search this blog

Wednesday, December 26, 2018

Get the popcorn ready for 2019


Occasionally I'm criticized for focusing too heavily on the Proto-Indo-European (PIE) homeland problem. But there are two very good reasons why I focus so much on this issue. First and foremost, this blog is a platform for original research related to European genetics, rather than just a review blog, so what I focus on largely depends on the data that are released by various labs. And a lot of the data released over the past three years have been directly relevant to the PIE homeland riddle.

Secondly, I enjoy the fact that this is a fairly popular blog, so to keep it popular I have to produce content that attracts a decent number of people. The ten posts with original content that attracted the highest number of visitors over the past twelve months (around 10,000 or more hits each, which I suspect is pretty good for such non-mainstream content) are linked below. Eight of the posts are about the PIE homeland and/or the early Indo-Europeans.

Central Asia as the PIE urheimat? Forget it

On the doorstep of India

Graeco-Aryan parallels

Indo-European crackpottery

The mystery of the Sintashta people

Likely Yamnaya incursion(s) into Northwestern Iran

Indian smoke and mirrors

First real foray into Migration Period Europe: the Gepid, Roman, Ostrogoth and others...

Modeling genetic ancestry with Davidski: step by step

Some German guy once said...

So guess what? In 2019 you'll be seeing a lot more about the PIE homeland and related topics at this blog, and I offer no apologies for that. Indeed, I actually feel obligated now to stay in the PIE homeland debate to steer it as best as I can in the right direction.

But no matter what I do, I'll be going out of my way here to keep things fresh and fun. Even when I'm being brutal and vicious (not literally, of course, because I don't condone violence or threatening behavior) I'll try to make it a highly entertaining experience for almost everyone.

See also...

Big deal of 2018: Yamnaya not related to Maykop

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

108 comments:

sykes.1 said...

As an Indo-European interested in history and origins, all I can say is keep up the good work.

And Merry Christmas and Happy New Year to you and yours!

Matt said...

All the best for 2019! At the end of the day, this blog is your show, so you do you and focus on what you want to. It's frankly IMO not been a year of much fruitful advance on this IE problem*, compared to other data that's come out (beginnings of some East Asian data), especially in the publicly released and published data, but if anyone is looking for anything different, the data that's out there is out there for them to look at.

*When I think 2019, I think we now know that excess ANE in Asia in one for another allows Steppe_MLBA to act as an ancestor for populations where it looked wrong and Steppe_EMBA right, and we know Maykop looks wrong for a early Chalcolithic fusion of Maykop+EHG to "generate" Yamnaya (the Caucasus contributions must have come from some earlier population, less merged with the Anatolian genepool). That's really all the new light we seem to have on this issue in 2019.

Them meee said...

Glad to see I’m not the only one who thinks we’re starved of Indo-European data.

I hope much more data comes though. Don’t lose hope. And catch wind of rumors, because we’re that starved.

Or just desperate to see where IEs in Europe came from, and we were spoiled around 2015-2017.

We did get some huge stuff on Indo-Iranians, but next year or so could be bigger.

Andrzejewski said...

I am delighted to read that Maykop has been debunked as a source population for Yamnaya. I personally think that the latter was only one of related Steppe populations bringing PIE languages into Europe and Asia.

It is also intriguing to find out a substantial EEF ancestry in Yamnaya, which probably trickled slowly over 2 millennia via CTC from the West.

Arza said...

BAM files from Wang et al. are already online:
https://www.ebi.ac.uk/ena/data/view/ERP111919

Genomic history of the Greater Caucasus region
Here we present data from the accompanying paper in Nature Communications: Archaeogenetic studies have described the formation of Eurasian ‘steppe ancestry’ as a mixture of Eastern and Caucasus hunter-gatherers. However, it remains unclear when and where this ancestry arose and whether it was related to a horizon of cultural innovations in the 4th millennium BCE that subsequently facilitated the advance of pastoral societies in Eurasia. To address this, we generated genome-wide SNP data from 45 prehistoric individuals along a 3000-year temporal transect in the North Caucasus. We observe a genetic separation between the groups of the Caucasus and those of the adjacent steppe. The northern Caucasus groups are genetically similar to contemporaneous populations south of it, suggesting human movement across the mountain range during the Bronze Age. The steppe groups from Yamnaya and subsequent pastoralist cultures show evidence for previously undetected farmer-related ancestry from different contact zones, while Steppe Maykop individuals harbour additional Upper Palaeolithic Siberian and Native American related ancestry.

Bastian Barx said...

Thanks for a great blog. Always interesting to read.

Janko Raven Johnson said...

Thanks, Dave!

Al Bundy said...

Congratulations on the blog and another great year.We finally saw Hittite Adna and only more samples will clarify but a Balkan route for Hittite looks pretty ridiculous at this point.Also, what if Ukraine Neolithic and not Yamnaya was the steppe ancestor for Bell Beaker and Corded Ware as Chad Rohlfsen suggested?

Ric Hern said...

@ Davidski

Thanks. This blog keeps me going. It is almost like a X-Mass Present each time you post something. Heheheeh.

JuanRivera said...

Nice blog, nice work and you often get things right. I know you can keep it up in 2019.

Matt said...

Arza, damn yeah, good spot, upload date 19/12/2018, looks like.

Browsing for the paper at Nature Comm to see if released only found: https://nature-research-under-consideration.nature.com/users/37265-nature-communications/posts/34186-the-genetic-prehistory-of-the-greater-caucasus (Under consideration since June?).

Inadvertently pointed me to a new blogspot by Stephen Schiffels, group leader at the Max Planck (best known to me through the work on Anglo-Saxon, Roman era samples), on the human mobility and language dispersal work: https://medium.com/stephan-schiffels/on-human-mobility-and-language-dispersal-fa9f8b499c2e

bellbeakerblogger said...

Eurogenes is one of a kind and I know many hours go into it each week. Thanks for all the work, it’s only getting better!

Dragos said...

There is nothing wrong with your focus on I.E. - its extremely topical. It's also very emotional for many western Europeans (& their American kids), as their society is collapsing, they feel a need to at least claim ''we wuz Indo-Europeans''.
Anyhow, my only request is you live up to the blogs claimed name ''Euro''- genes, and stop appealing to Tengri-worshippers like JuanRiviera and Andrezjewski

mzp1 said...

I think Indo European would be way more popular with the mainstream if it was called something simpler like 'Eurasian'.

I mean everyone knows it came from India, no need to be so obvious about it with the name and all.

paleog said...

Merry Christmas and best wishes from Russia.Thank you for your blog!

Them meee said...

@mzp1

I hope you’re joking, but seeing your previous comments, I doubt it.

Rod Johnson said...

Thank you so much for the information and the hard work that it takes

JuanRivera said...

