Population structure in Italy using ancient and modern samples

[Ygorcs]

There is no data on the ancient Italic tribes who morphed into the cities with the Greeks and Romans.

Central and Southern Italy is where the Italics lived, and both regions today are Greek in terms of paternal haplogroup. Not admixture or other charts.

If they are right:

Initially, historical linguists had generally assumed that the various Indo-European languages specific to ancient Italy belonged to a single branch of the family, parallel for example to that of Celtic or Germanic. The founder of this hypothesis is considered Antoine Meillet (1866-1936).

Gray and Atkinson come up by using their Bayesian phylogenetic model that the Italic branch separated from the Germanic branch 5500 years ago, roughly the start of the Bronze Age.

Then I am right. The Celts split the Germanic language speaking people to the north and the Italic speaking speaking to the south, into two different groups. Each developed their own new dialect, and thus language.

29% R1b in Latium; 29% R1b in Campania:
both 18% of J2 and both 11% of G:

https://www.eupedia.com/genetics/italian_dna.shtml

You should not presume that modern frequencies of haplogroups in specific regions are representative of how they were 2000 ago, far less that they are representative of the even much earlier population that brought the language that would eventually thrive in that region. That's bound to lead to disappointing conclusions. Consider the genetic drift that is especially fast and dynamic in the Y-DNA haplogroups, the progressive loss of close association between autosomal ancestry and Y-DNA haplogroups, the mixing over the generations, the fact that a language can be and is often established by a socioculturally powerful minority (and that may happen multiple times, for instance: the language of people A is adopted to region of people B with a genetic influx of ~30%; then the people B transmits its language to people C, with a genetic influx of ~20%; then it's passed to people D with a genetic influx of ~20%... in the end the people D will speak the language that was ultimately spoken by the people A, who contributed to just 12% of their genetics - that might well have been the case of Italic peoples between the BA and the modern era)...

These similarities in frequencies of Y-DNA haplogroups may mean something, but they also may be totally coincidental. You shouldn't assume two populations have exactly the same origins based on a similarity of Y-DNA distribution. Even populations that are basically identical can have very different proportions of Y-DNA haplogroups just due to genetic drift. Genetic relatedness has even much more to do with Autosomal DNA, not paternal markers.

Besides, the Gray & Atkinson Bayesian model of IE languages has been heavily criticized in its premises and methodology by a vast array of linguists, who pointed out that some of its more suspicious conclusions (for starters, the dating of the earlier splits itself) could only have happened as a result of a model that was wrong in several of its very basis, treating the evolution of languages exactly as if they were the evolution of biological living beings. Many linguists presented solid counter-evidences and counter-arguments that clearly demonstrate the mistake of trying to "reinvent the wheel" (a popular expression here) of an entire science with some hi-tech "mathematical method" supposed to solve it all using just a few algorithms based on a method devised for biology originally. If you want to understand some of the caveats that render these results unreliable, see here:

https://www.languagesoftheworld.inf...ics/atkinsons-theory-of-language-origins.html

http://literaryashland.org/?p=10433

The assumption that Italic and Germanic are more related is also extremely doubtful from the point of view of historical linguistics. In terms of isoglosses and shared grammatical innovations, Germanic probably shares less with Italic than with several other IE branches (even Balto-Slavic and Indo-Iranian), and most current scholars only see a direct relationship of Italic with Celtic.

 

[Ygorcs]

You failed to read my initial post which compared Latium with Campania, which was the Central and Southern I was comparing. Latium/Rome is best represented in Central. Think of Central Italy, and most people think of Rome. Think of Southern Italy and most people think of Naples. And Campania/Naples is best represented in South. Latium is where the Romans and Latins lived. And Naples Bay was founded by the Greeks (and others). The Rome-Naples-Sicily-Greece connection.

Yes, so? I had already understood that, but that changes nothing in what I'd written. Nobody's denying the relationship between Greek South Italians and Latins especially after the unification of Italy by the Romans.

 

[brick]

There are obvious differences between the people of Lazio and Campania: the first are Central Italians and the second are Southern Italians. It is also showed in the study discussed in this thread.

 

[markod]

I used to be convinced that major Balkanic migrations reached Italy, but IMHO that's untenable considering those leaks. If Greeks and Messapians had a major impact there, modern south Italians and especially the Romans would be more northern and western.

