An interesting paper published online in October 2010 (official publication date is however 2011) but that I had not visualized to date is:
G. Chaubey et al., Population Genetic Structure in Indian Austroasiatic Speakers: The Role of Landscape Barriers and Sex-Specific Admixture. Molecular Biology and Evolution 2011. Open access.
My head is still dizzy from the pains of quitting up heavy smoking without any help whatsoev (and without the real conviction of it being worth the effort, just bored of going through an unhelpful and demanding bureaucracy of self-righteous institutionalized robbers and drug dealers to get my dose), so I won't extend much upon this one.
Most importantly it comes up clear that Indian Austroasiatics are, in spite of often obvious East Asian male-mediated ancestry, mostly South Asian by overall ancestry:
 |
| Fig. 3 - PCA and K=7 analysis |
In the structure analysis, we see that the main component (black) of Indian Austroasiatic speakers is shared with Dravidian populations of South India (but not Brahuis, who are typical Pakistanis in spite of speaking a Dravidian language). This component is also shared at lower frequencies by South and SE Asian populations.
This is also apparent in the PC graph, where Indian Tibeto-Burman peoples cluster with East Asian ones but not Austroasiatic Indians, in spite of some lesser tendency in that direction.
In an aside, it is also notable to mention that the North Indian component (green) is clearly different from those in West Eurasia, even if there is a lesser presence of this South Asian component. This rather goes against those who claim that North Indian/Pakistani specificity is of West Eurasian origin: if anything some West Eurasian element would need to be explained as originating from South Asia. The only WEA population lacking it are, as usual, Basques, indicating it is at least a marker from outside Europe and, in Europe, surely marks Neolithic or Metal Ages (Indoeuropean) arrivals.
Back to Austroasiatics, it is worth quoting this:
The mtDNA haplogroup allocation of Munda speakers is similar to Dravidian and Indo-Europeans of the Indian subcontinent (Basu et al. 2003; Metspalu et al. 2004; Chaubey et al. 2007; Chaubey, Metspalu, et al. 2008; Chaubey, Karmin, et al. 2008; Thangaraj et al. 2009). We carried out a high-resolution analysis of those haplogroups of Munda speakers, which account for >4% of their maternal gene pool. All the seven maternal haplogroups found frequently in Munda speakers are autochthonous to India (supplementary fig. S5, Supplementary Material online) (Chandrasekar et al. 2009) and references therein, accounting altogether for 57% of the maternal gene pool of present Munda speakers. The extensive analysis of these haplogroups revealed relatively recent sharing of most recent common ancestors within these groups between AA and non-AA speakers (MRCA), suggestive of admixture; a similar result was observed recently for hg R7, which is the most frequent among these in AA speakers (Chaubey, Karmin, et al. 2008). The mtDNA lineages of Munda speakers do not cluster in basal parts of the tree (to founder haplogroups M, N, or R) but are spread among the derived branches that date to <10KYA (Supplementary fig. S5, Supplementary Material online) suggests that the mtDNA diversity found in contemporary Munda speakers is the result of admixture from neighboring populations of India.
Having these distinctions present is important to understanding what in these peoples related to SE Asia, is autochthonous from South Asia or recent (Neolithic) immigrant from SE Asia.
So, IEs moved out of India and spread their language in Europe?
ReplyDeleteI'm not saying that either. What I'm underlining is that there is a component that, in this study and at this depth (k=7), looks like genuinely South Asian (from the North) but not Indoeuropean-specific (Brahui and Burusho are very high on it).
ReplyDeleteRemember that what these graphs describe is affinity likelihood and South Asians are clearly "like" West Eurasians and vice versa to some extent, so they are going to show always some such affinity. Whether particular populations fall in or out of such components is probably often an artifact of the algorithms used rather than something categorically black or white.
It's not lineages but the nuclear genome what is being considered here and, in it, each marker is almost unavoidably going to have its own deep patterns. What we see is the overall synthesis for many markers, nothing else.
It might mark a late UP or Epipaleolithic migration, such as the one associated with R1a (rather mysterious from the archaeological viewpoint as of now) or LT. Maybe microlithism is a fashion that expanded from India (where it is clearly much older than elsewhere), carrying some genetic components, in addition to the overall genetic affinity... but I cannot say much for sure.
ReplyDeleteAnyhow, don't you think that the fact that Brahuis are clearly Balochis who retain Dravidian should revive the old theories about Dravidian being the IVC language and maybe related to Elamite?
ReplyDeleteI do not see room for pre-Iron Age Indoeuropeans there, though maybe some sort of superfamily can be hunched between Dravidian and IE? Or maybe Dravidian is an arrival from Iran (Elamite) and some sort of proto-proto-IE was spoken in the area too - but hard to support, right?
your guys' names always confuse me
ReplyDeleteanyway. blogged it last fall
http://blogs.discovermagazine.com/gnxp/2010/10/sons-of-the-conquerers-the-story-of-india/
It's a mere coincidence. I took up Maju because Sugaar (my older internet handle) sounded to "sugar" in English (when actually is a mythological name of clear fiery meaning). Maju is a variant name for the same God with unclear origins (maybe related to maypole celebrations, some say but these do not exist among Basques). Later I met up Manju elsewhere. The names are unrelated.
ReplyDeleteAnyhow, I may have read your article in the past, Razib, but I had forgotten. Though maybe that was the reason why Brahui-Balochi identity did not really surprise me when I saw it.
What is clear is that the "ANI" component, for whatever is worth, is not from outside of South Asia, but a regional duality that probably pre-dates Indoeruopean and even Neolithic flows. It does not look like there is a lot of light blue component ("Neolithic" or West Asian) anywhere, just among Balochi peoples what is normal.
There is more European (darker blue) component instead, among Pashtuns and Sindhis, however its presence among Burusho rather suggests it is pre-IE.
Take care of yourself. Quitting smoking is not easy. I have not had a cigarette in about a year and a half.
ReplyDeleteMaju : "There is more European (darker blue) component instead, among Pashtuns and Sindhis, however its presence among Burusho rather suggests it is pre-IE."
ReplyDeleteI understand why you would think that, nevertheless I can see a reason that would rather disprove that theory.
AFAIK the only haplogroup among the Burusho that could fit with a "European" element is R1a (at least _some of it_ if we consider that not all the south Indian R1a are exogenous to south India (at the moment it seems that R1a would rather originate in south Asia originally AFAIK)).