Tengri worshipper, seriously? I'm catholic, albeit a very liberal one, and that mention of Tengri is both out of place and time and an ad hominem.

JuanRivera said...

In addition, this era is increasingly globalized, so that is increasingly hard to adress Europe separately from Asia.

JuanRivera said...

Anyways, discussions should include data as much as possible.

Dragos said...

@ JR
Sorry I think I've had a few too many drinks. But seriously, as you say, maybe the blog should be more inclusive, more open minded and less of an ''R1b-club'' ? '
If some academics and posters have different perspectives, they shouldn't be charged with being nutcases or pandering to politics, as my counter-example demonstrated, it's not very nice.

Andrzejewski said...

@Dragos I didn’t know that my PIE ancestors were Tangri worshippers

Andrzejewski said...

I don’t believe that Yamnaya was the ONLY IE or PIE culture to spread its genes or languages; I think that lots of interrelated Steppe cultures were involved in the process of Indo-Europeanization.

JuanRivera said...

So, people at Molgen seem to be getting more hostile to international collaboration in ancient DNA, yet are finally accepting that Bell Beakers have steppe ancestry. Ancient DNA research without international collaboration leads nowhere. Also, research suggesting Iberia as a horse domestication center was incorrect.

JuanRivera said...

Ready for what the following year awaits.

Davidski said...

@All

One important thing that I forgot to mention in my post is that I expect everyone to be familiar with the blog rules and not break them.

So, especially for the new people here...

- no racial or ethnic taunts and insults

- avoid discussions with obvious trolls and/or mentally unstable people

- do some reading here before posting if unfamiliar with the relevant topics

More rules here...

New rules for comments

Anyone who consistently breaks these rules will be banned very quickly.

JuanRivera said...

Wonder when the horse paper will be published.

JuanRivera said...

Sure that 2019 is going to be a megayear for genetics, with all those paper abstracts we saw this year plus new ones.

Ric Hern said...

What will be especially interesting is if they can pinpoint a pure R1a Epi-Paleolithic or Mesolithic population somewhere similar to the R1b Balkan Mesolithics...and some Mesolithic samples of the Caucasus Piedmont Steppe and Lower Don.

Dragos said...

@ Andrzejewski
Yeah sorry to break it to you, your ice age ancestors spoke a now extinct paleosiberian language.
Subsequently your homeland became the centre of Turkic and Mongolian ethnogenesis

Ric Hern said...

And the whereabouts of Yamnayas elusive Brother...

Andrzejewski said...

@Dragos That’s what the Pre-Proto-Indo-Europeans spoke, a language which came from Lake Baikal and later became PIE on the Steppes. Whatever happened to their original homeland means squat to me. I am more interested in the Steppes. What I don’t like reading is when people misunderstand Wang et al. preprint and think that everything Indo-European has something to do with West Asia or the Caucasus instead of Eastern Europe.

If Steppe populations looked anything but European then most modern Europeans who carry 40% -50% of their aDNA would’ve looked East Asian. We all know it’s not the case.

Andrzejewski said...

@Ric Hern are you referring to Villabruna by any chance?

Andrzejewski said...

I am wondering where @Dragos thinks the PIE cake from? (Unless it might’ve developed completely independently on the Steppe itself). Last time I checked, Romanian was an IE language.

Andrzejewski said...

Not just “R1b club”: Steppe people were both subclades, R1b and R1a1. Both spoke dialects of an IE language. Former were Centum, latter were Satem.

Andrzejewski said...

R1a1 was indigenous to Ukraine post LGM. R* is from MA1/AG3.

Ric Hern said...

@ Andrzejewski

I'm talking about a whole R1a population not just singleton samples. At Karelia we see R1a and J. In Ukraine Mesolithic we see R1a, R1b and I2a. In Khvalynsk we see R1a, R1b and Q. So like I said something similar to Balkan Mesolithic with many R1b sample....

Dragos said...

@ Andrzecewski


I don't really understand half of what you're asking/ stating. But if I may propose what might seem to be a rather radical idea - following the evidence and reconstructing scenarios, instead of our picking our preferred narrative & bending the data to fit it, the former being the ''scientific method'', the latter being the ''R1-creationist'' approach.

With that in mind- the earliest I.E. languages do not have much steppe ancestry at all (sure, one of the Anatolian MBA's have 3% , and Myceneans have 5-10% which had been trickling in over thousands of years). And sure, there was an R1a founder effect in South Asia, but it's timing of 300 BC unfortunately does not make it particularly relevant for Indo-Aryanization. What binds these groups together ?
Well its obviously Iran-type ancestry which expanded along the northern frontier of the Mesopotamiam ''Great civilizations'' , and with that the prestige language amongst the ''semi-civilized'', haplotypically diverse and economically mixed border groups (as George pointed out).

I realise this might not sit well with you, because your (shallow) pride demands that it came from your R1 steppe ancestors. And of course, it's just a little unpalatable, even amongst the self-declared ''liberals'' here, that PIE came from the vicinity of northern Iran, visions of blonde Siberians -> steppe are much nicer, now matter how erroneous and foolish.

Sure, there's no Iran Chalc admixture in Yamnaya, but, frankly, I don't care about that. I won't discuss the specifics or Europe, because frankly, it's beyond all of you here, even though you styles yourselves as the Avant-Garde of Paleogenomics (LOL Mr BB blogger).

Ric Hern said...

@ Dragos

Where precisely is this Northern Frontier of the Mesopotamian "Great Civilizations" that you stated ? Was this not Hurrian Urartian and Kaskian etc. territories ?

Zarzian said...

Thank you for providing the place to discuss the PIE question using Archeogenetics data. I have visited this site multiple times a day for the last 5 or so years. It is amazing to read and learn from the many experts, past and present.

Jool said...

Indo-European coming from the steppe was already the leading theory before any genetic studies. It’s a quite straightforward explanation for a common language family from Europe to north India. It was vying against the other mainstream explanation: Neolithic expansion origin. Now that a huge genetic signal of 50% replacement is found in the ancient genes at the turn of Bronze Age, with specific signal from the steppe, you will have a hard time overturning the steppe narrative. Mind you, it was an unexpected signal. The linguists were supposing an elite mediated spread, not a majority replacement.
Now what I am readying my popcorn for in 2019, is the Reich paper about Spain. If they indeed sampled populations in Basque country and on the Mediterranean coast and found a 40% autosomal replacement and >90% for the Y, we might have of host of new languages coming from the steppe: Iberian, Tartessian, Basque. Hell, even Proto-Etruscan can come from the steppe. It’s not absurd to have a multilingual expansion. The Neolithic expansion from East Asia had Sinitic, Austroasiatic and Austronesian languages. The expansion from the Fertile Crescent had Semitic, Sumerian, probably Dravidian and whatever language the Anatolian and European first farmers were speaking. I think that having a multilingual steppe will calm the debate about ultimate origin of Indo-European. If you have one language you can still try to follow the footsteps to whatever place in the Caucasus or in the Taïga. If you have multiple footsteps, they are probably just lost in the mist of the Mesolithic.