Look at the Adriatic samples we have. They have the same Iberian-Alpine ancestry presumably also found in the Etruscans. The source population for the Italians on the other hand needs to have very significant Levantine and Iranian ancestry components.

 

[Ygorcs]

I used to be convinced that major Balkanic migrations reached Italy, but IMHO that's untenable considering those leaks. If Greeks and Messapians had a major impact there, modern south Italians and especially the Romans would be more northern and western.

How is that so if the Mycenaean Greek and the IA Empuries Greek samples are virtually as "southern" and "eastern" as the Romans in that PCA? I'm sure there was some extra Iranian and Levantine-enriched input into Italy, but I don't think those leaks discard major Balkanic migrations at all. Present-day Greeks and Balkanites as a whole are much more northern and western than those old DNA samples seem to indicate the ancient populations were (except for very northern Balkanic people, basically around Pannonia).

 

[Demetrios]

Wow, this is exciting.

Is this the same paper as the one that RYU reported on? In that one there were two "types" of Romans, yes? One was more "North Italian" like, and one was more "South Italian" like, but by the Imperial period it was definitely more "Southern Italian" like, with a further small change in the post Imperial period.

These look definitely more "South Italian" like. So, this is perhaps a different paper?

Any info on the dating of these samples or the context?

Just assuming for the moment, which I probably shouldn't do, that these samples are all Imperial Era Romans and from a different paper, then I think it just reinforces some of the conclusions we tentatively reached from that prior information.

The "original Romans", from the Republic, were definitely Italic speakers, and were probably more like Northern Italians. As time went on, more and more influence from "Greeks" infiltrated north from Southern Italy. That influence on Italy didn't begin in the first millennium B.C. with Magna Graecia. As I've been saying for ten years, and as recent papers are beginning to conclude, it started back in Mycenaean days.

So, those "more North Italian" Romans of the Republic probably had some of it too, as do modern North Italians. I would guess they were the predictable mixture of Italian MN (also known as Sardinian like) with some steppe admixed migrants, although if Parma Beakers are an indication of the type of admixture we're talking about, they would have varied in the amount of steppe they carried. To that would perhaps have been added a bit of "Mycenaean", carrying a bit of Caucasus/Iran like admixture.

After the incorporation of Magna Graecia in the last centuries of the first millennium BC that would only have increased.

As for the Etruscans, we knew for a long time that their mtDna was like that of most of southern Germany/Northern Italy, i.e. predominantly MN like, so predominantly "farmer" like but with some absorbed U5, either from the WHG, or from the steppe people. I wouldn't presume to judge. Some ancient MtDna experts will have to figure that out.

So, the question has always been, not only what were they like in terms of yDna, but what were they like autosomally. From these leaks, it seems they may have been like Parma Beakers, although which Parma Beaker I don't know. If it's a pretty steppe admixed one, I think we can probably finally put to bed any idea that there was a folk migration from Anatolia to central Italy in the first millennium B.C., an idea which so many have vociferously championed for so long, and which I have resisted for just as long. In the case of the Etruscans we have tons of archaeological evidence, and it just never supported that.

One of the arguments for that very late migration directly from Anatolia has been the "elevated" Caucasus like/Iranian like ancestry in modern Tuscans. What an irony if that came by way of the "Imperial/Classical" Romans, who got it by way of the Greek like people of Southern Italy.

One of the counter arguments has always been that there's a lot of R1b in Tuscans. I've always doubted much of it was "Galiic/Celtic", because other than the northwestern fringe, they really only raided into Tuscany proper, not settled. So, where did the R1b come from? One could say the Romans, but the R1b is unbroken all the way north.

Could it be that the Etruscans, like the Basque, are a case of an R1b but still farmer heavy group mixed with Sardinian like peoples, where, perhaps because it was mostly males by that point, and perhaps the culture was more matrilineal, the children adopted the "farmer" language?

Could there have been a small, elite movement from the Aegean into "Etruria" in the Iron Age? It's possible, I suppose. Y Dna will tell us what happened, although I'm starting to doubt it. Even if one or two samples carry J2, it could have filtered north or been adopted through the long contact between the Etruscans and the Greeks, both directly and through Magna Graecia. We would need a large number of samples.

I know it's unbecoming to say "I told you so", but I have to do it. I took such nonsense over the years from people on dna-forums, where I was virtually excluded, to 23andme forums and even here, where I was constantly harassed, and also saw my ideas ridiculed on theapricity, anthrogenica and by "he who most not be named", .

