In their recent paper, titled
Early contact between late farming and pastoralist societies in southeastern Europe, Penske et al. make the following claim:
By contrast, Yamnaya Caucasus individuals from the southern steppe can be modelled as a two-way model of around 76% Steppe Eneolithic and 26% Caucasus Eneolithic/Maykop, confirming the findings of Lazaridis and colleagues 47. This two-way mix (40% + 60%, respectively) also provides a well-fit model (P = 0.09) for the Ozera outlier individual, consistent with the position in PCA and corroborating an influence from the Caucasus.
Err, nope.
The Ozera Yamnaya outlier, a female dated to 3096-2913 calBCE, is, in fact, a ~50/50 mix between standard Yamnaya and Late Maykop. It's a result that is totally unambiguous.
There are a number of ways to demonstrate this fact. For example, with the qpAdm software that was also used by Penske et al., except with different outgroups or right pops. Please note that in my dataset the Ozera outlier is labeled Ukraine_Ozera_EBA_Yamnaya_o.
right pops:
Cameroon_SMA
Levant_N
Iran_GanjDareh_N
Iran_C_SehGabi
Georgia_HG
Turkey_N
Serbia_IronGates_Mesolithic
Russia_WestSiberia_HG
Russia_Karelia_HG
Latvia_HG
Russia_Boisman_MN
Brazil_LapaDoSanto_9600BP
Ukraine_Ozera_EBA_Yamnaya_o
Russia_Caucasus_EneolithicMaykop 0.554±0.031
Russia_Steppe_Eneolithic 0.446±0.031
P-value 0.00109868 (FAIL)
Ukraine_Ozera_EBA_Yamnaya_o
Russia_LateMaykop 0.512±0.035
Russia_Samara_EBA_Yamnaya 0.488±0.035
P-value 0.462447 (PASS)
I can also do it with the Global25/Vahaduo method. And you, dear reader, can too, by putting the
Target and
Source Global25 coords from the text file
here into the relevant fields
here.
Target: Ukraine_Ozera_EBA_Yamnaya_o
Distance: 2.9292% / 0.02929202
50.6 Russia_Samara_EBA_Yamnaya
49.4 Russia_Caucasus_LateMaykop
0.0 Russia_Caucasus_EneolithicMaykop
0.0 Russia_Steppe_Eneolithic
Moreover, here's a self-explanatory Principal Component Analysis (PCA) plot that illustrates why my Late Maykop/Samara Yamnaya combo is much better than the reference populations used by Penske and colleagues. It was done with the PCA tools
here.
I'm pointing this out for two main reasons. First of all, this is a fairly obvious mistake that should've been avoided, especially considering the level of expertise and experience among the authors (such as Wolfgang Haak and Johannes Krause).
Secondly, it's important to understand that the Ozera outlier comes out almost exactly 50% Samara Yamnaya because the standard Yamnaya genotype already existed well before she was alive, and thus she cannot be used to corroborate any sort of influence from the Caucasus in the formation of the mainstream Yamnaya population.
As for the Yamnaya Caucasus individuals, I don't know why Penske et al. attempted to model their ancestry as a group, because they don't form a coherent genetic cluster. RK1001 and ZO2002 are fairly similar to standard Yamnaya samples, while RK1007 and SA6010 resemble Eneolithic steppe samples from the Progress burial site. This is what happens when I try to reproduce the Penske et al. model with my outgroups.
Russia_Caucasus_EBA_Yamnaya
Russia_Caucasus_EneolithicMaykop 0.187±0.019
Russia_Steppe_Eneolithic 0.813±0.019
P-value 4.15842e-06 (HARD FAIL)
Oh, and Penske et al. modeled the ancestry of mainstream Yamnaya as a three-way mixture with Steppe Eneolithic, Caucasus Eneolithic/Maykop and Ukraine Neolithic (or Ukraine N). They succeeded, but with my outgroups it's another hard fail.
Russia_Samara_EBA_Yamnaya
Russia_Caucasus_EneolithicMaykop 0.177±0.017
Russia_Steppe_Eneolithic 0.706±0.026
Ukraine_N 0.116±0.014
P-value 4.73919e-07 (HARD FAIL)
Admittedly, proximal models aren't easy to get right. And if you throw enough outgroups into a model, a large proportion of plausible models will fail. But I'm somewhat taken aback by these poor statistical fits.
In my opinion, mainstream Yamnaya doesn't harbor any Caucasus ancestry that wasn't already present on the Pontic-Caspian steppe during the Eneolithic or even much earlier (see
here). But ultimately this problem can only be solved with direct evidence from ancient DNA, so let's now wait patiently for the right samples.
Citation...
Penske et al.,
Early contact between late farming and pastoralist societies in southeastern Europe,
Nature, https://doi.org/10.1038/s41586-023-06334-8
See also...
Understanding the Eneolithic steppe
On a slightly different tack, trying with some G25 data (which may have been obsoleted by the latest spreadsheet), I found that it seems like allowing Usatovo samples to act as a source for Yamnaya_Ozera allows for a superior model fit to what you can achieve via two-way between Afanasievo and Late Maykop: https://imgur.com/a/mSt20YL
ReplyDeleteThat seems to be logical given that in the plot by Davidski, a point "southwest" of the cline Davidskis identifies might lead to a better fit to the position.
@Davidski
ReplyDeleteIn your opinion, which regions and time periods are most crucial for resolving the issue of Caucasus related ancestry in Yamnaya and other Bronze age Steppe groups? Which cultures or zones do we most urgently need sampled?
With some specifications of samples in G25 anyway seems like you can get to a mixture of Ozera from almost equal proportions of the Eneolithic Ukraine samples, the Steppe_EMBA and Maykop, including some Steppe Maykop: https://imgur.com/a/MWX38hE
ReplyDeleteLinks between both Steppe Maykop and Usatovo in the same individual are not necessarily surprising; we know from David Anthony's presentation that Reich lab have a few more Usatovo samples than we see here because they cluster in different places, including one that seems like they fit on the Khvalynsk-Progress cline - https://imgur.com/a/MWX38hE, and we know from Ringbauer's presentation that there is some linkage between an Usatovo person and the Steppe_Maykop wagon rider - https://imgur.com/a/3Jl3nNW; David Anthony says this Usatovo person is 3500-3300 BCE so that compares with Maykop_Late, who we know are directly linked to Ozera, at 3200 BCE, and Ozera's date at 3000 BCE.
Yamnaya_Ozera is not very well archaeologically attested as to her culture, but is early relative to Yamnaya and was called Yamnaya. Clearly seems like she has some IBD linkages with "Yamnaya Proper" but maybe it is tempting to me to see her as transitional between contact cultures and the expanding Yamnaya/Corded Ware with a base of ancestry more solidly in the Steppe_Eneolithic network, where any linkages with SEE Neolithic are more diluted (and Caucasus_Eneo more so, if present).
But yeah still very much speculative on my part, and maybe this kind of modelling is "overfitting".
@Matt
ReplyDeleteRarely is it possible to get a straight 50/50 fit that passes in qpAdm and G25 and is supported by IBD data.
So the Ozera outlier looks to be a first generation Yamnaya/LateMaykop mix.
@Synome
ReplyDeleteIn your opinion, which regions and time periods are most crucial for resolving the issue of Caucasus related ancestry in Yamnaya and other Bronze age Steppe groups? Which cultures or zones do we most urgently need sampled?
We really need Neolithic and early Eneolithic steppe samples all the way from Moldova to the North Caucasus.
Otherwise what's gonna happen is that the authors of these papers will keep making up contrived steppe ancestral groups (like Penske's Steppe Eneolithic with two EHG-heavy Khvalynsk samples) and then reaching for samples from Armenia or Iran to get passable models.
Right now what we're seeing is the classic garbage in, garbage out.
ReplyDelete“ Yamnaya_Ozera is not very well archaeologically attested as to her culture,”
Although lacking accompanying pottery, her burial posture seems to be associated with Zhivotilovsk culture which in part was thought to represent a trickle of Majkop mobility toward the northwest
it’s actually fairly well described
Great, i m fascinating with the new Samples that came with this paper. Which Sample should we choose for "Forest Steppe" or North Steppe Like(i dont like Forest Steppe concept Here cause on Early Neolithic steppe was larger)?
ReplyDeleteThe Cernavoda samples are quite interesting, a detailed investigation of those might add significantly to the whole Proto-Anatolian debate.
ReplyDelete@Virgin
ReplyDeleteI haven't looked in detail at the Penske samples yet.
But there was no hard genetic border between the steppe and forest steppe. Some of the populations were basically like early farmers and others like Yamnaya and probably everything in between too.
@Copper Axe
Yep, the Cernavoda samples are nice. Quite a bit of steppe in some of them, I guess as much as we'd expect.
I have not been able to look at the autosomal data of the paper in detail, but is it fair to say that the Yamnaya & Corded Ware genetic profiles were not at all that common in the studied area during the Eneolithic and that Yamnaya & Corded Ware genetic profiles were likely on the periphery of the studied area (steppe/forest steppe transition area)?
ReplyDelete@Aram XAN030? That's from 1200BCE Crete LBA
ReplyDeleteHis best fits are with Unetice and Czech Bell Beaker
There's no admixture from Bell Beakers and Unetice in the Aegean BA/EIA or in the southern Balkans, maybe in some border zones near Hungary and Croatia, but that's not really important. Both XAN030 and XAN051 have Balkan Y DNA and are autosomally similar to other samples we've seen from Greece and surrounds. With G25 only one of the Krousonas women scores marginal amounts of WHG, the others are close to zero which is night and day with what we see in other places. There's also near 0% overlap between the Y DNA of BA central/western/northern Europe and what we see in southeast Europe, the Aegean and Anatolia so unless you suggest they were exchanging wives over a thousand mile distance I don't see how this theory makes any sense. Skourtanioti et al wasn't very good in their modelling and they totally phoned in the uniparental analysis so don't take their theories too seriously.
ReplyDelete@Davidski, OK; I'll check out what is the case when its easy for me to qpAdm with these models (i.e. with AADR is updated).
ReplyDeleteAlso: https://genomebiology.biomedcentral.com/articles/10.1186/s13059-023-03013-9 - "Genetic history of East-Central Europe in the first millennium CE"; this paper is released so we can get times and dates on these sample IDs.
Seems like for these Iron Age and Medieval samples (from "Genetic history of East-Central Europe in the first millennium CE"), there is a pretty clear cut split of IA samples being Germanic-like (even those from the SE of Poland) and the MA samples being like present day Slavic people - https://imgur.com/a/GvVq1x5
ReplyDelete@alex
ReplyDeleteSkourtanioti et al wasn't very good in their modelling and they totally phoned in the uniparental analysis so don't take their theories too seriously.
Orpheus has his head stuck firmly up his ass.
He believes that every paper that manages to get through peer review is right, and he doesn't have a clue how to analyze data.
The Cernavoda individuals are quite heterogeneous, with some EEF heavy some steppe heavy, but all shifted 'northeast' compared earlier/ more southern Neo/ early Chalc genomes and significant shift from EEF Y-dna to that of Dniester-Dnieper hunter-gatherers.
ReplyDeleteThey are probably the proximal source of EEF/UkrN shift in Yamnaya c.f. Khvalynsk/ progress
So at quick G25 look Yamnaya Samara, for ex, are :
~ 20% Cernavoda
~ 80 % Khvalynsk/ Piedmont cline
Interesting stuff...
ReplyDeletehttps://genomebiology.biomedcentral.com/articles/10.1186/s13059-023-03013-9
I'm going to start blogging about these Polish samples when the other paper is also published.
@Davidski
ReplyDelete"Yep, the Cernavoda samples are nice. Quite a bit of steppe in some of them, I guess as much as we'd expect."
That one sample with lots of steppe nodels quite nicely as something close to 50% Cernavoda and 50% Yamnaya with Yamnaya as a standin for pre-Yamnaya steppe ancestry. I should see if the bulgarian Yamnaya sample works as either a Yamnaya/Cernavoda or Yamnaya/Ezero because thats what I suspect the sample is mixed between, there was quite some interaction between the two down south apparently.
@Copper Axe
ReplyDeleteThat one sample with lots of steppe nodels quite nicely as something close to 50% Cernavoda and 50% Yamnaya with Yamnaya as a standin for pre-Yamnaya steppe ancestry.
What's the C14 dating for this sample?
@Davidski, yeah, re; Poland IA_MA paper (Stolarek et al) I think the other paper will be a useful complement (I assume you mean the one with the BCE samples?).
ReplyDeleteThe argument in this paper seems to me (although I skimmed it) to be that even the Iron Age samples from the SE have relatively little autochthonous ancestry (per G25 though there is one clearly "Slavic" individual at Maslomecz_IA), and mostly represent intrusive Wielbark/East Germanic culture.
So looking at earlier samples may be useful to validate these concepts.
Preliminary Labelling: https://pastebin.com/smnSER6Y
Some Vahaduo work to demonstrate what the preliminary labelling and identification of outliers is based on: https://imgur.com/a/sybLWaU (this outlier labelling isn't going to be perfect, but gives a rough first draft).
This may also benefit a fair bit from official genotypes.
@Davidski
ReplyDeleteI was thinking of this sample:
"KTL001. Grave 10. Belongs to the KTL A group.
The grave was uncovered in square A'/7, at a depth of 3.09-3.11 m, in the layer of the Cernavodă
I culture within the settlement. It was partially covered by grave 9. The young male was in a
crouched position on his right side with his skull to the southeast. His left arm was stretched to
the knees. His right arm was bent and the forearm stretched forth. Dating of the skeleton: 3502-
3343 calBCE (4609±26 BP, MAMS-48810)."
Its one of the younger samples hence why I think this sample has a recent steppe ancestor.
Two quotes from D.W. Anthony:
ReplyDeleteFirst from“The Horse, the Wheel, and Language”:
„The spread of the Usatovo dialect up the Dniester valley(…)the movements into the East Carpathians and up the Danube valley occurred in the right sequence, at the right time, and in the right directions to be connected with the detachment of Pre-Italic, Pre-Celtic, and Pre-Germanic—the branch that ultimately gave birth to English.”
Second from “The Indo-European Homeland from Linguistic and Archaeological Perspectives”:
“After that, a cluster of western European branches separated to the west, into the Danube valley on the south side of the Carpathians with the Yamnaya migration up the Danube about 3100–2800 BCE, and into southern Poland on the northern side of the Carpathians with the expansions of the Usatovo and the Tripolye C2 cultures about 3300–3000 BCE (Ecsedy 1994, Mallory 1998, Klochko & Kosko 2009, Heyd 2011, Anthony 2013). These last separations match the proposal that the ancestors of Italic and Celtic (and perhaps pre-Germanic) could have separated in a rather complex phase of migrations and language spreads.”
Pre-Germanic languages came from Usatovo. Now we have Usatovo DNA. Was Anthony right?
@EastPole
ReplyDeletePre-Germanic languages came from Usatovo. Now we have Usatovo DNA. Was Anthony right?
Nope.
@EastPole, there's no evidence of a direct migration from Usatovo to any Northern or Western part of Europe, so if that were so it would have to rely on some chain of Usatovo->unsampled Balkan EBA->Western European IA, I think, with greatly more linguistic than genetic influence (which would primarily come from early CWC and Beaker offshoot).
ReplyDelete@Matt,
ReplyDelete“cremation of the dead was the prevailing custom in Central Europe from the late Bronze Age until the Middle Ages (MA)”
“Presented here whole-genome analyses of the individuals from the IA group together with our previous observations [37, 38] and numerous archaeological findings consistently support earlier hypotheses assuming that the Wielbark culture was associated with immigrants from Northern Europe who spread within the region of present-day Poland and mixed with the autochthonous IA population. Most of the data collected for the IA and MA groups are in line with the hypothesis assuming genetic continuity from the IA to the early MA in East-Central Europe and suggest that the migration from east in the sixth CE was not necessary to form the genetic pool of the MA group”.
Slavs in IA cremated. There were some other groups present in Poland which didn't practice cremation. So it is natural that in the graves, which were not Slavic, not Slavic dna was discovered.
@Matt
ReplyDeleteSo Jordanes was right about where Goths came from.
RE: Poland
ReplyDeleteApart from one iron -age Masłomęcz individual (who is J2b2), there is a clear disjunction between Iron Age and medieval Poles
Seems like most of the autochthons were in fact pred. G2a, as they can't be Nordics (Jutland being G2a-poor). Could be Alpine/Halstatt -influence coming in during Lausits period ?
Fair few R1a- ?M458 in Iron Age but within the local autosomic profile
@ Matt
Poland_Masłomęcz_Iron_Age_oEastEurope:PCA0103, should be PCA0113 ?
@EastPole, absence of evidence is not evidence of absence of course. But what is actually "evidence of presence"? That is the question here.
ReplyDelete@Rob, I automatically matched up the sample IDs from Davidski's preliminary sheet with what was given in the paper, and added the outlier identification suffixies myself. PCA0103 is the label of the sample which seems to me to have the unusual level of shared drift for their IA culture with Slavic groups. PCA0113 doesn't seem to have made it through to what Davidski gave in the preliminaries, unless more were subsequently shared after the file I grabbed.
@ East Pole
ReplyDelete''Pre-Germanic languages came from Usatovo. Now we have Usatovo DNA. Was Anthony right''
We were expecting Usatavo to be full of I1, R1b-U106, and 'Nordic drift " ?
''Slavs in IA cremated. There were some other groups present in Poland which didn't practice cremation. So it is natural that in the graves, which were not Slavic, not Slavic dna was discovered.''
Even Italics cremated. Seems the switch to ingumation in Wielbark was of cultic significance.
@Rob, PCA0103 is a female. So no informative y-dna signals for her.
ReplyDelete@ Matt
ReplyDeleteok, maybe Im just missing it just cant find PCA0103 in their data list. But there is a PCA0113 and it is from Maslomecz.
