Genetics And Aryan Migration: Setting The Record Straight (Once Again)

Anil Kumar Suri

Jul 14, 2017, 07:53 PM | Updated 07:53 PM IST

South Asia (Wikimedia Commons) 
South Asia (Wikimedia Commons) 
  • The Hindu’s Readers’ Editor A S Panneerselvan has attempted to defend Tony Joseph’s article on genetics and the Aryan migration theory.
  • Panneerselvan in his defence of Joseph, and Joseph, in his response to the geneticists’ article, appear to evade fundamentally important points.
  • Here, Anil Suri scrutinises the arguments offered in both the articles and responds. 
  • The Hindu’s Readers’ Editor A S Panneerselvan has attempted to defend Tony Joseph’s article on genetics and the Aryan migration theory from some of the points made by me earlier in Swarajya. By now, senior geneticists Drs Gyaneshwer Chaubey (whose disagreement with Joseph’s article I had referenced in my piece) and K Thangaraj have published a detailed rebuttal to Joseph’s optimistic claims, principally that geneticists have now “converged” on the answer to whether an Indo-Aryan immigration to the subcontinent did actually happen, which is that there is now “sure-footed” proof of such immigration.

    Panneerselvan in his defence of Joseph, and Joseph, in his response to the geneticists’ article, appear to evade fundamentally important points that go against their narrative, and, when all else fails, bluster their way through. The arguments they use may perhaps confound the general reader. I therefore try to scrutinise their arguments afresh. Panneerselvan also tries to deflect attention from my criticism by levelling a bizarre accusation at me, and I shall set the record straight on this.

    Spin over facts

    For starters, Joseph had said, ...this theoretical structure (of Ancestral North Indian, ANI/Ancestral South Indian, ASI – parenthesis mine, for clarity) was stretched beyond reason and was used to argue that these two groups came to India tens of thousands of years ago, long before the migration of Indo-European language speakers that is supposed to have happened only about 4,000 to 3,500 years ago.” Now, the genetic affinity between some North Indians and Europeans has been known for a long time, and had been attributed to an ancient common ancestry rather than the more recent migrations of the Indo-Aryan type. So, when Joseph did not care to specify who exactly he was accusing of this “stretching”, he was choosing spin over facts. Tellingly, Panneerselvan does not address this criticism of mine either. As I wrote in my article, it was in a 2011 peer-reviewed scientific publication, whose authors included Thangaraj, the one who had originally proposed the ANI/ASI construct, and Chaubey, that this conclusion was arrived at. They have now pointed this out themselves in their article in The Hindu.

    Why cite selectively?

    The 2011 findings of Chaubey, Thangaraj et al are again corroborated by a recent paper by Partha Majumdar and co-workers (Mondal et al, 2017), who also differed from Reich’s 2016 hypothesis of recent migrations into the subcontinent, as I said in my earlier article. Interestingly, this paper was published at almost exactly the same time as the Martin Richards paper (Silva et al, 2017) that Joseph relies on to argue in favour of an Indo-Aryan migration. Significantly, the Majumdar paper has the virtue of analysing many Indian tribal populations for the first time in addition to data from the 1000 Genomes project, which makes their data set superior to the rather narrow database of Silva et al, who focused mostly on South Asians living in the United Kingdom and the United States (see the article by Chaubey and Thangaraj for why this is so). Since Joseph emphasises the importance of newer and better data, it is puzzling why he dwells on Silva et al and doesn’t even mention other studies like that of Majumdar. Indeed, Joseph may play up what he chooses to, but how exactly did he arrive at his claim of “unambiguous” conclusions and “convergence” of opinions?

    This is Joseph’s actual sin: the selective exclusion of divergent schools of thought from his narrative.