That's what happens, people, when you follow an agenda, an ideology, instead of looking at all the evidence. Assemble the facts and only the facts, drop all preconceptions and "ologies", and go from there.

@Cato,
I don't know if the more "northern" influence on the Etruscans was Parma Beaker like, or ancient "Ligurian" like, or something else; that's why I said "Parma Beaker like". It definitely seems to be a steppe admixed group to some extent.

You're right; this happened relatively late.

There is a clue about all these that ties to your narrative about Bronze Age/Mycenaean/Aegean migrations into Italy, and specifically Rome, which might shed a little more light, although it could be totally wrong. It's actually part of Roman mythology and has to do with the Arcadian hero Evander of Pallene. I know mythology is not absolute in terms of corroboration in a broader scientific context, but still i find it interesting that it relates with this leaked data. Per Roman tradition he supposedly founded the city of Pallantium on the future site of Rome and specifically on the Palatine Hill, sixty years before the Trojan War. He also instituted the festival of the Lupercalia. Evander was deified after his death and an altar was constructed to him on the Aventine Hill. I have also read of another tradition that presents the Capitoline Hill as originally having been founded by Hercules, although even this might be related to Evander bearing in mind the erection of the Great Altar of Hercules in the Forum Boarium, ascribed to Evander and situated on the plain between the Capitoline, Palatine, and Aventine Hills. Evander was also a protagonist in the Aeneid by Virgil, having been an ally of Aeneas. Do we have any information on the dates of the Roman samples, so we can determine whether this southern-like influence represents proto-Roman or subsequent migrations? Do we know when it's coming out in full?

 

[halfalp]

I used to be convinced that major Balkanic migrations reached Italy, but IMHO that's untenable considering those leaks. If Greeks and Messapians had a major impact there, modern south Italians and especially the Romans would be more northern and western.

Look at the Adriatic samples we have. They have the same Iberian-Alpine ancestry presumably also found in the Etruscans. The source population for the Italians on the other hand needs to have very significant Levantine and Iranian ancestry components.

I'm not sure about major expansion. But it's already a general rule that where there will be Bell Beaker archeological traces, you will have great chances to found Steppe / Balkans ancestry and R1b-P312, and Bell Beaker Culture is the likely only candidate to be related with such ancestry. Everything Anatolia, Levantine or Iranian in Italy probably came from the Sea.

 

[Demetrios]

How is that so if the Mycenaean Greek and the IA Empuries Greek samples are virtually as "southern" and "eastern" as the Romans in that PCA? I'm sure there was some extra Iranian and Levantine-enriched input into Italy, but I don't think those leaks discard major Balkanic migrations at all. Present-day Greeks and Balkanites as a whole are much more northern and western than those old DNA samples seem to indicate the ancient populations were (except for very northern Balkanic people, basically around Pannonia).

When it comes to modern samples, especially the Greek ones, keep in mind that the representations tended to be north-related mostly. Other modern autosomal studies do cluster for example Peloponnesians with Sicilians, therefore respectively close to the Mycenaean samples as well.

 

[markod]

How is that so if the Mycenaean Greek and the IA Empuries Greek samples are virtually as "southern" and "eastern" as the Romans in that PCA? I'm sure there was some extra Iranian and Levantine-enriched input into Italy, but I don't think those leaks discard major Balkanic migrations at all. Present-day Greeks and Balkanites as a whole are much more northern and western than those old DNA samples seem to indicate the ancient populations were (except for very northern Balkanic people, basically around Pannonia).

They are close, but the Romans are more eastern when under the Balkanic hypothesis they should be less eastern. South Italy had well established BA cultures before the Greeks came, so the demic impact of the Greeks would have been diluted at least.

 

[Angela]

Angela, so I suppose you agree with this.

Good thing this isn't Eurogenes. Such wrong comments are rarely read on Eurogenes.

No, I don't agree with it. If you've read any of my posts on the subject at all you would know that. I've been saying for ten years that I didn't think there was a folk migration from Anatolia to "Etruria" in the first millennium BC. The only possibility which remained was "perhaps" some movement of elites. What I have said and believe is that there was immigration to Italy from Greece starting very early, before the major colonization of the first millennium BC and that this kind of ancestry probably tricked slowly up the peninsula. If there's J2 in the Etruscans it could very well have come from that, not any direct migration from Anatolia.

