23andMe Share your 23andMe Ancestry Composition

[Riverman]

today I checked other results from users of 23 and me (Tuscans mostly, and I confirm that the story remains the same), so the more I consider the matter the less I understand. I suspect that just as 23andMe has "expanded" this central-southern Italian cluster up to the Po, it has done the same with the French-German towards the south, crossing the Alps, always with a view to what now seems to me a hyper -simplification, perhaps aimed at getting rid of the various “Broadly” components.

Well, everybody complained about the broadly component in the past, some even made 23andme look bad, in comparison to much worse ancestral composition results, because they had high broadly. Yet you can only improve the algorithm between two related populations to assign correctly up to a certain point, since their genetic variation overlaps. At some point you have to decide whether a segment present in both people being assigned to let's say Italian or German, or English or Norwegian. The direction of the gene flow might not even be know and hard to reconstruct, so they decided that if the neighbouring segments are X, the unclear segment should be assigned to X also. Simple as that.
The only way to improve their method is using better references I guess, otherwise what choice do you have? Its a simple question and problem, but the solutions are all complicated and won't be right all the time. There is no easy solution out there, otherwise it would have been long applied by all the companies.

Concerning Italians, you have to consider that even if they score at the same level as Southern French or Northern Spaniards, the reason might not always be exactly the same, because Italians are just more often the result of more extreme component admixture. They have both more extremely Northern and more extremely Southern ancestry.

So if 23andme wants to create an Italian component, they problem is again simple: Either they make it too Northern, then all the Southern Italians will get big amounts of MENA ancestry, or they make it too Southern, then the Northerners will all get a lot of Celtogermanic affinity. It doesn't matter at which end you push, someone will just drop out if a population is that diverse. So they decided for a reasonable compromise, using Central Italians as the reference and therefore both the North and South get some admixture in the results, but not too much, so all Italians should be largely dominated by one generic "Italian" ancestral component.

Think about it, what are the alternatives if they want to please Italian customers?

Concerning the Germanic admixture in Italians, I think it shouldn't be underestimated at all, but only more ancient and historical DNA can help to solve this. Because for a real evaluation, you need to get the right references first.

One aspect of this is, how many German genetic relatives do those Italians with about 1/3 German ancestry have? Could be a hint to more recent, historical migrations too, even after the end of Antiquity.

 

[Angela]

@Riverman
@ Regio X


today I checked other results from users of 23 and me (Tuscans mostly, and I confirm that the story remains the same), so the more I consider the matter the less I understand. I suspect that just as 23andMe has "expanded" this central-southern Italian cluster up to the Po, it has done the same with the French-German towards the south, crossing the Alps, always with a view to what now seems to me a hyper -simplification, perhaps aimed at getting rid of the various “Broadly” components.

The northern / continental genetic influence on northern Italy is undeniable, but in the first place I would attribute the bulk of these genetic contributions to prehistoric / protohistoric periods, and here the major suspect is the expansion of the peoples of the Urnfields culture in the late age of the European bronze. In this context, there is a tendency to overestimate the Lombard influence. Personally, I immediately avoid the positions of the Nordicists and the like who would see Lombards in every canton of Northern Italy but I find those of those who minimize it to the extreme just as laughable.
I believe that the Lombard one was a minority component in absolute numerical terms (although perhaps a little more consistent than the minimum estimates attributed to it), whose alleged genetic trace does not seem completely uniform, but rather discontinuously scattered over a territory that in any case from a demographic point of view it had suffered from the Greco-Gothic war and the Justinian plague. Which is why I wouldn't rule out some local bottleneck phenomenon. (For the same reason, wanting to be picky at the micro-local level, there could have been some fairly significant influences also from other populations similar or related to Goths and Lombards headquartered in the late imperial age). It therefore doesn't seem appropriate to speak of their impact equal to zero, but the general picture, whether we are talking about contributions from the Bronze Age or the early Middle Ages, in the end doesn't change much, because the Northern Italians in all analyzes and PCAs in circulation, however, fall within the cluster of Southern Europeans, somewhat shifted towards the Iberian regions and southern France.

Let's do a more detailed and local analysis. Excluding the Val d'Aosta which has a history of its own and is Italian only administratively, some Occitan valleys in Piedmont or the Alto Adige / South Tyrol, the culturally Italian and / or Italian-speaking northern regions for centuries, in which the transalpine or more specifically "Germanic" is really marked, recent and often also documentable under the historical aspect, it is located above all in the internal Veneto and Friuli (to the point that Friuli becomes in practice a sort of Germanic-Slavic outlier when it is compared with the average of the Italian samples). In Northern Italy, along the Po river, between Lombardy and Emilia-Romagna, autosomal make-up is much less extreme and less unbalanced towards continental Europe. Similarly to the Longobard presence on the same territories, even on the Gallic one there would be a long discourse to open, it is a very anomalous Celtization: the so-called Cisalpine Gauls deal with a strange melting-pot that involves them with previous and more numerous substrates still well present, descendants of proto-Villanovan or Villanovan groups and others. Briefly speaking, we have before us a sort of variegated “koinè” where previous presences of Etruscan or Ligurian extraction stand out, which therefore assimilates something transalpine, without being completely transformed (see in Bologna and surroundings).