I'm not to sure about the mtDNA hgs but it doesn't look filled to the brink with European-like hgs IIRC.
A few Burusho do look like typical "European" (apparently, a few could even pass for Europeans from the north half of Europe, or close) even if it's rather rare.
So IMO the only possible origin for this "European" element would be the (still hypothetical) bronze age IE migrations, carrying some R1a coming from central Asia.
Otherwise we would expect at least some y-DNA I or R1b1b2 (at the very least some very clear mtDNA traces). But AFAIK it's no the case, leaving only a recent arrival of (some?) y-dna R1a to associate with this European element (since we know there were "European"-like populations in central Asia and south Siberia during chalcolithic and bronze age with mostly R1a).
We can always find a reason as to why a south Asian population with this "euro" element would speak a non-IE language. It's not like there is a real impossibility, here.
PS : Good luck for your ordeal (quitting smoking).
- the way I see it all these autosomal data strongly correspond to mtDNA. So I believe these represent pre-civilizational identity (10,000 years?). I assume only after civilizational phase male specific migrations/invasions took place.
ReplyDelete- Andrew commented at Razib's blog that Madagascar uniparental lineages 50-50 between Asian and African. However, autosomally they are 66% African and 33% Asian. If this is true, I believe we need to have a clear understanding on whether native genome has some advantages over non-native genome.
- The picture I get is original Europeans (dark blue), first absorbed Near Easterners (light blue) then north-west Indians (light green). Linguistic history cannot be constructed from this model
- I suppose two Indian components shown in this figure are north-west Indian and central-east Indian. North Indian and South Indian is misleading.
- If you remember some of the older studies, Irula, a South Indian tribe completely branches out in autosomal studies. I believe Irula represents true South Indian component.
- The above two points clearly support Proto-Dravidian homeland in central-east India. We do know all Dravidian language branches are found in this region. South India has only one branch (SD).
- According to archaeology farming started in central-east Indian region which gradually spread to South India around 3500 years ago.
- I'm not sure whether original South Indians (represented by Irula) had different languages. The fact they don't much differ from other Dravidians in uniparental lineages I suppose that is unlikely. Also, none of the isolated languages are found in South India also there is no study that talk about any substrate in Dravidian languages of the South.
Wagg: Burushos look to R1a like Basques to R1b, mutatis mutandi: neither gave birth to the lineage but the lineage is old among them probably and not Indoeuropean/Neolithic IMO.
ReplyDelete"We can always find a reason as to why a south Asian population with this "euro" element would speak a non-IE language. It's not like there is a real impossibility, here".
For me there is, why would they absorb IE blood in similar amounts as their IE-speaker neighbors and not change language as they did? Isn't more logical that Pashtuns and Sindhis changed language without absorbing much European blood.
The Euro component is probably a Central Asian one, related to H and U in the area, which are certanly pre-IE stuff - also R1a maybe (but less clearly so, because R1a may be original from India). Anyhow the Pashtun and other Central Asian groups have more dark blue component than the Burusho but these have more than Sindhi and any other South Asian population shown, IE or Dravidian.
"A few Burusho do look like typical "European""
A lot of people from that area look more or less European (but not on average), typically with a Nordic (Dutch-like) tendency when they do. This is probably more because they are ancestral to Europeans (specially NE European) populations than because they being heavily penetrated by them.
After all West Eurasians had to come from somewhere and our genetic pool is rather narrow (essentially three mtDNA R subclades and three or four Y-DNA lineages: IJ, G and R1). All those must have coalesced somewhere between NW India and Iran, maybe Afghanistan and Kashmir in particular. For me people like the Burusho and the Kalash surely represent leftover diversity from those times, rather than unlikely back-migrations.
Though, even if I am wrong, any such back-migration from Europe surely happened before Indoeuropean migrations.
"So IMO the only possible origin for this "European" element would be the (still hypothetical) bronze age IE migrations, carrying some R1a coming from central Asia.
Otherwise we would expect at least some y-DNA I or R1b1b2 (at the very least some very clear mtDNA traces)".
This logic is correct but does not necessarily explain the blue component, which only indicates some genetic affinity not ascendancy. What I think is that while part of the blue component surely reflects the shy IE migration and assimilation of locals among Pashtun and maybe Sindhi, that should not be the case among Burusho, who have obviously remained refractory to Indoeuropeization.
So part of the component reflects an older connection with Europe and I think it may even speak of when the peoples began moving out of South Asia in Westward direction, some 50 Ka years ago. The blue component shows up as a "memory" of an old shared ancestry in spite of many tens of millennia of separation. But well, I can see how this is not easily accepted.
Alternatively it's all nothing but a mirage and Burushos would become a totally different and homogeneous component just a few Ks deeper in the analysis. The Hazara tend do that very easily, not sure why the Burusho take longer.
Manju:
ReplyDelete"the way I see it all these autosomal data strongly correspond to mtDNA. So I believe these represent pre-civilizational identity (10,000 years?). I assume only after civilizational phase male specific migrations/invasions took place".
Fino-Ugric peoples are also largely Y-DNA N, a Siberian and ultimately East Asian lineage, almost non-existent in Europe out of their area. Yet the autosomal DNA is like 85% or 90% genuine European, and the mtDNA is like 95% or more Euro. These peple only reached a "civilizational" phase with the Swedish and Russian conquests a few centuries ago, yet the case is extremely similar to that of Austroasiatic tribals (and the word "tribal" and the word "civilization" sound like opposites in any case).
"Andrew commented at Razib's blog that Madagascar uniparental lineages 50-50 between Asian and African. However, autosomally they are 66% African and 33% Asian. If this is true, I believe we need to have a clear understanding on whether native genome has some advantages over non-native genome".
Or rather the opposite: why immigrant lineages had an advantage in remaining when the overall input was obviously larger from native ones? I'd say this has to do with Austronesian tendency to constitute an aristocracy above Africans and mixed groups.
"I suppose two Indian components shown in this figure are north-west Indian and central-east Indian. North Indian and South Indian is misleading".
Technicality but probably true. Are bengalis "ASI" or "ANI"? If Bengal and Bihar are "South" genetically, then you are absolutely right. But if they are "North" then you are wrong.
