search this blog

Thursday, December 6, 2018

Europe's ancient proto-cities may have been ravaged by the plague


The Cucuteni-Trypillia culture of the Eneolithic Balkans and Eastern Europe is best known for its mega-settlements or proto-cities, each one featuring hundreds of homes, temples and other structures, and likely to have been inhabited by as many as 20,000 people. But from around 3,400 BC these mega-settlements were no longer being built, and a few hundred years later the Cucuteni-Trypillia culture vanished.

Two main explanations have been given for its rather swift demise: violent invasions by steppe pastoralists from the east and/or a massive out-migration by its people as a result of environmental impacts from rapid climate change (see here). However, these theories have failed to gain wide acceptance due to a lack of hard evidence in their support.

Now, another potential explanation is being offered, and it is supported by hard evidence. According to Rascovan et al., the plague may have been a key factor in the decline of not only the Cucuteni-Trypillia culture, but much of Neolithic Europe (see here). From the paper, emphasis is mine...

Between 5,000 and 6,000 years ago, many Neolithic societies declined throughout western Eurasia due to a combination of factors that are still largely debated. Here, we report the discovery and genome reconstruction of Yersinia pestis, the etiological agent of plague, in Neolithic farmers in Sweden, pre-dating and basal to all modern and ancient known strains of this pathogen. We investigated the history of this strain by combining phylogenetic and molecular clock analyses of the bacterial genome, detailed archaeological information, and genomic analyses from infected individuals and hundreds of ancient human samples across Eurasia. These analyses revealed that multiple and independent lineages of Y. pestis branched and expanded across Eurasia during the Neolithic decline, spreading most likely through early trade networks rather than massive human migrations. Our results are consistent with the existence of a prehistoric plague pandemic that likely contributed to the decay of Neolithic populations in Europe.

...

In this work, we report the discovery of plague infecting Neolithic farmers in Scandinavia, which not only pre-dates all known cases of plague, but is also basal to all known modern and ancient strains of Y. pestis. We identified a remarkable overlap between the estimated radiation times of early lineages of Y. pestis, toward Europe and the Eurasian Steppe, and the collapse of Trypillia mega-settlements in the Balkans/Eastern Europe.


Citation...

Rascovan et al., Emergence and Spread of Basal Lineages of Yersinia pestis during the Neolithic Decline, Cell (2019), https://doi.org/10.1016/j.cell.2018.11.005

See also...

Migration of the Bell Beakers—but not from Iberia (Olalde et al. 2018)

Late PIE ground zero now obvious; location of PIE homeland still uncertain, but...

"The Homeland: In the footprints of the early Indo-Europeans" time map

192 comments:

Davidski said...

I wonder if the Burned house horizon of Neolithic/Eneolithic Europe was a direct result of the plague?

I mean, why would you burn down settlements, including even huge proto-cities, regularly and rebuild them if you weren't trying to get rid of something at all costs and start over?

Burned house horizon

Them meee said...


Woah. You can’t describe how monumental this is.

Though we’ve known for a while that the plague ravaged Neolithic Europe to an ugly degree, the confirmation of strong links between the timeline of its spread and Cucuteni-Trypillia’s collapse, one of archaeology’s biggest misteries, and one that no earlier theory was truly able to decipher, is truly something else.

What else can I say?

AWood said...

Did the plague not hit Italy and Spain or something? How is it that there was a bounceback of female EEF lineages and ancestry, yet the bulk of the post-Neolithic male lineages seem to be derived from the Eurasian steppes?

Toby_P said...

Hhmm. I'm not sure the dates add up. The lineage shifts in western Europe occurred with Bell Beaker, and thats almost a thousand years after Cucuteni ends. It seems like they're trying to associate a series of events to one cause.

BTW anyone notice fig 5A & 5C. Are they suggesting that CHG expanded into the south steppe between 8000 and 5000 BC (''expansion of populations from Iran/Caucasus to Eursian steppe) ? Are these tjhe same Danish consortium working on Shuvaleri etc ?

Ric Hern said...

My guess is that those who survived were mostly herders in Mountainous areas. So any bounce back would have occurred from the Mountains...

Davidski said...

@Toby_P

I think the idea is that the plague caused a demographic collapse and a downward trajectory in demography among the European farmers that they could not get out of, and pastoralist groups from the steppes and their descendants took advantage of that over a long period of time.

In other words, the rise of R1b in Western Europe need not overlap with the major prehistoric plague pandemics.

Davidski said...

By the way, these Neolithic plague pandemics might also explain the bounce back in hunter-gatherer ancestry in much of Europe, perhaps because the farmer groups that lived further away from the main settlements and maybe had more diversified subsistence strategies survived better and were able to then occupy some of the empty land even before the steppe pastoralists showed up.

Ric Hern said...

I wonder if the Pile Dwellings were a result of trying to escape rodents which would have been attracted to grain and other foodstuffs ? Rodents are the main carriers of the fleas which carry the Plague....An extensive herding system had the least chance of attracting rodents.

Bogdan said...

@Davidski:

Quite possibly your best blog post. This is the big one folks....

Bob Floy said...

@Davidski

"I wonder if the Burned house horizon of Neolithic/Eneolithic Europe was a direct result of the plague?

I mean, why would you burn down settlements, including even huge proto-cities, regularly and rebuild them if you weren't trying to get rid of something at all costs and start over?"

FYI, Cucuteni's Anatolian ancestors were doing something similar at Catal Hayuk, and I think at a few other sites.

Folker said...

Burning of CT settlements was very likely linked with religion. The frequenty seems always the same periodicity (around 60/80 years). Unlikely to be linked to plague. Mind that the largests settlements could have ben far more than 20 000 (Maydanets could have been home to around 46 000 at its peak). Another remark: the assumption that those settlements where egalitarian with no mega structures is obsolete. Some mega structures have been found, meaning at least some kind of organization.

Fanty said...

"Burning of CT settlements was very likely linked with religion."

Isnt BUILDING "megastructures" not usualy linked to religion? Not burning them down. Except if its the megastructures build for a different religion of course. ;-)

Bob Floy said...

@Folker

"Burning of CT settlements was very likely linked with religion. The frequenty seems always the same periodicity (around 60/80 years)."

I'm inclined to agree. And like the habit of building megalithic structures, it's a practice that seems to have been brought from the farmers' ancestral homeland. Of course none of this rules out the idea that the plague was a factor in the decline of neolithic Europe, it might well have.

Matt said...

Good post.

Having skimmed the paper, I guess I would add a couple of question marks that:

- Gokhem4 is pretty late in the day for Funnelbeaker, at 2900 BCE. That's practically overlapping with early Corded Ware, so I don't know if it totally does for yersinia pestis being steppe origin.

The argument is that there is inferred branching between 4000 BCE - 3000 BCE, but I don't know if that can be shown that they are necessarily evolving in different populations for most of that time. I mean, most likely they would be, because otherwise they'd be in close direct competition with each other, but maybe not.

Even if these plague strains did evolve in separate populations though, I'm not sure the evidence that they evolved in CT is necessarily persuasive.

They've just made the argument that practically any level of population contact is sufficient for plague to evolve.

- The argument is the Gokhem strain is virulent enough - "At the genomic level, these strains contain the plasminogen activator gene that is sufficient to cause pneumonic plague—the deadliest form of historic and modern plague (Zimbler et al., 2015)". But surely the divergent then dominant steppe strains must have been more virulent or their clade would not be the dominant surviving clade? Rather some EEF version would be.

So there may still be some biological advantage from having a bit of preadaptation to a nastier version of plague. I don't then know if we can totally but away the idea of disease facilitating population movements in the manner previously theorised.

Also doesn't seem to favour the idea that the nastiest version evolves necessarily in the densest farmer populations (which itself throws a bit of uncertainty on the idea of most virulent plague evolving in CT).

Ric Hern said...

@ Folker

Are you referring to something like a townhall ? Well usually some practices were incorporated into religion in order for people to take them serious. It became the recipe for survival in certain circumstances....

George said...

Hi,

“The Black Death epidemic had run its course by the early 1350s, but the plague reappeared every few generations for centuries.” This fits the ~60 year cyclic burning of the C-T villages.
Quote from: The Air of History (Part II) Medicine in the Middle Ages, the Black Death Section: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3573364/

Fire was used to contain the Plague in 1900 Honolulu. “Plague and Fire: Battling Black Death and the 1900 Burning of Honolulu's Chinatown”, By James C. Mohr (2004).

There are of course other possible reasons. Deliberate house-burning in the prehistory of Central and Eastern Europe by John Chapman http://dro.dur.ac.uk/5987/1/5987.pdf?DDD6

“… the ritual burning of a Neolithic house in the middle or end of its life-history is a good example of domithanasia” from Weaving house life and death into places: a blueprint for a hypermedia narrative by Ruth Tringham. http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.461.1359&rep=rep1&type=pdf


Davidski said...

@George

This fits the ~60 year cyclic burning of the C-T villages.

Yep, I don't see any contradiction between the burning of the settlements being a response to a fairly regular plague pandemic cycle and a religious ritual.

I think that these sorts of major social upheavals like regular plague attacks would have had a profound effect on the beliefs, superstitions and religion of these people.

Them meee said...

@Toby_P

“BTW anyone notice fig 5A & 5C. Are they suggesting that CHG expanded into the south steppe between 8000 and 5000 BC (''expansion of populations from Iran/Caucasus to Eursian steppe) ?”

That’s interesting. Wonder how they came to that conclusion.

That is, if they have any new samples showing that, then it looks like Davidski was right on track once again.

It is best to wait though.

Leron said...

Indo-European since its inception has always been a language of trade and commerce. Cucuteni-Trypillia, the godfather of IE culture and language, met it's demise at the hands of the plague because they preferred inhumation for their burials. While their western steppe confederates survived in greater numbers because they opted for cremations instead. The early IE tribes absorbed as much as was necessary, including genes, from this advanced society and carried on the tradition of commerce after the epidemic subsided. Following more lucrative paths, they were led south and eventually came to Anatolia.

It was not wild and rambunctious steppe men on horses that spread Indo-European. This was the job of more enterprising people like the Hittites. That made Old Assyrian merchants cross Cappadocian mountain ranges, a feat later Neo-Assyrian armies still dreaded, just to make business with them. And it even led their army to touch Babylonian soil a thousand years before Alexander.

Through commercial prowess and just a fair amount of conquest, IE became the lingua franca over its other sister languages to the north that were cruder and slightly more reminiscent of Uralic (although they separated some thousands of years before).

Them meee said...

@Leron

“It was not wild and rambunctious steppe men on horses that spread Indo-European. This was the job of more enterprising people like the Hittites.”

Then how do you explain Indo-Iranian? Doesn’t look like just commerce did it. They were among the most “wild and rambunctious” steppe peoples, but yeah, it was totally spread by commerce only, and the Scythian were only known as merchants and not warlike nomads. Sure.

Davidski said...

@Leron

I guess you missed the part where the cultures of Eneolithic Europe west of the steppe collapsed, and were replaced by those derived from the steppe.

In other words, Cucuteni-Trypillia wasn't Indo-European. It vanished and was replaced by Indo-Europeans.

Leron said...

