As someone said that I am “some blowhard nobody”, but all what I wrote about Hg. R1, its origin and its expansion, is proved so far (look at the YFull tree and all is clear), I want to remember another point of my theory, that R-L23-Z2103 present massively in Yamnaya (and which is also my paternal line) was due to a migraton from West, and not being anyway my opponents able to demonstrate that R-L23-L51 did come from Yamnaya (but it is now discounted that they were the Bell Beakers), my theory about the Indo-European languages is that the oldest are the centum ones of the Alpine Region / Central Europe, and only the satem ones expanded from Yamnaya. I know that the question centum/satem is debatted (and a great contribute gave the mourned Woudhuizen), but centum are older and satem are younger.
It looks more-or-less correct and a great little summary of the genetics with the caveats that:
1) Genetic pulses are very different sizes (e.g. barely any West Siberian into Sintashta; high GAC into European farmers). 2) It’s still also shadow-y what happens with the Balto-Slavic group; there are indications of a high-HG ancestry subpopulation that may be important for them (e.g. there are many samples from East-Central Europe from Latvia as far down as Serbia that have elevated HG proportions and preferentially seem to share genetic drift with later Slavic and Baltic speaking populations today), and possibly Balkan farmers too. 3) Caucasus farmers and Iranian farmers are necessarily a bit more vague. This will get nailed down a bit. 4) Greek must be impacted by post-Neolithic groups in the Aegean, because the sampled people we know must have spoken Greek (not some other related para-Greek language) have related Minoan ancestry rather than just Greek Neolithic. 5) Not so clear whether CWC is a branch from Yamnaya or some slightly earlier cultural stage. Closely related tho. 6) Very importantly, there probably were continuous mutual influences between all these IE languages so long as they remained geographically close enough, esp. in the early stages of divergence. So proto-Armenian might have some influence on the language spoken in Iran when the true Iranians showed up, proto-Greek and proto-Balto-Slavic and proto-Indo-Iranian continuing to be in some mutual influence etc. There’s probably continuous horizontal influence in the true phylogeny, on some level. It’s not necessarily the case that the Yamnaya and Corded Ware expansions really left totally separate branches of IE linguistically because of this horizontal transfer; we can’t necessarily say something like people “Speak a descendent of a Corded Ware language, not a Yamnaya one” linguistically because the sound changes etc could spread, without this being just an approximation.
Of course you know all this already.
Matt’s second point is kinda underrated I feel so I’d like to emphasize it too. A few people have brought up this attraction Balto-Slavic speakers seem to have to certain peri-Carpathian Neolithic groups in the Eurogenes’ PCA-based method. We’ve discussed that point on here before as well. Much like the pulse of Western European Neolithic in the scheme, there probably needs to be an extra pulse from certain non-GAC (or at least what we have from it so far) eastern populations into Balto-Slavic, though hopefully that seemingly somewhat controversial and formally less explored point should be clarified more in the future…
I suppose with the Yamnaya-like Sredni Stog samples that Anthony showed, the model might also look as Sredni Stog -> Yamnaya and Sredni Stog + something HG-rich -> Corded Ware, per point (5). Or there’s already a bit of a cline in EHG-CHG proportions in the area like we see with some Sredni Stog samples deviating towards the ‘north’/HG zone, though all current early CW samples seem to also have that slight western, not just northern, pull so it’s also plausibly acquired from more western populations only after the separation of the two.
@Matt When you say CWC didn’t speak a language distinct from Yamnaya language do mean the distinction is below the language level and only on a dialect level? Like Punic and Hebrew and Edomite.
Where does Anatolian fit into your diagram?
@DaThang, I don’t think I’ve necessarily said that here, although I do think I said in the past that if there are relatively close relatives found within Corded Ware and Yamnaya (if that set of information, which isn’t even a preprint yet, still holds up), and they’re autosomally so similar, then likely they didn’t diverge very long ago or completely and the language would be likely not to be very different? It does seem most likely to be on a dialect level, I guess in those terms.
