A new
preprint at
bioRxiv by Kerdoncuff et al. makes the following, somewhat surprising, claim:
One of the individuals, referred to Sarazm_EN_1 (I4290) described above that was discovered with shell bangles showing affiliation with South Asia, has significant amount AHG-related ancestry, while a model without AHG-related ancestry provides the best fit for Sarazm_EN_2 (I4210) (Table S4.5).
First of all, the authors are actually referring to sample ID I4910 not I4210.
The aforementioned table, based on qpAdm output, shows that I4290 has 15.9% AHG-related ancestry and basically no Anatolian farmer-related ancestry. It also shows that I4910 has no AHG-related ancestry but 17.9% Anatolian farmer-related ancestry.
AHG stands for Andaman hunter-gatherer. The authors are using it as a proxy for South Asian hunter-gatherer ancestry.
However, I've looked at I4290 and I4910 in great detail over the years using ADMIXTURE, Principal Component Analysis (PCA), and qpAdm. And I'm quite certain that they do not show any obvious, above noise level South Asian ancestry. Indeed, I'd say that if they do have some minor South Asian ancestry, then I4910 probably has more of it than I4290.
Kerdoncuff et al. used the following "right pops" or outgroups: Ethiopia_4500BP.SG, WEHG, EEHG, ESHG, Dai.DG, Russia_Ust_Ishim_HG.DG, Iran_Mesolithic_BeltCave and Israel_Natufian.
This means they mixed data that were generated in very different ways (DG, SG and capture) and included some poor quality samples. For instance, the highest coverage version of Iran_Mesolithic_BeltCave offers just ~50K SNPs.
Mixing different types of data and relying on low coverage samples, even in part, often has negative consequences when using qpAdm. So I suspect that the above mentioned mixture results for I4290 are skewed by a poor choice of outgroups.
When I run qpAdm I try to stick to one type of data and avoid low quality singletons in the outgroups. This is the best qpAdm model that I can find for Sarazm_EN:
right pops:
Cameroon_SMA
Morocco_Iberomaurusian
Israel_Natufian
Levant_N
Iran_GanjDareh_N
Turkey_N
Russia_Karelia_HG
Russia_WestSiberia_HG
Mongolia_North_N
Brazil_LapaDoSanto_9600BP
Sarazm_EN
Kazakhstan_Botai_Eneolithic 0.113±0.017
Turkmenistan_C_Geoksyur_subset 0.887±0.017
P-value 0.06392
Sarazm_EN_1 (I4290)
Kazakhstan_Botai_Eneolithic 0.129±0.021
Turkmenistan_C_Geoksyur_subset 0.871±0.021
P-value 0.11019
Sarazm_EN_2 (I4910)
Kazakhstan_Botai_Eneolithic 0.104±0.021
Turkmenistan_C_Geoksyur_subset 0.896±0.021
P-value 0.07427
Also...
Sarazm_EN
Andaman_hunter-gatherer -0.018±0.020
Kazakhstan_Botai_Eneolithic 0.123±0.019
Turkmenistan_C_Geoksyur_subset 0.895±0.020
P-value 0.0298403
(Infeasible model)
Please note that Turkmenistan_C_Geoksyur_subset is made up of just three relatively high quality individuals: I8504, I12483 and I12487. That's because it's not possible to model the ancestry of Sarazm_EN using the full Geoksyur set, probably due to subtle genetic substructures within the latter.
Below is a PCA plot that, more or less, reflects my qpAdm model. I4290 and I4910 are sitting right next to each other in a cluster of ancient Central and Western Asians, and it's actually I4910 that is shifted slightly towards the South Asian pole of the PCA. Indeed, I can confidently say that there's no way to design a PCA in which I4290 is shifted significantly towards South Asia relative to I4910.
Citation...
Kerdoncuff et al.,
50,000 years of Evolutionary History of India: Insights from ∼2,700 Whole Genome Sequences,
bioRxiv, posted February 20, 2024, doi: https://doi.org/10.1101/2024.02.15.580575
See also...
The Nalchik surprise
A comedy of errors
561 comments:
«Oldest ‹Older 401 – 561 of 561@ Jaakko Häkkinen
This is what Dahl writes, and sorry for long quote.
''The language found in the later runic inscriptions (Late Runic) differs very markedly
from that in those written in the Older Futhark. It is highly questionable if a person
from the 11th century would have understood the speech of Hlewagastir, the author of
the famous Gallehus horn from around 400 B.C.E.''
''Here is in brief what I think is the most probable scenario for the origin of the
Scandinavian languages: Germanic-speaking groups arrived to the very western-most
corner of the Baltic (present-day Germany and Denmark) somewhat before the
beginning of our era. A little later they expanded eastwards as far as Uppland on the
north side and the Vistula estuary on the south side of the Baltic. During the ensuing
half millennium, the languages of the different Germanic groups became differentiated,
exactly how much we do not know. At the same time, such pre-Germanic groups as
still remained would slowly be Germanized, a process that we shall probably never find
out the details of.
As Denmark emerged as the major political power in Scandinavia, the language of its leading classes spread to critical parts of the other Scandinavian
countries, most probably in several waves. The result was a relatively homogeneous
language situation at the beginning of the historical period, which has been mistakenly
extrapolated backwards in the traditional account of the history of the Scandinavian
languages.''
He is speaking of language, or perhaps, dialect shift- 3 waves. Not exactly surprising when there were assompanying population shifts
In the late Bronze Age we already have the urnfielders and, as far as we know, the Unstrut culture (Thuringia...), 22 of the 25 men were identified as I2-M436>L38 and 3 individuals belong to haplogroups that betray an northern & eastern genetic inheritance: R1b-U106 (individual M9) and R1a-Y2395-Z282 (individuals M10 and M11)
We also have a sample of R1a-Z280 in Halsberstadt, while in the Knoviz culture, in addition to L2, we have R1b-Z2118, R1a-Z280, H2-P96, I2a1b-L1229, I2a1b-Z2069 & R1b-U106
Then in the Hallstat culture in central Europe we have the boom of G2a-L497 with samples at Mitterkirchen, Magdalenenberg, Hochdorf, Louny, Stradonice, Komárno, Chotín, together with R1b-Z2118, R1b-U106, R1b-Z2103, R1a-M458 (Singen), I2a1a-L161 (Dobromerice), I2a1b-S18331 (Chomutov Polaky)-
Then the Central European proto-Celts (urnfielders) and Celts (hallstattians) were a mixture of male markers some of which have never been documented during the chalcolithic, the Bronze or Iron Age neither in the British Islands nor in Iberia nor in Italy with which we can exclude Central Europe as the origin of the IE languages in those regions.
It is time to stop overestimating the importance of our P312 lineage and recognize the role played by I2a and G2a2b-L497 in relation to the IE languages (the same occurs in the Balkans with other markers involved, including J2b-L283 and several subclades of J2a)
We only have north-east France, Switzerland and perhaps the Straubing culture in Germany, but this requires a more elaborate comment and discussion.
@Gaska.
Nobodu may be sure of anything without old texts and witnesses but I doubt Celtic languages came from Central Europe, and as late as IA. It was older I think, or at least its proto-forms were spoke more westernwards. Hallstatt is not more considered as the ONE ethnic group thing, not more than was the too heterozygotous group of Urnfields of diverse depth of penetration and diverse origins.
THe larger extension of Celtic languages (and toponymy) includes almost too much Western Europe regions where Y-R1b-P312 was dominant and IS still well present. How a language spoken by a little "barbaric" not centralized group of tribes could have imposed in so little time its language allover the West and Northwest? Maybe not impossible but surely not evident!
The old western IE complex (before Celtic, Italic, Venetic, Lusitanian, Ligurian, Northwest Block, IE Rhaetian part) is the place where Celtic individualized itself, rather in Central Eastern France so in West, when Italic individualized itself closer to the East (Central) and to other cultures centered around East-Austria Carpathian Bassin.
@Gaska
The abundance of I2a-M26 in the Halstatt samples was interesting. This makes a nice connection to the only two Celt-Iberian samples in existence which were also I2a-M26.
What are your thoughts on the British Isles? From what I remember of Patterson's paper non-Local P312 groups did play a the primary role in the Celtic invasion there.
This Halstatt resurgance of Alpine, EEF-rich groups with 'Sardinian'-related (obviously now a misnomer) I2a-M26 and certain classes of G2a, almost recalls the old racial stratifications "Alpine' race, although they were often too broad & incorrect.
The classic textbook views of Halstatt being early Celts are challenged due to the age & divergence of Celtic-like languages & para-languages, but Halstatt might have stimulated their glottogenesis/ differentiation & cultural development.
@Moesan
The likely constraint on the traditional views would be due to the relationship between Celtic and Italic. Regardless of whether Italo-Celtic is a proper node of PIE or if it is a sprachbund doesn't matter so much in this view. You still put Proto-Celtic's diversification anywhere from Central Europe to the coast of France because the Tyrsenian(Etruscan/Rhatic) and Aquitanian/Basque language families don't narrow things down either. The parental chromosomal lines are already too blurred in these groups to help.
Without some sort of new evidence the opinions we get going forward are just as likely due to the fact the descendants of the Celts can all afford to buy a new book on the topic rather than any real conclusion.
@Matt
You made the choice to resort to a logical fallacy and recoil when called out on it. I'm sure you disagree with thousands of comments on this blog every year but you choose to respond so I felt it was only fair to point out.
Would you make your actual views clear? Specifically I want you to clarify what in the Southern Arc paper and the follow ups changed your mind? I've read this blog almost from the beginning and while I don't always read the comments I've read hundreds of yours alone. I could clearly see how you mostly agreed @Davidski's analysis, even helping out processing data and cogently elaborating on various points.
So what made you change your mind? You agreed with his analysis of the admixture that the Southern Arc theory bases everything on and clearly understood it then so what changed? The linguistic evidence is the same, the archeological evidence too is basically unchanged. I can't recall you ever explaining your reasoning.
Jaerl:
“He is speaking of language, or perhaps, dialect shift- 3 waves. Not exactly surprising when there were assompanying population shifts”
There is nothing here contradicting with what I wrote about the Runic inscriptions.
About his other views: he clearly believed at that time that Scandinavia was Germanicized very late, which is not the best argumented view nowadays. However, his view on the later Scandinavian development still seems valid:
Östen Dahl:
”I shall argue that the apparent homogeneity in the language in the central parts of Sweden, Denmark and Norway around the previous millennium shift was quite a recent phenomenon and due to the spread of a koiné from a political and economical centre in the south.”
That is: after the Proto-Scandinavian and Transitional Scandinavian, Old Scandinavian stage was still rather homogenous, which allows such interpretations as the spread of a levelling dialect to replace other regional dialects.
Moreover, the first entities to branch off within Scandinavian were Old Gutnish (in Gotland) and Elfdalian (https://en.wikipedia.org/wiki/Elfdalian). That Swedish and Norwegian generally look like later branchings, indeed supports the view that there occurred an expansion of such a levelling dialect. One more piece of evidence is the label “dǫnsk tunga” (Danish tongue), shared by the Viking Age speakers of Old Scandinavian even in nuclear Norway and nuclear Sweden.
@Romulus
Regarding the celtiberians, we have dozens of urnfielders sites in Catalonia and the Ebro valley. From the material culture, Spanish archaeologists think that they entered around 1,100 BCE and had their origin in the RSFO. These are the proto-Celtiberians who spoke an archaic Celtic (first branch that separated from the common Celtic)-In many of these sites both inhumation and cremation were practiced, which shows that the urnfielders mixed with the locals. Spanish and French geneticists have to work and analyze both RSFO and Iberian urnfielders to draw some interesting conclusions.
In this sense, the Iberian Iron Age samples we have do not help much because 95% are R1b-Df27, and we only have two I2a-M26 as you have pointed out. The oldest is a Basque from Navarre (a territory where Ibero-Basque was spoken in the Iron Age) and shares his marker with an Iberian from Occitania and the other one is a Celtiberian (Berones from Alava, current Basque Country)-These people also occupied La Rioja (I recommend trying the wine if you haven't done so yet) and spoke an Indo-European language. As these lineages are documented in Iberia at least since the chalcolithic, we have to assume that the Celtiberian from Alava is simply a Celticized Iberian not patriline descendant of urnfielders.However, there are new mitochondrial lineages from the north of the Pyrenees and it is also evident that the Celtiberians are more autosomally similar to the Gauls than the Iberians.
-As you said, downstream of M26 there are a lot of I2a-L160 Gauls but there are also, I2a1b-Z170, I2a1b-Y3670, I2a1b-S19763-Both the Iberian and Gaul I2a seem to descend from the local neolithic.
*CRU012 (697 BCE)-Alto de la Cruz, Cortes, Navarra, Spain-HapY-I2a1a/1a-M26-L158
*CLR35 (350 BCE)-Le Cailar, Gard, Occitaine-HapY-I2a1a/1a-L158
*ELT006 (3.805 BCE)-Els Trocs, Iberia- HapY-I2a1a/1a1-P37>CTS595>L158>Y3992>Z2049
*I3759 (278 BCE)-La Hoya, Celtiberians, Iron Age, Iberia-HapY-I2a1a/1a1-L158>Y3992>Z2049
@All
I'm working on a blog post about the new Germanic preprint.
It'll take me a few days, at least, because I think there are some serious issues with this manuscript that the authors should correct.
@Romulus
The history of the British islands is very similar to the Iberian one. The Atlantic connection remained uninterrupted from the neolithic until the Iron Age. The trade related to the Atlantic Bronze culture has been perfectly demonstrated thanks to genetics although there were no large migrations but rather commercial trips, marriages (exogamy), etc. This was also demonstrated by Patterson with his high_EEF samples from the English Bronze Age that could only have Iberian origin-Besides R1b-L21, I have only found HapY-R1b-DF27 (4) HapY-R1b-L2 (1), HapY-I2a-Y10720 (1) & HapY-I2a1b-L1195 (1) in the entire British Bronze Age-The latter has a clear origin in the islands but the others have their origin in mainland Europe-
*I6470 (1.651 AC)-Virgazal, Cogotas culture, Spain-HapY-R1b-DF27-FGC33092-Y30815>Y30814
*I7629 (1.067 AC)-Melton Quarry, BA_England-HapY-R1b-DF27>Y30815>Y30821>Y30814
*SUC009 (1.749 BCE)-Su Crucifissu Mannu, EBA_Sardinia-HapY-I2a1b/2-M436>S2555>Y10720
*I19859 (1.453 BCE)-Rowbarrow, Wiltshire, MBA_England-HapY-I2a1b/2-M436>S2555>Y10720
Regarding the linguistic issue, well I suppose everything will depend on the faith you have in the Kurgan theory, if you think that the BBC spoke IE then the British Isles spoke that language since chalcolitic (Not Celtic because it did not even exist as such)-In my opinion, in the Bronze Age there are no reasons to think about a language change because migrations are not significant. The Iron Age was different, because then there was an important entry of continental uniparental markers and of course a small change in the autosomal composition -
Enough to think about the entry of the Celtic from France?, in my opinion yes.
@Moesan
In the absence of written records we will always move in quicksand and rather than certainties we will only be able to develop more or less convincing theories. The cultures of the Iron Age in France and Central Europe are uniparentally very heterogeneous and will not serve to solve the linguistic doubts raised. Italy could help solve the problem because north of the Alps, southern Germany and Switzerland in the Bronze Age were overwhelmingly U152>L2.
