Multiple hearths of agriculture in ancient South Asia

Patrick Wyman’s Tides of History podcast is tackling South Asia and prehistory. He wrote up a Substack for it too, Ancient South Asia – Farming and People in India and Pakistan. I agree with Patrick here, though my confidence is low:

…It seems unlikely that a group living 1400 miles to the east would have chosen precisely the same suite of domesticated plants and animals as their related brethren in the Fertile Crescent. It’s intriguing that a fourth distinct group, as yet unsampled by geneticists, might have been living in the Fertile Crescent alongside their relatives 10,000 years ago or more. But the most likely, in my opinion, is that the group ancestral to later South Asians was living somewhere between the Indus Valley and the Zagros, perhaps on the Iranian Plateau: close enough to adopt some pieces of the Fertile Crescent farming package, close enough to head a short distance east, through the Bolan Pass, and into South Asia.

My confidence in this part is higher:

Yet they were not alone in South Asia, nor were they the only ones engaged in farming. Further to the east, along the Ganges River, the indigenous foragers were also experimenting with plant cultivation. In fact, there were no fewer than five places in South Asia where we see evidence of independent plant domestication. Mung bean, urd bean, horsegram, several varieties of millet, and rice were all cultivated extensively. These crops had the benefit of being able to grow during the summer monsoon season. South Asia was actually home to multiple Neolithics of its own.

In Southeast Asia and Europe, the hunter-gatherer populations contributed 20% or less to the ancestry of modern groups, who descend mostly from farmers and pastoralists. In South Asia the “Ancient Ancestral South Indian” (AASI) ancestry is ~50%. What’s the difference? I think the likelihood is that AASI populations were moving toward agriculture is a likely reason why they were much more demographically robust and impactful.


12 Replies to “Multiple hearths of agriculture in ancient South Asia”

  1. I read that substack with great interest. Some of it is obtained by reasoning on very thin evidence, but the most interesting part is this –

    But the most likely, in my opinion, is that the group ancestral to later South Asians was living somewhere between the Indus Valley and the Zagros, perhaps on the Iranian Plateau: close enough to adopt some pieces of the Fertile Crescent farming package, close enough to head a short distance east, through the Bolan Pass, and into South Asia.

    This is virtually the (Out of Iran/Bactria/IVC) theory. It has its own proponents based on linguistics but this is the first time I am seeing agriculture as a pivot for the theory.

    The arguments Patrick presents ties in with the linguistic differentiation for words for agricultural crops and cattle. Consider wheat (one of the crops he alludes to) – originally from the Fertile Crescent – they have been first attested in Mehrgarh in 5000 BC, then in Central India in 3500 BC and finally in the Deccan by 2000 BC. The great coincidence is that the words for wheat in Indian languages are all variants of Godhuma (Sanskrit) – in the Dravidian languages as well – implying that the crop was passing through the Arya hinterland before arriving in Dravidian etymology as a loanword.

    1. It was Johanna Nichols’ 1997 paper that postulated that Bactria could be the IE center based on loanwords loci (linguistics for agricultural livestock, metals and crops). Since then she has walked back that paper but the retraction was weaker than the original paper.

  2. It is worth recalling that several of the key parts of the South Indian Neolithic package of domesticated crops were originally domesticated from their wild types in the African Sahel (Dorian Fuller was particularly important in documenting this fact), although without much, if any, discernible demographic component coming with this culturally transmitted technology (Y-DNA T may be a legacy of this although it is hard to tell and the data available to me isn’t good enough to seriously determine the validity of these hypothesis). It looks to me as if Wyman isn’t familiar with this fact, and is as a result claiming South Asian domestication of some crops that were actually domesticated in Africa.

    Also, while the domestication of rice from wild types in Northern India may have begun earlier than the arrival of the Munda people, it didn’t really take off until Southeast Asian Austroasiatic Munda people (who had a better idea of what they were aiming for and how to farm it) migrated to the region.