Andrzejewski said...

@Jool improbable. Steppe pastoralists only spoke IE languages.

Andrzejewski said...

@Jool Every male in Spain, not just >90%, was wiped out by a Yamnaya/BellBeaker invasion 4500 years ago. It easily explains why Basques have a very high R1b.

https://elpais.com/elpais/2018/10/03/inenglish/1538568010_930565.html

The reason not EVERY man TODAY carries them R1b genes are due to later invasions: Romans, Germanic - chiefly Visigoths, Arabs and Berbers from North Africa, Phoenician traders and later immigrants.

Andrzejewski said...

CHG had something to do with Khvalynsk however because first we find Q1a1 Haplogroup (Maykop was rich in ANE), but also because CHG went up from virtually non-existent in Samara to 25%-30% in Khvalynsk

Samuel Andrews said...

@Davidski,

Are you able to work with the Wang 2018 BAM files?

BAM files from Wang et al.:
https://www.ebi.ac.uk/ena/data/view/ERP111919

JuanRivera said...

There was also maternal replacement, in which steppe people introduced H2a1, H6a, W, U2e, U4b, C4a3, C4a6, C5, etc. Not only is Iberia's R1b(xR1b-V88) steppe in origin, but also Iberia's R1a and Iberia's Q1a (seen as far south as Andalusia).

JuanRivera said...

I forgot mtDNA C1.

CleverPrimate said...

Your focus on the PIE question is the primary reason that I read your blog on a daily basis. I don't see any way to reasonably discuss the genetic makeup of Europe without addressing the PIE issue. I have been fascinated with the Indo-Europeans ever since I learned who they were. I have learned more about PIE and the people who spoke it from your blog than from any other source. Please keep up the good work.

Guy Tipton said...

Merry Christmas David,

Thanx for all the work you put into this blog. You're the Ancient DNA goto blog for many of us out here in intertron world. Here's hoping for another mind boggling year.

Cheers,
Guy

rozenfag said...

Interesting, it seems that popcorn thing has already started.

Anyways, would be interesting to see your analysis of Caucasus data. I hope New Year celebrations won't suffer because of this.

Dragos said...

@ Jool

-''Now that a huge genetic signal of 50% replacement is found in the ancient genes at the turn of Bronze Age, with specific signal from the steppe, you will have a hard time overturning the steppe narrative.''

A very Nordocentric & arrogant statement that shows utter ignorance & contempt for other regions. Here are your steppe values for Myceneans (5-10%), Hittites (0%), Indo0-Aryans (-15%;) minus because theyre 1000 years too late.
The big genetic shift in actual Indo-Europeans is Iran ancestry with Hg E-V13, J2a1, G2a2b.

- ''I am readying my popcorn for in 2019, is the Reich paper about Spain. If they indeed sampled populations in Basque country and on the Mediterranean coast and found a 40% autosomal replacement and >90% for the Y,''

That 40% figgure for Iberia is arbitrary. Its a made up statistic based on using GERMAN BB in the PLeft. You really ought to read supplements & data tables, or if not- Davidsky really ought to help the rest of you out. The reality of turnover in Iberia is 60-70%, essentially in line with the rest of Atlanto-Vasconica. Later, the majority of west Europeans language shifted to the Celtic & para-Celtic of east-central Europe, as the linguistic & historical data suggested.

@ Juan

''There was also maternal replacement,''

Exactly, as per above.


Davidski said...

@Dragos

Your estimate of steppe ancestry for the Mycenaeans is wrong, because obviously neither pure EHG nor Yamnaya from Samara migrated to Greece.

It's rather obvious that Mycenaeans have around 20% Europe_LNBA/Steppe_MLBA ancestry.

See here...

Steppe admixture in Mycenaeans, lots of Caucasus admixture already in Minoans (Lazaridis et al. 2017)

Dragos said...

Modelling Myceneans as LBNA steppe, or LNBA Central Europe is merely illustrative, but has the problems of artificially compacting steppe ancestry to a single event, or 'invasion horizon'' which did not in reality exist.

Rather, the steppe ancestry made its way to Greece over millenia and in waves and manner which remains to be determined, but most likely secondarily & indirectly. We can readily appreciate this due to the Wang paper showing a ''LN Greek'' looking Varna like, which you appear to be taking as LBA Steppe.
I do not think this should be news, as numerous people have already highlighted this on the Net.

Davidski said...

@Dragos

Of course you mean Varna outlier-like, not really Varna-like. The Varna outlier has a lot of steppe ancestry and is very similar to Europe_LNBA and Steppe_MLBA.

There is also evidence of unadmixed Steppe_MLBA people in what is now southern Bulgaria during the pre- or early Mycenaean period.

Although the few currently sampled Mycenaeans "only" have around 20% of this type of ancestry, the Crete Armenoi sample from Bronze Age Crete has much more, which suggest recent admixture into Greece, not gradual gene flow from the north over thousands of years.

So my point stands.

Orthogonal said...

" the steppe ancestry made its way to Greece over millenia and in waves and manner which remains to be determined, but most likely secondarily & indirectly. "

So you'd expect little variation between Minoan and Mycenae. And is that the case Dragos?

Samuel Andrews said...

@Dragos,

Steppe people moved in heavily into southeastern Europe. These movements were unrelated the R1b/R1a Bell beaker/Corded Ware movements that went into northern Europe.

-Single sample from Bulgaria dating 500 BC has 23% Steppe ancestry. Samples dating 1000BC in Croatia have 33% Steppe ancestry.
-All of the non-Lombard indviduals in the Lombard burial in southwest Hungary cluster pretty close modern Italians and all have at least 20% Steppe ancestry. They're best representatives we have of pre-Slavic people of Balkans.
-Modern Albanians & South Slavs clearly have pre-Slavic ancestry which already had lots of Steppe ancestry.

Southeast Europe was home to lots of IE languages: Illyrian, Thracian, Dacian, maybe more.

Samuel Andrews said...

@Dragos,
The Indo-Europeanization of Europe: the intrusion of steppe pastoralists from south Russia and the transformation of Old Europe
https://www.tandfonline.com/doi/pdf/10.1080/00437956.1993.11435900.

According to the Kurgan hypothesis, 'Kurgan wave I' happened in 4500 BC. Gimbutas was wrong that Kurgan people became elites all over Europe starting in 4500 BC spreading IE speech. However, she was right that they invaded and interacted with cultures on the west coast of the Black sea (Varna & others) in 4500 BC.