A migration to the south could have brought this marker to the Italian peninsula. In fact we have L2 in Broion (1492 BCE), but then the problem of the Etruscans who were mostly U152 and yet non-Indo-European appears. It seems a dead end because this marker is shared by IE and NON Indoeuropean cultures ergo in the absence of convincing evidence, I believe that neither the BBC nor the Bronze Age cultures of Iberia, Italy and France spoke Indo-European languages until the appearance of the Urnfielders.
Something similar happens in Spain, all the Iberians analyzed (Layetanos, Ilerkavones, Indiketes, Vascones) are overwhelmingly DF27 and are identical to the Argarians, so genetic continuity and common sense force us to think that this marker never spoke IE languages. However, we already have at least 8 DF27-Gauls so this marker at least in northern France (not in Aquitaine or southern Occitania) spoke Celtic.
What is interesting is the large share of ancestors from the Baltic BA source and R-Y35 among the Burgundians from Bornholm from the Roman period.
@ Jaakko
In light of this data, if it’s correct, when would you date the proposed loans and interactions between Norse, Saami & Finnic ?
@tea, if you wish to read my recent thoughts about Lazaridis's models in the Southern Arc paper, you can review the discussion between myself and gamerz_j in this comment thread from February 2024 - https://www.blogger.com/comment.g?blogID=4123559132014627431&postID=5726989184238938979&isPopup=true
How you have arrived as some strawman position that I am committing an "Appeal to Authority Fallacy" and saying that every specific argument by Reich lab is correct, from my generally offering a defense of the state of the field of archaeogenetics, and some specific assumptions about what I believe about the Southern Arc papers that I did not state in this comment thread, I do not know.
history and chronology of distribution of Algic Languages
https://youtu.be/yBYd_-E2HqY?si=9Hca1ro2aWi4cWrU
Rob:
“In light of this data, if it’s correct, when would you date the proposed loans and interactions between Norse, Saami & Finnic ?”
Norse is only the late end of these contacts. The early end are the Paleo-Germanic loanwords into Saami and Finnic sometimes around 1000–500 BCE:
(1) As nowadays all the best-argumented views assume the Germanic presence in Scandinavia already since the Scandinavian Bronze Culture (beginning ca. 1600 BCE) – no matter whether the language arrived only slightly earlier or already with the CWC and whether it came from the south or from the east – and
(2) because soon this culture spread to coastal Finland and Estonia, and
(3) because around 1000 BCE Early Proto-Saami arrived in Southern Finland and Early Proto-Finnic in Estonia, these regions appear to be credible locations for these early contacts.
There are still possible earlier Pre-Proto-Germanic loanwords in West Uralic, but these loanwords require modern critical assessment. Some of them have received other explanations.
From the Late Proto-Germanic stage onward (> Northwest Germanic > Proto-Scandinavian > Old Scandinavian) there are also Germanic placenames in Southern Finland, so it appears that this region was continuously a target of Germanic expansion waves, the last one being Swedish expansion since the 13th century CE.
Off topic but this seems unusual: https://www-inrap-fr.translate.goog/architectures-et-depots-funeraires-au-neolithique-pontcharaud-clermont-ferrand-17868?_x_tr_sl=fr&_x_tr_tl=en&_x_tr_hl=en-US&_x_tr_pto=wapp#
Describes a site being uncovered in France with individual burials in a flexed position around 4500-4000 BCE.
Maybe these samples need some adna testing... 50 burials (albeit some cremated) covering from late mesolithic to Bell Beaker.
https://www.biorxiv.org/content/10.1101/2024.03.15.585102v1 - "High-resolution genomic ancestry reveals mobility in early medieval Europe" - Leo Speidel of RELATE, and collaborators from Skoglund's group
They propose a new method which I think is about using f-statistics with inferred allele ages, such that they screen or weight for recent informative alleles?
Given my comments upthread I thought this was interesting:
"We apply this framework to 1,151 available ancient genomes, focussing on northern and central Europe in the historical period, and show that it allows modelling of individual-level ancestry using preceding genomes and provides previously unavailable resolution to detect broader ancestry transformations.
In the first half of the first millennium ~1-500 CE (Common Era), we observe an expansion of Scandinavian-related ancestry across western, central, and southern Europe. However, in the second half of the millennium ~500-1000 CE, ancestry patterns suggest the regional disappearance or substantial admixture of these ancestries in multiple regions. Within Scandinavia itself, we document a major ancestry influx by ~800 CE, when a large proportion of Viking Age individuals carried ancestry from groups related to continental Europe. "
Also they enter into the Wielbark debate:
"In the region of present-day Poland, our analysis suggests several clear shifts in ancestry. First, in the Middle to Late Bronze Age (1500 BC - 1000 BC), we observe a clear shift away from preceding ancestry related to Corded Ware cultures51 ~500CE. Second, in the 1st to 5th century CE, individuals associated with Wielbark culture represent an additional drastic qualitative shift away from the preceding Bronze Age groups, and can only be modelled with a >75% component attributed to the Early Iron Age (EIA) Scandinavian Peninsula, with multiple individuals at ~100% and a stronger affinity in earlier Wielbark individuals.
The Wielbark archaeological complex has been associated with the Goths, a Germanic-speaking group, but this attribution has remained unclear. Our modelling supports the idea that some early Germanic-speaking groups expanded into the area between the Oder and Vistula rivers, but since a considerable proportion of burials during this period were cremations, the possible presence of individuals with other ancestries can not be strictly rejected if they were exclusively cremated (and therefore invisible in the aDNA record)."
David, perhaps your blog post will need to cover both papers? Or maybe you'll do them separately?
Spiedel's new paper proposes that the Polish can be modelled this way:
"Instead, individuals from medieval Poland can only be modelled as a mixture of majority Lithuanian Roman Iron Age ancestry, which is similar to ancestries of individuals from middle to late Bronze Age Poland, and a minority (95% CI: 18.9%-26.0%) ancestry component related to individuals from Roman Italy (p = 0.015) (Figure 3b). These medieval individuals from Poland thus carry no detectable Scandinavian-related ancestry, unlike the earlier Wielbark-associated individuals. Instead, present-day people from Poland are similar in ancestry to these medieval individuals (Figure 3c)."
They exclude a model of Polish individuals as derived from Baltic-Like + Wielbark + some additional relatively more EEF rich population.
I'm not so sure about Poland as Lithuania_Marvele plus Italy_Rome_Imperial though. It seems reasonably clinal sensible at a broad sense of West Eurasia PCA, but I think the lack of presence of ancestries recently derived from Hellenistic Turkey doesn't really fit.
Second paper, this time by Pontus Skoglund, more or less on the same subject:
https://www.biorxiv.org/content/10.1101/2024.03.15.585102v1
@Matt
Yeah, I'll have to blog about both.
Skoglund's models look a bit crap: very basic and distal.
Something Daco-Thracian is a more intuitive southern source for Slavs. I'm still not sure how it would have gotten there although Moldavan Scythians I believe seem to have some of it.
There's no single source of the "excess" EEF in modern Slavs.
It's a function of Isolation-by-Distance since the Iron Age.
@ Davidski and all
New preprint:
https://www.biorxiv.org/content/10.1101/2024.03.18.585512v1
The Genomic portrait of the Picene culture: new insights into the Italic Iron Age and the legacy of the Roman expansion in Central Italy
Abstract:
"Background: The Italic Iron Age was characterized by the presence of various ethnic groups partially examined from a genomic perspective. To explore the evolution of Iron Age Italic populations and the genetic impact of Romanization, we focused on the Picenes, one of the most fascinating pre-Roman civilizations, who flourished on the Middle Adriatic side of Central Italy between the 9th and the 3rd century BCE, until the Roman colonization. Results: We analyzed more than 50 samples, spanning more than 1,000 years of history from the Iron Age to Late Antiquity. Despite cultural diversity, our analysis reveals no major differences between the Picenes and other coeval populations, suggesting a shared genetic history of the Central Italian Iron Age ethnic groups. Nevertheless, a slight genetic differentiation between populations along the Adriatic and Tyrrhenian coasts can be observed, possibly due to genetic contacts between populations residing on the Italian and Balkan shores of the Adriatic Sea. Additionally, we found several individuals with ancestries deviating from their general population. Lastly, In the Late Antiquity period, the genetic landscape of the Middle Adriatic region drastically changed, indicating a relevant influx from the Near East. Conclusions: Our findings, consistently with archeological hypotheses, suggest genetic interactions across the Adriatic Sea during the Bronze/Iron Age and a high level of individual mobility typical of cosmopolitan societies. Finally, we highlighted the role of the Roman Empire in shaping genetic and phenotypic changes that greatly impacted the Italian peninsula."
@ Carlos Aramayo
"Finally, five Picenes and two Etruscans are placed at the basal portion of the R1-L23 branch, together with other ancient Yamnaya, Balkan and Southern Caucasic samples" (p 18); but are you sure that Italians came always from elsewhere and in the Ronman Empire era as the Harvardians imposed everyone? I'll study the paper of course, and this time full of Italian scholars and don't forget the lab that Harvard made to Florence....
"On the other hand, it is worth noting that the trans-Adriatic distribution of the internal branches of J2-M172/M12 was previously interpreted as a clue of a BA expansion from the Balkans in the Italian area and a link between BA Balkans and BA Nuragic Sardinia, possibly with peninsular Italian intermediates that were not observed before [19,49]. Interestingly, two out of three of our J2-M12 Picene samples (PN91 and PN101), due to their phylogenetic position (Additional file 3: Fig. S10) in between the BA Nuragic and the BA Balkan clusters, could represent the descendants of the aforementioned Italian intermediates", but why the Sardinian samples are the oldest found so far?
@Davidski, that may be a worthwhile thing to do.
I think with that model of Poland_MA in Speidel 2024, that may highlight one problem with filtering for recent variants (if I'm correct that is effectively what they do when they describe twigstats) that are present in whole-genome shotgun samples (and missing in capture data).
Recent variants may lack the power to separate between a population that has the right recent relatedness but has the wrong deeper relatedness? The cosmopolitan Imperial_Rome pool might have enough recent variants in common to fit, but their deeper ancestry might be wrong?
This is a problem I think has been present in the past with authors who use methods that focus on reconstructing recent haplotype sharing - there's no cross-check that the model is also plausible for deeper ancestry.
Models which have high power to distinguish between relatively recent ancestry should be used to distinguish between models that are roughly plausible at a deeper level...
Looking at SITable3, this is maybe also their lack of relevant whole-genome shotgun samples from SE Europe in the Iron Age/Roman Era. They might get a better p (than 0.015), and a result more consistent with deep ancestry, with some of those samples, if they have them. The low numbers of these sorts of samples pose practical limitations to deploying their method right now.
("Twigstats takes the Relate output format as input and allows computation of f-statistics directly on genealogies, by using the inferred expected number of mutations on each branch as input, which is computed as the product of a prespecified average mutation rate per base per generation, the branch length, and the number of bases each tree persists. Importantly, Twigstats computes f2-statistics ascertained by an upper date threshold, such that only branches younger than this threshold are used. If a branch crosses the threshold, we only use the proportion of the branch underneath the threshold. Twigstats additionally allows specifying a minimum derived allele frequency and lower date-threshold. Twigstats can also compute f2 statistics on age-ascertained mutations, which is particularly convenient for individuals not built into the genealogies.")
@ Jaakko
That all makes sense. In fact, societes as far as Oleni Ostrov show evidence of contact with Germanic groups. So I would not be surprised if the expansion of Saami & Finnic occurred somewhat earlier than recently proposed.
Rob:
"That all makes sense. In fact, societes as far as Oleni Ostrov show evidence of contact with Germanic groups. So I would not be surprised if the expansion of Saami & Finnic occurred somewhat earlier than recently proposed."
That is impossible, as I have already told you many times:
1. You cannot see language from the DNA - that is impossible.
2. Saami spread north from Southern Finland only ca. 200 CE when Late Proto-Saami disintegrated.
3. Saami replaced earlier non-Uralic languages of Northern Fennoscandia.
what a wonderful, good person Davidski is!
@ Jaako
Obviously Saami isn’t native to northern Scandinavia
When they arrived to Finland isn’t clearly known. stratification of loans is a useful clue but not definitive dating. Only a study of this calibre focussing on the north will bring a solid platform.
And we know that paleo-Germanic existed by 2000 bc. By 1500 we have developed proto-Germanic
By 1500, early west Uralic was in Eastern Europe.
Do the Math
Rob:
"Obviously Saami isn’t native to northern Scandinavia
When they arrived to Finland isn’t clearly known."
True, but we know when Saami spread to the north: only after 200 CE. No traces of any earlier spread, because the people in there still spoke Paleo-European languages when the Saami arrived.
Rob:
"stratification of loans is a useful clue but not definitive dating. Only a study of this calibre focussing on the north will bring a solid platform."
It is definitive, when we know the absolute datings of the Germanic and Scandinavian sound changes, because we have Runic inscriptions since ca. 100 CE. Via these loanwords we can anchor the Saami and Finnic sound changes in the absolute chronology, too.
Rob:
"And we know that paleo-Germanic existed by 2000 bc. By 1500 we have developed proto-Germanic
By 1500, early west Uralic was in Eastern Europe.
Do the Math"
Your math is wrong here. Proto-Germanic is dated around 500 BCE or even later, not 1500. Paleo-Germanic can be dated around 1000 BCE, and around that time Saami and Finnic arrived in the Baltic Sea Region. Everything before that is generally just Pre-Proto-Germanic.
@Carlos
Thanks for sharing that. Did a little recap of the Y DNA from the Etruscan paper compared to this:
Previous Etruscans
7 G2a2b2b1a1
1 J2b2a1 L283
19 R1b1a1b1a (mostly)P312->L2
New Etruscans
3 R1-M269/L23
1 I2-M438/L460
New Picenes
7 R1-M269/L23
3 J2-M172/M12
1 J1-M267/F1614
1 J-M304
Looks like the Etruscans have a lot of G2a compared to the Picenes having J2, both having a lot of R1b-M269.
Seems like there is a relationship between the J2bs in the Picenes and those found in the Cetina culture and elsewhere in the BA Balkans.
@Davidski
Where do you think the authors proposed Italian-like source comes from? I've seen some people propose migrants from either the Balkans or Ashkenazi Jews as the reason for the Southern ancestry, but the former lacks any historical bases (to my knowledge) and the latter seems unlikely due to a lack of Levant-like DNA in Medieval Poles as well. It seems odd to even model Poles with an Italian-like source from the Roman period, but I'm open to surprises.
@ Jaako
“Your math is wrong here. Proto-Germanic is dated around 500 BCE or even later, not 1500. Paleo-Germanic can be dated around 1000 BCE, and around that time Saami and Finnic arrived in the Baltic Sea Region. Everything before that is generally just Pre-Proto-Germanic.”
These are just outdated guesses and they misunderstand the limitations of the craft (not science) of linguistics :)
@ Jaakko
“because we have Runic inscriptions since ca. 100 CE.”
As per above, early runes are single word inscriptions
They do not attest the prevailing diversity of Germanic languages at the time
@Noble Goth
The Polish Medieval analysis in the Speidel preprint doesn't make much sense, because it defies expectations based on other analyses like PCA, deep ancestry, and uniparental markers.