    I cannot help repeating the question I asked in my earlier article: did Joseph leave out the opinions of scientists that did not go with the narrative he was trying to present (as he seemed to have done with Chaubey), or did he just not even seek their opinion at all? Again, he is free not to interview experts with differing perspectives. But then, to claim that they have all converged on a conclusion would be taking liberties with the truth. Would this come under the ambit of due diligence that Panneerselvan talks about in his defence of Joseph? I sincerely hope Joseph does not claim he was not aware of the 2011 Chaubey/Thangaraj and 2017 Majumdar papers, for that would imply sloppy homework.

    Science fail

    Panneerselvan makes a truly bizarre statement when he says that I have not provided any statements from Chaubey “to validate his disapproval of the study”, and that “...Suri’s article does not throw up a single specific objection of Dr Chaubey either to the study of which he is a part, or to the extensive interview with Peter Underhill that was cited in Mr. Joseph’s article.” It is unclear why Panneerselvan (and Joseph?) think Chaubey disagrees with a scientific publication of which he is a co-author; if he had, he would probably have asked not to be included in the list of authors – as he indeed did in the case of Silva et al – or placed a different position on record in a subsequent publication. And if I had said such a thing, I am sure he would have written to Swarajya by now accusing me of attributing false statements to him.

    I have already highlighted this point in my previous article, and repeat it now: the 2015 Underhill paper says,“We caution against ascribing findings from a contemporary phylogenetic cluster of a single genetic locus to a particular pre-historic demographic event, population migration, or cultural transformation. ....our data do not enable us to directly ascribe the patterns of R1a geographic spread to specific prehistoric cultures or more recent demographic events.”

    Why would Chaubey disagree with these conclusions?

    Let me put it in another way: Joseph claims Underhill has now changed his views. To what, exactly?

    Joseph’s appalling response to the geneticists

    Joseph writes that the geneticists say “that ‘Out of India’ is a possible explanation for the genetic spread that we observe,” and further accuses them of making “no reference to a single peer-reviewed genetic study that makes a serious case for ‘Out of India’.” I am at a loss for whether this is a calculated straw man argument or just downright silly. One fails to see where the geneticists have backed one direction of migration or the other. They have only alluded to the well-known fact that genetic similarities exist between some Indian populations and Europeans, which imply some migrations in or out of India (we don’t know the direction, or even time, yet), but nothing so far suggests the hypothetical Indo-Aryan one (last para).

    Also, why does Joseph point fingers at Chaubey and Thangaraj? The very paper that Joseph tries to present as gospel, Silva et al, makes this statement: “Gene flow at this time (Bronze Age – parenthesis mine for clarity) was clearly bi-directional, as seen in the expansion west of lineages M5a2a4, U2c1b + 146 and M3a1b + 13105).”

    Indian R1a strand not derived from the Central Asian – the fatal flaw in Joseph’s narrative

    As I said in my earlier article, fundamentally, to argue that the predominant R1a strand found in India is the result of gene flow from Central Asia; the former should be derived from the latter, which is not the case. Chaubey and Thangaraj have stated this themselves in their article now (third-last paragraph). This effectively knocks the bottom out of pretty much everything Joseph was trying to argue. Although this is a very simple (and crucial) point, Joseph misses it altogether (second paragraph of his reply). Panneerselvan too beats around the bush while evading this basic point.

    Chaubey and Thangaraj have also re-emphasised that many tribal communities carry high frequencies of R1a, a fact I stated earlier in my article. For good measure, the geneticists add that their own (as yet unpublished) findings are that many R1a branches show a very early presence in India. This is one of the reasons why, as the geneticists said, sampling is so crucial – again, a point that eludes Joseph completely, leading him to write that “under-representation ceases to be a material issue”.

    It is worth repeating: we have better data today than we did years ago, but as it stands today, there is no case for an Indo-Aryan immigration into the subcontinent.