Them meee:

Indo-Iranian came very late in the development of IE, so there's no need to attribute them a quintessential character of IE as a whole. Perhaps they are the most romanticized in this field. Archaeology and history would let us know early IE people were best adapted to thrive in settled regions and in cities, interspersed by mountains to act as natural barriers. The open steppe allows for mobility but there are other groups that by nature seem to have been better at it, and hence why Iranic languages exploded in the south but barely few remain in the north. You should also be aware that "Scythian" was not exclusive to Iranians but could represent many distinct groups unless they were overtly Eastern in appearance. Even a group of Jews today are known as "Scythian" (Ashkenazi) which should let you know how loose has always been.

Davidski:

Collapse does not mean total extinction. The few that remained were enough to carry the C-T legacy forward. I also believe the survivors had significant steppe ancestry (although not evenly distributed) and spoke the proto-IE language that replaced the Kartvelian-like C-T language. This is where IE picked up those Kartvelian words that linguists have pointed out, rather than all the way in Georgia and surrounding areas. And it would be the western steppe folk that were primarily involved in the formation of IE. Those further east were later to receive IE culture, although adapted to fit their environment.

Them meee said...

@Leron

“Archaeology and history would let us know early IE people were best adapted to thrive in settled regions and in cities, interspersed by mountains to act as natural barriers.”

Which is why Indo-Europeans conquered half the world, from Ireland to India, in a few short pulses and their speech became the most widespread in the world.

Because they were settled urbanites, and not at all pastoralist nomads, let alone good at it.

Right?

Toby_P said...

@ Them meee

Yes exactly I was just wondering if they now have some sample to based it on, or are still jsut speculating ...

Ric Hern said...

Indo-Europeans are good adoptors and adaptors. That is I think what gave them the edge...

Bob Floy said...

@Leron

"Cucuteni-Trypillia, the godfather of IE culture and language..."

Huh?

Leron said...

Them meee:

To put it simply, most IE were looking for new homes to settle and new venues of trading to feed and expand those settlements. They were not Mongols. Finding grazing land was not their main objective. Although they were proud of their horses, so were Assyrians and other urban civilizations. The Rig Veda threw light into a relatively short span in time when cities were mostly not in their possession but quickly changed perspective when they got them. The Greeks and Hittites looked down on those who just wandered around without a large city to call their own.

Ric Hern said...

Leron did you in any way look at the Genetic evidence of Ancient samples ?

Bastian Barx said...

Getting seriously tired of Leron's unfounded extrapolations. Too much imagination, and too little intellectual honesty, is what it's about.

Davidski said...

Leron obviously hasn't heard of ancient DNA yet.

Them meee said...

Doesn’t seem terribly knowledgeable in archaeology either. Don’t know how he came to the conclusion the Indo-Europeans were city dwellers and not at all steppe nomads, who weren’t at all looking for new grazing lands, when the Greeks even described the arrival of the Mycenaeans or a similar group happening this way.

Also I can’t see how C-T could have spoken a Kartvelian-like language.

PF said...

By the way, these Neolithic plague pandemics might also explain the bounce back in hunter-gatherer ancestry in much of Europe, perhaps because the farmer groups that lived further away from the main settlements and maybe had more diversified subsistence strategies survived better and were able to then occupy some of the empty land even before the steppe pastoralists showed up.

This is a pretty great insight. Looking back, the HG resurgence was always a clue that farmer societies suffered some sort of internal collapse before the the steppe migrants came and mopped things up.

At the same time, this doesn't mean that there wasn't violent population replacement by the steppe migrants. It's difficult to write off facts like G2a disproportionately surviving in mountainous areas and almost nowhere else, the sex bias in uniparentals, etc.

epoch said...

I recall that La Brana 1 already had derived adaptive variants on loci involved in immune responses. I never saw anything similar checked in other papers, but it would be interesting to see if farmers had similar variants.

"For the remaining loci, La Braña 1 displayed the derived, putatively adaptive variants in five cases, including three genes, PTX4, UHRF1BP1, and GPATCH1, involved in the immune system (Table 1 and Extended Data Fig. 8). The latter is associated with the risk of bacterial infection."

Dmytro said...

A plague epidemic sounds like a plausible cause for the complete disappearance of classical CT in the areas which hosted the "megalopolises" (Majdanets et sim.) This seems to have occurred in the period subsequent to the last known "ritual town burning" shortly after ca. 3300 BCE. What is particularly interesting is that the ex-CT population which produced the "steppe" cultures of Serezliivka and Zhivotilovka neighbouring and /or mixed with PostStog and the "Late Trypilian" cultures (also steppe-mixed to a greater or lesser extent) does not seem to harbour cultural forms closely related to the "megalopolitic" Trypilians of right bank Ukraine. I suspect BTW (talking about fire) that the "disposal of the dead" in megalopolitan CT (and in much of CT actually) was some kind of above ground ritual burning which left no "inground" evidence such as we have for both cremations and inhumations in many other cultures.

George said...


For commentary on recurring pestilence see:
"Of Plagues And Prehistory" on www.gnxp.com

Carlos Aramayo said...

@Davidski,

Maybe it´s not the right place to comment this, but I wanted to let you know of the following recent "review" article:

"Haryana’s Rors brought Western flavour to the Indus Valley: New genetic study claims that the Rors came to the Indus Valley when it was flourishing during the Bronze Age and inducted West Eurasian genetic ancestry"

https://tinyurl.com/ycqb7bss

Maybe the academic publication mentioned in this review can add something important.

JuanRivera said...

The Indo-Europeans seem to be in many ways like the Mongols. Horses were important to them, they aided in the diffusion of plague, and they expanded over a huge area. The difference is that they left a significant impact where they went. Steppe ancestry shows from the Altai, NW China, Bangladesh and Sri Lanka to Ireland, Iberia and the mediterranean islands (except Sardinia). It shows up in BA Europe including Ireland and Iberia (the Bell Beaker samples with no steppe ancestry are highlighted as no_steppe for a reason), in Chalcholithic Armenia and NW Iran, in Anatolia, in the Altai and even Lake Baikal and Mongolia to a minor extent, in the Tarim Basin, in BA Central Asia, and Iron Age Iran, Swat Valley and likely other Indian samples. So, a better analogy for Indo-Europeans would be all-terrain amphibious tanks.

Matt said...

Off topic from this slightly but was listening to Reich's latest video presentation, which is pretty non-novel generally, but was surprised at 48:00 to hear "(The Yamnaya) were very successful. They expanded from where they initially uh, originated, all the way from Hungary in Central Europe, all the way to the Altai Mountains, on the boundary of Mongolia" - https://www.youtube.com/watch?v=990052wQywM. I guess that must be a strange phrasing, rather than some kind of statement that Yamnaya in Hungary actually ancestral to other groups?

Davidski said...

@Matt

I'm quite sure that Hungarian Yamnaya isn't ancestral to Yamnaya.

Interestingly, though, the oldest Yamnaya sites are located in the west in Ukraine and in the east in Samara, rather than in between.

Not sure what that means?

Matt said...

Yep, doesn't make a lot of sense to me, wanted to check that there wasn't anything at all that could possibly support it so probably just some strange language.

Not sure about distribution of Yamnaya sites in western Ukraine and east of Samara - intermediate land not so good for their subsistence method?

Bob Floy said...

@Davidski

"Interestingly, though, the oldest Yamnaya sites are located in the west in Ukraine and in the east in Samara, rather than in between.

Not sure what that means?"

I'm not sure what it means exactly either, but I am pretty sure that L51 came from those Ukrainian Yamnaya communities.

Them meee said...

@Davidski

Maybe a very rapid expansion? Would also explain some sort of mad rush eastwards that became Afanasievo.

Or am I talking nonsense?

Philippe said...

"all the way from Hungary in Central Europe, all the way to the Altai Mountains, on the boundary of Mongolia"

Sounds like he's describing the two extremes of the land they inhabited, from Hungary to the Altai.

JuanRivera said...

So, it's becoming clearer and clearer that IE came from the steppe. The "Iran_N"-like ancestry detected in the mediterranean is clearly non-IE, being associated with Minoan , Nuragic, Iberian, and possibly Etruscan. In the rest of its, it was associated with Sumerian, Hurro-Urartian (which is one of the substrates of Armenian, as well as Anatolian languages), Elamite, and possibly Dravidian. Paleogenetics, including that of Plague strains, seem to indicate that the manner of IE expansion over Eurasia was eerily (or disturbingly) similar to that to that of Iberians in Latin America, with a similar end result.

Them meee said...

@JuanRivera

“So, it's becoming clearer and clearer that IE came from the steppe.”

It has already been clear for the last three or so years. The biggest difference is this:

https://eurogenes.blogspot.com/2018/10/the-resistance-crumbles.html

JuanRivera said...

Ultimately and likely inexorably, the steppe hypothesis is going to become the definite theory, probably at the level of those in the natural sciences. The Anatolian hypothesis collapsed, so it did out-of-india, and the correlation of relatively high levels of Iran_N and CHG with non-IE languages, both historical and present, and other things, are going to collapse the Armenian hypothesis.

Them meee said...

It’s strange that people still talk about Iran this, Iran that, when the Wang paper is out and about. Some people can’t let go of Hajji Firuz.

The reprint may have new samples and that is something to pay attention to, but still.

Davidski said...

Well, the steppe hypothesis is not going to be totally accepted by some geneticists working on this problem as along as Hittite era samples from Anatolia supposedly lack steppe ancestry.

So where we're at now is that there's widespread agreement that late PIE came from the steppe, but the homeland of early PIE is still being debated, and might be for a while yet.

Davidski said...

By the way, I don't think it's a coincidence that image B in the figure below explicitly says that animal traction and wheeled transport spread without much human gene flow, including, and I suspect, especially onto the steppe.

Figure 5

In other words, expect more evidence soon backing this up.

Toby_P said...

Not sure about the pan-Turanist vibe people are putting out here...
Unlike Mongols and Turkics, PIEs had their own, developed Ur-Kultur. Let's not forget that, even if the northern / late or post-Late IEs which spread out relied heavily on pastoralism.

JuanRivera said...

Well, the PIE culture was unique. While it had some similarities with central and eastern steppe cultures, it had several differences such as the extensive use of farming (presumably in the rivers and forest-steppe), the existence of naval vocabulary, higher commercial activity, higher use of wheels, among others. Overall, it seems the PIEs were agropastoralist with almost equal use of its components, supplemented by hunting, gathering and fishing, and that knew both year-round houses and chariots.

Them meee said...

@Toby_P

True, but as steppe nomads they had a lot of similarities, and they were nowhere near exactly like Leron described.

Samuel Andrews said...

@Them mee,

i feel you.

When put model Hajj Chalcholithic (5000bc) you see it can't be ancestral to yamnya. Look at its ancient ancestry ratios. There's no room for such ancestry in yamnaya.

https://docs.google.com/spreadsheets/d/1_jr6EMoG_1eFm1x8hltNHW2bE5A0lW9N7XDjCU_jO_k/edit#gid=0

Haj isn't that different from modern assyrians and neighboring people. If yamnya had legitimate mesoptainian, west asian, whatever you wanna call it ancestry we would know by now.