This is to say that there is certainly a considerable internal variability in Northern Italy, especially between the Alpine border areas and the Po Valley proper, "more southern" than the other areas, but it is difficult for me to think to the point that its inhabitants suddenly become crypto-Frosinone / Abruzzo / Samnites with some northern nuances. Should we think of such an intense and widespread Romanization of the territory? The case of the Venetian woman of the lagoon is perhaps the most striking of those I have seen: the coastal Veneto / Polesine is certainly different from the internal Veneto and the Dolomite area, because it veers towards the Ferrarese area and is in line with the Po Valley cluster. . We certainly admit an ancient Etruscan component, some ancient Greek / Aegean influence (let's not forget that we are in the immediate vicinity of Adria) that make it more "Mediterranean". But - for God's sake - reaching almost the same values ​​as in Ciociaria or in Terra di Lavoro? If confirmed, there would be tons of more or less recent studies on the peninsula's cline to be thrown away, which instead tell us something quite different.


See the various contributions reported here

https://www.eupedia.com/forum/threa...ure-in-Italy-using-ancient-and-modern-samples


Now I'm absolutely aware that when a reference sample is established, short blankets are created that always keep something uncovered, but here it would be really nice to understand with what criteria 23andMe has developed this gigantic "Italian" cluster that travels practically homogeneous and undisturbed for over 700 km .... Is it actually something whose genetic signal is found more or less in every corner of Italy? the common denominator of all Italians? Is it representative of an area that has suffered less external influences / contaminations? Is it intended to be a reference to the ancient Latium, perhaps the Augustan one between Lazio and Campania, or in any case to that territory most involved in the ethnogenesis of the ancient Romans? Or is it trivially the area where most of the Italian users of 23andMe come from?
In some ways the choice seems even paradoxical to me because - again according to the usual studies of the Italian cline - it is precisely between Central and Southern Italy that the genetic gap becomes more sensitive (while North and Center are more contiguous, with Tuscany that even seems an offshoot of Northern Italy).

There are both advantages and disadvantages to using your own clients as reference samples. The advantage is obvious: lots more samples. The disadvantage is also obvious: the places of origin, given it's not a scientifically chosen random sample, may come disproportionately from one area over another.

I think that's part of what is happening here. There's no recent big swell in the numbers, to my knowledge, of Italians from rural mountain valleys of the north being tested. Imagine a huge 2 D plot. Thousands of samples with Italian ancestry are tossed onto it. Not very many land in the far north. What are they going to choose as the "centrum" for Italians? Clearly, southern Lazio and internal Campania.

So, if you share a good number of those alleles a lot of your genome will be labeled "Italian". What to do with any alleles that might be shared with neighboring "countries"?

Well, do you already have a well defined neighboring cluster?

In samples from some far northern valleys and particularly internal Veneto and Friuli which from what I'm understanding are the areas which undergo a big change, an interesting situation arises where some of these alleles also appear in southwestern Germany or Austria perhaps. Objectively speaking, they could be "Italian" alleles in more northern areas, or more "northern" alleles in Italy. It's a choice. There are undoubtedly more samples from Germany and Austria so that becomes the "centrum" of another cluster.

Everything depends on the choice of the centrum and the samples available. I saw this a bit in my own case when I used a calculator where the creator, for whatever dumb reason, didn't include Northern Italian samples. I suddenly became Bulgarian and then Albanian. Some companies used to group southern Italy and Greece (including its islands) as one entity. When splitting them, at first Greeks had a lot of "Italian". You could have done it the other way as well.

I see this in my own results. Is that 10% "French", which all seems to be from the Cote d'Azur really French? In that particular case, it's far more likely to be "Italian", or at least they are alleles shared by both groups and so could be either one.

Btw, as to the amount of "Longobard" ancestry in Italy I'm on record that it exists but isn't particularly significant, mostly based on the yDna evidence. I stand by it. Where I would expect it to be highest is precisely in the internal Veneto because that was the point of entry and where the largest number of settlements were recorded. I wouldn't have expected it to be high in Venezia and surrounding areas for the reasons you mentioned as well as the fact that population numbers would have been higher there. As the Lombards moved on, it would have been a case more and more of just warriors and even admixed warriors. Look at the Vikings and Scythians.

Friulans are highly admixed with Slavs; a lot more "Italian" in the "French" of southeastern France than in the people of Friuli.

As for some of those mountain valleys, we know some of them were settled late by Bavarians, some of it may be founder effect etc.

I think the most important thing to keep in mind is that 23andme is not doing anthropology. They're trying to give something to their clients in exchange for the all important dna they sell for medical research. What most of their clients want to know is "am I __". They don't want to know about historical processes and gradations within countries.

So, bottom line, do I think this means there is no Italian cline? Absolutely not. That's not what 23andme is investigating. If you want to know about that you're better off with calculators like those Jovialis has put together. You can see the cline there front and center not just in the results but in the PCAs.

Btw, we'd know a lot more if some of these people from the internal Veneto, or Friuli, or the far northern valleys or even regular Lombards from the valley of the Po and the foothills would stop being so shy and would post their results on those calculators. Jovialis could put them on the cline and we could see where x% of more "northern" ancestry lands you. Of course, it might show that the Lombards who aren't from the far mountain valleys are still very "Italian" indeed, despite being from north of the Po. The Po may not have been the barrier it was imagined to be.

 

[Pax Augusta]

@ Stuvanè

Quick comment. I don't think 23andme is investigating the Italian cline as Angela says.

