Mapping the Origins and Expansion of the Indo-European Language Family

[princenuadha]

@jean

Well to be fair I was just pondering whether or not the r1a1a came to GPCW with yamanaya. I know there were migrants, but there was also some continuity.

I'm really just talking about r1a1a, not the existence of yamanaya movements west. I already believe the latter.

[Jaska]

We have to be careful, because non-IE does not necessarily mean pre-IE.

Good point.

Basque appears to be a Copper Age language which may not have evolved where it is now spoken.
Etruscan appears to have arrived in Italy from NW Anatolia c.1,200 BC.

At least Basque still seems to be older in Western Europe than Indo-European languages. I'm not sure when the Italicization of Italy occurred - before or after Etruscan.

Correction - Siculan was an IE language usually grouped with Q-Italic.

Thank you, so it seems to be.

Well I didn't want to be picky, but Koch recently argued that Tartessian is Celtic, and if we assume that the language which left the -ossos and -inthus place-names is Pelasgian, then some linguists argue that was Luwian.

That's very interesting. :)
I only saw mentions that Anatolian and Greek may have related substrates.

I am curious though about what the language in the western forrest steppe of the Ukraine and west might have been.

Language of geminates has left traces at least in Germanic and Balto-Slavic, if I remember right what Schrijver wrote. Also western Uralic languages have traces of a language with un-Uralic nasal geminates *mm and *nn, but some shared substrate loanwords would be needed to identify if the language preceding Indo-European was the same preceding Uralic.

[Jean M]

At least Basque still seems to be older in Western Europe than Indo-European languages.

I'd love to see any evidence of this. It just seems to be an unquestioned assumption.

I'm not sure when the Italicization of Italy occurred - before or after Etruscan.

There is place-name evidence that the Etruscans entered a country that was already Indo-European. Specifically they entered the territory of the Umbrians. They created a linguistic wedge between IE-speakers north and south, which no doubt encouraged the separation of Italic from Proto-Italo-Celtic.

[Richard Rocca]

At least Basque still seems to be older in Western Europe than Indo-European languages.

I'd love to see any evidence of this. It just seems to be an unquestioned assumption.

I'm not sure when the Italicization of Italy occurred - before or after Etruscan.

There is place-name evidence that the Etruscans entered a country that was already Indo-European. Specifically they entered the territory of the Umbrians. They created a linguistic wedge between IE-speakers north and south, which no doubt encouraged the separation of Italic from Proto-Italo-Celtic.

They only created a partial wedge as the Umbrians still held the territories to the east of the Etruscan areas. The real dividing line between Celtic and Italic seems to have been slightly further north, with the Ligurian tribes speaking a Celto-Italic language and the Veneti speaking an Italic language with strong Celtic and Germanic influences. I think it is more likely that the Etruscans continued an already existing differentiation process between Italic and Celtic.

[Jean M]

@ Richard

Yes I think I say something like that. 

[Jaska]

I'd love to see any evidence of this. It just seems to be an unquestioned assumption.

The evidence does not come from Basque but from the Indo-European side:

1. Celtic has developed in continuous contacts with Germanic:
http://www.ling.upenn.edu/~kroch/courses/lx310/ringe-handouts-09/celt-loans.pdf

2. Germanic has developed in continuous contacts with Finnic and Saami:
http://www.mv.helsinki.fi/home/jphakkin/Jatkuvuus2.pdf

There are neither early Italic loanwords in Germanic, nor Celtic loanwords in Finnic and Saami, which shows that there are loanword layers only between adjacent languages. The development of Germanic must be placed in Scandinavia and the development of Celtic right to the south of it: Germany.

Other possibility would be to place the Germanic branch in Baltia, but Germanic shares less loanwords with Balto-Slavic than with Finnic and Saami, so Scandinavia must be the place.

No genetic or archaeological evidence can disprove these linguistic results. Celtic expansion to Western and Southwestern Europe must be later.

[alan trowel hands.]

@ princenuadha

No. I just don't know where to start.

Look - the TRB people were similar to Balkan Neolithic types genetically. The Balkan Neolithic types are nothing like the steppe foragers or the Siberian foragers with mtDNA U2e who were more robust types. The TRB farmer who was tested for a large chunk of her genome was like Southern Europeans, not modern Scandinavians. So modern Scandinavians with all that R1a1a cannot be unalloyed descendants of the TRB. Corded Ware types were more robust and rather like Mesolithic types.