Wang confirms yamnya has distinct paleolithic Caucasus ancestry which was not involved with early farming or any cultural advancement in west asia. Yet the idea yamnya's southern ancestry is linked to cultural advancements stubbornly remains the official narrative.

A part of the problem is they think the southern ancestry reached russia in a migration in the same way steppe ancestry reached europe in a migration.

What probably happened is one native hunter gatherer population in russia happened to carry lots of chg ancestry. Then with a new skills they learned from the south they conquered russia and ukraine (complelty replacing natives in the process) hence making chg ancestry widespread.

But actually they didn't completely replace ehgs, because we see excess ehg in EBA Europeans and lower CHG than what a purely yamnya population would give.

Ric Hern said...

Did I hear correctly ? "...Horses hitched to wagons..."

Ric Hern said...

I wonder if he suggests that Yamnaya related groups had a lot of interaction with each other over an extremely long distance after the initial spread ? Maybe the spread of certain strains of the Plague point to that ?

Ric Hern said...

The purples line 3.9 kiloyears in Figure 5 D is also interesting. The extent of Scythian Trade ?

Davidski said...

@Samuel Andrews

What probably happened is one native hunter gatherer population in russia happened to carry lots of chg ancestry. Then with a new skills they learned from the south they conquered russia and ukraine (complelty replacing natives in the process) hence making chg ancestry widespread.

More likely the population densities were higher in the south of the steppe, and so with higher mobility and more frequent contacts during the Eneolithic, the CHG ancestry spread very quickly throughout much of the Pontic-Caspian steppe until population densities evened out, and so did the level of CHG.

Davidski said...

Wang confirms yamnya has distinct paleolithic Caucasus ancestry which was not involved with early farming or any cultural advancement in west asia. Yet the idea yamnya's southern ancestry is linked to cultural advancements stubbornly remains the official narrative.

Yeah, I really hope Broad MIT/Harvard & Max Planck Jena get over this very soon and start interpreting the data more objectively.

EastPole said...

@Ric Hern

„The purples line 3.9 kiloyears in Figure 5 D is also interesting. The extent of Scythian Trade ?”

This is something I don’t understand. There is a comment to Figure 4.:
“Similarly, massive migrations of the Eurasian Steppe populations significantly changed the genetic background of Central Europe populations after the Neolithic decline, around 4.8 kya, which was later followed by a new migration from Central Europe back to the Eurasian Steppe at 4 kya.”

And on Figure 4. Central Europe is shown as:

https://i.postimg.cc/6qj2HW5L/screenshot-460.png

I am not aware of any massive migrations from Central Europe (as shown on the map) to the Eurasian Steppe at 4 kya.

Davidski said...

The authors probably think that Sintashta formed as a result of a back migration from Central Europe to the steppe.

This might be partly true, because there's evidence of contacts between the steppe and the Carpathian Basin. But most of the back migration to the steppe was from other parts of Eastern Europe, and probably especially from the forest steppe of what is now Ukraine.

Their map shows that this back migration originated in what is now Germany. Hmmm.

Matt said...

@Philippe:Sounds like he's describing the two extremes of the land they inhabited, from Hungary to the Altai.

Most likely, Reich's a good scientist, but an odd speaker, I have to say.

@Davidski:By the way, I don't think it's a coincidence that image B in the figure below explicitly says that animal traction and wheeled transport spread without much human gene flow, including, and I suspect, especially onto the steppe.

I mean, you have Funnelbeaker and Central European Neolithic axleed wheels without much evidence of change in geneflow from preceding pops (though this is hard to get at because of relatedness between EEF), simultaneous axled wheels appearing in Mesopotamia and steppe. I'm not sure they have new evidence so much as there's no signal of new migrations at the 3500 BCE dates when wheels become pretty ubiquitous. Wheel seems like something which spread without demic diffusion.

@Ric Hern: Did I hear correctly ? "...Horses hitched to wagons..."

Yeah, you did. I don't know if they actually used horses to draw wagons preferentially though, or if Reich is just simplifying matters for the audience.

Ric Hern said...

@ Matt

If true about the horses it is something significant because horses can not really be hitched to the wagons the same way as oxen. Some thought that horses needed a much more complicated hitching system than oxen and the innovation for this followed only much, much later....now I wonder if archaeologists discovered something....

Davidski said...

It seems unlikely that horses were used to for traction on the steppes until the invention of chariots by Sintashta during the Middle Bronze Age.

Yamnaya and Catacomb probably used oxen to pull their wagons.

Them meee said...

What’s with these sweeping conclusions and speculative maps? Are they hiding something (new samples or archaeological evidence) or are they just assuming too much?

Richard Rocca said...

See David Reich's new video starting at 50:25

https://www.youtube.com/watch?v=990052wQywM

Narasimhan's pre-print has 176 samples, but they now have 531 ancient genomes with 80 coming from BMAC city of Gonur, Turkmenistan.

Looks like they have narrowed down stepp ancestry there and in Pakistan to between 2000 and 1500 BC. Sounds like chariots and Rig Veda to me :D

Them meee said...

Noice, BMAC.

It also eereily coincides with the archaeological evidence. What a sad moment for any surviving OIT craziness there may be. In fact I wonder how some of the most adamant OIT supporters would react to this. Pure schadenfreude, anyone?

Very exciting and important overall.

JuanRivera said...

The figure of the spread of animal traction and wheels happens to look like the neanderthal range.

Grey said...

"Wheel seems like something which spread without demic diffusion."

maybe via family sized artisanal diffusion?


#

horses vs oxen

iirc the thing with this is horses can pull loads but until the invention of the horse collar it would press against their windpipe for some anatomical reason so they couldn't pull as *heavy* a load as oxen but they could pull light loads.

#

from reading up on the plague a while back

1) the fleas live in central asia and iirc once infected the rats die within a few weeks or so - you need a relatively fast vector to reach a target for infection while they're still alive (imo sugar ships in the middle age version)

2) iirc horses have some immunity/resistance to plague so i wonder if they could be infected but transmit it over longer distances cos immune?

steppe horse traders?

Ric Hern said...

Well a Breaststrap harness setup could have been used...

Bob Floy said...

@Them meee

I think that this is the point where OIT's pitiful holdout supporters give up entirely on trying to offer evidence, and just start accusing the scientific establishment of pushing a narrative that favors Europe(which is extra funny, because if anything that establishment is doing the opposite right now). The great thing about vague conspiracy theories is that they tend to do whatever you need them to.

Grey said...

"a Breaststrap harness setup could have been used"

yes - apparently they can't pull as much that way but they can some (according to the internet)

https://en.wikipedia.org/wiki/Horse_collar

Ric Hern said...

@ Grey

How big were the Yamnaya Wagons ? As far as I can remember some other ancient wagons did not have massively big wooden wheels. After all some four wheeled Sumerian "chariots/wagons" were pulled by Donkeys or Onagers with the primitive harnessing available...and large ponies/smallish horses certainly have more weight than donkeys.

And within that link you shared more recent studies proved that horses with a Breaststrap Harness did not pull much less weight than horses with a modern collar. It basically all depends on if you can stabilise the Breaststrap in order for it not to ride up to the horse's throat.

EastPole said...

Russell Greys’ PIE theory:

https://i.postimg.cc/yxDJph0x/screenshot-461.png

https://www.youtube.com/watch?v=pCbulVBNNUY

It follows that Slavic and Indo-Iranian languages were spoken in Armenia 8000 BC when they separated. Interesting.

Grey said...

@Ric

"four wheeled Sumerian "chariots/wagons" were pulled by Donkeys or Onagers with the primitive harnessing available"

sure, that's my point - they could pull loads just maybe not as heavy a load as oxen

https://spana.org/wp-content/uploads/2016/09/Banner.png

knowing any possible weight differences might influence what to look for archaeologically?

e.g. solid wheels vs spoked wheels? wooden side panels vs wicker?

#

also *if* it started with travois and some bright spark decided to stick some wheels on the end of the poles it might have started with two wheeled "wagons" i.e. some early chariots might not have been chariots per se but wheeled travois for carrying loads.

https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcRupLxEMU8zaRsdzB7GZEvOwHryGo8CYiT_yy88TGqRxjFg90WZ

thinking aloud

Grey said...

horses vs donkeys

random thought - for traction (pulling loads) is being lower to the ground better?

(i.e. better leverage with shorter legs?)

conflict between wanting bigger horses for riding and shorter legged ones for pulling might be relevant in some way (or not)

#

plague and horses

if you read around on bubonic plague you get a lot of comments like this

http://www.bbc.co.uk/history/british/middle_ages/blackdisease_01.shtml

"The plague generally left untouched the indigenous nomad population, because rat fleas do not like the smell of horses, with which the nomads lived in close proximity."

there's other tales of merchant caravans leaving a town and all dying leaving the horses coming back to town on their own.

imma gonna guess no one has actually ever tested and proved plague fleas "don't like the smell" of horses and imma gonna guess the actual explanation may be relevant to the question on this thread (and possibly medically useful as well).

Davidski said...

@EastPole

Russell Greys’ PIE theory:

It follows that Slavic and Indo-Iranian languages were spoken in Armenia 8000 BC when they separated. Interesting.


I wonder who was speaking them there at the time?

Definitely not our ancestors.

epoch said...

@Grey

One of the findings of the ancient pest investigation is that the older pest variants weren't transmitted by fleas. For that to work Y. Pestis must be able to survive flea intestines and if I recall correctly one of the pest papers found ancient Y. Pestis couldn't. The point is once Y.Pestis can survive there is clogs the intestines of the fleas so much it will starve and in the process will attempt to suck blood from any available warm blooded animal, in vain. This, however, will transmit the bacteria.

https://www.cam.ac.uk/research/news/plague-in-humans-twice-as-old-but-didnt-begin-as-flea-borne-ancient-dna-reveals

Matt said...

Hmm... Gray, looks like he's got a brand new tree - https://i.imgur.com/BePr17W.jpg. Shows coalescence about 4000 BCE (e.g. closer to the PC steppe time than late Anatolian time of 6000 BCE or beyond he normally goes for). It's triple structured with a split of southern from European, then Balto-Slavic, Western IE split. Given he's talking about a tripartite geographical separation of IE through Anatolia, PC-steppe and Iran, now, it looks like he's suggesting that Yamnaya/Corded Ware are proto-Balto-Slavic, while Western IE descends through some other channel.

Browsing around for anything Gray published in 2018 found this set slides - http://indoeuropean.wdfiles.com/local--files/abstract/SS2017_Gray_slides_Pavia2018v2.pdf.

This seems to have been given by at http://indoeuropean.wikidot.com/program - which links to his

"A new hybrid hypothesis for the origin and spread of the Indo-European languages - Russell Gray (MPI Jena)"

Abstract: For over 200 years scholars have vigorously debated hypotheses on the origins and spread of IndoEuropean languages. From Hindi to Icelandic, distinct cultures and populations all speak languages derived from a single source − Proto-Indo-European. But from where, when and why these languages spread remains an enigma. Recent archaeogenetic findings and contentious results from computational phylolinguistics have further fuelled these controversies.