"If you remember some of the older studies, Irula, a South Indian tribe completely branches out in autosomal studies. I believe Irula represents true South Indian component".
A deeper layer maybe or a purer group. Hard to say: I remember the paper but only useless PCA graphs, no structure by K layers, where the substance usually is.
PC analysis fails when more than three distinct populations are involved.
"The above two points clearly support Proto-Dravidian homeland in central-east India. We do know all Dravidian language branches are found in this region. South India has only one branch (SD)".
This is a very interesting. Would you say that Bahui then arrived from Chattishgahr or Bengal? How and why?
Why can't we discern a genetic or archaeological model for that? Isn't it possible that Dravidian languages penetrated the subcontinent from the NW, maybe even Iran, in the Neolithic, when the possibility for the expansion of a homogeneous language (now family) was optimal?
"According to archaeology farming started in central-east Indian region which gradually spread to South India around 3500 years ago".
What about the much older Neolithic of Pakistan, specially Balochistan, where the oldest Neolithic of the subcontinent is found?
"I'm not sure whether original South Indians (represented by Irula) had different languages. The fact they don't much differ from other Dravidians in uniparental lineages I suppose that is unlikely".
Languages are not genes. People can be close in genes and have very different languages, people can also be close in language and have very different genes.
"Also, none of the isolated languages are found in South India also there is no study that talk about any substrate in Dravidian languages of the South".
Maybe nobody has yet even considered that possibility? It's not so easy to detect such stuff, much less be half-sure about it, and sometimes preconceptions act as pycho-cultural barriers, blocking the gorilla in the middle of the screen from being perceived at all.
About my non-smoking, I'm doing ok, thanks for your interest. Just feeling much more electrical and less focused, the previous days were much worse.
ReplyDelete"For me there is, why would they absorb IE blood in similar amounts as their IE-speaker neighbors and not change language as they did? Isn't more logical that Pashtuns and Sindhis changed language without absorbing much European blood."
ReplyDeleteOK but we can imagine anything, especially since it's a few mountain valleys (3 valleys). Anything.
Say for instance, some IE warriors take over the valleys and settle there. 100-150 yrs later the locals rebel and eventually the old language definitively consolidate its position in that society. Being (in this ancient time) far from the big political and population centers, with then limited interactions with the rest of the Indo-iranian populations, it sounds viable.
Kind of cheap literature story I guess, but my point is that, especially in such an isolated community, many things are still possible to explain it and that I wouldn't use it as a definitive factor to rule anything out.
"The Euro component is probably a Central Asian one, related to H and U in the area, which are certanly pre-IE stuff"
ReplyDeleteAFAIK, there is no such hgs mtDNA among the Burusho (as for the U if they exist they should be of the south Asian flavor). For these haplogroups to disappear through time they would have to be very few and they wouldn't leave such a trace (the dark blue component) among the Burushos. Which leave us with (some?) y-dna R1a1a and since "Europeanized" R1a1a in south Asia in epi- or upper paleolithic time are difficult to imagine, (and that we have no I or r1b1b2) only R1a1a from bronze age remains.
That was my logic in my former post but I didn't explain it well.
The reason why I always had trouble to side with your hypothesis that the ancient south Siberian R1a (Keyser et al. 2009) were more representative of a very ancient paleolithic population rather than representative of the chalcolithic IE movement from north of the black sea is that - besides the cutltural connection that could only be the product of a minor flow from east Ukraine influencing a older substrate - the ancient south Siberian mTDNA hgs didn't fit well with your hypothesized ancient paleolithic population european-like IMO.
The recent LBK paper only conforted me in my idea that it was a rather "recent" (roughly chalcolithic) phenomenon, mostly.
In the LBK study, the map showed that the component, beside its orignal home (Anatolia) had high concentration of this component in northern France and .... south and east Ukraine, exactly where we would expect the south Siberian R1a1a (ancestral to the Tocharian language?) to come from, within the Kurgan model. This could easily explain mtDNA haplogroups (and their frequency) such as HV, T, K, (but also from other studies, hgs such as U1 and U3 for instance), etc... among these south Siberians).
> rest is following
Somehow these east European R1a1a, mostly "Europoids" (as that's how the ancient populations of these regions are qualified) - in this the dark blue component continuum - mixed with maternal lineages from Asia minor during or before Chalcolithic (which BTW could easily explain some of the ancient loanwords from Anatolian/west Asian languages in proto-Indo-european).
ReplyDeleteMy point : the ancient south Siberian population is probably not really representative of paleolithical central Asian populations (only in a superfical way, if so) and the "Europoid" R1as in Asia (but the south Asian R1as are probably more ancient, in big parts at least - even though interestingly, from memeory, the R1a1a diversity in Finland, Serbia, Ukraine, Turkey and some parts of Russia was higher than in north Pakistan and east India (the highest diversity being in west India and sputh Pakistan IIRC) are better explained by chalcolithic and bronze age movements.
"also R1a maybe (but less clearly so, because R1a may be original from India)"
And that's exactly why I had to come up with y-DNA I and R1b1b2, because R1a probably arrived "late" in the east european area and it had to take some time for the process of mixing for populations high in y-DNA r1a1a originally from south Asia to become autosomally European, and since you insist on some upper paleolithical origin for the south Asia component, I and R1b1b2 would be the easiest way to explain a typical "European component" (AFAIK there are some y-DNA I and mtDNA U5 and H in Iran which in this case could easily be paleolithic and probably are).
"it may even speak of when the peoples began moving out of South Asia in Westward direction, some 50 Ka years ago. The blue component shows up as a "memory" of an old shared ancestry in spite of many tens of millennia of separation. But well, I can see how this is not easily accepted."
ReplyDeleteYes, I think it's dubious. 50 KYO .... hmmm....
I have difficulties to follow you on that one.
"A lot of people from that area look more or less European (but not on average), typically with a Nordic (Dutch-like) tendency when they do. This is probably more because they are ancestral to Europeans (specially NE European) populations than because they being heavily penetrated by them."
To me the time you're referring to are too old for such a proximity in genes and phenotypes. Difficult to accept for me, indeed.
But who knows.
Sorry for the annoyingly long post. Conciseness is probably not be my best quality.
"... only R1a1a from bronze age remains".
ReplyDeleteAt this moment deciding if R1a1 is East European, Central Asian or South Asian by origin, unless it belongs to a well defined subclade such as R1a1a7, is impossible.