    1. @ohwilleke,
      It is worth recalling that several of the key parts of the South Indian Neolithic package of domesticated crops were originally domesticated from their wild types in the African Sahel (Dorian Fuller was particularly important in documenting this fact)

      I am confused by this, because I seem to see this paper of Dorian Fuller:

      which says “Our review of the evidence, leaves no doubt that
      horsegram (Macrotyloma uniflorum) was domesticated in ancient India. The evidence of remnant wild populations today suggest that the wild progenitors were distributed in the semi-arid savannah or scrub woodland zones, including margins of tropical dry deciduous woodlands, of western and peninsular India, and also through parts of the lower slopes of the western Himalayas. This suggests two main wild distribution areas that were geographically separated, one in the savannah corridor of western and peninsular India, and one in the western Himalayas”.

      Cursory searches suggest that lentils were domesticated in the near east, but for specific ones like mung bean Wyman could be right?

      By the way, I am not familiar with the basics in the subject – can someone tell me the difference between when a crop gets cultivated and when it gets domesticated?

    2. if he’s unaware of the sahelian crop package and its presence at the earliest stages of agricultural diffusion in the west/deccan, thats a big miss. Is domesticated water buffalo part of the rice package?

      1. @girmit If you understand what is going on can you tell me what aspect of Wyman’s post contradicts Sahelian crop transfer? It seems understanding this will help me understand the claims of this post better.

        Regarding your question on water buffalo, at least in China water buffalo seems to have been domesticated well after rice was – “cultivation”, “domestication” and “agriculture” seems to develop historically in that order; maybe waterbuffalo fits the third the best?

        (Anyway I looked up my own question about cultivation vs domestication and got the answer – obvious in retrospect, but was too inertial to guess it before looking up. But that helped me follow ohwilleke’s point about Mundas; domestication seems to have been needed much more technology than I naively expected, though now I am wondering why the Mundas couldn’t simply use the seeds they brought).

        1. I have no qualification to refute wymans post , but its an odd thing to elide if one were aware of it. But perhaps he’s fully cognizant of the sahelian origin and its irrelevant to his point about domestication. I think jowar was first attested near the Inamgaon (MH) chalcolithic sites. I don’t know how common that is, for the same crop to diffuse widely and be domesticated independently.

  3. I still think that the group which moved with the fertile crescent influence into South Asia was more so an early copper age group (dental change in the copper age of the indus region) while an older eastern Iran population moving in and disturbing the balance of the locals in the mesolithic lead to newer fragmented cultures out of which local aspects of the South Asian Neolithic arose. Not to say they are responsible for it directly, but the disturbance probably was.

    The copper age migration would have probably resembled the earliest copper age samples from northeast Iran (which I recall are less anatolian rich than the inner Iran copper age samples but not so sure about this last point). It may be linked to Shahr BA1.

  4. One minor problem: Around the 18th minute he mentions the Neolithic of the Indus having the classical fertile crescent package, this is generally true. However, he says that they are morphologically identical to the analogues further west, which isn’t really the case. The fertile crescent grain imports form under 10% of the total finds, the local grains were morphologically different in being wild undomesticated naked local grain. This differs from and is distinguishable from the imports because the imported grains and naked undomistacated ones. It isn’t even like domesticated grain isn’t found, it is but in smaller quantities. The picture, at least to me, points to the more domesticated crops as being imports as opposed to a major incursion in the early Neolithic. Imported animals likely represent a larger share of the imported material than imported grains. Although later on around the 28th minute he acknowledges the breadth of opinion on the matter.

    Another minor problem: The claimed damdama stature for males is 180 cm, and for females is 174 cm (23rd minute). These values are outdated. Evidently, his claim is from the 2003 Lukacs and Pal paper. However, in 2014 Lukacs published an update where Damdama stature was listed as following:
    • Male femur estimate = 176.3 cm (-3.46 cm vs older estimate)
    • Male tibia estimate = 178.1 cm (-7.08 cm vs older estimate)
    • Male femur + tibia estimate = 178.7 cm (4.88 cm vs older estimate)
    • Female femur estimate = 166.4 cm (-3.19 cm vs older estimate)
    • Female tibia estimate = 167.7 cm (-7.47 cm vs older estimate)
    • Female tibia estimate = 167.8 cm (-4.62 cm vs older estimate)
    Why is there such a difference? Because the older estimations used a white American reference for the equations while the newer ones used a more tropical population like Egyptians. This is important because the ganga Mesolithic group bones are obviously tropical in nature (from the 2003 paper which compared them to various modern groups).

    Minor comment: Some of the things he says are reminiscent of the 1986 Sarai Nahar Rai paper. So I got that reference.