You always have to be cautious about claims in papers that introduce super duper new methods, because often it's just hot air.
@ Jaakko
“1000 BCE, and around that time Saami and Finnic arrived in the Baltic Sea Region”
Nothing against that date. But specifics will be refined by aDNA
The Picene paper has shown that all the Italic peoples analyzed, i.e. Latins, Daunians and Picenes, share genetic markers with the Balkans. Some examples
*KUK002 (1.163 BCE)-Koukounaries, Paros, Mycenean, Greece-HapY-J1b-Y6313>Y6304>F1614
*I20231 (615 BCE)-Muğla-Değirmendere, Carian, Anatolia-HapY-J1b-Y6313>Y6304>Y19093
*PN172 (800 BCE)-Novilara, Picenes, IA_Italy- HapY-J1-M267>Y6313>Y6304>ZS3642>Z2223-Y19093
*I7231 (1.247 BCE)-Dimov Grob, MKD_BA, North Macedonia-HapY-R1b-Z2103>Z2110>CTS7556
*I8112 (666 BCE)-Lisicin Dol, MKD_Anc, North Macedonia-HapY-R1b-Z2103>Z2109>Z2110>CTS7556
*PN62 (725 BCE)-Novilara, Picenes, IA_Italy-HapY-R1b-Z2103>M12149>Z2106>Z2109>Z2110>CTS7556
*I24345 (950 BCE)-Velim-Kosa, Bronze Age, Croatia-HapY-J2b2a-L283
*PN101 (660 BCE)-Novilara, Picenes, IA_Italy-HapY-J2b-L283
*ORD014 (489 BCE)-Ordona, Apulia-Daunians, IA_Italy-HapY-J2b2a-L283
mtDNA-H35
Vodovratski Pat, Macedonia, Iron Age-I7233 (855 BCE)-Lazaridis, 2.022
Novilara, Oicenes Iron Age, Italy-PN56 (650 BCE)-F.Ravasini, 2.024
mtDNA-H46-
Bezdanjača Cave, BA_Croatia-I18079 (1.500 BCE)-Lazaridis, 2.022
Novilara, Picenes, IA_Italy-PN85 (650 BCE)-F.Ravasini, 2.024
+mtDNA-H1at
Velika Gruda, BA_Montenegro-14500/I14497 (1.150 BCE)-Lazaridis, 2.022
Bezdanjača Cave, BA_Croatia-I18733 (1.150 BCE)-Lazaridis, 2.022
Novilara, IA_Italy, Picenes-PN50 (650 BCE)-F.Ravasini, 2.024
In the PCA we see a clear Adriatic-Balkan cluster that includes the Picenes and Daunians, the Etruscans are closer to the Central European Celts, the South Gauls and the Iberians and the Latins are between Etruscans and Adriatics. IMO the Italic languages could be of Balkan origin and not related to the BBC or to Bronze Age migrations originating in Central Europe. After all, Etruscans unlike Picenes and Daunians were mostly U152.
@Davidski
Very much agree. I'm very confused why they thought using an Imperial Roman sample as a model was even a good idea with their current method as well, it seems a bit rash - quite similar to the recent Germanic paper as well modelling 'East Scandinavians' as 7% Latvian HG, claiming a Baltic migration. All seems extremely short-sighted and lacking any actual merit in some cases.
In my honest opinion, it seems far more reasonable to assume the Southern ancestry seen is mediated through interactions between less Eastern-shifted Europeans from around Poland. The current hype I've seen is reminding me of the entire Scythian debacle.
@Noble Goth
I'm going to look at this issue in detail over the next week or so, and then I'll post my findings here and at bioRxiv, and I'll also get in touch with the authors.
But I think it's already obvious that Speidel et al. have simply produced a statistical construct (with a relatively poor statistical fit) based on available data that doesn't reflect reality, and they've taken this too literally.
@Jaakko
"It is definitive, when we know the absolute datings of the Germanic and Scandinavian sound changes, because we have Runic inscriptions since ca. 100 CE. Via these loanwords we can anchor the Saami and Finnic sound changes in the absolute chronology, too."
One thing that occurs to me, if, as Dahl suggests, the Early Runic language doesn't exactly represent the language spoken by people locally at the time, then sound changes observed in loanwords in Saami or Finnic could be older than they seem when compared to runic inscriptions, couldn't they? Probably not by very much, but still.
There was a funny error in Speidel's preprint:
"Present-day populations of Hungary do not appear to derive their ancestry from early medieval Longobards, and are instead more similar to Scythian ancestry sources (Supplementary Figure 5), consistent with the impact of later arrivals of Avars, Magyars and other eastern groups"
The Scythian groups they compared to were the Hungarian "Scythians", which are just local Pannonians dubiously considered to have been Scythians by Turanophilic Hungarian scholars. As such these people have no eastern ancestry (unlike actual Scythians).
And since these people lack this eastern input, the similarity between them and modern Hungarians is not a sign of impact from Avars, Hungarian Conquerors and other groups as the authors suggest, but rather a lack of significant input from such populations.
@ Gaska
What do you want to demonstrate through that? Already the paper writes that there were reciprocal exchanges through the Adriatic, and that probably from at leasst the Palaeolithic, and we already knew that beginning from the oldesr subclades of hg R1b, but why not I-M223, clearly older in Italy and many others. To demonstrate something you'd need the terminal SNP, what the paper doesn't give. Only a few samples:
1) R1b>CTS7556* has the oldest survived samples in Belgium and the aDNA from the Balkans is only 3900 ybp but its formnation is 5300 ybp.
2) J2b2a-L283 has the oldest samples in Sardinia and Italy, and I have always written that the origin is around the Caucasus (but not through the Levant, but either Anatolia or eastern Europa), i.e. J-YP91 and above all J-Z600 and J-Z2509 etc. MOK15, J-Z615* from Serbia 3900 ybp, I demonstrated like other R-Z2118 having come from the eastern Alps.
3) mtDNA H1at (I looked at that because more defined than other we could get the FASTA file to say more) is deeply rooted in Italy and not in the Balkans etc etc
4) The R1-M269-L23 have to be seen until the terminal SNP.
5) I invited you all to read the paper which "demonstrated" that Albanian language is the oldest IE one, who wrote them and his/her scientific career...
@ Erik
''the Early Runic language doesn't exactly represent the language spoken by people locally at the time, then sound changes observed in loanwords in Saami or Finnic could be older than they seem when compared to runic inscriptions, couldn't they? Probably not by very much, but still.''
This is more complex than Jakko admits
Firstly, the defining sound shifts pg PGMc are simlpy said to have occurred c. 500 BC. Don Ringe for ex clearly states these are guesses. We have no inscriptions to set them, although there are some Roman source names which are said to show a late persistance of pre-PGMC, but this could be because they were mediated via non_GMc intermediaries (e.g. Celtic).
Secondly, the pervasive incursion of Norse loans into Saami doesn't preclude the presence of Saami -like languages or a dialect continnum before that event. I'd say we see west-Uralic groups in Finland c. 1000-500 BC, and this is when paleo-lapplandic was replaced. So, several hundred years earlier than assumed. A significant revision
@ Gaska
And about H35 and H46 what do you want to demonstrate? That the oldest samples are from the Balkans because there are older aDNA from up there? But only because they tested older samples, but it doesn't mean that there aren't even older in Italy, only that older samples from Italy aren't tested. You should look at the YFull tree, with all the issues I denounced because they didn't resolve the question of the heteroplasmies.
H35* is above all found in central or northern Europe and also Italy. No samplpe from the Balkans, and H35e and H35f, separated 9400 years ago, are all in favour of Italy.
H46* the same. The samples separated 7600 years ago are everywhere, but above all in Italy than the Balkans.
@ Gio
"...but are you sure that Italians came always from elsewhere and in the Ronman Empire era as the Harvardians imposed everyone? I'll study the paper of course, and this time full of Italian scholars and don't forget the lab that Harvard made to Florence..."
Of course we can wait, also, to more details when peer-reviewed paper comes out.
@ Romulus
"Seems like there is a relationship between the J2bs in the Picenes and those found in the Cetina culture and elsewhere in the BA Balkans."
Yes, I think there's even a sample, maybe published in Mathieson et al (2018), J2b2a1a1a1b2-Z38240, from Bronze age, around the 16th century BCE (Cetina culture).
@Gio
The most basal adna sample of J2B2-L283 is from Russia, north of the Caucasus, where it very likely originated, and was picked up by SS/Suvorovo, passing it to the Yamnaya tribes that travelled along the Danube and left traces of it (like Mokrin).
@DragonHermit
and J1-Y19093?
@Copper Axe
All the people of Hungarian descent I've met have no interest in what the internet has to say about their people. It's an entirely anecdotal account but is still refreshing.
@Matt
Honestly I asked a question that while not quite "rhetorical" perse, was still an unreasonably vague prompt. Asking from a place of respect can you please explain your thoughts on the genetic, linguistic, and archeological aspects on the Southern arc theory? I acknowledge my own flippant, even brusk attitude and can only hope you respond with sincerity in kind.
THank you
@ DragonHermit @ Carlos Aramayo
I probably was one of the first to say that hg J (both 1 and 2) came from around the Caucasus, already when the Harvardianlikes wanted to impose an origin in the Levant and that European J-s came from the Levant and I said that they came independently either through Anatolia or even eastern Europe. The same I wrote in thousands of letters about J-Y15223, probably entered the Jewish pool in Italy but come before from around the Caucasus. The same for the E-PH3893 of Napoleon very diffused in Lunigiana between Tuscany and Liguria. And against Hunter Provyn I wrote thousands of letters about the J-L283 probably even from central Europe and not Middle East as he thought (it is his Y) and his Phylogeographer as a great bluff. But what I wrote above also about the mt-s remains, and I challange everyone to demonstrate the other way around. More of course when we get the terminal SNPs and the FASTA files.
Erik Andersson:
“One thing that occurs to me, if, as Dahl suggests, the Early Runic language doesn't exactly represent the language spoken by people locally at the time, then sound changes observed in loanwords in Saami or Finnic could be older than they seem when compared to runic inscriptions, couldn't they? Probably not by very much, but still.”
Why would it not represent the local language? I did not see such a claim from him. Only around the Viking Age he assumes the expansion of the Danish koiné. That is much later than Early Runic.
I do not follow why you think such a thing would even theoretically affect the dating of the sound changes. Could you write open what you mean by this?
Rup:
“This is more complex than Jakko admits
Firstly, the defining sound shifts pg PGMc are simlpy said to have occurred c. 500 BC. Don Ringe for ex clearly states these are guesses. We have no inscriptions to set them, although there are some Roman source names which are said to show a late persistance of pre-PGMC, but this could be because they were mediated via non_GMc intermediaries (e.g. Celtic).”
We have enough evidence: the Proto-Germanic sound changes are later than the Proto-Celtic sound changes, and the Northwest Germanic sound changes can be seen in names in Latin inscriptions. One century here or there, but there is no way that the Proto-Germanic sound changes could be dated to 1500 BCE as you claimed. You are alone with your ungrounded beliefs.
Rabb:
“Secondly, the pervasive incursion of Norse loans into Saami doesn't preclude the presence of Saami -like languages or a dialect continnum before that event. I'd say we see west-Uralic groups in Finland c. 1000-500 BC, and this is when paleo-lapplandic was replaced. So, several hundred years earlier than assumed. A significant revision”
There is no point inventing from zero evidence some early Saami expansions. I already told you that Northern Fennoscandia was populated by Paleo-European speakers, and they were replaced by the Saami speakers only around 500 CE. You just cannot ignore the linguistic results and make your own claims out of nothing, that is unscientific.
@Noble Goth, well, in Table S3 they really did try every viable model that passed their criteria of involving samples from Europe from the Iron Age to Roman period.
It just seems that the Italy.Imperial(I).SG and Lithuania.IronRoman.SG was the strongest outcome.
Though it's also noteworthy that they seem to have in their supplement additional models with slightly higher p-values that allow some lower level input from any of Scandinavian / Western European Iron Age / Steppe populations (about 6-7%). The two way model is preferred as giving a lower number of populations not being the best p.
Table: https://i.imgur.com/eDGhQpa.png
I think there is still possibly some problem of being limited in resolution to resolve "southern" shift sources through the modelling choices and available sample though.
Using their method they used:
Fixed outgroups: Ekven_IA.SG, Ireland_BA.SG, Estonia_CordedWare.SG, Anatolia_EBA.SG, Yamnaya
Rotational sources: Scandinavian_Peninsula_EIA(I).SG, Scandinavian_Peninsula_EIA(II).SG, Britain.IronRoman.SG, Poland_BA(I).SG, Poland_BA(II).SG, Poland_BA(III).SG, Lithuania.IronRoman.SG, Italy.Imperial(I).SG, Italy.Imperial(II).SG, AustriaFranceGermany(I).IronRoman.SG, AustriaFranceGermany(II).IronRoman.SG, AustriaFranceGermany(III).IronRoman.SG, Portugal.IronRoman(I).SG, Portugal.IronRoman(II).SG, Russia_Sarmatian.SG, Saami
which means that when any of those samples in the rotational sources weren't being used as a source, they were an outgroup. So maybe there is enough power to distinguish some ancestry sources, but more Near Eastern IA shotgun samples would enhance things?
I think maybe they might do better if they include some Iron Age Levant and North African groups in the outgroups, and used any SE European Iron Age samples as rotational.
@ Jakko
“We have enough evidence: the Proto-Germanic sound changes are later than the Proto-Celtic sound changes, and the Northwest Germanic sound changes can be seen in names in Latin inscriptions. One century here or there, but there is no way that the Proto-Germanic sound changes could be dated to 1500 BCE as you claimed. You are alone with your ungrounded beliefs.”
Everything is derivative and inferred
Assumption based on assumption present as “indisputable evidence”
No shit proto-Celtic is older than proto-Germanic
But the development of proto-GMC is gradual , and I claimed that by 1500 we have PGMc
which are these Latin inscriptions you speak of ?
“There is no point inventing from zero evidence some early Saami expansions. I already told you that Northern Fennoscandia was populated by Paleo-European speakers, and they were replaced by the Saami speakers only around 500 CE. You just cannot ignore the linguistic results and make your own claims out of nothing, that is unscientific. ”
As per above, paleogenomic is science
cherry picked linguistics is spit & polish homeopathy
“and I claimed that by 1500 we have early PGMc
…….
We need to recall Jaakko’s buddies on FraudArchiver tried to prove a migration of Finns initiating the “protoGermanic” sound shifts with some sham qpAdm models. Didn’t work out to well in light of the present data
Röp, I'm truly sorry that either your knowledge or your mental capacity is insufficient for understanding linguistics and interdisciplinary studies and comprehending that you cannot see language from the DNA. I will go on discussing only with people who have sufficient knowledge and understanding about these questions. :)
@Jaakko
Page 8-9 + footnote 10 of his previously discussed "The origin of the Scandinavian languages", Dahl discusses Makaev's hypothesis of a "Runic koiné" in support of dialectal differences within Scandinavia at the time. I don't feel like quoting the whole thing here.
If innovations observed in Proto-Norse loanwords occured dialectally before becoming widespread, and/or the runic inscriptions reflect a slightly more archaic form of the language than what was spoken at the time they were made (which is common in a literary language), then the absolute dating of those loanwords could be off, giving the impression that they are more recent than they really are. I'm not suggesting that they're 1000 years older, but maybe as much as 150-200 is possible.