    Virtue signalling

    Panneerselvan tries to claim the high moral ground for Joseph when he says, “… one cannot cite earlier papers, when the discipline was at its formative phase, to disprove the contemporary findings when the discipline has become robust with substantial amount of additional data and analytical tools. This repudiates the basic rules that govern science and knowledge production.” However, as we have seen, Joseph appears to have relied on spin to downplay findings that do not agree with his narrative (Singh et al, 2011), and plain ignored other recent findings that he found similarly uncomfortable (Majumdar). He tries to spin the 2015 Underhill paper his way, but unfortunately Underhill’s very cautious and carefully worded statements in his interview to Joseph don’t quite acquire the desired colour. One is not entirely convinced by the simplistic ‘Recent Papers vs. Old Papers’ scenario that Joseph and Panneerselvan are trying to present it as – it appears to be a clear case of cherry-picking of evidence to build a narrative.

    Also, Panneerselvan does not bother to explain why Joseph came to regard one 2017 paper (Silva) to be worthy of note, but not another published at exactly the same time, which arrived at a different conclusion (Majumdar), and indeed how he went on to claim that one conclusion had been “unambiguously” arrived at.

    Panneerselvan’s comment also seems to betray an ignorance of the peer-review process. Recent papers do not automatically supersede older ones. They may use better data and/or technology, but may be flawed in their analysis or interpretations. We do not know if a publication has been previously rejected by other journals, and, if so, why peer reviewers had found it unsatisfactory. In any case, it takes a long time, and usually a good deal of corroboration by others, for the peer community to be convinced by any publication. Joseph’s enthusiasm in trying to paint the barely months-old Silva et al paper like it is the final word on the topic is touching, but a great deal of time needs to pass before we know how much acceptability it has gained.

    What the experts say

    It is precisely to gauge what the peer-community thinks of the paper that Joseph should have made an effort to obtain the views of a cross-section of scientists like Drs Lalji Singh, K Thangaraj and Richard Villems, who had previously concluded that there was no genetic evidence for an Indo-Aryan migration. Also, as with any open scientific problem, scientists may continue to pursue further studies to confirm or refine their own previous conclusions, and may still be in the process of publishing these new findings. Such new studies may well end up corroborating their earlier conclusions – indeed it is evident from the geneticists’ article in The Hindu that such is their case.

    If he had talked to other experts, they might also have pointed him to what appear to be rather obvious flaws in Silva et al. For instance, they regard only Y-chromosome haplogroups H, K2a1 and C5 as native to South Asia from before the Holocene. However, high incidence of the ancestral form of haplogroup R2, for instance, is found in both Indo-European and Dravidian speakers and tribal communities, and extremely low frequencies in Central Asia, ruling out the possibility of it having entered the subcontinent during the Holocene. But Silva et al claim R2 entered India from West Eurasia.

    Using similar arguments, if we classify haplogroups C5, H, H1, L1, R2, H0, H2 as South Asian, G1, G2, J1, J2, R1b as West Eurasian and C3, K2, N, O2, O3, Q1a, Q1b as East Eurasian, and, keeping R1a as unclassified for now, re-plot the ancestry contributions of the sampled South Asian populations shown in Figure 3(c) in Silva et al, notice how much difference it makes. The difference is so wide that one needn’t apply any statistical test. And this is without taking into account the problem of limited sampling of mostly expatriates referred to above. How did the peer reviewers who waved this paper through miss the prima facie questionable conclusion that the Dravidian (Telugu and Tamil) speakers sampled had nearly 70 per cent West Eurasian ancestry, indeed more than the Gujarati (Indo-European) speakers?

    The ancestries of South Asian populations as shown by Silva et al (left), and after correction (right).
    The ancestries of South Asian populations as shown by Silva et al (left), and after correction (right).

    Lastly, given the rather brief rebuttal from the geneticists, one would like to know if, in the interests of fairness, they were allowed a sufficiently detailed article by The Hindu’s editors – unrealistic word limits are an obvious editorial ruse for suppressing uncomfortable points of view.

    Also Read: Genetics Might Be Settling The Aryan Migration Debate, But Not How Left-Liberals Believe

    Anil Kumar is a materials scientist.

    Get Swarajya in your inbox.