The idea that CW was local and just absorbed ideas from Yamnaya was orthodoxy for a long while, but does not meet the test of genetics. It is Yamnaya-derived, not just culturally, but genetically.

I must admit that while not rejecting it completely I am highly suspicous of the robust/tall gracilee/short thing.    and stress just has so much to do with it.  Even in my own family the male line shrunk from someone who was remembered as a big strong guy then through urban poverty (and I imagine a bread and tea type  ) to a slight guy about 5ft 4 or 5 who then over the next three generations of improving economic position and better  grew about 3 inches bigger and more robust build each generation until well over 6ft.  You also see the effect of  world wide.  The young Japanese tower over their grandparents due to a more western .  The Dutch went from being very short to being the biggest people in Europe in one generation again down to .  Height and build is one of the things I trust least in the archaeological record.  I strongly suspect skull and face shape is influenced by this too although I dont have an real data on this.  I think I read that the lower face is underdeveloped in poverty.  

NB- for some reason the word d..i..e..t is being deleted from my post

[Mikewww]

At least Basque still seems to be older in Western Europe than Indo-European languages.

I'd love to see any evidence of this. It just seems to be an unquestioned assumption.

I'm not sure when the Italicization of Italy occurred - before or after Etruscan.

There is place-name evidence that the Etruscans entered a country that was already Indo-European. Specifically they entered the territory of the Umbrians. They created a linguistic wedge between IE-speakers north and south, which no doubt encouraged the separation of Italic from Proto-Italo-Celtic.

They only created a partial wedge as the Umbrians still held the territories to the east of the Etruscan areas. The real dividing line between Celtic and Italic seems to have been slightly further north, with the Ligurian tribes speaking a Celto-Italic language and the Veneti speaking an Italic language with strong Celtic and Germanic influences. I think it is more likely that the Etruscans continued an already existing differentiation process between Italic and Celtic.

By Veneti, you mean the Adriatic Veneti, right?

[Richard Rocca]

At least Basque still seems to be older in Western Europe than Indo-European languages.

I'd love to see any evidence of this. It just seems to be an unquestioned assumption.

I'm not sure when the Italicization of Italy occurred - before or after Etruscan.

There is place-name evidence that the Etruscans entered a country that was already Indo-European. Specifically they entered the territory of the Umbrians. They created a linguistic wedge between IE-speakers north and south, which no doubt encouraged the separation of Italic from Proto-Italo-Celtic.

They only created a partial wedge as the Umbrians still held the territories to the east of the Etruscan areas. The real dividing line between Celtic and Italic seems to have been slightly further north, with the Ligurian tribes speaking a Celto-Italic language and the Veneti speaking an Italic language with strong Celtic and Germanic influences. I think it is more likely that the Etruscans continued an already existing differentiation process between Italic and Celtic.

By Veneti, you mean the Adriatic Veneti, right?

Correct. There are some attempts to connect the Veneti of the N. Adriatic with the Atlantic Veneti, but it seems to be a stretch.

[Mikewww]

I'd love to see any evidence of this. It just seems to be an unquestioned assumption.

The evidence does not come from Basque but from the Indo-European side:

1. Celtic has developed in continuous contacts with Germanic:
http://www.ling.upenn.edu/~kroch/courses/lx310/ringe-handouts-09/celt-loans.pdf

2. Germanic has developed in continuous contacts with Finnic and Saami:
http://www.mv.helsinki.fi/home/jphakkin/Jatkuvuus2.pdf

There are neither early Italic loanwords in Germanic, nor Celtic loanwords in Finnic and Saami, which shows that there are loanword layers only between adjacent languages. The development of Germanic must be placed in Scandinavia and the development of Celtic right to the south of it: Germany.

Other possibility would be to place the Germanic branch in Baltia, but Germanic shares less loanwords with Balto-Slavic than with Finnic and Saami, so Scandinavia must be the place.

Germanic and Celtic languages, or some elements of them, may have developed in these spaces as you suggest. However, what would be most useful would be to understand where proto or pre- Germanic, Italo-Celtic and Balto-Slavic languages were spoken. They may have originated in different places than where elements of them interfaced and further developed.

No genetic or archaeological evidence can disprove these linguistic results. Celtic expansion to Western and Southwestern Europe must be later.

Sounds like a preconceived notion when new evidence can't change a perspective. Perhaps the challenge will not be in the linguistic results but your interpretation of them.

Back to Jean's question, I don't see how your commentary on IE languages provides evidence of your assertion.

At least Basque still seems to be older in Western Europe than Indo-European languages.