The debates have focused on two leading hypotheses: an origin either on the Pontic Steppe c. 6000 BP associated with horsebased pastoralism or in Anatolia c. 9000 BP associated with the spread of farming. In this talk I will summarise the state of the art in both genetics and linguistics, and outline a novel hybrid hypothesis that proposes an origin in the eastern Fertile Crescent c. 8000 BP.


From the slides I'd gather that the Jena team seem to have adapted to try and beat Chang's method that revised the dates substantially (although still placing them too deep for Yamnaya as such). They seem to have some criticism of Chang's data coding and forcing of written known varieties to be ancestral (I'd gather one of these is date compression among known written varieties).

I still can't believe Gray's got it right - the idea that three varieties of IE would split off in different directions, that the presence of a variety of non-written sources in different regions is simply random, as well as the wheel reconstruction and so on, seems just far too much to bear (plus the genetic aspects as we've all discussed). Still, let's see him try his luck with his model.

(Of course, the key figures about both what Gray and Jena's hybrid model and the what the new methods in language phylogenetics are, are unfortunately absent from the above link! No spoilers I guess...).

Matt said...

The French dub of the video linked upthread seems to work better than sticking auto-translate on the German dub - https://www.youtube.com/watch?v=a34TgLS9Tj8, if anyone is interested in what is actually said. (Still not very well, and non-dubbed version would be more ideal for English speakers of course! Or actually speaking the language...).

Also extended a bit so he shows the genetic basis for their Armenian hypothesis (of course this is perhaps more doubtful given Wang's paper?)

Davidski said...

What really strikes me as odd about these proposals from Gray and others from Max Planck is how they don't have any real support from ancient DNA.

It seems like Gray doesn't understand the ancient DNA data that's being generated at Max Planck Jena, and his colleagues there, namely Haak and Krauze, don't want to explain it to him.

EastPole said...

@Davidski
“What really strikes me as odd about these proposals from Gray and others from Max Planck is how they don't have any real support from ancient DNA”

Maybe genes are not so important. But I have a question then:
If Indo-Iranian languages (red arrow) and Balto-Slavic languages (blue arrow) separated around 5000-4000 BC:

https://i.postimg.cc/NjZSfPQb/screenshot-462.png

and Balto-Slavs went north to the steppe and Indo-Iranians went east to Iran and India, then what language was spoken by Andronovo pastoralists who came from Central-Eastern Europe (CWC>Sintashta>Andronovo), migrated to India around 1500 BC and ,as we know from aDNA, were autosomaly still close to Balto-Slavs at on the border of India?

epoch said...

@Davidski

Also, the slides seem to suggest that it is between him and Chang 2014. But he seems to miss entirely the point that this whole Bayesian modeling both of 'm use could be completely useless.

JuanRivera said...

As the Anatolian hypothesis collapsed, its proponents shifted to the Armenian one. It was a predictable thing. They still don't get that the PIE speakers were most likely inner eurasians north of the Black and Caspian seas.

Matt said...

@Davidski, Yeah, I think it's really hard to see how it fits easily with the genetics; it seems like you need a lot of big language switching despite minority component and dilutions to happen, and that's assuming that we've just missed some Eneolithic signal in the steppe. IE from the steppes has less of that on the whole (India, Mycenae, Hittite? where genetic influence is fairly low and language switching seems high).

And it's hard to see how the Western branch even works at all, in the sense of having a genetic correlate to its expansion (unless we are to say that the correlate is the CHG like pulse through Greece to Sardinia, and Mycenean reflects a further pulse out from Anatolia of the southern branch, which all seems quite difficult).

@EastPole, I'd assume that following Gray's model, if I understand right how it works, Sintashta and Andronovo (and indeed all early EMBA-MLBA steppe societies that are assigned as Indo-European) were not proto-Indo-Iranian, but they spoke a very early form of proto-Balto-Slavic. Then they would've language switched as they were absorbed into more numerous Indo-Iranian speakers to the south (probably with some influence from this very early Balto-Slavic dialect and religion etc. into Indo-Iranian?).

Under this model the existing Balto-Slavic languages would only represent a linguistic subset of a larger related family which once dominated NE Europe / Western Russia and the Pontic-Caspian steppes, rather than proto-Indo-Iranian.

JuanRivera said...

I recently ran models of iberians, the Seh_Gabi_ChL component used is higher in the south, mediterranean coast and Galicia, whereas it's lower in the interior and absent in the portuguese. It doesn't bode well for the Armenian hypothesis, as the south and mediterranean coast are known to have been non-IE-speaking until the romans came, and the interior was IE-speaking.

JuanRivera said...

The same could be said for non-steppe CHG. Overall, I don't know how the Armenian hypothesis proponents reconcile that distribution in Iberia, as well as the fact that Sardinia was non-IE-speaking until the romans, and that Germanics nor Balto-Slavs have none of it.

bellbeakerblogger said...

There was another European population collapse in animals at this time: domestic pig 1.0.

Since the authors show the phylogenetic root of yersinia pestis in the Nordic-Baltic area, and since yersinia pestis apparently wasn't spread by the flea...
http://www.sciencemag.org/news/2015/10/bronze-age-plague-wasnt-spread-fleas

I wonder if European domesticated pigs of that time were being infected by Eastern Baltic boars.

http://bellbeakerblogger.blogspot.com/2017/03/and-bronze-age-pigs-caliebe-et-al-2017.html

Anyhow, It'd be interesting to compare the human decline with that of Neolithic farmer pigs. Might be a little tautologous, unless those pigs were infected at considerable rates.

And also, I think Razib predicted a similar situation in Uruk, and again I believe this pig turnover happened there as well.


a said...

Davidski said...

"I wonder who was speaking them there at the time?"

Definitely not our ancestors.

Have you seen any studies showing the survival rates/ancient plague- blood type ratios?
One of the dominant Steppe blood types A-.

George said...

Hi,

Blood Type Biochemistry and Human Disease:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5061611/pdf/nihms804234.pdf

From page 8
"Type O blood group is associated with increased incidence of plague, cholera, mumps, and tuberculosis infections; type A blood group is associated with increased incidence of smallpox and Pseudomonas aeruginosa infection; type B blood group is associated with increased incidence of gonorrhea, tuberculosis, Streptococcus pneumoniae, E. coli, and salmonella infections; and type AB blood group is associated with increased incidence of smallpox, E. coli, and salmonella infections"

Please see the summary in Table 2 on page 28.



a said...

George said...

"Blood Type Biochemistry and Human Disease:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5061611/pdf/nihms804234.pdf"

I remember vaguely AE. Mourant's work-frequency of A+ regions Armenia/Balkans etc...-. It would be interesting to plot the mutations of Yersinia pestis;and heavy urbanized areas within Europe-Anatolia[Justinain plague for example]and the evolution of blood types.

Samuel Andrews said...

New Neolithic DNA available from Poland. Published October 2018.

A genomic Neolithic time transect of hunter-farmer admixture in central Poland
https://www.nature.com/articles/s41598-018-33067-w#Sec21

Who were the farmer ancestors of Northern Bell Beaker, of Sintashta, of proto-Germans, of Slavs. These are important questions for the origins of the bulk of modern Europeans.

At this point it is unknown. There's a huge collection of candidates. So this new data from farmers in Poland is important.

Davidski said...

@EastPole & Matt

I'm kind of shocked how incompetent this supposedly new interdisciplinary Indo-European homeland theory from the Max Planck linguists looks.

It's nothing more than a revised Anatolian theory, although still mostly at odds with the latest archaeogenetic results.

But hey, the positive thing is that this leaves it to others to really work things out.

JuanRivera said...

I kind of noted that before. Indo-Iranian and Germanic from outside the steppe? Seriously? It ignores interactions between both of them and Uralic. And genetically, the CHG in Steppe groups is different from the one south of the Caucasus. And there's the fact that all groups retained some nomadism, and specially Indo-Iranians, Germanics and some Italic groups.

Bob Floy said...

@Davidski

"But hey, the positive thing is that this leaves it to others to really work things out."

More and more I'm beginning to think that this is just the way it has to be right now, these guys have really, really lost the plot and it stinks to high heaven, I don't care what anyone says. Has anyone looked at the wiki page for Yamnaya lately?

https://en.wikipedia.org/wiki/Yamnaya_culture

"R-M269, one of the most prolific paternal lineages across western Eurasia. R-M269 arose roughly 10,000 years ago, as the people of the Fertile Crescent domesticated plants and animals for the first time. Around 8,000 years ago, the first farmers and herders began to push east into Central Asia and north into the Caucasus Mountains. Some of them eventually reached the steppes above the Black and Caspian Seas. There, they lived as pastoral nomads, herding cattle and sheep across the grasslands..."

Bob Floy said...

Unless they know something really big that NO ONE outside the academic community has any inkling of(unlikely), and are planning to reveal it next year or whatever, something is very wrong.

Davidski said...

@Bob Floy

Has anyone looked at the wiki page for Yamnaya lately?

Yeah, it's just pure propaganda. I don't think even the person/people who wrote that believed it. They were just hoping to dupe the readers and cause a bit of confusion for the time being.

I guess they think that when ancient DNA finally totally debunks their fantasy narrative they'll cross that bridge when they come to it, or something.

Ric Hern said...

@ Samuel Andrews

That is very interesting. What is especially interesting and boggles my mind a bit is Figure 5. Ukraine GAC has more WHG than Polish Neolithic samples. So I wonder if GAC maybe brought the extra WHG seen later ?

The other thing is the CWC samples that apparently do not fit Yamnaya as well as later CWC samples...Mmm...

Davidski said...

@Ric

The other thing is the CWC samples that apparently do not fit Yamnaya as well as later CWC samples...Mmm...

The most Yamnaya-like CWC samples are the earliest CWC from the Baltics.

After that they become more and more admixed with farmers, and start resembling Bell Beakers. All of the Polish CWC sampled to date have a lot of farmer admixture.

Ric Hern said...

@ Davidski

Sorry I think I phrases it wrong. Apparently if I understood correctly those CWC samples did not point to Yamnaya as their origin....?

Davidski said...

@Ric

How did you figure that out?

Ric Hern said...

@ Davidski

Sorry I think I phrases it wrong. Apparently if I understood correctly those CWC samples did not point to Yamnaya as their origin or source of their origin...?

Ric Hern said...

@ Davidski

I just read the Paper Samuel Andrews shared earlier.

Ric Hern said...

@ Davidski

https://www.nature.com/articles/s41598-018-33067-w#Sec21

Davidski said...

@Ric

I'm not sure what you mean, because there's nothing original in this paper in regards to the CWC.

The data show the same thing as in previous papers: Yamnaya ancestry is missing in pre-CWC populations (in this case GAC) and then suddenly shows up in CWC.

The two CWC males in this paper belong to Y-haplogroups I2, so they probably have GAC ancestry, but they still have significant Yamnaya ancestry, which suggests that migrants from the steppe founded the CWC in Poland. This is what the authors of the paper argue as well.

More sampling of CWC remains from Poland is likely to show exactly the same patterns as in Czechia, Germany, and the East Baltic, with very Yamnaya-like CWC populations expanding into the area, and then mixing with the locals and becoming less Yamnaya-like.

Bob Floy said...