I'm reminded of Heikillä 2014 (Bidrag till Fennoskandiens förhistoria i tid och rum), where he demonstrated that the pre-Northwest Germanic sound shift *ē > *ǣ occurred before Grimm's law in some loanwords in Saami and Finnic, implying that it existed dialectally for a long time before becoming widespread.
I'm not trying to push some specific scenario, just keeping an open mind.
@Gio
Well, I am not an expert in Italian archeology but I believe that these results demonstrate the influence of the Cetina culture at least on the Italian Adriatic coast. J2b-L283 is one of the few markers that can certainly be related to Indo-European languages because there are samples in the Mycenaean culture. I agree on its Caucasian origin but I do not agree on the route followed to reach the Balkans, not through SS-Suvorovo (unless someone find some sample of this marker in the steppes) but through Asia Minor as so many markers below J2a & J1 that reached the Balkans.
While the Etruscan mitochondrial markers are very similar to the Spanish and French ones from the Iron Age, those of the Picenes and Daunians are identical to the Balkan ones. More examples
-mtDNA-H13a2/a
Tiryns, BA_Mycenean culture, Greece-TIR001 (1.326 BCE)- Skourniatioti, 2.023
Velim-Kosa, BA_Cetina culture, Croatia-I24344 (950 BCE)-Patterson, 2.021
Novilara, Picenes, IA_Italy-PN125 (650 BCE)-F.Ravasini, 2.024
+mtDNA-X2m
Matkovici, BA_Croatia-I5074 (1.447 BCE)-Lazaridis, 2.022
Novilara, Picenes, IA_Italy-PN90 (650 BCE)-F.Ravasini, 2.024
Do you think it is a coincidence? If you add to this that both Latins, Picenes and Daunians share the Z2103 & J2b-L283 markers with the Balkans, then no one can deny the Balkan genetic and even linguistic (Mesapians) influence on the Italic peoples.
I am not a linguist either and I don't know if it is possible to relate the Balkan Indo-European languages (including Mycenaean) with Latin, Osco-Umbrian, Messapian etc…. but maybe ome specialist in Indo-European languages should think about this genetic data.
And of course, I do not think that establishing this genetic connection is an offense to Italy, the Italians or you. If you think that these markers are older in your country you can only do one thing to prove it, find samples older than the Balkan ones.
@Jaako
Of course we can see language from DNA.
We have written evidence of Paleo-european languages related to certain cultures and now thanks to genetics we can establish with complete certainty which uniparental markers spoke those languages.
If you are lucky enough to have enough ancient samples to determine the geographical origin of a male marker and you are able to demonstrate the existence of genetic continuity & what language that lineage spoke (with written evidence of course) then you can perfectly trace the origin of the language that marker spoke.
It is not so difficult to understand, perhaps the absence of written records during the Iron Age in Finland (and Scandinavia in general) has led you to be so skeptical about the ability of genetics to draw conclusions regarding the geographical origin of your languages.
@Matt
Could agree on an issue for the Southern shift being a sample problem. The odd-ball is to me is why the hell they chose to stick with Roman Imperial as a model for it. The problem is, I think as Davidski said earlier, it defies earlier analyses which I could agree with. The problem with Roman Imperial as a sample is it should indicate recent MENA ancestry along with it, considering the Imperial period saw a large population shift to MENA-related peoples, mainly from the Levant. I've never seen a single analyses of any Polish sample, modern or not with any Levant-like admixture (unless they have recent ancestry from say, Ashkenazi Jews) - not to mention any bog-standard analysis or even a simple PCA would have shown such admixture years ago. If I read the pre-print correctly they suggested the Roman Imperial-like ancestry was roughly 18-25% of Medieval Poles, which is enough to shift them noticeably on a PCA. Yet on PCA, they're more Eastern than Modern Poles. Confused as well?
It's more of a head-ache when they then didn't suggest Ashkenazis or East-Meds as a source, but Balkan people from some sort of back-migration after the Slavic expansion southwards considering their sample was Imperial Roman. I saw another poster here mentioning an issue with the Hungarian analysis as well, so in hindsight I'm not surprised, and after reading the pre-print a bit more it seems the authors models are shitty or they're just confused.
Either that or I'm just as silly.
These people don't know anything about Polish population structure and history.
They're basically just statisticians going with the best stats based on the samples they have available to them.
But their model, even though it doesn't reflect reality, can be put in a useful context by explaining how it came about and why it's wrong.
I'll do that soon here and at bioRxiv, probably this weekend, and I think it'll make a pretty good blog post about Polish population history.
Look forward to the read - I really do wish Polish population genetics wasn't cluster fucked by Western scholars awfully misunderstanding it. Hopefully once we're done brainrotting we can actually get a good paper on Polish and even other Eastern European populations histories.
Gaska:
“Of course we can see language from DNA.
We have written evidence of Paleo-european languages related to certain cultures and now thanks to genetics we can establish with complete certainty which uniparental markers spoke those languages.”
That is not seeing language from DNA. That is comparing linguistic results to genetic results and finding matches. That is the scientific way.
The unscientific way is to predict language from DNA like Rob does. He ignores all the linguistic results and thinks for some reason that he can see language better directly from DNA. Of course that is impossible, because language is not inherited in DNA. Do you understand the difference now?
Gaska:
“If you are lucky enough to have enough ancient samples to determine the geographical origin of a male marker and you are able to demonstrate the existence of genetic continuity & what language that lineage spoke (with written evidence of course) then you can perfectly trace the origin of the language that marker spoke.”
Here you go wrong. You have no way to know without linguistic results whether or not there is a linguistic continuity alongside the genetic continuity. Let us assume that the birth mutation of paternal lineage X occurred in 2000 BCE. In every generation the carriers of this paternal lineage marry women and get children. Their sons carry the same paternal lineage, but did they speak their father’s language or their mother’s, or both? In 1000 years, there are ca. 40 generations = 40 possibilities for a language shift occurring in that lineage.
So, there is absolutely no basis for you to claim that you can see language from this paternal lineage! You can only say something about the language spoken by the carriers of this paternal lineage, if you have linguistic results from that place in that time where these men lived.
Gaska:
“It is not so difficult to understand, perhaps the absence of written records during the Iron Age in Finland (and Scandinavia in general) has led you to be so skeptical about the ability of genetics to draw conclusions regarding the geographical origin of your languages.”
It has nothing to do with early written records, because linguistic results are not dependent on written records. It has all to do with understanding how languages are inherited (not in DNA) and replaced by new languages. Apparently this is rather difficult to understand, because this must be explained to many people, and for some of them over and over again. Can you understand it now, after reading carefully my message?
It'll be entertaining seeing Jakko writhe when we get aDNA from eastern Finland, for ex, showing the arrival of Uralic speakers c. 700 BC. Undoubtedly, it'll be his patented disinformed sermons about how DNA doesn't speak and that Uralic actually expanded from a mystery Corded Ware group.
@ Gaska
"And of course, I do not think that establishing this genetic connection is an offense to Italy, the Italians or you. If you think that these markers are older in your country you can only do one thing to prove it, find samples older than the Balkan ones".
I already answered your questions. Your proof is the same used from the kurganists for R-L23-Z2103: no doubt that 5300 years ago this haplogroup was in Yamnaya and expanded from up there, but 5400 years ago is 700 years after that this haplogroup formed and very likely it came from elsewhere, because if older samples were in Yamnaya, largely tested, they would have found it, but they didn't. Your "proof" is that these hgs you quoted were found in the Balkans in 3300 ybp and in Italy 950/650, thus you conclude post quem, propter quem, but those hgs formed many thousands of your before and you don't know where they were then. Of course you can ask me to find older samples of them in Italy. The same happened for villabruna, which is the oldest R1b1 found so far, in fact my opponents like Rich s (but where is he now?) have to say that villabruna is a dead end line, but you accepted that that clan could be at the origin of the expansion of this haplogroup.The samples found at Mokrin did come from the eastern Alps before, because also the archaeological culture was the same, and all the oldest R-Z2118 have been found in Italy or close to it. You ask for oldest samples in Italy, but you know how did happen the case of Adria 1 and 2, thrown away from the discover tree of FTDNA after the letter of Rich S to Goran, and don't forget that these Italian scholars, among them Cruciani and Trombetta, were those who supported against me that R-V88 did come from Near East, and you know how it ended: the oldest samples are found so far in central Europe, but also in Italy, and none doubts about its European origin. And perhaps you know that the Harvardians built the lab for aDNA at Florence (look at Caramelli) forbidding them to test the Y but the mt and before that they gave them tons of old bones found by Francesco Mallegni of Pisa University. Perhaps I am not the only one to think these things now.You post other samples of mt now. To the previous ones I answered before, demonstrating my truth not through the aDNA (it is difficult to find what these scholars don't want to find), but through the presence of the oldest separated samples in modern DNA, and to me this is the same proof I used for forseeing Villabruna, and Villabruna was found, of course by chance by whom didn't think to find it.About J-L283 I discussed a lot with Hunter Provyn, and the oldest and rare subclade (YP91) seemed to have come more from central Europe and even Poland than Anatolia etc etc but the oldest sample is in Italy (ORC007).
"-mtDNA-H13a2/aTiryns, BA_Mycenean culture, Greece-TIR001 (1.326 BCE)- Skourniatioti, 2.023Velim-Kosa, BA_Cetina culture, Croatia-I24344 (950 BCE)-Patterson, 2.021Novilara, Picenes, IA_Italy-PN125 (650 BCE)-F.Ravasini, 2.024"
For H13a2a we have to say that it is above all in India, Pakistan, Caucasus. I24344 from Croatia is 2900 years old. The sample in Italy may have come from everywhere and I'll say more through the FASTA file, but Italy also has H13a1, certainly not from the Balkans.
"+mtDNA-X2mMatkovici, BA_Croatia-I5074 (1.447 BCE)-Lazaridis, 2.022Novilara, Picenes, IA_Italy-PN90 (650 BCE)-F.Ravasini, 2.024"
Look at the YFull tree: the oldest X2m'n 10300 ybp is YF019625 ITA, and the second X2m1a, 4800 ybp is JX153046.1 ITA etc etc. I5074 aDNA from Croatia is 3300 ybp.
@Romulus
But the La Tene French samples are almost all P312+. Either the Halstatt period was less stratified, or we're looking at a French population with an earlier Neolithic layer that was preserved in this particular sample. I-M26+ is most certainly a SW French/Iberian hunter-gatherer lineage that spread with the Atlantic farmers. Unlikely anything to do with Central Europe.
@All
Does anyone have a fairly comprehensive list of the Y-chromosome haplogroups in the Polish Medieval and Polish Viking samples?
@ Awood
'' I-M26+ is most certainly a SW French/Iberian hunter-gatherer lineage that spread with the Atlantic farmers. Unlikely anything to do with Central Europe.''
It's the opposite. I2a-CTS595 moved north to south. All the earliest sampels are found in central -northern Europe, arriving to Italy & Iberia only during the Late Chalcolithic
@ Rob
"It's the opposite. I2a-CTS595 moved north to south. All the earliest sampels are found in central -northern Europe, arriving to Italy & Iberia only during the Late Chalcolithic".
https://www.yfull.com/tree/I-CTS595/
Ahahah, all the oldest separated samples are in Italy!
@Gio
catacomb theory of some IE languages https://postimg.cc/MfvhHLT6
@Jaako
1-I understand the difference perfectly, your “scientific way” is the only irrefutable proof that a certain marker spoke a certain language in a certain historical period. For example, we know that Df27>Z195 spoke Iberian because the men analyzed in the villages where texts written in Iberian have been found belong to that marker. Linguistic results can be linked to a certain genetic marker and no one can prove otherwise
2-In Europe, during the Bronze Age, linguistic results that can be considered as irrefutable proof can only be found in the Peloponnese and Crete because only in those regions are written texts that can prove beyond doubt what language (or languages) was spoken by those cultures-Then in the Iron Age we have only Aquitanian, Tartessian, Iberian, Celtiberian, Ligurian, Etruscan & various Italic and Balkan Indo-European languages. Thanks to genetics we already know which were the male lineages linked to these cultures, although it is evident that the more sites that are analyzed, the more we will know about the genetic composition of the different European cultures.
3-Leaving aside the cultures that have left written texts, the language or languages spoken in the rest of Europe from the Paleolithic to the Iron Age will always be a mystery to us, no one will ever be able to prove scientifically that the BA Britons spoke Indo-European, the Iron Age Scandinavians Germanic, the Urnfielders proto-Celtic or the Trzciniec culture Balto-Slavic. The possibility of joining efforts between linguistics and genetics would have been completed and we would not be able to move forward
4-So, on the one hand, linguistics can try to reconstruct paleolanguages, establish relationships between different languages, study loanwords, etc., and thus elaborate more or less sophisticated theories about common homelands, geographical origins, etc. ..... but in the absence of written records we will always be talking about more or less convincing theories and in most cases simply fairy tales
5-On the other hand, genetics has the ability to trace the origin of any uniparental or autosomal marker until determining without a doubt its exact geographical origin. In this sense it is a science immensely superior to linguistics because its evidence can become irrefutable.
To be continue
@ Gio
“https://www.yfull.com/tree/I-CTS595/
Ahahah, all the oldest separated samples ar”
“Oldest separated samples”? I don’t think that’s a scientific deduction
The earliest I2a1b is in Goyet , Belgium 14000 Bc
The earliest in Italy is 3000 Bc.
Big difference
Erik Andersson:
“Page 8-9 + footnote 10 of his previously discussed "The origin of the Scandinavian languages", Dahl discusses Makaev's hypothesis of a "Runic koiné" in support of dialectal differences within Scandinavia at the time. I don't feel like quoting the whole thing here.”
There seems to be no actual evidence supporting Makaev’s view, and it seems actually contradicted by Dahl in page 9, as well as in Antonsen’s review in 1968.
Erik:
“If innovations observed in Proto-Norse loanwords occured dialectally before becoming widespread, and/or the runic inscriptions reflect a slightly more archaic form of the language than what was spoken at the time they were made (which is common in a literary language), then the absolute dating of those loanwords could be off, giving the impression that they are more recent than they really are. I'm not suggesting that they're 1000 years older, but maybe as much as 150-200 is possible.”
OK, thanks for clarifying. You are right in that innovations reach different regions at different times (or they do not reach some regions at all, in which case we get a new dialect boundary). It is still difficult to believe a delay of centuries – that is so long a time that sound changes have already stopped spreading long before that. Namely, sound changes only affect a short period of time, because they change the pronunciation in every word and every name in which that particular sound appears. Instead, single words can spread continuously step by step for a long time from region to region. Sound changes do not act like that.
Moreover, sound changes do not affect on mutual intelligibility like new words or new meanings in old words. Sound changes even today still show the ancient dialect boundaries, even when convergence has led to shared developments and innovations in vocabulary and all the other levels of language. South Estonian is (or was until recently) seen as a part of Estonian language, but based on sound changes, South Estonian was the first unit to branch off from Proto-Finnic.
Archaicity of literary languages is due to a long literary tradition: spoken languages change organically faster, but written languages change slower. In the Early Runic era there was no literary tradition yet to mention, but writers just put down the words how they pronounced them.
