Plagues and Peoples

[stoutfellow]

This is going to be rather long and geeky, I'm afraid.

I'm currently reading Who We Are and How We Got Here, by David Reich. It's about discoveries, over the last twenty years or so, concerning the DNA of past populations of humans around the world. It's a fascinating book, and I may give it a review later on, but the chapter I just completed triggered a series of thoughts.

About 5000 years ago, a population originating in the East European steppe, the Yamnaya culture, expanded hugely into Europe, displacing a resident farming population. Part of their success can be attributed to the discovery of the wheel; charioteers made a formidable fighting force against a sedentary, wheel-less population, But there may have been a second factor. Analysis of the teeth of the indigenous farmers, before the Yamnaya expansion, shows no indication of exposure to Yersinia pestis - the plague bacterium - but afterward, both populations show traces of its presence. It is entirely possible that the farmers were decimated by plague, easing the way for the intruders.

What follows is my own reasoning, not Reich's, and if anyone wants to rebut, I'll be glad to read it.

A digression: for a long time, the received wisdom was that disease pathogens undergo evolutionary pressure towards reduced virulence - a host who survives for a long time after infection is more likely to infect others. A generation or so ago, Paul Ewald pointed out that this applies only to diseases spread by personal contact, not to vector-borne diseases; for those, a host does not need to be out and about to spread the disease. (His Evolution of Infectious Disease is also fascinating.)

The reason this comes to mind is this: the Yersinia DNA recovered from the farmer and Yamnaya populations lacks the genes that permit transmission by fleas. The plague outbreak (if such it was) was pneumonic plague - which is directly transmitted, not vector-borne. At that time, then, plague would have been subject to pressure towards reduced virulence, which would allow for mutual adaptation, given enough time, between the Yamnaya and the pathogen, while the farming population would have faced a virgin-field epidemic, rather like the Native Americans at the time of the Columbian Exchange.

Most of the genome of modern Europeans is derived from the Yamnaya; there is relatively little trace of the ancient farmers.

Comments

[jsburbidge]

I've seen speculation that the pre-IE population of Europe, culturally and linguistically, might have resembled New Guinea - fragmented into little micro-regions (low mobility, long time for dialect divergence). Not a context in which you get a lot of development of immunity. On the other hand, the Yamnaya were pastoralists, which also tends towards low levels of disease; so I'm not sure they would have had the sort of immunity build up that urban populations tend to get by selection. Those populations were in India, the Middle East, and Egypt at that time: are there any traces of plague involved in the Indo-Hittite movement into Asia Minor?

All of which goes to say that although the Yamnaya might have picked up Yersinia pestis from an endemic area -the steppe area is, if I recall correctly, a possible source for plague later on in history - I'm that they would have had a big advantage over the Europeans. I would expect both populations to have had a similar die-off rate.

[stoutfellow]

Your points are well-taken. However, though the Yamnaya were pastoralists, they emerged from an agricultural population after, and probably as a result of, the increased mobility due to the invention of the wheel. (Cf. the shift from agriculture to nomadism in the Plains Indians after the introduction of horses.) Prior to that time, most of the steppe population lived in river valleys, and - conceivably - the ancestors of the Yamnaya acquired partial immunity then.

Speculation, of course. Reich devotes only a single paragraph to the possibility.

[stoutfellow]

I managed to track down the article that Reich mentions (S. Rasmussen et al., "Early Divergent Strains of Yersinia pestis in Eurasia 5,000 Years Ago", Cell 163 (2015): 571-82. At the end of the article, the authors write "These early plagues may have been responsible for the suggested population declines in the late 4th millennium BC and the early 3rd millennium BC" and "In light of our findings, it is plausible that plague outbreaks could have facilitated—or have been facilitated by—these highly dynamic demographic events" (meaning the large-scale population movements of the Bronze Age).

The underlying problem is the fact that the lower-density Yamnaya population was able to largely replace - not just conquer - the higher-density farming population. The hypothesis concerning immunity offers an explanation; if the die-off rates had been similar, the explanation fails.

[graydon]

Farming populations are not inherently resilient. They're inherently vulnerable until you've got some concept of a state and a pretty large geographic area. (This is in large part the Roman innovation, a generalized concept of the state.) Once you've got that, fighting any of them is fighting all of them, at least in principle, and there are inevitably a lot of them.

Prior to that point when you've got a fragmented population of near-subsistence agriculturalists, the low-density/high-mobility pastoralists come in and burn the food stores in the early winter. Nobody really has surplus and getting to an area with unburnt food is hard anyway. If the pastoralists have wool technology and move comfortably in the early winter and the agriculturalists do not, well. It's even easier.

