July 05, 2020
Home  »  Magazine  »  National  » Cover Stories  » Opinion  » Opinion »  After Meluhha, The Melange

After Meluhha, The Melange

India was never isolated. It’s seen a number of migrations, including the Vedic ‘Aryans’—and the Veda was no continuation of Harappan religion.

Google + Linkedin Whatsapp
Follow Outlook India On News
After Meluhha, The Melange
The Pashupati seal
After Meluhha, The Melange

Slice Of Pie

  • Proto-Indo-European was spoken in the Pontic-Caspian steppe around 4000 BC.
  • Spread through Eurasia. One line went west, into Europe. One, Indo-Iranian, came south.
  • The Indo-Aryan family branched off from that.



Michael Witzel is the Wales Professor of Sanskrit at Harvard University. At a conference earlier this year, he gave a talk tit­led Beyond the Flight of the Falcon: Early ‘Aryans’ within and outside India, to be published in Kumkum Roy et al. Excerpts from a version edited and extended for Outlook:

A major problem was, and to an extent still is, dating the early arya texts, especially the timeframe of the Rigvedic period. Archaeology alone cannot yet deliver relevant dates for northwest India (Greater Punjab), that is, for the end of the Harappan civilisation at 1900/1300 BC and the beginning of the Vedic civilisation. Rather, it is a combination of textual and linguistic data that indicates the Vedic period’s beginning. Increasingly, fine-grained genetic data, especially ancient DNA, may substantiate these results.


For a potential beginning of the (Rig) Vedic period, only a small area of Harappa has been stratigraphically (the study of rock layers) studied, providing data for around 1300 BC. Importantly, while the general Harappan pottery design is maintained, the vessels reveal some new designs and a shift to cremation with subsequent urn burial; more recently, additional data have emerged, like the extensive Harappan graveyard at Farmana, and the rec­ently found burials at Sinauli, allegedly dating to 1800-2000 BC—well before the immigration of the Indo-Aryans (IAs) to  Gre­a­ter Punjab. Thus, any overlap between Harappan and Vedic civilisations is as yet unclear, though it can be exp­ected for the Haryana/Delhi area.

Read Also: We Are All Harappans

Importantly, the people of the Vedic civilisation were semi-nomadic and did not dwell in the post-Harappan agricultural villages of Haryana; instead, they were constantly on the move with their cattle. We still need to find clear pastoral rem­ains of the period. Still, pottery remains and linguistic data ind­icate ext­ensive commun­ication between the two populations: there are many non-IA loanwords from before, during, and after the Rigvedic period.

As for clear repercussions of the Harappan civilisation on the Vedic populations, we must study: 1. the few clear archaeological indications (continuing pottery style [always the norm in successive cultures], depictions on vases, the red parting line in married women’s hair on some figures, etc.), 2. apparent remnants, merely in low-level strata of religion and 3. the impact of the (northern) Indus language on Rigvedic and later Vedic. All these indicate only minor continuation of Harappan elements, compared to the major rupture in civilisation beginning with the immigration of IA speaking populations around 1200 BC.


There was a northern and a southern (“Meluhha”) Indus substrate (a language that influences a more prestigious one) that influenced the IA language. Research, including evidence from Sindhi place names, points to a (later?) Dravidian settlement in that area (and Maharashtra), before IA speakers introduced the ancestor of modern Sindhi. It remains unclear when the Dravidians moved into the Indus area.

Read Also: 'Harappans United Regions Across 2 Million Sq KM’

In this connection, the question of the Indus inscriptions on seals and small tablets is relevant. Some, like Parpola, assume the underlying language is Dravidian. However, no “decipherment” has been accepted by serious scholars. Farmer, Sproat and Witzel have indicated that these signs must not indicate a script that can depict spoken language, but can be symbols.

The arrival of IA speakers in Greater Punjab is heralded by many loanwords derived from the substrate language of this area, the Northern Indus language. I have called this northern Indus language “Para-Munda” as it shares only part of typical Munda traits. There are about 300 loanwords in the Rigveda, even discounting loans that occurred earlier during the IA migration. Importantly, the oldest Rigvedic loans are not from Proto-Dravidian.

There are no loans reflecting the Harappans’ inte­rnational trade, seals, staple cereal (wheat, which appears only from the Atharvaveda onw­ards), towns, mythology (e.g., involving a tree goddess and a tiger, etc.). Clearly, the loans come from the post-Harappan rural population. They increase in post-Rigvedic times and involve other language families, incl­uding Dravidian.


A revolution has occurred, in the past 10 years or so, with the possibility of sequencing ancient DNA (aDNA), allowing us to specify population history in finer detail. We had no Indian aDNA until very recently, and recent excavations at Indus sites (Farmana cemetery, etc.) have yielded no genetic results so far. But we now (May 2018) have aDNA from the Swat Valley. Further, reports of aDNA from the Harappan site of Rakhigarhi have finally (June 2018) been announced in newspaper reports. Both are preliminary reports that await publication.