_______
Trying a similar set-up to the paper, the Wielbark mainstream can be modelled as Unetice + Nordic IA, although anachronistic and not coterminous
The Gaski lass (Przeworsk attribution) is indeed shifted more toward western Europe
medieval Poles come in as Baltic_LBA + HUngarian LBA
Feeding Kowelko back in as a source doesnt seem to do much
@ Matt - yeah I see. Thanks
ReplyDelete@Rob, also it's possible that PCA0113 would be shifted in an unusual direction, if he got through to G25... But the paper's Figure 2 https://genomebiology.biomedcentral.com/articles/10.1186/s13059-023-03013-9/figures/2 supports that only PCA0103 and PCA0001 among Iron Age have a very larger significant shift to East_Europe relative to North-Western_Europe, in the f4 stats that they are able to produce.
ReplyDeleteAlthough it might be good to double check these f4 stats against G25 where we are able. There are many samples who are intermediate, from the Middle Ages, while the Vahaduo North Europe PCA (on G25 data) is quite strong about a split. The f4 stat may be confounded by EEF:Steppe:WHG ratios.
It's also possible that PCA0113 is shifted in some other direction orthogonal to the East_Europe:North-Western_Europe f4 stat (like SE shifted or something).
In their supervised ADMIXTURE two-way model, the sole J2a (who is going to be PCA0113) clusters as completely NW European with zero East Europe, but again this may be due to some other shift that is not captured by these two components. (He seems to have not made it through to their PCA either, so we can't tell from that).
@Rob; Stolarek's f4 looks like it has a pretty decent relationship with Baltic_BA proportion from a very rough and ready G25 model - https://imgur.com/a/clMWYtQ
ReplyDeleteThere is a remarkable symmetry in migrations to Balkans and South Caucasus. In the same period as Cernavoda forms Steppe ancestry appears in South Caucasus ( Areni cave ) .
ReplyDeleteIn EBA Yamna expands to Balkans but not to South Caucasus. Most probably because Kura-Araxian people were blocking the Daghestan pass.
But in MBA another wave of migration from Steppe. This time it reaches Greece in Balkans and deep in South Caucasus / Historic Armenia / North Iran.
MAJ017 and MAJ019 have a genetic profile typical for early Corded Ware , but they probably belong to the Budjak Yamnaya variant. They also have quite a bit of HG ancestry. It seems to suggest that the CW culture evolved from a Budjak group.
ReplyDelete@Davidski @Matt
ReplyDeleteBeen a bit busy to follow this paper more closely, but seeing the good fit Usatovo samples have in both qpAdm and G25 in modelling Yamnaya ancestry, is it not via them that Yamnaya has actually some "Southern Arc" admixture?
Usatovo seem to have varying amounts of Maykop admixture, in one of them (USV005) reaching 30+%.
And like Matt mentioned there is some IBD link between Usatovo and Steppe Maykop.
@Rob
ReplyDeleteHow common was cremation among Goths and Vandals in Italy, Spain, North Africa i.e. on territories where there was no R1a, no Slavs? What percentage?
@ Matt
ReplyDeleteNeedless to say, I’m pretty skeptical of their one-way qpAdm models or their explanation of “cremation”.
@ Aram
No there was no 2200 bc wave to Greece from the steppe
It came from Thrace , as many kurgans fell into disuse at that time
Hello Davidski, this is completely unrelated to this post and I apologize for that, but recently it has been announced that the Anthrogenica forum is closing permanently on August 17th, 2023. Do you have any thoughts on this? How will this effect amateur/lay genetics?
ReplyDelete@gamerz_J
ReplyDeleteUsatovo clearly does have Maykop ancestry, which shows up in PCA and in IBD.
But the IBD links between Yamnaya and Maykop/Steppe Maykop are weak, and Yamnaya doesn't show any Maykop links in PCA.
@WSH
ReplyDeleteHello Davidski, this is completely unrelated to this post and I apologize for that, but recently it has been announced that the Anthrogenica forum is closing permanently on August 17th, 2023. Do you have any thoughts on this? How will this effect amateur/lay genetics?
I haven't been active at Anthrogenica for a long time, so this doesn't change anything for me personally.
Another forum will probably appear to take its place.
@ east Pole
ReplyDelete“ How common was cremation among Goths and Vandals in Italy, Spain, North Africa i.e. on territories where there was no R1a, no Slavs? What percentage?”
Low. But when these groups entered Roman territory in 5th century, Przeworsk culture & Wielbark had ceased.
At this point, they were Arian Christians, but inhumation was patchily used in Germania since 100 BC, due to returning mercenaries bringing this idea in
However, back in the lower Elbe region, western Pomerania, old Frankia, cremation continued used until 500AD
NB: R1a-L664 and Z284 is seen as a minority lineage in some Germanic regions
@Davidski
ReplyDeleteThanks for the reply.
I brought this up because that paper seems to get the best p-values in its qpAdm setup with modelling Yamnaya as a mix of Usatovo+Progress_N, as I can't really run qpAdm myself would you say that using your outgroups such a model would also fail as proximal sources for Yamnaya?
@WSH, on that tangent, I'm a lot more pessimistic on that than Davidski. Sadly its part of a wider trend; the internet of open and anonymous forums (in terms of real life identity) is being closed down in favour of walled and non-anonymous social media, with restricted conversations. What's happening on the big platforms is obvious, as they get turned into machines to fuel their owner's egos and/or marketplaces for advertising and to sell products.
ReplyDeleteWhile for small forums and blogs, that cater to small audiences and remain unmonetized, storage costs are increasing and the people with the big money and power are favouring that trend. Further, the people (on the left and right) who moderate and own these spaces often no longer believe in the ideals of blogging or open forums with free speech, and just believe in making money and controlled speech that furthers their political aims. The younger generation often don't really like to read but watch low-information density videos that are frankly more like entertainment than learning. Blogs such as they still exist are increasingly becoming like podcasts; something where a writer puts out some well-established ideas as entertainment for a uninformed lay-audience to subscribe to as a paid reader, and not something to generate or be part of a serious conversation.
AI scraping and DDoS attacks make some forums also more difficult and expensive to maintain as well; anthrogenica is now unusable to me (and probably others) to browse as a casual having a look every now and then (lost my login details ages ago and didn't bother to re-register) due to whatever Cloudflare solution they have implemented to stop scraping, which doesn't seem to work well at all.
Internet search on the web outside the "walled garden" of the platforms is getting trashed and becoming unsable due to search-engine optimisation by clickbaiters (for malware and commercial reasons). Search engine companies don't seem to really be doing too much to combat this trend, and would seemingly prefer people to get information in future through via their proprietary AI chatbots that have scraped the internet, rather than directly linking to websites or talking to other people on them.
A lot of these problems are the result of trends in technology and finance, but I also blame the generation who grew up online for trashing the promise of the internet with devolving into trolling, flaming, irony sh*tposting, clickbait and activism (left and right wing), all for personal advantage and short-lived feelings of superiority and "clout". That makes everyone cynical about even bothering to defend it or preserve it.
The open internet that for the generation born about 1978-1999 was a defining feature of young adulthood, with communities that had conversation, is probably gonna go dark. A lot of that information and conversation, like on anthrogenica, is probably gonna be lost (and, yes, like 99% of anthrogenica is just people talking about their personal haplogroup or family tree or 23andme results, in ways that they find interesting yet don't mean much to posterity, but there is probably some stuff on there that would be of interest to the future). If not entirely then much so. At least for a while. Appreciate spaces that are left that we have, for as long as they're present.
All that said, I'd love to be proved wrong in this prediction though, and to look back in 5 years and think "Oh, wow, why did I think that?".
@gamerz_j, yeah, there is that model. But I don't know that they give a really strong reason to prefer it. In terms of actually there are samples who not explainable any other way.
ReplyDeleteAlso there is the case that the Ukraine_Eneo guys are all pretty heterogenous, and so it is hard to get a true grasp of how representative each is of the underlying population or not.
For example if we look in the Vahaduo North Eurasia PCA, which has more power to distinguish Progress_Eneo from the Steppe_Maikop cline and distinguish between varying levels of Ancient North Eurasian related ancestry: https://imgur.com/a/eg7gqx7
Then it's looks like the overall clines of Usatovo/Usatovo_Majaky and Cernavoda are different, and MAJ023 and USV004 are displaced away from a cline between SEE_CA type sources and Yamnaya. And in the Cernavoda, I'm not sure there is one cline of that KTL001 has some Ukraine_N admixture that takes it away from a separate cline.
@gamerz_J & Matt
ReplyDeleteIf you guys can list the samples that they used in each model of interest then I can try and reproduce them with my outgroups.
Which Usatovo samples did they use?
Target: Ukraine_EBA_Yamnaya - I2105 (one sample, kind of weak!)
ReplyDeleteSource: USV (USV003, USV006, USV004, USV005), Steppe_Eneolithic (PG2001, PG2004, VJ1001, I0433, I0122)
@gamerz_J & Matt
ReplyDeleteStrange choice of running just one sample, so this doesn't prove anything about Yamnaya origins. Also, the model is a fail...
Ukraine_EBA_Yamnaya_I2105
Ukraine_Usatove 0.319+/-0.039
Russia_Steppe_Eneolithic 0.681+/-0.039
P-value 0.000350011
A couple more, with horrible fits...
Ukraine_EBA_Yamnaya
Ukraine_Usatove 0.456+/-0.031
Russia_Steppe_Eneolithic 0.544+/-0.031
P-value 9.75673e-07
Russia_Samara_EBA_Yamnaya
Ukraine_Usatove 0.201+/-0.021
Russia_Steppe_Eneolithic 0.799+/-0.021
P-value 4.63917e-09
Maybe their OGs need review...
ReplyDeleteOn another topic, one of the downsides of this study when it comes to identifying AncIBD links is they only looked within the study's samples: https://i.imgur.com/hZfA4IZ.png
There are some (although some of those date gaps for 20cM links make me nervous about the dates BCE provided in supplements!), but do these samples have linkages to contemporary Maikop/Steppe_Eneo etc at all? We don't know.
However, I'm guessing that Harald R will make a revision of his preprint at some point later in the year that may provide some hints on these.
Their outgroups are pretty standard as far as peer reviewed work goes.
ReplyDeleteBut these outgroups are weak, because outgroups are generally weak in such work.
This makes it easier for shit models to get passing P values. Hehe.
Rob
ReplyDeletePossible. But something must have triggered their migration to South. Why they migrated at 2200 BC? A push from North that displaced them? A climatic event?
@alex All qpAdm sources are proxies by definition, Unetice/BB giving the best fit when rotated with Yamnaya, early CWC etc sources simply function as a proxy. Even with other Balkan sources that give a lower but still good fit, they're always preferred over some pure Yamnaya source
ReplyDeleteCombined with Logkas IBD with CWC instead of Yamnaya gives us a pretty clear picture on the ancestry of these populations. Agrees with Clemente et al 2021 as well
"I don't see how this theory makes any sense"
I don't think they wrote any theories in the paper. Uniparentals are easily explained by later admixture with Balkan populations before they reach Greece, or an altogether different group (Unetice/BB-like) that remains unsampled with those uniparentals. Anyway my main point was that the specific R1b Aram brought up is found in LBA Crete centuries or even a millennia after proto-Greek appeared, Yamnaya vs CWC ancestry is relevant for Balkanic languages (see Lazaridis et al 2022 for that too)
@Aram MBA steppe ancestry in Greece doesn't seem to be from the steppe, but EBA and pre-EBA ancestry probably is from steppe groups (even if displaced from the steppe).
CWC groups did displace Yamnaya ones so it could be that. Or just traveling, which was common for Yamnaya-rich groups
Orpheus
ReplyDeleteNot only Xan030 but other samples also from Mugla, Pylos, also from Cinamak in Albania. There is pretty much PF7562 in ancient Greeks to link it with Proto Greeks.
And I don't think it came from BB/Unetice related people. That is quite unlikely imho. Thrace mentioned by Rob is far more realistic.
Here a reliable source
https://discover.familytreedna.com/y-dna/R-PF7562/tree
Hmm... Replicating their full model for Ozera - https://pastebin.com/kKwteaft (at p=0.03)
ReplyDeleteBut indeed it doesn't work with even small and reasonable changes to outgroup (inclusion of Serbia_Iron_Gates_Mesolithic etc).
Using a compromise between your OGs and their OGs, and a reasonable compromise to maximise coverage (from yours: drop Iran_Seh_Gabi_C, drop Boisman, drop Brazil_LapaDoSanto; from theirs drop MA1, their "WHG"; include Israel_C, include France_MontAime_MLN to identify any specific drift in the founder EEF), I got a comparable model fit to yours for Ozera from the same groups, at p=0.58 (https://pastebin.com/y7vTqQ3r). There's nothing about their preferred inclusion of Turkey_Ikizitepe_LC and Russia_Steppe_Maikop that would drive a model of Russia_Yamnaya_Samara and Russia_LateMaikop to fail.
Also found that a comparable fit at p=0.37 was found from a 3-way model of Ukraine_N (10%), Russia_Steppe_Eneolithic (Progress+Vonyuchka) (28%) and Caucasus_LateMaikop (60%) (https://pastebin.com/Aki7GjpF). That possibly provides some evidence that Ukraine_N plus Russia_Steppe_Eneolithic can act equivalently to Russia_Yamnaya_Samara in this context (although IBD indicates direct match with Yamnaya in reality!).
@Aram
ReplyDeleteAround 2200 bc you had the 4.2 kiloyear event which had a significant impact on the steppes. The lack of rainfal turned it very arid and the region got severely depopulated during that period. Hence the sharp distinction between Catacomb and Srubnaya in autosomal profiles. Something similar happened during the last few centuries of the 2nd millennium bc although not as drastic.
In other news, as part of the recent deluge of dna, the paper on the Gurgy site with the large G2a family has been published - https://www.nature.com/articles/s41586-023-06350-8
ReplyDeleteThere is a phenotypic artistic reconstruction (see https://www.mpg.de/20653021/0721-evan-family-trees-from-the-european-neolithic-150495-x) which is based on Hirisplex. Maybe Sam Andrews will be pleased to see they found a site where 10/94 samples have red hair (possibly a greater % from the main family, on which basis the reconstruction more or less gave all of samples a reddish-brown colour, although its probably overdone).
(They are the standard sort of people with about 1/6 HG ancestry).
OT
ReplyDeleteDavid,
Have you seen that new Nature paper about the Neolithic family grave in France which has been analysed?
How do they compare to moderns?
@ Orpheus
ReplyDelete''Combined with Logkas IBD with CWC instead of Yamnaya gives us a pretty clear picture on the ancestry of these populations. Agrees with Clemente et al 2021 as well''
Not ncessarily. How strong were the IBD links ?
The archaeology & Y-DNA links to Balkan Yamnaya. So if the IBD links are true, it could be CW women being brought into the community
Good 2022 archeological paper complementing this paleogenetic study
ReplyDeletehttps://www.journals.vu.lt/archaeologia-lituana/article/view/31845/30685
@John Thomas; the data is here if anyone wishes to convert them to genotypes - https://www.ebi.ac.uk/ena/browser/view/PRJEB61818
ReplyDeleteAlthough I don't think we'd see anything new from existing sites from France from the same age and location.
Also if anyone wants to convert these Sino-Tibetan / Hoabinhian admixed individuals from late first millennium BCE to early 2nd millenniium AD: https://www.ebi.ac.uk/ena/browser/view/PRJEB59488
One can’t go past Manzura for northwest Pontic steppe
ReplyDelete“History carved by the dagger: the Society of the Usatovo Culture in the 4th Millennium BC”. Igor Manzura
Readily available on web search
@Aram
ReplyDeleteI don't know what you think the Proto-Greeks are but to say that PF7562 was an important Proto Greek marker and that it demonstrates the connection of Mycenaean culture with Yamnaya is more wishful thinking
1-These are the true Proto-Greeks (or Proto-Mycenenans, I don't know if the Greeks use this terminology)-J-Y6313>ZS5071 (Lazarides-2.787 BC), I2-L596 (Nea Styra-2.515 BC), J1-Z2215>L620 (Nea Styra-2.487 BC), J1b-ZS50 (Perachora-2.450 BC & Nea Styra-2.413 BC), G2a2b-FTC60521 (Perachora-2.450 BC), L2-L595 (Nea Styra-2.412 BC), J2a-Z36800 (Koufonisi-2.407 BC), G2a2b-PF3378 (Nea Styra-2.400 BC), J2a-L26>Y113289 (Sarakinos-2.312 BC), I2a-Y87044 (Theopetra-2.237 BC)-With the genetic data we have (uniparental and autosomal) the Mycenaean did not come from the steppes but from Anatolia.
2- You have not been able to prove that there is PF7562 in Yamnaya and neither in Armenia (chalcolithic & bronze Age)-At the moment R1b-PF7562 has not been found in the chalcolithic (3.000-2.000 BC) and early Bronze Age (2.000-1.400 BC) nor in the Balkans nor in Greece or the Peloponnese. Three cases in the Pylos palace when the Mycenaean culture had already collapsed, one case in the Late Bronze Age in Crete (where we do not know if an Indo-European language was spoken), one sample in the Iron Age of Albania and late Anatolian samples do not serve to prove the connection between Yamnaya and the Mycenaean language. Orpheus is right, XAN30 best fits are Unetice and Czech Bell Beaker.
@Aram said- Smyadovo is not from Yamna but from Sredni Stog, and it migrated with Cernavoda culture. That is quite obvious imho”
How can you say such nonsense? Smyadovo M269 was a neolithic farmer of the Gumelnita-Karanovo culture dating back to 4,500 BC, at that time there was not even Cernavoda culture (4.000-3.200 BC) or Yamnaya culture (3.300-2.600 BC)-Cernavodá is a mixture of Ukraine neolithic and steppe eneolithic and occupied the territory of Gumelnita when this culture disappeared.
@Rich
Try to accept the existence of Smyaodovo M269 and its autosomal composition, the longer you take to accept it the greater the ridicule. The only funny thing is to see how you and your friend Rocca keep on defending the same nonsense as 5 years ago. Now you have also genomes from the western steppe (Usatovo, cernavodá etc) and L51>L151 still does not appear. Oh my God, I forgot that we have P310 in Mongolia, Ha Ha Ha.
@East Pole
Anthony has long since become disconnected from reality, not only has he been wrong with Usatovo, he has been wrong with almost all of his reinterpretation of Gimbutas' theory.
