August 01, 2008

Origin of Hindu Brahmins

Sengupta et al. (2006) suggest a Paleolithic/Neolithic, pre-Indo-European, origin of most Hindu Y chromosomes. This inference is, however, a house of cards resting on the assumption of a slow evolutionary mutation rate, which I have criticized elsewhere.

In my previous post, I suggested a Bronze Age origin for many Indian Y chromosome haplogroups. In this one, I use Ken Nordtvedt's Generations2 program to estimate ages for upper-caste Indians (Brahmins). As Brahmin status is elevated within the Hindu caste system, and is conferred patrilineally, Brahmins are excellent candidates for determining whether or not an exogenous Bronze Age population is responsible for the introduction of Indo-Aryan languages and the establishment of the caste system in India, in accordance with the much-challenged traditional opinion.

I report ages for haplogroups with more than 7 observations:


N Generations
Years (25y/gen)
Years (30y/gen)
H1 13 112
800 BC 1360 BC
L1 12 148
1700 BC 2440 BC
R1a1 55 139
1475 BC 2170 BC
R2 20 141
1525 BC 2230 BC
J2a (with DYS388)
17 214
3350 BC 4420 BC
J2a (without DYS388) 17 174
2350 BC 3220 BC

DYS388 tends to mutate very slowly in most human lineages, but quite fast in haplogroup J, therefore I report results with or without its inclusion.

The three most populous haplogroups (R1a1, R2, J2a) are the best candidates for lineages of exogenous origin, and have Bronze Age coalescence time, in accordance with the traditional theory. However, indigenous lineages (H1 and L1) and others with fewer numbers did enter into the Brahmin gene pool. The majority (64%) does appear to trace its ancestry to the early Indo-Aryans.

Thus, the congruent distribution of haplogroups R1a1 and R2 in India, as well as the limited occurrence of haplogroup J2a mostly in upper castes are easily explained.

An argument against R1a1's Indo-Aryan affiliation is that it is more diverse in tribal Indians. Yet R1a1 occurs at very low frequencies in tribal Indians, who in turn form a very small part of Indian society. A better explanation is that R1a1 in tribals is due to multiple founder effects, and its elevated presence among Indo-European tribals supports this scenario.

The pattern of J2a and R1a1 occurrence and age suggests different roles for these two lineages. J2a seems older, consistent with a larger effective population size, whereas R1a1 is found at a higher frequency and is not limited to Brahmins.

Perhaps, J2a formed the nucleus of the priestly caste, allowing it to diversify for some period of time, with a later founding of a succesful Brahmin lineage by a group of R1a1 males of different caste origin. Such conjectures will have to remain speculative for the time being.

17 comments:

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

And where is J2b2?

Dienekes said...

Not enough observations.

Unknown said...

India ..... 5.7 (3700BC) // Sengupta et al 2006
India ..... 4.9 (2900BC) // Jobling et al. 2006

AP said...

"An argument against R1a1's Indo-Aryan affiliation is that it is more diverse in tribal Indians ... A better explanation is that R1a1 in tribals is due to multiple founder effects."

That is the only logical explanation since diversity is low 'within' tribes but high 'among' them as seen from the R1a1 haplotypes in the Sengupta paper. Whithin tribes there are sigletons, repetitions, and closeness. Most of the 'among' diversity can be attributed to geographical distance, ie, multiple founder.

UncleTomRuckusInGoodWhiteWorld said...

The rulers in INdia did not tend to be Brahmins but Kshatriyas. Wouldn't we also need to look at them for a better understanding of Indo-European immigration?

Richard Parker said...

I posted the reference to this post to a group called
http://groups.yahoo.com/group/Indo-Eurasian_research

and got back the following message:"The thread is based on what all four moderators agree was that amateurish blog post, Richard. It slipped by us and didn't really belong on a professional List".

Previously, I got the answer, when I first referenced your post:

[Mod. note. One of the main uses of population genetics is to infer the direction of migrations. But you certainly can't do so in the simplistic way claimed in that blog post, which, e.g., tries to associate specific haplotypes
with "the nucleus of the priestly caste," etc. Quite silly - SF.]

They're so goddam condescending that I'd almost even giving them something to tak about.

Hope you'll go in there with asses' jaws or waving cudgels, and slay them

regards
Richard

Dienekes said...

I think I made clear what is speculative in my post and what is not. The comment about the priestly caste was speculative.

What is NOT speculative is that applying the 3.6 Z.U.F. rate which is based on simulations where haplogroups have an average size of less than a thousand men today to haplogroups of India is a joke.