In principle this is a case like J1, where most of the lineage hangs from a haplotype in North Africa and this one from haplotypes clearly West Asian. Replace J1 by R1a, North Africa by "Central Asia or Europe" and West Asia by South Asia and the case is the same. Africanists will try to imply that J1 is therefore "African" and Europeanists will try to imply that R1a is therefore European (or Central Asian).
But the reality is that the oldest part of the tree lays in West Asia and South Asia respectively, so only a subclade can be original from North Africa and Europe-or-Central-Asia respectively.
So we have to understand R1a1 in Europe as a derivative from South Asia, arriving in the Paleolithic, in a flow that is also apparent (though distinct) in its relatives R1b (in West Eurasia in general, and even Sahelian Africa) and Q (in West Eurasia, Central Asia, Siberia and even America).
"The reason why I always had trouble to side with your hypothesis that the ancient south Siberian R1a (Keyser et al. 2009) were more representative of a very ancient paleolithic population rather than representative of the chalcolithic IE movement from north of the black sea is that - besides the cutltural connection that could only be the product of a minor flow from east Ukraine"...
You probably mean Samara Valley in East Euro-Russia, not "East Ukraine" which is conquered territory like all former Dniepr-Don.
"... influencing a older substrate - the ancient south Siberian mTDNA hgs didn't fit well with your hypothesized ancient paleolithic population european-like IMO".
I do not understand the reason after reading the paragraph once and again, sorry.
My thesis is in fact that Central Asian population could be an ancient West-East hybrid, as reflected by shared autosomal components from the Urals to Pamyr, which must reflect an older homogeneity, which is often misread because of that classification of lineages between West or East.
Continue from above, Wagg, my mouse is playing tricks on me.
ReplyDelete"In the LBK study, the map showed that the component, beside its orignal home (Anatolia) had high concentration of this component in northern France and .... south and east Ukraine"
The paper could not justify those maps. I was also trickled at first but then had to add a warning in red colors because the map was meaningless on account of the data. Hate it really when people distort the data in order to create a map or otherwise conclusions that fit their preconceptions (but not the data). :(
"... where we would expect the south Siberian R1a1a (ancestral to the Tocharian language?) to come from, within the Kurgan model."
First of all we have to solve how R1a in Europe came from South Asia, where the oldest diversity seems to lay. And the obvious answer is through Central Asia, because R1a is otherwise low in numbers and diversity.
And we have to solve where proto-IEs (Samara culture) came from, because the archaeology of the area AFAIK never touched virgin soil, so it's an unsolved matter before c. 5500 BCE. The finding of U2 in the earliest inhabitants of the region suggests a South Asian and Central Asian connection overall. But the matter is far from fully clear, specially for the earliest IEs of Samara valley and their neighbors from the East: the Botai culture.
All this needs much greater research but by the moment I suspect at least partial origins in a Central Asian group with South Asian connections (R1a specially but not only).
"... such as U1 and U3 for instance"...
Neither U1 nor U3 are particularly "European". U1 is found in various places (from Egypt to India AFAIK but seldom Europe), U3 is mostly circum-Pontic, probably with a West Asian origin. Correct me if I'm wrong.
And anyhow, East Ukraine has so much to do with PIEs as Madagascar, so to say: the genuine PIE homeland is further East at Samara district, East of the Volga, which is the only natural barrier of some size in the area (and until Peter the Great it was where Asia began).
"mostly "Europoids""
Europoid is synonym of Caucasoid: West Eurasian, not Mongoloid. It does not mean a European origin automatically: it can have a West or Central or even South Asian one as well.
"And that's exactly why I had to come up with y-DNA I and R1b1b2, because R1a probably arrived "late" in the east european area"...
It is a good idea but do we find R1b in Pakistan or India? AFAIK not except among the Lambadi, which are not relevant.
...
"and since you insist on some upper paleolithical origin for the south Asia component, I and R1b1b2 would be the easiest way to explain a typical "European component""
ReplyDeleteExactly the opposite. My point is that the "European component" is nothing but a diffuse indicator of some shared ancestry at the level when it also compares with West Asia and South Asia. A multi-component pop. is not automatically a sign of admixture but just of intermediate position, of affinity. Often "multilayered" populations end up having their own distinct homogeneity at deeper Ks. Only in some cases such a recently admixed groups this component=ancestral population identity stands, in all the other cases it's just a vague indicator of affinity in comparison with the rest.
Pakistan pops. cannot be described only as South Asian, because they are also more directly related, since some 50 Ka ago, to West Eurasians, because these spawned from that area surely. This must show up in the structure analysis somehow and does as European affinity band.
I do not mean a migration from Europe in the Paleolithic (though in the case of West Iran it happened for sure, this is a different region altogether, separated by vast arid provinces), what I mean is the origin of West Eurasians as such, which must have happened in Afghanistan, Pakistan and NW India.
"Yes, I think it's dubious. 50 KYO .... hmmm.... I have difficulties to follow you on that one".
Yes you do: what I say is that there should be a trace indicating affinity in NW South Asia where West Eurasian populations must have originated somehow. After all it was a small group or a few of those which colonized all the West of the continent, much like happened in America at a later date (notwithstanding the lesser inputs from Africa and NE Asia). This does not need to be sign of back-migration, just of memory of those origins.
"To me the time you're referring to are too old for such a proximity in genes and phenotypes. Difficult to accept for me, indeed".
I understand it's difficult to accept, even for myself. But there must be a "genetic memory" of the specific population from which the colonists of West Eurasia sprung some 50 Ka ago. This should not have been totally erased, right? Genes do not die easily: they just evolve.
There was probably a founder effect A towards Afghanistan and then, from this A population, either a B (West Eurasian in general) or A1, A2, etc. (several WEA subgroups) spawned. A would then have got a lot of influence from South Asia overall, so they do not keep this distinction anymore (except the Kalash???) while B or A1, A2, etc. did not get almost any other influence from SA, so they still look neatly different, actually even more, as the in-group evolution has enhanced their distinctiveness.
"Sorry for the annoyingly long post. Conciseness is probably not be my best quality".
It's OK. I do the same all the time. No big deal.
