Why Bronze Age steppe people replaced the farmers they conquered

One of the major revisions in my own mind about the demographic and historical processes of the Holocene in relation to humans has been the reality that large and dense agglomerations of agriculturalists could be marginalized by later peoples, to the point of having a smaller genetic footprint in the future than anyone might have imagined. If you had asked me ten years ago I just wouldn’t have believed that the first farmers of Europe or South Asia wouldn’t account for the vast majority of the ancestry of the contemporary populations of the region. By “first farmers” I don’t even mean migrants. At that point, I had assumed a primarily Pleistocene indigenous hypothesis for the origin of Europeans and South Asians, with farming diffusing through a mixture of a few migrants along a demographic wave of advance.

That’s not what it looks like according to ancient DNA. In Northern Europe, it seems that around half or more of the ancestry is due to the incursions of a pastoralist steppe population during the Bronze Age. In Southern Europe and South Asia, the fraction is closer to 10-25%. But even in the latter case, the fraction of steppe ancestry is far higher than I had expected.

I had assumed that the steppe migrants would contribute 1-5% of the ancestry of Europeans and South Asians and that the spread of Indo-European languages was a matter of elite transmission and emulation. Think the Hungarians, for example, as an example of what had assumed.

So what explains what really happened?

During the Mongol conquest of Northern China Genghis Khan reputedly wanted to turn the land that had been the heart of the Middle Kingdom into pasture, first by exterminating the whole population. Part of the motive was to punish the Chinese for resisting his armies, and part of it was to increase his wealth. One of his advisors, Yelu Chucai, a functionary from the Khitai people, dissuaded him from this path through appealing to his selfishness. Chinese peasants taxed on their surplus would enrich Genghis Khan far more than enlarging his herds. Rather than focus on primary production, Genghis Khan could sit atop a more complex economic system and extract rents.

Most of you at this point can see the general framework then. For thousands of years, pastoralist people of the Inner Asian steppe and forest would extract rents out of the oikoumene by threatening them with force. The reason the East Roman Empire did not face the Hunnic onslaught during the lifetime of Attila is that they paid the horde tribute. Imperial China did the same during some periods. In other instances, civilized states found in the barbarians of the steppe useful confederates. The Tang dynasty did not collapse during the 750s because of the intervention of the Uyghurs, who suppressed the rebellion of An Lushan. In 9th century Baghdad the rise of the Turks was enabled by their usefulness in court politics and distance from any given faction.

The rise of the “gunpowder empires” during the 16th century and the eventual closing of the Inner Asian frontier with the crushing of the last embers of the Oirat confederacy between the Russian and Chinese Empires in the 18th century marked the end of thousands of years of interaction between the farmland and pasture.

But this makes us ask: when did this dynamic begin? I don’t think it was primordial. It was invented and developed over time through trial and error. I believe that the initial instinct of pastoralists was to turn farmland into pasture for his herds. This was Genghis Khan’s instinct. The rude barbarian that he was he had not grown up in the extortive system which more civilized barbarians, such as the Khitai, had been habituated to.

In these situations where pastoralists expropriated the land, there wouldn’t have been an opportunity for the farmer to raise a family. Barbarian warlords throughout history have aspired to be rich by plundering from the civilized the peoples…but would the earliest generations have understood the complexity of the institutions that they would have to extract rents out of if there wasn’t a precedent?

Instead of conventional historical dynamics of predatory elites and static peasantry, a better way to understand what occurred with the incursion of steppe pastoralists during the Bronze Age might be a simple ecological model of intra-specific competition. In a pre-state society defined by clan and tribal ties, steppe elites may have seen the farmers who were earlier residents in the territories which they were expanding into as competitors rather than resources from which a life of leisure might be obtained. In other words, instead of conquest, the dynamic was of animal competition.

Of course, pre-modern societies did not have totalitarian states and deadly technology. Rapid organized genocide in a way that we would understand was unlikely to have happened. Rather, in a world on the Malthusian margin, a few generations of deprivation may have resulted in the rapid demographic extinction of whole cultures. You don’t need to kill them if they starve because they were driven off their land.

In fact, we have some precedent of this historically. The Spaniards were intent on extracting rents out of the native peoples of the New World and living a life of leisure, but in many areas disease and exploitation resulted in demographic collapse. Imagine a conquest elite as vicious as the Spaniards, but without thousands of years of precedent that conquered peoples were more useful alive rather than dead. 