Don't forget that French & German in 23andme still seems to have a peak in the Swiss Alps. It's not really a Northern European component, but more of a Central European component. For some strange reason, Central Europe doesn't seem to exist in 23andme's ethnicity estimate.

 

[Regio X]

Adding up to previous posts, I'd answer the following.

@Riverman
@ Regio X



today I checked other results from users of 23 and me (Tuscans mostly, and I confirm that the story remains the same), so the more I consider the matter the less I understand. I suspect that just as 23andMe has "expanded" this central-southern Italian cluster up to the Po, it has done the same with the French-German towards the south, crossing the Alps, always with a view to what now seems to me a hyper -simplification, perhaps aimed at getting rid of the various “Broadly” components.]

It may have expanded the Italian cluster up, but it didn't do the same with the NW towards South (I see you mentioned F&G though). Some %s have been always there. In our case in specific, it looks like 23andMe re-accommodated most of the Broadly components and also those minor that usually showed up (at very low %s). For example, my mother's NW hasn't changed substantially (it decreased from 42.2 to 41.9); on the other hand, the F&G per se jumped from 34.8 to 41.9, i.e., now all the NW % are F&G. My father's increased a bit though, whereas mine decreased.
I see that you're also referring mostly to the "dissonance" between results from relatively close areas, and to how these results are established. Good question! One of the issues of myOrigins imo is precisely this lack of coherence between results from people with very similar ancestries. I don't think it helps in interpretations.

The northern / continental genetic influence on northern Italy is undeniable, but in the first place I would attribute the bulk of these genetic contributions to prehistoric / protohistoric periods, and here the major suspect is the expansion of the peoples of the Urnfields culture in the late age of the European bronze. In this context, there is a tendency to overestimate the Lombard influence. Personally, I immediately avoid the positions of the Nordicists and the like who would see Lombards in every canton of Northern Italy but I find those of those who minimize it to the extreme just as laughable.
I believe that the Lombard one was a minority component in absolute numerical terms (although perhaps a little more consistent than the minimum estimates attributed to it), whose alleged genetic trace does not seem completely uniform, but rather discontinuously scattered over a territory that in any case from a demographic point of view it had suffered from the Greco-Gothic war and the Justinian plague. Which is why I wouldn't rule out some local bottleneck phenomenon. (For the same reason, wanting to be picky at the micro-local level, there could have been some fairly significant influences also from other populations similar or related to Goths and Lombards headquartered in the late imperial age). It therefore doesn't seem appropriate to speak of their impact equal to zero, but the general picture, whether we are talking about contributions from the Bronze Age or the early Middle Ages, in the end doesn't change much, because the Northern Italians in all analyzes and PCAs in circulation, however, fall within the cluster of Southern Europeans, somewhat shifted towards the Iberian regions and southern France.

Agreed. That's roughly what I was suggesting in this thread:
https://www.eupedia.com/forum/threa...ments-for-Central-Europeans-of-German-Descent

Let's do a more detailed and local analysis. Excluding the Val d'Aosta which has a history of its own and is Italian only administratively, some Occitan valleys in Piedmont or the Alto Adige / South Tyrol, the culturally Italian and / or Italian-speaking northern regions for centuries, in which the transalpine or more specifically "Germanic" is really marked, recent and often also documentable under the historical aspect, it is located above all in the internal Veneto and Friuli (to the point that Friuli becomes in practice a sort of Germanic-Slavic outlier when it is compared with the average of the Italian samples). In Northern Italy, along the Po river, between Lombardy and Emilia-Romagna, autosomal make-up is much less extreme and less unbalanced towards continental Europe. Similarly to the Longobard presence on the same territories, even on the Gallic one there would be a long discourse to open, it is a very anomalous Celtization: the so-called Cisalpine Gauls deal with a strange melting-pot that involves them with previous and more numerous substrates still well present, descendants of proto-Villanovan or Villanovan groups and others. Briefly speaking, we have before us a sort of variegated “koinè” where previous presences of Etruscan or Ligurian extraction stand out, which therefore assimilates something transalpine, without being completely transformed (see in Bologna and surroundings).

This is to say that there is certainly a considerable internal variability in Northern Italy, especially between the Alpine border areas and the Po Valley proper, "more southern" than the other areas, but it is difficult for me to think to the point that its inhabitants suddenly become crypto-Frosinone / Abruzzo / Samnites with some northern nuances. Should we think of such an intense and widespread Romanization of the territory? The case of the Venetian woman of the lagoon is perhaps the most striking of those I have seen: the coastal Veneto / Polesine is certainly different from the internal Veneto and the Dolomite area, because it veers towards the Ferrarese area and is in line with the Po Valley cluster. . We certainly admit an ancient Etruscan component, some ancient Greek / Aegean influence (let's not forget that we are in the immediate vicinity of Adria) that make it more "Mediterranean". But - for God's sake - reaching almost the same values ​​as in Ciociaria or in Terra di Lavoro? If confirmed, there would be tons of more or less recent studies on the peninsula's cline to be thrown away, which instead tell us something quite different.

See the various contributions reported here

https://www.eupedia.com/forum/thread...modern-samples

I believe in some Roman influence, shifting them a bit South(east), followed by some influence from the Germanics, shifting them back a bit North.
And I wonder if there was an additional Southern input to Venezia at the times of Serenissima. Possibly.