@Ric

This paper looks a little funny to me. They're just flat out referring to CHG as "Iran_Neolithic", Anatolian farmer ancestry as "Levant_Neolithic", and talking about GAC as if it had been considered an IE culture until just now? What the hell is going on here?

Ric Hern said...

@ Davidski

In the Discussion "The CWC’s affinity to WHG, however, contrasts with results from published CWC individuals that identified steppe ancestry related to Yamnaya as the major contributor to the CWC genomes"

Or maybe I understood this wrong...

Davidski said...

Yeah, the mixture models in this paper are crap, but the authors clearly point out that CWC came from the steppe.

They also point out that there seems to be some substructure in CWC, because one of their Polish CWC samples needs extra WHG ancestry to be modeled correctly.

What this means is that he had extra WHG because his steppe ancestors mixed with locals with excess WHG in Poland.

Ric Hern said...

@ Bob Floy

Yes indeed.

Davidski said...

Here's the pertinent quote from the paper. Nothing really ground breaking.

We were able to obtain significant models (p > 0.05) with the Yamnaya by analysing the two CWC dyads separately, with the results suggesting that while the N44:45 dyad can be modeled as a result of admixture between the GAC and the Yamnaya (p = 0.340), the N47:49 dyad requires also a small contribution from a WHG source (0.596 < p < 0.765) (Supplementary Table S5).

The culprit was probably a GAC ancestor with an unusually high level of WHG.

Ric Hern said...

@ Davidski

Thanks. I thought it meant Steppe Ancestry not directly related to Yamnaya but some other Steppe source.

Ric Hern said...

So I could be right about the extra WHG coming from Ukraine GAC into Poland...

Davidski said...

@Ric

GAC probably entered Ukraine from Poland, not the other way around.

And the reason that GAC was so rich in WHG ancestry may be related to the fact that pure WHG individuals still existed in Poland well into the Neolithic, and the aforementioned paper actually has a sample like that.

Ric Hern said...

@ Davidski

Thanks.

Folker said...

In a certain watten, Gray’s hybrid theory may appear attractive if you try to find PIE South of Caucasus, and it can explain the odd conclusion of Wang et al paper. If you are postulating that the CHG-like admixture found in the Steppe came during Mesolithic or early Néolithic from a population A, which also admixed with local Caucasians and Anatolians, you can find some backing in data. But only if you don’t look too closely, or postulating that the results showing a difference in the source of CHG are wrong. It could explain why Wang et al has been postponed. If the CHG-like found in the Steppe is really different from those found elsewhere, all this will fall apart. But given the Mt hg, the answer should be obvious.

Bob Floy said...

@Davidski

What is all this lamenting the no steppe ancestry in GAC, as if they had been expecting it? Did I miss something? Isn't GAC's non-IE/neolithic farmer status old news?

Davidski said...

@Bob Floy

Well, Gimbutas thought that GAC was IE, but I was skeptical, especially after seeing the first DNA results from a GAC sample.

Here's a rather popular post from my other blog from way back...

The Globular Amphora man from Late Neolithic Poland

Ric Hern said...

About the Y.Pestis. In Figure D it looks like it dispersed around Belarus while in Figure A it shows Human Migration through Belarus. Now I wonder if Belarus was less affected by the Plague and if it served as some kind of refuge area before later expansion towards the West ? Or is it just my day for seeing things that aren't there ? Heheheeh..

Bob Floy said...

@Davidski

"Gimbutas thought that GAC was IE"

Yes, but that was a long time ago, I had the impression that Gimbutas's ideas weren't really much of a topic anymore, what with the big influx of data from the last few years. Half of her basic ideas seem to have been vindicated while the other half just seem kind of quaint now.

And that is quite an old blog post, I was still paying attention to Maciamo back in those days, lol.

EastPole said...

@Davidski
“But hey, the positive thing is that this leaves it to others to really work things out.”

It looks like they will start looking into joint modeling of genetic and linguistic data next year:

“Co-estimating Human Mobility and Language Dispersal with Ancient DNA and Linguistic Data”

https://www.findaphd.com/phds/project/co-estimating-human-mobility-and-language-dispersal-with-ancient-dna-and-linguistic-data/?p104661

In my opinion joint modeling of genetic and linguistic data will produce more probable results than looking only at languages or only at genes.

Davidski said...

@Folker

In a certain way, Gray’s hybrid theory may appear attractive if you try to find PIE South of Caucasus, and it can explain the odd conclusion of Wang et al paper.

Yeah, clearly, the Max Planck Jena PIE homeland theory and the odd conclusion in Wang et al. are closely related phenomena.

And since the data in Wang et al. showed that Yamnaya was not founded by migrants from south of the Caucasus or from Iran, Max Planck will have to come up with a new, more nuanced and complicated way in which language may have been acquired by Yamnaya from the south. I think this is a clue...

This region thus allows to study in detail the mixing and interdigitation of people, their materiality and cultural systems and challenge many of the too simple models developed for another interface of the Eurasian steppe zone those directed towards Europe.

How should we interpret the movements of people throughout Bronze Age Europe?

Them meee said...

I remember a few years back everyone thought GAC was IE somehow. So it is still very recent a development (only became totally mainstream after last year’s samples) and thus a hard pill to swallow for some.

Even things have changed in many ways in the last year, so seeing folks who are slow at adapting in academia (can’t let go of Renfrew or GAC as IE) is lamentable but sorta expected since things were slower paced back in the day before the advent of ancient DNA. It is not a justification but it is why it happens.

Bob Floy said...

@Them meee

You're right about all of that, but, I dunno, it seems to me that one should be able to expect a little more from these very well funded crack research teams, they're supposed to be the best of the best.

Matt said...

Folker: But only if you don’t look too closely, or postulating that the results showing a difference in the source of CHG are wrong. It could explain why Wang et al has been postponed. If the CHG-like found in the Steppe is really different from those found elsewhere, all this will fall apart.

This is where I suspect that it will be hard to actually demonstrate what really different really means (because you can have composites of earlier and later events, mtdna subject to drift and selection, date estimates from site frequency data hard and experimental) and they will just need a transect covering the period labelled in the graphic from Rascovan from this post as "10 kya - 7 kya - Expansion of populations from Iran/Caucasus to Eurasian Steppe". Rather than just picking up samples at the tail end of that. If there's an increase, even of similar amounts to Mycenaeans' steppe (15% from memory?) hard to see a scenario without a movement, and if there's not, there's not.

Davidski said...

@Matt

All we've got to work with right now are those three Eneolithic steppe samples from the North Caucasus steppe, so let's do that.

If you look at the ADMIXTURE bar graph from Wang et al., it seems like the oldest sample, VJ1001, has the most of the green ancient Caucasus-related component. The other two, and especially the youngest sample, PG2004, has the highest levels of the European blue and Siberian-related light green.

Wang et al. bar graph

The differences are subtle, to be sure, and the dates for these three samples actually overlap somewhat, but there appears to be a fairly clear trend nevertheless.

So what do you reckon that a sample from that area of the steppe dating to a few hundred more years than VJ1001 might look like? I think it'll be almost totally green.

In other words, right now, to me it looks like there was an influx of EHG ancestry into this region during the Eneolithic and/or a bit earlier, rather than a migration from the south.

Matt said...

I think I'd rather have earlier dna and more samples.

Bob Floy said...

What does the little sliver of red in the steppe Maykop samples represent?

Davidski said...

@Matt

Me too, but I'm seeing a pattern there.

@Bob Floy

I think it's the cluster that peaks in Amerindians.

Bob Floy said...

@Davidski

Thanks.

Unknown said...

aDNA from Poland (Brześć Kujawski (BKG) G2a, Funnel Beaker (TRB) C1a, Corded Ware (CWC) I2a, Early Bronze Age (EBA) R1a) https://docs.google.com/spreadsheets/d/1yg8N3s0KOEfYMg6pWKGLkUK6w_IfcKFqUrQCenXx0JQ/edit?fbclid=IwAR1CCAfa5cNMjKtoOzQfRnfoJT06Rn9ErbOdYOF7ajRMDaYaKZypXzbZP90#gid=0

Ric Hern said...

Something interesting.

Why did Balkan Neolithic/Chalcolithic Cattle MtDNA Haplogroup T6 basically disappear from Europe ? Was the Plague responsible ? A bit of flight of the imagination makes me think that when the Plague struck many people were unable to plough and plant and maybe relied heavily on animals to provide nutrition ? Did Neolithic people eat Haplogroup T6 into close extinction ?

https://www.researchgate.net/publication/289335710_An_independent_event_of_Neolithic_cattle_domestication_on_the_South-eastern_Balkans_evidence_from_prehistoric_aurochs_and_cattle_populations

Toby_P said...

@ Davidski

''By the way, I don't think it's a coincidence that image B in the figure below explicitly says that animal traction and wheeled transport spread without much human gene flow'

That's sems to be the case. This paper suggests that the idea of the wheel was transported in components of a package, diffusing, and allowing local groups to reformulate according to their own needs

https://www.academia.edu/35839438/Transforming_Technical_Know-how_in_Time_and_Space._Using_the_Digital_Atlas_of_Innovations_to_Understand_the_Innovation_Process_of_Animal_Traction_and_the_Wheel

Philippe said...

have you seen the wiki page on the genetic history of europe?

https://en.wikipedia.org/wiki/Genetic_history_of_Europe

Davidski said...

Credit to Wikipedia for putting up these warning notices...

This article's factual accuracy may be compromised due to out-of-date information. Please update this article to reflect recent events or newly available information. (May 2017)

This article may need to be rewritten entirely to comply with Wikipedia's quality standards. You can help. The discussion page may contain suggestions. (January 2018)


But people should still complain. That might help to clear the page of all that garbage.

dsjm1 said...


Just watched the new David Reich Video on Youtube - in particular wanted to hear the comment about Yamnaya in Hungary.

How I might interpret this is if recent ancient burial data were able to establish that an early branch of Yamnaya were indeed already well established in the Hungarian Plains well before a later set of migrations (involving L51-L11 sub-branches.

I gather there is already evidence for the much older R1b-Z2103 sub-clade being in the Carpathians (Vučedol Culture in particular) and my reaction to this is to wonder if it is what Dr Reich may be alluding to. Aren't there some new publications due any day now that go deeper into ancient burial finds in the Hungarian Plains ?.

At the same time I see in his (Reich's) migration slide an arrow showing migrations both into the Carpathians & just to the north of the Carpathians. If that north migration arrow through today's Ukraine & Poland is not specifically R1a (which I doubt it is based on the start of the arrow) then I can accept that older Yamnaya were already in the Carpathian Basin prior to the later R1b-L11 based Yamnaya migrations.