The places where the argument has always been that the Etruscans were formed through a mass migration from Anatolia have been precisely Eurogenes and Anthrogenica, although lately Anthrogenica has modified it somewhat to opt for an "elite" takeover.

I think you've rather got things backwards.

As for the Italics I have never even considered that they weren't steppe admixed.

It's just that here we don't shut down opposing viewpoints. We attempt, where possible, to civilly show people why they are incorrect. If they start insulting other people or posting ultra-nationalist or nordicist propaganda that's another story.

 

[Angela]

I used to be convinced that major Balkanic migrations reached Italy, but IMHO that's untenable considering those leaks. If Greeks and Messapians had a major impact there, modern south Italians and especially the Romans would be more northern and western.

Look at the Adriatic samples we have. They have the same Iberian-Alpine ancestry presumably also found in the Etruscans. The source population for the Italians on the other hand needs to have very significant Levantine and Iranian ancestry components.

There's a big difference between the Illyrians and the Mycenaean like Greeks. I have no problem with an Illyrian or Thracian like population forming part of the ethnogenesis of Italians. Heck, they're some of my highest matches in terms of ancient peoples.

They may even have trickled into Southern Italy.

Greeks are a completely different story. We're not talking about modern mainland Greeks here. We're talking about the Mycenaeans. The Mycenaeans plot near modern Southern Italians/Sicilians, Ashkenazim, and near Cypriots. Did they have a lot of "Armenian" like, South Caucasus like ancestry? I don't know what "a lot" means. What was it? 20%. OK. On Admixture Southern Italians and even modern mainland Greeks have about the same amount of that "component", as I've pointed out numerous times. Levantine? Was it about 5% in the Mycenaeans? I have to go back and check.

I really don't understand your basis for this claim. We have TONS of archaeological and written evidence for ancient Greek migration into southern Italy. We have that kind of ancestry showing up in the Bronze Age.

Now there's suddenly some phantom population which explains things better? Which population? Where is the evidence?

Is this still about the four or five "ancient Romans" who drift toward Cyprus on that leaked PCA? The ancient populations of Crete, Rhodes, and most probably Cyprus as well fed into southern Italy and Sicily. They probably plot in that space. In fact, didn't one of the PCAs above show that?

Ed. Sorry, Ygorcs, we cross-posted.

 

[berun]

it's somehow vintage, but the fourth map provides an idea about the Italian case.

View attachment 11101

 

[Ygorcs]

They are close, but the Romans are more eastern when under the Balkanic hypothesis they should be less eastern. South Italy had well established BA cultures before the Greeks came, so the demic impact of the Greeks would have been diluted at least.

I don't think that's a necessary conclusion. Do we have EBA and MBA samples from Central Italy and South Italy? Do we really know when these IA Roman samples come from and date from? I really don't know. It might be possible that we're dealing with successive waves in Central/South Italy: first a basically Sardinian-like EEF; then heavy Iranian and perhaps Levantine-rich waves; then waves bringing heavily diluted steppe-related input (from BB and later Balkanic IE, i.e. Greek, Messapian, perhaps Liburnian); and finally migrations during the imperial era (could they have brought more Levantines, North African/Punic and Anatolians than we thought?). By the way, I don't think these Romans look significantly more eastern than Cretans, who also have quite a bit of Iranian and Levantine ancestry in my calculations. What if most of the Greek flux came from the Aegean islands or even from West Anatolia? In sum, I have more doubts than answers after all these leaks. ;-P

 

[markod]

@Angela, Ygorcs:

I suppose the Mycenaean samples we have thus far have too much EEF. but if the Greek settlers came from a more exotic place (Asia Minor etc.) that could work, provided the Greeks managed to replace the previous inhabitants. I personally doubt this because to me it looks like the native Fossa culture made up the bulk of the population. We'll see once we get early Iron Age samples.

 

[Jovialis]

There are obvious differences between the people of Lazio and Campania: the first are Central Italians and the second are Southern Italians. It is also showed in the study discussed in this thread.

Why is it that these amateur PCAs seem to violate one another? Especially compared to PCAs from official papers? Brick, I'm pretty sure we have been over this before, with the broad generalizations of using simply "Lazio" by itself as a sample.

