bengal, Bihar, etc.) and that R also looks originated in South Asia (and then expanded mostly Westward (to West Asia and beyond) and Northwards (to Central Asia and surroundings). It's likely that Q expanded more or less at the same time as R1 did initially but because of founder effects and such ended up in different proportions and ethno-geographic routes to the future (which is today).

      I find it difficult to discern the plausible origin of Q1a. Because while Q1a*-Koryak and Q1a1 are oriental, Q1a3 is widely scattered (WEA, Central Asia, South Asia and America) and Q1a2 is West Asian. One could well argue here for an Oriental origin of Q1a but that would seem to imply a migration eastward (at the Q1>Q1a level) and then a return westward (as fully evolved Q1a3 and Q1a>Q1a2, what has low parsimony value. For me a Central Asian origin seems more likely (and would fit well with the archaeology we know of, where Altai was pivotal, after c. 40 Ka, and was so mostly in a West->East direction). It still needs to be explained how it reached Yemen.

      Delete
  23. "R7b is mostly Dravidian (more common among tribal than caste) and located further south in Andrah Pradesh. Only R7a is as you say (mostly Austroasiatic tribal but also present at low frequencies in IE and Dravidian speaking groups)"

    So? R7 is still eastern Indian. And I pointed out that the association with Munda/Austo-Asiatic speakers was later than was its entry into India. So the concept of a primarily water-borne entry into India from the east still stands.

    "We have discussed this before and you agreed".

    I have never agreed that N and R coalesced in the same region. I have always maintained that R coalesced somewhere in SE Asia, if not further east, and N coalesced somewhere near Africa.

    "For example L3 in Eurasia+ shows a pattern of expansion Eastwards, so, after estimating the geometric centroids of M and N we may wish to correct them towards East Africa".

    Very much towards East Africa. The 'centroid' has nothing to do with it, especially if the particular expansion has been 'in a predominant direction'. If an expansion is unidirectional the origin will lie far from the centroid. I agree witht the research that suggests a Persian Gulf/moist Arabian Peninsula as origin for both M and N. But they do not both come from the same region within that single region.

    "Inversely the centroid of R would need to be slightly corrected towards Indochina or Bengal instead".

    To me the evidence suggests at least as far east as Indochina, and probably even further east.

    "No. To where the ancestor is".

    No. To where you like to believe the ancestor is.

    "And you agreed with it just weeks ago".

    In what context?

    "And you even proposed such a correction as if it was your own idea".

    I will have always given credit for any change in my opinion. Don't lie.

    "No. There's no Austronesian-speaking people in Thailand and Cambodia".

    That's not what those worthy people at Wikipedia say. Your advice is to never 'trust Wikipedia at face value', but here goes:

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

    Quote:

    "They are also found in Singapore, the Pattani region of Thailand, and the Cham areas of Vietnam (remnants of the Champa kingdom which covered central and southern Vietnam), Cambodia, and Hainan, China".

    "There's very little Q in South Asia, however West and Central Asia... that's another story altogether!"

    Then you go and contradict yourself:

    "So IMO Q1 is from West and Central Asia, whether Q as a whole is from South Asia is less clear - but it's plausible considering where P(xR,Q) is found"

    Thanks for the information, though.

    "For me a Central Asian origin seems more likely (and would fit well with the archaeology we know of, where Altai was pivotal, after c. 40 Ka, and was so mostly in a West->East direction)".

    That's basically what this ancient paper claimed:

    http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1377800/pdf/10053017.pdf

    ReplyDelete
    Replies
    1. "I have never agreed that N and R coalesced in the same region"...

      Am I saying that? Nope. I say that you agreed and even argued for a correction towards the ancestral centroid once the raw centroids had been calculated. Either we apply the correction to all or none.

      For you method is not important just what you have decided a priori to be "the truth". That's not science but religion.

      "In what context?"

      I can't know because comments are not searchable. It does not matter anyhow because what applies to one branch of the phylogenetic tree, applies to all others (unless you have some sort of Jesuit casuistic by which the rule is bent or even totally changed at whim - what is not scientific methodology).

      "the Pattani region of Thailand"

      That's not Cambodia/Thailand but the part of Malaysia that by historical accident fell in the borders of Thailand. Grasping for straws!

      The Nicobarese are also NOT Austronesian but Austroasiatic.

      "Then you go and contradict yourself"...

      I was not aware of any Q* in Pakistan. As of today I do not know which is the paper that establishes that - only a vague mention which may be wrong.

      But I am in the right to contradict myself: to change my opinion. Even in the same day. You should try it.