@ maju :
ReplyDelete"So we have to understand R1a1 in Europe as a derivative from South Asia, arriving in the Paleolithic"
I never said otherwise. I just implied that it doesn't seem to be very ancient in east of Europe (-> the proposed date for the apparition of R1a1a) and that paleolithic europoid individuals would have probably to be I or R1b1b2 and thus, as they are not found in significative presence, this leaves us with R1a1a and so with more recent date for this "European" dark blue component arrival (The theory of the ancestral west eurasian component reflected in the dark blue color is difficult to accept to me. See below).
"It is a good idea but do we find R1b in Pakistan or India? "
I wasn't really considering it possible (actually R1b is found in north Pakistan, Afghanistan and Nepal in low quantity (though among the Hazaras it's (mostly?) R1b1b1 IIRC and it probably arrived there from central Asia I guess with Turkic or Mongol movements)) I was just setting the background to show that only (at least some) R1a1a could fit with the dark blue component.
"I understand it's difficult to accept, even for myself. But there must be a "genetic memory" of the specific population from which the colonists of West Eurasia sprung some 50 Ka ago"
The peoples would have diverged so much in the populations of west/north south Asia (before the peopling of the west Eurasia) - making them quite separate from other south Asians and spawning a specific group at the orgin of the Europeans - but their affinities wouldn't have changed much for the 50k-35k time period of separation and would be still visible today? Hard to accept for me indeed.
Beside, shouldn't the European be mostly derived from the light blue component? Shouldn't the light blue component be also much more present in the west of south Asia as the west asian populations also originate from the west of south Asia if so much tracks of ancestral west eurasian populations had to remain visible somehow in modern south Asian populations?
"The finding of U2 in the earliest inhabitants of the region suggests a South Asian and Central Asian connection overall"
Indeed but in this case we're talking of something very ancient. Not sure it can be used as an indication concerning the presence of the dark blue in north/west south Asia. As you say it's not clear.
"You probably mean Samara Valley in East Euro-Russia, not "East Ukraine" which is conquered territory like all former Dniepr-Don. "
ReplyDeleteSamara is seen as the ancestral region of this culture but that doesn't mean that the proto-indo-european language would appear there (proto-IE is a specific moment of the evolution of the language).
For a pre-yamna/yamna flow, this is a good place to consider, I think (for the ancestral speakers of proto-tocharian (once again it fits w/ the archeological data as well, AFAIK) or proto-indo-european language).
"I do not understand the reason after reading the paragraph once and again, sorry."
I think haplogroups such as T, K and a few others (all very present in the mtDNA lineages) doesn't fit well with paleolithic south siberian/north central Asian populations and better with a later flow, and coming from the west. The maps from the LBK study fit very well with this pattern of an IE east Ukrainian flow. I remember that you hinted at the dubious nature of these maps, I just have trouble accepting that specialists could do such a mistake. It has to fit with some of their data (some missing or non-published data?).
The fact is that the matches of the mtDNA hgs of south Siberia (in Kayser et al 2009) in modern peoples (including the ones often seen as neolithic and originally more clearly west Asian mtDNA hgs) were generally found in Europe.
As for U3 and U1 I didn't mean they were "european", I was more thinking of the Caucasus (and as such, the region near it, such as Ukraine and south russia). I just think they were part of the lot that went eastwards in chalcolithic and not before (the west Eurasian paleolithic samples generally always point to the same mtDNA hgs).
Anyway it's basically useless to try to unravel all this (except for the pleasure to solve a riddle), it's too soon to know. We won't reach the solution by ourselves and the answer will come from other persons than us :)
Have a nice day.
Maybe I am totally wrong about this component and is after all a IE marker. But I do consider other options.
ReplyDeleteAnyhow:
"Samara is seen as the ancestral region of this culture but that doesn't mean that the proto-indo-european language would appear there (proto-IE is a specific moment of the evolution of the language)".
I understand that PIE is what was spoken in Samara valley between c. 5500 (or maybe earlier) and c. 3500 BCE, when IEs began expanding and new creating creole dialects all around. What was spoken in Ukraine after the IE expansion began (i.e. in the Seredny-Stog II phase) is a derivate and not anymore just PIE but one or more branches under the root node.
So I do not understand why your insistence on Ukraine of all places, when it's obviously a "province" or corridor, not the core area.
It is not really possible that Tocharian or Anatolian spread from the Ukranian or Don basin extension because they are clearly older than the node producing the European dialects. When IEs were invading the Dniepr-Don area, they were also in Altai already, as well as in Azerbaijan: the three expansions are roughly contemporary and therefore one does not originate the other.
"I think haplogroups such as T, K and a few others (all very present in the mtDNA lineages) doesn't fit well with paleolithic south siberian/north central Asian populations and better with a later flow, and coming from the west".
At this moment I am of the opinion that mtDNA JT, K and W spread from West Asia in various directions with the Neolithic. There was also a West Asia to Central Asia and Pakistan Neolithic flow. It probably affected most strongly the area of Uzbekistan, as is the most favorable for agriculture in all Central Asia and can support a more stable population, which should resist the nomads' migrations to a great extent till present day.
They may have "hitchhiked" the IE migrations somewhat but they do not properly belong to them.
"The maps from the LBK study fit very well with this pattern of an IE east Ukrainian flow. I remember that you hinted at the dubious nature of these maps, I just have trouble accepting that specialists could do such a mistake. It has to fit with some of their data (some missing or non-published data?)".
I need context (a link pefereably) to be sure of what I meant and if I can sustain that opinion or must change my mind. Probably I emphasized that the tiny Elbe group was hyper-oversampled, distorting surely the understanding of all the matter. Also there are issues because Haak et al. have only used HVS-I, which is a quite confusing.
"The fact is that the matches of the mtDNA hgs of south Siberia (in Kayser et al 2009) in modern peoples (including the ones often seen as neolithic and originally more clearly west Asian mtDNA hgs) were generally found in Europe".
But this may be two different things:
1. An old connection.
2. A new IE input.
AFAIK the whole Upper Paleolithic of the area is of Western Eurasian affinties, including early Aurignacoid and later Gravettian offshoots (Ma'alta), so the logical thing is to find West Eurasian and even East European lineages. It's just normal even without IEs.
And yes, we cannot understand always all the facets. Enjoy.
"dizzy from the pains of quitting up heavy smoking"
ReplyDeleteHang in there. My father-in-law did the same a couple of decades ago and his quality of life improved dramatically as a result. It really does pay off in the not nearly as long as you'd think run.