[stoutfellow]

Perhaps so; but the question was still being argued among archeologists as recently as 2013. (Peter Bellwood, First migrants: Ancient Migration in Global Perspective. I haven't read it - it's mentioned in Reich's endnotes - so I can't evaluate its claims, but the issue was up in the air.)

[graydon]

I might think the European specialists would benefit from work that was done in the 70s and 80s about the Iroquois expansion into territory held by "Hurons" -- Algonquian speakers -- and the Iroquois had an organizational advantage, not a mobility one or a toolkit one. All three would presumably work better still!

(Though I would be very curious if there's anything more recent about better access to wool blankets!)

And yes, the whole "what, you mean the Yamnaya hypothesis is plausible?" thing is still up in the air but I think that's in large part because it's the true heir of Gimbutas' Kurgan hypothesis and has immense political baggage.

[jsburbidge]

Given that the main alternative to the Kurgan hypothesis, the placing of PIE in the Anatolian highlands, has major problems with timing and the distribution of later languages, my sense was that the Kurgan / Yamnaya hypothesis is pretty widely accepted these days even prior to this sort of evidence.

But, yeah,when you come in with fire and sword on an essentially immobile population, probably moving your own population in behind the front wave in wagons, the net effect tends towards overwhelming the prior population, especially when this is the first time this has ever happened. When you add in status effects over the next few generations,the proportions can only increase.

And, if course, this isn't like the Huns,sweeping over Europe in one lifetime. The splintering of PIE in Europe into Celtic, Italic, Germanic, Hellenic, Lithuanian et al. argues for a gradual inundation, like the tide coming in.

(Edited to reverse autocorrect errors.)

[graydon]

My understanding is that the initial Yamnaya advance is regarded as quick; a couple-three hundred years to from the steppes to the Atlantic. The language splits come with the transition to the bronze age and if not new populations then definitely new political centres and cultural traditions. (But are fundamentally lurking in misty pre-history and there's an awful lot we don't know about back there, e.g. the remnants of the big neolithic battle where everyone would have thought there couldn't be enough people for a battle.)

[jsburbidge]

"Quick" on a Huns parallel would be one generation. Two to three hundred years gives plenty of time for a by-stages advance with subtly different groups of people being left in different areas.

Which is entirely reasonable when people don't know where they're going (the only people who might tell you speak and unrelated language) and you're, collectively, interested mainly in spreading out so as many chieftan-level males can have as much territory as possible to set up as small wanax-es. Each generation's younger warriors keep moving forward with a subset of the PIE population until they hit the sea.

(PIE culture seems definitely to have been more male-centred, top-down-authority, siy-god oriented than what they replaced. It makes for a mindset suited for conquest.)

[graydon]

It helps a lot if there already waggon-roads, instead of donkey tracks! (Well, and a pile of other stuff; people managed to do an amazing amount of stuff with the neolithic toolkit, but they didn't do it as fast.)

I don't disagree at all with that pattern of movement; my (questionable, non-specialist) understanding is that the different words for things show up later, though.

[jsburbidge]

Maybe not even donkey tracks, though the larger the track the more tempting the possibility of what's at the other end seems. One problem with the wheel is that once you leave flat open country your problems multiply

Language drift takes time, but it builds on either or both of relative isolation (boundaries becoming isoglosses) and divergent values existing prior to the division between dialects. (That is, natural or political boundaries.) My guess (for which there is no evidence, really, other than plausibility, so it's a just so story) is that each generation of the movement to the West left a set of people settled behind them. Where there were no major boundaries and easy communication a common identity was reinforced; where there was the Rhine or a major forest or the Alps splitting must have started almost as soon as settlement. (There's certainly later evidence for the Rhine as a major isogloss marker. (And, very roughly, the Rhine, the Alps, the Balkans and the Adriatic look like dividers between Hellenic, Italic, Celtic/Gallic and Germanic.)

The common vocabulary gives us people with sheep, horses, wheels, carts, a certain distribution of male/female roles (*duhitar, "the little milker", becomes daughter), a sky-god religion. After that...well, we can see the process of turning metaphor into dead metaphor happening in a way we can examine in Hometic Greek, as well as borrowing to enrich the language ("tyrannos" seems to have been Lydian for "king"). (Interestingly, I can't think of many words for chieftan or king which have a common root - wanax, basileus, cyningas, rex.)

Conquest and politics are fast, language change is slow, and human memory over even a generation is chancy. (If you had asked an English person in about 1860 which European nationality he felt closest to, he would have said "German". If you asked the same thing sixty years later, not only would you get a different answer, but your later interlocutor would think of his view as the way it had always been.)