Albanian and Greek are descended directly from the Yamnaya. We see a cline from Yamnaya, all the way to Greece_N, and Albania_N. With samples like Yamnaya_Bulgaria_o, and Logkas_MBA intermediary. Mycenaean clearly on the cline as well closer to the original Neolithic people in Greece. I think it is possible this same dynamic played out in southern Italy as well, during the Bronze Age. My own halopgroup is from the Yamnaya in the Balkans EBA, which also shows up in Mycenaean Greece in the Palace of Nestor.
Mostly agree but positioning Bell Beaker as derived from Corded Ware rather than parallel to it seems off.
I also note the omission of Anatolian languages. We may not have a ready solution for their origins that is widely accepted, but they really do belong in the chart, if only to have a question mark with no lines in or out of that box.
Sorry, but I’m quite lost in the graph here. I always thought that the bell Beakers did not come to form the Corded-Ware, but were a parallel population (also coming from the Yamnaya) to the Corded-Ware. Am I right or wrong?
@Roberto Santos
Strictly speaking, CW and Yamnaya themselves likely share a recent common ancestor, not necessarily CW deriving directly from Yamnaya. Unpublished Sredni Stog samples that Anthony has shown plot exactly like Yamnaya on PCA, with our earliest/most steppe-rich CW samples deviating towards HG-rich populations to the west/north in comparison. We have found early CW populations rich in the western R1b lines now in the east. It does seems plausible now that Beaker arose as a local northwestern variant of CW subgroups rather than from a separate migration from the east overtaking CW from its direct south, both theories having being considered in the past.
I do wonder why @ohwilleke also thinks Beaker should be mapped as parallel too because I think he’s aware of these CW results as well. Curious to hear his thinking on the issue.
@Jovialis
To me, the Italian BA-IA (excluding the ones with African or Near Eastern like admixture) set so far seems explainable as on a west-central European Beaker to Italian Neolithic-Chalcolithic cline, excluding some Balkan and Aegean influenced populations like the Messapic-speaking Daunians that seem to combine Balkan and Aegean influences alongside the local Italic-like substratum, the Sicily MLBA cluster that shifts towards the Aegean compared to the EBA one, an Adriatic Proto-Villanovan individual with Balkan ties, an Etruscan with the Adriatic kind of J2b.
What else do you see?
“We have found early CW populations rich in the western R1b lines now in the east. It does seems plausible now that Beaker arose as a local northwestern variant of CW subgroups rather than from a separate migration from the east overtaking CW from its direct south, both theories having being considered in the past.
I do wonder why @ohwilleke also thinks Beaker should be mapped as parallel too because I think he’s aware of these CW results as well. Curious to hear his thinking on the issue.”
Yes, you have isolated CW with R1b, but it still doesn’t ring true or sound convincing. In addition to the genetics, the anthropology isn’t a good fit – different artifacts, different strategies in warfare, probably different branches of IE linguistically, etc. CW and BB are just too distinct, pretty much from the outset (further complicated by the fact that the earliest BB people are not steppe individuals and appear to have adopted a Mediterranean culture by another population genetic people).
Likewise, the Yamnaya as the source population for CW also seems off. . . . a sister population with a remote common ancestor, sure. But not an ancestor-descendant relationship.
@ohwilleke
I am curious to know who I have to thank.
@Gigolo Moronogni – Your self-congratulation seems more than adequate.
I do not use twitter, but someone should ask Lazaridis if the new Southern Arc paper violates the idea that “the lack of any discernable Levantine or African DNA found in Minoans in Mycenaeans” from the 2017 paper brings back the Phoenician and Egyptian origins hypothesis. Because it seems like EVERYONE has some Levant_PPN with the model they are using in the Southern Arc paper. It seems like you cannot have one without the other. Moreover, isn’t Anatolia_N and Levant_PPN already very similar, something like 70%+? Wouldn’t than somewhat confound the results?