[Jean M]

@ Jaska

If we accept the existence of Proto-Italo-Celtic, as Don Ringe and colleagues do, things fall into place. It is my contention that Proto-Italo-Celtic spread initially up the Danube and then split into two streams from the Carpathian Basin.

One went south into Italy and across to Iberia. There we have evidence of Lusitanian, which seems Italo-Celtic, and Ligurians are mentioned. Ligurian is another seemingly Italo-Celtic language.  

The other stream moved north to the head of the Danube, north of the Alps. There Proto-Celtic developed. This group became dominant already in the mid-Bell Beaker period. (Some of them moved into NE Iberia. The language Celtiberian seems to reflect the early Proto-Celtic stage.) In the Bronze and Iron Ages there were migrations out of this area in pretty well all directions.

Proto-Germanic developed c. 500 BC in the Jastorf culture in contact with Celtic to the south.

If Basque had been the pre-IE language of Iberia, we would expect that to show in hydronyms at least. It doesn't. Nor are there borrowings from Basque in all the archaic Hispanic languages. It appears in fact that the precursor to Basque was Aquitanian, which was only spoken in part of Navarre in the Roman period. It entered the present Spanish Basque Country in the Post-Roman period.

[Jean M]

I must admit that while not rejecting it completely I am highly suspicious of the robust/tall gracile/short thing.

Me too believe me. But in this case the genetic results match up. The TRB farmer is not like modern Scandinavians. Of course we want lots more ancient DNA to completely sort all this out. But so far we have no reason whatsoever to believe that R1a1a was Neolithic in Europe or anywhere else. See http://www.familytreedna.com/public/R1a/ for tree and estimated dates.

for some reason the word d..i..e..t is being deleted from my post

It does that to me.

[alan trowel hands.]

I must admit that while not rejecting it completely I am highly suspicious of the robust/tall gracile/short thing.

Me too believe me. But in this case the genetic results match up. The TRB farmer is not like modern Scandinavians. Of course we want lots more ancient DNA to completely sort all this out. But so far we have no reason whatsoever to believe that R1a1a was Neolithic in Europe or anywhere else. See http://www.familytreedna.com/public/R1a/ for tree and estimated dates.

for some reason the word d..i..e..t is being deleted from my post

It does that to me.

lol Think the word 'grub' or 'nosh' will have to do instead of .  I am looking forward to a paper called 'The effects of nosh on skeletal structure and how grub changed over the Neolithic-copper age transition'.

[princenuadha]

I must admit that while not rejecting it completely I am highly suspicous of the robust/tall gracilee/short thing. and stress just has so much to do with it. Even in my own family the male line shrunk from someone who was remembered as a big strong guy then through urban poverty (and I imagine a bread and tea type ) to a slight guy about 5ft 4 or 5 who then over the next three generations of improving economic position and better grew about 3 inches bigger and more robust build each generation until well over 6ft. You also see the effect of world wide. The young Japanese tower over their grandparents due to a more western . The Dutch went from being very short to being the biggest people in Europe in one generation again down to . Height and build is one of the things I trust least in the archaeological record. I strongly suspect skull and face shape is influenced by this too although I dont have an real data on this. I think I read that the lower face is underdeveloped in poverty.

I take this as unwavering support for my "neolithic" speculation : )

In all seriousness though, despite CW being reported as robust like the mesolithic Europeans, the recent map by DiMarco suggests that there are different types of CW peoples and that the GPCW group (high in r1a1a?) is more neolithic-like. That's essentially what kick started my wandering.

Is it at all feasible for r1a1a to have been tied to some neolithic European movements, and latter spread east from eastern Central Europe based the haplogroup itself?

Also, back to r1b, how do you propose that r1b would have come to D-D without the C-T and earlier milk drinkers from around marmara? Or have you abandoned that idea?

[Jean M]

Is it at all feasible for r1a1a to have been tied to some neolithic European movements, and latter spread east from eastern Central Europe based the haplogroup itself?

The idea that R1a1a sprang up in Central Europe or on the North European Plain or in India, or wherever happens to be closest to the hometown of the person presenting the case, is the cherished idea of romantics who happen to carry the haplogroup. Wouldn't it be lovely, they think, if my very own ancestors had lived within sight of the glorious (insert favourite landmark) from the Palaeolithic to the present.  

Practical persons want proof. Haplogroups cannot move without people carrying them around. That does not always mean that we see a clear trail in archaeology. But let's say the evidence is that movement which ended up with R1a1a in ancient DNA all took place from the Pontic-Caspian steppe from c. 4000 BC onwards.  

