brown is a state of mind. :-)

So the BJP was right, and there is no Aryan invasion?

So the BJP was right, and there is no Aryan invasion?

my impression is that the hindu nationalists make a few assertions:

1) no aryan invasion
2) aryans indigenous to india

#1 is tenable, #2 is almost certainly not. there is a lot of evidence that the *indo-aryans* come from the "andronovo" culture of the eastern black sea. there was a group of *indo-aryans* who formed part of the warrior class in the syrian kingdom of mitanni circia 1500 BCE. i say *indo-aryan* because the gods indo-aryans worship have slightly different renderings than those indo-iranians worship, which indicates that BOTH groups had already differentiated by in the ur-heimat (homeland) of the indo-european tribes (that is, the indo-europeans of syria likely spoke a dialect closer to vedic sanskrit than avestan). the ur-heimat of the indo-europeans was almost certainly a very cold place far from the large bodies of water. this seems to exclude india. additionally, a lot of words for indian animals are neologisms in indo-aryan languages. example: elephant is ~ the animal with the hand in many of the languages (in bengali, elephant is 'hathi,' and hand is 'hath'). this indicates they had to make up the names as they met these animals, rather than being endemic basic elements of their world-view.

point 2 is under dispute. there is some evidence that 3500 years ago a very successful male lineage (that is, the Y chromosomal line) spread throughout northern india. this lineage is extent in smaller frequencies in southern india, and there is a strong positive correlation with 'higher' caste and this lineage. the implication is that a group of alien males simply prevented indigenous males from reproducing. north indian elites have traditionally practiced female infanticide at a rather high level, and the genetic data shows far more 'social promotion' of 'indigenous' female lineages than male lineages. an analogy might be much of latin america, where the male lineages tend to go back to iberia (or to a lesser extent africa) and the female lineages are amerindian or african.

the main dispute is whether the lineage in question is exogenous (related to a lineage in eastern europe & central asia) or indigenous (that is, an indian version of an ancient genetic line that made it big for whatever reason). if it is indigenous it is unlikely that this lineage (sometimes called M17) is directly associated with aryans. if it is exogenous, the 'aryan invasion theory' holds up far better. the data is really unclear now.

hope that helps.

p.s. short answer: there was an incursion of 'aryans.' the question is was it on the order of a few thousand over a short period of time or many many thousands over a longer period of time. recall that the spaniards conquered the new world with only a few tens of thousands of troops in the first hundred years (the toppling of the aztec and inca empire being on the order of 1000 spanish men who used native allies). de facto polygamy by criolla males for several centuries meant that a small number of iberians ended up contributing about half the genetic heritage of latin america, and almost all the male ancestry of 'mestizos.'

Then what is it that makes us modern Indians so incredibly varied, even within the same caste, class, and sub-cultural groups? "All Asians look the same," or so the saying goes. But even glancing at my East Asian friends (whose DNA is probably far more homogeneous amongst one another than Indians), I can discern the differences: number of epicanthal folds, physical stature, cheekbones, etc. Are the differences I'm pointing to simply small idiosyncrasies of humanity, thereby making this line of questioning futile?

Then what is it that makes us modern Indians so incredibly varied, even within the same caste, class, and sub-cultural groups?

1) many groups are clearly "admixes." the evidence is clearest from both indigenous indian jewish groups: the bene israel display the 'cohen modal haplotype' which has its origins 3000 years ago and unites the kohaneim (priestly) lines of the world jewry. tyoe 'lemba jews cohen' into google for some really interesting stuff. the cochin jews (according to a study that i cite on my site [follow link above]) have south asian female ancestors. this is obvious in that many are called 'black jews' and look similar to other keralites (to some extent, the bene israel also look indian). mixing between various groups is pretty common, the northeast and northern fringe of india shows a 'grade' between east asian and south asian.

2) when you mix different groups the variance in physical type increases because you have more genetic combinations at play.

3) india might have some very ancient lineages, which might have resulted in the development of a whole host of adaptive phenotypes which also recombine. there is a line of thinking which points to south asia the source of west and east eurasian lineages 50-80,000 years ago. they would argue that the 'european' or 'non-european' appearence of indians is an artifact of the fact that europeans are an adapted subset of south asians (the more convential theory is that aryan europeans invaded india and brought that physique, so it turns that theory on its head).

4) recall the large size of the subcontinent, it displays at least as much physical diversity as europe itself.

5) inbreeding will lead to novel adaptations and perhaps sexual selection that sets one group off another. indian groups are somewhat inbred. i have read for example that many south indian brahmin groups show evidence of population explosions starting from a small initial base. the initial base with be a random sampling, and since it is smaller than the source population one would assume it is atypical in some ways.