Addendum: The fraction of haplogroup M, which is probably derived from Pleistocene South Asians, is greater than 50% in places like Sindh. This indicates that the steppe migrations were strongly male biased in the initial generations.


25 thoughts on “Why Bronze Age steppe people replaced the farmers they conquered

  1. Nice opener to the topic (competition vs extractive “symbiosis”/parasitism). My thinking, slightly in contrast (to at least offer another possibility):

    1) In terms of likely Indo-European complexes moving into Europe*, at least for North and Central Europe, in specifics, we have the Corded Ware Culture and the Bell Beaker Culture.

    Though heterogenous, the Corded Ware Culture appear to been quite mobile and pastoralist, a little more sedentary than the Yamnaya, while the BBC are more sedentary and agricultural yet, and isotope analysis indicates are roughly comparable in mobility to the EEF Funnelbeaker / Globular Amphora Culture which the Corded Ware replaced in North and Central Europe. So BBC perhaps not exactly these very mobile pastoralists really?

    2) David Reich notes that the entry of Indo-Europeans looks to be around the time of clearance of forests and transformation of the European environment to being more steppe like, and offers a tentative suggestion that the big success of IE groups in Europe may be their worldview and specifically how this affected that they used much more of the environment (which early farming groups made more marginal use of).

    So putting these together I would guess something like that:

    – Corded Ware first expand into the Northern zone of Europe where farming was very marginal, with the Central German Corded Ware we have as probably the most major incursion into serious farming territory. The CWC horizon really doesn’t include many places we’d expect to have serious depth of agriculturalists at this time in history.

    Largely the CWC are clearing forests, turning them into pasture and introducing a new form of pastoralism into lands where hunter gatherers, if anyone, subsisted, rather than turning farmland into pasture. They may take some farmland, but farmers just aren’t that well established across most of the Corded Ware zone, with its very cool, forested climates that are rather ill fitting to Anatolian derived agriculture.

    – Bell Beaker may form from a complex process of CWC+other agricultural Europeans from Central Europe (+possibly steppe) groups. They are a more agricultural complex, but still with the forest clearing behaviours seen in CWC that distinguish them from early farming groups in Northern and Central Europe, who were more likely to seek out preferential microclimates for farming than to transform the landscape, and hence had limited demographic size.

    The Beaker expansion into Britain then, is another story of forest clearing and new habitats, rather than violent takeover of old ones as much. This kind of seems to fit to me, in that archaeologists seem to talk about continuity with pre-Beaker and post-Beaker at any given site, and a major population reduction just prior to the Beaker period. The way to square this could be if the “continuous” sites are not telling the whole story about population change for the Isles as a whole, because the big transformation is the opening up of new niches in the landscape, not any kind of discontinuity at given sites. Archaeologists are right about the sites they looked at, but those are less informative about the overall picture than they thought.

    – Once fairly large population sizes are established by the Corded/Beaker wave in North Europe, clearing forests and establishing agriculture (and more agricultural BBC tend to replace CWC – “farmer power”), they started to look south to the much more primo farming land there, and generally you then get a pattern of North->South expansions from worse->better farming land (along with a bit of raiding of surpluses). Which eventually by the time of the Roman Empire we see a group in SW-SC Europe try to put a stop to, expanding political control into the north and marginalising the Celtic groups (whether this long term worked, or just loosed Germanic groups on the border…).

    So in this interpretation, IE groups are less conquerors who re-purposed deforested North-Central European farmland for pasture, and much more like fairly humble peoples who were distinguished from the farmers who came before them by clearing forests for pasture, and then for themselves to farm.

    (At the same time, some phenomena go on at the eastern end, with Corded Ware forest clearing IE groups from the forest zone of western Russia expanding to the steppe as more sedentary Sintashta to replace more pastoralist, Yamnaya-like groups there, and probably the formation of early proto-Balto-Slavic people from a mix of Corded Ware groups and people with high levels of hunter gatherer ancestry around the Baltic).

    Population size estimates from ancient genomes and archaeobotany (when exactly did those forest clearances happen?) will sort this all out. In your interpretation, popsize should be very high among farmers and forests cleared before IE expansion to Europe, whereas in mine, popsize probably quite low and forests mostly cleared after late Neolithic to early Bronze Age.