Once this mistake is realized, it becomes apparent that haplogroups J2a and R1a1 coalesce to the Bronze Age, the period traditionally associated with the Indo-Aryan migration.

J2a is manifestly a non-Indian lineage, concentrated in the upper-castes. It is absurd to think that it was "Epipaleolithic" or "Neolithic" in India in the first place. How can a "Neolithic" Y-chromosome lineage that pre-dated the arrival of the Indo-Aryans by several thousand years end up almost completely in the upper castes?

As for R1a1, we now have direct archaeogenetic evidence linking it to the Andronovo culture, which was postulated as having an Indo-Iranian identity.

Hope you'll go in there with asses' jaws or waving cudgels, and slay them

Why bother.

interested bystander said...

In my personal page at FTDNA, there is a section titled (Y) Haplogroup. Under it are matches to a 12 marker selection.

It currently shows, for my R1a & R1a1, 2 two-step mutations from Inia; also, 7 four-step mutations from India & 2 additional four-step mutations from India-tribal; along with one each from Sri Lanka & Uzbeckistan.

My line came from Norway in the late 1800's.

interested bystander, Aug 2 '08

interested bystander said...

correction to my above post.

2 three-step mutations, instead of two-step mutations. All one-step and two-step mutations are fro Europe.

interested bystander.

McG said...

My expertise, if I have any, is within R1b. I have applied Zhiv's work and been satisifed with most of it. I have two major areas of interest: 1. How to count mutations, I do not think ASD is appropriate. 2. The value of 12 at 393.(which I have). I am also s116+. In R1b 393 = 12 is sparse, even though it is Paleolithic/mesolithic (my opinion). When a group of 12's are converged you get old TMRCA's and it is apparent in the range of values they have at the various dys loci. They are obviously much older than a random set of 393 = 13. The question is why are they so sparse??? If they're the oldest, all the population studies conclude that the older inhabitants dominate the gene pool. They clearly don't. In a separate study I concluded that events accompanying the great flood (MWP meltwater pulse 4)may have innundated doggerland and parts of the british isles at large loss of life??? It is not apparent that iberia was hurt but the mediterranean was hammered and, of course, the black sea flooded. I would like your opinion on the discontinuity in 12 in R1b and my thoughts above.

Anonymous said...

Dienekes:

When you started the right" mutation rate article,. I was expecting it to lead Indian Angle as it is one of the controversial topic.

One can agree A global population expansion happened in Bronze age period.

As you agreed you comments are speculative about priestly cast. Before publishing just simple wiki search on Indian demography might have helped you a little.

India has mainly 4 types of people:

1. Upper castes: warriors, trader and priests.(mostly live in plains)-15%
2. Tribes( people who live in forest and lives isolated like African Inca tribes) 7%
3. Lowers caste – Live in plains as part of society. 14% so called untouchables
4. other casts( who do professional trade.) 45%- some times 10% of this can go to upper casts.
And the other
5 Muslims 14%

R1 a is dominant in lower casts also whose population is close to 150 million., R1a spread across cast spectrum

Cast phenomena is different , it gives advantage to rich and priestly class. People who migrate and with exogenous origin can also end up in priestly class. The other part of your Greek army (J2) is one example historically they became Hindus.

Not sure how you can make argument for origin based on age. Your argument is purely speculative. The diversity and availability of multiple R branches clearly indicates the origin is close to India. Don’t forget present day Afghanistan is close and part of Indian gene pool. Pakistan and Afghans are clear examples of J2 and R1a differentiation. Just the borders separate J2 dominant population and R dominant population.

What is is your definition of exogenous origin?. And what palce you consider to make it exogenous to India?

No haplogroup originated or developed in Siberia either it is C, F , K, P or E-V13 they all came from places where human inhabitation is supported. Sibera is place where they can find human bones along with mammoths because of ice. No civilization. No people . No origin. They are just some transients. Only now because of technology last 1000 years people expanded there.

Your blog got popular and your hypothesis are slowly dominated by your cultural agenda..

Cheap Shot for your shot:
An article about Indians from one of these quarters is expected: Jewish, Greek, Chinese and Muslim as it is hard for them to accept Indian antiquity either culturally or genetically. I agree they dont look like Euro Aryans.
Also:
If your argument is valid. It will be definitely taken up by geneticists across the globe. Lets see.

Dienekes said...

The other part of your Greek army (J2) is one example historically they became Hindus.

If J2 in Indians was due to the "Greek Army" then why are Indian Brahmins lacking in any of the other haplogroups that were likely carried by the Greeks?