What, Gimbutas called "Kurgan wave I" explains the Varna_outlier. Varna_Outlier is the result of sporadic admixture via immigrations-raids-whatever not a full scale migration. Steppe ancestry didn't become normal in southeast Europe until a few thousand of years later.

Only, a migration of Kurgan people can explain Steppe ancestry in Myceneans.

Jool said...


@Dragos
thanks for the ad-hominem attacks. I did not even explicitly take a stance in the debate. I just remarked that when a theory is already mainstream and get unexpected strong support from a new field, overturning it is going to be hard.

40% autosomal replacement is a quote from Olalde
https://elpais.com/elpais/2018/10/03/inenglish/1538568010_930565.html
I would have a hard time reading the supplements of a paper which preprint is not public yet.

So what's your narrative Dragos ?


@Andrzejewski

Steppe pastoralists only spoke IE languages.

What's your evidence for that ?
And why does all the people of Spain spoke non-IE languages when the romans arrived, except for the Celtibers (Celts: a very likely Iron Age addition).

Davidski said...

@Dragos

There are good reasons why Lazaridis et al. didn't consider the idea that steppe ancestry entered the Mycenaean gene pool via gradual diffusion over thousands of years as a plausible hypothesis.

- the practically zero level of steppe ancestry in Minoans

- the much higher level of steppe ancestry in Crete Armenoi

If the arrival of steppe ancestry in the Aegean was a very gradual process, then we'd expect at least a few per cent of steppe ancestry in the Minoans, rather than the significant differences that we see between the Minoans, Mycenaeans and Crete Armenoi.

Dragos said...

@ Davidski

''There is also evidence of unadmixed Steppe_MLBA people in what is now southern Bulgaria during the pre- or early Mycenaean period.

Although the few currently sampled Mycenaeans "only" have around 20% of this type of ancestry, the Crete Armenoi sample from Bronze Age Crete has much more, which suggest recent admixture into Greece, not gradual gene flow from the north over thousands of years.

So my point stands.''

Yes, but how does that change the fact that steppe ancestry was present there almost 2000 years before the MLBA steppe migrant in Bulgaria ?
You seem be be hanging all your eggs in that one sample.
NB R1a-Z93 is virtually absent in Greece.

My figure of 10% was qouting ''raw'' Yamnaya ancestry, to be in line with Jool's original statement of 50% in northern Europe.

@ Orthogonal

''So you'd expect little variation between Minoan and Mycenae. And is that the case Dragos?'

Minoans were based in Crete, which is an Aegean island 300 km or so off the mainland.
The absence of steppe ancestry therefore makes sense, does it not ?

People migrate along familiar networks.
Crete and proto-Minoans were very much linked to the a maritime Dodecanese network, and certian parts of SW Anatolia, whilst the mainlad had been linked to the Danube region via highland pastoral systems since the LN (ones still used by Vlahcs to this day); as well as (other parts of) western Anatolia. Perhaps, the highland pastoralists could cross rivers and straights, but not open oceans.

@ Sam
Thanks, I know Dacians & Thracians quite well.
But I suspect youve missed my point. - We all know steppe ancestry is prolific but it is not self evident that it brought IE to Anatolia and Greece, or even Italy for that matter.
Please note: the male MLBA Brnze Age Croatian sample with lots of steppe is J2b2.

Dragos said...

@ Davidski

''There are good reasons why Lazaridis et al. didn't consider the idea that steppe ancestry entered the Mycenaean gene pool via gradual diffusion over thousands of years as a plausible hypothesis.''

Its because Wang data was not availabel to them
Anyhow, you seem to be arguing against direct data. You have no leg here

Dragos said...

@ Jool
The same methodology is used for the 'steppe Iberian BB"' in the 2018 publication by Olalde et al. It available on the net

Davidski said...

@Dragos

Yes, but how does that change the fact that steppe ancestry was present there almost 2000 years before the MLBA steppe migrant in Bulgaria?

Because you can't assume that every outlier sample is evidence of gene flow that had a perceptible population-level impact.

How do you know that the Varna outlier-like so called Neolithic sample from Greece contributed any ancestry to the Mycenaean gene pool?

If it did, then why didn't the Minoans have any of this type of ancestry? Are you claiming that the sea between the Greek mainland and Crete was an insurmountable barrier? That's obviously not true, otherwise how did steppe ancestry suddenly appear there during the Mycenaean period?

And if it did, then why wasn't steppe ancestry widespread in the Balkans until the Bronze Age rather than the Chalcolithic?

Dragos said...

@ Davidski

''Because you can't assume that every outlier sample is evidence of gene flow that had a perceptible population-level impact.


Yet you have made that assumption for Greek Peloponessus _outlier (the ''CHG shifted individual'') and are doing the same with Crete Armenoi. So lets assume ceteris peribus

''How do you know that the Varna outlier-like so called Neolithic sample from Greece contributed any ancestry to the Mycenaean gene pool?'''

Neolithic here coul dbe something like 3000 BC, same as the Pelop_outlier.
But the short answer is I don't know if she/ he made an impact any more than you know that Myceneans derive from that one Bulgarian -steppe outlier
haha


''If it did, then why didn't the Minoans have any of this type of ancestry? Are you claiming that the sea between the Greek mainland and Crete was an insurmountable barrier? That's obviously not true, otherwise how did steppe ancestry suddenly appear there during the Mycenaean period?''

Obviously not insurmountable but difficult without the know-how. By the Mycenean period, mainlanders had co-opted the Minoans trade-networks.

''And if it did, then why wasn't steppe ancestry widespread in the Balkans until the Bronze Age rather than the Chalcolithic?''

The Bronze Age goes for over 2000 years (3500 - 1000 BC)
In the EBA, there are plenty of individuals without steppe ancestry, and where present it was piecemeal
By the MLBA, Steppe ancestry had diffused more widely due to flourising trade and the passage of time.

Them meee said...

@Dragos

Just a reminder...

https://eurogenes.blogspot.com/2018/12/the-hajji-firuz-fiasco.html

Matt said...

I don't agree with any of Dragos' statements in this thread (esp. not his "political" digs), and Samuel seems quite right to me that the Balkan data seems to suggest its unlikely for steppe related ancestry to build up slowly over millennia.

But there is certainly a gap in sampling in Greece at the moment, and I don't think it's for sure that steppe ancestry couldn't be in the mainland in about the same amounts as in Mycenaean samples by even 2500 BCE and not influence Crete for another millennia. The Mycenaean samples are only the first we have to show steppe related ancestry, but we have no earlier samples since neolithic.