Erik:
“I'm reminded of Heikillä 2014 (Bidrag till Fennoskandiens förhistoria i tid och rum), where he demonstrated that the pre-Northwest Germanic sound shift *ē > *ǣ occurred before Grimm's law in some loanwords in Saami and Finnic, implying that it existed dialectally for a long time before becoming widespread. I'm not trying to push some specific scenario, just keeping an open mind.”
Open mind is great! And too rare, unfortunately.
Heikkilä seems to be alone with that view. He assumes that the sound substitutions were exceptionless: Finnic *KK <-- Germ. *K and Finnic *K <-- Germ. *G. But there are exceptions in the clearly later loanwords, so it is possible that there were exceptions also in the earlier loanwords. There is no compelling reason to deny the possibility that Finnic *K could also come from Germanic *K, and therefore no need to assume that the Northwest Germanic change *ē > *ǣ > *ā would have occurred much earlier in some pre-dialect than in the others.
@David
Mr David ,Is it possible that the people of the Cernavodă culture from the Eneolithic of Ukraine were the ancestral speakers of the Hittite languages?
@Davidski
Do you know of any papers/theories that allow "sanity" checks on individual samples or the pops? Unless there is a method that allows us to discard all the garbage I don't think we will ever see the end of cherry picked confirmation bias in the literature.
I do have a few ideas but if there are better proposals in the literature I could avoid consulting statisticians whose expertise goes beyond the purview of your average data scientist.
@ Rob
Of course I have always said that an only one sample isn't sufficient for supposing its presence from the beginning where it is found now, but:
I-L1274FT63794 * Z27360 * BY21561+66 SNPsformed 11800 ybp, TMRCA 7000 ybpinfo
I-L1274*
id:ERS256268ITA [IT-CA]
I-L160PF4089 * PF4060 * PF4053+39 SNPsformed 7900 ybp, TMRCA 5800 ybpinfo
id:ERS256306ITA [IT-CA]
I-Y21970Z27151 * Z27153 * Z27138+32 SNPsformed 5800 ybp, TMRCA 3500 ybpinfo
I-Y21970*
id:ERS256298ITA [IT-CA]
I-Y21971Y22899 * Y21971 * PH4419+3 SNPsformed 3500 ybp, TMRCA 3100 ybpinfo
I-Y21971*
id:YF016890ITA [IT-PA]
I-PF4088PF4088/S5312/V2447formed 5800 ybp, TMRCA 5800 ybpinfo
I-PF4088*
id:HGDP01069ITA [IT-CA]srd
I-Y36693A17276 * A17275 * A17271+4 SNPsformed 5800 ybp, TMRCA 5700 ybpinfo
I-Y36693*
id:YF010340IRL [IE-RN]
id:ERS257169ITA [IT-CA]
I-PF4104PF4112 * PF4114 * PF4115+24 SNPsformed 5800 ybp, TMRCA 2400 ybpinfo
I-PF4104*
I-Z27054Z27054 * Z27056 * Z27052+2 SNPsformed 2400 ybp, TMRCA 1500 ybpinfo
id:HGDP00671ITA [IT-CA]srd
I-PF4144PF4144info
id:ERS256323ITA [IT-CA]
id:ERS256321ITA [IT-CA]
I-S18051S18051 * Y125866formed 5400 ybp, TMRCA 2400 ybpinfo
id:YF014488
id:ERS256301ITA [IT-CA]
I-PF4188PF4193 * PF4195 * FGC93139/FT63759+6 SNPsformed 4700 ybp, TMRCA 3600 ybpinfo
I-PF4188*
id:HGDP01071ITA [IT-CA]srd
I-PF4211PF4211 * PF4218 * PF4219+3 SNPsformed 3600 ybp, TMRCA 2300 ybpinfo
id:YF063112FRA [FR-42]
id:ERS256437ITA [IT-CA]
I-PF4223PF4225 * PF4223formed 3600 ybp, TMRCA 1750 ybpinfo
I-PF4223*
id:ERS256453ITA [IT-CA]
I-Y3945Y3948/Z26571 * Y3954/Z26577 * Y27472+10 SNPsformed 1750 ybp, TMRCA 50 ybpinfo
id:HGDP00665ITA [IT-CA]srd
id:ERS256444ITA [IT-CA]
Of course you are right about the oldest I-Y4192Y4241 * Y4240 * V7790/S24016+21 SNPs formed 14600 ybp, TMRCA 12100 ybp, but why not Tagliente 2 17000 years ago or Ostuni etc etc. Haplogroup I was among the WHG and probably at least from 17000 years ago also R1b1.
@A.Wood said-But the La Tene French samples are almost all P312+.
That is not true, both the Hallstatt and LaTene cultures in France are very heterogeneous in their male markers & several subclades of I2a are surprisingly abundant
Gaska:
“1-I understand the difference perfectly, your “scientific way” is the only irrefutable proof that a certain marker spoke a certain language in a certain historical period. For example, we know that Df27>Z195 spoke Iberian because the men analyzed in the villages where texts written in Iberian have been found belong to that marker. Linguistic results can be linked to a certain genetic marker and no one can prove otherwise”
Exactly.
Gaska:
“3-Leaving aside the cultures that have left written texts, the language or languages spoken in the rest of Europe from the Paleolithic to the Iron Age will always be a mystery to us, no one will ever be able to prove scientifically that the BA Britons spoke Indo-European, the Iron Age Scandinavians Germanic, the Urnfielders proto-Celtic or the Trzciniec culture Balto-Slavic. The possibility of joining efforts between linguistics and genetics would have been completed and we would not be able to move forward”
Now you do not know what you are talking about. Sure, Paleolithic languages are beyond our reach, but from the timeframe of the expansion of the extant language families we do have many kinds of linguistic evidence. Take for example the Saami languages. We know when and where Late Proto-Saami was spoken, because we have layers of loanwords from Baltic, Germanic and Finnic, fitted in the framework of phonological history of each language lineage, anchored in the written attestations of Early Runic language (representing Northwest Germanic).
Therefore we know that Late Proto-Saami began to disintegrate ca. 200 CE in Southern Finland, and we know that the Saami proto-dialects replaced Paleo-European languages during the following centuries in Northernmost Fennoscandia – at the same time when they established contacts with Scandinavians in the Norwegian coast.
Gaska:
“4-So, on the one hand, linguistics can try to reconstruct paleolanguages, establish relationships between different languages, study loanwords, etc., and thus elaborate more or less sophisticated theories about common homelands, geographical origins, etc. ..... but in the absence of written records we will always be talking about more or less convincing theories and in most cases simply fairy tales”
In most cases? Absolutely wrong. You just do not understand the methods of historical linguistics, so your opinion is based on ignorance. You should always try to comprehend the methods behind the results, before you try to disprove or dismerit the results. There are plenty of books titled “An Introduction to Historical Linguistics” or something similar; just pick one and gain some knowledge on the topic.
Gaska:
“5-On the other hand, genetics has the ability to trace the origin of any uniparental or autosomal marker until determining without a doubt its exact geographical origin. In this sense it is a science immensely superior to linguistics because its evidence can become irrefutable.”
That only concerns uniparental markers, and even those results can still change, when new discoveries are done. Tracing autosomal ancestry components is more of a lottery: despite of methods like qpAdm, every study still gives different results, and there is always a possibility that a more suitable model or ancestry components will be found in the future.
Most importantly: even if genetics was the God’s Own Words, DNA still cannot tell us about language, and genetic continuity cannot prove about linguistic continuity, as I demonstrated in my earlier message. So, even if we know the origin of paternal or maternal haplogroups, we can never say anything about their languages in different moments of time.
That certain haplogroup is in one region at one moment of time associated with certain language can never be taken as a proof that in another region or at another moment of time it was associated with the same language lineage. Only linguistic methods allow us to say anything about languages. Do you understand this?
@ Gio
''Of course you are right about the oldest I-Y4192Y4241 * Y4240 * V7790/S24016+21 SNPs formed 14600 ybp, TMRCA 12100 ybp, but why not Tagliente 2 17000 years ago or Ostuni etc etc. Haplogroup I was among the WHG and probably at least from 17000 years ago also R1b1."
Tagliente, San Teodoro, Continenza, Oriente are all I2-M223, I2a2 by ISOGG 2018
Ostuni are females only, Paglicci was C1.
@Jaako-
What is truly interesting is that geneticists have invaded the field of linguistics, developing theories about the geographical origin of languages using the genetic data at their disposal-And this is obviously very risky because the inability to produce irrefutable evidence causes the different theories to become ideology, that is, each person will choose as valid the theory that best fits their agenda or their political, racial, social interests, etc. ...in short, there can be nothing less scientific than a dogma of faith, which is what certain genetic-linguistic theories have become.
So in my opinion, if we accept the game proposed by geneticists of giving an opinion on linguistic origins, we have to take into account several essential principles
-1-Accept that all prehistoric European societies were patrilineal so that only male markers can serve to trace the origin of a language, ergo the genetic continuity of male markers is proof of linguistic continuity
-2-Reduce the importance of exogamy to a survival strategy of homo sapiens and consider that women, despite being in charge of raising children, did not transmit their language but were culturally incorporated into the host society
If one person does not agree with these basic rules, then all discussions are absurd because no one will ever accept the theories proposed by other people, that is, there will always be people (as you have done in your last comment) who will deny that genetic continuity through the male line serves to establish linguistic continuity due to the possibility that women are the ones who transmit the language to their children, others will say that it is not the uniparentals that must be taken into account but rather the different autosomal components.
That's why I always talk about fairy tales and unproductive discussions. In this sense, that is what you are doing in this and other blogs, ie. developing theories that you will never be able to prove because you will never have written texts to demonstrate that a certain culture spoke a certain language.
I suppose you will agree with me that, in the absence of written texts, your "linguistic results" are too late in time for genetics to be able to resolve the origin of that language.
Or do you think that by studying the linguistic results of historical times (Germanic, Balto-Slavic, Uralic etc....) you will be able to demonstrate what language was spoken by the CWC, BAC, Yamnaya, Unetice or the Urnfielders?.
Without a doubt, an expert like you will be able to determine how a language is inherited, but in the absence of written texts you are subject to a time limit that becomes an obstacle impossible to overcome, even for people as intelligent as you.
@Gio
I am not talking about the ultimate origin of these lineages, because obviously some may have Balkan origin, others Levantine or Anatolian, and even some of them could have their origin in Italy, what I am saying is that all of them are shared by the Balkans and Italy and that at the moment, all of them are older in the Balkans.
@Noted Goth, I think a couple of caveats for me, not defending their presentation of their model as it is but in some limited sense offering explanations:
1) Italy Imperial, as a set, in my quick Vahaduo model is still primarily Turkey Barcin farmer derived - https://i.imgur.com/huFAPqd.png. 50% Barcin, 15% Afanasievo, 35% Iran+Levant Neolithic+CHG.
2) So you do pretty much get a position just about within Polish variation on all Vahaduo's PCAs (usign G25 data) from 77% Lithuania_Marvele_Roman+22% Rome Imperial - https://imgur.com/a/QVG1eb2 . It's at the edge, because the vector of shift away from Marvele is not quite EEF rich enough (as we'd intuit from a more realistic sense of the chain of populations) but the synthetic model is touching Polish variation in all the PCAs (and most underlying genetic variance).
A Vahaduo model aiming to hit their synthetic population vs the real Polish population has limited difference in deep components; the real Polish only trading about 2% Barcin_N for Levant_PPNB and being ever so slightly less rich in WHG and Steppe ancestry - https://imgur.com/a/DlFd2fq.
So I think I have some understanding of why this model just about passes (and it's obvious why this should be a better model than e.g. replacing Imperial Rome with Iron Age Spain or something like this; too WHG rich for one), but I really think they should comment that:
1) It's not necessarily historically plausible and Imperial Rome is *only* a proxy for PaleoBalkan populations that we *know* exist from the capture genetic data, and that they only use it are limited by what they can use in their method.
2) They are limited in their models power to detect differences in flow from outside Europe based on the qpAdm setup they have. I personally think they should correct this in their models, but at least they should note it, if not.
@Арсен
The Cernavoda culture is more plausible.
@Tea
Many of these people are high-end statisticians, and much better at it than me. So stats aren't really the issue.
In fact, I would say that population genetics is probably too statistically driven and not organic enough.
@ Rob
Tagliente M223?
In Yfull tree:
I-M223CTS10125/S2441 * DFZ77/Z77 * CTS11545+43 SNPs formed 17400 ybp, TMRCA 15100 ybpinfo
I-M223*
I-Y344225Y344263 * Y344289 * Y344420+60 SNPs formed 15100 ybp, TMRCA 12700 ybpinfo
I-Y344225*
id:R7ITAage
I-Y344262Y344559 * Y344262 * Y344269+15 SNPs formed 12700 ybp, TMRCA 12500 ybpinfo
I-Y344262*
id:R15ITA [IT-RM]age
and upstream:
I-S2555*
id:R11ITA [IT-AQ]age [11858 ybp]
id:I4915SRB [RS-14]age [8070 ybp]
Why does Gaska say that Balkans always have older samples than Italy?
Let's look at the Discovery tree of FTDNA. Ahahah, I am seeing "e 2", probably for Tagliente 2, out of the field and older than I-S2555*. Ahahhhh
@Jaako said "but from the timeframe of the expansion of the extant language families we do have many kinds of linguistic evidence"
Not only the Paleolithic languages, but also the neolithic, chalcolithic, and most of those from the Bronze and Iron ages. In the absence of written records everything is speculation. The example of the Saami people is valid but you are talking about historical times, you simply have no idea what language or languages were spoken in Finland in the chalcolithic. If you know it and can prove it, you deserve the Nobel Prize.
@Jaako said.In most cases? Absolutely wrong
Yeah in most cases in periods before the Iron Age unless there are written records. That is to say, you will only be able to develop more or less convincing theories about the language that the CWC spoke, but you will never be able to reliably demonstrate it. You are simply participating in a game and doing exactly what you criticize other people for.
@Jaako said- "genetic continuity cannot prove about linguistic continuity, as I demonstrated in my earlier message"
You have not demonstrated anything in your previous message, because IMHO in specific circumstances genetic continuity through the male line can be considered evidence of linguistic continuity.
Iberia is an obvious example because there is genetic continuity since the chalcolithic (R1b-P312>DF27), cultural-archaeological continuity, absolute genetic uniformity (because we have 300 samples from the Bronze and Iron Age, 294 are R1b-M269 and 6 I2a) and absence of migrations that reached the territory in which Iberian was spoken during the Iron Age. So although we cannot prove it either because we do not have writings from the Bell Beaker culture, common sense makes us think that in chalcolitic Df27, spoke the same language as in the Iron Age.
Do you understand? Maybe you have a better theory and know what language the BBC spoke.
@ Arsen
RE: Anatolian , read the comments of Ethan, Epoch & myself
@Matt
Even though Italy Imperial is mainly "Turkey Barcin" in terms of distal ancestry, the problem is that it's also largely of recent Middle Eastern origin.
So any European population with a significant input from Italy Imperial will show significant recent Middle Eastern ancestry, and possibly even some African admixture, both in terms of genome-wide ancestry and uniparental markers.
But we don't see that in the Medieval Poles or present-day ethnic Poles.