Obviously R1a1a is older than that and some might have travelled quite widely in the Holocene. Some might have reached India as the deserts in between bloomed. Some might have moved north. But if so, those movements did not leave a lot of descendants it seems, as the bulk of R1a1a tested in the subclades project is falling into the 4000 BC + subclades that match up with IE languages. That is what we would expect. The Copper Age economy was more productive and supported a rising population.  

[alan trowel hands.]

I must admit that while not rejecting it completely I am highly suspicous of the robust/tall gracilee/short thing. and stress just has so much to do with it. Even in my own family the male line shrunk from someone who was remembered as a big strong guy then through urban poverty (and I imagine a bread and tea type ) to a slight guy about 5ft 4 or 5 who then over the next three generations of improving economic position and better grew about 3 inches bigger and more robust build each generation until well over 6ft. You also see the effect of world wide. The young Japanese tower over their grandparents due to a more western . The Dutch went from being very short to being the biggest people in Europe in one generation again down to . Height and build is one of the things I trust least in the archaeological record. I strongly suspect skull and face shape is influenced by this too although I dont have an real data on this. I think I read that the lower face is underdeveloped in poverty.

I take this as unwavering support for my "neolithic" speculation : )

In all seriousness though, despite CW being reported as robust like the mesolithic Europeans, the recent map by DiMarco suggests that there are different types of CW peoples and that the GPCW group (high in r1a1a?) is more neolithic-like. That's essentially what kick started my wandering.

Is it at all feasible for r1a1a to have been tied to some neolithic European movements, and latter spread east from eastern Central Europe based the haplogroup itself?

Also, back to r1b, how do you propose that r1b would have come to D-D without the C-T and earlier milk drinkers from around marmara? Or have you abandoned that idea?

Well it seems unlikely that R1a was in the earliest Neolithic spread but there is a heck of a long time between 6 or 7000BC and 3000BC and a lot could have happened in between in farming zone of SE Europe.  Probably the only chance for R1a to have been among later Neolithic farmers in SE Europe YET also have got somehow into the steppe populations who later moved east would be if it somehow was present, entered  or got encorporated in SE or east-central Europe in some farming culture in the middle Neolithic and became part of the Cuc-Tryp culture.  I am not supporting that but its probably the only possible wriggle room area I can think of for a farming origin of R1a.

One thing I do get a bit troubled by though is that there are a lot of parallels between R1a and R1b other than their later distributions but the explanation sought tends to be very different.  I would not be surprised if they both had a similar histories pre-4000BC and we are maybe being influenced too much by the modern distributions.   
 

 

[Jaska]

Germanic and Celtic languages, or some elements of them, may have developed in these spaces as you suggest. However, what would be most useful would be to understand where proto or pre- Germanic, Italo-Celtic and Balto-Slavic languages were spoken. They may have originated in different places than where elements of them interfaced and further developed.

Indeed there are Northwest Indo-European, Balto-Slavic, Proto-Baltic, Middle Slavic, Dialectal Baltic, Palaeo-Germanic, Early Proto-Germanic, Late Proto-Germanic etc. loanwords in Finnic and Saami. And yet we have only few late loanwords between Germanic and Balto-Slavic.
http://books.google.fi/books?id=RqkBXIJkkuEC&printsec=frontcover&hl=fi#v=onepage&q&f=false
(Chapter 67: Koivulehto)

Conclusion: Germanic development could not have occurred anywhere near the Southeast corner of Baltic Sea, but it must have occurred in Scandinavia. Only at some later stage Germanic spread to the area of Jastorf culture. Jastorf area cannot explain (1) the lack of Germanic–Balto-Slavic contacts, and (2) Germanic–Finnic/Saami contacts.

Sounds like a preconceived notion when new evidence can't change a perspective. Perhaps the challenge will not be in the linguistic results but your interpretation of them.

Please take a look at the linguistic evidence above: Germanic developed in the area adjacent to Finnic and Saami. You cannot find another interpretation for those results, can you?

Read this if you disagreed with my claim that archaeological or genetic continuity cannot testify for linguistic continuity:
http://www.mv.helsinki.fi/home/jphakkin/Uralic.html

Back to Jean's question, I don't see how your commentary on IE languages provides evidence of your assertion.

1. Celtic language is the earliest Indo-European languages in Western Europe.
2. Celtic language cannot have spread to Western Europe until the first millennium BC, because it was developing adjacent to Germanic, which was developing adjacent to Finnic and Saami.
3. Basque probably was there already before them, if we have this wide area of ancient Vasconic placenames (although they seem to be disputable).