Dravidian as a widespread subcontinental language is probably less than 4500-5500 years old, and there is good reason to think, given what we know of the time line of Austro-Asiatic expansion together with South Chinese agricultural developments to think that its presence in India may be only a thousand or two years earlier than Dravidian. The lack of population strong genetic structure between Indian Dravidans and Austro-Asiatic populations is consistent with this scenario, so long as the language shift did not involve any proportionately strong outside demic components.
ReplyDeleteThis suggests that the original ASI languages of India are probably lost or survive only in a handful of tiny linguistic communities whose link to the past can't be coroborrated.
The ANI-ASI divide probably pre-dates Indo-European, Dravidian and Austro-Asiatic languages in India, with modest levels of admixture within the subcontinent between the two groups. My intuition is to think of ANI as congruent with Harappan society and its predecessors (probably with a significant 7,000 year old Mesopotamian-Persian Neolithic founder population component), and to think of ASI as autochthonous to the rest of India, or at least, as having roots deeper than anyone in India other than some tribal populations.
It is hard to know if the Khasi-Aslain AA speakers are remnants of the original Munda wave or a subsequent one without more background that I have on this population.
* * *
Your point on ANI Green in Indo-European but not Basque populations in Europe, is a good one. Its presence in the Near East and lack of pink North Asian components in the Near East and Europe, makes a metal age IE expansion seem an unlikely source of ANI Green in Europe and the Near East. If the Green in the Near East had come from admixure with European IE it would have had more dark blue relative to it in the Near Eastern mix.
It would make more sense for green in Europe and the Near East to be due to Neolithic expansion from the Near East and represent Persian introgression into that Neolithic founder population. This would suggest that Harappan might very well have been akin to Elamite genetically and linguistically, but that Dravidian was not - which seems increasingly likely IMHO.
In this view the green in the Indo-Aryan mix is Harappan substrate, and the blue the West Eurasian PIE speaking substate, with Brahuis is not genetically Dravidian, only lingustically due to language shift.
"It really does pay off in the not nearly as long as you'd think run".
ReplyDeleteI'd like to have access to cheap inhalable nicotine without cigarettes but specially cigarettes without nicotine for the joints. I smoke tobacco only for the joints, I only began smoking cigarettes when I was 18 or so and I have quitted twice before but I always come back because I like my joints now and then.
So what I'd need is something I could smoke with 0% nicotine, so it doesn't become addictive.
Alternatively it'd be nice to get my fixes of nicotine without all the tar of cigarettes (and cheaply because it's tyrannical and quasi-racist that we smokers are almost single-handedly sustaining the burden of the state that represses our habit). But these options, which are the smart ones are not really available because of medical totalitarianism.
"The above two points clearly support Proto-Dravidian homeland in central-east India. We do know all Dravidian language branches are found in this region. South India has only one branch (SD)".
ReplyDeleteThis is a very interesting. Would you say that Bahui then arrived from Chattishgahr or Bengal? How and why?
Why can't we discern a genetic or archaeological model for that? Isn't it possible that Dravidian languages penetrated the subcontinent from the NW, maybe even Iran, in the Neolithic, when the possibility for the expansion of a homogeneous language (now family) was optimal?
"According to archaeology farming started in central-east Indian region which gradually spread to South India around 3500 years ago".
What about the much older Neolithic of Pakistan, specially Balochistan, where the oldest Neolithic of the subcontinent is found?
Witzel's paper on the lack of a Dravidian substrate in early Indo-Aryan, and the fact that the crops used in the South Indian Neolithic were different from the crops used in the much older IE Neolithic with indications of trade relations between the two only at a Harappan fringe trading post in SW Indian, as well as an absence of cultural carryovers between the two all point to an non-ANI source for Dravidian.
The distribution of Y-DNA hg T in India does show congruence with the proto-Dravidian region suggesting a possible source population for language subsequently carried by ASI natives, and my bet remains with a small group of men with a substantial Y-DNA hg T component as the source of both Dravidian and the South Asian Neolithic crop set which is mostly Sahel African domesticates.
Y-DNA hg T has one of the most odd geographic distributions of any of the Y-DNA hgs. Sister Y-DNA hg L in contrast corresponds geographically to ANI.
The widespread distribution of Y-DNA hg T in Europe (even Iberia) and the Fertile Cresent screams old Neolithic or Epi-Paleolithic. It isn't even remotely a good match for Y-DNA hgs I or J or R1a or any of the MNOPS hgs (other than R1b). Y-DNA hgs geographically so probably wasn't part of the same mix as any of them. It also doesn't have the widely distributed and mixed in pan-Indian distribution of Y-DNA hgs H and R2 taht seem to be South Asian natives. It looks more like Y-DNA hg G or R1b but still isn't too similar in proportionate distribution to either. The geographically and sub-ethnically concentrated distribution of Y-DNA hg T in parts of Africa and India suggests recent arrival in each place compared to Fertile Cresent and European ages - this hasn't had time to blend into the background.
"So what I'd need is something I could smoke with 0% nicotine, so it doesn't become addictive."
ReplyDeleteTime for a road trip to Amsterdam. ;)
"The ANI-ASI divide probably pre-dates Indo-European, Dravidian and Austro-Asiatic languages in India, with modest levels of admixture within the subcontinent between the two groups".
ReplyDeleteProbably.
"My intuition is to think of ANI as congruent with Harappan society and its predecessors (probably with a significant 7,000 year old Mesopotamian-Persian Neolithic founder population component), and to think of ASI as autochthonous to the rest of India, or at least, as having roots deeper than anyone in India other than some tribal populations".
Instead I think that the divide is pre-IVC, it is probably something that happened in the Paleolithic, as some groups took the Coastal-Krishna River route and others took the Narmada-Son one, converging maybe at Bengal.
I see no particular reason of IVC alone being able to create such a genetic divide, specially when it's obvious that the ANI component is as autochtonous as the ASI one. It's more like the internal divides we see in other subcontinental regions: South and North West Asians, West Asians and Europeans, or the ones also obvious in East Asia.
A common characteristic is that these subdivisions within major regions always overlap, they are clinal and never clear cut. And they usually cluster together when compared with outgroups.
"It would make more sense for green in Europe and the Near East to be due to Neolithic expansion from the Near East and represent Persian introgression into that Neolithic founder population".