(One thing that might throw doubt on that would be the reported absence of steppe ancestry and presence of CHG related ancestry in Middle Bronze Age Sardinians (2100–1550 BC), suggested to be from the Aegean... but on the other hand that can't be much of ancestry, judging by how Sardinians aren't exactly Minoan or Mycenaean like today. And so it might be hard to tell the difference between a Minoan or Mycenaean like proximate source, might be that MBA Sardinians could equally have been affected by a Minoan or Mycenaean like wave, as in either case in "ultimate" terms they're mostly Neolithic like (79-84%) with some CHG (15%), with about 4% EHG in Mycenaeans).

Davidski said...

@All

Apparently, the Eneolithic steppe samples from Wang et al. belong to R1b-V1636.

https://www.yfull.com/tree/R-V1636/

This is the same subclade of R1b that is found in the late Kura-Araxes sample, which is strong evidence that there was gene flow from the steppe into Kura-Araxes.

Them meee said...

@Davidski

Too bad someone will turn this into a Hajji Firuz Two.

But that’s huge either way.

Dragos said...

@ Matthew
LOL. Take a look at steppe admixture vs time-assorted-individuals

https://ibb.co/3YbVc93

There are distinctive and several 'blips' followed by redistribution. This process occurrs over 3000 years.

Around 45-4000 BC there is 5-10% average.
There is a big missing of sample between 4000-3000 BC
Then 3000-2000 BC (although biased sample set from Bulgaria) average of 20-30%
Post 2000 BC (biased toward Croatian set) average of 30-40%, but in males with non-R1. (probably due to northern Europeans trying marry off their daughters to J2 and G2 elites)

But please dont get me wrong, the steppe input was much appreciated.

Also, the Balkans had numerous IE groups often quite divergent - Illyrians, Thracians, Dacians, Phrygians, Greeks, and other lost groupings.
What did NW Europe have ? Celts in the Iron Age.

Andrzejewski said...

I hope that some people wouldn’t misinterpret the movement in the opposite direction.

Davidski said...

@Dragos

There's no plausible way to explain the rather sudden appearance of widespread steppe ancestry in the Balkans during the Bronze Age than to say that it, well...appeared there rather suddenly during the Bronze Age.

Moreover, its rather sudden appearance there at that time suggests that there were migrations from the steppe into the Balkans during the Bronze Age, and this simply corroborates the steppe hypothesis about the spread of Indo-European languages into the Balkans based on linguistics and archeological data.

Nothing that you've said in this thread will change my mind, or the minds of the vast majority of the commentators and visitors here, about this.

To change our minds you would need to produce convincing new data that contradicts the presently available data in that it shows the slow build up of steppe ancestry in the Balkans prior to the Bronze Age. So till you have that contradictory new data, you might want to save your energy here and enjoy the holidays.

Dragos said...

Yes movements occurred from steppe all the time, over thousands of years, and this ancestry built up gradually over 3000 years/
But were they IE ? Looking doubtful Im afraid.
in fact, they seem to have got their ass handed to them in Greece.

Andrzejewski said...

@Dragos "Yes movements occurred from steppe all the time, over thousands of years, and this ancestry built up gradually over 3000 years/
But were they IE ? Looking doubtful Im afraid.
in fact, they seem to have got their ass handed to them in Greece."

???????????????????????????????

Ric Hern said...

@ Dragos

By this time it is clear that you for some or other reason dislikes the West especially Northern and Northwestern Europeans. Everything you said had a smell of that and for me personally it seems like your hypothesis is built upon that...

Ric Hern said...

@ Dragos

Did you ever consider that Steppe people did not see Greece as a Tourist destination before Sun Block Cream was invented ? Heheheeh...

Bob Floy said...

@Everyone

You might as well stop arguing with Dragos now, he knows that all you want to do is fantasize about blond supermen conquering the world on horseback, and he's not going to discuss the specifics of Europe, because, frankly, it's beyond all of you. We should all just be grateful that he deigned to come here and grace us with his presence for the last few days. Now let's all go back to our Tengri shrines and have a nice cry.

Ric Hern said...

@ Bob Floy

Whahahaha.. !!!

Dragos said...

@ Ric

I dont follow your riddles. Nor, actually, my best trade partners are from NW Europe.
I guess just you particular guys here seem to be having a problem with 'facts.
The data really isnt doing you any favours.

Population (averg) Balkans_N Yamnaya_Kalmykia Anatolia_BA
Balkans_ChL (4500 :400 BC) 90.4 || 9.6|| 0
Bulga_EBA (3000-2000 BC) ||68 || 32 || 0
Croat_MLBA (1600-800) ||61.2 || 38.6 || 0.2
Balkans_IA ||54.6 ||21.4 ||24
Mycenaean 41.4 ||10.4 ||48.2


Obviously Sam & Matt are wrong.

10% Steppe in Chalcolithic
30% in EBA, 40% in MBA
What will 4000-3000 BC when samples come ? Do we bet about 20% ??


Now, English is not my first language, but that seems like a G-R-A-D-U-A-L increase to me

I mean, we even have samples from Late Neolithic Greece showing steppe in ADMIXTURE, and you guys are sayin "No. It all came at once''

Yet for some reason, when (the real) IE start appearing in Europe, steppe ancestry drops . HHmm.

Orthogonal said...

"Minoans were based in Crete, which is an Aegean island 300 km or so off the mainland.
The absence of steppe ancestry therefore makes sense, does it not ?

People migrate along familiar networks.
Crete and proto-Minoans were very much linked to the a maritime Dodecanese network, and certian parts of SW Anatolia, whilst the mainlad had been linked to the Danube region via highland pastoral systems since the LN (ones still used by Vlahcs to this day); as well as (other parts of) western Anatolia. Perhaps, the highland pastoralists could cross rivers and straights, but not open oceans. "

Sorry, I must have been thinking wrongly because I looked at Google maps and it shows Crete much closer to mainland Greece than 180km, so it must floated over in the past couple thousand years to about 50 miles off from the nearest headland on the coast in that Mediterranean Ocean there.
I would have assumed if steppe people were gradually flowing into the area for millennia as you say that a short hop of water wouldn't stop gene flow to getting to the non steppe Minoan but obviously they were hygrophobic.

Orthogonal said...

^Make that: "much closer than 300km"

Dragos said...

@ Orthogonal

Maybe they were
Anyhow, your arguements are rather moot- we know steppe ancestry reached Greece thousands years before 1300 http://eurogenes.blogspot.com/2018/07/a-corded-ware-related-proto-greek-from.html
I presume when the full paper is release, we'll know its exact date.

I'm sorry that somehow sours the Steppe MLBA narrative you fellas are running, and kidding yourselves that your going to somehow sway the professionals LOL
I really don't udnerstand the dishonesty here. I guess my tin-foil hat was right.

epoch said...

@Dragos

"Yet for some reason, when (the real) IE start appearing in Europe, steppe ancestry drops"

You compare Bulgaria with the Peloponnese to make this claim. That is intellectually dishonest, to put it mildly. If you need this kind of data manipulation to make your point you don't have a point.

