my point is linguistics cannot tell anything about the tribal names that I am relating into groups of same origin, because we do not know the languages of e.g. Sherdana. Scordisci .... or Serians, Scirii, Cimmerians, white Syrians...or of Gutians who ruled over Sumer.....
Actually, we do not need to know the languages strictly speaking. If you say that these names are related (hypothetically speaking, we can assume that for the moment), we would have to postulate sound laws existing that are responsible for changing these names from one to another, and also, what would be the original form, hypothetically speaking?
š-r-d-
n
skordisci
Serian
Skiri
Cimmeri
Syria
So according to this, *ʃ- (š) corresponds to *sk-, corresponds to *s, corresponds to *k (keep in mind that originally in Greek, the name is rendered as "
Kimmeroi"). Which one is the original form? I would personally argue *sk-, because it's possible to derive all other forms from this (*sk > *s > *ʃ, or *sk > *k).
Likewise, *d is corresponded by a Ø (missing) in all names except Sherdana and Scordisci. It's more likely that *d is the original form and that it disappeared than a *d suddenly appearing.
Then we have *-mm- in "Cimmeri", which has no equivalent whatsoever elsewhere, yet we must assume it to be also the original form.
We also have a final *n in "Sherdana" which we have no where else.
Amongst the vowels, we have *a = *e = *i = *o = *y. Which one is the original form? It's impossible to tell.
Based on the above, I would personally probably propose that the original form is something along the lines of *
sk(V)mm(V)rd(V)n (with "V" representing unknown vowels), which appears just too unlikely to me.
it is clear that linguistics cannot claim that those tribal names are related, nor that they are not...
No, I obviously can't. But that's perhaps my point. As you can see above, it's
very unlikely that such a multitude of (in some cases bizarre) sound shifts would have occured,
even if such a name was transmitted.
hence I use other tools: geographical spread, genetic correlation and history....
No. You have no way of demonstrating that these names are related, but
ad-hoc assume that a relation exists anyways (without knowing wether this is the case or not), and
ad-hoc attempt to correlate this with patterns that you see on maps.
finally, a question for you: if you imagine PIE langauges as a tree, where do sound laws exist
1) only between parent branch and a child branch
2) also among child branches and grandchild branches...how?
Sound laws exist everywhere. And, I do not think that the tree model does necessarily explain a language family reasonably well (but neither does the wave model in the same manner), at least, not in the conventional sense: one "node" may have multiple "twigs" sprouting from it simultaneously (the Romance languages are a good example of that).
As I said, sound laws do exist everywhere, multiple "twigs" can also have common sound laws due to an area effect. You have to consider that sound laws also work on dialect level. The German dialect lines are a good example of that: the various sound laws of the second germanic sound shift affect German dialects to varying degrees, and as a result you have the Benrath and Speyer lines, for instance.