Analysing Iron Age Italian samples

[Angela]

your conclusion then reads that
North_Italian should represent = NorthWest Italy plus lombardy and
N_Italian should be = NorthEast Italy

for the others.......my guess is
O_Italian = Adriatic Italians, inc modern croatia and coastal slovenian lands

TSI = northern Tuscany .................Florence, Pisa and Lucca areas
Tuscan = southern Tuscany .................sienna area

I guess you didn't see my post. How could N_Italian be North East Italy when I just related that two of the five members of that group were Lombards? If the other three were from Northeast Italy, I don't know, but probably at least one was from Friuli-A. Squecco. Perhaps you're another one if you volunteered your data to Dienekes. I don't know about the 5th one.

So, all we could say is mixed North Italian.

[Angela]

As for Crete Armenoi, as I mentioned upthread we indeed don't know her provenance. To repeat, given how many "TSI like" samples pop up all over, she could have been from Italy, the Balkans etc.
Then there's the fact that it's a low quality sample to consider.

I think that to get a clearer picture of affinities, both the Northern Alpine, German Bavarian samples and the Balkan samples should be included in analyses. To use just one or the other may skew the analysis.

[torzio]

I guess you didn't see my post. How could N_Italian be North East Italy when I just related that two of the five members of that group were Lombards. If the other three were from Northeast Italy, I don't know, but probably at least one was from Friuli-A. Squecco. Perhaps you're another one if you volunteered your data to Dienekes. I don't know about the 5th one.

I will speak with Adriano as I have known him for many years ...............he is related via my wife's line somewhere and we are still looking, maybe from Villach southern Austria near italy............last chat December 2019

As for lombards ...........are they all Bergamo based ...........my guess would be Brescia or Cremona ones are different

I tried many years ago ( 6) with Dienekes ...........he was more or appeared to be more interested in O_Italians that came via Vienna lands and beyond

[Angela]

I will speak with Adriano as I have known him for many years ...............he is related via my wife's line somewhere and we are still looking, maybe from Villach southern Austria near italy............last chat December 2019

As for lombards ...........are they all Bergamo based ...........my guess would be Brescia or Cremona ones are different

I tried many years ago ( 6) with Dienekes ...........he was more or appeared to be more interested in O_Italians that came via Vienna lands and beyond

I only remember one of the surnames. It appears to be from just northwest of Milano.
http://www.gens.info/italia/it/turis...1#.XjMI-WhKhPY

[Angela]

Has anyone worked with the Irlach and Alburg Bell Beaker samples from Southern Germany?
https://www.biorxiv.org/content/10.1...944v1.full.pdf

[Dou]

Has anyone worked with the Irlach and Alburg Bell Beaker samples from Southern Germany?https://www.biorxiv.org/content/10.1...944v1.full.pdf

Recently I discovered that one of my great-great-grandparents has this last name apparently of North-Italia:
http://www.gens.info/italia/it/turis...0#.XjMWf2hKjIU

Do you know of any website where you could find information?

[torzio]

I only remember one of the surnames. It appears to be from just northwest of Milano.
http://www.gens.info/italia/it/turis...1#.XjMI-WhKhPY

There are only 11 households in italy with that surname.....6 come from lomvardy , of which Varese is the main town,...especially places clivio and viggiu

[Angela]

Recently I discovered that one of my great-great-grandparents has this last name apparently of North-Italia:
http://www.gens.info/italia/it/turis...0#.XjMWf2hKjIU

Do you know of any website where you could find information?

Cognomix sometimes has information. In this case, Salina does appear. It's sort of a mixed bag, however. There are Lombard, Campanian, Sicilian branches etc. I don't know if all the Lombard Salina's actually have all their ancestry from Lombardia or some are of Southern Italian ancestry who moved north in the great migration which began in the 1950s.

At any rate there's a branch in Messina, one in Campania which might be traced to the Gens Sallia, as are the ones in Lombardia, and which can be found in Varese, Pavia, and Milano.

https://www.cognomix.it/origine-cognome/salina.php

[Ygorcs]

Here is a modeling of the Dodecad K12b populations with ADC 1x, with the latest version of all of the samples I've composed. R437 has a strong presence from South to Center. While Northern_Italian have a stronger affinity to R1016. While N_Italian gets a strong affinity with one of the Balkan Bronze Age samples. O_Italian gets a lot of Protovillanovan, and Tuscan, and TSI seem to get the Armenoi_Crete sample; I think Angela is right that this person probably originally came from Italy.