Razib – if you wouldn’t mind explaining for a second, why are the Maykop Culture and the Kura Araxes culture ruled out as possible candidates for origin of Indo-Anatolian languages?
It seems to a layperson like they’re placed smack dab in the right location and at the right time to be the source for these languages to spread north to the Yanmnaya.
I come from a place called Kerala where there is a community of muslims called mapilla’s (literally: son in law) – these are Arab traders who came to Kerala 1200 years ago and established a community that adopted the language of the women they married locally.
I know its not an exact comparable, but why couldn’t the case be that rural less developed steppe men have come to a more urban culture for trade and business and married local women to strengthen those ties? adopting the women’s culture and eventually take it back north?
I wonder if things would change up if they use the Dzudzuana and Taforalt samples, rather than Barcin and Levant_PPN. Despite the fact that the Dzudzuana samples are not out yet, I wonder if they considered using it.
Considering Natufian is modeled as 72% Dzudzuana and rest Taforalt, I wonder if one would find Taforalt in Yamnaya. If not, I think it could be a chink in the armor of this Southern Arc theory.
@Mohan, at the moment in the online conversation around this topic, people are tending to rule them out because of some presentations by David Anthony and Nick Patterson have been uploaded online. DA’s one was taken down unfortunately, but you can watch Dr Patterson’s here: https://www.college-de-france.fr/site/en-lluis-quintana-murci/symposium-2022-05-31-12h00.htm.
These provided some currently unpublished evidence that the Yamnaya are genetically the same as an earlier population at a site called Sredny Stog in Ukraine (see 24 minutes in NP’s presentation).
Though we don’t know the date of these samples, people have generally assumed that they’re early enough that this precludes Yamnaya being genetically formed from Maykop+earlier steppe people. We’ll have to see the actual dates to make sure that they are early enough to preclude this possibility of course.
So that’s why people tend to assume that the admixture must have come earlier from either the European direction or an earlier Caucasus wave.
Although, David Anthony also has a note on this in this paper – https://reich.hms.harvard.edu/sites/reich.hms.harvard.edu/files/inline-files/Anthony_proof3_Khvalynsk%20pz-2022-2034_v1.3_21_3_2022%20final-compressed%5B3%5D.pdf – “The Sredni Stog culture succeeded and replaced the Dnieper-Donets culture in the strategic Dnieper Rapids and throughout the steppes of Ukraine beginning around 4500–4300 BCE and ending in the late fourth millennium BCE with the appearance of Yamnaya. Unpublished Sredni Stog male genomes exhibit admixture ‘cocktails’ with the same basic elements as Yamnaya (EHG & CHG & AF). The CHG & EHG component was like Khvalynsk/ Progress-2, suggesting an eastern origin for at least part of the Sredni Stog population, and the AF component could have come from either the early Maikop or Tripol’ye populations. Sredni Stog introduced into the Ukrainian steppes new funeral customs (the Khvalynsk or ‘Yamnaya’ position), ceramic types (shell-tempered like Khvalynsk), and economies (large numbers of horse bones) that had appeared earlier on the Volga. Sredni Stog has for decades been recognized as an Eneolithic ancestor of Yamnaya influenced by late Khvalynsk, early Maikop, and the Tripol’ye and Varna cultures. “
So note here he doesn’t totally exclude “early Maikop”, so maybe something is possible.
But we don’t have any samples to show the admixture actually happening, in either direction so far, so it’s very difficult to prove anything. We do find at least one outlier in Yamnaya in Ukraine who is admixed between the main Yamnaya group and Makyop (Yamnaya_Ozera) who is dated 2950 BCE. There are some other individuals that show admixture between Maykop and the steppe, who might represent re;migration back north, but they’re mixed with the Central Asian Tarim/ANE type people and don’t seem to be able to be ancestors for Yamnaya.