Interestingly, Prof Kenneth W Harl (Tulane Uni) just posted a new lecture on PIE and its spread, and he shows a map that is in-line with the map of Reich (but adds an arrow higher up showing what is obviously R1a 'Baltic' & 'Slavic' PIE movement and he also has a lower arrow adding 'Greek' PIE movement from the Black Sea west coast. Prof Harl's arrow going just Nth of the Carpathians is labelled 'Germanic' PIE. The arrow going into the Carpathians & along the Danube is labelled 'Celtic' & 'Italic' PIE - The link to this (go to 16:40 mins to see Harl's map) = https://www.youtube.com/watch?v=xkVc_WX35eE&t=112s&fbclid=IwAR3UCA3Cbmol0p9YatDd6xZsHRYCK7pxpiyhVnRbyWPlitoYT_tosygvcKs

So my comment is (and I am new to this type of deeper analysis so may be way off target) is that Yamnaya as R1b-Z2103 were in the Hungarian Plains long before the R1b-L51-L11 derived migrations which appear to be around 3000BCE-2800BCE.

I hope this makes sense.

Doug M

Them meee said...

Man I just hope they have new samples to back all of that up or something. It would be nice to have a hint at the source of R1b-L151 just like R1a-M417, and at that of CWC. The new stuff about Indo-Iranian is nice though.

Ric Hern said...

Could this be a clue ?

https://www.researchgate.net/publication/323684421_The_domesticated_horses_from_the_submerged_prehistoric_village_of_Urdoviza_Kiten_on_the_Bulgarian_Black_Sea_coast-among_the_oldest_known

Ric Hern said...

What is interesting about the size of these horses is that they basically fall into the size limits of the Bosnian Mountain Horse, the Altai Horse and the Connemara Pony....and they seemed a bit more robust than the Botai type.

Apparently this type also ocurred earlier in Transylvania...

EastPole said...

@dsjm1

“Prof Kenneth W Harl (Tulane Uni) just posted a new lecture on PIE and its spread”

His maps are outdated. Indo-Iranians didn’t migrate from Yamnaya but like Balto-Slavs are derived from Corded Ware Culture (CWC):

CWC>Sintashta>Andronovo>Indo-Iranians

Here is a better map, based on prof Eske Willerslev genetic study:

https://i.postimg.cc/j2cYRQ1D/screenshot-404.png

https://www.academia.edu/36689289/Invasion_aus_der_Steppe

dsjm1 said...

@EastPole...

Thanks for those links. Is the message in Eske's chart, that R1a migrated to CWC area, then later migrated all the way back past the Steppes homeland?.

I listened very carefully to Prof Harl warning us not to mix race with language. So, am now just a bit confused.

Cheers Doug

Davidski said...

@dsjm1

Is the message in Eske's chart, that R1a migrated to CWC area, then later migrated all the way back past the Steppes homeland?

There's a post and a major discussion on this topic here...

The mystery of the Sintashta people

Toby_P said...

That article Ric linked is interesting. ''pact on the progress of the civilizations, is disputable.
The horse was probably domesticated rather late and
the opinion about an early domestication in the 5ft
millennium B.C. (Sredniy Stog culture, Derievka) is recently
strongly contested. According to the recent data,
the domestication has happened probably in the western
part of the Eurasian steppe, between the Northern
Black Sea region, Iran and Kazakhstan. It seems that
this process was developed not earlier than the first
half, and most probably during the middle (even the
last third) of the fourth millennium BC''

We know that Botai horses were possibly the earliest to be domesticated in 3500 BC, although a more western line actually became successful. So I guess horses did not come from Eneolithic steppe, but perhaps from somehwat further west ?

Davidski said...

@Toby_P

So I guess horses did not come from Eneolithic steppe, but perhaps from somehwat further west?

Not sure what this guess is based on exactly, but I'm not guessing when I say that a sister clade of the main domesticated horse Y-chromosome was present in horse remains from a site not far from the one where the human remains with the first R1a-M417 were found. See here...

Of horses and men

So the chances that the modern domesticated horse lineage doesn't come from the Eneolithic steppe north of the Black Sea are pretty small.

Wait for the next big ancient DNA paper on horse domestication and you'll see.

dsjm1 said...

@Davidski ...

A-ha.!. Thanks for that link. It explains what @EastPole was mentioning.

Doug

Toby_P said...

@ Davidski.
Thanks ! Any ideas when that paper is coming ?

Davidski said...

Early next year?

Major horse paper coming soon

Ric Hern said...

@ Toby

This type of horse was also found at Kostenki during the Paleolithic. So it could still be that some related type was domesticated on the Steppe and from there spread to Bulgaria. 3300 BCE seems to be the time of arrival there. Interesting is also the Salzmünde Tabiano Horse dated to roughly the same time...So someone spread domesticated horses during this time from somewhere...most probably the Steppe or somewhere closeby, maybe Western Ukraine or the Transylvanian Plateau...relatively big horses in Hungary during the Bell Beaker period.

Ric Hern said...

@ Toby

The Cavepaintings at Kapova Cave/Southern Urals seems to resemble this type of horse. This is not far where the Samara Culture originated. So when looking at the spread of Horseheaded Septers starting within Khvalynsk > Novodanilovka > Suvorovo, the origin looks like somewhere near Samara.

Ric Hern said...

The problem with todays horses is that most were "upgraded" with Arabian Stallions and Turkmene Horses. So genetically most will genetically pull towards the Middle East. So yes ancient DNA is needed to solve the puzzle.

Davidski said...

Arabian and Turkmen horses have the Y-HT-1 Y-chromosome haplotype, so they're from the steppe originally, at least in terms of paternal ancestry. They're probably derived from Scythian horses.

But yeah, the fact that Y-HT-1 is fixed in Arabian and Turkmen horses has even confused some horse genetics experts who claimed that Y-HT-1 was from Asia.

It's not from Asia. Like I said in my blog post linked to above, it's the sister clade of the older Y-HT-4, which has been found in Eneolithic horse remains from the North Pontic steppe.

Ric Hern said...

@ Davidski

Yes indeed.

Toby_P said...

@ Ric

I recall that Marsha Levine dismissed such paintings as ''circumstantial evidence'', at best, just like cave paintings of any other wild animal.
Outside Indo-Europeanist circles, not many scholars accept Anthony's interpretation of the horse-sceptres, because they interpret their ''basicness'' as indicating imitation of the more sophisticated forms in the West, rather than being older/ more archaic.
Certainly, after seeing Alan Outram's interview, he doesn't seem to be jumping toward horse domestication at Samara Eneolithic, or even the Yamnaya period. However, he could just be holding his Aces..

Davidski said...

Outram can't reveal the results of a new paper in an interview, but he gave strong hints that the horse remains from Yamnaya burials belong to the modern domesticated horse line.

And it's already a fact that the modern domesticated horse line was present on the Eneolithic North Pontic steppe, because horse remains from there have been tested and they belong to the Y-HT-4 haplotype, which was one of the main domesticated horse Y-chromosome lineages until historic times, when it was bred out by selective breeding.

These horse remains from the Eneolithic North Pontic steppe are supposed to be from a domesticated horse, so if the modern domesticated horse wasn't domesticated on the North Pontic steppe during the Eneolithic, then it was domesticated somewhere else earlier and imported onto the North Pontic steppe. But this doesn't sound very plausible.

There's a chance, I suppose, that this was actually a wild horse, but again this would mean that the same horse line was domesticated somewhere else, and imported onto the North Pontic steppe where wild horses just happened to have belonged to that line. Again, not very plausible.

So it seems like that there was a horse domestication center on the North Pontic steppe and this is where modern horses come from.

Ric Hern said...

@ Toby

https://www.york.ac.uk/biology/news-events/news/cave-paintings-of-horses/

Ric Hern said...

@ Toby

http://www.pnas.org/cgi/doi/10.1073/pnas.1108982108

Ric Hern said...

@ Toby

https://www.nature.com/articles/ncomms13158

Ric Hern said...

@ Toby

So according to these studies, ancient people portrayed their surroundings relatively accurately in Cave Paintings. Ancient DNA confirmed this.

Toby_P said...

Thanks. It makes sense I guess. When mammoths died out, horses and bison became the main prey. If there was a region for the horse to then domesticate, that would be it

Ric Hern said...

The interesting thing about that site in Bulgaria is the amount of remains from Stallions. Could this be an early sign that selection took place as early as 3300 BCE.or even earlier ? This could also explain the amount of Male lines in horses that is far less than the Female lines.

JuanRivera said...

There's an interesting word in the PIE vocabulary. That word is *sámh₂dʰos, which means sand. It helps narrow the PIE homeland, as it indicates either a coastal location or a dry one. The Pontic-Caspian steppe turns out to be both, as human-environment interaction would have forced most people to live near big rivers and the seas to the south, where sand is more likely to ocurr. Also, there's a sandy semidesert north of the Caspian. Other proposed homelands don't have much sand. It adds up nicely with horses, wheels, copper/bronze, beaver, birch, snow but not rain, the two words for cold and hot each, and significant water-related vocabulary (fishes for example) and some naval words (*néh₂us, *mori, etc), along with other things.

JuanRivera said...

And if I'm correct, PIE has a word or root that means to ride.

Ric Hern said...

@ JuanRivera

Yes. Personally I feel that the most accurate reconstruction of PIE will be by only using the Northern Branches from Tocharian to Celtic. Basically all languages at the same latitude as the Pontic Caspian Steppe...That is just my personal feeling. Too many other peoples and languages to the South that could screw up the original picture...

Arza said...

Was it already posted (news from 3.5.2018)?

An­other ERC Ad­vanced Grant se­cured for the Uni­versity of Helsinki: Ar­chae­olo­gists to col­lab­or­ate with natural sci­ent­ists

With his research group, Heyd wants to map out how the Yamnaya culture, also known as the Pit Grave culture, migrated from the Eurasian steppes to prehistoric south-eastern Europe approximately 3,000 years BCE.

Field studies and sample collection for the project will be conducted in Romania, Bulgaria, Hungary and Serbia.


helsinki.fi

MAP

JuanRivera said...

I found the word. It's *gʷer-, and means (to) mount. It's clearly in the context of mounting an animal, likely a horse.

Them meee said...

@Arza

Well looks we’re finally getting somewhere. I especially can’t wait to see results from Romania and Hungary.

It doesn’t matter if it’s a dead end or not, it will be huge and if it’s a dead end we will get a better picture and idea of where to look anyway.

Matt said...

For the most part, I think the consensus around here is (fairly enough) the new Jena work around applying phylogenetics to IE is doubtful at best, but if anyone wants to get a sense of what they think they're doing, various links:

https://www.york.ac.uk/language/news/events/events2018/pllt-16-nov/ - "15/11 - 11:30-12:30 Paul Heggarty (Max Planck Institut, Jena) The challenge of tailoring cognate data-sets for phylogenetic analysis: illustrating the new IE-CoR database. 16/11 - 11:00-12:00 - Russell Gray (Max Planck Institut, Jena) - The origin of Indo-European – solved?"

https://imgur.com/a/r8Askk4 (sorry lost the web link) - Historical Linguistics Workshop, 7th December 2018 - Jena

https://www.shh.mpg.de/438157/cobldatabase - "CoBL is a new database structure for exploring how languages within a family relate to each other in Cognacy in Basic Lexicon."

http://lingulist.de/calc/events/slides/dot_heggarty.pdf - recent(ish) 2017 presentation by Heggarty on CoBL and his opinion on what he believes to be the flaws in steppe hypothesis (basically an update of his 2015 blog post - https://dlc.hypotheses.org/807, and I'm probably a little less sympathetic to his PoV for having read the presentation to be honest).

https://i.imgur.com/2VT6K0m.png - CoBL: The Germanic Test Case (abstract, 2016). Cuts to the quick of what they believe wrong with the dataset that the previous models were using was wrong in.

https://www.researchgate.net/profile/Simon_Greenhill/publication/328696049_Bayesian_Phylolinguistics/links/5bdc51db92851c6b27a177d0/Bayesian-Phylolinguistics.pdf

https://consultingphilologist.wordpress.com/2018/10/01/indo-european-etymological-dictionaries-a-guide-for-the-perplexed/ - "Over the last three years I have worked more-or-less intensively and widely on Indo-European etymological problems as a research associate in Indo-European comparative linguistics on the Cognacy in Basic Lexicon project with the Max Planck Institute for the Science of Human History. "

JuanRivera said...