I have found the results of genetic ancestry models in Vahaduo far less reliable and plausible than using nMonte.R software directly. I'd recommend a bit of caution using it. Maybe that's because unlike nMonte2 it doesn't show a restricted mode in which it'd try to achieve only the best fit at the same as using a more moderate number of reference populations instead of using all possible reference samples to have the best fit possible.

[Angela]

I have found the results of genetic ancestry models in Vahaduo far less reliable and plausible than using nMonte.R software directly. I'd recommend a bit of caution using it. Maybe that's because unlike nMonte2 it doesn't show a restricted mode in which it'd try to achieve only the best fit at the same as using a more moderate number of reference populations instead of using all possible reference samples to have the best fit possible.

I'll let Jovialis speak for himself if he chooses and when he has time, but could you be a bit more specific? In particular, how do you know it's more reliable? What's the objective standard as to whether a certain result is "better" than others?

I'm not trying to be provocative at all. I think we need objective criteria. Present company excepted, of course, it has too often seemed to me that the "modeling" has tended to support the creator's "intuition" or "hypothesis" if "bias" is too negatively loaded a word. It may all even be unconscious.

As to "best fits", I often think that in trying to get really low fits, people throw anything into the hopper, which may not only result in "over-fits", but in historically implausible scenarios.

In terms of using more "moderate" numbers of reference populations, I'm not sure that's such a good idea. Whenever someone is choosing the samples some bias can creep in, or it may just be a case that they don't know of, or have thought of other possibly good sources. If we are interested in exploring the source of the steppe admixture in Italy, and we use only South German steppe admixed samples, and not Balkan steppe admixed samples as well, some important clues might be missed.

Or, take the case of the Parma Beaker samples. If the analysis only used the most steppe admixed sample it might skew the analysis one way; if the one with very little steppe is used it skews it another. Instead, it might be the third which is the better fit. Now, when the samples from a certain site are very homogeneous it wouldn't matter, but in some cases that isn't the situation.

This is why, imo, so much of the modeling has been really wrong, mostly, it's true, when using modern samples, but also even when using ancient samples.

My instinct has always been that the more samples (within reason) the better, and the less human "interference" the better.

[brick]

I have found the results of genetic ancestry models in Vahaduo far less reliable and plausible than using nMonte.R software directly. I'd recommend a bit of caution using it. Maybe that's because unlike nMonte2 it doesn't show a restricted mode in which it'd try to achieve only the best fit at the same as using a more moderate number of reference populations instead of using all possible reference samples to have the best fit possible.



The main cause of implausible results are the models themselves. With wrong models that are not based on ancestral components, one can try to support the less plausible theories.

Just by changing the models a little, the peaks of the ancient samples change a lot.

Dodecad K12b

Italian Bell Beaker + Mycenaeans + Minoans + Crete Armenoi

[brick]

Here is a modeling of the Dodecad K12b populations with ADC 1x, with the latest version of all of the samples I've composed. R437 has a strong presence from South to Center. While Northern_Italian have a stronger affinity to R1016. While N_Italian gets a strong affinity with one of the Balkan Bronze Age samples. O_Italian gets a lot of Protovillanovan, and Tuscan, and TSI seem to get the Armenoi_Crete sample; I think Angela is right that this person probably originally came from Italy.

You've chosen ADD DIST COL 1. It's too much, in my opinion. It inflates some results.

By reducing this value, the results change a bit.


DIST COL 0.5






DIST COL 0.25






DIST COL 0

[Angela]

You've chosen ADD DIST COL 1. It's too much, in my opinion. It inflates some results.

By reducing this value, the results change a bit.


DIST COL 0.5






DIST COL 0.25






DIST COL 0

Your last chart has R850 Iron Age Ardea and then 851 as well.

[brick]

Your last chart has R850 Iron Age Ardea and then 851 as well.

Print Zeroes - No

There are always the same ancient samples in the source, but only in the last chart R851 appears to a sample in the target (North_Italian), so you can't see R851 in the other two charts.


Print Zeroes - Yes

0.5x




0.25x




0

[Jovialis]

I have found the results of genetic ancestry models in Vahaduo far less reliable and plausible than using nMonte.R software directly. I'd recommend a bit of caution using it. Maybe that's because unlike nMonte2 it doesn't show a restricted mode in which it'd try to achieve only the best fit at the same as using a more moderate number of reference populations instead of using all possible reference samples to have the best fit possible.