    *I wish there was some easy geographical term for specifically “That part of Europe that excludes European Russia”, which would be more accurate.

  2. where farming was very marginal
    It doesn’t even have to be VERY marginal, just sufficiently ineffective to offer no great advantage to collecting a portion of excessive farming output vs. getting the whole grazing output. One needs to feed both the farmers and their government and taxation superstructure to collect taxes hands-off, so in the end only a meager fraction of the land’s riches may be available as plunder. Even in the societies which developed more efficient farming, the disruptions of wars, conquest, and resulting anarchy, starvation and epidemics may have turned farming into marginal use of land (as it probably happened in Central Asia in the Middle Ages, with its collapse of government and irrigation systems)

    Not to mention that in pastoralist-on-pastoralist conflicts, their inherent mobility already made self-exile of the defeated the norm of warfare, so they historically expected their vanquished rivals to pack up and leave, rather than to offer rich tribute.

  3. Matt, my recollection is that hunter-gatherer ancestry actually went up with the arrival of pastoralists, since they had more of it than the inhabitants of Europe at the time.

  4. There are potential biological reasons for male-biased admixture from Yamnaya unrelated to any social ones. Steppe populations were substantially Rhesus-negative and India/Europe were not. Steppe women and local men could be expected to be selected against even in a random mix – in addition, Steppe men were less likely to be immune-rejected by local women than their own women.

  5. @TGGP, what’s that in relation to? Early Yamnaya / Corded Ware have more HG ancestry than MN farmers, obvious less than HG rich people who persisted around Baltic for sure (see Mittnik 2017) and poss. Northern Forest Zone (and so there replacement by pastoralists = reduction on HG ancestry).

  6. Agricultural states in close proximity to each other in Mesopotamia did prefer conquest and/or tribute from very early times, no? Thus the growth of empires, very modest ones at first

    With the Mayan city-states it is not so clear. I would be interested to know more about how they interacted in war. The “First Civilizations” series on PBS is not terribly informative in this regard. The word “conquest” only appeared one time in the first segment, which included discussions of what looked like obvious conquest states further north.

  7. “Rather, in a world on the Malthusian margin, a few generations of deprivation may have resulted in the rapid demographic extinction of whole cultures. You don’t need to kill them if they starve because they were driven off their land.”

    In addition, diseases. Not only, the farmers were dispossessed, they communicated new diseases and also displaced. A good model for India might be that the dispossessed farmers migrated to central and southern India, but an inability to farm their agricultural package might have made the population less impactful. ASI, either from Iranian farmer or AASI is still significant beyond the Vindhyas.

  8. The pastoralists didn’t even have to drive people off of their land. There were no fences to protect cropland. So they could simply allow their animals to graze on farms, and the farmers were instantly in trouble. You can’t undo the damage.

  9. Very interesting. So if the migrations were male-mediated, and they were initially planning to simply drive the farmers off their land so they could use it for pasture, were they also initially planning to steal women for wives from the farmers? Or do you think they would have been planning to bring women from their steppe territories over later?

  10. @Rick
    Good analysis, but what about resistance? Weren’t there Harappan/EEF armies?

    In the Mannerbund model, in the Foundation-of-Rome model, bands of youthful, strong and militarized men would join together, leave their commumity and route another settlement, killing the men and taking the women to create a new community for themselves.
    This, in my opinion, might have been the core IE modus operandi.
    They liked to spread themselves, their “seed”, their people. Colonisation is in the center of IE thought (imho), including their expansionist behaviors and mindsets. We can see that in every IE society.
    This is not universal to them, nor unique of them, though.

  11. Razib,
    What is the source for addendum in your post regarding 50% haplogroup M in Sindh? Just curious if there is more data on Sindh in the same source.