Also, if there were migrations of Italian and/or Balkan groups into Poland during the Medieval period, then this would plausibly be reflected in the shape of the Polish Medieval cluster. That's because recent admixture from coherent groups usually leaves clines in PCA.
But again, that's not what we see, and instead all we have are some outliers that are relatively more southern and western compared to the main Polish Medieval cluster. To me this looks like Isolation-by-Distance captured in real time, and it's one of the main reasons for the supposed Italy Imperial-related signal.
There are also some other things to consider, like the fact that models based on formal stats favor mixture sources from the extremes of mixture clines, but I'll lay it all out in a couple of days in a new blog post.
Matt, David
In my opinion, a good clue are the Dacians, who migrated to Poland under the pressure of Rome, which is well documented in archaeological material. Especially since, in Ralph's study, Poles share a lot of Iron Age ancestors with Balkan populations.
Poles share a lot of Iron Age ancestors with Balkan populations because many Balkan populations are Slavic speaking, and their shared ancestors with Poles lived during the Iron Age and earlier.
Dacians might explain some of the southern shift in northern Slavs relative to Balts. But it's hard to be so specific.
Gaska:
“So in my opinion, if we accept the game proposed by geneticists of giving an opinion on linguistic origins, we have to take into account several essential principles
-1-Accept that all prehistoric European societies were patrilineal so that only male markers can serve to trace the origin of a language, ergo the genetic continuity of male markers is proof of linguistic continuity”
It depends on a case from which parent the language is learned, so it would not be realistic nor credible to claim that it always comes from the father.
Gaska:
“-2-Reduce the importance of exogamy to a survival strategy of homo sapiens and consider that women, despite being in charge of raising children, did not transmit their language but were culturally incorporated into the host society”
Again, conditions vary from case to case, and no exceptionless rules can be claimed.
Gaska:
“If one person does not agree with these basic rules, then all discussions are absurd because no one will ever accept the theories proposed by other people, that is, there will always be people (as you have done in your last comment) who will deny that genetic continuity through the male line serves to establish linguistic continuity due to the possibility that women are the ones who transmit the language to their children, others will say that it is not the uniparentals that must be taken into account but rather the different autosomal components.”
You certainly understand that only such rules should be accepted which are true and realistic. If we accept false rules, then we form a cult which has nothing to do with science. No matter how hard we believe, it cannot make our results credible and scientific, because our premises are false.
Gaska:
“That's why I always talk about fairy tales and unproductive discussions. In this sense, that is what you are doing in this and other blogs, ie. developing theories that you will never be able to prove because you will never have written texts to demonstrate that a certain culture spoke a certain language.”
Fairy tale is to claim that genetic continuity could testify for linguistic continuity, when everybody know that it is not true in this reality. Only the members of the cult believe in such an unscientific method.
I already explained to you about the linguistic methods and results: it is a network formed on the basis of all languages in contact with each other. Therefore it is enough that some languages were written early – other languages can borrow the absolute chronology via the loanword layers.
Naturally, the more distant time, the less certain are the linguistic results, but since the Late Bronze Age we have already many strong results (see below). The linguistic results in any case win the genetic results 100–0 when considering language, no matter what (unless you belong to the cult of Rob).
Gaska:
“I suppose you will agree with me that, in the absence of written texts, your "linguistic results" are too late in time for genetics to be able to resolve the origin of that language.”
What do you mean by this? Seems somewhat internally contradictory sentence.
Gaska:
“Or do you think that by studying the linguistic results of historical times (Germanic, Balto-Slavic, Uralic etc....) you will be able to demonstrate what language was spoken by the CWC, BAC, Yamnaya, Unetice or the Urnfielders?.”
No, but we can study the linguistic situation long before written texts. Take Late Proto-Indo-Iranian for example: phonologically regular vocabulary in Indic and Iranian concerning chariots and certain ritualistic practices are compared to dated archaeological remains of chariots, graves and ritual places. Therefore we know that Late Proto-Indo-Iranian was spoken within the Sintashta Culture around 2000 BCE. Written records are later, but they agree well with this: archaic Mitanni Indo-Aryan was written already ca. 1400 BCE.
Gaska:
“Without a doubt, an expert like you will be able to determine how a language is inherited, but in the absence of written texts you are subject to a time limit that becomes an obstacle impossible to overcome, even for people as intelligent as you.”
Unfortunately you still know nothing about the methods of historical linguistics. We have methods to study language BEFORE written records, see the example above.
Gaska, I am sure you can read and write my name correctly, if you try. Only Rup has no capacity to do that.
Gaska:
“Not only the Paleolithic languages, but also the neolithic, chalcolithic, and most of those from the Bronze and Iron ages. In the absence of written records everything is speculation. The example of the Saami people is valid but you are talking about historical times, you simply have no idea what language or languages were spoken in Finland in the chalcolithic. If you know it and can prove it, you deserve the Nobel Prize.”
In Northern Europe Chalcolithic is not usually in use, but Late Neolithic continues to ca. 2000 BCE. But I agree: we cannot have certain knowledge about so early languages, but that is only because there is no linguistic continuity in the region from that time to the historical times. In such cases in which linguistics cannot help us, we can indeed make guesses based on archaeological or genetic results. For example, it has been proposed already long time ago that the people who brought the Corded Ware Culture to Southwestern Finland were probably Indo-European speakers.
Gaska:
“Yeah in most cases in periods before the Iron Age unless there are written records. That is to say, you will only be able to develop more or less convincing theories about the language that the CWC spoke, but you will never be able to reliably demonstrate it. You are simply participating in a game and doing exactly what you criticize other people for.”
No, you misunderstood that. I only criticize when people ignore the linguistic results and claim to know language better on the basis of genetic or archaeological results alone. But when there are no linguistic results to lean on, then we can make guesses based on genetic or archaeological results.
Gaska:
“You have not demonstrated anything in your previous message, because IMHO in specific circumstances genetic continuity through the male line can be considered evidence of linguistic continuity.”
How could you know when those specific circumstances are fulfilled? You cannot know that. There is always a chance for language shift, unless the population lives in a remote island without any contacts to any other population. So that is the only circumstance which would allow the interpretation you want: a remote island and no external contacts.
Gaska:
“Iberia is an obvious example because there is genetic continuity since the chalcolithic (R1b-P312>DF27), cultural-archaeological continuity, absolute genetic uniformity (because we have 300 samples from the Bronze and Iron Age, 294 are R1b-M269 and 6 I2a) and absence of migrations that reached the territory in which Iberian was spoken during the Iron Age. So although we cannot prove it either because we do not have writings from the Bell Beaker culture, common sense makes us think that in chalcolitic Df27, spoke the same language as in the Iron Age.
Do you understand? Maybe you have a better theory and know what language the BBC spoke.”
Genetic continuity cannot prove linguistic continuity, unless there is a total lack of external contacts like in the remote island. That is the only acceptable law for genetic continuity.
Think about the whole wide Bell Beaker Region – coming to the historical era, there are already many different languages spoken in the former Bell Beaker region: Celtic, Italic, Greek, Illyrian, Basque, Iberian, Baltic, Slavic, Germanic… You can see the genetic continuity to some degree in all of these regions, right? Still only one or two language branches can descend from the original Bell Beaker language (the most popular seemingly being Italo-Celtic).
@Davidski, I didn't say anything to confirm their model was correct, and in fact had caveats and doubts about it representing a literal ground truth, and some of the problems have been noted. Just noting that the amount of such components is relatively lower than people might expect in the Imperial Rome, and their model is not as incompatible with broad genetic affinities, allele frequencies and PCA positions as might be the intuition.
As to whether isolation-by-distance or a pulse model is more compatible to explain changes in ancestry, time series data will help to throw light on such possibilities as it comes to light. It could well be more akin to an isolation by distance model.
How much clines reflecting a direction pulse movement are preserved depends on how assortative reproduction is post any such pulse.
@Jaakko
"No, but we can study the linguistic situation long before written texts. Take Late Proto-Indo-Iranian for example: phonologically regular vocabulary in Indic and Iranian concerning chariots and certain ritualistic practices are compared to dated archaeological remains of chariots, graves and ritual places. Therefore we know that Late Proto-Indo-Iranian was spoken within the Sintashta Culture around 2000 BCE. Written records are later, but they agree well with this: archaic Mitanni Indo-Aryan was written already ca. 1400 BCE."
You see this whole methodology is flawed.
Sure, chariotry appears in the oldest Indo-Aryan texts but only appears in the Younger Avesta. The chariot warrior class is entirely missing from the Older Avesta and was most likely adopted from Indo-Aryans at some later time. Unfortunately, this is usually ignored but Boyce did bring it up in much of her works, "The warrior is usually known as rathaestar "chariot rider", a term evolved evidently after the Iranians had adopted the war-chariot instead of fighting on foot... but the vocabulary of his poetry still reflects an older state of affairs, when probably only the weak and old travelled in heavy ox-drawn carts, and men would have made their way, and fought, on foot(Boyce, 1996).
So, without the chariot constraint you can only guesstimate when Proto-Indo-Iranian was spoken. You really can't reliably date most languages.
Dacians could be a possibility, however, they were mostly reserved to Southern Poland. The vast majority of Medieval Poles were in current Wielkopolskie or Masovia. In other words, most of the populations seems to have been living in Central/ West Poland in the early Medieval era.
The Southern shift could very simply be due to mixing with groups, such as Balkan people, but also possibly Germans who were less Eastern and Northern shifted. It's all speculation for now, really.
In the end, the population that caused the Southern shift is largely trivial and is unlikely to be one group, but rather Poles simply mixing with Europeans who had a sizeable portion of it already, whether from the South or the West. I think all of us, even though the paper was a let down, can agree none the less a blog post demonstrating it's issues is still a pro, and hopefully these sorts of silly mistakes won't occur in the future.
Note: All sequenced Y-DNA and mt-DNA on Medieval Poles is shown in Stolarek's earlier paper from last year.
See here: https://genomebiology.biomedcentral.com/articles/10.1186/s13059-023-03013-9
figure: https://genomebiology.biomedcentral.com/articles/10.1186/s13059-023-03013-9/figures/1
@ Gio
Yes, J2b2 moved into Italy c. 2200 bc from western Balkans & ultimatly the Caucasus, I2a1b moved into Italy c. 3000 BC from the Rhone region.
This is all old news, you should look at ancient DNA data instead of pending sequencing modern results from YFull
@ Gaska
Don't waste your time. Jaaski is still struggling with understanding his own life thesis - where Uralic comes from.
I dont think he's a useful resource for Europe. Not until he gains a better understanding of the current evidence base instead of parrotting cherry picked quotes from yesterday's textbooks.
@ Gaska
Yeah the new Hallstatt and La Tene samples from the McColl paper are R-P312, G2, R-U106, I2 so pretty diverse just like Unetice and bronze age central Europe in general
Vara:
You see this whole methodology is flawed.
Sure, chariotry appears in the oldest Indo-Aryan texts but only appears in the Younger Avesta. The chariot warrior class is entirely missing from the Older Avesta and was most likely adopted from Indo-Aryans at some later time. Unfortunately, this is usually ignored but Boyce did bring it up in much of her works, "The warrior is usually known as rathaestar "chariot rider", a term evolved evidently after the Iranians had adopted the war-chariot instead of fighting on foot... but the vocabulary of his poetry still reflects an older state of affairs, when probably only the weak and old travelled in heavy ox-drawn carts, and men would have made their way, and fought, on foot(Boyce, 1996).
So, without the chariot constraint you can only guesstimate when Proto-Indo-Iranian was spoken. You really can't reliably date most languages.”
1. Interestingly, the word ‘charioteer’ is mentioned in Older Avesta.
https://www.academia.edu/106978888/Indo_European_and_Indo_Iranian_Wagon_Terminology_and_the_Date_of_the_Indo_Iranian_Split
2. Even the word ‘chariot’ is phonologically regular in Iranian, so it must be older than the Iranian sound changes. It cannot be borrowed from Indic.
3. Chariot vocabulary is not the only piece of evidence, see here:
https://www.academia.edu/106979217/Fire_and_Water_The_Bronze_Age_of_the_Southern_Urals_and_the_Rigveda_with_Andrey_Epimakhov_
4. Please define “most languages”? We can date many Indo-European and Uralic proto-languages quite reliably.
I doubt there were ever any 'Dacians' in southern Poland, let alone in the Middle Ages, apart from perhaps temporary influence during the times of Burebista & Decebalus.
Worth mentioning that there was also a J2b L283+ sample discovered in the Celtic (western Hallstatt) tumulus at Magdalenenberg, Germany.
@Jaakko
"1. Interestingly, the word ‘charioteer’ is mentioned in Older Avesta.
https://www.academia.edu/106978888/Indo_European_and_Indo_Iranian_Wagon_Terminology_and_the_Date_of_the_Indo_Iranian_Split
2. Even the word ‘chariot’ is phonologically regular in Iranian, so it must be older than the Iranian sound changes. It cannot be borrowed from Indic."
But check this part out:"At the same time, it is conspicuous that we cannot reconstruct
the PIIr. terminology for certain parts of the chariot, especially for
its most essential part, the spoked wheel (including ‘spoke’,
‘felly’, ‘rim`). This may partly be due to the paucity of Old and Middle Iranian texts, but the fact is that Skt. ará- ‘spoke of a wheel’, nemí- ‘rim (of a wheel)’, and paví- ‘metal felly (of a wheel)’ have no Iranian counterparts."
The word "ratha" is mentioned in the Gathas and I already quoted Boyce's. Here again: "Zoroaster himself seems to have made use of a wheeled vehicle in his journeyings; but the vocabulary of his poetry still reflects an older state of affairs, when probably only the weak and old travelled in heavy ox-drawn carts, and men would have made their way, and fought, on foot". So, yeah doubtful that it was a battle chariot.
In any case, the "ratha" terminology in the Gathas ultimately goes back to PIE itself so not really useful in dating the Proto-Indo-Iranian.
"3. Chariot vocabulary is not the only piece of evidence, see here:
https://www.academia.edu/106979217/Fire_and_Water_The_Bronze_Age_of_the_Southern_Urals_and_the_Rigveda_with_Andrey_Epimakhov_"
Just a dubious interpretation of Vedic texts. The Iranian Ahura Apam Napat is the counter to the Varuna and his association with water and fire is related to the firmament which Lubotsky stretches to include some well.There are other interpretations from others like Massimo Vidale, Parpola...etc that point out to Vedic rituals in BMAC or the Allchins and the IVC.
Anyways, assuming this interpretation is correct then it shows that some sort of Indo-Iranian ritual took place there and not the exact age of Proto-Indo-Iranian. In fact, since Lubotsky is basing his interpretation mostly on the Veda why can't it be some Old-Indo-Aryan ?
"4. Please define “most languages”? We can date many Indo-European and Uralic proto-languages quite reliably."
As you can see from above, we really can't say with absolute certainty that Proto-Indo-Iranian dates to 2000CBE even though we have the Avesta and the Veda + Mitanni texts to help us with that. So I imagine it might even be harder to do that for languages that were attested much later than that.
Let's look at Indo-European. There is no agreement on the date even by linguists who assume the same homeland. Neolithic out of Anatolia vs Chalcolithic out of Anatolia, Ivanov and Gamkrelidze Chalcolithic/BA PIE from south of the caucasus vs Heggarty's late neolithic/early chalcolithic...etc.
Historical linguistics is pretty useful but it's not the be all and end all.