@ Gaska
ReplyDelete“ These are the true Proto-Greeks (or Proto-Mycenenans, I don't know if the Greeks use this terminology)-J”
Yea that’s the problem- you don’t know what you’re taking about, Although several people have tried to explain it to you
Your list is from southern Greece; pre-Greeks
The Theopetra individual is part of a proto-Greek community
This comment has been removed by the author.
ReplyDelete@Rob
ReplyDeleteWell, those are the chalcolithic samples that we have in the current Greek territory, whether they are Proto-Greeks, Pre-Greeks, Pre-Proto Greeks, Proto-Pre Greeks, Para-Greeks, Proto-Para Greeks Pre-Proto-Para Greeks or whatever you want to call them.
And you know what?, to date, there is no R1b-M269 (Z2103, PF7562 or L51>L151) and no R1a-M147 either in chalcolithic Greece
Even if there had been invasions or migrations from the north in the early Bronze Age, the uniparentals of the Mycenaean culture are still overwhelmingly Balkan-Greek-Anatolian-Levantine and autosomically the steppe signal is so weak that attributing a language change to it seems impossible to me.
I know that for you anything is enough to link the Mycenaean culture with Yamnaya and the Indo-Anatolian languages with the Balkans. I think that this attitude of yours is more a matter of faith than a scientific conviction and the same goes for your theory that M269 originates from the Don River, do you have any proof? or is it just a hunch.
Gaska
ReplyDeletePylos is marked as Mycenaean. Three cases of M269 are found there.
GRC_Mycenaean_Palace_of_Nestor_BA
Pylos, Palace of Nestor
Lazaridis discuss one M269 sample which is negative to L23 in ancient Armenia. Chances are high that it is PF7562. I simply do not refer to that Chinese theytree service. I consider it unreliable. But if You look there You will notice a PF7562 in ancient Armenia.
Sooner or later PF7562 will be found in Steppe. That is just a matter of time.
@Gaska
ReplyDelete"@Rich
Try to accept the existence of Smyaodovo M269 and its autosomal composition, the longer you take to accept it the greater the ridicule. The only funny thing is to see how you and your friend Rocca keep on defending the same nonsense as 5 years ago. Now you have also genomes from the western steppe (Usatovo, cernavodá etc) and L51>L151 still does not appear. Oh my God, I forgot that we have P310 in Mongolia, Ha Ha Ha."
Gaska, try reading again what Mathieson et al wrote about Sample I2181 (Smyadovo) on page 8 of their paper, "The Genomic History of Southeastern Europe" (2018):
"In two directly dated individuals from southeastern Europe, one (ANI163) from the Varna I cemetery dated to 4711-4550 BCE and one (I2181) from nearby Smyadovo dated to 4550-4450 BCE, we find far earlier evidence of steppe-related ancestry (Figure 1B,D). These findings push back the first evidence of steppe-related ancestry this far West in Europe by almost 2,000 years, but it was sporadic as other Copper Age (~5000-4000 BCE) individuals from the Balkans have no evidence of it."
Clearly, Smyadovo (I2181) had steppe-related ancestry. If R1b-M269 was indeed a Balkan Neolithic farmer lineage, why is it that Mathieson et al found no Balkan Neolithic farmers belonging to it (see that last sentence in the quote above)? Why is it the only R1b-M269 in their data set had steppe-related DNA? If I2181 was a farmer, he was obviously one whose Y-DNA line came from the steppe.
I'm pleased as punch about the existence of Smyadovo (I2181).
Rob pointed out to you that his own analysis showed that I2181 had 15% Khvalynsk ancestry (i.e., steppe-related ancestry).
And yes, there are two R1b-L51 samples from the ancient steppe pastoralist Afanasievo culture. One of them was R1b-L52 (P310): Sample I6222 from Wang et al (2020), rc dated to 3316-2918 calBC. The other was R1b-L151: Sample C3341, from Kumar et al (2022), rc dated to 2815-2526 calBC.
Then, of course, there are the oldest known Corded Ware samples to date, which are R1b-L51 (I'm including the subclades of L51 among them), and all the other steppe pastoralist and steppe pastoralist-derived R1b-M269, including Yamnaya.
There is no evidence that R1b-M269 has any real connection to the Neolithic farmer cultures of Europe west of the steppe. Thus far, no R1b-M269 has been found among them, despite the abundance and ever growing number of European Neolithic farmer samples, and no ancient R1b-M269 and subclades remains have ever been found that completely lack steppe DNA.
Add to that abundance of Neolithic farmer samples the newly published data set from Gurgy les Noisats in France from Rivollat et al, "Extensive pedigrees reveal the social organization of a Neolithic community". The researchers got Y-DNA from 57 of them, all G2a and H2m, and no steppe DNA.
Makes one wonder where all the western and central European R1b-M269 was hiding during the Neolithic Period.
There were at last 7 cases of M269 in ancient Greece.
ReplyDelete4 cases from Skourtanioti 2023
3 cases from Lazaridis 2022.
So I simply don't understand what we are discussing here?
AID009 1 (ss) 1 98403 0.08 M R1b1a1b Aidonia LBA 3 65 0.21 -NA- -NA- -NA- 64 576 -0.013 0.00598 0.001 0.013
GLI002 1 (ds) 1 60375 0.05 M R1b1a1b GlykaNera LBA 1 111 0.34 0.02 0.01 0.03 112 1.01E+03 0.037 2.64E-02 0.001 0.027
TIR008 1 (ds) 1 385098 0.31 M R1b1a1b Tiryns IA 4 710 " 2.19" 0.01 0 0.02 489 4401 0.002 3.56E-03
XAN030 1 (ss) 1 481327 0.39 M R1b1a1b J2b1 Chania LBA 4 3157 8.72 0.01 0 0.02 639 5751 0.017 0.00968 0.001 0.015
Well, look, the way I've mostly been using quotes and graphics from Anthony's presentations is because he is really the only one who has done presentations (or a presentation anyway) that include information on what unpublished material Harvard's got for the pre-Yamnaya North Pontic / steppe region. And generally only in the context of what he's reporting or showing about genetic ancestry, inferences about what increases in genetic outliers tell us about mobility, and shared IBD, which I think we can have a reasonable trust that he's passing on accurately. If there is another preferred source for this information (to my knowledge, none), we can use it.
ReplyDeleteOff-topic but @ulfing and LGK if you're reading: https://historicengland.org.uk/whats-new/news/new-scientific-study-solves-mystery-2000-year-old-grave/ - another finding where a burial thought to be unclear/male and which resisted dna analysis has been found by tooth enamel protein to be female...
ReplyDeleteI think the evidence presented by ulfing in the other thread seemed persuasive, but it does seem odd to be two counter-intuitive results in a row here (although not totally counter-intuitive I admit).
https://phys.org/news/2023-07-insights-indo-european-languages.html -
ReplyDelete"New insights into the origin of the Indo-European languages"
"To solve these problems, researchers from the Department of Linguistic and Cultural Evolution at the Max Planck Institute for Evolutionary Anthropology assembled an international team of over 80 language specialists to construct a new dataset of core vocabulary from 161 Indo-European languages, including 52 ancient or historical languages. This more comprehensive and balanced sampling, combined with rigorous protocols for coding lexical data, rectified the problems in the datasets used by previous studies."
"These results are not entirely consistent with either the Steppe or the farming hypotheses. The first author of the study, Paul Heggarty, observed that, "Recent ancient DNA data suggest that the Anatolian branch of Indo-European did not emerge from the Steppe, but from further south, in or near the northern arc of the Fertile Crescent—as the earliest source of the Indo-European family. Our language family tree topology, and our lineage split dates, point to other early branches that may also have spread directly from there, not through the Steppe." "
Heggarty and his CoBL (Cognacy of Basic Lexicon) project strikes back... (Against Kassian et al - https://www.degruyter.com/document/doi/10.1515/ling-2020-0060/ - "Rapid radiation of the inner Indo-European languages: an advanced approach to Indo-European lexicostatistics")
Feel free to discuss how much obviously cooked up bollocks this either is or isn't ;)
„Language trees with sampled ancestors support a hybrid model for the origin of Indo-European languages”
ReplyDeletehttps://www.science.org/doi/10.1126/science.abg0818
https://www.mpg.de/20666229/0725-evan-origin-of-the-indo-european-languages-150495-x
Wolfgang Haak and Johannes Krause are coauthors.
Lazaridis is skeptical:
https://twitter.com/iosif_lazaridis/status/1684642571342209026
I am also skeptical.
@Davidski, the paper below is a classic example of bad modelling - shit in, shit out:
ReplyDeletehttps://www.science.org/doi/10.1126/science.abg0818
@ Gaska
ReplyDelete“Proto-Greeks, Pre-Greeks, ””
These mean two completely different things. One is the (simplified) meaning of folk who Brought early forms of Greek, the other is non-Greek .
“And you know what?, to date, there is no R1b-M269 (Z2103, PF7562 or L51>L151) and no R1a-M147 either in chalcolithic Greece”
And why did you ascribe the Cernavodă / Yannaya- I2a with “did not come from the steppes but from Anatolia”
Is it basic lack of knowledge or do you bend the truth ?
Also, you should not be making predictions because you can’t even digest published data staring you in the face.
“I know that for you anything is enough to link the Mycenaean culture with Yamnaya and the Indo-Anatolian languages with the Balkans. I think that this attitude of yours is more a matter of faith than a scientific conviction and the same goes for your theory that M269 originates from the Don River, do you have any proof? or is it just a hunch.””
my framework have been proven irrefutably correct. The fact that you can’t understand that is symptomatic of your own mental shortcomings. You’re happy to claim that R1b-M269 is from Balkans, which it’s not, but I’m not allowed to highlight that Greeks came via the Balkans, which they did ? You’re self -contradictory
So I’m going to drink my champagne and let a “Rocca et al” put you out of your final misery about M269 when the time comes
I guess regarding Heggarty's paper, my thoughts are that it's fascinating that, even after correcting their dataset for the problems they identified, the lexicon still follows the pattern that Indo-Iranian splits before (Balto-Slavic(Italic-Celtic(Germanic))).
ReplyDeleteBut the authors cannot seriously claim a model where they postulate that Balto-Slavic separates from (Italic-Celtic(Germanic)) at 5000 BCE is a model with a "secondary steppe homeland"! What is the archaeological/cultural correlate of this 5000 BCE split? It ain't Corded Ware culture. This is a "We still have an Early Neolithic date for the root and essentially no friggin idea how the structure correlates with any population movements" model.
The tree structure may be interesting in some sense, but essentially estimating the rate of differentiation rate of the linguistically unattested proto-languages from the attested ones seems like it just cannot work, if this is result we are getting.
(Like, this is the tree - https://imgur.com/a/2Q663gv. Proto-BS node splits at 4800 BCE. So unless you want to say that the Yamnaya-CWC is not the point of the split of these groups, and that BS comes from some mysterious unattested steppe-related culture, which does not make any sense, I don't know how the authors can present the narrative in their map as having any plausibility...)
ReplyDelete@Matt,
ReplyDeleteTa, I did see the headline this morning. The photo of the bone shows it is in atrocious condition, I wonder if there is some taphonomic factor(s) which degrades amelY faster than X? Without getting into special pleading it just doesn't seem as confident an identification when dna can't be used as additional verification.
@All, regarding R1b in Greece:
Don't forget that in the late and terminal LBA numerous Italians, some Central Europeans (from or related to Urnfield), and probably some Balkaners (for whom an origin area is less precisely known) had established themselves in/around the Mycenaean palaces including in Crete. It would seem they were primarily males (craftsmen and mercenaries). During the collapse/transition most of them were either killed, went to Cyprus/Levant/Egypt as sea peoples or otherwise left with very little to no linguistic legacy on the recovering population of IA Greece. Either way they could be responsible for some R1b and "elevated steppe" ancestry in LBA and EIA Greece. We will eventually find out from isotopes how this occurred.
In any case as always I advise waiting for the (re)sequencing of the Grave Circle people to put to rest most reservations about who is a real early Greek speaker, who is a descendant of locsl conversion/recruitment within or before Greece, who is a plainly non-Greek immigrant from Europe etc.
@ LGK
ReplyDelete“In any case as always I advise waiting for the (re)sequencing of the Grave Circle people to put to rest most reservations about who is a real early Greek speaker, who is a descendant of locsl conversion/recruitment within or before Greece, who is a plainly non-Greek immigrant from Europe etc.”
Not really . By grave circle times, and given the porous nature of Mycenaean society , grave circle sequences aren’t going to tell us who proto-Greeks were. These elites could very well be G2a or J2-
It’s very simple as we already know how proto-Greek arrived- a fusion of Cernavoda and Yamnaya wave from Bulgaria. Classic free -linguistics is bullshit; that’s why people are hallucinating non-existent migrations from Babino just because they “feel” that’s what should have occurred based on alleged cladal links with Armenian or whichever
^ Tree-linguistics
ReplyDelete@Rob
ReplyDeleteMy point is that they are inarguably Mycenaean Greek speaking elites not far removed from the first written Greek so people can't say (in Gaska sense) oh they are just foreigners from post palatial collapse or whatever. Assuming that a complete male turnover hadn't occurred and that the GC guys aren't some local upstart dynasty, there is is also a greater chance that they would preserve any original (proto-Greek) Y lines not found otherwise in the already sampled elites from elsewhere in Greece. I agree at this time at least some are very likely going to be J2, but since they also have a novel foreign anthropological type (to MBA Greece) I think there is a good chance they are not predominantly recruited locals but preserve a migrant signal in some way.
Luckily the project is also looking at isotopes to help iron this out
@Aram
ReplyDelete"There were at last 7 cases of M269 in ancient Greece.
4 cases from Skourtanioti 2023
3 cases from Lazaridis 2022.
So I simply don't understand what we are discussing here?
AID009 1 (ss) 1 98403 0.08 M R1b1a1b Aidonia LBA 3 65 0.21 -NA- -NA- -NA- 64 576 -0.013 0.00598 0.001 0.013
GLI002 1 (ds) 1 60375 0.05 M R1b1a1b GlykaNera LBA 1 111 0.34 0.02 0.01 0.03 112 1.01E+03 0.037 2.64E-02 0.001 0.027
TIR008 1 (ds) 1 385098 0.31 M R1b1a1b Tiryns IA 4 710 " 2.19" 0.01 0 0.02 489 4401 0.002 3.56E-03
XAN030 1 (ss) 1 481327 0.39 M R1b1a1b J2b1 Chania LBA 4 3157 8.72 0.01 0 0.02 639 5751 0.017 0.00968 0.001 0.015"
Can you tell me which ones are R1b-PF7562? I know XAN030 from Chania in Crete is, but Lazaridis mentions two LBA Mycenaeans from the Palace of Nestor who are PF7562, as well (see the research article, "A genetic probe into the ancient and medieval history of Southern Europe and West Asia", page 6 of 12). In that same article, also on page 6, Lazaridis mentions Sample LYG001, an R1b-PF7562 from Lysogorskaya in the North Caucasus from Wang et al (2019) and says he is "genetically similar to the Yamnaya".
Lazaridis cites those two R1b-PF7562 guys in the Palace of Nestor as examples of the impact of northern migrants on mainland Greece.
Do you know which samples are the two R1b-PF7562s from the Palace of Nestor? I can't tell from the paper or the accompanying spreadsheet.
I'm not sure why a certain knucklehead here thinks it's a big deal that thus far no R1b-M269 has turned up in Chalcolithic (Copper Age) Greece. Why would we expect it to? Greece was not Indo-Europeanized during the Chalcolithic Period.
@ LGK
ReplyDelete''My point is that they are inarguably Mycenaean Greek speaking elites not far removed from the first written Greek so people can't say (in Gaska sense) oh they are just foreigners from post palatial collapse or whatever.''
Yes agree. My main point however is that by then, the PGk question will already be resolved (due to more Logkas & Theopetra-like data from Thessaly and surrounds), and GC data will be a more interesting reflection of how society was functioning several hundred years later.
I suspect that it would *not be the strict patrilineality of post-Beaker western Europe but one more based on acquired status. As you know, post-Yamnaya groups were entering a 'Dark Age' after the mid Helladic collapse, and intuituvely they might have positioned themselves at the top of a new society in the recovery phase although it might not be a given that rule was restricted to them by
@Aram
ReplyDeleteI found the three R1b-PF7562 Mycenaeans from the Palace of Nestor from Lazaridis' Southern Arc paper:
I13506
I13518
I19364
Evidently the first two were father and son and the last one was a third degree relative of theirs.
I have not had time yet to really take a close look at them, but it's interesting that Skourtanioti says that XAN030, that R1b-PF7562 from Chania in Crete, has high Western Eurasian Steppe herder (WES) ancestry.
Do you know of any other R1b-PF7562 among the ancient Greek samples?
Thanks.
@ LGK
ReplyDelete“Don't forget that in the late and terminal LBA numerous Italians, some Central Europeans (from or related to Urnfield), and probably some Balkaners (for whom an origin area is less precisely known) had established themselves in/around the Mycenaean palaces including in Crete. It would seem they were primarily males (craftsmen and mercenaries). During the collapse/transition most of them were either killed, went to Cyprus/Levant/Egypt as sea peoples or otherwise left with very little to no linguistic legacy on the recovering population of IA Greece. Either way they could be responsible for some R1b and "elevated steppe" ancestry in LBA and EIA Greece. We will eventually find out from isotopes how this occurred.
In any case as always I advise waiting for the (re)sequencing of the Grave Circle people to put to rest most reservations about who is a real early Greek speaker, who is a descendant of locsl conversion/recruitment within or before Greece, who is a plainly non-Greek immigrant from Europe etc.”
I thank you for having hinted to the Italians among the “Sea People”, because people who don’t know the million part than me about history, linguistics and all the rest and offends me from more than 10 years tried to hide that and didn’t studied either “The Ethnicity of the Sea People” of the great Woudhuizen, unfortunately dead too soon. I demonstrated more than 10 years ago that among the old languages of Crete there was some Italic one in the “Dienekes’ Anthropology blog” he is or not Lazaridis. And about R-PF7563 (it isn’t true that R-PF7562 is older, it is a recent little survived subclades) had its presence and perhaps origin in Sardinia thus in Italy because what we find in Sardinia was before in Italy.