As for the internal variability in North Italy, agreed, but it's also true there would be a considerable similarity. And there's also some similarity between N. Italians and, say, Central Europeans or Central Italians, but I think that in genetic sense North Italians are simply... North Italians. As such, they must be distinguishable from Central Europeans for the most part, and now we must mention Ancestry.com. In my opinion it's doing a great job with North Italians, including NE Italians. For example, I know about a guy from the mountains of Udine who scores almost 80% for N. Italy (69% Northern Italy plus 10% Northeast Italy, Croatia & Bosnia and Herzegovina). It seems to reflect the fact he fits better to the North Italian cluster than to any other, or that he's much similar to other North Italians than he is to other pops, in genetic sense. Still in AncestryDNA, "Germanic Europe" in North Italians looks more realistic, apparently more in line with actual Germanic influence into the area.

Now I'm absolutely aware that when a reference sample is established, short blankets are created that always keep something uncovered, but here it would be really nice to understand with what criteria 23andMe has developed this gigantic "Italian" cluster that travels practically homogeneous and undisturbed for over 700 km .... Is it actually something whose genetic signal is found more or less in every corner of Italy? the common denominator of all Italians? Is it representative of an area that has suffered less external influences / contaminations? Is it intended to be a reference to the ancient Latium, perhaps the Augustan one between Lazio and Campania, or in any case to that territory most involved in the ethnogenesis of the ancient Romans? Or is it trivially the area where most of the Italian users of 23andMe come from?
In some ways the choice seems even paradoxical to me because - again according to the usual studies of the Italian cline - it is precisely between Central and Southern Italy that the genetic gap becomes more sensitive (while North and Center are more contiguous, with Tuscany that even seems an offshoot of Northern Italy).

I don't think these results defy the Italian cline concept, if we consider the (commercial) goals of the company, the purpose of identifying very recent ancestries. For example, AncestryDNA does distinguish certain Italian clusters, but they could unify them in one (Italian) cluster if they wanted. For the good or not, 23andMe just has a different approach. We could think on a calculator working with continental level. The fact this calculator is able to distinguish most of, say, European and Native American influences doesn't imply that Europeans are genetically the same, nor Native Americans.
If we think 23andMe should divide Italian cluster into three clusters, as AncestryDNA does, that's ok. What I'm saying is basically that what 23andMe is doing doesn't seem "magic". Now, I agree that it could work better some North Italian groups, make more sense of them, which would demand more reference samples. That seems a consensus here.

Btw...A couple of curiosities related to close relatives:


- I notice that 23andMe greatly raises the share of Italians in Regio X, by at least 10 points compared to his parents. Is it a difference due to updates not yet made on his parents' oracles or what else?

I have 4 siblings, some of them could be more Northern shifted than my parents, some could be in between, and some could be more Southern, as myself. It pretty much depends on which parts of their DNA were inherited. Seeing results from several calculators, it does seem I'm a bit shifted South compared to them, but I think this difference in 23andMe is partly explained by phasing, which would increase accuracy.

@Angela
Despite some of these Venetians are more shifted South or North in 23andMe or other commercial calculators, sometimes substantially, they don't necessarily plot that differently in some PCAs. For example, that woman with 4 grandparents from San Vito di Cadore scores 48.5 % Italian and 50.5% F&G in 23andMe, however, her K15 results are not that different from my mother's. They plot closely at the related PCA. Then she belongs to N. Italian cluster. The neighbor Austrians from Tyrol must be much more Northern shifted than her.

 

[Angela]

If 23andme's French/German chosen "centrum" is the Swiss Alps, with which I would disagree by the way, as it causes a lot of confusion, then a lot of these results make sense, and, in fact, you could do an analysis using that and get higher percents further south into Italy.

Another factor is that given how few Italians from Italy test, I would think a lot of the samples 23andme have with all four ancestors from far northern mountain areas and the internal areas of the Veneto are from people who believe or even have reason to believe that they have ancestry precisely from, say, the Alps on the other side of the border. I highly doubt they are representative of all Lombardia or even the Veneto and, say, Piemonte. These may just be outliers.

That's not to say, I once again state, that there isn't an Italian cline. You're just not going to find it on 23andme.

You can even find it more on ancestry. My time line picture is nothing like that of Salento or Jovialis. If people really want to understand this, why don't some people from Lombardia with these "high" French/German percentages, i.e. more than 25%, post their pictures to aid in our understanding.

Where I actually "am" is sort of where Piacenza, Genova, La Spezia, and Northwest Toscana (i.e. is it Pisa/Livorno?) meet?

 

[torzio]

Adding up to previous posts, I'd answer the following.

It may have expanded the Italian cluster up, but it didn't do the same with the NW towards South (I see you mentioned F&G though). Some %s have been always there. In our case in specific, it looks like 23andMe re-accommodated most of the Broadly components and also those minor that usually showed up (at very low %s). For example, my mother's NW hasn't changed substantially (it decreased from 42.2 to 41.9); on the other hand, the F&G per se jumped from 34.8 to 41.9, i.e., now all the NW % are F&G. My father's increased a bit though, whereas mine decreased.
I see that you're also referring mostly to the "dissonance" between results from relatively close areas, and to how these results are established. Good question! One of the issues of myOrigins imo is precisely this lack of coherence between results from people with very similar ancestries. I don't think it helps in interpretations.

