frisgo

Old French <fregon>, now <fragon> 'petit houx, houx-frelon, pungitopo, butcher's broom, knee-holly, Ruscus aculeatus', along with Norman <frégon>, Poitevin <fringon>, and dialectal forms in Charente [frego:], Vienne [frgo:], and Haute-Vienne [fre^gú, fre.ge:], continues Gallo-Latin <frisgo(:)> attested in glosses. Another gloss <brisco> (for *brisgo:?) is represented by Limousin <bregou>, in Basses-Alpes [bre.gu:], in Charente again [brigu:]. A protoform *grisgo: is indicated by <grigon> along the Soudre (North Gironde), [gri:gu:] in Gironde, and [grisku:n] in Les Landes.

Now, we do have evidence that native Gaulish *fr- developed regularly from Celtic *sr-, which is preserved in Irish. Thus French <(se) re(n)frogner> 'to frown, scowl, knit one's brows' can be traced to Gallo-Latin *refrogna:re 'to turn up one's nostrils', from Gaul. *frogna: 'nostril', continuing Celt. *srokna: (Welsh <ffroen>, Breton <fron> 'nostril'; cf. Old Irish <srón> 'nose'). And Lombardic <froda> 'stream, waterfall' (attested since 1280) continues Gallo-Latin *fruta, with suffix-change from Celt. *srutu- 'stream' (We. <ffrwd>, Br. <froud>, Old Ir. <sruth>). However, there is no basis for supposing that original *sr- could yield *br- or *gr- in any part of Gaul.

I believe that what we have here is a very important isogloss indicating the linguistic situation in Gaul before its conquest by the Gauls. If the protoform of <fregon> and the rest involved *gWHris-, we expect Celtic to have *gris-, Ligurian to have *bris-, and *fris- is presumably Venetic. The attestations of *bris- do indeed occur in Liguria Transalpina, and those of *gris- are consistent with Gallicization proceeding from the southwest.

Recently I suggested that the Durance (Manche) and Drouance (Calvados), both continuing *Druantia, could be taken as evidence of Ligurian settlement in northwestern Gaul, with the Gaulish conquest of central Gaul splitting these NW Ligurians from their Transalpine fellows. It is time to change this hypothesis. The <fregon> forms instead indicate that the pre-Gaulish population of NW Gaul was Venetic, with the Ligurian boundary running through Charente. Likewise, the *Druantia of Poland, now Drwe,ca/Drewenz, probably represents Venetic settlement rather than Ligurian (which ought to make Torsten happy).

Wherever the Common Celtic homeland was, the P-Celtic/Gaulish homeland might have been in Aquitania, SW of the Garonne. Here presumably Celt. *sr- became *fr-, and *wr- became *br- (beside existing *br- from earlier *bHr-). Thus when the Gauls conquered the Veneti to their north, Venetic *fr- would have been maintained in loanwords. Huge numbers of Gauls migrating across the Garonne to greener pastures would have allowed the back-filling of Aquitania by the Vasconic Aquitani, leading to the linguistic situation described by Caesar.

DGK