IDK, we do not have a "Persian" reference in that sample and we know that there was East European Epipaleolithic flow to the Zagros area via the Caucasus instead.
Still Basques show the light blue component, even if in reduced amounts, and this one can be associated to Y-DNA J2, which is the only Neolithic marker with a West Asian origin showing at some levels among Basques.
Anyhow, look at the Adigey (high in green, the highest of West Eurasia by far!) Maybe it's more an Anatolian kind of link (as Caucasus peoples are usually closest to Anatolians). Maybe the green cluster is associated to G2a, which is important among some populations (but not Basques) and would have a more Caucasian (proxy of Anatolian) kind of connection?
"In this view the green in the Indo-Aryan mix is Harappan substrate, and the blue the West Eurasian PIE speaking substate"...
I have already discussed a lot with Wagg about this so I won't bother repeating. Just that I don't think this is a correct interpretation (or most probably not).
Amsterdam is an expensive urban myth. We do not need to go to Amsterdam, if anything we travel to Morocco... but as I say no need. I can just get out to the street and ask or grow my own or... and most of it is legal or so barely illegal that does not matter.
ReplyDeleteThe problem are the effects it has in driving me to addiction to tobacco (nicotine) once and again. I'm not really interested in nicotine because it barely gets you any high of any sort: just addicted. It's a stupid drug, even more stupid than alcohol.
"The problem are the effects it has in driving me to addiction to tobacco (nicotine) once and again".
ReplyDeleteDo you mix your marijuana with tobacco? Why?
From the conclusion of the article:
ReplyDelete"The analysis of autosomal data suggests bidirectional gene flow across the Bay of Bengal restricted to Austroasiatic-speaking and Tibeto-Burman–speaking populations. The presence of a significant (approximately one-quarter) southeast Asian genetic component among Indian Munda speakers is consistent with this model, implying their recent dispersal from southeast Asia followed by extensive admixture with local Indian populations".
I remember you arguing with me that Austro-Asiatic was not the product of a movement east from SE Asia. And do you still maintain that there was complete separation between East/SE Asian M haplogroups and South Asian versions with no M haplogroups between the two? More from the conclusion:
"The strongest signal of southeast Asian genetic ancestry among Indian Austroasiatic speakers is maintained in their Y chromosomes, with approximately two-thirds falling into haplogroup O2a".
Again you were very sure not so long ago that Y-hap O2 was not associated with Austro-Asiatic-speaking people in India. And certainly did not come from anywhere in SE Asia. Yet the article claims:
"Geographic patterns of genetic diversity of this haplogroup are consistent with its origin in southeast Asia approximately 20 KYA, followed by more recent dispersal(s) to India".
"Most importantly it comes up clear that Indian Austroasiatics are, in spite of often obvious East Asian male-mediated ancestry, mostly South Asian by overall ancestry"
It's actually more complicated than that. From the conclusion:
"Comparison of mtDNA and Y chromosome data reveals that the 'import of local genes,' at least in case of the Munda speakers of India, has likely been biased toward the female sex, resulting in a situation where the southeast Asian ancestry signal in the mtDNA lineages of Indian Munda speakers has been entirely lost".
"Indian Tibeto-Burman peoples cluster with East Asian ones but not Austroasiatic Indians"
Presumably because they each entered India from a different region.
"In sharp contrast, among the geographically proximate Khasi-Aslian–speaking Khasi population, approximately one-third of the mtDNA lineages have southeast Asian ancestry"
Well at least you're now aware of the fact that there are such people as 'Khasi'. And they're from exactly where I said they were from, the borderland.
Razib made a comment at his blog that South Asian M is basal to East Asian M. He made a reference to Stephen Oppenheimer's 'The Real Eve'. I don't have the book. However, I haven't come across any studies in this regard. I suppose if South Asian M is basal to East Asian M, then that settles the issue.
ReplyDeleteNot marihuana, hashish. Traditionally here marihuana was relatively hard to find because Moroccans (and other more mythical sources like Lebanon or Afghanistan) produced hashish instead. Hashish is the pressed resine of the plants' flowers and is usually smoked mixed with something else, usually tobacco.
ReplyDeleteHowever quality hash may be smoked directly but with some art on the part of the user and some special pipes. Your usual "macaco" street mix is seldom good enough for that and anyhow it's simpler to mix with a cigarette in your usual joint.
"I remember you arguing with me that Austro-Asiatic was not the product of a movement east from SE Asia.... Again you were very sure not so long ago that Y-hap O2 was not associated with Austro-Asiatic-speaking people in India. And certainly did not come from anywhere in SE Asia".
ReplyDeleteWhy do you LIE? Either show me the source of this LIE or apologize. I am aware that AA tribals came from SEA since I know of their existence.
"Presumably because they each entered India from a different region".
More like in different times and in different processes. AAs are clearly Neolithic while TBs may be of more recent origin and related to the general flow southwards of TB language/culture, which is of Metal Ages maybe.
"... South Asian M is basal to East Asian M".
ReplyDeleteThere must be a misunderstanding here or some sort of elaboration that I cannot understand from these sentences alone.
The only M ancestral to all 43 M basal sublineages is M-root. Some of these lineages are South Asian, others are East Asian and a few are from other areas. My whole reasoning of M having coalesced in SA derives from most basal lineages existing there. That's what Terry is challenging but he's cheating, bending and twisting so much that I cannot follow his thought anymore without deep suspicion.
Give me a couple of months, maybe just weeks and I'll probably produce something more clear.
Cheers.
"Why do you LIE? Either show me the source of this LIE or apologize. I am aware that AA tribals came from SEA since I know of their existence".
ReplyDeleteI'm referring to when we were arguing over the Austronesian expansion and its associated haplogroups where you denied that Y-hap O2a was associated with the expansion of the Austro-Asiatics into India. I'm sure you'll be able to find it if you care to search.
"Some of these lineages are South Asian, others are East Asian and a few are from other areas".
Especially from those areas between Northeastern India and East Asia.
"My whole reasoning of M having coalesced in SA derives from most basal lineages existing there. That's what Terry is challenging but he's cheating, bending and twisting so much that I cannot follow his thought anymore without deep suspicion".