      Delete
  24. "Now, there clearly is a gap (paucity of wadis) until about Pakistan, if moving along the shoreline. But all it would take is a few years of exceptional rainfall (within a wet period) for this gap to become easily traversable by lots of motivated people fleeing over-population".

    Eurologist and I both seem to be thinking exactly the same on this subject.

    "I believe most emigrations happen under population pressure when conditions are good - not long after they have turned bad - especially if the adjacent regions are climatically more challenging".

    Obviously so, to me. By the time conditions have turned bad it is usually too late to move.

    ReplyDelete
    Replies
    1. "Eurologist and I both seem to be thinking exactly the same on this subject".

      So you now think that the OoA migrant population exited Arabia and arrived to South Asia by the coastal route through Balochistan? I'd congratulate you for your change of mind but I'm sure you are confused and saying what you don't really think.

      Delete
  25. Maju, thanks for your explanation on Q. Not to beat a dead horse, but do you have a hypothesis on how Q got to West Asia or the Altai?

    It seems clear that Q is very old, when even Q1a3 etc. must be very old. Perhaps one needs finer resolution (and more data), here, to understand where the highest diversity is at each sub-level.

    ReplyDelete
    Replies
    1. I always give as granted that Q (Q1) and at least some R1 (R1b) reached West Eurasia (incl. Central Asia) in the genesis of the Upper Paleolithic, since c. 55 Ka. ago (based on the finding of H. sapiens bones in association with Aurignacoid industries in Palestine, which is the oldest known date for the West Eurasian colonization). This was probably also the case for IJ and G (derived from F but whose known ancestry is a thin line until their expand in the West). And of course is the case of the relevant mtDNA clades (R0, U, JT, N1, X, etc.)

      The technological aspects of the Aurignacoid industries (blades) have precedents in South Asia AFAIK and the phylogeny also seems to be rooted or at least transited through that region.

      I'd say that most clades which succeeded in this colonization process did in West Asia or Europe but Q (Q1 or Q1a) appears to be the lineage of those who succeeded in the remote North: Altai and Siberia (and later in America).

      Like mtDNA X2, Q1 is not a European clade (contrary to the Stanford clan of academic deceivers) but it does belong to the somewhat looser West Eurasian set of populations that formed in and after the Aurignacoid period (c. 55-30 Ka ago) and that were connected to each other at the origin if we are to judge from the similar technologies and the relatively limited (but not too narrow) gene pool (rather than single founder effect, array of a few founder effects).

      That's what I think.

      Delete
  26. "That's not Cambodia/Thailand but the part of Malaysia that by historical accident fell in the borders of Thailand. Grasping for straws!"

    In what way is that relevant? We were talking about mt-DNA R22's northern presence being a product of Austronesian expansion rather than being a product of ancient R expansion. Surely the modern political boundaries did not exist in either period. Because R22 is 5 mutations from basal R its presence on both the SE Asian mainland and in the Lesser Sunda Islands is extremely unlikely to be a product of early R dispersal. It is much more likely to be a product od more recent dispersal, probably Austronesian but perhaps Hoabinhian.

    "The Nicobarese are also NOT Austronesian but Austroasiatic".

    At no stage did I claim they were either Austronesian or Austroasiatic. You're imagining things again. The presence of R14 shows that the Nicobarese certainly have some connection with the Lesser Sunda Islands though. And New Guinea. The Austro-Asiatic language suggests an arrival in the Nicobars during the Hoabinhian. Some have seen a minor connection in New Guinea to the Hoabinhian also.

    "So you now think that the OoA migrant population exited Arabia and arrived to South Asia by the coastal route through Balochistan?"

    Eurologist's comments do not necessitate a coastal migration through anywhere. Merely a movement through what is now Iran during a time of diminished aridity.

    "I say that you agreed and even argued for a correction towards the ancestral centroid once the raw centroids had been calculated".

    Centroids are only a guide. The distribution of basal haplogroups and their branches is much more useful in determing haplogroups' origins. I'm sure you will find much to disagree with in my latest alterations to mt-DNA R:

    http://ourorigins.wikia.com/wiki/Mt_R_east_to_west

    ReplyDelete
  27. "We were talking about mt-DNA R22's northern presence being a product of Austronesian expansion"...

    Not at all. Why would a basal subclade of R, whose existence (at least in "pre" form) probably began c. 60 Ka ago or more be involved in full in Austronesian expansion. And how does Cambodia or Nicobar fit in? You are just obsessed about Austronesizing everything, as if there was no people on that part of the World before Austronesians!

    "http://ourorigins.wikia.com/wiki/Mt_R_east_to_west"

    And this means? I can just see a list of R subclades with some comments about their location (not sure if correct - didn't bother looking at everything).