Matt said...

Mathieson's own words on the topic are:

"However, in two directly dated individuals from southeastern Europe, we find far-earlier evidence of steppe-related ancestry (Fig. 1b, d). One (ANI163) from the Varna I cemetery was dated to 4711–4550 bc and another (I2181) from nearby Smyadovo was dated to 4550–4450 bc. These findings push back the first evidence for steppe-related ancestry this far west in Europe by almost 2,000 years, but it must have been sporadic because other Copper Age (approximately 5000–4000 bc) individuals from the Balkans have no evidence for such ancestry. Bronze Age (approximately 3400–1100 bc) individuals do have steppe-related ancestry: we estimate that they have about 30% (confidence interval: 26–35%), with the highest proportions in the four latest Balkan Bronze Age individuals in our data (later than roughly 1700 bc) and the least in earlier Bronze Age individuals (3400–2500 bc; Fig. 1d)"

I don't think it's totally honest to aggregate the samples together as showing a trend of increase within the Balkans Chalcolithic and an aggregate of 10% in Chalcolithic enriched to 30% in the Early Bronze Age. There's nothing like that in the temporal series data within the Chalcolithic series (which is fairly well populated), and quite a few data points to suggest unadmixed individuals in the Balkans even within the Vucedol and the Early Bronze Age that argues against it.

Dragos said...

@ Matt

''I don't think it's totally honest to aggregate the samples together as showing a trend of increase within the Balkans Chalcolithic and an aggregate of 10% in Chalcolithic enriched to 30% in the Early Bronze Age. There's nothing like that in the temporal series data within the Chalcolithic series (which is fairly well populated), and quite a few data points to suggest unadmixed individuals in the Balkans even within the Vucedol and the Early Bronze Age that argues against it.''

People aggregate samples all the time, Davidski, academics, alike
The Balkan Chalcolithic aggregate set i composed excludes Smyadovo and Varna outliers, not to skew the data set with too much steppe. For transparency I2509-I0785, most dated 45-4000 BC

Of course, there is individual variation, but there is nothing wrong with looking at averaged trends, so your claims of of dishonesty are absurd.

Chalcolithics (overall).
Balkans_ChL
Balkans_N 90.4
Yamnaya_Kalmykia 9.6

Bulgaria EBA (33-2000 BC)
Balkan Neol 68
Yamnaya 32

Croatia MLBA
Balkan Neol. 61
Yamnaya 39

So steppe admixture pattern:

4500-4000 BC : ~10%
4000-3300 BC : ???
3300-2000 BC: 30 %
2000 - 800 BC : 40 %


???= Cernavoda, Cotofeni, Hotnica_Vospada
Well, Im sure you can do the math.


@ Epoch
I'm not sure what your ranting about.
It's entirely appropriate to analyse Thrace with the Aegean because they both experienced the same movement from Anatolia c. 2500 BC., although you appear to be ignorant of this

But Ill outline what is dishonest, your suggestions that a female in Varna-Karanovo VI culture grave represents ''proof'' of steppe invasions of Anatolia; or your plea that all R1a in Bronze Age India went up in smoke :)

epoch said...

@Dragos

"It's entirely appropriate to analyse Thrace with the Aegean because they both experienced the same movement from Anatolia c. 2500 BC., although you appear to be ignorant of this"

If that were the case, the why isn't there any steppe in Anatolia Chalcolithic? According to your idea steppe gradually increased in Bulgaria, which according to you apparently must mean that it did also in the Peloponnese. So why didn't it gradually increase in the must closer Anatolian samples?

Dragos said...

@ Epoch

By gradual i mean it was a long process, overall, throughout the Balkans. Bulgaria experienced more forceful pulses because much of it is in the steppe, thus stands apart from Dalmatia or Greece.


Whilst steppe admixture continued to increase in the northern Balkans (think of it as the inverse correlate of increasing MNE back in CWC & BB), this process was holted in Greece and certain parts of Thrace due to the a reversal of flow - a migration from Anatolia to Thrace and the Aegean. This has been well described since the 1900s.

And note- the current Bulgaria BA data set does not depict this process because theyre specifically chosen (almost exclusively) from the ''Barrow cutures'', not the other 3 or 4 macro-groupings which also existed there, especially the elites associated with hoards and citadels.

Also, note we only have 1 (propper) sample from west Anatolia post-4000, so future samples might show some steppe admixture.

So feel free to interpret the data in a different way, that's entirely your right, but you shouldn't accuse of dishonesty.

epoch said...

@Dragos

"Bulgaria experienced more forceful pulses because much of it is in the steppe, thus stands apart from Dalmatia or Greece."

Exactly. Therefore it is intellectually dishonest to use those levels to measure a perceived *decrease* of steppe at the onset of the IE languages in the Peloponnese. You could have suggested it, state that it was possible. But it is by no means any proof of it, for exactly the reason you stated there.

Them meee said...

@Dragos

So what do you think steppe spoke, or was more closely related to?

epoch said...

@Dragos

"The Balkan Chalcolithic aggregate set i composed excludes Smyadovo and Varna outliers, not to skew the data set with too much steppe"

BTW, did you aggregate the that Balkans Chalcolithic? I mean, you show some nMonte runs and David has a data sheet with Balkans_Ch in it. When I run it I get this:

[1] "distance%=1.1592"

Balkans_ChL

Balkans_N,90.4
Yamnaya_Kalmykia,9.6

I have a hunch you didn't compose anything.

Dragos said...

@ Epoch

''Exactly. Therefore it is intellectually dishonest to use those levels to measure a perceived *decrease* of steppe at the onset of the IE languages in the Peloponnese. You could have suggested it, state that it was possible. But it is by no means any proof of it, for exactly the reason you stated there.''

okay ill rephrase it.
Why was steppe ancestry steadily but surely increasing throughout the Balkans, and even reaching Greece (by virtue of the mysterious ''Corded Ware like LN Greek''), yet a relatively large migration of Hurrians appears c. 3000-2000 BC in those parts were the earliest IE-speakers of Europe appear.
A slightly confounding pattern.

''BTW, did you aggregate the that Balkans Chalcolithic? I mean, you show some nMonte runs and David has a data sheet with Balkans_Ch in it.
I have a hunch you didn't compose anything''

No I never claimed i invented the G25
Davidski's data set has averaged Balkan-Chl as well as individuals, right ?
But i have no idea which actual samples he's averaged out to get that.
So to be sure that it doesn't have outliers, I pooled ''regular'' Chalcolithics together and averaged them using a simple function command. I just spelled this out above, try reading carefully next time.
It turns out theyre the same as Davidski's, so he used a the very same data set.


c) ''When I run it I get this:

[1] "distance%=1.1592"

Balkans_ChL

Balkans_N,90.4
Yamnaya_Kalmykia,9.6''

The same as me ? Terrific.

