--- In
cybalist@yahoogroups.com, Rick McCallister <gabaroo6958@...> wrote:
>
> ***R
> It would seem that there are maybe 3 to 5 known branches of Celtic
> 1. Ibero-Celtic
> 2. Gallo-Brythonic, including Lepontic and possibly Belgic & Ligurian
> 3. Goidelic
> 4. Maybe Tartesso-Celtic
> 5. Maybe Noric-Balkan-Galatian Celtic (unless an extension of  G-B)
Strabo was correct in distinguishing Ligurians from Celts. Several phonological developments show that Ligurian cannot be a Celtic language:
1. PIE *p > Lig. /p/. The river Porcobera (Sent. Minuc.), Latinized as Porcifera (Plin.), now Polcevere, evidently means 'salmon-bearing' (*pork^o-bHer-); cf. Old Irish <erc>, <orc> 'perch, salmon', OHG <ferhana> 'trout', etc.
2. PIE *gWH > Lig. /b/. Lucus Bormani (Itin. Ant.) and other Lig. place-names (and god-names) associated with warm springs, collected by Kretschmer, evidently reflect PIE *gWHorm- 'warm', Lat. <formus>; cf. Skt. <gharmas>, Old Pruss. <gorme>, OIr <gor> 'heat', etc.
3. PIE *r. > Lig. /ar/. Bargae (Tab. Vel.) evidently continues the PIE zero-grade *bHr.g^H- (Celt. *brig-, Gmc. *burg-) from *bHerg^H- 'to protect, defend'.
4. PIE *eu > Lig. /eu/ against *ou, which Matasovic' considers pan-Celtic. He explains Neviodunum (sic) as an orthographic variant of Gallo-Latin Noviodunum, but in fact Nevidunus (Tab. Vel.) is Ligurian (in form possibly a Caland-bahuvrihi), also Leucomelium (Tab. Vel.), the personal name Nevius (Lat. inscc. of Liguria), and the Salluvian king's name Teutomalius (Liv. Epit. 61).
Ligurian agrees with Celtic in rendering the non-labiovelar mediae aspiratae as mediae. It also shows *gW > /b/ and *kW > /p/.
DGK