    ReplyDelete
  28. "Why would a basal subclade of R, whose existence (at least in 'pre' form) probably began c. 60 Ka ago or more be involved in full in Austronesian expansion".

    Why ever not? Any number of different SE Asian clades were involved with the Austronesian expansion. In fact it is a bit debatable as to which were the first involved. Anyway R22's expansion cannot have taken place at the same time as basal R's because of its mutation 'stem'.

    "And how does Cambodia or Nicobar fit in?"

    We have very little information concerning Cambodia but it would be fairly reasonable to suppose its haplogroups were closely related to those from Thailand and Laos, and further afield Vietnam and Malaya. As for the Nicobar Islands, R14 is found there, as well as in the Lesser Sunda Islands and in New Guinea. That spread is unlikely to have been a product of Austronesian expansion but rather Hoabinhian.

    "And this means? I can just see a list of R subclades with some comments about their location (not sure if correct - didn't bother looking at everything)".

    It means that it is very difficult to make a plausible case for mt-DNA R having coalesced anywhere other than in Southern Wallacea, or very, very close to it.

    "but do you have a hypothesis on how Q got to West Asia or the Altai?"

    I've been waiting for Maju to follow this up, but here goes:

    Q is a branch within P. Further back P was a branch within MNOPS. Two of the four branches within MNOPS (M and S) are found entirely beyond Wallace's Line, and probably coalesced in New Guinea/Melanesia. That places the other two (NO and P) close by. Especially when we consider that MNOPS is a branch within K(xLT). K1 is a very rare South Asian haplogroup. K2 and K4 are found in Wallacea and further east, and K3 is found only beyong Wallace's Line. To me it is obvious that MNOPS, like mt-DNA R, must have originated somewhere very near Wallacea.

    NO moved north up the east coast of Eurasia forming N and O, and P moved west with mt-DNA R. Up the Ganges, where Y-DNA R formed leaving R2 behind. P and the remaining R carried on. R1 formed from basal R somewhere near the border between South and West Asia. Baluchistan? Q also looks to have formed directly from P around that time.

    "I always give as granted that Q (Q1) and at least some R1 (R1b) reached West Eurasia (incl. Central Asia) in the genesis of the Upper Paleolithic, since c. 55 Ka. ago"

    Fits exactly.

    ReplyDelete
    Replies
    1. "Any number of different SE Asian clades were involved with the Austronesian expansion".

      And Papuan clades and what not. Maybe but you'd have to demonstrate in each case: just because it's SE Asian it's not "Austronesian" by origin.

      "... it is very difficult to make a plausible case for mt-DNA R having coalesced anywhere other than in Southern Wallacea"...

      Go see your doctor, seriously.

      "To me it is obvious that MNOPS, like mt-DNA R, must have originated somewhere very near Wallacea".

      Agreed form MNOPS (not with the same logic as yours however): in Indochina or Sundaland (but surely not in Wallacea of all places). But mtDNA R is probably older and, unlike MNOPS has a centroid much farther West. While R(mt) and MNOPS(y) are largely coupled, their origin is not the same but rather they both belong to a wider process which is the Great Eurasian Expansion overall, which may well be much more mixed and confused that you imagine.

      It's very possible that R expanded Eastwards with IJK before the genesis of MNOPS as such and that the P men returning to South Asia, gathered with their relatives of other IJK and that included women of mtDNA R (but different subclades than the ones we find in East Asia and Australasia).

      The exact details... damn, we again lack of a proper documentary and we have to imagine them. :/

      "I've been waiting for Maju to follow this up"...

      I replied long ago: Q/Q1 spread westward with R1/R1b c. 55-50 Ka ago.

      Delete
  29. "just because it's SE Asian it's not 'Austronesian' by origin".

    I didn't claim it was 'Austronesian by origin'. Far from it. It's a basal R haplogroup so its origin goes way back. What I suggested was that its eventual expansion from its place of origin was related to the Austronesian expansion. It must have coalesced first somewhere in Indonesia (including the Lesser Sunda Islands) or in Malaysia, before it spread to all those places. It cannot have coalesced everywhere. I presume you've seen this at Dienekes:

    http://dienekes.blogspot.co.nz/2012/04/first-look-at-dna-of-neolithic.html

    Quote:

    "As I have argued elsewhere, the past seems to have been much more dynamic than many had suspected, and the people that walked and rowed to the ends of the Earth during the Upper Paleolithic did not suddenly grow fetters or decided to stay put during the Neolithic as many 'Paleolithic continuity' adepts had proposed".

    Very much the problem with your interpretation of the evidence here.

    "they both belong to a wider process which is the Great Eurasian Expansion overall, which may well be much more mixed and confused that you imagine".