Dragos said...

@ Them meee
I must eat my words and bow down, because Jool made some very clever remarks.

epoch said...

@Dragos

"Why was steppe ancestry steadily but surely increasing throughout the Balkans, and even reaching Greece (by virtue of the mysterious ''Corded Ware like LN Greek''), yet a relatively large migration of Hurrians appears c. 3000-2000 BC in those parts were the earliest IE-speakers of Europe appear."

Even rephrased it means your graph is meaningless because you compare regions that, by your own admission, are incomparable.

The only place where a language shift is seen, Lineair A to proto-Greek, is Crete and there it is clearly accompanied with steppe. Your theory would require Lineair A to be IE and while that has been suggested most people consider it unlikely. But even in the unlikely case that Lineair A was an IE language we can be absolutely sure it isn't Greek (Because Lineair B was). And therefore the shift to Greek is highly correlated to steppe admixture.

But it looks like you already noticed yourself that another language can be associated with the uptick of Iranian admixture: Hurrians.

Dragos said...

LOL. I did not state theyre not comparable. Theyre Perfectly comparable because the same processes were occurring in both regions. Obviously there were no Hurrians in Greece, so do pick up on parody next time. As others have told you, your theory & approach is rather simplistic.
How do you know Minoan, if indeed non-IE, was not the Native Neolithic language ? And there were multiple streams of migration across the regions.
In fact, the works of Balkan linguists have pointed out the huge stratigraphic complexities of langauge strata including possible pre-Greek IE, Anatolia-type IE, and a multitude of other IE langauges.
But one cannot and never will grasp these if the basis of their knowledge is ad-hoc anecdotes.

Look, ill elaborate - there's a reason why there's a precipitous drop in steppe ancestry

''Material culture of the Yamnaya group is still the dominating factor in assemblages dated to the first quarter of the 3rd millennium BC, and one might well envisage a relatively balanced exchange system to have been in place and a kind of symbiosis to have been established between predominantly agricultural societies using Ezero A and B1 material culture and inhabiting tell sites, and the pastoralists using Yamnaya material culture that were living in the wider landscapes around. However this situation changes in the second quarter of the 3rd millennium when Yamnaya-type kurgans and burials sharply diminish in numbers, and Ezero Tell sites expand regionally and locally.''

Why ?

'' try to explain this situation, firstly by highlighting and contextualising the dataset of several lavishly equipped graves and hoards, and prestigious and exotic single finds of this period, discovered mostly in Bulgaria in the last two decades. It will then describe the outstanding fortified settlement sites of Kanlıgeçit, Selimpaşa and Mikhalich, foreign in their design, construction and material culture compared to local settlements, ''

It becomes a little hard to imagine that Myceneans elites which emerge after 2000 BC derive from a group which appears to have been vanquished 500 years earlier, whilst the new incoming elite assumed a 'substratum'' status.
But we can all agree that it's not impossible. So let's wait and see

epoch said...

@Dragos

"I did not state theyre not comparable. Theyre Perfectly comparable because the same processes were occurring in both regions."

You already made clear with this Remark: Bulgaria experienced more forceful pulses. If you don't understand that this means that graph is meaningless because the start conditions in the Peloponnese are different it's worse than intellectual dishonesty.

"How do you know Minoan, if indeed non-IE, was not the Native Neolithic language ? And there were multiple streams of migration across the regions."

You fail to see the point. Whatever the language was, it wasn't Lineair B. So we are sure the change from Lineair A to Lineair B was a language change. We also are sure that samples from the time of Lineair A show no steppe admixture. We are also sure that samples from the users of Lineair B do. And we know that a Pelloponese sample of Mycenaeans do. So in Crete we see language change to Greek strongly associated with the coming of steppe admixture.

Could you, if you quote something, put a link with a source in the post? Mind you, the sample of Ezero and contemporary EBA samples are from the period in your graph where steppe admixture is actually ramping up. Some of these samples have similar Y-DNA as the Bulgarian Yamnaya sample.

Dragos said...

@ Epoch


The hyperlink failed, for some reason, so here's the full title (Kanlıgeçit-Selimpaşa-Mikhalich and the Question of Anatolian Colonies in Early Bronze Age Southeast Europe)
Ezero, looking at the ADMIXTURE graph from Mathieson, has a a bit of steppe ancestry, not surprising. But it's coverage was very poor, and as a lone sample not much more can be said. In fact, archaeologists are still not sure about the origins of Ezero, because it appears after a several hundred year hiatus.

epoch said...

@Dragos

This is what Mathieson states on the Bronze Age samples, of which the Ezero sample is part:

"Bronze Age (approximately 3400–1100 bc) individuals do have steppe-related ancestry: we estimate that they have about 30% (confidence interval: 26–35%), with the highest proportions in the four latest Balkan Bronze Age individuals in our data (later than roughly 1700 bc) and the least in earlier Bronze Age individuals (3400–2500 bc; Fig. 1d)."

That doesn't sound like a Ezero was a successful migration wiping away the gradually increasing steppe admixture. It sounds more like more than one migration from the steppe influenced Bulgaria. Which fits the theory that both Greek and Anatolian languages originally came from steppe populations in separate migrations over the Balkans.

epoch said...

@Dragos

Thanks for that article, by the way. Hugely interesting read. These settlements seem connected to Ezero but also Anatolia and the suggestion that Efe's Great Caravan Route has something to do with it made me think how ideas and migrations exchanged on the Silk Route: Languages such as Turkish went West, whereas religions such as Nestorism and Islam went East.

Arch Hades said...

Do we even know a Minoan like population ever even existed on the Greek mainland? Basically a population who derives 80% of it's ancestry from Neolithic Anatolian and with the remainder being CHG...but who totally lacked Steppe admixture. I'm not sure a population like that ever existed in the Greek mainland. It could be steppe admixture came simultaneously with CHG ancestry in mainland Greece.

Another thing, the Mycenaeans were modeled as having 13-14% Early Bronze age steppe admixture or alternatively 20% Middle-Late Bronze age steppe admixture. But any population with high Neolithic Anatolian ancestry will score higher in MLBA steppe than EBA steppe..as the MLBA steppe component is just mostly EBA steppe mixed with some Central European farmer ancestry. In my opinion it's unlikely 20% of the ancestry of the Mycenaeans will literally stem from the LMBA steppe.

Davidski said...

@Arch Hades

Do we even know a Minoan like population ever even existed on the Greek mainland? Basically a population who derives 80% of it's ancestry from Neolithic Anatolian and with the remainder being CHG...but who totally lacked Steppe admixture. I'm not sure a population like that ever existed in the Greek mainland.

Yes it did.