Agreed. That's roughly what I was suggesting in this thread:
https://www.eupedia.com/forum/threa...ments-for-Central-Europeans-of-German-Descent

I believe in some Roman influence, shifting them a bit South(east), followed by some influence from the Germanics, shifting them back a bit North.
And I wonder if there was an additional Southern input to Venezia at the times of Serenissima. Possibly.

As for the internal variability in North Italy, agreed, but it's also true there would be a considerable similarity. And there's also some similarity between N. Italians and, say, Central Europeans or Central Italians, but I think that in genetic sense North Italians are simply... North Italians. As such, they must be distinguishable from Central Europeans for the most part, and now we must mention Ancestry.com. In my opinion it's doing a great job with North Italians, including NE Italians. For example, I know about a guy from the mountains of Udine who scores almost 80% for N. Italy (69% Northern Italy plus 10% Northeast Italy, Croatia & Bosnia and Herzegovina). It seems to reflect the fact he fits better to the North Italian cluster than to any other, or that he's much similar to other North Italians than he is to other pops, in genetic sense. Still in AncestryDNA, "Germanic Europe" in North Italians looks more realistic, apparently more in line with actual Germanic influence into the area.

I don't think these results defy the Italian cline concept, if we consider the (commercial) goals of the company, the purpose of identifying very recent ancestries. For example, AncestryDNA does distinguish certain Italian clusters, but they could unify them in one (Italian) cluster if they wanted. For the good or not, 23andMe just has a different approach. We could think on a calculator working with continental level. The fact this calculator is able to distinguish most of, say, European and Native American influences doesn't imply that Europeans are genetically the same, nor Native Americans.
If we think 23andMe should divide Italian cluster into three clusters, as AncestryDNA does, that's ok. What I'm saying is basically that what 23andMe is doing doesn't seem "magic". Now, I agree that it could work better some North Italian groups, make more sense of them, which would demand more reference samples. That seems a consensus here.

I have 4 siblings, some of them could be more Northern shifted than my parents, some could be in between, and some could be more Southern, as myself. It pretty much depends on which parts of their DNA were inherited. Seeing results from several calculators, it does seem I'm a bit shifted South compared to them, but I think this difference in 23andMe is partly explained by phasing, which would increase accuracy.

@Angela
Despite some of these Venetians are more shifted South or North in 23andMe or other commercial calculators, sometimes substantially, they don't necessarily plot that differently in some PCAs. For example, that woman with 4 grandparents from San Vito di Cadore scores 48.5 % Italian and 50.5% F&G in 23andMe, however, her K15 results are not that different from my mother's. They plot closely at the related PCA. Then she belongs to N. Italian cluster. The neighbor Austrians from Tyrol must be much more Northern shifted than her.

We need to revisit history

Venice began after the Roman empire circa 450AD....from escaping people from Treviso and Padua areas ............they stayed isolated for 600 plus years

The Ostrogoths where the first "germanics" and settled mainly in North Italy with their capital Ravenna
The Lombards entered 80 years later, defeated the Ostrogoths and took over
For Veneto/friuli/ Trentino areas......in the middle ages came the swabians and Ottonian Franks .......settling up Castelfranco type of cities......names of families like Ezzelini, Scalinger, etc

The Lombards, Goths, Vandals, Burgundians etc all came from the same area of the modern German/Polish border......from the baltic sea to basically north of modern Silesia

 

[Angela]

The elephant in the room is the yDna. If the "Germanics" who entered Italy were mostly U-106 and I1, there just weren't that many of them. You could try to add some mtDna but I doubt you could be very precise in saying certain mtDna is "only" Germanic. Maybe you'd just have to factor up the yDna a bit.

Most of "this" ancestry has been in Italy for thousands of years, and that includes the "Gallic" branch. Southern Germans have a lot of "Gallic" in them.

That's why northern Germans in this scheme wind up with significant Scandinavian, and eastern Germans with a lot of Slavic.

There's just been too much admixture over the millennium for people who want "pure" clusters to ever be satisfied.

 

[Pax Augusta]

The elephant in the room is the yDna. If the "Germanics" who entered Italy were mostly U-106 and I1, there just weren't that many of them. You could try to add some mtDna but I doubt you could be very precise in saying certain mtDna is "only" Germanic. Maybe you'd just have to factor up the yDna a bit.

I'm going by memory, but apart from some specific cases (communities with known Germanic migrations dating back to the Middle Ages - such as the Cimbri of Veneto and Trentino, Walser of Piedmont and Aosta Valley, Mocheni of Trentino - isolated communities, founder effect, drift effect, etc. ...) it seems to me that Germanic Y-DNAs in Italy generally do not exceed 5%.


Dal Friuli alla Valle d’Aosta: le minoranze germaniche in Italia (From Friuli to Aosta Valley: Germanic minorities in Italy)

https://www.treccani.it/magazine/lingua_italiana/speciali/minoranze/Caria.html

 

[Angela]

^^That's my recollection as well. So maybe up it a bit for the sake of the mtDna and some stray other yDna, and you still have probably way under 10%.

I don't know why some people just can't accept it. Most of this "Alpine" ancestry has a very ancient presence in Italy.

Btw, I also don't see how such a relatively small intrusion could "return" Northern Italians to what they "were" in the parlance of anthroforums.