The Rakhigarhi data, published in the Economic Times on June 13, 2018, purports to show, in the words of the excavator, V. Shinde, that “the Rakhigarhi human DNA clearly shows a predominant local element… There is some minor foreign element which shows some mixing up with a foreign [Iranian] population… This indicates quite clearly, through archaeological data, that the Vedic era that followed was a fully indigenous period with some external contact.”

Niraj Rai, the genetic specialist of the Rakhigarhi report, echoes this: “skeletons at Rakhigarhi point to a predominantly indigenous culture that voluntarily spread across other areas, not displaced or overrun by an Aryan invasion…It will show that there is no Steppe contribution to the Indus Valley DNA… The Indus Valley people were indigenous, but in the sense that their DNA had contributions from Near-Eastern Iranian farmers mixed with the Indian hunter-gatherer DNA”.

Yet, conversely, he sustains Vagheesh Narasimhan’s May 2018 paper: “A migration into [ancient] India did happen… It is clear now, more than ever before, that people from Central Asia came here and mingled with [local residents]. Most of us, in varying degrees, are all descendants of those people.” Importantly, the R1a genetic marker, typical of the Western Central Asian Steppes, is missing in the Rakhigarhi sample. Rai adds: that “the analysis of the DNA sample will be of a period before [my italics] the Steppe people supposedly arrived in India. If R1a is absent in the Indus Valley sample, it suggests it was brought into South Asia, perhaps by a Proto-Indo-European (PIE) speaking group.” (Obv­iously, PIE is  too early).

Later newspaper reports add more materials and are more balanced. The initial conclusions about “no Aryan invasion” echo the cultural politics of the past 30-odd years and are blatantly weird, as one cannot expect the genetic materials of IA speakers in the Harappan Civilisation: they entered the subcontinent only after its dissolution, or at best, during its final phase around 1300 BC—not at time of the excavated materials, said to be around 2600 BC. We need confirmed dates for the one or two skeletons with recovered aDNA before we can make definite statements about the contemporaneous Indus population, keeping in mind that people moved around during the Harappan period as isotopes indicate. Archaeological data had already revealed that the population was not homogeneous and individuals had moved into Harappa from distant parts of the Greater Indus area (as seen in teeth enamel, etc.).

So, after decades of denial of any migration into India by arc­h­a­eologists, the pendulum is swinging back to constant contact, migration and population mixture—which would at  least allow for the migration of IA speakers into the subcontinent.

Precisely this is maintained in the recent pre-publication of a massive genetic paper by Vagheesh M. Narasimhan et al. According to this study, in the late second millennium BC, a large-scale middle/late Bronze Age Steppe migration entered the Indus periphery, apparently at least in part via the Inner Asian mountain corridor, where up to 30 per cent Steppe DNA is found with the Kalasha in Chitral, in westernmost Pakistan. This movement includes, from 1250 BC onwards, one to the Swat Valley where, for the first time, Indian aDNA has been retrieved. This precisely fits, both in time and space, the migration of the IA speakers of the Rigveda, visible in archaeology and linguistics. These migrants have also left a clear imprint on the genetic setup of the modern Brahmin and Bhumihar population of North India, where Central Asian traits are up to 57 per cent, while with other populations this amounts only to 11 per cent, and it is hardly seen in South India.

Religion, Mythology, Ritual

When comparing the texts on Vedic religion, mythology and ritual with those of the Indus civilisation and of those of the Indo-Iranian and Indo-European ancestors of Vedic culture, it is clear that the Veda is not a continuant of the Indus Civilisation (or even an overlap)—with the possible exception of some low level deities, spirits and demons (kimidin, mura-/shishna-deva) that frequently are isolated linguistically and belong to the Greater Punjab (“Para-Munda”) substrate.

The evidence for Indus mythology (visible on small tablets and some seals) is not reflected at all in the Vedas, and any link with later Hinduism, thousands of years later (championed by Parpola), is a phantasy: for example, the famous “Pashupati” seal reflects a widespread Northern Eurasian deity, the Stone Age Lord of the Animals; likewise, the killing of the buffalo demon Mahisha by a deity, etc., is separated from supposed Hindu continuants by chasms in time and space. Other supposed continuities belong to “low-level” cultural features: the red parting line in married women’s hair, and the Namaste gesture, which is found in the Bactria-Margiana Archaeological Complex around 2000 BC, and…in Jomon time Japan, 1900 BC: certainly not the spread of an Indus feature.

In Conclusion

In sum, neither was India ever isolated, nor did all facets of its archaeological, linguistic, textual, genetic/somatic data arise “on their own” locally; instead, they look back to some 60,000 years of Out-of-Africa history. The subcontinent presents a fascinating array of internal developments and external influences that only ­patient and unbiased study can reveal.

Next Story >>
Google + Linkedin Whatsapp

More from Michael Witzel

The Latest Issue

Outlook Videos