    Some other comments: I did not know about the specific architecture that he mentioned. What he said sounded more like a west Asian Neolithic thing. However, based on what I know, even in the earlier levels, there were common four roomed buildings which did have openings that connected the rooms to each other and outside. I do not think he mentioned that, and instead mentioned connected houses which were navigated by rooftops (more of a catahoyuk thing). Maybe his comments were regarding the later Neolithic/early chalcolithic layers of Mehrgarh. However, as far as the earlier layers are concerned: Here is a quote from Jarrige 2006 “This layout with its symmetrically disposed houses, with rather regular open spaces in-between, forms a marked contrast with the plans of several Neolithic settlements from Western and Central Asia, where the houses cluster tightly together and where there is no evidence for alleys, doorways or large open spaces. The plans of the houses from early villages so far recorded in the Neolithic of Western or Central Asia often show rather irregular combinations of small cubicles of various sizes.” Then again, the stuff that I do know about is from the aceramic/period 1, which by my estimation is prior to a notable demographic input in the copper age. Maybe the differences in latter vs earlier architecture, with later architecture coming to resemble Neolithic west Asian styles indicates a migration of people at that time. Goes well with dental change happening around that time in Mehrgarh.

  5. There is a poor understanding of how agriculture began in South Asia. There is a major bias towards the fertile crescent as far as looking into the origins of Neolithic is concerned.

    People here need to be calm and think rationally. The fertile crescent is a crescent shaped patch in the midst of an ocean of desert. West Asia and Near East cannot sustain a large population unlike a region such as South Asia. South Asia is blessed with monsoon, abundance of rivers and a very fertile land. Just compare population size of India, Pakistan and Bangladesh with those of Near Eastern or West Asian countries. Populations of Turkey and Iran are around 83-85 million each. Saudi Arabia – 35 million, Iraq – 40 million, Syria – 17 million. Their combined population is less than 300 million. They are puny countries compared to even Pakistan or Bangladesh.

    It is an artifact of the skewed and biased standpoint of western scholarship that tries to put the fertile crescent as the origin of everything. Their research into this region began in the late 18th and early 19th centuries because of the importance given to this region in the Bible.

    It is extremely dubious that people from thinly populated relatively dry desert wastelands taught agriculture to South Asians, where water is plentiful.

    Just for the record – the farmers at Mehrgarh have houses which are similar to the houses of early Iranian farmers who lived in the eastern part of the fertile crescent. The ancestry found across South Asia is derived from an ancestral population very similar to those of these early Iranian farmers. If farming was brought to South Asia, these early Iranian farmers can be the only source for it.

    But these Early Iranian farmers were foragers and goat herders. They did not have cattle. On the other hand, South Asians farmers in Mehrgarh as well as in Kunal, Bhiranna etc in Haryana, were already using Indian origin humped Zebu cattle and Buffaloes right from before 7000 BC. The dominant crop at Mehrgarh was barley and wheat made up less than 10 % of the crop remains. The range of wild barley does extend upto the northwestern regions of South Asia.

    Now we also know from genetics that the Neolitihic South Asian farmers did not come from Iran but were distant relatives of these early Iranian farmers and separated from them before 12,000 ybp and likely imho around 15-18 kya.

    It is unfortunate that people are not aware of these basic facts.

    Coming to rice, both agriculture and genetics support the independent domestication of rice in South Asia. The Munda did not bring it here. Early evidence of rice prior to 6000 BC have been found at multiple sites in the Ganga plains such as Lahuradewa, Jhusi, Koldihwa, Hetapatti etc. Here is the genetic evidence –

    Scholars like Fuller try to ignore or downplay the importance of this because it is strongly against at odds with their theory. Dont blindly believe them.

    Finally, the water buffalo was also domesticated in South Asia. The river buffalo in the Sarasvati-Indus valley while the swamp buffalo in a region stretching from Eastern India to Indochina. All buffaloes west of India, I.e. those found in Iran, Iraq, Turkey, Syria as well as in Egypt, Italy etc. have their origins in the western part of the subcontinent.

    1. The important aspect about the barley is that it looks morphological undomesticated like local wild grain. There are domesticated (and imo imported/traded) varieties of barley present in that 10% to which the wheat also belongs to. The idea is that if people were really moving in the early Neolithic in large numbers (compared to the natives) then why play it on hard mode?


Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.