Maybe northern Italians changed, of course, with movement from the south of the peninsula, or even actual migrants to places like Luni, but not to the extent that some would think or like to think. I'm not all that far from the Villanovan, or from the more Neolithic like Parma Beaker, closer by far than I am to a lot of "Imperial" samples. When civilizations are destroyed, the people in the cities die like flies. Those who flee to the mountains survive, and when the dust has settled some of them come back down and start all over again. It happened as recently as during the second World War. I grew up on the stories of the exodus of people from La Spezia (many, of course, died in the massive bombing raids which literally flattened the ancient city; so much so that La Spezia is really a 20th century city now) up into the hills of the Lunigiana and then up further into the Apennines themselves. Some died on the way, some ultimately returned to the lowlands. It's an old story.

We just need lots more samples from Bronze Age and early Iron Age Italy north of Rome.

 

[Riverman]

What is Southern German? Its mostly Germanic with Celtic plus some Imperial Roman, Balkan and Slavic ancestry being thrown in. What are Northern Italians in comparison to the Central and Southern ones? They have both more Celtic AND more Germanic ancestry, as well as less of the Greek and more exotic MENA ancestry. Sum all of this up, and you get close to it. Yet its nearly impossible, without a detailed ancient DNA analysis, to tell in detail, when which components came. Like how much historical times Celtogermanic ancestry, first with Germanic tribes, then with Germans during the Holy Roman empire, really came down.
To answer that, you need to make a time transect for various places, like the population before the migration period, during, afterwards, early Medieval and up to modernity. Without such comparison, everything is guesswork, because you can end up at the same point of the PCA for very different reasons and migrations.

On Anthrogenica ph2ter did a R1b-U152 map, weighted and with the FTDNA data, I think this clearly shows the Celtic-Italic and related Celto-Germanic influences in Italy:
https://i.imgur.com/LAUcxCY.png

Add to that the direct, heavily Northern shifted Germanic influences in the migration period and counter that with the more Southern and South-Eastern influences in Southern Italy and you come to the result. But yes, its totally relative, either you give Southern Germans more Italian or Northern Italians more F & G. Many Southern Germans-Austrians have still Italian by the way, usually around 0-6 percent, in some more mixed places like Carinthia it might be more. This is also similar to the AncestryDNA results.

I also wouldn't underesimate the Germanic influence in Southern Italy, without which the shift South East would be way stronger, but this got reduced both by direct admixture as well as later internal Italian and Balkan migrations.

 

[Regio X]

We need to revisit history

Venice began after the Roman empire circa 450AD....from escaping people from Treviso and Padua areas ............they stayed isolated for 600 plus years

The Ostrogoths where the first "germanics" and settled mainly in North Italy with their capital Ravenna
The Lombards entered 80 years later, defeated the Ostrogoths and took over
For Veneto/friuli/ Trentino areas......in the middle ages came the swabians and Ottonian Franks .......settling up Castelfranco type of cities......names of families like Ezzelini, Scalinger, etc

The Lombards, Goths, Vandals, Burgundians etc all came from the same area of the modern German/Polish border......from the baltic sea to basically north of modern Silesia

Which part of my post you're addressing to?

I believe you'd not disagree that there was South-North flow too, before the Germanic. We also descend from Romans.
One of my points was that NE Italians seem to fit pretty well in N. Italian cluster, which is a real one, and part of the Italian cline. The fact that there are internal differences in North Italy, or that North Italians could score a bit differently in certain calculators (for "exotic" pops), doesn't change it, the same way a French scoring for North Italy would not mean the non-existence of a cluster related to France.

F&G are modern, not ancient. I don't think these people perfectly correspond, i.e., moderns don't equal ancient. At the end, the Germanic influence, whatever they are, is already part of N. Italian gene pool since much time ago. I believe we would not disagree on this. Now, I naturally understand that Germanics came, if it is what you're saying. More directly, what I meant is that those relatively high F&G in part of North Italians apparently is not explained mostly by this ancient Germanic influence, i.e., the F&G% wouldn't equal that Germanic input. "Germanic Europe" in AncestryDNA seems to correlate better to the Germanic influence we're discussing, if we assume Germanics did come in somewhat low numbers, as Y-DNA would suggest.

 

[Pax Augusta]

On Anthrogenica ph2ter did a R1b-U152 map, weighted and with the FTDNA data, I think this clearly shows the Celtic-Italic and related Celto-Germanic influences in Italy:
https://i.imgur.com/LAUcxCY.png

I don't find it very accurate and lumping all the subclades of R1b U152 together can be confusing.

These were done very recently by another user, on two specific clades, Z36 and Z56, also based on FTDNA projects and I guess also on some academic studies.

Z36 could be indeed a signal of relationships from the late Bell Beaker period between Switzerland and northern Italy and Z36 may indeed have some relationship with the contacts between northern Italy and Hallstatt and La Tène reaching as far as the north of Tuscany; as Z56 can be really originated in the Bell Beaker period but I disagree here with his assumptions attributing Z56 later to the Ligurians, for example. One must have a deep understanding of the archaeology of prehistoric and protohistoric Italy before making assumptions.

R1b U152 Z36

R1b U152 Z56

 

[Angela]

We may not know "precisely" what percentage of "Germanic" came when without a lot more ancient samples, but if we're going to be objective and follow the dictates of how genetic research is supposed to be done, we have to look at the yDna and that gives us the general parameters of HOW MUCH came, and it isn't very much.

There's not a lot of "Germanic" anywhere in Italy and certainly not in the south. Period.