@ Rob
“@ Gio Yes, J2b2 moved into Italy c. 2200bc from western Balkans & ultimatly the Caucasus, I2a1b moved into Italy c.3000 BC from the Rhone region.This is all old news, you should look at ancient DNA data instead of pending sequencing modern results from YFull“
I of course take into consideration what you say and invite you to read what I have been saying from all my life, and before writing about genetics I wrote books of poetry, critics and many other disciplines. What have what you say to do with Greeks that pretend that Southern Italy is “Greek”, with Arabs who pretend it is Arab, Albanians who pretend that Italy is Albanian (I think funded by Turks and you should know why), Celts who pretend it is Celt, Jews who pretend it is Jewish (and not the other way around as I have been writing in all my life) etc etc. By my studies I found uniparental markers in my wife from Sicily of probably Norman or German descent, some of my ancestors have a probably Longobard Y, I am always inquiring if my Y (probably from Yamnaya) is in Italy for 3000 years or come with the medieval migrations, and of course I think that its oldest origin is from at least 17000 years in the Alpine region. For these questions we probably are close to a massive destruction of the earth, et pour cause.
@ Gio
''is in Italy for 3000 years or come with the medieval migrations''
You'd need to do high-resolution sequencing for that, and you'll figure it out. You could be Balkan, heaven forbid ;)
''and of course I think that its oldest origin is from at least 17000 years in the Alpine region..''
And how did it get to the Alps ? Did it just emerge from the soil or arrive from somwhere ? If so what is the path and why ?
I’m interested in what the foreign media are saying about the terrorist attack in Moscow, who is suspected and who is declared guilty
@ Rob
''and of course I think that its oldest origin is from at least 17000 years in the Alpine region..''
And how did it get to the Alps ? Did it just emerge from the soil or arrive from somwhere ? If so what is the path and why ?
I have been writing from the beginning of my inquieres that they were the hunter-gatherers of the Siberian corridor, as also the language of theirs, probably the ancestor of IE, was linked to all the Nostratic language and the closer were the Uralic ones. Following Gaska's hypothesis, I thought that they began to speak the Caucasian languages of the Alps as the ancestors of Sardinian and Basque, to me linked with hg I-M26. All said and written for years.
Vara:
“In any case, the "ratha" terminology in the Gathas ultimately goes back to PIE itself so not really useful in dating the Proto-Indo-Iranian.”
Naturally. Therefore linguists concentrate on words, derivations and meanings which are only seen in Indo-Iranian, when they date and locate Proto-Indo-Iranian.
Vara:
“Anyways, assuming this interpretation is correct then it shows that some sort of Indo-Iranian ritual took place there and not the exact age of Proto-Indo-Iranian. In fact, since Lubotsky is basing his interpretation mostly on the Veda why can't it be some Old-Indo-Aryan ?”
The age comes from the natural scientifically datable archaeological remains being referents for this phenomenon. This phenomenon seems to be well documented in both Avestan and Rigveda.
Vara:
“As you can see from above, we really can't say with absolute certainty that Proto-Indo-Iranian dates to 2000CBE even though we have the Avesta and the Veda + Mitanni texts to help us with that. So I imagine it might even be harder to do that for languages that were attested much later than that.”
Not with absolute certainty, but with confident credibility. There really are no serious challenging views. Do you have some different favorite view you would like to share with us?
Vara:
“Let's look at Indo-European. There is no agreement on the date even by linguists who assume the same homeland. Neolithic out of Anatolia vs Chalcolithic out of Anatolia, Ivanov and Gamkrelidze Chalcolithic/BA PIE from south of the caucasus vs Heggarty's late neolithic/early chalcolithic...etc.”
Nowadays we must distinguish between Proto-Indo-Anatolian (Early PIE) and the Core-IE (Late PIE) – naturally these stages get different datings. But we can outright reject all views which claim the Core-IE expansion before ca. 4000 BCE, because the vocabulary with wagons and pastoralist livelihood do not allow such datings.
Methodologically the quantitative computational calculations based only on the number of shared words are much weaker than the qualitative paleo- and archaeolinguistic reconstructions (that is: comparing reconstructed words and meanings to datable extralinguistic reality), because there are many possible explanations for some branches sharing more or less words than expected with other branches. Such a number can never be translated directly to the family tree model, and even less to centuries since the divergence. This restriction of the method is unfortunately often ignored by the computational phylolinguists and other lexicostatistic linguists.
@ Arsen - I don’t think this was Ukraine, nor “ISIS”.…
Rob
The youngest Dacian barrows in Poland date back to the turn of the 5th and 6th centuries AD.
Noble Goth
When in the middle of the Middle Ages, after three millennia of cremation, the Polish population returns to the skeletal ritual, we see essentially the same population from three thousand years ago (approx. 80%), but slightly genetically shifted towards Southern Europe. This should not be surprising, considering the archaeologically documented inflow to Poland in the Iron Age and the early Middle Ages of people from the south, such as Celts, Romans, Dacians, Vlachs and inhabitants of Pontus - Scythians, Sarmatians, Goths and Greeks.
@ Ambron
“
The youngest Dacian barrows in Poland date back to the turn of the 5th and 6th centuries AD ”
I don’t understand what that means . There were no Dacians in existence in 400/ and 500 AD.
@Jaakko said "It has been proposed already long time ago that the people who brought the Corded Ware Culture to Southwestern Finland were probably Indo-European speakers"
Yeah “Probably”
@Jaakko said-"Genetic continuity cannot prove linguistic continuity, unless there is a total lack of external contacts like in the remote island"
You should have said “like in Iberia”
@Jaakko said-"Many different languages spoken in the former Bell Beaker region"
I don't think you have thought about what you have written, in which BBC domain was it spoken, Balto-Slavic Yllirian or Greek?
@Jaakko said-"The most popular seemingly being Italo-Celtic"
If you say so we will have to believe you, apparently you are also a BBC expert.
I said, “in the absence of written texts, your linguistic results are too late in time for genetics to be able to resolve the origin of that language”
It is not at all contradictory, in the absence of written records, your linguistic results will always be very late (historical time), but even if we talk about the Iron Age the genetic markers were already very heterogeneous (example in the Hallstatt and LaTene cultures in France), so even genetics cannot be used to solve the problem.
Rob
They were and lived in southern Poland:
https://naukawpolsce.pl/aktualnosci/news%2C220245%2Ctajemnica-kurhanu-pradawnego-plemieniu-dakow.html
@ Ambron
“ They were and lived in southern Poland:
https://naukawpolsce.pl/aktualnosci/news%2C220245%2Ctajemnica-kurhanu-pradawnego-plemieniu-dakow.html”
This is an issue of knowing basic history rather than posting irrelevant websites by amateurs
Dacians were decimated by Romans. They didn’t even exist in Dacia let alone southern Poland.
Yes it also means the Daco-Roman theory is non viable
@Matt
Do you know if Speidel et al. removed any outliers from the Poland Medieval and Italy Imperial sample sets?
Rob
Yes, yes... Science in Poland - the body of the Ministry of Science and Higher Education and the Institute of Archeology of the Maria Curie-Skłodowska University are amateurs. And Rob knows better, because he is an outstanding professional.
Gaska:
“You should have said “like in Iberia””
If you claim that there were no external contacts in Iberia, you must prove it. Give us archaeological and genetic sources which state that there were absolutely no contacts with neighbors.
Gaska:
“I don't think you have thought about what you have written, in which BBC domain was it spoken, Balto-Slavic Yllirian or Greek?”
Perhaps you are looking at different map? I looked at this:
https://www.eupedia.com/genetics/bell_beaker_phenomenon.shtml
Gaska:
“If you say so we will have to believe you, apparently you are also a BBC expert.”
I said it is most popular. Check here:
https://en.wikipedia.org/wiki/Bell_Beaker_culture#Postulated_linguistic_connections
@Davidski, in all cases they use subsets of individuals who pass a threshold, and then group them into cluster.
Projected onto Vahaduo West Eurasia by PCA: https://imgur.com/a/zET6Bj2
G25 for the subclusters: https://pastebin.com/kG55MAr1
As you can see, most models use the subset who are relatively more shifted towards Europe (I) and no models use the subcluster that sits over the Near East (IV).
The following samples are also included in the following groups:
Poland_Middle_Ages: PCA0124, PCA0167, PCA0194, PCA0340, PCA0344, PCA0354
Lithuania_Marvele_Roman.SG: R10830, R10832, R10836, R10838
Poland_Middle_Ages_o: PCA0143, PCA0148, PCA0157 (this outlier population is not modelled).
@Jaakko
"Therefore linguists concentrate on words, derivations and meanings which are only seen in Indo-Iranian, when they date and locate Proto-Indo-Iranian."
But as per Lubotsky, the words related to the spoke wheeled chariot are missing from Iranian yet it's still reconstructed for Proto-Indo-Iranian.
"This phenomenon seems to be well documented in both Avestan and Rigveda."
No ritual dedicated to Apam Napat was documented in the Avesta hence Lubotsky also focused mostly on the Rigveda. And again he reconstructed it to Proto-Indo-Iranian.
But I guess that's the fault of the linguist(s) not the method itself?
"There really are no serious challenging views. Do you have some different favorite view you would like to share with us?"
I don't know if there are no serious challenges nor does it matter when the main view itself is full of holes. But there are already different dates for Proto-Indo-Iranian and its split, from Heggarty recently to Eric Hamp who treats PII as an early split in a Steppe homeland context. There's also Parpola who I think dates the Proto-Indo-Aryan split to 2400BCE.
If you want the short version of my view; much of what is reconstructed for Proto-Indo-Iranian is only found in Indo-Aryan. Witzel highlights the presence of Iranians in the Rigveda and the presence of Indo-Aryans in the Younger Avesta. To me a lot of the later shared innovations were borrowed by Iranians sometime after the composition of the Gathas much like how they allegedly borrowed the river names in Witzel's old scenario.
"But we can outright reject all views which claim the Core-IE expansion before ca. 4000 BCE, because the vocabulary with wagons and pastoralist livelihood do not allow such datings."
I don't want to repeat every boring argument on the wagon vocabulary whether it was a potter's wheel or not especially when I lean towards PIE being a copper age language. However, the pastoralism argument also depends on where PIE was spoken. If PIE did not originate on the steppe then this 4000(4400?) BCE date also doesn't hold.
"This restriction of the method is unfortunately often ignored by the computational phylolinguists and other lexicostatistic linguists."
I agree. Except Ivanov and Gamkrelidze were not computational phylolinguists and they arrived at a different date and location for the PIE homeland than many who support the steppe homeland theory.
Vara:
“But as per Lubotsky, the words related to the spoke wheeled chariot are missing from Iranian yet it's still reconstructed for Proto-Indo-Iranian.”
False claim. See page 259. You know well that they are not missing from Iranian.
Vara:
“No ritual dedicated to Apam Napat was documented in the Avesta hence Lubotsky also focused mostly on the Rigveda. And again he reconstructed it to Proto-Indo-Iranian.”
Another false claim. See page 269. Why you make these false claims? You cannot make the evidence disappear even if you wanted to.
Vara:
“If you want the short version of my view; much of what is reconstructed for Proto-Indo-Iranian is only found in Indo-Aryan. Witzel highlights the presence of Iranians in the Rigveda and the presence of Indo-Aryans in the Younger Avesta. To me a lot of the later shared innovations were borrowed by Iranians sometime after the composition of the Gathas much like how they allegedly borrowed the river names in Witzel's old scenario.”
Weird that your beliefs prevent you from seeing and reading the evidence which contradicts your beliefs... (See above.)
Vara:
“I agree. Except Ivanov and Gamkrelidze were not computational phylolinguists and they arrived at a different date and location for the PIE homeland than many who support the steppe homeland theory.”
You should always read also the critique and try to compare the presented arguments. That someone is a linguist unfortunately does not make them intelligent nor give them sufficient knowledge about the topic they study. Let the best arguments win. :)
@Jaakko
" False claim. See page 259. You know well that they are not missing from Iranian."
Page 259: "At the same time, it is conspicuous that we cannot reconstruct
the PIIr. terminology for certain parts of the chariot, especially for
its most essential part, the spoked wheel (including ‘spoke’,
‘felly’, ‘rim’). This may partly be due to the paucity of Old and
Middle Iranian texts, but the fact is that Skt. ará- ‘spoke of a
wheel’, nemí- ‘rim (of a wheel)’, and paví- ‘metal felly (of a
wheel)’ have no Iranian counterparts."
You either don't read what you link or are pretty dishonest. Hint: "no Iranian counterparts".
"Another false claim. See page 269. Why you make these false claims? You cannot make the evidence disappear even if you wanted to"
There's nothing about ghee or a well ritual in page 269.
I don't think you understand how this is supposed to work. Here's what your argument boils down to; "fire is worshipped by both Iranians and Indo-Aryans so 500BCE Persepolis is Proto-Indo-Iranian".
Like I said nothing we can use to precisely date Proto-Indo-Iranian. If Lubotsky's interpretation is correct there's nothing stopping it from being an Indo-Aryan ritual rather than Proto-Indo-Iranian.
"Weird that your beliefs prevent you from seeing and reading the evidence which contradicts your beliefs... (See above.)"
Weird that your beliefs have clouded your vision to the point where you can't read the paper you linked.
"You should always read also the critique and try to compare the presented arguments. That someone is a linguist unfortunately does not make them intelligent nor give them sufficient knowledge about the topic they study. Let the best arguments win. :)"
I'm pretty sure Ivanov and Gamkrelidze are pretty knowledgeable as are their opposition. On the other hand, it's quite evident you are working based on preconceived notions.
Going back to the main point, here's a quote by Stefan Zimmer: "It must be stressed, and cannot be said often enough, that whatever date is given for 'PIE,' it is necessarily no more than pure speculation". Speculation doesn't necessarily mean incorrect though.
@ All
How to merge .bed, .bim files into 1240K dataset, help me pls
@Jaakko said-Perhaps you are looking at different map? I looked at this:
https://www.eupedia.com/genetics/bell_beaker_phenomenon.shtml
If you want to have a serious discussion about Bell Beaker culture you should not use maps or information from Eupedia, not only are they outdated but the site is a joke. There are dozens of published works and many much better maps to get an idea of the geographic extent of this culture. If you are interested I can recommend you some because your knowledge is very limited.
@Jaakko said-“I said it is most popular. Check here:
https://en.wikipedia.org/wiki/Bell_Beaker_culture#Postulated_linguistic_connections
You should have copied the exact quote from wikipedia because it is precisely what I have said in my previous comments
“As the Beaker culture left NO written records, all theories regarding the language or languages they spoke remain CONJECTURAL. It has been suggested as a candidate for an early Indo-European culture, or as the origin of the Vasconic substrate. James Mallory (2013) notes that the Beaker culture was associated with a HYPOTHETICAL cluster of Indo-European dialects termed "North-West Indo-European," a cluster which includes the (predecessors of) Celtic, Italic, Germanic and Balto-Slavic branches”.
In other words, conjectures, hypotheses & fairy tales.
@Matt
So they only model the ancestry of the cluster made up of these samples?
Poland_Middle_Ages: PCA0124, PCA0167, PCA0194, PCA0340, PCA0344, PCA0354
Vara:
“You either don't read what you link or are pretty dishonest. Hint: "no Iranian counterparts".”