Though, the models are all dependent on the samples that are being used. I just recently uploaded new coordinates for Ancient Near Eastern populations. I like to personally verify each sample, by downloading them straight from the official source on ebi.ac.uk. Which I convert to usable raw data, all in a consistent format (AncestryDNA). The real fruits of my labor are the raw data notepad files I am producing. Dodecad K12b is just one of the calculators I can utilize for analysis of these files on Admixture Studio. I hope that one day I can use an up-to-date calculator that is based on all of the genetic science on Ancient DNA. Basically one, which uses Ancient DNA sources as the components of which to go by. I haven't used nMonte. R software, but if there is a way for the program to read the raw data files, I could utilize them there too for analysis.

[Ygorcs]

I'll let Jovialis speak for himself if he chooses and when he has time, but could you be a bit more specific? In particular, how do you know it's more reliable? What's the objective standard as to whether a certain result is "better" than others?

I'm not trying to be provocative at all. I think we need objective criteria. Present company excepted, of course, it has too often seemed to me that the "modeling" has tended to support the creator's "intuition" or "hypothesis" if "bias" is too negatively loaded a word. It may all even be unconscious.

Basically I noticed that because, using many possible reference populations in certain models, including populations that are extremely unlikely to have contributed to a certain DNA sample (say, an early Subsaharan African or an early Southeast Asian sample vis à vis a DNA sample from North Italy), the Vahaduo results would give a few percentages to such highly implausible reference populations in their best fit models, whereas using nMonte2 software those really weird results (from a geographical and historical point of view) appear far, far less frequently, and generally using the same reference populations the results given were far more credible, with less numerous, but much closer and more plausible reference populations picked by the software to give a model with a better, but still credible fit.

[Angela]

Interesting. I've yet to see any result like that showing up, but perhaps that's because Jovialis has kept the input to West Eurasian mesolithic samples, when he's included them at all.

Of course, the more tools the better. No one tool is going to be perfect or provide all the answers.

[Ygorcs]

Interesting. I've yet to see any result like that showing up, but perhaps that's because Jovialis has kept the input to West Eurasian mesolithic samples, when he's included them at all.

Of course, the more tools the better. No one tool is going to be perfect or provide all the answers.

Yes, I agree. These models - even those made by professional geneticists and published in their papers - should always be seen as clues to be interpreted in combination with other data and non-genetic evidences, as hints that help us get a clearer picture, and not as some kind of objective received truth.

[Jovialis]

I posted this in the Vahaduo thread, but it is also relevant here. Italian Iron age, Medieval, and Modern samples, using the Dodecad K12b coordinates in Past3 PCA software.

[A. Papadimitriou]

The people from Civitavechia could have had admixture from groups other than the Etruscans and Carthaginians.

In general I am skeptical about ALL the labels, but either way no one should assume that the people from Civitavecchia were similar to proto-Etruscans.

Personally I assume that proto-'Etruscans', as an ethnic group formed inside Tuscany but they descended from people further north.

The ancient sources were pointing to an early movement of Italic speakers AND to other movements from the southeast during the Bronze Age.

Osco-Umbrian samples would be more informative than Latin or 'Latin'.

[Angela]

The people from Civitavechia could have had admixture from groups other than the Etruscans and Carthaginians.

In general I am skeptical about ALL the labels, but either way no one should assume that the people from Civitavecchia were similar to proto-Etruscans.

Personally I assume that proto-'Etruscans', as an ethnic group formed inside Tuscany but they descended from people further north.

The ancient sources were pointing to an early movement of Italic speakers AND to other movements from the southeast during the Bronze Age.

Osco-Umbrian samples would be more informative than Latin or 'Latin'.

I don't understand your terminology. The people in Civitavecchia were Etruscans. There's no reason to assume they were anything else. Now, whether those Etruscans will prove slightly different from Etruscans in other areas I don't know. I think we've seen from other situations that it can take a long time for admixtures to stabilize. Maybe further north they were a bit more "northern".

Regardless, they present as a combination of EEF "like" farmers admixed with a steppe admixed population. To know what their steppe ancestors were like upon arrival in Italy we'd need older samples, and we'd probably find them further north.

As for the North African admixed Etruscan, we don't have enough information to label that ancestry "Carthaginian". It was perhaps a grandparent, so back date the admixture accordingly. Nor should we assume that this was other than a one off.

The fact remains that Herodotus was wrong, and everyone who relied so heavily on him was wrong. The ancestors of the Etruscans were the local farmers and steppe admixed people from the north.

Yes, there was most probably Bronze Age migration from the southeast into Italy before the Iron Age, but that wasn't an important element in the Etruscans from what we can see so far. No do we know the size of that migration.

[torzio]

The people from Civitavechia could have had admixture from groups other than the Etruscans and Carthaginians.