Peloponnese_N I3920 clusters with Minoans and is dated to 3933-3706 calBCE, so by the Minoan period, around 2400-1700 BCE, there may have been more CHG in parts of mainland Greece than in Crete.

But this was rather obvious anyway, because the Mycenaeans could be modeled very well as Minoan_Lasithi and Europe_LNBA/Steppe_MLBA.

Read this and have a close look at the PCA. Peloponnese_N I3920 is the most eastern Peloponnese_N sample, but almost all of the Peloponnese_N samples show significant excess CHG without anything from the steppe.

Main candidates for the precursors of the proto-Greeks in the ancient DNA record to date

Slumbery said...

@Davidski

Where is that 20% coming from? An estimation for an upper limit? I tried this with G25 nMontes (yes, I have a hammer and...) and could not get more than 14% for any of the samples.

Also, how much of it is actually CHG and how much of it is Iran? nMontes results suggest it is a mixed bag with more Iran than CHG. You usually emphasize that these are not the same thing.

Davidski said...

@Slumbery

The point I was making was that Peloponnese_N I3920 was very similar to Minoan_Lasithi. I don't know the proportions of CHG for each of them.

But I do remember that when modeling them as Barcin_N and something eastern, both came out mostly Barcin_N plus around 20% Armenia_EBA. I can't check this right now, but try these reference pops and see what you get...

Armenia_EBA
Barcin_N
CHG
Ganj_Dareh_N
Tepecik_Ciftlik_N

Happy New Year!

Slumbery said...

@Davidski

I already did some tests, that is why I asked. The online nMonte accept only four reference populations at once, but of course it is not always necessary to include all five in the same test for meaningful results.

Global 25 nMontes Minoan_Lasithi
Barcin_N 87.5%
CHG 3.33 %
Ganj_Dareh_N 8.33 %
Tepecik_Ciftlik_N 0.83 %
Fit 1.8195

Global 25 nMontes Peloponnese_N_o
Barcin_N 76.67 %
CHG 0.83 %
Ganj_Dareh_N 12.5 %
Tepecik_Ciftlik_N 10 %
Fit 2.0304

Now, I am not sure Armenia EBA is OK to use in such a test. It is much younger than the other references (even much younger than the tested Peloponnese_N) and have ancestries from all over the map, including from Neolithic Anatolia (and a few percent from the steppe even). Here they are anyway:

Global 25 nMontes Minoan_Lasithi
Armenia_EBA 5.83 %
Barcin_N 85%
Ganj_Dareh_N 7.5 %
Tepecik_Ciftlik_N 1.67 %
Fit 1.8182

Global 25 nMontes Peloponnese_N_o
Armenia_EBA 10 %
Barcin_N 74.17%
Ganj_Dareh_N 8.33 %
Tepecik_Ciftlik_N 7.5 %
Fit 1.7411

But:
Global 25 nMontes Armenia_EBA
Barcin_N 10 %
CHG 35.83 %
Ganj_Dareh_N 19.17 %
Tepecik_Ciftlik_N 35 %
Fit 2.8478


Happy New Year for you too!

Although I am not in Australia, so I am still deeper in the old year. I won't be here at midnight :)

Davidski said...

@Slumbery

Those are fairly similar results for eastern admixture for Minoan_Lasithi and Peloponnese_N_o in your models, and overall they reflect the PCA results.

The discrepancies you're seeing are due to overfitting, and it doesn't matter that the CHG figures don't reach 20%.

Matt said...

@Arch Hades, could even be that a wave of Anatolia+CHG ancestry without any steppe ancestry came after steppe ancestry arrived in Greece, and reduced the proportions of steppe ancestry somewhat. Would fit with the dominance of J2 haplogroups I guess?

I think that's difficult though in that it requires either a relatively "pure" source of CHG ancestry or a far larger pre-CHG rich steppe component than the Mycenaean samples had, and that's not yet so clearly demonstrated.

Happy New Year all!

Bob Floy said...

Happy new years to all, and thanks to Davidski for the valuable service that he provides.

Dragos said...

@ Davidski

Well, CHG type acnestry was even present in 5000 BC Bulgaria (Krepost).

You often highlight the Greek Pelo outlier, but there is also the contemporarnoues Peloponnese_N:I3709 which has much less (the same as the other Greek Neolithics). So it is difficult to elucidate representativeness.

I think the philosophy here is to blanket categorise anything CHG or Iran related as non-IE, and furthermore link it into a Minoan-Hurrian stratum, or the like.
However the problem is that it is difficult to distinguish between all the West Asian ancestries with nMOnte or qpADM.

Therefore there are some important points to consider.

* Minoans are predominanly Greek-Aegean Neolithic in origin (70s % vs 30-40%) and (more speculatively) their language might be too.

* The post-Anatolia/ CHG/ Iran ancestries in Minoans & Myceneans is different. Mainland Balkans/Greece is from colonists with a ''Trojan'' material culture, whilst Crete (Minoan) is from SW Anatolia, Dodecanese, etc

These are important differences to consider.

Davidski said...

@Dragos

Peloponnese_N:I3709 certainly has more CHG/Iran_N ancestry than any of the other Greek Neolithic samples.

Not as much as Peloponnese_N_o:I3920, but that's OK, because it seems like this is roughly the time when the CHG/Iran_N ancestry began to really stream into mainland Greece (and IMO likely also into Crete).

Peloponnese_N:I3709
Barcin_N,91
CHG,6
Ganj_Dareh_N,3

Peloponnese_N_o:I3920
Barcin_N,83.6
Ganj_Dareh_N,10
CHG,6.4

I'm not sure how you came to your conclusion that Peloponnese_N:I3709 was a standard Greek Neolithic sample, because Klei10, Rev5 and Peloponnese_N:I5427 are practically 100% Barcin_N, while the rest of the Peloponnese_N samples show varying amounts of CHG/Iran_N.

The reason I pointed out the excess CHG/Iran_N in Peloponnese_N_o:I3920 is because Arch Hades was under the impression that such people didn't exist on the Greek mainland. So it's important to note that obviously they did.

Slumbery said...

@Davidski

"The discrepancies you're seeing are due to overfitting, and it doesn't matter that the CHG figures don't reach 20%."

I am not sure what you mean by discrepancies, the tests came back with fairly expected and consistent results.
There is surely over-fitting, but Armenia_EBA is not correct even for a two-way model. You cannot estimate CHG+Iran_N admixture into an Anatolia Neolithic base population by a reference population that is itself almost half Anatolia N. At best it gives an upper limit estimation.
BTW, he groups that brought CHG/Iran_N admixture into Peloponnese_N were probably not very similar to Armenia_EBA, because Armenia EBA is much more CHG shifted (vs. Iran).

I agree however that is not very important CHG+Iran_N does not actually reach 20%, it is just arbitrary round number that came up. This kind of admixture was present and significant (and there was no steppe with it).