As someone said that I am “some blowhard nobody”, but all what I wrote about Hg. R1, its origin and its expansion, is proved so far (look at the YFull tree and all is clear), I want to remember another point of my theory, that R-L23-Z2103 present massively in Yamnaya (and which is also my paternal line) was due to a migraton from West, and not being anyway my opponents able to demonstrate that R-L23-L51 did come from Yamnaya (but it is now discounted that they were the Bell Beakers), my theory about the Indo-European languages is that the oldest are the centum ones of the Alpine Region / Central Europe, and only the satem ones expanded from Yamnaya. I know that the question centum/satem is debatted (and a great contribute
gave the mourned Woudhuizen), but centum are older and satem are younger.
It looks more-or-less correct and a great little summary of the genetics with the caveats that:
1) Genetic pulses are very different sizes (e.g. barely any West Siberian into Sintashta; high GAC into European farmers).
2) It’s still also shadow-y what happens with the Balto-Slavic group; there are indications of a high-HG ancestry subpopulation that may be important for them (e.g. there are many samples from East-Central Europe from Latvia as far down as Serbia that have elevated HG proportions and preferentially seem to share genetic drift with later Slavic and Baltic speaking populations today), and possibly Balkan farmers too.
3) Caucasus farmers and Iranian farmers are necessarily a bit more vague. This will get nailed down a bit.
4) Greek must be impacted by post-Neolithic groups in the Aegean, because the sampled people we know must have spoken Greek (not some other related para-Greek language) have related Minoan ancestry rather than just Greek Neolithic.
5) Not so clear whether CWC is a branch from Yamnaya or some slightly earlier cultural stage. Closely related tho.
6) Very importantly, there probably were continuous mutual influences between all these IE languages so long as they remained geographically close enough, esp. in the early stages of divergence. So proto-Armenian might have some influence on the language spoken in Iran when the true Iranians showed up, proto-Greek and proto-Balto-Slavic and proto-Indo-Iranian continuing to be in some mutual influence etc. There’s probably continuous horizontal influence in the true phylogeny, on some level. It’s not necessarily the case that the Yamnaya and Corded Ware expansions really left totally separate branches of IE linguistically because of this horizontal transfer; we can’t necessarily say something like people “Speak a descendent of a Corded Ware language, not a Yamnaya one” linguistically because the sound changes etc could spread, without this being just an approximation.
Of course you know all this already.
Matt’s second point is kinda underrated I feel so I’d like to emphasize it too. A few people have brought up this attraction Balto-Slavic speakers seem to have to certain peri-Carpathian Neolithic groups in the Eurogenes’ PCA-based method. We’ve discussed that point on here before as well. Much like the pulse of Western European Neolithic in the scheme, there probably needs to be an extra pulse from certain non-GAC (or at least what we have from it so far) eastern populations into Balto-Slavic, though hopefully that seemingly somewhat controversial and formally less explored point should be clarified more in the future…
I suppose with the Yamnaya-like Sredni Stog samples that Anthony showed, the model might also look as Sredni Stog -> Yamnaya and Sredni Stog + something HG-rich -> Corded Ware, per point (5). Or there’s already a bit of a cline in EHG-CHG proportions in the area like we see with some Sredni Stog samples deviating towards the ‘north’/HG zone, though all current early CW samples seem to also have that slight western, not just northern, pull so it’s also plausibly acquired from more western populations only after the separation of the two.
@Matt
When you say CWC didn’t speak a language distinct from Yamnaya language do mean the distinction is below the language level and only on a dialect level? Like Punic and Hebrew and Edomite.
Where does Anatolian fit into your diagram?
@DaThang, I don’t think I’ve necessarily said that here, although I do think I said in the past that if there are relatively close relatives found within Corded Ware and Yamnaya (if that set of information, which isn’t even a preprint yet, still holds up), and they’re autosomally so similar, then likely they didn’t diverge very long ago or completely and the language would be likely not to be very different? It does seem most likely to be on a dialect level, I guess in those terms.
