Vasco-Caucasian

The Vasco-Caucasian hypothesis states that Basque is related to Nakh-Daghestani (NEC) and Abkhaz-Adyghe (NWC).

The proposed time-depth for this hypothetical macro-family is the early Neolithic. Some authors like John Bengtson group Vasco-Caucasian and Burushaski into Macro-Caucasian.

The main source for Vasco-Caucasian etymologies is the Proto-North-Caucasian reconstruction made by Seregi Starostin.

Basque and Iberian form the Vasconic language family. But my colleague Marco Moretti and myself have been able to identify two more Vasco-Caucasian families different from Vasconic.

Latin posseses some substrate loanwords like aquila, cucullus, cucutium, cumulus, lutum, mucus, solum, turris which aren't of IE origin but have a Vasco-Caucasian etymology. I call the source language Apenninic because it was possibly related to the homonymous archaeological culture of the Bronze Age.

Apenninic had a relative in the Iberian Peninsula called Cantabrian, which left loanwords in Basque like burki, kokor, kokotz, kolko, kukur, konkor, moko, mokor, morgoi, morroi, moxo, mozkor, muki, mukurru, tontor, topina.

But there was also another Vasco-Caucasian which left loanwords in Basque. I call it Pyrenaic and was responsible for words like ate, dardala, ilinti, ondiko, samats, simaur, urde, zamar. It seems that Pyrenaic was closer to NEC than the other two families and it was also related to Pre-IE Germanic, as it show a number of etymologies.

As an example of reflexes in the three groups, I'll choose  PNC *k'aldlZ:wV: (˜ -tL':w-) 'cheek; chin' (there's no resontrcution of Proto-Vasco-Caucasian yet). 

The Cantabrian protoform would be *k'o(l)X:wV:-S:V (˜ -s:V) 'chin', with Basque outputs kokots (HN, L, LN, Z, R), kokotz (L, LN), okots (B) 'chin'.

For Vasconic I reconstruct *k'a(l)G:wV:, which gives Basque aho, ago 'mouth'.

For Pyrenaic I reconstruct the protoform *k'alt:wV: 'cheek', which gives Catalan galta id.