@Aram & Rob
ReplyDeleteCan you understand that those few cases of PF7562 in the Minoan and Mycenaean cultures are irrelevant when discussing the time and the way in which the Indo-European language reached the Peloponnese? You need M269 (in any of its branches) or R1-M417 between 3,000-1,500 BC (first writings in Mycenaean) to justify the arrival of those lineages from the Yamnaya culture or from the north of the Balkans (Bulgaria etc). Until you find these markers in the Peloponnese everything is, as always, wishful thinking.
@Aram
Try to deal with it, there is no PF7562 in Armenia and Yamnaya.
@Rich
I have saved all your posts on Anthrogenica, it will be fun to reread them. Take my advice, find another hobby. People like you, Rocca, "Chiricaua" Hulan, "Borrico" Ruderico, "Vaudeville" Anglesqueville, "Mr Bean" JDean etc etc, you will be happy that anthrogenica is dead forever, so many of your posts will disappear forever. I know you never admit your mistakes, but at least you could have the balls to apologize for your behavior
@LGK-
The R1b-M269 marker is irrelevant to find out who were the first speakers of Mycenaean because in the Peloponnese there is genetic continuity in both uniparental and autosomal markers since the chalcolithic, this makes it absolutely impossible that some newcomers of whatever lineage could impose their language on the natives. What would be the reason? again the absurd theory of the R1b-M269 elites?, look what has happened with the Griffin warrior. The Mycenaean seems to be an Anatolian affair, the steppe has very little to say in this matter.
@Rob-
I stopped toasting with champagne a long time ago to celebrate my victory over Rocca et al, all of them, including Harvard were wrong and continue to be wrong because they are obsessed with a theory that never had much of a chance. It has already been proven that M269 does not originate from the Don river so I have also toasted with vodka to your health. There is also no I2a-L699 in the Peloponnese so this lineage is also irrelevant when we talk about the Mycenaean. And calm down, your ability to reason is getting smaller and smaller
@Matt
ReplyDeleteCheers. Interestingly they float a new reliability number, 96% this time. It will be interesting to see on what they base that.
@Gaska
ReplyDelete"This makes it absolutely impossible that some newcomers of whatever lineage could impose their language on the natives. What would be the reason"
Why is it impossible, because you can't comprehend it? This sort of local elite recruitment happened throughout IE expansion and is also quite normal in history. As to how an introduced proto-Greek became ascendant over whatever was spoken there earlier by a larger native population, some answer must be had regardless of whether they came from the steppe via Balkans, or Anatolia. You seem prepared to accept the latter unconditionally but somehow become very sceptical when considering the former, its ok we all know why LOL
@Gio
Regrettably some of the key theories of western-central Mediterranean origins for some sea peoples (Sherden, Shelekesh) are contained only in Italian language literature that is several decades old now, and largely obscure to new anglophone-only students.
A reminder very relavent to this new linguistic paper.
ReplyDeleteLinguist Don Ringe in his recent paper wrote:
“The general rule about extrapolating into the unobserved past still applies: results are comparatively secure when different lines of evidence converge on the same result. Computational cladistics yields only one line of evidence; therefore, it must be used in conjunction with traditional methods, archaeology, ancient DNA evidence and everything else that might be relevant”.
Using traditional methods, archaeology, ancient DNA evidence and everything else that might be relevant we can date the separation of Indo-Slavonic languages into Proto-Balto-Slavonic and Proto-Indo-Iranian. It was the time when R1a-Z645 split into R1a-Z283 and R1a-Z93, and CWC groups rich in R1a-Z93 migrated east to become later Indo-Iranians. It was around 3000 BC.
https://postlmg.cc/QHtfTPq3
@LGK
ReplyDelete“@Gio
Regrettably some of the key theories of western-central Mediterranean origins for some sea peoples (Sherden, Shelekesh) are contained only in Italian language literature that is several decades old now, and largely obscure to new anglophone-only students“.
That could also be true (partially), but I wrote probably 20000 letters from 2007 (perhaps in my bad English) about these arguments, and the genetic papers of Morelli et al (unfortunately dead too soon), Francalacci, Boattini etc were published in English and upon them (above all Francalacci et al 2013 /2015) I based my analyses. Certainly if the owners of the blogs write that I am an antesemite and ban me or don’t publish my letters (I criticized not only the Harvardians but also the Phoenician-driven papers of Zalloua et al, which, at my advise, demonstrated the other way around as to what they believed), all that may happen. A Bulgarian Pomak deleted his post in the R1b group at FB because I wrote that his R1b could be also a Turkish back migration of haplogroups migrated to central Asia from Yamnaya (probabvly), but a scientist who searches for the truth, should take into consideration all the possibilities and verify them. Woudhuizen about the Sea People took into account all the hypotheses.
Not only SHRDN and SHKLSH, but the WKSH were very likely the Oscans, and of the type Osco-Umbrian was the Italic language in Crete, thus probably R-PF7562, not present in the large amount of Sardinian R-PF7563, could be present in continental Italy beyond the large presence in aDNA in the Balkans (Italy and the Balkans, with the emerged Adriatic, were the same land).
This comment has been removed by the author.
ReplyDelete@ arame
ReplyDeleteThe Smyadovo outlier is too old to be related to Cernavoda by several hundred years. Moreover his *cultural* context is Karanovo VI
@All
ReplyDeleteAre there any redeeming features about this paper?
https://www.science.org/doi/10.1126/science.abg0818
If a method recovers humans to be most closely related to kangaroos, not chimpanzees, then the problem lies with the method.
DeleteLikewise, I cannot claim to understand the maths behind this study, but any quasi-skilled linguist can immediately recognize that the results are non-sensical, despite the fact that the authors have been working on this paper for 8 years.
This study creates a big problem in the field, not only because it further alienates many linguists from aDNA, but also due to the fact that all future papers working on the origins of IE will have to cite this paper as the latest development in the field.
Glad to see Lazaridis expressing his concerns about this paper, but as long as there is no published rebuttal, this study will reign supreme.
The mistakes of a fool need forty wise men to be solved, as we say where I come from...
@ East Pole
ReplyDeletealthough I agree with a Balto-Slavic-I-A clade in broad terms, much of that tree you posted is wrong from a population (Ie reality) perspective. Albanian -German clade is BS, & according to the CW model, Germanic and many other languages have digenic origins
So linguistics alone cannot handle internal branching within languages, and the classic family tree approach is outmoded
@Davidski, Chundra Cathcart says - https://www.science.org/content/article/new-language-database-narrows-search-first-speakers-indo-european
ReplyDelete"“Some aspects of the tree’s branching structure and chronology are improbable or impossible,” says Chundra Cathcart, a University of Zürich linguist who was not involved with the study. Still, he says the data collected for the study are “really, really top notch” and will be “an invaluable resource to at least a generation of researchers.”"
"Since the analysis was completed, Cathcart says language tree-building methods have continued to improve. Now there are more ways to detect words borrowed long after languages split, for example. Applied to this study’s data set, the newer methods might return a completely different timeline. It’s “a very important contribution,” he says, but “the debate is not 100% settled.”"
Heggarty discusses the Ringe IE typology and the question of Indo-Balto-Slavic in the paper's supplement:
ReplyDelete"There is clarity and consensus on all of the main well-attested branches of Indo-European: Anatolian, Tocharian, Indic, Iranic, Baltic, Slavic, Greek, Armenian, Albanian, Germanic, Italic and Celtic, and on the pairing of Indic with Iranic. There is also strong, if not universal, support for Baltic pairing with Slavic.
Beyond that, however, there is very considerable uncertainty and debate as to which of the major branches relate most closely to which others, and thus on the higher order branching sequence in the early stages of Indo-European divergence. Reference classification catalogs like Glottolog default to a rake structure. "
Also:
"In our main results, the most significant difference with the Ringe tree is, not unexpectedly, the placing of the Balto-Slavic branch. Unlike in the Ringe tree, where Balto-Slavic forms a clade with Indo-Iranic (as the last major node in the tree), in our results support is highest (posterior probability 0.63) for Balto-Slavic forming a clade most closely with Germanic-Italic-Celtic instead; support for Balto-Slavic + Indo-Iranic is far lower (0.12).
Notwithstanding the difficulties discussed in §7.6.2.1 above, the Ringe tree is nonetheless widely discussed in the linguistic and aDNA literature, so as an additional sensitivity analysis (SA6b) we applied clade constraints to force its topology, including the controversial Indo-Iranic + Balto-Slavic node. As presented in Fig. 2 in (5), the Ringe topology does not in fact include a definite position for Albanian or Germanic, however.
We therefore implemented a series of backbone constraints to enforce a topology of (((((Indo Iranic, Balto-Slavic) (Greek, Armenian)) (Italic, Celtic)), Tocharian), Anatolian), while allowing Germanic and Albanian to move freely.
This SA6b analysis produced a median root date of 8564 BP (7228‒9795), earlier than our main analysis by 444 years (5.47%). So constraining to the Ringe topology in fact produced results chronologically less rather than more in line with the shallow chronology of the Steppe hypothesis. In part this is because this topology forces Indo-Iranic+ Balto-Slavic together, despite the scale of the difference between them. It is not unexpected that enforcing the Ringe topology does not greatly affect our overall results, however, because as noted above, the Ringe tree is found within our posterior distribution of possible trees, although with a very low probability of 0.07%."
and also:
"So the only three characters in (Ringe's model) that ostensibly support a unique [Indo-Iranic + Balto-Slavic] node are all highly problematic, much debated in Indo-European linguistics, and not widely accepted as probative that the node did actually exist.
Ringe himself warns that overall, even his preferred tree is supported by very few characters, and contradicted by more: “18 characters are incompatible with the ‘best’ tree returned … in computational terms our result is a total failure” (p. 85 in (50)). Ringe concedes that “the higher-order subgrouping of the lE family has remained an unsolved problem for so many generations partly because the evidence is genuinely meagre” (p. 98 in (50)). In his own analysis, “the evidence for virtually all the larger, non-traditional subgroups that our algorithm posits is so slender” (p. 104 in (50)): no more than a handful of binary characters for any higher node. This is as reflected in the very low (>0.3) support values for the first three splits in our results here, and only 0.15 support for [Indo-Iranic + Balto-Slavic]. "
Full comments here: https://pastebin.com/Xp1enQr9
It doesn't seem difficult or impossible to imagine as compatible with a Steppe Hypothesis a model where the splits are Anatolian, then SE European IE (Greek, Armenian, Albanian), then Indo-Iranian, then North and Western IE (Balto-Slavic-Celtic-Germanic-Italic); that could simply follow a Pre-Yamnaya->Yamnaya->Eastern CWC->Western CWC split of some form (whatever pre-Yamnaya represents). CWC is somewhat diverse enough that in some groups R1a-M417 descendants could rise to high frequency, R1b-M269 L23 in others (and I1 too) with founder effects and reduction of y-dna diversity that we see in CWC.
ReplyDeleteThe problem with these trees is the very early date given than that they break with a really strong consensus as to the higher-order topology.
(Unless we accept that there was two-thousand year old split between the ancestors of Balto-Slavic and Celtic-Italic-Germanic maintained in steppe culture from virtually at the time of Khyalynsk, at stage up to expansion about 3000 BCE... Which is about as unbelievable, if not more so, as the "Basque from the Steppe via Yamnaya; IE from the CWC" idea that was bubbling around a few years ago).
This comment has been removed by the author.
ReplyDeleteHeggarty is in effect asking us to believe that Corded Ware, when it acts as the vector spread IE to Northern and Western Europe in his model, is already split between languages that are slightly diverged as today are Icelandic and German, or standard Italian and French, or Slovenian and Polish, or Welsh and Breton. Do we really think that such a genetically narrow culture (small population size and linkages among early Corded Ware individuals found and Yamnaya and CWC individuals) could have harboured such a level of diversity as this?
ReplyDelete
ReplyDelete@LGK
It is impossible because it is absolutely undemonstrable. Elite theory is a failed attempt to hide the inability of geneticists to prove the genetic continuity between Yamnaya and the Mycenaean culture.
Look, everything is much simpler, just find any branch of M269 in northern Greece or the Peloponnese between 3,000 and 1,500 BC, which lead you to the PF7562 Greek samples (Palace of Nestor and others) and then Lazaridis will be right and the Mycenaean have its origin in the steppes. Otherwise, you will only be able to tell fairy tales (sea peoples, small migrations of explorers, traders, warriors etc). On the other hand, the genetic relationship between Anatolia & Greece is evident since the mesolithic, we can even speak of a genetic continuum, its very hard to think that these regions spoke different languages when their genetic markers are so similar.
You seem prepared to accept a Mycenaean steppe origin and to deny an Anatolian origin, its ok, we all know why LOL
Max Planck agrees with Harvard on the origin of IE south of the Caucasus although fixes its origin around 6,000 BC. Heggarty and his team disagree on another essential issue, the Balkan languages (including Albanian and Greek) did not originate in Yamnaya but in Anatolia which is precisely what I was saying.
ReplyDeleteThe hybrid theory once again attributes a secondary role to Yamnaya, although I do not quite understand its reasoning.
Is Heggarty saying that in the steppe (5.000 BC) only Balto-Slavic, Germanic & Italo-Celtic were spoken? and that 2.000 years later the CWC brought this common language to mainland europe?
And how do they explain the genetic relationship between CWC and India (R1a marker?) or Balto-Slavic and Indo-Iranian, now linguists say they are unrelated?
if this is true, the Kurgan theory in either Gimbutas version or Anthony version can rest in peace.
Rich
ReplyDeleteHere the link about PF7562 in ancient Greece. I posted earlier it.
Kydonia 30= XAN030
https://discover.familytreedna.com/y-dna/R-PF7562/tree
@Gaska
ReplyDelete"@Rich
I have saved all your posts on Anthrogenica, it will be fun to reread them. Take my advice, find another hobby. People like you, Rocca, "Chiricaua" Hulan, "Borrico" Ruderico, "Vaudeville" Anglesqueville, "Mr Bean" JDean etc etc, you will be happy that anthrogenica is dead forever, so many of your posts will disappear forever. I know you never admit your mistakes, but at least you could have the balls to apologize for your behavior"
I'm flattered. Too bad you didn't learn anything from them.
Instead of reading my posts, you should take the time to read important books you still haven't read, like Reich's Who We Are and How We Got Here, Mallory's In Search of the Indo-Europeans (a little dated now but still good), and Anthony's The Horse The Wheel and Language.
Actually, my behavior toward you has been rather restrained. You are the one who owes everyone an apology, especially your own Y-DNA ancestors, whose steppe pastoralist origin you signally fail to acknowledge, despite the literal mountain of genetic evidence. Read Reich's book.
". . . @Rob-
I stopped toasting with champagne a long time ago to celebrate my victory over Rocca et al, all of them, including Harvard were wrong and continue to be wrong because they are obsessed with a theory that never had much of a chance. It has already been proven that M269 does not originate from the Don river so I have also toasted with vodka to your health. There is also no I2a-L699 in the Peloponnese so this lineage is also irrelevant when we talk about the Mycenaean. And calm down, your ability to reason is getting smaller and smaller"
The only victory you're truly entitled to celebrate, Gaska, is a victory over good sense, rationality, and self control.
And that last sentence, addressed as it is to Rob, who has forgotten more about population genetics than you'll ever know, is hilarious. It is yet another example of your victory over good sense, rationality, and self control.
It's amazing that you are still blazing away, despite the fact that you have been publicly humiliated time after time after time. It's mind boggling.
@LGK
ReplyDelete"@Gaska
"This makes it absolutely impossible that some newcomers of whatever lineage could impose their language on the natives. What would be the reason"
Why is it impossible, because you can't comprehend it? This sort of local elite recruitment happened throughout IE expansion and is also quite normal in history. As to how an introduced proto-Greek became ascendant over whatever was spoken there earlier by a larger native population, some answer must be had regardless of whether they came from the steppe via Balkans, or Anatolia. You seem prepared to accept the latter unconditionally but somehow become very sceptical when considering the former, its ok we all know why LOL"
What you wrote above makes perfect sense, LGK, given the fact that Greek was pretty obviously imposed upon a large non-Indo-European substrate, which left its mark on the language, as Mallory so ably recounts on pages 66-71 of his book, In Search of the Indo-Europeans.
Interesting that R1b-PF7562 skeletal remains, with high steppe DNA, have turned up among the Bronze Age Mycenaeans, not only at Pylos on the Greek mainland but in Crete, as well. PF7562 is fascinating to me, since it is a branch under R1b-M269 and parallel to R1b-L23, the father of R1b-L51 and R1b-Z2103, both of which are also well represented among steppe pastoralists.
@Gaska
ReplyDelete"Heggarty and his team disagree on another essential issue, the Balkan languages (including Albanian and Greek) did not originate in Yamnaya but in Anatolia which is precisely what I was saying."
You are utterly clueless when it comes to languages. Greek and Albanian are not archaic branches of PIE, like Anatolian or Tocharian. This is easy to tell from present day linguistic analysis. They are rather late offshoots from the steppe. There is about a 0.00% chance they come from Anatolia.
The Balkans (Both western and eastern) had samples showing 70 to 85% steppe admixture (accompanied by kurgans) since the EBA. This is as high or higher than even Corded Ware. There is no argument here for what you're saying linguistically, archeologically, or genetically.
As for Anatolian itself, it just has to split at 4,000 to 5,000 BC. There is no evidence it was actually spoken IN Anatolia then. The earliest attestation we have is Hittite ~19th century BC.
For all we know, Proto-Anatolians were just distant eastern cousins of the Yamnaya that migrated south through the Caucausus and were diluted genetically in the 4th millenium BC. Look at that royal R1b Arslanteppe sample. This was ALSO presented as an option by Lazaridis if you read the paper. Lazaridis didn't say he was 100% sure that PIE originated south of the Caucausus. That was just 1 option.
@Gaska
ReplyDelete“And how do they explain the genetic relationship between CWC and India (R1a marker?) or Balto-Slavic and Indo-Iranian, now linguists say they are unrelated?”