The other stream moved north to the head of the Danube, north of the Alps. There Proto-Celtic developed. This group became dominant already in the mid-Bell Beaker period. (Some of them moved into NE Iberia. The language Celtiberian seems to reflect the early Proto-Celtic stage.) In the Bronze and Iron Ages there were migrations out of this area in pretty well all directions.

It could be possible that some Pre-Celtic dialect spread to Iberia already at the second millennium BC (if we have evidence?). But Celtic proper hardly could.

Proto-Germanic developed c. 500 BC in the Jastorf culture in contact with Celtic to the south.

That one is an axiom. Proto-Germanic cannot have been developed there: it arrived there from Scandinavia; see the evidence above. Contacts with Finnic and Saami were continuous since Northwest Indo-European. Archaeological continuity cannot prove about linguistic continuity, and that is all the support there is for the Jastorf hypothesis.

If Basque had been the pre-IE language of Iberia, we would expect that to show in hydronyms at least. It doesn't. Nor are there borrowings from Basque in all the archaic Hispanic languages. It appears in fact that the precursor to Basque was Aquitanian, which was only spoken in part of Navarre in the Roman period. It entered the present Spanish Basque Country in the Post-Roman period.

How about the wide area of Vasconic placenames? Although it seems that they are disputed; I haven't assessed them myself, so it is hard to say anything conclusive yet. But do we have any linguistic evidence for the early spread of Pre/Para-Celtic languages to Iberia, either?

[MHammers]

Here is an excerpt from Zsuzanna K. Zoffman from Human remains from the kurgan at Hajdunanas-Tedej-Lyukas-halom and an anthropological outline of the Pit-Grave ethnic groups that supports a mixed population of Yamnaya people that might have implications for R1b and proto-Beaker people.

This is talking about the physical types found in Yamnaya graves that are different from the common steppe Cro-magnon type.

'Besides them, the representatives of a more gracile variant with narrower cranium and hypsiconch orbits also occur in the material from the Carpathian basin (Marcsik 1979; Zoffman 2000; 2004; 2006).  At the same time, leptodolichomorhous (gracile Mediterranean) and brachymorphous type variants appear at the Brailita Pit-Grave site in Moldova (Necrasov and Cristescu 1957).  According to the Romanian authors (Necrasov and Cristescu 1957, 1973; Haas and Maximilian 1958), the Cro-magnon type arrived from the east , while the gracile Mediterranean and the brachycranial types represent the autochthonous population.'

[Jean M]

Germanic development could not have occurred anywhere near the Southeast corner of Baltic Sea, but it must have occurred in Scandinavia.

Yes that all fits with the archaeology. I see pre-Germanic speakers being sucked into Scandinavia in the Bronze Age when the climate was right for farming there, then moving south into Jastorf when the climate turned bad for farming further north. Proto-Germantic strictly speaking (i.e. the language immediately preceding the break-up into North, East and West Germanic) was therefore developed in Jastorf, which probably displaced some Celtic speakers. The expansion of the Germani appears to have been at the expense of Celts. But this was the extreme edge of the vast Celtic-speaking zone.

[Jean M]

Basque probably was there already before them, if we have this wide area of ancient Vasconic placenames (although they seem to be disputable).

What wide area of Vasconic placenames? If you are thinking of Venneman's theory, that has little/no support. Larry Trask didn't accept it all.

[Jean M]

It could be possible that some Pre-Celtic dialect spread to Iberia already at the second millennium BC (if we have evidence?). But Celtic proper hardly could.

Celtiberian is the most archaic form of Celtic, but it is generally seen as Celtic. I don't see any problem with that. As I said, Lusitanian appears earlier. It presumably arrived with Bell Beaker. Later there were waves of Gauls into parts of Iberia, bringing their own Celtic. In Galicia they overlaid a previous linguistic layer similar to Lusitanian. See http://www4.uwm.edu/celtic/ekeltoi/volumes/vol6/6_16/index.html

[princenuadha]

@alan trowel hands.

I'll try to squeeze this in that tiny space : )

Here is a senario, or a working hypothesis. Note that ill will presuppose that the given dates on the subclade map, here[1], are correct.

M417 originates around Central Europe or eastern Central Europe, near to where r1a1a* is found in abundance [1]. A group of M417 people wander west and latter become L664 folks. A mutation, z645, takes place somewhere around the balkans or central Europe. Some of those descendants stay in the area and become z283 while some move east (along the forest, not southern steppes) to become z93. Note that the ones who moved east mixed with locals, assuring that they were "meso-European-like" in their autosomal composition.