6) a lot of these things based on looks are subjective. i am pretty sure that europeans would be less likely to assert that 'indians are white people with dark skins' if indians had not produced some significant cultural output. many south asians have broader features that overlap with africans at one end (at least northeast africans like ethiopians and somalis) and yet rarely do people say indians are 'blacks with straight hair' (though ancient greek geographers did make this assertion about south indians, while comparing north indians to egyptians in color, though they asserted there were distinctions between the two groups).

whose DNA is probably far more homogeneous amongst one another than Indians

don't bet on it. east asians can distinguish each other pretty well (the japanese are about 25% 'ainu' genetically, which makes them hairier for one) and i think punjabis and tamils all look the same. india is somewhat more substructured than east asia from what i recall (cavalli-sforza?), but not nearly as much as one would think.

point 2 is under dispute. there is some evidence that 3500 years ago a very successful male lineage (that is, the Y chromosomal line) spread throughout northern india. this lineage is extent in smaller frequencies in southern india, and there is a strong positive correlation with 'higher' caste and this lineage. the implication is that a group of alien males simply prevented indigenous males from reproducing... if it is exogenous, the 'aryan invasion theory' holds up far better. the data is really unclear now.

Hmmmm. I'd have to check out if there's been an objection in the literature recently, but both the Majumder and Bamshad studies I reviewed on the blog here indicate high homology to Eastern Europeans at many loci (esp. Y chromosomal, but also some autosomal) for upper castes. As you say there is a different life history along the mtDNA for the maternal lineages. This is very similar to the emerging picture for the Jews, actually - invader/immigrant males & local females. It seems unlikely to me that homology to Eastern Euro chromosomes could come about b/c the invasion went in the *reverse* direction, and AFAIK the base pair homology is high enough that we're not talking chance similarity.

Have you seen someone who's lodged an objectiona gainst the sequence homology to Eastern Euros at *both* autosomal & Y loci?

The Genetic Heritage of the Earliest Settlers Persists Both in Indian Tribal and Caste Populations disputes the m17 Y marker as being evidence of recent male mediated migration (control-f Ra1). the courdaux paper i referenced came out later and seems more robust (kind of an amplification of what wells was saying-that is, he thinks m17 is a signature for male mediated migration). basically, toomas kisivild is making a persistant case for a predominantly 'indigenous' origin for south asians. the case is obviously strong via mtDNA, and i suspect he'll try to say autosomal isn't to be trusted and the NRY coalescence can't be calibrated correctly.

Fascinating stuff guys. What books would you recommend to laymen curious about the current thinking on the Indo-Europeans?

What books would you recommend to laymen curious about the current thinking on the Indo-Europeans?

books are always outdated now, this is a reawakening field in the way of nazi assocations from world war II. in search of the indo-europeans is pretty good, but if you want to keep update date, you need to read archeology, philology and genetics journals.

also, go to my blog and control-f "historical population genetics" and check out those books. i rec. 'the real eve' and 'the journey of man' over the others.

Where in the world do you think he's from (no cheating!):

[If I can't get comments to load the image directly, here's the link:
http://www.bbc.co.uk/radio4/today/reports/archive/international/images/abdulah_yusef2.jpg]

Is he desi? Latin American? What do you think?

Razib: Australian aboriginal people fit the black people with straight hair stereotype a bit better, but yes, as I pointed out earlier, at least w.r.t. American blacks, there isn't alot of difference between Vijay Singh and Tiger Woods

googling says somalia? be cautious to judge and individual book by its cover.

Razib, that's cheating. I meant, look at the face and guess ...

i would have guessed maldives or something. the JPG has a muslim name, and he doesn't look bangladeshi or pakistani or "african."

Interesting article and interesting link.
thanks for sharing.
here are some more papers related to this discussion.
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=14761656
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=11988631
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=12536373

prak, all three papers are referenced above. you can find links to the full text the first two papers on our site, and the last one is linked up there.

Thanks for the pointers.

Razib - I've been reading GNXP for a while and have gotten tons of good information out of it. I'm still looking for something that can crystallize and give me a good overview of the current thinking. Sadly, I am not an expert and don't have the time, access or technical training to read the scientific journals. I do browse the human biodiversity websites however.

It does strike me that the research has gone far beyond what is currently out in book form. I'm sure that there would be a huge market for a general interest book on Indian origins and human biodiversity that sums up the recent research. It sounds like a lot of conventional thinking has been upended and good layman's intro might make something of a splash with the NYT book-reading set.

Oh well, I'm off to browse the GNXP archives.

BTW - another good link to follow if other laymen are into this subject is http://dienekes.blogspot.com (it's also linked from GNXP).

Bamshad and kisivild et als recent papers are contradictory.

Bamshad trys to bolster AIT based on homogeneity of Y polymorphisms in caste groups vs tribal groups.

Whereas Kisivild cites greater diversity of the m17 mutation(supposed Indo Aryan marker) in india-pak-afgh-eastern iran compared to central Asia.

The diversity argument is compelling. What is unfortumate is that for all the linguistic theories that people are trying to fit their data to, thay completely ignore the impact of the harappan civilization and the archeolological evidence of a population shift from western india to the gangetic plains and to a lesser degree towards south india.

Considering this the homogeneity observed in caste populations is not very remarkable. The contribution towards a 2nd phase of urbanization/population explosion will be greater for people with an urban history as compared to scattered tribal populations. Also social stratification is more a function of urban societies and has lesser utility for smaller tribal populations. What is to be remembered that the harappan civil would have drawn from many surrounding gene pools and caused their homogenization for close to 3 millenia.

Its impact cannot be ignored.