  12. Related may be the fact that only after Iron age farming could be productive enough for generating lootable surplus.

  13. Most successful people of the crucial time were clan based and based their networks on kinship. They were bio-logically successful and those which won the fights had more offspring.
    Slaves and subjugated people would have lost the good wives and low rates of offspring, if they kept slaves to begin with. In many cases they would have just killed the foe’s males and a lot, but the most interesting females.
    The book from Lawrence H. Keeley is great in describing primitive warfare. The main reason why large scale genocides were not that common is that a total victory of complete strangers is the exception. If such victories happened, because one group had a significant edge, the result was almost always, and logically so, the large scale decimation to extinction of the losers.
    The same can be observed in some farmer wars before Indo-Europeans. Actually the new clans with hunger-gatherer lineages in Europe point to such conquests of farmer communities by them. New, even more warlike hybrid cultures emerged and spread. That’s why the original G2a domination of the EEF seems to disappear even before the IE conquest.
    IE were probably just the most successful of the hunter gatherers adopting to the new conditions and had a more favourable starting pointing from starting with.

    I think the decimation of the forein males is absolutely logical behaviour and if you look at things from that perspective, a lot of people finally got overrun by foes they didn’t cripple when they had the chance to. Imagine you spare a tribe which is your bitter enemy, how can you know that this very same tribe won’t be stronger the next time and if its in 100 years from now? Obviously, in a context in which bitter tribal fights were the rule rather than the exception, you culturally learn your lesson or become one of the vanquished. That’s why we see so fast and radical changes in the male yDNA.

    I rather wonder why the Indo-Aryan input into South Asia is so low and my best guess is that even at that time diseases deadly to more Northern European like people were rampant in the region. So the most effective, probably even the only successful way for the newcomers to establish themselves on the long run was mixture and even then the selection might be in favour of the locals.
    Along those lines, it would be very interesting to see whether the tropical climate functions like a barrier for the higher West Eurasian influence in India.
    Comparing with this map: https://en.wikipedia.org/wiki/Climate_of_India#/media/File:India_climatic_zone_map_en.svg
    The relationship of Indo-Aryan vs Dravidian based on climate alone is striking. And the higher proportions of IA ancestry are mostly confined to the drier and definitely not tropical regions of South Asia.

    As for the change in later times: Gold and money were crucial. Without money to control and tax a large population would have been much less lucrative for the conquerors. Gold and money really tamed the steppe conquerors in later times and made their biological success much smaller because of their greed. So the human greed saved later civilizations.

  14. Another aspect is that pastoralists don’t need as many slaves or workforce to begin with. Its rather about feeding your people from the limited resources you have. More people don’t mean a proportionally higher productivity. Even on the contrary, if you have a high natural growth, which the successful, expanding pastoralists must have had, you had troubles feeding your own people and children with the resources available.
    So slaves would have been much more of a burden than a profit. That was different in those areas where the agrarian productivity was high enough for producing large tributs for the conquerors.
    In Northern Europe crop farming was not very productive for quite some time, while in Southern Europe and South Asia things were different. So even before the introduction of money, keeping local males alive seems to be more profitable there while cost ineffective in the North. But still things were much more complicated and difficult to exploit before the dawn of more advanced administrations and money.

  15. Before Columbus, east of the Mississippi, the aboriginal population consisted of farmers: the Abenaki in what is now New England and the Canadian maritimes, the northeastern Iroquois – also the Hurons – further south, the Choctaw. My understanding is that they have contributed little, genetically, to the contemporary populations of these regions. Perhaps they were not sufficiently dense agglomerations of agriculturalists, though, if memory serves, Mann in 1491 indicated that the population was quite large and relatively dense by pre-modern standards before European contact. Was it really less dense than that of the European or S. Asian agriculturalists under consideration?

  16. They had significant losses because of diseases which disrupted everything for them. Also, the newcomers were, unlike the Iberians, close knit communities of religious people which came en masse, were for the most part complete farmers with a wider range of crops and animals and didn’t need the Indians – with the exception of some learning like potatoes.
    But overall, the European cultural package was much more complete (domesticated animals in particular!) and the Indians culturally worlds apart. I think the cultural, also biological, differences were much, much smaller in the European context, even in South Asia at the time in question, in comparison.
    Europeans had all the advantages and could take what they needed culturally without having to integrate the Indians in large numbers, which was a very difficult when tried anyway.
    In Meso- and Southern America some conquerors had a completely different view on things, were not primarily farming communities and could overtake more complex, state or state-like organisations. I don’t think such a “head swap” was the goal nor even possible in Northern America.