OK. So Balto-Slavic and Indo-Iranian are unrelated and the original proto-Indo-Iranian language and culture was closer to Hittite than to Slavic. But something happened in the Bronze Age. Massive migrations of CWC related R1a rich Slavic tribes to Central Asia, and to India and Iran very strongly influenced languages, cultures and religions of Indo-Iranian tribes, which originally, as we know, were closer to Hittite. Is it acceptable to you?
On the Heggarty paper: https://twitter.com/nezih_seven/status/1684936978284892161?t=MvYYBDjJxaAp7SaNDe825w&s=19
ReplyDeleteIt's funny because Lazaridis & co. literally talked about Armenian, Albanian, etc... being brought over by steppe ancestry, and Heggarty is like "cool story bro" while completely ignoring and STILL QUOTING Lazaridis. Imagine quoting a paper that literally disagrees with 95% of your model, because there is a 5% in there that you want to use for your own purpose.
ReplyDeleteSo yeah, Heggarty's methodology is dog shit, and will never garner consensus. Like David said, not every paper published is the end all be all. Plenty of kooky academics that publish kooky shit no one takes seriously. You need the consensus of your community for something to be considered fact.
The Heggarty paper is, in my opinion, an huge misstep by Max Planck SSH and they should either retract it or refute it with their own future studies. As Matt said, putting aside any questions of internal classification, the chronology alone is so absurd and inconsistent with all available evidence as to be easily rejectable out of hand.
ReplyDeleteTrying to overturn a century of IE studies and recent aDNA evidence with a single unproven linguistic software analysis is simply the tail trying to wag the dog, and it won't work.
@Aram You completely missed my point, non-Yamnaya ancestry in samples eliminates them as proto-Greek speakers regardless of ydna, especially when they're pre-Mycenaean. CWC ancestry has no relation to Balkanic languages.
ReplyDeleteIndeed Thrace is far more likely, as is Anatolia, although influence from places like Italy has been found in LBA Crete (Skourtanioti mentions that there are archaeological finds about this in her paper, others mentioned it here too). So there was population movement with CWC-related ancestry into Greece at points
@Rob Stronger than Yamnaya's which were absent iirc
@Gaska Might interest you, Mycenaeans have elevated Levant-like anceatry compared to Minoans. Obvious NEastern geneflow. Griffin Warrior has this too
Anyway as you said until we get early Mycenaean samples, and see with what samples they're linked to, we can only just speculate. Ultimately even the assumptions that proto-Greek was spoken in North Grece is an assunption, following Ven Beek 2022 one can assume that they were spoken in South or South-Central Greece. Plenty of option but limited clues so far.
Entertaining to see strongly opinionated people bicker with eachother tho ngl
@Matt Didn't read the paper yet but don't they say that the "split" date isn't a language split but just the time divergence began? The actual split (languages not mutually intelligible anymore) could occur much later and theoretically doesn't disturb language splits that can be linked to aDNA/archaeology.
ReplyDelete"It is important to clarify what a split represents in the methodology followed here. A lineage split does not by any means entail fully distinct languages. Even if any two language varieties (taxa) differ in cognacy for just a single meaning in the basic set of 170 reference meanings, then in the phylogeny they are already necessarily distinct sublineages, split from each other. A split does not, then, correspond to the much later stage by which so many further differences have accumulated between those divergent lineages that they have become two fully-fledged different languages and are no longer mutually intelligible."
Also yes I-Ir being unrelated to B-S seems confirmed again given Olander 2022 and Kroonen 2022 traditional methods. And the early I-Ir split (Kroonen 2022).
Balkanic via Anatolia also works geographically but not sure about this one.
Overall their tree grouping is pretty close to the recent groupings with traditional methods.
@Orpheus
ReplyDeleteYou completely missed my point, non-Yamnaya ancestry in samples eliminates them as proto-Greek speakers regardless of ydna, especially when they're pre-Mycenaean.
Y-DNA trumps autosomal DNA when it comes to language change in ancient times, because such changes were usually brought about by males.
The order is this: Y-DNA, autosomal DNA, mtDNA. Surely you can understand this?
But anyway, can you list the Balkan samples that you believe have Corded Ware ancestry as opposed to Yamnaya/Balkan ancestry?
Here's something interesting from pages 332-333 of the Supplementary Material of the Lazaridis et al paper, "The genetic history of the Southern Arc: A bridge between West Asia and Europe":
ReplyDelete"Given that within the phylogeny of R-M269 (R-PF7562, (R-L51, R-Z2013 [sic - Z2103 is meant])) both R-PF7562 and R-Z2013 [sic, Z2103] have their earliest examples in the North Caucasus and steppe to the north, the most likely hypothesis is that the entire R-M269 clade originated there as well, with R-L51 representing a lineage that eventually became highly successful in mainland Europe, R-PF7562 a lineage that did not achieve the prominence of its relatives, and R-Z2013 [sic, Z2103] became highly successful (briefly) as part of the Yamnaya culture and its offshoots . . . "
Makes sense. It seems to me if one backs down the tree a bit toward the root to the take in the picture of the whole R1b-L389 clade, with P297 and its scions, M73 and M269, and with V1636, an origin somewhere in Eastern Europe is indicated.
Why does Lazaridis keep using the term "mainland Europe"?
ReplyDeleteHow is the steppe not mainland Europe?
@Davidski
ReplyDelete"Why does Lazaridis keep using the term "mainland Europe"?
How is the steppe not mainland Europe?"
Yes, it is weird. Maybe it's because the steppe is a prairie sea of grass. Once you go west beyond the sea of grass, you're on the mainland?
He probably should have written "Europe west of the steppe" or "peninsular Europe" or something like that.
It doesn't matter how they choose to define branching, there is no conceivable way to accommodate Slavic and Celtic diverging from each other even slightly ~5000BC with the steppe hypothesis (and considering how robust our evidence for a steppe origin of proto-II and Greek is, we may as well extend this to ~5500BC). It betrays any basic understanding of the steppe's archeology.
ReplyDeleteIt's also essentially inconceivable genetically, as we know the bulk of the "CHG" on the steppe is much older. Admixture with a south caucasus source can only reasonably happen closer to 4000BC (in line with Lazaridis). Are we also supposed to believe these divergences maintained themself as they crossed the caucasus and indo-europeanized the Steppe?
I don't really understand why one even feel the need to contemplate a hybrid model under this tree, if you truly believe in these results. I guess they find some of the Steppe hypothesis evidence overwhelming.
@All
ReplyDeleteI've skimmed through the Heggarty paper, and I don't have the strength to write a critique.
The main point is that it offers nothing tangible, and instead it ignores tangible evidence.
I'm sure that over the next few months and years we'll see some good rebuttals, but the paper should never have been published in the first place.
@Davidski
ReplyDeleteThe methodology employed by Heggarty was already dealt a death blow by the linguist Dr. Asya Pereltsvaig and the historian Dr. Martin W. Lewis in their 2015 book, The Indo-European Controversy: Facts and Fallacies in Historical Linguistics.
In a nutshell, IE cannot be as old as Heggarty claims.
@Rich S.
ReplyDeleteRight, but since Heggarty has tried it again at this level, we now require someone to totally debunk him again at the same level.
The question is, why wasn't this done during peer review at Science?
@Davidski
ReplyDelete"@Rich S.
Right, but since Heggarty has tried it again at this level, we now require someone to totally debunk him again at the same level.
The question is, why wasn't this done during peer review at Science?"
You're right, of course. It's kind of like a game of Whack-A-Mole. Wherever and whenever the mole of error pops up, he has to be whacked down again.
That's a good question, too. Maybe they had no competent historical linguists reviewing Heggarty's paper, or, if they did, perhaps they let it go as a speculative argument?
Heggarty seems to be on a personal crusade for a Fertile Crescent origin of PIE because his central tenet has been - language dispersals are associated with a productive economy (e.g. you can have a look at his work on the Andes region), and thinks some variant of the steppe hypothesis threatens that. Apart from being obtuse for being unscientifically rigid (instead of saying 'oh well if that's what the evidence says..''), he should realise that steppe pastoralists did have a productive economy, often moving into sub-neolithic, partially de-neolithicized and/or recently depopulated regions.
ReplyDeleteBut his other issue is the trend to want to 'sexy' up linguistics by fusing it with statistics. Linguistics should remain the traditional craft of analysing cognates and deciphering ancient inscriptions, etc . You cant reduce language change and complex human behaviour to a Y_DNA analysis programme. His methdology is null & void.
@ Oprheus
Do you seriously think that Greeks having ancestry from CWC instead of Yamnaya with a language vector coming from the Levant is 'equally as likely' as deriving from a post-Yamnaya groups further north ?
@Orpheus-
ReplyDeleteYes, the Mycenaeans also have Levantine blood and some of them (Lazaridis) have no hint of steppe ancestry. We also know that the steppe component had not reached the Peloponnese by the early Bronze Age (Perachora, 2,000 BC), so only a good handful of samples from that region between 2,000-1,500 BC can help solve the mystery. The turn of events regarding the Indo-Europeanization of Greece has been radical and everything points to Asia Minor. We'll see what happens, I have always been surprised that Z2103 has not yet appeared in Mycenaean culture.
ReplyDelete@East Pole
I am not a linguist and have no idea whether or not Balto-Slavic is related to Indo-Iranian, I only know that there is an obvious genetic relationship between CWC-derived cultures (overwhelmingly R1a) and South Asia. I don't know if that genetic relationship brought there the Indo-European languages or if they predate the arrival of R1a and frankly I don't care.
I just wanted to emphasize some inconsistencies in Heggarty's paper that seems to link the Indo-European languages of South Asia to Anatolia and not to the steppes. You should ask him
Apparently, they had this statistical link analysis ready for a long time, but they decided to publish it after Lazaridis' article. By itself, this methodology cannot say anything about the geography of the origin and spread of languages. Conclusions about the homeland of Indo-Europeans to the south of the Caucasus are based on the arbitrary synchronization of this linguostatistical scheme with paleogenetic data on the distribution of CHG. And if linguists will argue with linguistic conclusions, then fans of paleo DNA can also argue with this synchronization of CHG. For example, take Greece. In the Neolithic of Greece, either there is no CHG at all, or in some samples it will amount to 5%. Up to 15% of CHG appears only in Minoans and more than 30% only in Logkos about 2000 BC. So their theory crumbles from the very beginning. The second problem is their claim that Iran is Neolithic is Indo-Europeans. In this case, the language of Elam should be Indo-European, which is also not the case. The third problem. How did the Indo-Iranian vocabulary get into the Uralic languages, if the Indo-Iranians went to India from Iran without entering the steppe.
ReplyDeleteI do break a little from most of you guys, in thinking that there is actually a purpose in assembling such a database and trying to address why trees built on these methods seem to work for an array of other language familes. Papers after all have been published giving sensible time depth results for Austronesian (Greenhill), Sino-Tibetan (Sagart / Zhang), Dravidian (Kolipakam), Uto-aztecan (Greenhill) and more (and likely a growing number of families).
ReplyDeleteThe method doesn't just simply return implausibly much earlier dates for all languages to which its applied but often generally the "right" (expected) sort of time. The whole argument of "This method is at root flatly inapplicable and useless" would be more compelling (or persuasive at all) if such results didn't exist outside IE.
But perhaps in Indo-European there is some extremely rapid phase of massive lexical differentation that just isn't common among these other families, due to some dynamics of competition and admixture that aren't common, and this makes inferring time depth from a body of recent languages evolving at a slower rate something that does not work well. And if that's so, that tells us something about early Indo-European speaking societies compared to later ones.
Or, "Have the sampled languages systematically evolved lexical substitutions much more slowly than an unsampled ancestral phase?".
It seems to me like the way forward challenging this conclusion may be less about linguistic arguments, and more about simulating data where the rate of lexical substitution varies over time across trees, and only later parts of the tree are sampled. Do the methods then actually still work? Do they really recover the faster substitution rates in unseen parts of the tree?
But this is more for mathematicians and computer scientists to do than linguists.
Of course, the longstanding complaints about the implausibility of Heggarty's argument about linguistic paleontology still stand. Effectively it seems to me he's still in a position of saying "All these languages, which diverged thousands of years ago, then systematically applied their term for 'turning' to wheel terminology, in systematically an identical way, when wheels were introduced". Can we really believe this? (In addition to the lack of archaeological correlates of the (Baltic-Slavic)-(Celtic-Germanic-Italic) split and expansion, in their model at ~5000 BCE).
Re; Pereltsvaig, IIRC her critique focused on whether the analysis got it right when it came to the relationships among and time depth of modern languages, and also on the lack of direct historical relationship between historical languages and their modern descendants. Although I remember her comments only from her blog post - https://www.languagesoftheworld.info/bad-linguistics/malformed-language-tree-bouckaert-colleagues.html.
ReplyDeleteSo H and his large body of supporting expert historical linguists at the Planck, who have put their names to producing the data for this, have reworked the dataset and methods to essentially get those relationships and their time depth right (all her little whinges about "Well, the method isn't putting Polish with Czech so it's not right" and so on), by removing the imbalances in cognates per language that drove errors. (This is things like, as they discuss, where in the previous IELex data-set, Ancient Greek had large numbers of variants per term included, some of which were little used in practice, which drove the algorithm to estimate that such terms made it a deep branching side branch of the ancestor of modern Greek, because such was necessary to evolve a large body of terms which were supposedly not found in the descendent). Or as right as they can be got under conditions of ongoing contact areas.
(On a tangent I expect Pereltsvaig's comments fell somewhat flat among non-linguists; from that blogpost she employed some acidic and withering style - "malformed" and so on - that perhaps literary critics would find adds to the argument, but I expect editors in scientific journals would almost always view as attempt to hide a lack of substance, whether that's the case or not.)
@ Matt
ReplyDelete“The whole argument of "This method is at root flatly inapplicable and useless" would be more compelling (or persuasive at all) if such results didn't exist outside IE.”
How do we know that it’s not due to chance ? Which languages were they and how was it verified ?
language change is non-constant . How did they cater for that ?
I know some have an empassioned need to “do stats stuff”, as if it’s legitimising results, but imo it just hides a lack of real knowledge, loses granularity and can sometimes be downright BS
Tinkering with stats for these kind of things should be icing on cake sort of stuff, not the basis of a model in place of analogue data
You can only have a “hybrid model” when you have 2, or more, adjacent societies in contact.
ReplyDeleteYou can’t have one speaking Greek and Albanian (lol; a medieval language in 5000 bc Mesopotamia), then other parts of the family spoken by a wholly different society 5000 miles away
@Dragon Hermit
ReplyDeleteNot only I am utterly clueless when it comes to languages, the same thing happens to me with astrophysics, quantum mechanics, medicine etc etc, that's why I participate in this forum to learn from people as intelligent as you.
But whenever someone says that there is a 0% chance of something happening (in this case that the Mycenaean has its origin in Anatolia) or when someone says that there are no arguments for what I am saying (linguistically, archaeologically or genetically) I always think of the number of people who have made, do and will make a fool of themselves.
I don't want to convince you of anything, we all know that the Kurganists are experts in interpreting the available genetic data in a peculiar way. That leads you to think that an R1b in Arlanstepe and a 0,5 % of autosomal steppe ancestry in Anatolia are good arguments to think that the homeland of the Indo-European is the Pontic steppes, and yet you do not find enough arguments to defend an Anatolian origin of the IE, when the Mycenaeans and the Anatolians are overwhelmingly NOT R1b and not R1a. Ok, no problem, you can think what you want, but do not try to convince anyone because you are only showing that you do not defend a scientific argument but a dogma of faith
The steppe hypothesis always saw all Indo‑European, Anatolian included, originating on the steppe. Prospects for solving the Indo‑European enigma are brighter now, free from the outdated preconception that the Steppe must have been its earliest, original source. And, what about the steppe ancestry? None in Anatolia, a decided minority in Armenia, and just a small fraction in Mycenaean Greece and generally across the Iranic and Indic branches. You may continue to think that the Yamnaya culture is the homeland of Indo-European, to me that is just a bad joke
@Rob, I listed the language families and the specific authors in the previous paragraph. Should be identifiable by searching the author and language family name.
ReplyDeleteWhether results can arise from chance and the relevant p values for getting the right date frames by chance would be another mathematical question that could be answered by simulations, I expect.
@Matt
ReplyDeleteI do break a little from most of you guys, in thinking that there is actually a purpose in assembling such a database and trying to address why trees built on these methods seem to work for an array of other language familes.
But if they're not getting coherent results that align with basic facts then they should just publish the dataset at the Max Planck website, not as part of a paper in Science.
Papers after all have been published giving sensible time depth results for Austronesian (Greenhill), Sino-Tibetan (Sagart / Zhang), Dravidian (Kolipakam), Uto-aztecan (Greenhill) and more (and likely a growing number of families)
I don't think any of these language families can be compared to the Indo-European family in terms of the number of different populations speaking Indo-European and the massive area across Eurasia that these groups inhabit.
This implies that there is something special about the Indo-European family and potentially how it expanded and grew.
@Davidski, right - contact, competition and rapid dispersal might lead to an especially rapid period of change - "We're not going to use that ancestral word, because that's what our competitors over there are using, so we're going to use this word instead, which we made up from some other terminology or adopted from other people who are around here", etc. These kind of dynamics, and I speculate a bit here, might lead IE to unusually rapidly differentiate relative to languages that are doing a more island-to-island dispersal or taking up isolated residence in mountainous areas and dispersing slowly as farmland gets used up.
ReplyDeleteSo that's possible and why I think statisticans should respond to this paper, to look at the mathematics of how confident we can be that the rate of the unsampled part of the tree can be inferred from the sampled tree, under simulations where they definitely differ. But I think if they'd just published it on a website on a blog post with a dataset, that probably wouldn't happen, and should happen. YMMV on the thresholds that science magazine as a journal should be applying.
@Rob, also Bantu expansion was tackled in this paper (https://www.pnas.org/doi/10.1073/pnas.2112853119). I can't remember if there was a paper on Niger-Congo as a whole.
ReplyDeleteThis comment has been removed by the author.
ReplyDelete@Rich
ReplyDeleteReally? What can I learn from those individuals? nothing about genetics, linguistics, archeology or anthropology. Now I understand why you think the way you think, if what you know about genetics you have learned from them, then it is easy to understand that conversations with you are boring and monothematic.