You wrote: “the words related to the spoke wheeled CHARIOT are missing from Iranian.” That is not true, because there are words for chariot and charioteer. If you only meant words denoting spoked wheel, you should have written so.
Vara:
“There's nothing about ghee or a well ritual in page 269.
There is text about Apam Napat in Avesta.
Vara:
“I don't think you understand how this is supposed to work. Here's what your argument boils down to; "fire is worshipped by both Iranians and Indo-Aryans so 500BCE Persepolis is Proto-Indo-Iranian".”
I already explained it to you: it is Proto-Indo-Iranian when it concerns words descended from Proto-Indo-Iranian into Indic and Iranian. But I see you have chosen your beliefs, and facts cannot involve in that.
@Jaakko said-“If you claim that there were no external contacts in Iberia, you must prove it. Give us archaeological and genetic sources which state that there were absolutely no contacts with neighbors”
Regarding the burden of proof although I am still waiting for Harvard to be able to prove definitively that my lineage has its origin in the Pontic steppes or somewhere in Siberia, I do accept that the one who has to prove genetic continuity in Iberia (and also in Aquitaine) is me.
1-Genetic continuity-The DF27>Z195 marker has been found in the Iberian BB culture, all Iberian Bronze Age cultures including the Argarians and of course the Iberians and Vascones (from Navarra) analyzed to date.
2-Genetic homogeneity-Not only is there continuity in the different branches of DF27 but we are also talking about an amazing genetic homogeneity, because if we except a few Greek settlers from Ampurias, out of 200 Iberian male samples-97% M269>P312, 3%-I2a. If the situation does not change, and we find some surprises (I have seen unpublished results and it seems that this is not going to happen) in Spain we only start to see genetic diversity with the Romans and Goths.
3-Of course there were external contacts, for example the urnfielders entered Iberia in 1.100 BCE, but they only settled in the Ebro valley and in the Castilian plateau, never in Iberian territory (two thirds of the peninsula at the arrival of the Romans)-There was also exogamy (British Isles, Central Europe, Sicily) and foreign cultural customs were adopted (e.g. pithoi burials of the El Argar culture). The truth is that these cultural contacts did not affect the genetic composition (on the male line) of the Iberians (for example in the culture of El Argar, all the men of the Almoloya, ruling class and peasants were P312, and we only found one Argarian E1b-L618 in La Bastida)
Now you are the one who should explain with linguistic or archaeological arguments why the Iberians being overwhelmingly DF27 spoke a non-Indo-European language, do you have any theory about it?-
I imagine what your answer will be-THE language is not transmitted through the male line but the female line…., the Iberian cultures were matriarchal…., the small percentage of steppe ancestry detected is sufficient proof that the Iberians originally spoke IE…, it does not matter that mass migrations did not occur, it was a small elite of warriors that imposed their language on the rest of the population etc etc
All desperate arguments that we have been hearing for years and that are impossible to demonstrate using one of your favorite phrases in a “scientific way.”
But when you do this, remember that those same arguments can be used against any genetic-linguistic theory that is proposed (including yours) and that therefore at the end of the day everyone will believe what they want to believe regardless of the scientific results obtained and the discussion will have once again been a waste of time.
Target: Turkmenistan_Gonur_BA_2:I2123
Distance: 2.2637% / 0.02263731
73.6 Turkmenistan_C_Geoksyur
26.4 SAHG
Is this the closest sample we have to central Asian Farmer + aasi profile? 10% extra Iran_GanjDareh_N is required but without much change in distance.
Other IVCp samples need excess Iran_GanjDareh_N over the TKM_C pointing to a complex history of Indian archaeogenetics where; somewhere between 10000-5000 bce IranN entered India through Mehargarh. Slowly expanding mixing with AASI. And between 5000-3500 bce Central Asian farmers mixing with the mehargarh and Birhanna population to form IVC population.
Rob
Contrary to Rob's nonsense, all scientific sources unanimously report that the Carpathian Tumuli culture is Dacians.
And the Slavs come biologically from the population of Bronze Age Poland. And that's a fact. And Rob's insults won't change that.
@Davidski, yes, the outlier samples are not included in the model (as I read the Supplementary Data).
These are the samples on Vahaduo PCA btw - https://imgur.com/a/alMwr6z
The six Poland_Middle_Ages are a fairly tight set of folks that sit on the "Balto-Slavic cline" but slightly north of present day Poles. The three outliers that meet their thresholds but do not use in the model are shifted south and NW-Europe-wards.
I assume they only use a small number of samples from the study due to highly limiting quality thresholds...
David
Looking at Figure 3 from Speidel's work, we see that if he had used the Poland BA source instead of the Lithuania IR to model Poland MA, he would have needed less of the Italy Imperial source, or even none at all. Looking at PCA, it should not be ruled out that the one-way model would not be rejected.
I hope you will run some Poland MA model using Poland BA...
TKM_C + SAHG vs SiS_ba2
TKM_C+AASI gives better model than SiS_ba2. Why? Any answers?
Target: Kashmiri_Pandit:BR_Kashmiri_Pandit-1
Distance: 1.2901% / 0.01290100
51.6 Turkmenistan_C_Geoksyur
23.6 SAHG
19.4 Russia_MLBA_Sintashta
5.4 Mongolia_EIA_SlabGrave_1
Target: Kashmiri_Pandit:BR_Kashmiri_Pandit-1
Distance: 2.0227% / 0.02022726
69.4 Iran_ShahrISokhta_BA2
25.4 Russia_MLBA_Sintashta
5.2 Mongolia_EIA_SlabGrave_1
Also models with SiS_ba2 are giving inflated steppe ancestry, why? (Extra 6-9% steppe_mlba)
Edit: add 1000 more years to central Asian agro-pastrolist movements. (Upto 2500bce)
@ambron, they ran 112 separate models which aimed to model Poland_Middle_Ages using Poland_BA (which they divided into three different groups of samples).
Of these, only one obtained a passing p-value, which is the model I reported in the table in my comment upthread ( https://i.imgur.com/eDGhQpa.png ), where they supplemented a model of Italy_Imperial+Lithuania_Marvele with 1% Poland_BA ancestry.
There may be some problems with their model (I've noted my concerns and let's see Davidski's post too) but they certainly tried models, using the samples that met their threshold.
(I don't have a problem with thoughtful criticism of their methods, that should be apprecitaed. But I think in these instances where researchers have just gone out and without any bias just random tested all available permutations that fit their date constraints, then reported the ones that work, it's important not to suggest that they have pre-conceived biases that only led them to test certain models).
@Matt
At the top of page 6 Speidel et al. actually do say that the Italy.Imperial-related ancestry might instead be part of the proto-Slavic package.
"...could be consistent with models of admixture taken place further south and arriving in Poland through north-westerly Slavic expansions."
So they're aware of the possibility that this southern admixture might be very old and very different from Italy.Imperial.
But that still leaves some important points and criticism for me to make.
Matt
I am absolutely not suggesting the authors' bias. I just focus on the technical aspects. And here we know that the authors did not try models with Poland BA (II), which is, according to PCA, the most likely source of Poland MA ancestors.
@ambron, here are all the models they tried using Poland_BA(II).SG as a source:
https://i.imgur.com/FZKcjWu.png
The layout is unusually placed so that they can provide a lot of data produced by qpAdm, but I've separated it using borders to group the output relating to each model. This from Table S3.
Does anybody know how much Frankish autosomal ancestry do modern French people have?
@ Ambron
“Contrary to Rob's nonsense, all scientific sources unanimously report that the Carpathian Tumuli culture is Dacians”
Nope. The “north carpathian group” and “carpathian kurgan culture” are different entities. Your link describes two different phases, which you haven’t understood
The NCG is the reflection of Przeworsk (Vandals) migrating south . The CKC is indeed described as ‘north Dacian”, but has also been described as proto-Slavic. Whatever the case, it is completely different group to actual Dacians in Dacia, who furthermore didn’t exist after their military aristocracy/ theocracy was destroyed by Rome. Are you too stupid to understand basic chronology ?
“ And the Slavs come biologically from the population of Bronze Age Poland. And that's a fact.”
Your statement is nonsensical because there were several different groups in BA Poland.
You should learn basics rather than making categorical statements founded on your personal fanaticism. Halfwit
This Speidel et al. analysis of Polish population history is much worse than I thought.
I'm just trying to cover everything in a relatively brief, coherent blog post.
@Jaakko
Alright my final comment.
" That is not true, because there are words for chariot and charioteer. If you only meant words denoting spoked wheel, you should have written so."
I already quoted Boyce saying ratha of the Old Avesta is a bull cart.
"There is text about Apam Napat in Avesta."
Are you stupid? Apam Napat is a deity not a ghee ritual.
"I already explained it to you: it is Proto-Indo-Iranian when it concerns words descended from Proto-Indo-Iranian into Indic and Iranian. But I see you have chosen your beliefs, and facts cannot involve in that."
And these words mostly descended from PIE.
Yeah, you have chosen your beliefs. It's pretty clear from your discussions with Gaska and Rob. The lack of critical thinking is something else though.
@Rob
There of course were still Dacian tribes after Trajan conquest. Romans conquered only part of Dacia. Beyond Roman borders still existed Free Dacians (Dakoi prosoroi), Carpi and Costobokoi.
Matt
Thanks for the tips!
I admit, I didn't have time to study the supplementary materials because I'm usually very busy during the week.
I now see that the Poland MA with Poland BA (II) and Italy Imperial models are true and statistically significant, as are those with Lithuania IR and Italy Imperial. As I predicted, Italy Imperial's share is lower in them. They seem better suited to the MDS space and historical realities.
Gaska:
“I do accept that the one who has to prove genetic continuity in Iberia (and also in Aquitaine) is me.”
It is not evidence about genetic continuity that we need here, because to some degree there exists genetic continuity practically everywhere and always. Instead, you should prove total lack of contact, influence, and admixture with neighbors. That would be the only way to claim that there is a linguistic continuity without any chance for language shift.
And total lack of contacts with neighbors is not seen in continents – only in remote islands. So you can save your effort and just admit that there is no total lack of contacts in Iberia. And this you just did:
“Of course there were external contacts - - There was also exogamy”
To conclude: we cannot exclude the possibility of language shifts in Iberia.
@ambron
Those models are barely passable.
But more importantly they're statistical constructs that don't reflect reality.
@Matt
Did they do any models of Poland_Viking?
@ Zardos/ Riverman
''There of course were still Dacian tribes after Trajan conquest. Romans conquered only part of Dacia. Beyond Roman borders still existed Free Dacians (Dakoi prosoroi), Carpi and Costobokoi.''
That's not correct.
Romans conquered all of Dacia - the only Dacia which existed. After the Roman conquest, the Dacians disappeared, because what actually defined them was gone- their men killed or enslaved, and sanctuaries destroyed. And they made sure of this because thats what they did to those who had bitterly opposed them. Even if some people survived here and there, that doesnt change anything, because they were no longer "Dacians'.
New towns were established, undoubetdly from Imperial colonists, incl East Med, as we've seen elswhere. After the 275 AD evacuation, there is a hiatus of settlement within the province, filled in by Sarmatians, then Goths from the East, then Vandals, etc.
There was no ''Dacia'' in the Trans-Carpathian reigons. That idea is a modern invention back-projecting isolated Roman references to 'free Dacians', largely due to lack of criticality and narrratives of 20th century Romanian nationalism. Those being (i) there were homogeneous ancient Dacians conveniently corresponding to the modern borders of modern Greater Romania (ii) there is a continuity between Iron Age (or even bronze age, as you persistently falsely claim on GimpArchiver), Roman times, and medieval to modern. Both contentions are false.
Thracians propper, Getae, TransCarpathian Carpi and Dacians are all different - genetically and culturally. You might not be aware of that, but its certainly the case. You cant; make arguements about Portuguese on the basis of French from Burgundy
@ Davidski Noble Goth
In southeastern Poland there was Dacian influence which could explain the Italian component, as they had previously partially absorbed the Celtic Boii and later by arriving Slavs from the east.
There's no Dacian or any sort of Italian component in these Medieval Poles.
It's all a bunch of crap. I'll explain in my new blog post.
@ Davidski
I think it is present in southeastern Poland but the Dacians were there in the Roman Empire and not in the early Middle Ages because from that time on, the Slavic migration began, the paper should definitely be revised
@Jaakko
Try checking the facts a little before putting out an idiotic answer, I am not your kindergarten teacher. In the absence of mass migrations, conquests, etc., sporadic cultural or genetic contacts, trade, etc., are not capable of producing changes in the language of a society. So genetic continuity and genetic homogeneity have been perfectly demonstrated.
@Jaako said-“we cannot exclude the possibility of language shifts in Iberia”
Yeah, you're right you also forgot that it is very possible that there was an extraterrestrial invasion and that's why the Iberians were half Martians and didn't speak IE.
This is not to deny some Iron/Roman Age Transylvanian groups ventured beyond the Carpathians, they almost certainly. I just making the arguement we cant really call them Dacians, and it is not even sure they account for whatever these signal are.
David
I don't know if you remember that Arza dealt with this problem on his blog:
https://slavicorigins.blogspot.com/2021/05/surplus-eef-ancestry-in-modern-day-slavs.html
Gaska:
“Try checking the facts a little before putting out an idiotic answer, I am not your kindergarten teacher. In the absence of mass migrations, conquests, etc., sporadic cultural or genetic contacts, trade, etc., are not capable of producing changes in the language of a society. So genetic continuity and genetic homogeneity have been perfectly demonstrated.”
My answer was right on spot. You cannot make such false claims, because we know well that language shift is possible if there are cultural contacts – it does not require mass migrations. You just have to accept this fact.
Genetic continuity cannot disprove language shift, because genetic continuity is apparent practically everywhere, and still the linguistic situation is very recent. You just have to accept these rules of reality. There is no point for you to make up your own reality with its own laws and rules.
Languages shifts don't occur like that, not so easily. If you haven't a replacement or at least a good taste of demic introgression, then you need a ruling elite military strong enough PLUS kind of a centralized organisation (administration). Rome was able to do that, Franks have not been spite they lft more demic inut in some places. Gaulish took at least 400 year before to be swept off definitively.
"Norman" Frenchs obtained only a mixed language where Germanic syntax is still dominant even if far to be pure. At most a tight network limited to trade can create a pidgin. Religion iself isn't enough to change in depth the natives languages or to sweap them of the everyday life.
Moesan:
"Languages shifts don't occur like that, not so easily."
You cannot make such a claim, because language shifts really can occur without remarkable genetic or cultural traces. Language shift is a sociologically conditioned process.
"Often, languages that are perceived to be higher-status stabilise or spread at the expense of other languages that are perceived by their own speakers to be lower-status."
https://en.wikipedia.org/wiki/Language_shift
In conclusion: Jaakki the “linguist” still doesn’t have a clue about historical linguistics.
Rob commenting linguistics is like an ant commenting an elephant. Hilarious!
Jaako has more talent at comedy than linguistics and he’s not even funny, except for when he rights articles on Uralic languages.
I do think there is a substantial possibility that the East Scandinavian cluster with Baltic HG ancestry represents not a mass migration but contacts across the Baltic. The authors state that there was a genetic barrier between East and Southern Scandinavia before ~2000bce, the differential HG ancestry could just represent the different admixture networks those regions existed in before merging after the Late Neolithic.
Post a Comment