U2e, U4b, C1e, C1f, C1g, C4a3, C5, CH2a1, H6a, some other H clades, some T clades, I, W, etc, were mtDNA clades that distinguished steppe populations from EEF populations to their west. U2e and U4b are from ANE, in which they developed from U2 and U4 from their Kostenki-Sunghir-like ancestors (who probably went extinct, leaving Yana and their ANE descendants the only populations carrying them). They are found in EHG and WSHG samples, and they both transmitted it to their steppe descendants. C1e, C1f, C1g, C4a3 and C5 also seem to be from ANE, but this time they got their ancestor clades from their Tianyuan-like ancestors. They were attested in EHG, WSHG, and Baikal_HG. The rest of the clades are from steppe CHG (the CHG in steppe, which remained north of the Caucasus since the Late Upper Paleolithic), which in turn got it from their Dzudzuana-like ancestors. On the paternal side, R1a, R1b(xR1b-V88) and Q1a clades (such as Q1a2b) seem to be EHG and WSHG, and in turn ANE; and being additionally found in WSHG descendant Botai and EHG (and WSHG to a very minor extent) descendant steppe peoples. All are reduced or absent in non-IE populations, excluding the paternal markers.

JuanRivera said...

And a note. Paul Heggarty apparently thinks IE is related to the other near eastern languages, judging by his dislike of the Eurasiatic hypothesis, even though IE's core is more related to other north eurasian language families than to near eastern ones. Also, he was a supporter of the Anatolian hypothesis, and shifted to the Armenian one after it became untenable. All to support his view that IE had to bring civilization with them, which somewhat disparages nomadic and pastoralist peoples by implying the common argument of "IE couldn't have been spoken by nomads (or even barbarians, said by other people) because it's too complex", which is false, especially when one looks to Taa, Yelî Dnye, indigenous australian languages, etc.

JuanRivera said...

What's funny is that they're still insisting in the Gray & Atkinson variation (with very few modifications), which completely excludes the steppes, and all miltidisciplinary evidence supporting the steppe hypothesis, including even the evidence of contact of IE branches with Proto-Uralic and its descendants.

Dragos said...

''For the most part, I think the consensus around here is (fairly enough) the new Jena work around applying phylogenetics to IE is doubtful at bes''

May I ask if that is based on the combined expertise in cladistics, linguistics and anthropology of the ''us'' here, or does it hinge on the emotivenss of preferred narratives ?

Thanks in advance

Them meee said...

I wonder how will these academics react if (when) the evidence blows up in their faces.

JuanRivera said...

Yamnaya's CHG ancestry is different from the one south of the Caucasus, being drifted and EHG-admixed, and was present in preceding cultures. It imparted mtDNAs H, T, I, W, etc, and probably some Y-DNA J, but it imparted neither mtDNAs U2e, U4b, C1e, C1f, C1g, C4a3, C4a6 and C5, nor Y-DNAs R1a, R1b and Q1a. And archeolinguistic evidence alone supports the steppe hypothesis.

JuanRivera said...

And let's not forget Y-DNA I2a.

Matt said...

Dragos: May I ask if that is based on the combined expertise in cladistics, linguistics and anthropology of the ''us'' here, or does it hinge on the emotivenss of preferred narratives ?

Can't speak collectively, but doubts seem mostly around here I think from synthesis of models with ancient and modern dna and known data on distribution of ancient languages attested in writing.

For myself; the eastern agriculturalist expansion seems hard to argue as IE speaking on the basis of non-IE Elamite and Dravidian, signs of an expansion across the Caucasus to steppes at 7000 BP are not clear (could be preexisting ancestry), and there's no genetic sign of a western expansion of IE by people with CHG/Iran_N ancestry into Europe at 7000 BP at all, and where it does show up later, it's in places where we don't see clearly early, clearly IE languages when writing emerges. None of these are fatal on their own, but they add up to my doubts, as explaining each one seems to add complexity. If they can be explained away without making the hypothesis much heavier and more complex, then great.

Personally, the linguistics themselves are difficult to have a clear position on even though it is at root a linguistic question in a sense. Even people who are expert in the subject and understand language far better than I ever will don't seem to clearly agree on many of the main questions. "Linguists don't do dates", except by these sort of vocabulary reconstruction archaeolinguistic arguments that are by nature selective and subject to cherry picking and circularity (linguists who are persuaded by one hypothesis or the other can find what they're looking for). Expert trees designed by hand based on their perceptions of which grammatical, phonological and morphological features are important are questionable on the same grounds of circularity. But the phylogenetic, computational methods are also questionable in being able to take into account language contact, though I admire that the people involved are actually doing lots of projects to try and account for this problem and doing testing in other language families and so on.

E.g.:

"There's clear wheel vocabulary that makes a post-5.5 kyr split essential.", "No there's not; that's just frequent homoplasy derived from a shared word for turning."

"There's vocabulary for horses, so they must have lived in an environment where horses were very frequent and important.", "Er... you don't need an animal to be about at high frequency for people to have a word for it, just for them to see it now and then enough in their lifetime to share a word for it with each other."

"There is evidence of high degree of contact with Uralic in the very core of the family." "No, there's not, that's peripheral and only solid evidence of contact in late branches, at best."

"High structure and diversity in cognacy and root meaning in early attested branches cannot have evolved in mere hundreds of years and implies much longer time depth for pIE than the Yamnaya or even an earlier Kurgan hypotheses can provide.", "That's all due to language contact."

and so on.

@JuanRivera: All to support his view that IE had to bring civilization with them

??? He's proposing a Late Neolithic-Eneolithic hypothesis though. Well before anything like civilization, spread by people who were by any measure less complex in material culture than Yamnaya.

But that does raise another question about/problem this Eastern Fertile Crescent hypothesis, which is that there is no processual reason about why it should happen. (Anatolian hyp had "They've got farming", Steppe had "They've got wagon based pastoralism", this Eastern Fertile Crescent one has "They've got ????").

Davidski said...

Eastern Fertile Crescent one has "They've got ????"

The almighty teal component.

Dragos said...

@ Matt
I see. I agree with you about lingusitics, which is somewhat frustrating give its centrality in some of the questions being posed.

To follow on ; ''Anatolian hyp had "They've got farming", Steppe had "They've got wagon based pastoralism", this Eastern Fertile Crescent one has "They've got ????''

They have aspects of those things, don't they ? Keep in mind that research in the eastern Fertile crescent has been lagging behind Europe and west Asia.

''For myself; the eastern agriculturalist expansion seems hard to argue as IE speaking on the basis of non-IE Elamite and Dravidian, signs of an expansion across the Caucasus to steppes at 7000 BP are not clear (could be preexisting ancestry), and there's no genetic sign of a western expansion of IE by people with CHG/Iran_N ancestry into Europe at 7000 BP at all, and where it does show up later, it's in places where we don't see clearly early, clearly IE languages when writing emerges. None of these are fatal on their own, but they add up to my doubts, as explaining each one seems to add complexity. If they can be explained away without making the hypothesis much heavier and more complex, then great.''

I'd tend to agree, but as you say, nothing is definitive. Certainly, I think the evidence from places in Europe where inscriptions do exist complicate any linear narratives. Indeed, in places where the clearest evidence for steppe expansion - we don't see any written evidence for languages until the Roman Era. As many people have pointed out, we need to be led by data from Greece, Anatolia and Asia, not western & northern Europe (as per further thoughts in Italian thread).

Them meee said...

@Dragos

Where do you think steppe fits in here? Given it appears in the Mycenaeans who spoke an Indo-European language and modern day IEs show steppe ancestry, as well as the lower steppe ancestry in Southern Europe, what do you think of that?

Davidski said...

@Dragos

As many people have pointed out, we need to be led by data from Greece, Anatolia and Asia, not western & northern Europe.

The problem currently is that the Max Planck linguists (and to a certain extent geneticists) aren't doing this properly.

Heggarty doesn't appear to be at all aware of the unusually close genetic, and especially paternal, relationship between Indo-European-speaking Eastern Europeans and South Asians.

Considering the geographic distance involved, the unrelated Neolithic farmer substrata in both of these regions (because in spite of what Heggarty claims Iranian farmers did not migrate to Eastern Europe!), and the late entry of the genetic signals into South Asia that link it with Eastern Europe, the idea that Eastern Europeans and South Asians don't share Indo-European ancestry via the Pontic-Caspian steppe can't be taken seriously.

Data from Greece are also backing the steppe hypothesis, with the Mycenaeans (but not the Minoans) sharing basically the same steppe component with Eastern European and South Asian Indo-European speakers.

Moreover, Minoans were already rich in the Caucasus/Mesopotamian component that Heggarty sees as the Proto-Indo-European genetic signal. So he better hope that the Minoan language is some day classified as Indo-European.

Anatolia is yet to be debated properly. But the idea that a dense sampling of Hittite era sites won't reveal samples with clear steppe ancestry seems absurd to me considering that such signals have already been picked up in Bronze Age Greece, Armenia and Iran.

On the other hand, where's the evidence of any population movements from the Near East that may have given rise to Yamnaya? There is none currently, in spite of what Max Planck Jena might want to claim. Their own data clearly show that Yamnaya wasn't related to Maykop, the earlier Meshoko, or any currently sampled ancients from the Near East.

How this isn't a problem for Heggarty and co. I have no idea, especially since they claim that ancient DNA is backing their theory.

Matt said...

Dragos: They have aspects of those things, don't they ? Keep in mind that research in the eastern Fertile crescent has been lagging behind Europe and west Asia.

Might not have been clear here. It's not so much that Eastern Fertile Crescent peoples would not have aspects of agriculture or pastoralism. But there's no proposed package or advantage which would drive their spread, such as the Anatolian theory goes with for agriculture, and the Steppe theory (at least in Anthony's version and as Reich talks about it) with that variety of pastoralism. Dienekes once suggested that bronze metalurgy was a driver of a spread from West Asia, although that seems doubtful (his theory didn't work on a genetic level that inspired it once adna came through, despite containing a germ of truth in its "West Asian" ancestry component).

Re; very late IE in some places (NW Europe?) it seems difficult to reconcile with deep structure that seems to be found in lexicon / root meaning traits. Not impossible, but more difficult. Don't really have much more to say on it than that.

dsjm1 said...


A thanks to all those commenting on IE spread and the issues surround the fine detail. The discussions have been very helpful and IMHO very well put. Refreshing !.