Albanian and Greek are descended directly from the Yamnaya. We see a cline from Yamnaya, all the way to Greece_N, and Albania_N. With samples like Yamnaya_Bulgaria_o, and Logkas_MBA intermediary. Mycenaean clearly on the cline as well closer to the original Neolithic people in Greece. I think it is possible this same dynamic played out in southern Italy as well, during the Bronze Age. My own halopgroup is from the Yamnaya in the Balkans EBA, which also shows up in Mycenaean Greece in the Palace of Nestor.
https://imgur.com/QUHtQBF
Mostly agree but positioning Bell Beaker as derived from Corded Ware rather than parallel to it seems off.
I also note the omission of Anatolian languages. We may not have a ready solution for their origins that is widely accepted, but they really do belong in the chart, if only to have a question mark with no lines in or out of that box.
Sorry, but I’m quite lost in the graph here.
I always thought that the bell Beakers did not come to form the Corded-Ware, but were a parallel population (also coming from the Yamnaya) to the Corded-Ware.
Am I right or wrong?
@Roberto Santos
Strictly speaking, CW and Yamnaya themselves likely share a recent common ancestor, not necessarily CW deriving directly from Yamnaya. Unpublished Sredni Stog samples that Anthony has shown plot exactly like Yamnaya on PCA, with our earliest/most steppe-rich CW samples deviating towards HG-rich populations to the west/north in comparison. We have found early CW populations rich in the western R1b lines now in the east. It does seems plausible now that Beaker arose as a local northwestern variant of CW subgroups rather than from a separate migration from the east overtaking CW from its direct south, both theories having being considered in the past.
I do wonder why @ohwilleke also thinks Beaker should be mapped as parallel too because I think he’s aware of these CW results as well. Curious to hear his thinking on the issue.
@Jovialis
To me, the Italian BA-IA (excluding the ones with African or Near Eastern like admixture) set so far seems explainable as on a west-central European Beaker to Italian Neolithic-Chalcolithic cline, excluding some Balkan and Aegean influenced populations like the Messapic-speaking Daunians that seem to combine Balkan and Aegean influences alongside the local Italic-like substratum, the Sicily MLBA cluster that shifts towards the Aegean compared to the EBA one, an Adriatic Proto-Villanovan individual with Balkan ties, an Etruscan with the Adriatic kind of J2b.
What else do you see?
“We have found early CW populations rich in the western R1b lines now in the east. It does seems plausible now that Beaker arose as a local northwestern variant of CW subgroups rather than from a separate migration from the east overtaking CW from its direct south, both theories having being considered in the past.
I do wonder why @ohwilleke also thinks Beaker should be mapped as parallel too because I think he’s aware of these CW results as well. Curious to hear his thinking on the issue.”
Yes, you have isolated CW with R1b, but it still doesn’t ring true or sound convincing. In addition to the genetics, the anthropology isn’t a good fit – different artifacts, different strategies in warfare, probably different branches of IE linguistically, etc. CW and BB are just too distinct, pretty much from the outset (further complicated by the fact that the earliest BB people are not steppe individuals and appear to have adopted a Mediterranean culture by another population genetic people).
Likewise, the Yamnaya as the source population for CW also seems off. . . . a sister population with a remote common ancestor, sure. But not an ancestor-descendant relationship.
@ohwilleke
I am curious to know who I have to thank.
@Gigolo Moronogni – Your self-congratulation seems more than adequate.
I do not use twitter, but someone should ask Lazaridis if the new Southern Arc paper violates the idea that “the lack of any discernable Levantine or African DNA found in Minoans in Mycenaeans” from the 2017 paper brings back the Phoenician and Egyptian origins hypothesis. Because it seems like EVERYONE has some Levant_PPN with the model they are using in the Southern Arc paper. It seems like you cannot have one without the other. Moreover, isn’t Anatolia_N and Levant_PPN already very similar, something like 70%+? Wouldn’t than somewhat confound the results?