I remember when you said that we R1b-P312 Basques gave up our Indo-European mother tongue because of our matrilocal culture. You know that this theory has even been discussed in a Spanish university as an example of nonsense and ignorance?
By the way, the Reich, Mallory & Anthony trilogy may be useful for some rednecks, but here in Europe they are just one more of the many options that are being debated. And of course I will continue celebrating my victory over your beloved Yamnaya, it represents the triumph over dogmatism and lack of scientific criteria.
There is definitely a tradeoff between "This paper has a substantial chance of being wrong, but offers some substantial new scientific work and methods that need to be published and become part of the official scientific record to gain response that moves things forward" vs "If this paper is wrong, its damaging enough that we should not print it". For IE linguistics I think that threshold is a lot lower than say, cancer drugs or something, but yeah, there definitely is an argument either way.
ReplyDelete@Gaska
ReplyDelete“I am not a linguist and have no idea whether or not Balto-Slavic is related to Indo-Iranian, I only know that there is an obvious genetic relationship between CWC-derived cultures (overwhelmingly R1a) and South Asia. I don't know if that genetic relationship brought there the Indo-European languages or if they predate the arrival of R1a and frankly I don't care.”
Linguists say Balto-Slavic and Indo-Iranian are unrelated. But surely there are lots of similarities and common features. You don’t have to be a linguists to see it:
“The list of common words and other features which are special to the two groups is clearly impressive, and the whole of the material must be referred to the period of Primitive Indo-Iranian. When on the contrary we look for signs of special contact between Iranian itself and Slavonic (or Baltic) we find that there are practically none. It is true that some of the words that are listed above are found only in Iranian and not in Sanskrit, but it is equally possible to point out others in which the reverse is the case.”
https://books.google.pl/books?id=cWDhKTj1SBYC&printsec=frontcover&hl=pl#v=onepage&q=%22%C2%A74.%20Indo-Iranian%20and%20Balto-Slavonic%22&f=false
The only way to explain it are the Bronze Age migrations of Slavic tribes to Central Asia, India and Iran which influenced Indo-Iranian languages and religions there.
@EastPole
ReplyDeleteLinguists say Balto-Slavic and Indo-Iranian are unrelated.
The consensus among linguists is that they are closely related.
You have to differentiate between a couple of papers by Heggarty and Gray and the academic consensus.
@Davidski Ah yes the famous IE speakers basques, iberians, etruscans and the famous uralic speakers lithuanians. Or, the famous non-IE R1a clans dominated by R1b early CWC clans, no? Not ro mention this logic fails once the assumed initial haplogroups of a population are wiped out by newer and often novel ones that come to dominate (but the language magically remains).
ReplyDeleteSamples and models on them can be found in the supplementals of the papers that first analyzed them. Skourtanioti et al 2022, Clemente et al 2021, Ringbauer for IBD (Logkas)
@Rob Not at all, the ancestry in Mycenaeans isn't CWC. It's 1:10 Yamnaya:Minoan-like (Lazaridis et al 2022)
Obviously no IE language came from the Levant, Levant-like ancestry is simply an indicator of Near Eaatern geneflow in post-Minoan times (which also apparently the Skourtanioti samples lacked)
@Rich S. What's their argument for PIA (not PIE) not being able to be ~6000BCE old?
@Orpheus
ReplyDeleteYou obviously know that the academic consensus is that proto-Greek came from Yamnaya, and there's nothing you can do about it.
I'll just leave you to cope and seethe, along with the likes of Gaska.
Orpheus
ReplyDeleteNow I see what You mean. Kind an unusual opinion.
I analyzed those Mycenaean samples for a possibility that they came from Anatolia or from Steppe via Caucasus like Eric Hamp was proposing.
I didn't found an evidence of that
First most of those G2 and J2 in Mycenaeans are Neolithic lineages. They are not new migrants. Not post Neolithic lineages like in Crete and Minoans. Quite contrary Mycenaean Greeks have _less_ post Neolithic Anatolian ancestry than in previous periods. Compared to EBA for instance.
More ever the MBA sample from Yassitepe is probably a Greek and as expected he has more European ancestry than in Yassitepe EBA. And ironically it has less Steppe ancestry than in Yassitepe EBA. Which btw is not surprising given that Myceneans overall have lower Steppe.
Btw Yassitepe MBA corresponds well with Ahhiyawa
The low level of Steppe ancestry in LBA Greece is not the result of dilution but rather a secondary reexpansion which altered the Y dna structure of MBA. So those R1b-PF7562 can't be new migrants in LBA/IA. They were there since MBA.
Maybe some J2b2a from Migdalia came from West Balkans but that is not crucial.
One of the authors of the IE language paper addressing some recent criticisms and defending the paper:
ReplyDeletehttps://nitter.net/SimonJGreenhill/status/1685093904964837376
@Gaska Wasn't that R1b from A.tepe a sacrificial offering too? Arguments like that can be made for some Caucasus/West Asian haplos found in Yamnaya/Khvalynsk too, ofc on their own they mean nothing unless there's a demonstratable change in the culture (or at the very least autosomes) as well. Languages are spoken by people and a relatively abrupt chane in language is a massive change to the people's ethnic identity, with all the things that accompany it. Equating haplos to language is purely magical thinking and already refuted by many examples (basques etruscans etc etc unless someone wants to claim beakers didn't speak IE in which case other people with beaker R1b did speak IE so the argument still fails).
That being said I too am suprised Yamnaya R1b isn't found anywhere in Greece in early Mycenaean or pre-Mycenaean times depsite their anestry being deteted there (Skourtanioti also modeled it in Tiryns IA).
After all if we go by the haplogroup argument alone, then proto-Greek never entered Greece until very late after Mycenaeans had formed, if it really comes from the steppe (Yamnaya). Which is obviously a dumb claim regardless of where it arrived in Greece from
Population wise they might be taking a view like our friend Gaska, all that j2 in Cetina and Aegean correlates with the introgression of CHG migration into Europe. But there's are several levels there, some move from north to south, which cannot be condensced into one event. Same with steppe ancestry in Southeast Europe, there are a few movements.
ReplyDelete@ Matt
I like the Bantu paper, havent yet read the others. I agree that such approaches should be utilised as an additional line of analysis. Network analysis has been another mode used in linguistics. Again it is used in DNA, i personally used it in Y-STRs and it almost perfectly predicted sequence-based phylogenies of Y-haplogroup I (I hope to publish it). Im not sure how this translates to other types of data, e.g. linguistics. I did not bother attempting to infer chronology from it, instead wiating for aDNA. The boys here can tell you stories about some early historical constructions based on STRs.
Ultimately big questions such as language spreads can only be concluded via a manual synthesis of all lines of evidence in a manner which enlists theoretical frameworks to explain how population turnovers, elite conequest, trade networks etc might impact language change.
@Gaska
ReplyDeleteYou must be very proud to be the descendant of Smyadovo (he spoke the purest Basque)
I think some Myceneans do have extra pull toward Bronze Age Anatolia, probably the effects of the Lefkandi horizon.
ReplyDeleteOverall, there were three CHG wavelets from the southern route.
@Gaska Sidenote, there is a way for language to change with culture and genes not changing: Small elite (which is very rarely demonstrated in ancient times, Magyars come to mind) or a small population subjugated by a later and powerul one within the same territory, in which case assimilation never occurs. A third modern option is multicultural empired/countries (in which case there's culture shift anyway since a monoculture dominates).
ReplyDeleteOf course the standard IE dispersal model is culture shift+autosomal shift.
@Davidski Yup and I haven't contested this position once, and find it the most likely. Not sure what you're on about but you do you
@Aram Wait why do you think I'm talking about a 100% autosomal migration? Minoans have a very strong Anatolia ChL shift observable in haplos too but they're 50% local. Likewise Mycenaeans are mainly of local Helladic/Minoan origin. I'm talking about influence, and how there's observable influence from Anatolia as well, not that Mycenaeans or Greek speakers actually came from there lol
An example is that proto-Greek could have moved from the Balkans into W Anatolia as Graeco-Phrygian and then split into Greece from Anatolia. This actually resolves both influences in Mycenaeans (steppe autosomal, anatolian autosomal + cultural) and still ultimately comes from some Yamnaya-like group (although without Yamnaya R1b).
There are several options on the table and not every one has to contradict the origin another proposes
Orpheus
ReplyDeleteHow do You differentiate a Neolithic haogroup from an Anatolian CHL haplogroup. They are very different but differentiating them is not easy for a person who is not an expert in Near Eastern haplogroups.
The J2-Z6064 which is prevalent in Mycenean Greeks is a Neolithic lineage present in Greece since the Neolithic. You can't find it in Anatolia Chl.
The same story with G2-s. If someone shows me a genuine Anatolian CHL haplotype in ancient Greeks I will be very grateful to him.
Another thought on this Planck language paper: I've talked about the way that mathematicians and statisticians could respond to this, but perhaps for one way the linguists could respond:
ReplyDeleteJust contrasting the results of Kassian 2021 that are found over and over again with these databases, even after this overhaul. These models from Planck (and including Chang's model here) repeatedly find a branching order like (Anatolian(Tocharian(Albanian-Armenian-Greek(Indo-Iranian(Balto-Slavic(Celtic-Germanic-Italic))))))...
That Kassian doesn't find this seems like it may potentially be driven by the fact that Kassian use not always directly attested languages (they use some), but often reconstructed proto-languages with estimated dates.
So effectively Kassian et al seem to be saying "We know that proto-Iranian, proto-Germanic existed at this time in this form, so therefore..." whereas the trees here are just inserting attested ancient known varieties like Pali and Avestan, Old Gothic, alongside modern-day attested varieties and allowing the tree to then infer a position, state and date for Proto-Iranian, Proto-Indic, and Proto-Indo-Iranian, and Proto-Germanic etc. That may then lead to the results that they see.
This should mean that in theory the state of the inferred Proto-Indo-Iranian found by Heggarty's model should be available for linguists to look at directly. Then they can test if it's either "No, no, no; this isn't consistent with our view of proto-Indo-Iranian" and give specific reasons (such as arguments about sound changes) or whether it is. Although a single inferred state is not necessarily available for Heggarty's tree, they should be able to provide a range of them.
The Indo-Iranian and Celtic nodes seem particularly important for this comparison - https://i.imgur.com/ANHL8NP.png
Also since there are alternative trees that they can make exist when they "force" topologies like Kassian's and Ringe's, in theory perhaps the same thing could be done.
@Orpheus
ReplyDeleteI believe that you should not disregard uniparentals when studying language changes. In this sense Iberia is a special place because we have genetic continuity in both autosomal and uniparental markers from the chalcolithic (2,500 BC) to the arrival of the Romans. We have hundreds of male genomes and except for 3 samples in the Bronze Age and one in the Iron Age all others are P312>DF27. Since all the Iberian Iron Age peoples spoke non-IE languages, the conclusion is clear, in the absence of external factors (conquests, migrations, etc) there is no reason to think that the BB culture spoke IE-
This has also been demonstrated in southern France where Aquitanians and Occitanians also spoke NO-IE languages, it is a pity that we do not have written records from the Iron Age in the British Isles, they would serve to solve the enigma of the language spoken by the BBC.
Apparently a 10-14% steppe ancestry in the Mycenaeans was enough to change their language and yet the Iberians kept their NO-IE language despite being 100% R1b-Df27, doesn't make much sense does it? Data are from Patterson, 2.021, qpAdm
Iberia_LIA_Cantabria
57.60-TUR_Barcin_N
21.20-Western_HG
21.20-RUS_Yamnaya_Samara
Iberia_BA (1.543 BC)
59.40-TUR_Barcin_N
21.40-RUS_Yamnaya_Samara
19.10-Western_HG
Iberia_Chalcolithic (2.220 BC)
64.50-TUR_Barcin_N
20.60-Western_HG
14.90-RUS_Yamnaya_Samara
On the higher order tree structure, and Balto-Slavic-Indo-Iranian, maybe a scholarly "show of hands" might be useful though, as Heggarty's supplement may or may not over-emphasize the degree to which there is not a consensus. I think it's really hard to get a sense of what most (informed) linguists think - on the internet people seem to go around saying "Oh, the Ringe tree is just the consensus among the vast majority of linguists" but I can't really get a sense if this is true, or whether linguists are generally more agnostic about the relatedness between each primary branch and higher-order subgroupings. (Apologies to any of the explainers in this comments section in the past).
ReplyDeleteFor example, Pronk, Kümmel & Kroonen thought not really for example, and as the paper notes Glottolog presents the IE tree as a rake.
Hence Olander - https://www.cambridge.org/core/books/indoeuropean-language-family/introduction/F143855FAE64FB937D00ECC9A4A0A08C - "When it comes to the higher-order grouping" (above the uncontroversial families such as Germanic, Italic, etc) "however, the situation is much less straightforward. A recurring theme throughout the book is that most of the individual subgroups are very difficult to place in the overall family tree, except Anatolian, for which the idea of an early separation has gained much traction in recent decades and is further supported in this book by Kloekhorst (Chapter 5). The position of Tocharian, often regarded as the second subgroup to separate, cannot be established with any certainty since shared innovations of the remaining subgroups are difficult to determine, as argued by Peyrot (Chapter 6).").
Pronk's chaper from Olander's text - "A closer relationship between Balto-Slavic and any of the other branches" (the context here is the preceding part of the text addressed isoglosses between BS and Germanic)" is difficult to demonstrate as well. According to Kortlandt (2016a) "the closest relatives of Balto-Slavic are Albanian and Indo-Iranian", but shared innovations are few".
@Aram
ReplyDeleteDiscussing the genetic relationship between Greek neolithic-chalcolithic, Mycenaeans and Minoans & Anatolia (and nearby regions) is like spitting against the wind. Besides the coincidences in the mitochondrial markers (if you want, we can discuss them calmly) there are some examples in the male uniparental ones (in addition to J2b-L283 which has also been found in the Caucasus).
I14798 (2.500 BC)-Oylum Höyük, Kilis_EBA, Anatolia-HapY-J1a2a/1a2-P58>Z1853>Z1865>ZS2652
NST001 (2.487 BCE)-Nea Styra, early Helladic, Euboea island-HapY-J1a-P58>ZS12519
HGC002 (2.209 BCE)-Hagios Charalambos cave, Crete-HapY-J1a-P58
TIR001 (1.326 BCE)-Tiryns, Argolis-HapY-J1a-Z2215>P58>CTS9721>Z643>Z2313>Z2292>FT137403
I1792 (2.300 BC)-Gonur Tepe, Turkmenistán-HapY-J2a-L26>L24>Y22662
HGC053 (2.100 BC)-Hagios Charambolos, Minoan culture, Crete-HapY-J2a-L26>L24>Y22662
I7079 (2.460 BC)-Devret Höyük, Anatolia-HapY-J2a1a/4b-L26>L25>F3133>BY38010>Z40171
I10950 (590 BC)-Himera, Sicily-HapY-J2a1a/4b-M410>L26>L24>L25>F3133>Z7706
I4353 (973 BC)-Hasanlu, IRN_Hasanlu_IA, Iran-HapY-J2a1a/1b1a/1a-L26>PF5197>PF5174>PF5252
GLI003 (1.367 BC)-Glika Nera, Attica-HapY-J2a1a/1-L26>PF5087>Z7314>PF5174>PF5177>PF5252
ART020 (3.300 BC)-Arslantepe, Anatolia-HapY-J2a-L26>PF5087>CTS4800>L558>Y5014>M319
HGC032 (2.100 BC)-Hagios Charalambos, Crete-HapY-J2a-L26>PF5087>CTS4800>L558>M319
I10945 (590 BC)-Himera, Sicily-HapY-J2a1a/1a2b/1b-L26>PF5087>CTS4800>M319>Y151557
Re: Heggarty's paper
ReplyDelete"1. The paper correctly dates Celtic split into Brittonic and Goidelic to 1000bce (borne out by adna from Britain).
2. Correctly dates proto romance split to around 300-500CE.
3. Correctly dates English split from Frisian to around 300-500CE, when Anglo-saxon settlements started in Britain.
4. Correctly assigns Proto-slavic split in daughters around 200-800CE, mean of ~500CE.
5. Correctly assigns split of proto-Scandinavian into daughters to around 750CE when Vikings expanded to Iceland, faroe islands etc.
6. Correctly assigns Anatolian and Tocharian as first 2 to branch off."
A #7 point was also given by @Vas, iirc it was them getting the Balkanic group right (Greek, Armenian, Albanian) like in Olander 2022
Has anyone who has read the paper found any sinificant errors besides "well this doesn't agree with the theory I'd like to be true"? Keep in mind their branching dates aren't denoting language splits (these come later than the calculated date).
@Aram Anatolia ChL is related to Minoans, not Mycenaeans
@Gaska Not disregarding it, just saying that it is one indication out of many and often has been demonstrated to be inadequate in explaining language change (or lack of change) on its own. As an indicator it's good, as actual proof on its own not that much.
@Matt Fwiw Barbieri et al 2022 gives a split between Balkanic and I-Ir at ~4000 BCE iirc
@Orpheus, well, judging from the paper that is I think simply extracted from (for some reason) Bouckaert's 2012 (https://www.science.org/doi/10.1126/science.1219669), which has been subject to criticism, and does vary in placing Greco-Albanian (not Armenian) with I-I. (So all these trees are probably not quite as similar as I may have suggested upthread).
ReplyDeleteGaska
ReplyDeleteThanks for the effort. I was asking specifically Mycenean period that is why I will look only on those cases that are from LBA excluding Crete for obvious reasons.
TIR001 (1.326 BCE)-Tiryns, Argolis-HapY-J1a-Z2215>P58>CTS9721>Z643>Z2313>Z2292>FT137403
A good case..Except that it's a J1-P58, which is a prominent Levantine - Mesopotamian haplotype.
GLI003 (1.367 BC)-Glika Nera, Attica-HapY-J2a1a/1-L26>PF5087>Z7314>PF5174>PF5177>PF5252
Another good example which was also found in Bronze Age Sicily. Ultimate origin somewhere in North Mesopotamia.
Devret Hoyuk is not F3133. It's an error on theytree. FTDNA assign it to M92 which is more plausible for North Anatolia / Amasya EBA. .