In the mean time, TRB people migrate into Scandinavia, immediately from Central Europe and possibly originating in the (late) neolithic balkans. The TRB people carry higher frequencies of M417 and "neolithic-like" autosomal DNA than the pre TRB Scandinavians. The migrating TRB people have a greater demic impact on contemporary Sweden than Denmark/North Germany. This results in the Swedish TRB being more "neolithic-like" in terms of mtdna [2] and craniometry [3] than the Danish-German TRB, along with having greater frequencies of M417 which is still causes some of the M417 disparities to this day.

By the time the German-Polish CW (GPCW) appears, the area is relatively high in M417 [2] and "neolithic-like" heritage [3]. Around 3k bc mutual interactions take place between CW, including GPCW, and yamanaya which certainly involves the spread yamanaya people and culture [4] (language included). I propose this exchange also involved the spread of forms of M417 into the southern steppes and yamanaya people. (Non-CW, M417, forest people may have done the same around the same time)

The exchanges in this part of the theory are especially convenient because it leaves room for r1b yamanaya in the southern steppes, largely exempt of r1a. In my theory, the migration of southern yamanaya people to GPCW could also explain what seems like an old presence of r1b in Eastern Europe.

Here are my sources:

[1] "R1a1a and Subclades Y-DNA Project - Results", familytreedna.com map "1. Old European (DYS392=13) ("C"/"OE")" Sept. 30 2012

[2] "Ancient Western Eurasian DNA", buildinghistory.org Sept. 30 2012

[3] DiMarco S. (2011) - Gente di Rame: Variabilita morfometrica craniofacciale e relazioni fenetiche in gruppi umani eneolitici dal territorio italiano. Attu della XLIII Riunione Scientifica. L'Eta del Rame in Italia. Florence 2011. pp 375-381.

Or  http://www.u152.org/images/stories/Copper_Age_Skulls.png courtesy Richard Rocca

[4] Richard Harrison and Volker Heyd, The Transformation of Europe in the Third Millennium BC: the example of ‘Le Petit-Chasseur I + III’ (Sion, Valais, Switzerland), Praehistorische Zeitschrift, vol. 82, no. 2 (2007), pp. 129–214.

Page 73 in article

 

[Jean M]

@ MHammers

I think you have cracked it.

[Jean M]

The exchanges in this part of the theory are especially convenient because it leaves room for r1b yamanaya in the southern steppes, largely exempt of r1a.

I suspected that this was R1b wishful thinking on your part, the counterpoint to the Rah-rah-R1a1a cheer-leading. :) Understandable maybe, but wishful thinking tends to get dashed on the rocks of reality sooner or later, so I find it generally best to follow the evidence rather than try to find "wiggle-room" for convoluted explanations that would make you happy for five minutes more.

The evidence suggests that R1a1a left the Repin culture at the east end of the steppe before Yamnaya developed, moving east to Afanasievo, and then south into the Tarim Basin, where it is found in mummies that we presume were the ancestors of the Tocharian speakers who show up there later in written records.  

The complete absence of any evidence of any cultural or genetic flow south from Central Europe into Yamnaya has clearly not deterred you, but I feel that I should point it out once more. The "exchanges" that you picture were a one-way flow northward.

[IALEM]

It could be possible that some Pre-Celtic dialect spread to Iberia already at the second millennium BC (if we have evidence?). But Celtic proper hardly could.

Celtiberian is the most archaic form of Celtic, but it is generally seen as Celtic. I don't see any problem with that. As I said, Lusitanian appears earlier. It presumably arrived with Bell Beaker. Later there were waves of Gauls into parts of Iberia, bringing their own Celtic. In Galicia they overlaid a previous linguistic layer similar to Lusitanian. See http://www4.uwm.edu/celtic/ekeltoi/volumes/vol6/6_16/index.html

In fact that is not the conclusion of the paper, but that there was some Lusitanian influence in Galicia but that there was an IE layer previous to Celtic that was not Lusitanian. Furthermore, Celtici would have arrived not only from Celtiberian area but also from Andalucia.
I have to say I am very sceptical about those theories of linguists that see layer over layer of prehistorical IE languages, or alternatively Non IE using just a few names of dubious orthography. Curchin, for instance, see 2 different prehistoric IE non Celtic layers previous to "Iberian Colonization" in SE Spain, it is just going insane...