Monday, July 26, 2010

Origins of Indians: Version 6.3

Semitic Introgression Theory:
My model of the Indo-Aryan caste society envisages a situation where the IE languages were brought in by the nomadic bands of R1a1 people to north-western India from Central Asia/Eastern Europe migrating in search of the greener pastures and who were superimposed by Aryan priestly families belonging to J2a and G2 lineages from West Asia. Instead of Semitic introgression theory, I could have called it Aryan Intrgoression Theory (AIT) for priestly family migrations and Steppe migrations for IE migrations. But I'll stick with Semitic Introgression Theory for the overall development of the caste Indian society.

As I argued before, the priesthood most likely was created in West Asia with its earliest literate civilizations. This is the reason we find R1a1 across the castes and erstwhile outcasts in North India but J2a is more restricted to Brahmins.

However, I also commented that this model need not be completely true in the case of the South India as there could be earlier J2a migrations from West Asia which could be seen in tribes like Toda.

Curiously, it appears we have another earlier migration from West Asia to North India too. A village in Himachala Pradesh called Malana is where you find this isolated community. According to a recent paper by Rajiv Giroti and Indu Talwar (2010), nearly 60% of this population belong to J2a. The community speak a Tibeto-Burman language (probably, we can get better idea if their matrilineages are tested).

But the interesting part is Wikipedia article on them;
People in Malana consider all non-Malani to be inferior and consequently untouchable. Visitors to Malana town must pay particular attention to stick to the prescribed paths and not to touch any of the walls, houses or people there. If this does occur, visitors are expected to pay a forfeit sum, that will cover the sacrificial slaughter of a lamb in order purify the object that has been made impure. Malani people may touch impure people or houses as long as they follow the prescribed purification ritual before they enter their house or before they eat. Malanis may never accept food cooked by a non-Malani person, unless they are out of the valley (in which case their Devi can't see them). Malanis may offer visitors food but all utensils will have to undergo a strict purification ritual before they can be used again.
I have already argued that purity-pollution rules were ritualized in West Asia and the core Brahmin Varna belonged to that society. Now, we have a population in India which though not part of the Varna system has all the ritualized purity system and patrilineally closer to West Asia ((Iran) in its orientation.

One more proof that, the caste system was basically a creation of West Asian priestly class and not Shamanic IE people from Central Asia/Eastern Europe.

Reference:

The Most Ancient Democracy in the World is a Genetic Isolate: An Autosomal and Y Chromosome Study of the Hermit Village Malana (Himachal Pradesh, India)
RAJIV GIROTI AND INDU TALWAR

9 comments:

bachodi said...

What are R1a1 and J2a etc ? different races or tribes ?

Maju said...

@Bachodi: R1a, J2a and others are Y-chromosome DNA (genetic) lineages, also called haplogroups. They are real purely paternal lineages to remote ancestors. They are important in understanding the origins peoples, though the interpretation is not always straightforward.

R1a is specially important in northern South Asia, some parts of Central Asia and all of Eastern Europe, it's often speculated to be related to Indoeuropeans but recent research (not yet fully confirmed) placed its overall origins in India, what surely implies an older time-frame (Neolithic or Paleolithic).

J2a is one of several J sublineages, which are dominant in West Asia and the Caucasus. Unlike J2b that is more evenly spread in India, it seems that J2a has a NW and Brahmin distribution mostly.

However, and this goes to Manjunath:

The typical Semitic lineage, if any, is J1, not J2a. Both J2 have a "highlander" or "northern" distribution in West Asia and are hence probably pre-Semitic.

It is possible that J2a does not represent a "Semitic" but other type of West Asian pre-Semitic flow (Elamite?, Sumerian? - even maybe Indo-Aryan, what's the apportion of J2a in Afghanistan? Nobody knows, AFAIK).

manju said...

Bachodi:
No, nothing to do with race (which I suppose is a mere convenient identifier at its least harmful self). Tribes, maybe, with multiple identities though belonging to the same paternal lineage.

Based on their distribution, we identify them belonging to either West Asian (J2a) or Central Asian/East European (R1a1). Along with it, we assign predominant cultural and linguistic identities of those regions to these lineages.

manju said...

J2 have a "highlander" or "northern" distribution in West Asia and are hence probably pre-Semitic.

Yes, I do understand that it's pre-Semitic. I'm more inclined to believe Sumerian (and probably part of Caucasian language families just like pre-Semitic J1).

Elam as far as I understand did not develop anything original and just copied their neighbours.

There are more Pashtuns in Pakistan than in Afghanistan. Probably, Pakistani Pathan population structure gives better idea (predominantly R1a1 around 40-50%).

Maju said...

"Elam as far as I understand did not develop anything original and just copied their neighbours".

Have you heard of Jiroft Culture? They made truly original stuff. Of course they were also influence by Sumer: everybody was, even Egypt probably. But not everything has to be strictly/straightforwardly "Sumerian".

"There are more Pashtuns in Pakistan than in Afghanistan. Probably, Pakistani Pathan population structure gives better idea (predominantly R1a1 around 40-50%)".

Fair enough but what about J2a?

manju said...