  17. Another crucial point I forgot is that the Protestant settlers in Northern America came with their wives largely and both female slaves for reproduction, kept women or more legal wives were no option for them. So they had their close knit communities of religiously defined groups which transferred completely.
    In those areas where primarily males went, mixture was the rule and in some cases even new tribal like groups based on a male founder came into existence in America.

  18. Didn’t the archaeology already tell us that farming collapsed?

    This doesn’t seem like much of an explanation to me. In the short term, killing the farmers leads to demographic replacement, but what about the long term? The ultimate winners were farmers, so farmers were capable of displacing herders.

    I would have guessed that the original farmers were displaced to the margins, but developed ways of pushing back and ultimately displaced in turn. So I would have guessed that in the long term this was not a good strategy for survival of the pastoral genes. I would have guessed that taking over the top of the farmer hierarchy and imposing downward mobility would be more effective at spreading genes, although it didn’t work in some examples, like the Hungarians. It worked fairly well for the Mongols, but not as well as the opposite worked for the Yamna.

    So something is wrong with my model. We can conclude that displacement was more effective than assimilation. But that doesn’t answer questions like “why” or “how.”

  19. One thing to keep in mind is that, prior to the formation of (settled) states, wars of extermination (or conquests, for that matter) were rare and unlikely. Standing/regular armies did not exist, and tribal warfare tended to be timid, if vicious, little affairs, mostly involving raids and counter-raids over food, goods, and women.

    One of the main advantages, among many others, that semi-nomadic pastoralists had over farmers in warfare of this type was that they could engage in warring year-around, which meant the ability to raid constantly and make life miserable for settled peoples and, in effect, depopulate agricultural areas by making them dangerous places to inhabit.

    Later barbarians were able to do this against the likes of Rome and China and make the border areas difficult to settle and garrison (the antidote to which was a periodic, massive mobilization by these states and launching of punitive expeditions, at times of exterminatory nature). Now imagine earlier farming communities without the cohesive centralized governments and the surplus resources (read “standing armies”) they made possible.

    I suspect that once the pastoralists arrived on the periphery, the lives of the border zone agriculturalists became perilous rather quickly and led to the depopulation of those areas. Rinse and repeat over generations and massive areas would be emptied of agriculturalists.

    Indeed, it’s even possible that the arrival of the pastoralists created the impetus that led to larger state-formation among the surviving agriculturalists, perhaps because they began to realize that centralization and the supporting of standing armies were their only salvation against the predations of these roving tribes of raiders, thieves, and bandits.

  20. What you describe happened in the Near East, where the farmer’s states were more sophisticated, yet even there in the long run pastoralist warrior overtook the states. This final step was done by Semites and Indo-Europeans alike.

    As for this:
    “One thing to keep in mind is that, prior to the formation of (settled) states, wars of extermination (or conquests, for that matter) were rare and unlikely.”

    That is not true, especially not for higher cultured farmers and pastoralists even less so. Even on the contrary, in pre-state cultural environments large scale exterminations were more common. Because states with their administration and most of the time money were a good target to exploit and more likely to survive a conquest, even if it was in a somewhat different shape and with new rulers on top of it. I think people constantly underestimate the organisational height of the Chalcolithic and Bronze Age people, just because they lacked writing. They were not like Papuans or the like, and even those exterminated close kin (!) on a regular basis actually. Not talking about foreigners. Like Chinese communities being killed off completely by Melanesians at various times in the historical period.

    The idea that humans didn’t fight in a Darwinian fashion as communities vs. each other is an ideologically based, flawed view. They always did. Whenever we have reliable data, they warred each other to the death. Unless the other side is too strong for being killed off, then it might look like a ritual thing or even become one over time. But especially if there are no cultural ties between the war parties, genocidal wars could be only prevented by both sides being strong enough or one side completely submitting to the other.
    The Mongol armies spared those cities which submitted, even if they completely destroyed those which resisted.

    Also the farmer economies which appeared after the Indo-European conquest in the Bronze Age were different from those before in Western Europe. They relied more on domesticated animals and their products than the less successful ones before.
    But the Indo-Europeans were no pure pastoralists to begin with and we see the transition back to more crop planting in the Unetice culture for example.
    But that was a step taken by the Indo-Europeans themselves, which, at that time, had incorporated more farmers blood indeed. Yet the whole macroregion was Indo-Europeanised and after such a large scale conquest, there is no coming back. A coming back was only possible for super successful individual male lineages which made it in the new cultural frame. That’s what we might see in the cases of I1 and E-V13 in particular. They were integrated into the Indo-European society and spread from within.