Anything else is just special pleading for one agenda or another.


The only way of arguing against that is by counting all the U-152 in Italy and labeling that Germanic. It isn't. U-152 is much older and broader than that. It came into Italy in the Bronze Age and perhaps again with Gauls in the first millennium BC. It can't be used to measure "Germanic" impact in Italy for the simple reason that we have the complete dna of the actual "Germanics" who invaded Italy, and they didn't carry U-152. They carried U-106 and I1. Period.

So, the graphic is meaningless to show Germanic influence in Italy. It's also meaningless to show Celto-Germanic influence because the Italics carried it too.

23andme has created a Gallo/Germanic component centered in the Alps. That's their choice, but it's wrong in the sense that we know what "Germanics" looked like, and it's not the Swiss or even the Bavarians. They should do a "Germanic" cluster centered in the North, and two French clusters in France. Then all these results might make more sense, even if the Bavarians and Swiss Germans might wind up not very happy.

By using this Celto/Germanic component they're confusing some Northern Italians into thinking they have far more "German" in them than they actually do. A lot of that ancestry could have been in Northern Italy since the Bronze Age before there "were" any Germans.

And no I'm not going down the rabbit hole with people who just keep special pleading and ignoring unignorable facts in order to peddle their own agendas.

ED. Per Pax' graphics there isn't a hell of a lot of U-152 in Germany anyway, or France either for that matter.

 

[Regio X]

We may not know "precisely" what percentage of "Germanic" came when without a lot more ancient samples, but if we're going to be objective and follow the dictates of how genetic research is supposed to be done, we have to look at the yDna and that gives us the general parameters of HOW MUCH came, and it isn't very much.

There's not a lot of "Germanic" anywhere in Italy and certainly not in the south. Period.

Anything else is just special pleading for one agenda or another.


The only way of arguing against that is by counting all the U-152 in Italy and labeling that Germanic. It isn't. U-152 is much older and broader than that. It came into Italy in the Bronze Age and perhaps again with Gauls in the first millennium BC. It can't be used to measure "Germanic" impact in Italy for the simple reason that we have the complete dna of the actual "Germanics" who invaded Italy, and they didn't carry U-152. They carried U-106 and I1. Period.

So, the graphic is meaningless to show Germanic influence in Italy. It's also meaningless to show Celto-Germanic influence because the Italics carried it too.

23andme has created a Gallo/Germanic component centered in the Alps. That's their choice, but it's wrong in the sense that we know what "Germanics" looked like, and it's not the Swiss or even the Bavarians. They should do a "Germanic" cluster centered in the North, and two French clusters in France. Then all these results might make more sense, even if the Bavarians and Swiss Germans might wind up not very happy.

By using this Celto/Germanic component they're confusing some Northern Italians into thinking they have far more "German" in them than they actually do. A lot of that ancestry could have been in Northern Italy since the Bronze Age before there "were" any Germans.

And no I'm not going down the rabbit hole with people who just keep special pleading and ignoring unignorable facts in order to peddle their own agendas.

ED. Per Pax' graphics there isn't a hell of a lot of U-152 in Germany anyway, or France either for that matter.

I wonder if it would collaterally affect clusters such the Scandinavian.

 

[Angela]

I wonder if it would collaterally affect clusters such the Scandinavian.

I think Northern Germany could perhaps wind up forming part of a larger "Germanic" cluster which includes Scandinavians, at least Danes, Norwegians and most of the Swedes. That might happen to some people like the Dutch as well. The Bavarians and south and south east areas would be more mixed with Gallic, maybe even a little Italian and Slavic.

Eastern Germans would wind up with a "lot" of Slavic, and maybe a few might even land in an East European cluster, which I think is the actual fact, whether they like it or not.

In many ways, not just genetically but in terms of culture and even political behavior, there's a big divide between those Germans you might label Northwestern European like and the East Germans.

Maybe there should be a "Central European" cluster, but it's probably too nebulous. Hungarians have so much "Southern like" German, the Czechs are a mix etc.

Also, I think people forget what a massive "Germanization" campaign was conducted by the German central authorities even in the 1800s during the time of unification. They convinced a lot of people they were "German" when they weren't, or at least when that was only one strand of their ancestry. The French did the same thing to some extent, and we know now there's a big difference between north of the Loire and south of the Loire. Up until the early 1900s I've been told the people of the Midi referred to vacationers from the north as "the French". Anyone who has spent any time in France knows the differences are obvious.

 

[Regio X]

I don't find it very accurate and lumping all the subclades of R1b U152 together can be confusing.

These were done very recently by another user, on two specific clades, Z36 and Z56, also based on FTDNA projects and I guess also on some academic studies.

Z36 could be indeed a signal of relationships from the late Bell Beaker period between Switzerland and northern Italy and Z36 may indeed have some relationship with the contacts between northern Italy and Hallstatt and La Tène reaching as far as the north of Tuscany; as Z56 can be really originated in the Bell Beaker period but I disagree here with his assumptions attributing Z56 later to the Ligurians, for example. One must have a deep understanding of the archaeology of prehistoric and protohistoric Italy before making assumptions.

R1b U152 Z36

R1b U152 Z56

Do you have a map for R-L2? Apparently it has an old presence in peninsula. R851 and R437 (Latin_IA) would have belonged to R-L2, for example.