Can't help but agree that Max Planck Jena case seems to be a matter of holding out against superior data and a changing world brought on by the aDNA revolution and the admixture analysis it is capable of.

Am still though intrigued at why David Reich so recently defined the pre-'invasion' homeland of the 'Yamnaya' as being the Steppes regions from 'Hungary to the Urals'. He has to be basing that on new data he knows of that is not yet widely available.

If indeed we include Hungary as the western edge of an 'established' Yamnaya region, then it could possibly help explain why L51 (down to L11) are missing from our current understanding as to origin (Hungary ? or Black Sea Coast at mouth of Danube ?) but Z2103 (and lower clades) aren't a mystery, just the question why so few of them today despite their obvious spread around 5000BCE-4500BCE.

Doug M

Them meee said...

@dsjm1

Maybe he’s saying that because Yamnaya did reach that kind of distribution, though if did it early in its history (unlikely to be ancestral to the rest of Yamnaya) maybe it was due to a super rapid expansion like I said earlier here, which would also explain Afanasievo as being contemporaneous with Yamnaya despite being so far east.

Dragos said...

@ Matt

''Might not have been clear here. It's not so much that Eastern Fertile Crescent peoples would not have aspects of agriculture or pastoralism. But there's no proposed package or advantage which would drive their spread, such as the Anatolian theory goes with for agriculture, and the Steppe theory (at least in Anthony's version and as Reich talks about it) with that variety of pastoralism. ''

I see what you mean, that's true. Although I personally would steer away from ''cookie-cutter'' type models, so I see no issue with it.

''very late IE in some places (NW Europe?) it seems difficult to reconcile with deep structure that seems to be found in lexicon / root meaning traits. Not impossible, but more difficult. Don't really have much more to say on it than that.''

I might be misunderstanding you again, however, the reconstruction for proto-Celtic is that of the chariot-sword-iron complex seen throughout western Europe and The Isles c. 1200 BC.
For all the attention to Italo-Celtic unity, Schmit & Isaacs have pointed to eastern IE languages having some impact on Celtic (I-A, Balto-Slavic), or at least shared developments (although these developemnts did not directly involve R1a or ''Slavs'', as our friend Them meee seems to be worried about). On the other hand, we have the widely recognised non-IE susbtratum in Celtic languages, incl Insular Celtic as its own case (although many of these earlier formulations need adjustment as some of them hinged on outdated notions of Ice Age Pan-Vasconicism). Finally, we have a very shallow split within Insular Celtic (Turn of Era - time frame).

As for Germanic, its expansion through NW Europe was very recent indeed.

Them meee said...

Just for clarification, I didn’t mean to say R1a = Slavs (seems more like EastPole’s speciality) or that Slavs impacted Celts, but that it used to be believed that R1a was the source of Western IE and R1b was not an IE marker.

That basically boils down to thinking only R1a groups are truly Indo-European. That’s about it.

Also, I believe the Celtic expansions were probably rather recent since they probably replaced or mixed with a lot of groups who speak now extinct Indo-European languages or dialects. And let’s not forget the Hallstatt-like signal in Britons...

EastPole said...

@Them meee

„Just for clarification, I didn’t mean to say R1a = Slavs (seems more like EastPole’s speciality)”


I have never said that R1a = Slavs.
From genetic and linguistic trees it seems now very probable that R1a-Z645 = Indo-Slavs, and R1a-Z283 = Balto-Slavs.

Matt said...

@dragos, yeah, I understand those paleolinguistic arguments about core vocab have been made, and without much I'm not very firmly opposed to the idea that Celtic IE expanded late to NW region where the 2500 BCE influx didn't speak an IE dialect.

(Perhaps they'd language switched at some point on their expansion, though I think it seems less straightforward than something like "Northern Bell Beaker" speaking a dialect that is ancestral to Celtic+Italic. As you say we have no direct evidence of what language those people spoke.)

I'm thinking more that if we take something like Heggarty or Gray's graphs on lexicon seriously, those suggest divergence from Italic around 3000BCE and intra Insular Celtic divergence to begin with interchanges around 2000BCE and to be established by 1000BCE. That seems like something to be explained, whether that's through contact there or not, in a way that should be generally consistent with our preference for early / late models (e.g. we can't "cherry pick" early dates for the whole tree and then late dates for Celtic, without a solid footing of data to do so, etc.). (And generally that's the case for IE, even where no written sources, we must still explain deep divergences in languages that were unwritten until about 1000-2000 years after the first written evidences).

Ric Hern said...

Just a few questions. Are there close similarities between Slavic/Baltic and Sanskrit ? When did the split happen ? If we propose that it happened during the Corded Ware period or earlier and many similarities remain between them, why then insist that Celtic had to have started evolving Late from a common source in Central Europe ? Why propose that One IE branch seperated for more than 4 thousand years by thousands of kilometres show more similarities to others than one that proposedly only started separating from common source 3000 years ago and a few hundred kilometres ?

EastPole said...

@Ric Hern

„Are there close similarities between Slavic/Baltic and Sanskrit ? “


http://eurogenes.blogspot.com/2018/04/on-doorstep-of-india.html?showComment=1523740548383#c6191530895381495941

Ric Hern said...

At the end of the day R1b in Basque makes up only +-1% of the total R1bs in Western and Northwestern Europe. We know that Spain had many other Non-Indo-European influences during ancient times. Eg. Carthaginians and maybe even Minoans etc...most of which spread along the Coastline...

Ric Hern said...

@ EastPole

Yes I know. That is precisely my point. Why should Celtic and Italic have totally different Indo-European ancestor languages and all kinds of suggestions about long term contact etc.to explain similarities while the Slavic and Sanskrit similarities can survive thousands of years of separation from their common ancestors ?

Them meee said...

Because people don’t want to accept the Romana are related to the barbarian Celts? Or because modern Celtic languages are too different from Ronance languages in their eyes to form an Italo-Celtic clade and deny a common origin due to this? Maybe linguists are just really stubborn (they kind of are)? It’s almost like Germanic and Italic are the forgotten branches of Western IE, treated as almost unique.

Also don’t Indo-Iranian languages retain some obvious similar with Balto-Slavic?

Matt said...

@Ric Hern: Are there close similarities between Slavic/Baltic and Sanskrit?

My linguistics-fu is limited, but...

In cognates in the lexicon, it seems like there aren't in the Jena 200 "independently entered by experts" data or the tree models couldn't really work and turn out as they do (not very / particularly close within the family).

(I would tend to give more weight to that kind of mass analysis than expert argument than one set of languages has particularly close lexical relationships by recruiting a smaller set of favourite shared cognates and arguing for their importance. Those just seem like I can't subjectively say who is right or wrong. Opinion may vary.)

In the morphology and phonology, a paper I was looking at by Gray justifying their methodology states

"The experience of Ringe et al. (2002), in their (non-Bayesian) search for a “perfect phylogeny” for the Indo European family, is instructive here. Ringe et al. start out with the historical linguist’s instinctive preference for data characters in phonology and morphology. But they find only 22 or the former, and 15 of the latter, that they consider validly usable, and retain even “fewer morphological characters” in their follow-up study (Nakhleh et al. 2005). This not only makes for a very small data-set, but as Ringe et al. (2002) themselves admit, “the worst news is yet to come: the vast majority of our well-behaved monomorphic characters simply define one or more of the ten uncontroversial subgroups of the family, contributing nothing to their higher-order subgrouping”. In practice, the main higher-order nodes in their output trees turn out to be supported by just two, one or even none of these phonological and morphological characters. Most of the phylogenetic information ends up being provided by their lexical cognacy characters after all, thanks to the sheer number of them".

As Ned Pegler from Armchair Prehistory also stated more sympathetically I think on Ringe's tree and morphology and phonology - http://armchairprehistory.com/2018/01/30/words-and-rules-and-the-contrasting-family-trees-of-indo-european/ - "His (Don Ringe's) frustration was that there were loads of shared words between languages but not that many shared morphological or phonological features. Yet his hunch was that the latter features were more important than the words.".

NP's own list from combining phonological characters from Ringe and Gamkrelizde & Ivanov to beef up the number of characters and strengthen the analysis finds some similar features shared between Indo-Iranian and Slavic/Baltic, but not particularly many compared to say, Greek or Armenian - http://armchairprehistory.com/wp-content/uploads/2018/01/list.png. 12/44 characters shared II and Slavic, against 15/44 for Greek and II, and 9/44 for II and Armenian, 7/44 for II and Italic, 8/44 for II and Germanic, 8/44 Baltic and II, reading that. And then he does some tree modelling using those features. I believe he may read here from time to time and correct me if I've misrepresent him!

So possibly some of those morphological and phonological features between the families, but not an amazingly strong relationship? And no special relationship in the lexicon at all. Though I can't say 100% if this is a fair surmise.

Ric Hern said...

@ Matt

Thanks.

Matt said...

No problem Ric. To go on a bit more, because I fear I may have not done full justice, if you're interested in the Armchair Prehistory bit, def. recommend reading his full post, as it lays out issues and methodology in a straightforward way.

Particularly I mangled the bit about shared features a little. Actually the full 44 features are not available in all comparison, and there's a different subset of total comparable features per language. E.g. if you look here for example - http://armchairprehistory.com/wp-content/uploads/2018/01/comparison-chart-raw.png. Indo-Iranian has 23 possible features as indicated by the diagonal, of which Slavic shares 13 and so on, and so that gives rise to a normalized comparison chart - http://armchairprehistory.com/wp-content/uploads/2018/01/comparison-chart.png.

He then talks about trying to find a tree that is compatible with the most feature retention (most parsimonious set losses and gains of features is how I might put it) as distinct from just using the comparison chart as a set of distances. As well as taking into account mutually exclusive features that must have developed in different later branches.

Doing it the dumb and wrong way and just using the normalized chart as a set of distances in PAST3 and then computing the tree (results - https://imgur.com/a/PBbHcCa) seems to give basically the same tree phylogeny as his final preferred tree though - http://armchairprehistory.com/wp-content/uploads/2018/01/morph-tree.png. For what it is or isn't worth. Albeit he adds a connection between Indo-Iranian and Slavic-Baltic to take account for the centum-satem isogloss (now suspected to be an areal feature or independent change / retention and not a morphological or phonological sign of shared descent?).

(Going on further, compared to Ringe's tree -http://armchairprehistory.com/wp-content/uploads/2018/01/morph-tree-ringe-1.png - difference would be that emphasis Germanic as an outgroup to a core IE clade, while these trees emphasise connections between Germanic to Baltic to Slavic in morphology and phonology, as well as stronger connections between Indo-Iranian to Greek and Armenian.

While compared to lexicon based trees like Chang's - http://cdn.sci-news.com/images/enlarge/image_2516_2e-Indo-European-Languages.jpg (but Grey and Atkinson, etc. are all the same in basic phylogeny, differing only in timescales) - those place more emphasis on a) early split of Tocharian and then early split of Greek-Armenian-Indo-Iranian, rather than placing the former with Italo-Celtic and the latter together, and b) tend to put Germanic more with Italic-Celtic. As is clear in the rest of his post and the tree that aims to be maximally compatible with lexicon and morphology and phonology).