  21. Barbarism loves genocide and ethnic cleansing, whether it be in the Bronze Age or colonial/Old West times. They want that land, they get rid of who lives there.

  22. Even on the contrary, in pre-state cultural environments large scale exterminations were more common.

    On what do you base this?

    I think people constantly underestimate the organisational height of the Chalcolithic and Bronze Age people, just because they lacked writing. They were not like Papuans or the like, and even those exterminated close kin (!) on a regular basis actually.

    Actually, they were rather like “Papuans,” who may have harbored the desire to exterminate their enemies, but until rather recently have lacked the means to do so. More on this below.

    The idea that humans didn’t fight in a Darwinian fashion as communities vs. each other is an ideologically based, flawed view. They always did. Whenever we have reliable data, they warred each other to the death.

    Again, this is just an assertion without evidence. Indeed, if you watch animals fight “in a Darwinian fashion,” they don’t often fight to the death. They fight until one gives up and yields dominance to the victor.

    Unless the other side is too strong for being killed off, then it might look like a ritual thing or even become one over time. But especially if there are no cultural ties between the war parties, genocidal wars could be only prevented by both sides being strong enough or one side completely submitting to the other.

    Pre-historic warfare was largely one of raids and counter-raids. This makes sense given that people who were not imbued with an ideology of the state were opportunistic fighters (more robbers and bandits, really), meaning they picked on soft, unguarded targets and avoided pitched battles. Indeed, there was likely no such a thing as a “pitched battle” in which one side never gave up and fought to the death (not as a cornered individual, but as an organized, coherent group).

    Note the following points. First, intra-pastoral conflicts were unlikely to be ones of extermination, because pastoralists possessed moveable properties and means of production. Losers may be evicted from desirable pasturage, but they could flee rapidly and avoid annihilation, if only to start over elsewhere. In point of fact, this was one of the mechanisms of pastoralist expansion, a major one in fact (Central Asia is replete with groups being expelled by victors, only to conquer new lands for themselves elsewhere in Europe and East Asia).

    In intra-(subsistence) farmer conflicts, the scale of wars was also limited by the fact that subsistence farmers lacked the expertise, weapons, and time necessary to wage extensive warfare. They could at best campaign seasonally and pretty ineffectively at that. And in an era when the population density of a region was very low and land was plentiful (if not of always the best quality), those farmers who lost their lands in wars could conceivably re-start elsewhere (rather like the mythical warriors of Troy fleeing to new lands)

    In a farmer-pastoralist conflict, the farmer was at a severe disadvantage given that he could only campaign seasonally with very limited numbers. The pastoralist, if outnumbered, could always flee, only to return and harass the farmer later. Extermination would have been virtually impossible.

    Pastoralists, of course, could overwhelm farmers in a set-piece battle, but unlikely to have done so given that they preferred a war of (less costly and less dangerous) attrition, i.e. constant raids. And also because they lacked the means to attack directly barricaded or fortified towns, especially those on hilltops. Why take the risk of a pitched battle and fighting into a town when they could harass the farmers, make their lives dangerous, and gradually depopulate the area?

    The Mongol armies spared those cities which submitted, even if they completely destroyed those which resisted.

    Mongols were not early pastoralists. They benefited from long-term contacts with civilized states. They had developed a powerful, unifying ideology of a state centered around Genghis Khan and his Golden Family and they had the organizational and technical means to annihilate their enemies on the battlefields AND storm their cities and towns.

    Furthermore, the Mongol ability to exterminate urban populations had one great assistant – paradoxically enough, the high density and complexity of the cities, which required extensive surrounding farmlands and support infrastructure to support the residents. In such an environment, unlike in a much earlier era of simple, minimal “infrastructure” to maintain small farming populations, losers who were expelled from their homelands were likely to experience mass die-offs. As an extreme example, imagine New York City cut off from grocery supplies.

    Cities tend to experience dramatic population growths during years of prosperity and peace, and tend to develop complicated infrastructure designed to maximize the number of people who could be supported. However, such complexity and maximalist population expansion tends to have a potentially lethal flaw – fragility in the face of a crisis.


Comments are closed.