In sum You have found two Mycenean haplotypes that almost certainly are a result of post Neolithic migrations to Greece. Their ultimate origin is somewhere in North Mesopotamia via Southern coasts of Anatolia. Even more remarkable, those two haplotypes were almost certainly present in Crete in Minoan contexts.
I hope that now You understand why Your theory of Proto Greek origin is not convincing at all. Isn't it more rationale to think that Greeks are from same place than other IE people. Rather than from South Anatolia and North Mesopotamia. Most of IE branches have nothing to do with North Mesopotamia / South Anatolia.
PS J2b-L283 is not via Anatolia, but via Steppe. Sooner or later that will be proven.
The data in this latest study is quite open, so if anyone is interested to look at what cognates are present in IE languages - https://iecor.clld.org/values
ReplyDeleteSome examples contrasting II, Germanic, Baltic and Slavic - https://imgur.com/a/sO5ljlr
Terrific, their methodology was able to produce accurate results for branching at a time depth of 2500 years.
ReplyDeleteAnd the results of the paper being impossible to correspond to any prehistoric detail is not an insignificant problem. If you care so much then you're free to provide solutions. There is no excuse for a paper trying to supplement their main results with aDNA (or any other field) in such a way. If they truly believe in these results then they should abandon the Steppe as a dispersal zone.
Although reading Heggarty's supplement, he seems critical of the possibility of comparing Kassian's method of using identified proto-languages as the starting points in the analysis:
ReplyDelete"The Kassian database is much smaller still, with just 13 language taxa. Moreover, of those 13, five are not even actual documented languages, but ‘reconstructed’ proto-languages.
Reconstruction in historical linguistics is prototypically understood in terms of phonology, however; to determine original lexemes for target meanings one has to attempt to recover precise lexical semantics instead. Unlike orthodox reconstruction on the level of sound changes, methodology for ‘semantic reconstruction’ is far weaker, vaguer, more controversial and open to subjectivity (68).
For time calibration too, the use of proto-languages in (74) is a serious weakness. With actual, attested ancient languages as in IE-CoR, time-depth calibrations can be based on dated texts (as documented here throughout §4), but no such historical data are available for the concept of a proto-language. The time-depth of a proto-language is generally at best an educated guess, open to subjectivity. And for the specific research purpose here there is a clear risk of circularity, in that existing unquantified inferences as to the time-depths of the proto-languages of the main branches have typically been estimated to fit with the premise of the ‘short chronology’ of the Steppe hypothesis, assumed in advance. Finally, the Kassian database is also much smaller in covering many fewer meanings per language, only 110 compared to IE-CoR’s 170."
So it may be that this limits new linguistic challenges from this angle.
@Synome Technically, MPG SHH no longer exists. It is now MPG Institute of Geoanthropology.
ReplyDelete@ Matt quoting Olander
ReplyDelete''"When it comes to the higher-order grouping" (above the uncontroversial families such as Germanic, Italic, etc) "however, the situation is much less straightforward. A recurring theme throughout the book is that most of the individual subgroups are very difficult to place in the overall family tree, except Anatolian, for which the idea of an early separation has gained much traction in recent decades and is further supported in this book by Kloekhorst (Chapter 5). The position of Tocharian, often regarded as the second subgroup to separate, cannot be established with any certainty since shared innovations of the remaining subgroups are difficult to determine, as argued by Peyrot (Chapter 6).").''
That's the point Ive been making for a couple years now. Linguistsics is incapable of reconciling internal sub-branching. Just sticking on some unverified statistical package onto a word list isn;t going to help things. At least linguists are sober about that.
On the other hand, we have this monstrosity, or some Uralicists claiming that Uralic spread by invisible/ extinct people from the middle Volga Eneolithic.
This is a job for population genetics, which (when accurately done) can chronologically & culturally rationalise the history of respective speech-communities and established cases when 2 or more IE lects converged to final language state, as in the case of the case of proto-Germanic & possibly Iranic.
^ At least some linguists are sober about that.
ReplyDelete@Gaska
ReplyDeleteThere are hundreds of books written about Greek, Albanian, Armenian, etc... These are clearly NOT archaic PIE languages. In fact, they are probably LATE Yamnaya languages. Even ignoring something like Anatolian, we can even spot that something like Tocharian is archaic. Out of the current PIE branches, Italo-Celtic has a near-total consensus of being the 3rd most archaic. Apart from Greece, Albania itself was filled with countless of kurgans, and there is a clear ~35% steppe ancestry component surviving since the Middle Bronze Age. Saying this language is from Anatolia is bizarre nonsense.
Lazaridis himself has come out against this paper saying it makes no sense, and I suspect linguists as well will soon follow suit.
If anyone thinks academics are going to abandon obvious shit like CW -> Sintashta, or Yamnaya -> Balkan IE, for the sake of Heggarty you've got another thing coming.
ReplyDeleteModels are supposed to solve problems, not try to upend dozens or even hundreds of years of established research.
So either Heggarty is right and everyone else is wrong, or this is just a bad model.
@Gaska
ReplyDeleteBased on their contexts nearly all of the samples you listed are irrelevant and inconsequential to the question of original male lines for Proto-Greek
HGC002, HGC053, HGC032
Middle Minoans in Crete itself, they are absolutely irrelevant to the question of Proto-Greek
I10950, I10945
The site is a Greek colony outside of Greece, from well after Greek colonization and assimilation of people from Ionia & other parts of western Anatolia. Moreover the study found dozens of mercenaries from all over the world. Even for you it is highly egregious to cite anyone from here as relevant to proto-Greeks (including the "Greeks" here c.f. commonness of "Greek" people of mixed Greco-Anatolian heritage in the Iron Age such as Herodotus)
These 5 can be written off immediately. The more interesting ones:
NST001 (2.487 BCE)-Nea Styra, early Helladic, Euboea island-HapY-J1a-P58>ZS12519
The site history, architecture and pottery suggests a very strong impression of Cycladic settlement. A Cycladic origin for these people is very plausible. Nobody should count on it being relevant to proto-Greek by any means.
TIR001 (1.326 BCE)-Tiryns, Argolis-HapY-J1a-Z2215>P58>CTS9721>Z643>Z2313>Z2292>FT137403
Probably the most relevant. But its a burial in the lower citadel with no goods. The authors state regarding the lower citadel burials:
"The burials without grave goods in the Lower Citadel markedly deviate from the Palatial
and Postpalatial period burial tradition of interring the dead with grave goods in extramural
chamber tombs dug into the slopes of hills around a settlement, such as the chamber tomb
necropolis at the nearby Prophitis Elias hill. Kilian (1980) (Kilian, 1980) interpreted such
Mycenaean intramural burials of the Late Palatial period as those of members of low social
status groups within the palatial society to whom the right was denied to be buried with grave
goods in such chamber tombs. This interpretation may very well be correct, but fact is that the
practice of depositing the dead without grave goods continued after the destruction of the
palace, when the social conditions must have been considerably different to those from the
Palatial period. In the early 12th century BC, it seems that the entire northernmost part of the
Lower Citadel was even temporarily transformed into a burial ground for such burials (Maran,
2008). Therefore, we may be dealing with funerary traditions linking the Palatial and
Postpalatial period that were not only practiced because burial in chamber tombs was
prohibited, but because there were social groups that did not identify with the normative funeral
traditions and wanted to bury their dead differently."
Who knows who he was?
Also worth remembering traditions which explictly associate Tiryns' construction with western Anatolians (Lycians).
GLI003 (1.367 BC)-Glika Nera, Attica-HapY-J2a1a/1-L26>PF5087>Z7314>PF5174>PF5177>PF5252
Very late LBA after conquest of Crete. The site is full of significant Minoan (not Anatolian) goods.
Basically these samples attest to 1) the active Mycenaean recruitment of Minoans and mainland Middle Helladic counterparts (an indisputable fact based on derivation of Linear B script, palatial organization, maritime competencies)
2) settlement of other Aegean groups on the mainland of Greece; unsurprising as some of they are historically attested and some persisted into mid-1st millennium BC before their final assimilation or expulsion by Greek city-states
3) Small numbers of migrant craftsmen, mercs or or workers from the east living in major Mycenaean settlements (as did Italians, Sardinians etc.)
All in all very unconvincing
If they really wanted to create a more or less hybrid theory, then they should have developed the Bomhard hypothesis, proceeding from the fact that specific languages arise through the hybridization of ancestral languages, and not from the need to hybridize theories about the homeland of the proto-Indo-European language. The emergence of a proto-Indo-European language as a mixture of EHG and CHG languages fits perfectly into Bomhard's hypothesis. The emergence of all the main branches of the Indo-European language can also be interpreted as a mixture of Proto-Indo-European and local Aboriginal languages. Such a mix will probably lengthen the chronology because mathematical analysis will perceive the vocabulary of the Aboriginal language as an Indo-European language proper.
ReplyDeleteThis comment has been removed by the author.
ReplyDelete@Rich S.
ReplyDeleteHere's something interesting from pages 332-333 of the Supplementary Material of the Lazaridis et al paper, "The genetic history of the Southern Arc: A bridge between West Asia and Europe":
"Given that within the phylogeny of R-M269 (R-PF7562, (R-L51, R-Z2013 [sic - Z2103 is meant])) both R-PF7562 and R-Z2013 [sic, Z2103] have their earliest examples in the North Caucasus and steppe to the north, the most likely hypothesis is that the entire R-M269 clade originated there as well, with R-L51 representing a lineage that eventually became highly successful in mainland Europe, R-PF7562 a lineage that did not achieve the prominence of its relatives, and R-Z2013 [sic, Z2103] became highly successful (briefly) as part of the Yamnaya culture and its offshoots . . . "
I wish I had noticed that before. I especially enjoyed seeing that Lazaridis et al recognized R-L51 when a lot studies would mention R-M269 without mentioning the detail that R-L51 was the offshoot of R-M269 that became highly successful in western Europe. It also toook a long time for posters here and in forums to start commonly specifying R-L51. R-M269 has way too many phylogenetic equivalents and too much time passed from the first of the phylogenetic equivalents to the last. The very first person born with the R-L51 mutation, that is a descendant of R-M269, is an ancestor to all of us that are derived for R-L51. Hopefully some 6000 year old L51 specimens or 5500 year old L52*P311*P310 specimens can be found soon. Then maybe the delirious one will stop his senseless posting.
@ Vladimir
ReplyDelete''If they really wanted to create a more or less hybrid theory, then they should have developed the Bomhard hypothesis, proceeding from the fact that specific languages arise through the hybridization of ancestral languages, and not from the need to hybridize theories about the homeland of the proto-Indo-European language. The emergence of a proto-Indo-European language as a mixture of EHG and CHG languages fits perfectly into Bomhard's hypothesis. The emergence of all the main branches of the Indo-European language can also be interpreted as a mixture of Proto-Indo-European and local Aboriginal languages.''
sounds like nonsense. Neither PIE nor its daugher languages are 'mixed lanaguges' or creoles. There might have been substratunm effects, but that's a different thing altogether.
Moreover, even for an infidel like yourself, it should be abundantly obvious that PIE would not emerge from a culturally peripheral & chronologically late place like the Caucasus - Volga region. East Romania & west Ukraine is where it all began, which were socially hybrid, but that's a different thing to lingusitic mixed languages.
@Rob
ReplyDeleteThis hypothesis of yours about the Balkans is known, but not only not proven, but refuted by a recent article about the north-western Black Sea region. It is unlikely that we will learn anything new from this region in the future. Besides Bomhard, there is also Kortlandt.
Indo-European between Uralic and Caucasian
Frederik Kortlandt, 2017
Indo-European is an Indo-Uralic language that borrowed directly from North-East Caucasian, but not from North-West Caucasian or Kartvelian, and indirectly from Semitic.
https://www.academia.edu/40834017/Indo_European_between_Uralic_and_Caucasian
Do Lazaridis and Reich really think Yamnaya is responsible for Tocharian? That just seems absurd.
ReplyDelete@DragonHermit
ReplyDeleteThat is only your opinion, to confirm it you have to prove it genetically and at the moment you are very far from achieving it. Good luck, you will need it. I think it is wonderful to think that the Proto-Greek, in the absence of genetic markers proving the contrary, arrived in the Peloponnese by parachute.
@Aram you said, “If someone shows me a genuine Anatolian CHL haplotype in ancient Greeks I will be very grateful to him”
ReplyDeleteand now it has been demonstrated that the Anatolian connection not only existed in the mesolithic-neolithic, but also in the chalcolithic and Bronze Age. What I believe is that there are more genetic arguments to affirm that the Mycenaean or Proto-Mycenaean (or Proto-Greek) came to Greece from Anatolia than from the Yamnaya culture. Do you want more examples?
I4622 (3.580 BC)-Tatika, Şırnak_ChL_B, Anatolia-HapY-J2a2-PF5008>Y182822>FT184963
I7421 (1.800 BC)-Sappali Tepe, Uzbekistán-HapY-J2a2-PF5008>Y188822>MF10501
XAN024 (1.300 BC)-Malefakis, Chania, Crete-HapY-J2a2-M410>PF5008>Y182822>MF10501
I10411 (2.200 BC)-Gonur Tepe, Turkmenistán-HapY-J2a2-PF5008>L581>Z37823>PF4993
AID007 (1.438 BC)-Aidonia, Corinthia-HapY-J2a2-PF5008>L581>PF5000>PF5033>PF4993>Y24510
PS-KDC001 (1.864 BC)-Kabardino-Balkaria-J2b1-L283, You cannot know the route that this marker followed because no samples have been found either in the steppes (northern route) or in Asia Minor (southern route) so we will have to wait, but in any case it is a marker of Caucasian origin. The samples in Mygdalia (1.500 BC aprox) belong to a different branch (xZ597) from those of the northern Balkans so they may well have come from Anatolia.
@LGK
ReplyDeleteAncient Greeks definitely had more of what is called Eastern Mediterranean-Anatolian and Levantine admixture than steppe ancestry-The Mycenaeans were an extremely mediterranean population; ANF ancestry (70%) and non existent WHG. And yet the cases of PF7562 in the Palace of Nestor when the Mycenaean culture did not even exist seem to you a good argument to say that the Mycenaeans have their origin in Yamnaya? Wishful thinking again-There is no real association with higher steppe ancestry in these samples (no different autosomally than their J2a and G counterparts)-
Regarding the Minoans-I think you always forget the genetic and cultural connection of the Minoan culture and the Mycenaean culture, the Mycenaeans did not only inherit the writing system. The Anatolian-Minoan connection is stronger and there is a possibility that Minoan was an Indo-European language, in which case its origin can only be in Anatolia. Of course, since there is also a PF7562 in Crete you would immediately say that Minoan has its origin in Yamnaya, wouldn't you?
Indo-Aryan etymologies of 16 Greek words-Geoffrey Caveney (2.022)-This paper presents the analysis of 10 Minoan Linear A inscriptions as grammatical Indo-Aryan statements. The paper demonstrates the historical plausibility of the presence of Indo-Aryan speakers on Crete circa the 17thc. BCE.
Try to deal with it, even Lazaridis has recognized what many people (including Arza in this blog) have been saying for years, that is, Smyadovo is the oldest M269 we have, and you know what?, Bulgaria is not in the steppes nor the Gumelnita-Karanovo culture is the Yamnaya culture.
ReplyDeleteNow you have no choice but to accept it, that is, after having insulted, harassed and banned all those people who said that this sample was undoubtedly M269. That also happened with ATP3 and with any R1b that has appeared in mainland europe in recent years.
Who has to stop his senseless posting? those who are right (although it is annoying for some) or the fanatics who have no idea of genetics and are dedicated to defending dogmas of faith without scientific sense?.
@Gaska
ReplyDeleteThe Pontic-Caspian steppe actually extends into Bulgaria, and the Smyadovo sample obviously has steppe ancestry.
Also, the idea that the steppe was the vector for Indo-European speech and culture in the Mycenaeans came about before ancient DNA was around.
Ancient DNA just confirmed that consensus that proto-Greek came from the steppe.
@ Vlad
ReplyDelete“This hypothesis of yours about the Balkans is known, but not only not proven, but refuted by a recent article about the north-western Black Sea region. It is unlikely that we will learn anything new from this region in the future”
We don't need your Fake News lectures. Given you don’t even know what a creole is, or can’t even perform formal analyses, I’m not sure how you see yourself qualified to advise me of my own theories. Quite frankly, you dont even understand Russian literature, although you're incessantly gas-bagging about one article or another you only half-digested. Aside the point, we’ve already established that you’re a dishonest troll who engages in menial lies, like your compatriot Archi.
Yes, this NW Black sea article established early kurgans emerged in the frontier of EEF and hunter-gatherers not Siberia or the Urals. For some reason you seem to equate EEF admixture or proximity with some kind of generic Balkan model, but the exact cause of your stupidity/ dishonesty is irrelevant.
“Besides Bomhard, there is also Kortlandt. Indo-European between Uralic and Caucasian
Frederik Kortlandt, 2017 Indo-European is an Indo-Uralic language””
nice try, poor blambamir. Aside from the fact that a majority of linguists rightly dont support such a clade, Uralic and PIE come from opposite sides of Eurasia. If you actually understood the genetic data over the past few years rather than pontificating on AG, you'd have realised that Uralic spakers emerged in the Lena basin, which has nothing to do with PIE. We might point out that Kortlandt’s linguistic analysis showed that Slavic is from 3000 bc, and Goths came from Bavaria rather than Wielbark. So he’s about as in touch with reality as you & your friend Jaska.
@Rob
ReplyDeleteI don’t understand why you are against the link of Indo-European and Uralic. It was evident to me as soon as I studied my first Hungarian grammar perhaps 60 years ago. These languages come from the Siberian corridor and are all related more and more we go further in time.
“Indo-Aryan etymologies of 16 Greek words-Geoffrey Caveney (2.022)-This paper presents the analysis of 10 Minoan Linear A inscriptions as grammatical Indo-Aryan statements. The paper demonstrates the historical plausibility of the presence of Indo-Aryan speakers on Crete circa the 17thc. BCE”
Don’t forget that the Greek dynasties of Ulixes, Agamemnon, etc etc derived very likely from the charioteers of Sintashta as many dinasties in the Levant and even Egypt, whereas a few others were due for their Y to introgressed Levant man as the Kadmo dinasty.