I'm certainly not claiming a coalescence for M anywhere other than South Asia. What I am claiming is that the distribution of M-derived haplogroups argues strongly against a coastal route east from South Asia into East Asia and Southeast Asia. The evidence supports a migration over land through the region of Assam, Northern Burma and Yunan, and a spread from that region. The only 'coastal' element in the expansion is that required to reach Australia/Melanesia/New Guinea and the one (probably later) needed to reach the Andamans.
"Give me a couple of months, maybe just weeks and I'll probably produce something more clear".
I am very much looking forward to it.
"I'm sure you'll be able to find it if you care to search".
ReplyDeleteI'm sure I won't find it. I know what I think about all that stuff. You probably misunderstood me and then further distorted what I meant.
So quote, please.
"What I am claiming is that the distribution of M-derived haplogroups argues strongly against a coastal route east from South Asia into East Asia and Southeast Asia. The evidence supports a migration over land through the region of Assam, Northern Burma and Yunan, and a spread from that region".
Coastal route or Assam route is a trivial difference. I have always sustained that "coastal route" just means "tropical route", as opposed to migration via Altai. The lesser details (probably all routes at the same time) are not that important nor can be inferred from the haplogroups easily at all.
"I am very much looking forward to it".
Don't get anxious. It will come in due time. Good things need time.
My whole reasoning of M having coalesced in SA derives from most basal lineages existing there.
ReplyDeleteWhat does this mean? Or what do you mean by basal lineages?
As far as I remember almost all studies have found coalescence age of M is greater in East Asia than in South Asia. What is you mean by 'coalesced' here?
Basal sublineages are those that hang directly from the ancestral node, M in this case (which has 43 basal sublineages per my latest count). M8 is a basal sublineage of M but C is not (it's twice derived).
ReplyDelete"As far as I remember almost all studies have found coalescence age of M is greater in East Asia than in South Asia".
"Most" not. One in particular. However Oppenheimer, co-author of that paper, later seems to have changed his mind and now seems to support an older age in SA. His latest chart is used by Petraglia in that sense (link).
"What is you mean by 'coalesced' here?"
Where it became M from pre-M and/or where it first became post-M from M. Specially the later, as we never know when a particular mutational state exists until it begins producing "children".
His latest chart is used by Petraglia in that sense (link).
ReplyDeleteWhich figure you are looking at? At least from figure 4, it appears Australian/Papuan M is the oldest followed by Chinese and Indian. This in line with the other studies.
Nah, my bad. I got the wrong idea or something.
ReplyDeleteWhatever the case when you say "most papers" you mean Salas 2009, which is "unsmokable", to transliterate a Spanish slang expression (also "unswallowable", same meaning).
The authors average the data regionally instead of properly addressing each haplogroup individually.
A quick review at PhyloTree shows that the following basal lineages are detached from M-root by a single (coding region) mutation:
·M1'51 (SA, WEA, Africa)
·M3a (SA)
·M3c (SA)
·M4"64 (SA, the "R" of M)
·M5 (SA)
·M9 (EA, specially SEA)
·M12'G (EA: SEA, NE India, G only more widespread)
·M13'46'61: SEA and SA
·M23'77 (no idea right now)
·M25 (no idea)
·M29'Q (PNG)
·M32'56 (SA and Andaman)
·M33 (SA)
·M34'57 (SA)
·M35 (SA)
·M40'62 (SA)
·M42'74 (Oceania?)
·M49 (no idea)
So of all basal subclades that evolved from M in short notice (and assuming all "no idea" are Oriental), we have:
SA-WEA: 10 clades
EA-Oce: 6 clades (max.)
Intermediate: 1 clade
So it does look like M short-stem sublineages are concentrated in South Asia, with at least 62% of all (almost 2/3).
This clearly indicates that SA's M is as old as that from anywhere else. And that most of the oldest M sublineages are in South Asia.
"Coastal route or Assam route is a trivial difference. I have always sustained that 'coastal route' just means 'tropical route', as opposed to migration via Altai".
ReplyDeleteTo most people 'coastal' means somewhere near a coast of some sort. How come you're so different?
I've just noticed something else interesting about M's distribution, and have had a reshuffle to see if any M haplogroups may have coalesced in South India: Andhra Pradesh, Karnataka, Kerala, and Tamil Nadu. There are a few possibilities, so I've altered my list accordingly. For a start 'South and East India' becomes 'South India' and haplogroup M41 moves out, into Central India. That's one haplogroup eliminated from the new category.
But M35 enters from Central India. According to Chandraseker M35 is one of the few haplogroups in the Karnataka tribal group, the Betta Kuruba. M35 is the first new entrant.
The most interesting new entrant, though, is M2, from West India. If we confine our view to just the Indian tribals there it is. At its peak also in the Betta Kuruba, at 64%. It's very easy to envisage a spread of M2 from the south. Wikipedia has its highest concentration in Bangla Desh but its distribution is almost pan-Indian.
So that leaves just two M haplogroups (M4''64 and M5) in Northwest India, the supposed region of M's coalescence. And, according to Wiki, M5a is centred on Orissa. So M5 may not be Northwest Indian after all. As for M4''64, I'm only giving a Northwest origin for it the benefit of the doubt. M4''64 could have spread from anywhere in India, except for the south.
Back to the south. Chandraseker places M54 as one of the main haplogroups in the Jenu Kuruba of Karnataka, 29%. It is really M3c1b. But any, or perhaps all, of the South Indian haplogroups could have entered from the north, by land. There goes the last possible evidence for a great southern coastal rapid migration. M's expansion was slow and painful: from Northeast India into the mountains and valleys of East and Southeast Asia. Eventually various Ms managed to cross the water to the Andamans and several other islands in SE Asia, including New Guinea, Melanesia and Australia. But I'm sure that Haplogroup N had beaten M to at least the last island. I suppose we could claim that N coalesced in Australia. Technically N's greatest basal diversity is on that island.
"To most people 'coastal' means somewhere near a coast of some sort. How come you're so different?"
ReplyDeleteFor me it's always been a matter of the South Asian route vs. the Siberian route. Siberia is farther from the coast. But I also know from the beginning that the "coastal route" was at least partly riverine, "inland". It does not change a thing.
"If we confine our view to just the Indian tribals there it is".
We cannot do that.
Anyhow, the comment format alone is not good enough to discuss this complexity. I'd suggest you start your own blog and/or join the wiki when it's ready (soon - TM).