Fair enough but what about J2a?
Didn't find any J2a in Sengupta et al.(2005/6) study (a pioneer study in J2a). But Qamar et al. (2002) study showed around 7% (6/93) belonging to J2 (Haplogroup 9). So that should be divided between J2a and J2b. So, I guess nothing to do with early West Asian society.

Maju said...

Well, anyhow... I'm looking for more references on the matter and, while I can't find right now anything on India subcontinent, I did stumble upon Abu-Amero'09, where we see that:

Iraq, ancient Mesopotamia, includes 27% of unspecified J2 (little of which is J2b per Semino 2004). So, maybe as much as 25% J2a. But it also includes 30% J1, which is AFAIK very rare in India, plus some 9% of E1b1b1 (again virtually non-existent in South Asia). Furthermore, considering the overall distribution of J1 and J2 in West Asia, where J1 has a southern tendency and J2 a northern/highlands one, it's very likely that ancient Sumerians had (and modern South Iraqis have) more J1 and less J2 than the Iraqi average. A good control could be Kuwait but I see no data for the little realm in this paper.

Iran displays 20% J2a and 11% J1, which is more in line with India but not yet. J2b is quite rare (3%), as is E1b1b1. Instead they have quite a lot of R1 (R1a, R1b) and G (G1, G2), in line with Anatolia-Caucasus.

Pakistan has some 10% J2a and 3% J2b. This is important because almost without doubt that's where J2a arrived to South Asia.

So we do have a cline Iraq-Iran-Pakistan in the line of maybe 25%-20%-10% J2a. How does this look "Sumerian"? Beats me. It does look West Asian but I don't see why it should be specifically Sumerian.

Was there a J2a founder effect at the origin of the Brahmin castes in IVC or later in Vedic India? Probably. Did it come from Sumer specifically? You'd have to prove via haplotype structure. Other reasonable options are from Iran or even a local founder effect in Pakistan/NW India, where J2a is relatively common in general. Or even a mix of some of these scenarios.

Anyhow, for me the arrival of J2a to South Asia stinks to Neolithic founder effect or mere gradual flow from Iran with clinal pattern.

manju said...

J1 almost doesn't exist in India. J1 is either part of Afro-Asiatic languages or Caucasian languages. The situation is complicated in West Asia as such. However, that is made clearer by J2a distribution in India.

I won't deny J2a Neolithic presence in India. I have already told that Dravidian tribes and now even a North Indian isolated community have J2a.

We get clear picture if consider J2a structure in North Indian castes. Why isn't it common among the so-called lower castes and IE tribals?

We can untangle the things based on these features. We also need to consider J2a distribution along with G2 (again generally observed among Brahmins). This again gives clue for West Asian distribution pattern that as you said is more northern West Asian.

What we really want to do is to check isolated and/or endogamous communities and base our observations on them. West Asia indeed has multiple migrations and because the region is small and not densely population that would change the demography considerably. But this is not the case with South Asia.

Maju said...

"We get clear picture if consider J2a structure in North Indian castes. Why isn't it common among the so-called lower castes and IE tribals?"

In truth I don't know because at this moment I lack the specific reference to consider the caste distribution detail. What paper(s) are you taking as reference? Whatever the case 10% of Pakistanis is much higher than it'd be expected by mere presence in upper castes, which, as we know, are just a small minority (and that must also have been the case in the past).

As the caste system almost without doubt arose in NW South Asia, I'd imagine that those haplogroups common in the area such as J2a or R1a had better opportunities to be co-opted to these upper layers of society.

However there are other haplogroups also common in Pakistan (the core IVC area) such as L (13%), C3 (7%), R1b (7%), R2 (7%), G2 (6%) and H1 (6%) which (with the exception of G2 that you just mentioned) don't see you mention in your analysis of Upper Caste Y-DNA (for good reason, I presume). The question is for me: are these lineages present in Hindu upper castes at comparable levels to J2a or not? And, if not, whether there is an apparent reason for it, such as being concentrated in some peculiar ethnicity or region (C3?), having a presumably later time of arrival (R1b?) or maybe suspected backflows from India proper (H1? R2?) What do you think?

"West Asia indeed has multiple migrations and because the region is small and not densely population that would change the demography considerably".

I don't think that is a correct assumption. West Asia after all is the oldest Neolithic region of the World, so it must have got comparably high densities very early on, making it relatively immune to major post-Neolithic demographic changes. Back then instead Europe or South Asia were surely sparsely populated in comparison, just because they were still in the hunter-gatherer period, what made them susceptible to some founder effects from West Asian agricultural immigrants carrying haplogroups like J2 and G2.

By the way, what about Göbekli Tepe (map)? It seems to be much older and, in a sense, more influential than Sumer, existing at the very beginning of Neolithic at the very crossroads of the three Neolithic regional groups. The very nature of the monument clearly implies some sort of priestly class, maybe caste, which may have left traces such as the ones you attribute to Sumer. The region (NW Kurdistan) is between Turkey, Iraq and Iran, where J2a and G2 abound.

Maybe you are onto something after all, i.e. a Neolithic priestly class/caste from West Asia with such Y-DNA markers but actually from an older and much more intriguing source: Göbekli Tepe and its strange monumental religion, which we have just recently begun to know about.