4
Well, here's my go to model for Western Europeans+new Swiss samples. Every single sample labeled Celtic has their closest distance to a N. Iberian or S. French population. I still use Collegno_o1 to represent Imperial Romans in my go to model over Imperial_Rome because the Imperial_Rome sample was likely too southern shifted for the average Italian colonist, a plurality of samples were Cyprus like, not to mention all the literal fresh off the boat MENA immigrants, South Italian-Greek islander seems more about right to me. Based on K36 results south of the Rhine Dutch are pretty different from all other Dutch and Flemish are pretty different from Walloons, so there's definitely some regional variation there, but those pops aren't on G25.Originally Posted by Creoda
model(same model for all the pops, if you use the model yourself I recommend using DEU_MA as individuals rather than the average, not a good idea comparing averages of pops vs individuals of pops, I was just too lazy):
Spoiler!
Target: English_Cornwall
Distance: 1.2712% / 0.01271192
43.4 Pre-Celtic-Briton/Beaker
29.6 Germanic
27.0 Celtic
(darkest eyed British Isles pop scores the most S. French-like Celt, coincidence?)
Target: English
Distance: 1.1496% / 0.01149641
45.2 Pre-Celtic-Briton/Beaker
34.0 Germanic
19.2 Celtic
1.6 Italic
Target: Irish
Distance: 1.3893% / 0.01389258
72.6 Pre-Celtic-Briton/Beaker
16.4 Germanic
11.0 Celtic
Target: Dutch
Distance: 1.0855% / 0.01085496
71.2 Germanic
15.0 Celtic
13.8 Pre-Celtic-Briton/Beaker
Target: Belgian
Distance: 0.8557% / 0.00855729
35.8 Germanic
24.8 Pre-Celtic-Briton/Beaker
23.6 Celtic
8.6 Italic
7.2 ImperialRoman
Target: French_Nord
Distance: 0.8087% / 0.00808733
34.2 Celtic
28.8 Germanic
21.6 Pre-Celtic-Briton/Beaker
8.6 ImperialRoman
6.8 Italic
Target: Spanish_Castilla_La_Mancha
Distance: 0.9247% / 0.00924716
39.8 Celtic
25.0 Iberia_Central_BA
17.4 ImperialRoman
8.0 Italic
5.0 Mozabite
4.8 Pre-Celtic-Briton/Beaker
Maybe the Empuires samples were even all Celts, it's still silly, imo, to use samples from Iberia instead of from the actual Alps, regardless of age. If anything the older age of the Alpine samples should make them prefer any newer Italic samples but they don't. I do not disagree with late period Alpine Celts being 34-40% steppe. The only thing I disagree with is people with however much % steppe N. French have making up the average of people in the Hallstatt/La Tene period Alps.I don't believe that Celts were 45-50% Steppe, rather I imagine they were exactly midway between Italics/Iberians and pre-Celtic Britons and Scandinavians, that is around 40% Steppe, if that were the case the homeland model for Celtic expansion would work better, although obviously there is no inherent reason why it should.
The "favoring" Italic bit was referring to all the models, despite you using Italic and Celtic sample very close to one another you ended up with Central and Southern French scoring 30+% Italic, that's what I mean by "favoring", for some reason the model prefers Etruscans to the Swiss/Lech samples.
Empuries is basically on the French-Spanish border, if the locals were not Celts I think they can say something about how Southern French Celts looked and to me it would be surprising if Celts/Celtized southern France ended up looking more Southern than those N-E Iberians.
I'll try to look more into the archeology of early Hallstatt to understand better how it expanded, at least we can agree that the early Celtic expansion must have involved a fair amount of assimilation to preserve at least partially the northerness of Northern France and the British isles and to even spread it to Southern France.
Did you consider any other theory on how Celts spread outside the classical Hallstatt-La Tene theory?
My model was just to compare using northern and non-northern samples for Celts, I just wanted to show that both concepts fail at explaining both Northern France and the Alpine region in one single swoop.
I based my claim on ancient components based model, when you use later ones for some reason the non-European adxmiture becomes more evident. I noticed it some time ago with Iberian models already but I'm not sure why G25 works like this.
You are not wrong but the problem with Rome_Imperial it's that it's a bit of an abstract proxy, because we can safely say that within Italy it was an extreme outlier that in effect didn't directly influence all of the territories you showed(especially not Northern Italy) while on the other hand more MENA populations directly influenced the gene pool of northern Roman territories somewhat, I'm not sure how the models would look if we adjust for using iron age North Africans and Anatolians/Levantines.
I don't know much about Y-DNA, but didn't we find E around Bronze age Europe?
I mean how do we explain Empuries1? The 3 non northern samples and BylanyA111 are already 34-38% Steppe and given how modern France looks, having early Celts be 30-35% Steppe implies that there was little replacement insofar as Northern France goes, or maybe Celtic was spread into southern France after the mixing and Celtization of Northern France(EDIT:like you said, I reply non-chronologically, don't mind this entire part).
BTW this leads me to question, were Basque at time basque-like themselves? All Iberian samples are quite a bit less Steppe, although not especially so.
No samples from the Basque area, but as far as I know, Iberia, from the North-East all the way to Portugal, was Basque-like pre-Celtic/Roman invasions, going on PCA plot position. Not just "around Basque", but very specifically Basque. Celts likely shifted them north as Empuires shows but North Africans and Imperial Romans back down to modern Iberians. Southern France clearly got more southern via Imperial Roman admixture, perhaps even modern Spanish, but may also likely have extra northern admixture from post-medieval northern French migrating, etc, which can explain why they still cluster around my hypothesized Celts. It's like N. Italians, they are almost identical to Etruscans but obviously they are not 100% Etruscan, nowhere close to it, despite circumstances of later migrations making them extremely similar.
E has been found in neolithic Europe, nearly all of it in the south-east, nearly all of it the specific clade E-V13, and still in tiny tiny amounts, nowhere close to the rate modern Europeans have it. J has even been found in EHGs and Yamnaya, that does not mean the vast majority of modern European J isn't from later Middle-Eastern migrations.
I believe La Tene areas in N. France and Belgium are dated older to anywhere else in France(but of course oldest are in La Tene/Switzerland) along the Rhine, maybe that supports Alpine Celts moving north first before they spread to southern France/Iberia.
The Celtic speakers that crossed the English straight definitely weren't S. French like, they undoubtably picked up admixture from the Rhine area/NW France area, a region which I do think was N. French-like, if not even more northern similar to England/Scotland MBA/LBA, so while English/Irish may show low amounts of S. French-like admixture(don't forget the case of elite language conquest in Hungary/Finland btw), that wasn't the entire % of the population movement by people who spoke the language.
The Celts were pretty civilized and urbanized for their time, I would be more surprised if they were overwhelmingly Bell Beaker descended rather than heavily neolithic farmer. The Celtic language is literally closer linguistically to Italic than it is to Germanic. Keep in mind back then Latium or anywhere in Italy should really be considered the same thing as northern Italy, Slovenia, etc geographically, the Mediterranean sea used to be a barrier between gene flow, not a conduit. It's not really that far fetched for populations on both sides of a mountain range(albeit big) to be pretty similar(Celts were more northern anyway, again I'm not arguing for the Celt average to be around N. Iberians lol, the distance between Iberians and S. French is pretty big).
There's also one important point I forgot earlier, populations like Czechs, Slovaks, southern Poles show this southern shifted Celt signal. Are we really going to say Czechs and Poles have Italic and Roman admixture now?
oldschool anthropology 
Reply With Quote
Bookmarks