    I can see that it is certainly much more 'mixed and confused that you imagine'.

    "Go see your doctor, seriously".

    So let's hear your evidence for an origin elsewhere.

    "But mtDNA R is probably older and, unlike MNOPS has a centroid much farther West".

    And I keep telling you that the 'centroid' is irelevant. Look at the distribution of the basal clades.

    "It's very possible that R expanded Eastwards with IJK"

    Are you now claiming that mt-DNA R coalesced in SW Asia? Surely you must accept that Y-DNA IJ coalesced there. If anything it was mt-DNA M that moved east with Y-DNAs K and F.

    "The exact details... damn, we again lack of a proper documentary and we have to imagine them"

    Rubbish. We now have reasonably detailed information about the distribution of most haplogroups. And more is being discovered every few weeks.

    ReplyDelete
    Replies
    1. You know well that I seldom agree with Dienekes.

      ... "the 'centroid' is irelevant".

      Then we disagree on how to interpret the data that we have. For me the centroid of a star-like node (haplogroup) is, if not the origin as such, highly informative re. the true origin of that haplogroup.

      You may want to correct it, you may want to state uncertainty ranges... but you can't declare it irrelevant, much less with an alternative model. And you have no alternative model but your whim.

      "Are you now claiming that mt-DNA R coalesced in SW Asia?"

      No, I think that F, IJK and K (and largely LT but concentrated only in part of the original scope) are a sequence of nodes from the same ancestral population in Northern South Asia. From this one branched out:(in this order even if it looks counterintuitive): H (South), MNOPS (East), and IJ and G (West, but only late, upon the return of the P clan).

      You may disagree and I can understand that many may disagree but that's how I perceive the Y-DNA side of things in the Great Eurasian expansion.

      We do not need to discuss this: we can simply agree to disagree. Not fun? There are many forums over there that will surely enjoy much more a hothead argumentative personality like yours. A blog is not exactly the same as a forum but even in a forum the potential of a debate eventually collapses once it's clear that no agreement will be achieved or no new evidence will be put forward for the time being.

      You asked for my point of view and I explained it, no need initiate another endless discussion at this point. I seriously prefer to let things rest for weeks, months... whatever is needed. It's more pleasant and usually fruitful.

      "We now have reasonably detailed information about the distribution of most haplogroups".

      Indeed but we will never know if Mary married Tom or if Tom Jr. is the son of Tom or that of Lucas (and what not!) We lack the soap-opera level of detail and unless you invent a time machine now we will always lack it.

      I accept that uncertainty. Can you?

      Delete
  30. "You know well that I seldom agree with Dienekes".

    Perhaps so, but on this occasion there is no doubt that Dienekes is absolutely correct.

    "For me the centroid of a star-like node (haplogroup) is, if not the origin as such, highly informative re. the true origin of that haplogroup".

    It is so easy to prove that to be incorrect. If members of a particular haplogroup move into a so-far unexploited region numbers will grow dramatically, throwing the centroid miles off.

    "And you have no alternative model but your whim".

    The distribution of basal haplogroups is far more reliable than is the centriod of a distribution.

    " I think that F, IJK and K (and largely LT but concentrated only in part of the original scope) are a sequence of nodes from the same ancestral population in Northern South Asia".

    Once again you're looking for a single origin 'garden of Eden'. JK can hardly be described as a 'South Asian' haplogroup. But I agree that LT can be described as such. So I agree that F, IJ and LT 'are a sequence of nodes', but from a shifting population. One moving from SW Asia through South Asia towards the east.

    "From this one branched out:(in this order even if it looks counterintuitive): H (South), MNOPS (East), and IJ and G (West, but only late, upon the return of the P clan)".

    Not just 'counterintuitive' but almost certainly wrong. The phylogeny shows that IJ branched off from IJK before LT did. Certainly long before MNOPS was formed.

    "We do not need to discuss this: we can simply agree to disagree".

    OK. I won't follow up on any reply you might care to make.

    ReplyDelete
  31. "JK can hardly be described as a 'South Asian' haplogroup".

    Sorry: IJ.

    ReplyDelete
    Replies
    1. I did not say that IJ is any South Asian haplogroup, I said that, in my understanding, IJK (and its parent F and its larger and probably older descendant K=LT+MNOPS) coalesced in South Asia. Only that explains, in my opinion, the resulting distribution of their descendant haplogroups.

      Delete

Please, be reasonably respectful when making comments. I do not tolerate in particular sexism, racism nor homophobia. Personal attacks, manipulation and trolling are also very much unwelcome here.The author reserves the right to delete any abusive comment.

Preliminary comment moderation is... ON (sorry, too many trolls).