Razib – if you wouldn’t mind explaining for a second, why are the Maykop Culture and the Kura Araxes culture ruled out as possible candidates for origin of Indo-Anatolian languages?
It seems to a layperson like they’re placed smack dab in the right location and at the right time to be the source for these languages to spread north to the Yanmnaya.
I come from a place called Kerala where there is a community of muslims called mapilla’s (literally: son in law) – these are Arab traders who came to Kerala 1200 years ago and established a community that adopted the language of the women they married locally.
I know its not an exact comparable, but why couldn’t the case be that rural less developed steppe men have come to a more urban culture for trade and business and married local women to strengthen those ties? adopting the women’s culture and eventually take it back north?
I wonder if things would change up if they use the Dzudzuana and Taforalt samples, rather than Barcin and Levant_PPN. Despite the fact that the Dzudzuana samples are not out yet, I wonder if they considered using it.
Considering Natufian is modeled as 72% Dzudzuana and rest Taforalt, I wonder if one would find Taforalt in Yamnaya. If not, I think it could be a chink in the armor of this Southern Arc theory.
@Mohan, at the moment in the online conversation around this topic, people are tending to rule them out because of some presentations by David Anthony and Nick Patterson have been uploaded online. DA’s one was taken down unfortunately, but you can watch Dr Patterson’s here: https://www.college-de-france.fr/site/en-lluis-quintana-murci/symposium-2022-05-31-12h00.htm.
These provided some currently unpublished evidence that the Yamnaya are genetically the same as an earlier population at a site called Sredny Stog in Ukraine (see 24 minutes in NP’s presentation).
Though we don’t know the date of these samples, people have generally assumed that they’re early enough that this precludes Yamnaya being genetically formed from Maykop+earlier steppe people. We’ll have to see the actual dates to make sure that they are early enough to preclude this possibility of course.
So that’s why people tend to assume that the admixture must have come earlier from either the European direction or an earlier Caucasus wave.
Although, David Anthony also has a note on this in this paper – https://reich.hms.harvard.edu/sites/reich.hms.harvard.edu/files/inline-files/Anthony_proof3_Khvalynsk%20pz-2022-2034_v1.3_21_3_2022%20final-compressed%5B3%5D.pdf – “The Sredni Stog culture succeeded and replaced the Dnieper-Donets culture in the strategic Dnieper Rapids and throughout the steppes of Ukraine beginning around 4500–4300 BCE and ending in the late fourth millennium BCE with the appearance of Yamnaya. Unpublished Sredni Stog male genomes exhibit admixture ‘cocktails’ with the same basic elements as Yamnaya (EHG & CHG & AF). The CHG & EHG component was like Khvalynsk/ Progress-2, suggesting an eastern origin for at least part of the Sredni Stog population, and the AF component could have come from either the early Maikop or Tripol’ye populations. Sredni Stog introduced into the Ukrainian steppes new funeral customs (the Khvalynsk or ‘Yamnaya’ position), ceramic types (shell-tempered like Khvalynsk), and economies (large numbers of horse bones) that had appeared earlier on the Volga. Sredni Stog has for decades been recognized as an Eneolithic ancestor of Yamnaya influenced by late Khvalynsk, early Maikop, and the Tripol’ye and Varna cultures. “
So note here he doesn’t totally exclude “early Maikop”, so maybe something is possible.
But we don’t have any samples to show the admixture actually happening, in either direction so far, so it’s very difficult to prove anything. We do find at least one outlier in Yamnaya in Ukraine who is admixed between the main Yamnaya group and Makyop (Yamnaya_Ozera) who is dated 2950 BCE. There are some other individuals that show admixture between Maykop and the steppe, who might represent re;migration back north, but they’re mixed with the Central Asian Tarim/ANE type people and don’t seem to be able to be ancestors for Yamnaya.