It would peak in La Spezia.

http://r1b.org/wp-content/uploads/2013/06/U152_Only_Map.png

https://www.eupedia.com/forum/threads/29062-R1b-U152-L2*?p=415833#post415833

 

[Riverman]

I think Northern Germany could perhaps wind up forming part of a larger "Germanic" cluster which includes Scandinavians, at least Danes, Norwegians and most of the Swedes. That might happen to some people like the Dutch as well. The Bavarians and south and south east areas would be more mixed with Gallic, maybe even a little Italian and Slavic.

Eastern Germans would wind up with a "lot" of Slavic, and maybe a few might even land in an East European cluster, which I think is the actual fact, whether they like it or not.

In many ways, not just genetically but in terms of culture and even political behavior, there's a big divide between those Germans you might label Northwestern European like and the East Germans.

Maybe there should be a "Central European" cluster, but it's probably too nebulous. Hungarians have so much "Southern like" German, the Czechs are a mix etc.

Also, I think people forget what a massive "Germanization" campaign was conducted by the German central authorities even in the 1800s during the time of unification. They convinced a lot of people they were "German" when they weren't, or at least when that was only one strand of their ancestry. The French did the same thing to some extent, and we know now there's a big difference between north of the Loire and south of the Loire. Up until the early 1900s I've been told the people of the Midi referred to vacationers from the north as "the French". Anyone who has spent any time in France knows the differences are obvious.

Actually we don't know the exact amount of Germanic admixture in Southern Germany also, because of the same reasons. Because again the same result could be created with a big part being Celtic, and therefore smaller ones being Germanic, Roman and Balkan, or the latter components being larger. The studies on Bavarians and Lombards done so far should caution us in this respect. We will see.
A large portion of the Germanics and later Germans/Franks was in any case not like North Sea Germanics also, but more mixed and with relations to the North.

 

[Regio X]

I think Northern Germany could perhaps wind up forming part of a larger "Germanic" cluster which includes Scandinavians, at least Danes, Norwegians and most of the Swedes. That might happen to some people like the Dutch as well. The Bavarians and south and south east areas would be more mixed with Gallic, maybe even a little Italian and Slavic.

Eastern Germans would wind up with a "lot" of Slavic, and maybe a few might even land in an East European cluster, which I think is the actual fact, whether they like it or not.

In many ways, not just genetically but in terms of culture and even political behavior, there's a big divide between those Germans you might label Northwestern European like and the East Germans.

Maybe there should be a "Central European" cluster, but it's probably too nebulous. Hungarians have so much "Southern like" German, the Czechs are a mix etc.

Also, I think people forget what a massive "Germanization" campaign was conducted by the German central authorities even in the 1800s during the time of unification. They convinced a lot of people they were "German" when they weren't, or at least when that was only one strand of their ancestry. The French did the same thing to some extent, and we know now there's a big difference between north of the Loire and south of the Loire. Up until the early 1900s I've been told the people of the Midi referred to vacationers from the north as "the French". Anyone who has spent any time in France knows the differences are obvious.

That's what I thought. If the Scandinavian cluster were integrated, that could work.

Out of curiosity, I'd like to see how the new version of AncestryDNA deals with these situations.

 

[Riverman]

That's what I thought. If the Scandinavian cluster were integrated, that could work.

Out of curiosity, I'd like to see how the new version of AncestryDNA deals with these situations.

Practically the same. The best representative for German is the South West, Germans which are very North Western get higher English, those very Northern Norwegian and Swedish, the ones to the East get Eastern European and those to the East and South East more often Balkan. The German sample is however wider, so the percentages for EE and Balkan decrease in my opinion rather on Ancestry DNA, while with less smoothing the transition to the North and North West involves higher British and Scandinavian percentages.

I have very little information about Italians on AncestryDNA, but I guess it should be similar, but probably better differentiated because of the larger samples AncestryDNA seems to have?

And again, people should not forget that the core Germanic ancestry was largely restricted to Jastorf originally, the admixture with para-Celtic/old Indoeuropean in the North West block and Central to Eastern Germany started early on, unless we assume a large scale replacement and no significant assimilation of the locals. Archaeology would say otherwise, but only ancient DNA can answer that in detail. The results so far look like admixture was widespread but the Germanic ancestry spread significantly to the South too.

 

[Angela]

Do you have a map for R-L2? Apparently it has an old presence in peninsula. R851 and R437 (Latin_IA) would have belonged to R-L2, for example.

It would peak in La Spezia.

http://r1b.org/wp-content/uploads/2013/06/U152_Only_Map.png

https://www.eupedia.com/forum/threads/29062-R1b-U152-L2*?p=415833#post415833

I still think it's a possibility that L2 was spread beyond Italy by the soldiers of Gallia Cisalpina, although it is possible, yes, that it arrived late with Gauls? Although if that's the case we have to deal with the low percentages in Genova and Savona.

What I don't get is why there are two hotspots like Treviso and La Spezia/Massa and so much lower numbers elsewhere. With La Spezia/Massa many of the current inhabitants have come down from the mountains of places like the Garfagnana where R1b as a whole reaches levels of 70-80% in some places, and there's a "lot" of L2, which could be founder effect typical of mountain refuges. I don't know enough about Treviso to venture an opinion.

So maybe we're getting a slightly distorted view because Boattini didn't have a lot of samples, and there's that mountain founder effect to take into account.

Anyone know if there's a good amount of L2 in Provence?