@Gaska
ReplyDelete"Ancient Greeks definitely had more of what is called Eastern Mediterranean-Anatolian and Levantine admixture than steppe ancestry-"
Ok, but we knew that already. And it has nothing to do with the samples you cited as Greek speakers of Anatolian origin which were refuted plainly, and everything to do with ongoing recruitment of locals in the Balkans and in Greece
"And yet the cases of PF7562 in the Palace of Nestor when the Mycenaean culture did not even exist"
It has been explained to you billions of times what the difference between Mycenaean and Achaean/proto-Greek is, even though you pretend to not understand.
When the Mycenaean palaces collapsed the entire population of Greece did not magically cease to exist. So it is entirely possible and likely these are Greeks, even though they might well be newcomers of more northern origin than Pylos. Incidentally Pylos/Palace of Nestor is one of the very few Mycenaean settlements with a Linear B text that hints at pre-collapse interactions with Doric speaking Greeks, perhaps a function of its westerly orientation relative to the other Mycenaean centres.
"Of course, since there is also a PF7562 in Crete you would immediately say that Minoan has its origin in Yamnaya, wouldn't you?"
Don't project your fantasies for Iberia onto others, I would say it is an individual with a paternal origin in the steppe, nothing more unless further demonstrable
@ Gio
ReplyDeleteThe resemblances are not enough to be definitive, however I dont discount that some are indeed genuine and apparently at the level of 'basic lexis'.
This requires a level of explanation beyond a simplistic Indo-Uralic node, which is simply a not a viable model.
As above, PU emerged from a ENA-rich, Hg N1c dominated population whilst PIE emerged in a WHG-shifted EHG groups in contact with EEF groups over a broad front from Bulgaria to the Don (the 'border' ran well toward the east along steppe/fprest steppe, as we know).
The idea of some far-reaching PIE-PU SIberian corridor does not qualify as a model.
BTW the Nostratic & Indo-Uralic ideas are very prominent in Russia and propagandized by trolls like Vladimir because of its role in 'propaganda of Empire'
Rob, your verbal diarrhea is not going to help the case in any way. Version in the Balkans or the north-western Black Sea region
ReplyDeleteclosed by the article under discussion. The desired population was obviously to the east. And taking into account the fact that the culture of Chernavoda is usually associated with the culture of Nizhnyaya Mikhaylovka, then the entire territory west of the Dnieper most likely also disappears. There remains Zhivotilovka-Volchansk, as Heyd suggests, or even further east.
Is it always possible to support that EEF did come from Northern Anatolia and not the other way around?
ReplyDeleteAnd what about my hypothesis that they were the Etruscans from the Aegean Sea and Southern Europe?
https://www.nature.com/articles/s41598-023-39250-y?fbclid=IwAR0gpJ3NmtCuIsDK7Z4PIhPV9_u7ybOpXDwEzOb_DbWNah0Zqg0Rfwl1S-U
@ Gio
ReplyDeleteAs per above, completely different geography - PU emerged in an ENA-rich groups dominated by Y-hg N1c, PIE emerged in WHG-shifted EHG clans close to EEF (Varna, Cucuteni, etc)
Nostratic -type hypotheses are uncouth & refletive of poor resolution/ lazy analyses, but are prominent in Russia w.r.t the 'propaganda of Empire'
@ Rob
ReplyDeletePerhaps you know, from my old letters that we also exchanged privately, that I dedicated my youth, among many other things, to the study of the works of Alfredo Trombetti about the monogenesis of languages, unfortunately written in Italian even though he knew more than 30 languages, and some of these modern linguists took in serious account that.
I am not Vladimir and am not a Russian, even though I never denied my friendship with them and also about this last tragic war, but Russia had and has some of the greatest linguists of the last centuries, beginning from Illich-Svitych, and about Albanian the great works of Orel/Oryol. Many are Jews, even though Dolgopol’skij never convinced me about the link of Indo-European and Semitic.
@ blabamir trollishkin
ReplyDelete''The desired population was obviously to the east. And taking into account the fact that the culture of Chernavoda is usually associated with the culture of Nizhnyaya Mikhaylovka, ''
More Fake News
Ill dumb it down for you. Cernavoda & Usatavo are dominated by I2a-L699, found in preceding
Neolithic in the Carpatho-Danube-Dnieper zone. Mikhailovka is jsut Cernavoda by different name.
Balkan EBA, also high levels of I2a-L699, as well as in western Anatolia
+ They have high levels of EEF
Further East your ass.
New preprint claims that archaic admixture models have got it all wrong
ReplyDeletehttps://www.biorxiv.org/content/10.1101/2023.04.05.535686v1
Wild!
I-L701 may not be firmly steppe as previously suggested, given that one of its ChL carriers from this paper can be modeled as EEF+WHG (or interestingly, EEF+iron gates HG).
ReplyDeleteI don't think there is anything to suggest downstream I-L699 doesn't have a slightly more eastern orientation though. I'd wait for higher coverage Y reads as I assume some of the many samples labeled I-L701 will be given more precise reads by FTDNA among others soon.
Complicating matters, I-L699 isn't the one found in BA Western Anatolia, parallel lineage I-P78 is. Hopefully its origin will gain more clarity, although nothing will change the fact that it represents clear Balkan/european ancestry in Western Anatolia.
@Rob
ReplyDeleteL699 is an interesting marker and, apparently, marks the Mariupol culture. Perhaps they had a cultural influence and served as a trigger for the transformation of the steppe from fishermen-hunters to cattle breeders. But they did not have any genetic influence on the steppe, autosomally they may have contributed 5% WHG and 5% EEF, and then Harvard disputes this. In all other respects, the steppe is EHG/CHG, which is even east of the Lower Mikhailovka, not to mention Chernovoda. In all L699 sites and cultures, the minor group is nowhere in the majority.
@Gaska
ReplyDeleteYou should consider that some of the Greek samples you mentioned are of local origin, that is possibly pre-Neolithic even. Personally I favored something like the Anatolian hypothesis but not exactly, because if the split between Anatolian and "nuclear-IE" is early, this could have happened during the Neolithic at least, either in Anatolia or in SE Europe (We could consider thus for all intents and purposes Anatolian to be "native" in Anatolia and "nuclear IE" to be "native" in Europe).
Then, Greek can theoretically be from the southern end of the ANF-Steppe Eneolithic continuum. Personally I see more similarities culturally between Greeks and CWC related groups by the way, not any close connection though.
Now in Greece there were definetely languages that did not descend from 'nuclear IE'. Still, I never believed Greek had anything to do with Yamnaya and in general I agree with you on many things you say. I always considered irrational the idea that all IE languages descended from Yamnaya.
Many so called pre-Greek words have 'Eurasiatic' parallels by they way. One possible native Cretan word for the meaning 'Sun', seems to be something like *tal-. It seems quite obvious to me that we could compare it to IE words for 'Sun' and Uralic words for 'fire'.
Greetings from Greece.
@Gaska
ReplyDeleteIt's not my opinion. It's fact. Every other linguist states that the two most archaic branches are
(1) Anatolian
(2) Tocharian
and Tocharian is already a Yamnaya offshoot.
Apart from Heggarty, no one will buy the bullshit that Armenian, Greek, and Albanian are from Anatolia 6000 BC or whatever nonsense. These are languages very closely related to CW languages to the point where Italo-Celtic, a CW-derived language, is listed as more archaic. Albanian/Armenian are also accompanied by noticeable steppe ancestry.
And don't even get me started on Indo-Aryan. The existence of a site like Arkaim proves everything you need to know.
@ Matt
ReplyDeleteJust trying to think what’s salvageable.
Aside from the muddled migration paths, Perhaps they’re detecting pre-existing dialectical diversity …
Although I still think linguistics is not something which can be readily analysed via a mathematical algorithm
@Dospaises
ReplyDelete"@Rich S.
Here's something interesting from pages 332-333 of the Supplementary Material of the Lazaridis et al paper, "The genetic history of the Southern Arc: A bridge between West Asia and Europe":
"Given that within the phylogeny of R-M269 (R-PF7562, (R-L51, R-Z2013 [sic - Z2103 is meant])) both R-PF7562 and R-Z2013 [sic, Z2103] have their earliest examples in the North Caucasus and steppe to the north, the most likely hypothesis is that the entire R-M269 clade originated there as well, with R-L51 representing a lineage that eventually became highly successful in mainland Europe, R-PF7562 a lineage that did not achieve the prominence of its relatives, and R-Z2013 [sic, Z2103] became highly successful (briefly) as part of the Yamnaya culture and its offshoots . . . "
I wish I had noticed that before. I especially enjoyed seeing that Lazaridis et al recognized R-L51 when a lot studies would mention R-M269 without mentioning the detail that R-L51 was the offshoot of R-M269 that became highly successful in western Europe. It also toook a long time for posters here and in forums to start commonly specifying R-L51. R-M269 has way too many phylogenetic equivalents and too much time passed from the first of the phylogenetic equivalents to the last. The very first person born with the R-L51 mutation, that is a descendant of R-M269, is an ancestor to all of us that are derived for R-L51. Hopefully some 6000 year old L51 specimens or 5500 year old L52*P311*P310 specimens can be found soon. Then maybe the delirious one will stop his senseless posting."
When "the delirious one" began posting in DNA discussion venues years ago, it was as a defender of the old idea that R1b weathered the LGM in the Franco-Cantabrian LGM refuge and emerged from there to repopulate much of Europe. He delighted in the luster of glory this seemed to impart to the Basques, back when many geneticists were laboring under the weight of 19th century ideas about the Paleolithic antiquity of the Euskaran language. One might call that error the "Basquetardization" of European prehistory, the mother of a now defunct bad genetic idea.
As ancient DNA systematically dismantled the old "we are all Basques" paradigm, step by inexorable step, the delirious one became the seriously bitter and angry one. He developed an implacable hatred for the advocates of the steppe hypothesis of IE origins, especially for those who had argued that R1b-M269 (in the form of the temporally appropriate subclades) accompanied the advent of the steppe pastoralist Indo-Europeans. He hated us because we had been right where he had been so demonstrably wrong. We had offended his national pride and the honor of his ancestral homeland. It was as if we had insulted his mother.
The Franco-Cantabrian LGM refuge was a gone goose as far as R1b was concerned. Even the delirious one, despite his delirium, saw it. Instead of admitting the truth in a good natured and self effacing manner, he embarked on a campaign of sour grapes, determined to prove that R1b-M269 came from just about anywhere but the steppe. He would show us, the despised "Kurganists", just how wrong we were.
Needless to say, that isn't working out well at all, because the truth is arrayed against it. The grapes are sweet, and "In Vino Veritas".
I am still hoping against hope that someday he'll come around, admit he was wrong, and begin to enjoy his own ancestral share in the saga of R1b-M269. Most of us would be happy to welcome him into the club of those with good sense. However, his frequent departures from the path of honesty militate against that hope.
Ah, well. Almost no one buys his bilge anyway.
@ Vladimir
ReplyDeleteYour suggestions are wildly innacurate. Its concerning that you have not read or understood a single paper in the last few years.
(1) '' In all other respects, the steppe is EHG/CHG,''
This is an autosomic profile, it doest not mandate what uniparental lineages are within it
And the western steppe - which is what is actually important was not monolithic, but even had a individuals with EEF ancestry, some with WHG, etc
(2) ''But they did not have any genetic influence on the steppe, autosomally they may have contributed 5% WHG and 5% EEF, and then Harvard disputes this''
Not true, I2a-I-CTS10057 shows a diverse autosomic profile. Secondly, Harvard has not utilised the new data, so Im not sure what you're talking about. Harvard did not say there is no UKr-N related ancestry, it was Anthony, citing pers comm of Mathieson. This is essentially nothing - Anthony cant analyse data, just tells stories of his own invention.
However, a median of the Cernavoda will provide ~ 20-30% input into eastern Yamnaya.
(3) ''In all L699 sites and cultures, the minor group is nowhere in the majority.''
more fake news.
I2a-I-CTS10057 is the dominant lineage in Cernavoda, Usatavo, Bulgarian Yamnaya, lower Dnieper Neo, Gordinesti.
Then, we even have EEF-related J2a, H2, G2a in Cernavoda too.
CWC & eastern Yamnaya are different groups with different origins, but you seem to not understand that.
(4) ''. In all other respects, the steppe is EHG/CHG, which is even east of the Lower Mikhailovka, not to mention Chernovoda. ''
nonsensical statement. Try google translate
@ EthanR
''I-L701 may not be firmly steppe as previously suggested, given that one of its ChL carriers from this paper can be modeled as EEF+WHG (or interestingly, EEF+iron gates HG).
''
If a lineage in is linked to steppe zone, then its steppe related. There are steppe-reelated G2a lineages which originally come from Syria.
It is misguided to assume a preconceived notion that only R1b-Z2103 is steppe related, and everything else was a fluke, or underclass, as one archaeologist tried to claim.
The key significance of this lineages is precisely because it falls under various autosomc profiles.
''Complicating matters, I-L699 isn't the one found in BA Western Anatolia, parallel lineage I-P78 is. Hopefully its origin will gain more clarity, although nothing will change the fact that it represents clear Balkan/european ancestry in Western Anatolia.'
It doesnt complicate it. It's broadly analogous to R1b-L151 and R1b-Z2103, just slightly older
@Rob
ReplyDeleteI-L701 is by definition Steppe related, but I don't see how it can be claimed that it is exclusively Steppe with the information we have, as we now have two early samples (I2377 in Hungary and PIE060) that can be modeled without an eastern contribution. Ukraine receives western geneflow between the mesolithic and neolithic, which serves as a possible conduit. Upstream I-CTS10057 is pan-european.
If it isn't exclusively steppe, then this means how we approach its downfield markers requires additional attention, although I doubt a mutation like I-P78 with a ~4100BC TMRCA (incidentally matching the estimated branching of Proto-Anatolian) lacks a connection with the Steppe and therefore IE.
Of course at any time a lineage can become partially or wholly acculturated into a different linguistic community. So the most important thing is that it's european.
@Rob
ReplyDeleteAll these cultures of Cernavoda, Usatavo, lower Dnieper Neo, Gordinesti were the substrate during the formation of populations of the Bronze Age, the main population that formed the cultures of the Bronze Age was east of Chernavoda and the Neolithic Dnieper
This comment has been removed by the author.
ReplyDeleteThis comment has been removed by the author.
ReplyDelete@ Vlad
ReplyDelete''All these cultures of Cernavoda, Usatavo, lower Dnieper Neo, Gordinesti were the substrate during the formation of populations of the Bronze Age, the main population that formed the cultures of the Bronze Age was east of Chernavoda and the Neolithic Dnieper''
The problem is you dont know what you;re reffering to, just making bland statements
Do you mean CWC, then everybody figured that out a week ago, not the least that there is no R1a in Cernavoda. But that's got nothing to do with southeast Europe or Anatolia, which had several waves which at times fused at times remained distinct
So there are groups like Boyanaovo, Cinamak or the younger Majaki which are new eastern and R1b-m269 derived, but others , even post-3000 BC are I2a and primarily interred in kurgans
@ Ethan
ReplyDeleteyour assumptions are unfounded. If I2a-701 introduced a vital package of innovations which inaugurated the Mariupol horizon and remained as the chiefs of the heirarchic Usatavo chiefdom buried with Cooper daggers, why would they adopt the language of Yamnaya commoners ?
I’m
@DragonHermit
ReplyDeleteSorry but it is only your opinion, the Kurgan theory is one of many that try to explain where the homeland of Indo-European was - Other linguists, archaeologists or geneticists think differently. In fact a Paleolithic origin has been proposed (Cavalli Sforza, Thomas, Hausler, etc), Anatolia (Renfrew, Gray, Atkinson, Bouckaert), Armenia (Gamkrelidze), Fertile crescent-nostratic homeland (Bomhard), Serbia (Todorovic), Baltic (W.Schmidt), Central Asia (A. Kozintsev), India (Igor Tonoyan-Belyayev), Pontic steppes (Gimbutas, Anthony, Chang), northern Caucasus (Patterson), North Pontic region-Sredny Stog (Davidski), Southern Arc (Harvard, Reich, Lazaridis), northern Iran (Krause, trappe, Max Planck), Anatolia-8000 BC (Heggarty) etc etc. I have no idea what the homeland of IE is, I simply try to relate genetic markers and compare them with all these theories. Of course some make more sense than others, but this is how science advances, proposing solutions and criticizing what you consider inadequate. I find the Kurgan theory acceptable in certain aspects (especially the relationship of R1a to CWC and Balto-Slavic), and totally unacceptable in others (R1b-L51>L151). But this is just my opinion, I don't claim to be right, I leave that to people as intelligent as you.
@LGK
ReplyDeletePlease help me understand your reasoning; PF7562 in the Minoan culture is simply a man with paternal origin in the steppe ergo nothing else until proven otherwise. However PF7562 in Mycenaean culture brought Indo-European to the Peloponnese? So this marker is part of the so-called ongoing recruitment of locals in the Balkans and in Greece?
Don't worry, I can perfectly differentiate a Mycenaean from an Achaean-Proto-Greek the one you don't seem to understand is you, don't you think that those PF7562 samples are too late to be linked to the arrival of the IE in the Peloponnese?
Don't you think you need that marker between 2000-1.500 AC in that region to be able to affirm that? Scenarios that linked the appearance of Greek with the end of the Mycenaean palaces (1,200 BCE) have been largely abandoned since the decipherment of Linear B established the use of the Greek language, from at least the mid 2nd millennium BCE.
And yet, you have many markers of Anatolian origin in the Greek neolithic, chalcolithic & Bronze Age in both Mycenaean and Minoan cultures (even Himera because those samples I posted have a classical Greek profile and are descendants of the Mycenaeans, or do you think they appeared in Sicily directly from Anatolia)?
Here the only one projecting fantasies, doing gymnastic mental exercises to deny the evidence and telling fairy tales is you my friend because today no one can deny the genetic connection of Anatolia and Greece from the Neolithic to the Iron Age, and that this relationship is much stronger than the genetic relationship with the steppes.
Where Proto-Greek was spoken?, I leave that to experts like you and Lazaridis.