I know you had a theory about Z631 being Celtic (I'm Z631 too), but it's simply a southern Illyrian marker. That's it. It's spread is certainly mostly linked to the Roman Empire. It likely entered Italy with Messapian migrations.
That's a "strong" position considering the data we have.
Thus you say that Z631 is a southern Illyrian marker, when diversity is not in southern Illyria and where the only ancient sample we have is from Italy and present a Celtic admixture. Interesting conclusions.
You say there's "no data". Look in Albania. It's all Z638.
That's what I find exhausting about such topic.
The main arguments that can be used rely on:
--> ancient samples location,
--> ancient admixtures,
--> modern diversity.
and most of the time, you see peoples flipping from one argument to another without much reasons other than showing their interpretation is the good one.
On my side, I stick on diversity analyses (and build my model from a combined analysis a the philogenic tree and the diversity), thus I just verifies that ancient samples/admixtures are not incompatible. I prefer to define my methodology and then apply it to find a conclusion, than flipping methodology to support my favorite theory (which produces well known validation biases).
Logically speaking, Albanian being mostly Z638 is no proof that the haplogroup originated there.
Because there is other region with heavy Z638 diveristy. Then finding from which place Z638 migrate, is not that easy.
Going by that reasoning, R1b whould have originated in western Europe, I think we would agree it is not the case ?
If you find a sample by ~1400BC somewhere, if this sample is linked to a clade appearing around 2200BC (Z2197 for exemple), and if in the same area there is good reasons to think that there has been a L283 expansion around 1800-1700BC (Y15058 & Y38240), this 1400BC is of little help to identify the direction of the migration.
Such sample could be a strong argument, but it need dates to "align" in a more clear way, for exemple: if this sample was a J-Y23094 (defined in 1200 BC, post expansion of Y15058/Y38240) from 1000BC ... then it would be a way stronger argument. But Z2197 (defining ~2200BC) found somewhere in 1400BC where a L283 expansion occured around 1800-1700BC, is nearly informationless about migrations.
Truth to be said, same apply to the Z2507- sample, Z2507 got define around 2400 BC, it is found in 666 BC in an area where Y15058 expansion occured ... thus it can fit either with a northern or a southern origin.
Interestingly, the southern Z638 sample would have been a "stronger" argument if the Y15058 samples didn't showed up. Because with these samples, we now know where the 1700-1800BC expansion occured. Before that, it was hard to validate a pre-1300BC expansion in this area, and thus a 1400BC sample would have had way more impact on our reasoning.
This illustrates how "blind" we are with ancient sample when a single one can modify significantly our view ... which encourage to be carefull with strong and definitive statements, because conclusions based on one sample are easily affeceted by statistical noise.
Beyond that, if someone claims that the southern Z638 is proof of an origin of Z638 there, it also have to consider that Z2507 originated in Slovenia (I guess you agree on that ?).
If not, the methodology varies with what the author try to demonstrates (and we exit the world of sience).
Such reasoning imply a Slovenian --> Albania movement from 2400 to 2200BC. I find that migration direction unlikely at that time.
But if we drop the Slovenian sample argument, we have to drop the Z2197 at the same time.
Interpretations of the current data could fit in many models :
--> Z597 around Slovenia, with a southward 1800-1700BC migration containing Y15058 and some Z638 branches.
--> Z597 in the south, with a migration northward in 1800-1700BC containing Y15058 and Z2507-, and some Z638 branches.
--> Z597 in the eastern Balkan, with Z2197 being left on the way when migrating from lower Danube to central Europe around 2200 BC
.
Yet, the new data are saying nothing to separate those models. But we have modern diversity. And modern diversity seems to favored a northern center of diffusion during EIA.
Diversity of Z631 and Z1043 would have hard time to fit with a Roman-based diffusion from Illyria, because the expansion of this clades predates the Romans. Thus, the Romans would have to come in Albania, mainly gather all Z631/Z1043 branches, but without touching other already diffused clades established here. It didn't sound very convincing to me.
Up to now, I was more align with the thinking that western Balkan got "colonised" by some L283 around 1300-1200BC. There is now good arguments to push this date back to 1800-1700BC. But data are not really saying much more than that.
Then, a second wave of L283 probably arrived around 1300-1200 BC (Y86181 for exemple). By that time, one would expect a north --> south migration instead of the opposite.
It's not a coincidence the southernmost Z638 aDNA marker is from southern Dalmatia, just a few hundred kms from Albania.
If you read me, you'll read that I currently think some Z638 arrived in southern Balkan around ~1800-1700BC (before this paper I would have said ~1300-1200BC, but this paper solves the question of the 1800-1700 BC expansion of Y15058 & Y38240).
That is basically what is favoring the diversity map I presented. In fact, this diversity map didn't says anything about the "arrival" date of Z638. But with the 1400BC sample and the Z2197 mutation, we can braket the arrival between 2200-1400BC, and there is now good evidence for a L283 expansion in the western Balkans around 1800-1700BC.
As I said, you jump on the Z638 sample, and you place it in Albania in 2200BC ... but you completely ignore Z2507- & Z2507 samples, that would place Z2507 in Slovenia in 2400.
Sounds like cherry picking to me. In fact these data are unsufficient to conclude definitely.
If you want to read more about it, read Hunter Provyn's blog (who is also Z631). Google this
First Ancient J-Z631 Found in the Via Salaria Necropolis in Rome
Hunter Provyn is doing a great job ... but what he says reflect his own interpretation. Sometimes I agree, sometimes not.
Note that the "Roman" J-Z631 present a western admixture ... in fact, in term of ethnicity his autosomal DNA look more Celtic than Roman.
Thus yet, from the data I know of, we have:
--> Z631 diversity center in central Europe
--> a ~2000 years old Z631 in Rome with a Celtic-like admixture
Here again we enter the magical world of cherry picking:
-When people look at the old J-L283 at Mokrin, if they are supporters of an indo-european origin (which I found far from being convincing with what we have), they will use autosomal DNA to show this dude was carying Steppe ancestry.
-When you look at Sardinian sample, the same peoples won't consider that much the admixture and will get satisfied with a complex admixture cleaning process.
-When looking at Z631 sample, here again, admixture is ignored and only location is retained.
For me, there is something strange about such methodology. It didn't look like science to me, it sounds like ideology, because the criteria that are used are place to validate a model, not to derive a model from the data.
Maybe Z631 will show up as not having its root in central Europe ... but current data are not saying that.
Personnally, I consider modern diversity as a very good proxy. Honestly, I don't know a single case where modern diversity completely fails.
Where, with ancient DNA, you are affected by sampling bias in a very significant way.
Most of the time Ancient sample are found ~500-1000 years post-split, implying an already strong level of migration-based convolution. Ancient samples are like modern diversity, with a bit less convolution but with a huge undersampling.