 

[Angela]

There's a big difference between the Illyrians and the Mycenaean like Greeks. I have no problem with an Illyrian or Thracian like population forming part of the ethnogenesis of Italians. Heck, they're some of my highest matches in terms of ancient peoples.

They may even have trickled into Southern Italy.

Greeks are a completely different story. We're not talking about modern mainland Greeks here. We're talking about the Mycenaeans. The Mycenaeans plot near modern Southern Italians/Sicilians, Ashkenazim, and near Cypriots. Did they have a lot of "Armenian" like, South Caucasus like ancestry? I don't know what "a lot" means. What was it? 20%.* OK. On Admixture Southern Italians and even modern mainland Greeks have about the same amount of that "component", as I've pointed out numerous times. Levantine? Was it about 5% in the Mycenaeans?* I have to go back and check.

I really don't understand your basis for this claim. We have TONS of archaeological and written evidence for ancient Greek migration into southern Italy. We have that kind of ancestry showing up in the Bronze Age.

Now there's suddenly some phantom population which explains things better? Which population? Where is the evidence?

Is this still about the four or five "ancient Romans" who drift toward Cyprus on that leaked PCA? The ancient populations of Crete, Rhodes, and most probably Cyprus as well fed into southern Italy and Sicily. They probably plot in that space. In fact, didn't one of the PCAs above show that?

Ed. Sorry, Ygorcs, we cross-posted.

I did go back and check. The separate Iran Neo/CHG in Mycenaeans was 18% as a high. They had no Levant Neo. Bronze Age Anatolians did, however, up to 6% in some analyses, and 11 to 14% in others, indicating there was definitely a move north into Anatolia. If Bronze Age Anatolians moved from there to the Greek Island, Greece, and Southern Italy, then they would have taken that ancestry along with them.

When we finally get Classical Era Greek samples, or even better, samples from the earlier parts of the first millennium BC from not only the mainland but the islands, I'd like to see them modeled with Levant Neo as one of the choices.

Apparently there are rumors that some of these ancient "Roman" samples which seem to drift toward Cyprus are from Pompeii. That's at the latest 79 AD, so I think we can say good-bye to all those elaborate fantasies that hordes of Byzantine Levantines and Anatolians moved to southern Italy without leaving a trace in the archaeology or the historical record.

I'd also like to point out that these researchers had better be very sure they're not picking up transient merchants and craftsmen and did some isotope analysis to help in that endeavor. The location of the burials would also be very important.

I've also been reading "Carthage Must Be Destroyed" in light of all the commentary about the Phoenicians/Carthaginians being responsible for "Levantine" ancestry in southern Italy. One of the problems is that in a run upthread, while supposedly Southern Italians like Calabrians have it, Sardinians do not. Yet, Sardinia had a substantial Phoenician/Carthaginian settlement and southern Italy did not.

When I have a chance I'll post some of the pertinent information from the book. I highly recommend it.

 

[davef]

I saw a map and the phoenicians took a small corner in northwest Sicily but that's about it in terms of settling in what is now Italy. It would be strange for them to have gone to the mainland without recording that they did so in the first place. This plus the fact that (stated above) Sardinians don't score levant-neo yet Sardinia had far more Phoenician colonies.

Btw Im not against phoenicians in case someone wants to accuse me of being so. I'm against t-rolls who make assumptions about the origins of certain people without evidence due to hatred against them

 

[markod]

Moreover, why would Roman Anatolians/Levantines have relocated to Apulia/Basilicata etc. where their signal is strongest. It makes no sense.

The people who settled those regions must have thought it preferrable to live as poor mountain shepherds rather than stay in their previous homes.

 

[Jovialis]

Moreover, why would Roman Anatolians/Levantines have relocated to Apulia/Basilicata etc. where their signal is strongest. It makes no sense.

The people who settled those regions must have thought it preferrable to live as poor mountain shepherds rather than stay in their previous homes.

Where are you getting Levantine from? The study says Anatolians BA has a strong signal there.

 

[MOESAN]

What I find fascinating about Italy is that it has a clear break between north and south. It is also seen in the Y-DNA.

R1b is mostly U152 in the north center, E1b1b is mostly E-V13.

Thanks for these pictures - pity the categories are changing 10% by 10% (5% would have been more precise and avoid the impressions of vacuum where it's not the case). for Y-R1b,subclades would have been useful, by instance, when someones argue about Lazio and Campania.