In general I am skeptical about ALL the labels, but either way no one should assume that the people from Civitavecchia were similar to proto-Etruscans.

Personally I assume that proto-'Etruscans', as an ethnic group formed inside Tuscany but they descended from people further north.

The ancient sources were pointing to an early movement of Italic speakers AND to other movements from the southeast during the Bronze Age.

Osco-Umbrian samples would be more informative than Latin or 'Latin'.

Carthage and its phoenician settlers only began 810bc.....do you mean these people or earlier

[A. Papadimitriou]

I don't understand your terminology. The people in Civitavecchia were Etruscans. There's no reason to assume they were anything else. Now, whether those Etruscans will prove slightly different from Etruscans in other areas I don't know. I think we've seen from other situations that it can take a long time for admixtures to stabilize. Maybe further north they were a bit more "northern".

Regardless, they present as a combination of EEF "like" farmers admixed with a steppe admixed population. To know what their steppe ancestors were like upon arrival in Italy we'd need older samples, and we'd probably find them further north.

As for the North African admixed Etruscan, we don't have enough information to label that ancestry "Carthaginian". It was perhaps a grandparent, so back date the admixture accordingly. Nor should we assume that this was other than a one off.

The fact remains that Herodotus was wrong, and everyone who relied so heavily on him was wrong. The ancestors of the Etruscans were the local farmers and steppe admixed people from the north.

Yes, there was most probably Bronze Age migration from the southeast into Italy before the Iron Age, but that wasn't an important element in the Etruscans from what we can see so far. No do we know the size of that migration.

Herodotus is misinterpreted AND wrong in that case but either way, I was always in favor of a local or better a Central European origin.

But I place importance on the account of Dionysius of Halikarnassus. As I have written here in the past close reading of the text points to a movement of 'Pelasgians' from the South-East during the Bronze Age (following a route along the Adriatic I think) before the movement of Tyrrhenians-proper from the north, certainly after the so called 'Bronze Age collapse'. // On the other hand, the Osco-Umbrians are perceived as a very old ethnic group in Italy, while in Northern and Central Italy there appear to have been other groups of uncertain linguistic affiliation like the Sicels and the Ligurians too.

I do NOT take ANYTHING written in ancient texts literally, but I read them carefully.

I believe the Pelasgians of the ancient sources were an IE group, at least broadly similar to 'Mycenaeans' genetically (Herodotus places the origin of BOTH 'Pelasgians' and 'Hellenes' in Thessaly). And one speculation that I had made is that classical Etruscans descended patrilineary mainly from the people of the prehistoric pile dwellings around the Alps.

Dionysius believed that they took their name from the WOODEN structures they were making ('high wooden palisades resembling towers'), supposedly the FIRST to make these in Italy. The most important thing about ancient sources is not if they are 100% factually correct.

Maciamo talked about 'Carthaginians'. It was a response to him. I don't know anything about the history of Civitavecchia in particular but I know that there were other people in the region other than those he mentioned.

I am not willing to make a debate on this issue currently.

[Angela]

^^I don't see anything there but speculation upon speculation, and built partly upon another ancient author who was no doubt also just relaying rumor and, in his case, his own biases and his desire to reconcile the Greeks to Roman rule. So, I think it's a good idea for you not to debate it.

To expect ancient authors to be ethnographers or geographers in the sense we know them today is, imo, a very mistaken notion.

[A. Papadimitriou]

^^I don't see anything there but speculation upon speculation, and built partly upon another ancient author who was no doubt also just relaying rumor and, in his case, his own biases and his desire to reconcile the Greeks to Roman rule. So, I think it's a good idea for you not to debate it.

To expect ancient authors to be ethnographers or geographers in the sense we know them today is, imo, a very mistaken notion.

Either way, there were people with a tradition of lake dwellings, stilt houses etc. That is archaeologically supported. These were essentially 'Italici' of Pigorini.

One interesting thing about the account of Dionysius of Halikarnassus, that I noticed myself while reading the PRIMARY source, is that according to him the Tyrrhenians *likely* took their name as a result of them being the FIRST to build "high WOODEN palisades resembling towers" in Italy. The most important thing about that is NOT if that is factually correct.

The speculation of mine is that the Etruscans would descend patrilineary primarily from these people. Essentially Andrea Cardarelli tries to support something similar to what I was thinking without knowing anything about him or his work but he does it in a way that it is bad in my opinion but you can read his arguments.

Maybe what I say is wrong an Pigorini was correct. Or Edoardo Brizio? I am not willing to debate it because I am waiting for more data.