The absence of other West Eurasian haplogroups besides R1a1 and J2, and indeed of any more derived clades within J2 such as J-M67 strongly suggests to me that these were part of the original gene pool.

India has mainly 4 types of people:

I follow the categorization of Sengupta et al. in speaking about "upper castes". The penetrance of the West Eurasian haplogroups in Brahmins are also evident in the wider sampling of Sahoo et al. who does not have Y-STR data.

Your blog got popular and your hypothesis are slowly dominated by your cultural agenda..

My hypothesis is based on my examination of the assumptions of the evolutionary mutation rate.

A Paleolithic Out-of-India model for R-M17 expansion would have been plausible. But, a Bronze Age OOI model for it is implausible.

An article about Indians from one of these quarters is expected: Jewish, Greek, Chinese and Muslim as it is hard for them to accept Indian antiquity either culturally or genetically.

The antiquity of all _Y-chromosome_ lineages (whether Greek or Indian) is affected if the ZUF rate is rejected.

Anonymous said...

House of cards:

1. The three most populous haplogroups (R1a1, R2, J2a) are the best candidates for lineages of exogenous origin,

From where?.
And you said again

The absence of other West Eurasian haplogroups besides R1a1 and J2, and indeed of any more derived clades within J2 such as J-M67 strongly suggests to me that these were part of the original gene pool.

A Paleolithic Out-of-India model for R-M17 expansion would have been plausible. But, a Bronze Age OOI model for it is implausible.

Nobody proposed Bronze Age OOI model you are saying it.

Even though J2a is observed in North Indian Brahmins. it is also observed south Indian Middle class who look dark and also in Nepal who look mongoloid. Brahmin observation is one isolated case of J2. J2 is in India but it’s distribution is not large and uniform like Other groups. By numbers it is more in South Indian Middle class( more H people). It is less in north India ( more R* people)

2. Bronze Age population is responsible for the introduction of Indo-Aryan languages and the establishment of the caste system in India:

Caste system is Tribal origin It is pre Indo – Aryan these groups might have cemented it.
. Nobody Introduced language at least. The language of R1a+R2 dominated genetic branch is Indo Aryan in India, Pakistan, Afghanistan, Across the border desert starting from Iran it is Semitic in J* dominated races up to Atlantic.
R* branch spread in non desert, mountainous regions upto Europe.. All these have common words through contact or origin.

Dienekes said...

Nobody proposed Bronze Age OOI model you are saying it.

If R-M17 is of Bronze Age origin in India, as the Y-STR data imply, then it can't have been involved in Paleolithic migrations out of India.

>> it is also observed south Indian Middle class who look dark and also in Nepal who look mongoloid.

That isn't saying much, since haplogroups don't determine physical appearance.

>> Brahmin observation is one isolated case of J2. J2 is in India but it’s distribution is not large and uniform like Other groups. By numbers it is more in South Indian Middle class( more H people). It is less in north India ( more R* people)

I spoke of J2a not J2.

>> Caste system is Tribal origin It is pre Indo – Aryan these groups might have cemented it.

If the caste system was of indigenous origin, then how come the West Eurasian haplogroups end up concentrated in the upper castes? So, either it was introduced by the Indo-Aryans, or it pre-existed in India, but the Indo-Aryans took over the upper caste role.

>> Nobody Introduced language at least.

The fact that Indo-Aryan languages are very closely related to Iranian languages which in turn are related to Indo-European languages implies that ultimately the Indo-Aryans had ancestors who came from elsewhere, and the genetic picture is consistent with this hypothesis.

Patarames said...

Dienekes is right that any Greek J2a would at least have introduced some E3b into India but that’s not the case.

Furthermore J2a is present among Indo-Aryan population in Iran which is another factor that should be taken into account by people who want to connect J2a with Neolithic expansion.

Nirjhar007 said...

The above sentence was the cause of my error in this thread:-D:http://dienekes.blogspot.com/2012/07/fastibd-over-2257-europeans.html?m=1#comment-form
No Dienekes as you can see Bengali Brahmins don't have y-dna J2 though others do: http://www.nature.com/jhg/journal/v54/n1/fig_tab/jhg20082t1.html#figure-title
http://www.sciencedirect.com/science/article/pii/S0002929707623532
So you say Hg J2a represents PIE with R1a1,R1b, R2a the result of assimilation but the ages here are above 1500b.c for J2 so was I.E. was present in IVC before? You say J2 was the nucleus of the priestly caste but tell me if J2 was present from IVC times then was that J2 an non I.E. West Asian influx? Or the Presence of Swastikas in IVC seals gets the reason?.
Good times.