ISCA : radical de noms de rivières.

La rivière EXE, en Devon

Sujet qui semble assez compliqué, et qui a provoqué une longue duscussion chez Rivet & Smith :

------------------


* Isca : radical de noms de rivières
* Rivet & Smith, p. 376 et suiv.
ISCA (river)
SOURCE
- Ptolemy II, 3, 3 : "Iska potamon ekbolai ( = ISCA FLUVII OSTIA), var. "Isaka (= ISACA, MS U)
DERIVATION. The name is British *isca 'water', in the sense of 'river', which underlies not only this and the following names but also Esk and Usk. The word seems to heve left few traces in Continental toponymy (see below). Jackson mentions early Irish esc, recorded in glossaries but not in texts, explained as meaning 'water'; to this Williams adds Gaelic easg 'fen'.
One must not, however, equate Isca 1 (Exe) and Isca 3 (Usk) simply and uncritically with *ïsca 'water', as has been done in the past. It is best to quote in full the analysis of Jackson in Britannia, I, (1970), 74-75 :
' The difficulty is that Caerleon is on the Usk, which is Welsh Wysg, Old Welsh Uisc, and needs British *Esca, earlier *Eisca; and that Exeter was Old Welsh *Caer Uisc (Asser : Cair Uuisc), from the same British form; though matters are further complicated by the seventeenth-century Cornish Karêsk, i.e. Kar Esk, given by Lhuyd for Exeter, which must be from *Isca not *Eska, in spite of appearances. A British *Eisca 'fishy river ' can be explained as from Indo-European *peisko- 'fish', as in Irish iasc ' fish ' which cornes from this ; as also, with different ablaut-grade, do Latin piscis and English fish. But if so, the Romano-British town-names should be Esca (i.e. Ëscâ) not Isca; and attempts by Forster, Williams, and others to explain this away, and also to account for the contradiction between Cair Uuisc and Kar Esk, are not convincing. The question must be left open.' 
The etymologies in question are reviewed by Nicolaisen in BZN, VIII (1957), 241-42, following in the main differing proposals of Forster. He does not favour the *peiska 'fishy water' mentioned by Jackson (now better *p(e)ik-sk-a, perhaps 'trout' rather than 'fish in général': Hamp in JIES, I (1973), 507-11), nor *peid-ska (cf. Greek pidax 'spring'). As between *eiska (with -k- suffix well documented in Celtic names) representing Indo-European *ei- ' gehen ' and the same representing *eis- ' (sich) heftig, ungestüm, schnell bewegen', Nicolaisen favours the latter, because it fits better with other river-names, including the type *Is- found in Isara and British *Isura. Nicolaisen then szparates from all these those names which he thinks directly descend from *Isca : a North and South Esk in Kincardine / Angus, an Esk in Kerry and Lough Easg of Donegal, seeing these as équivalent to Irish esc and Gaelic easg mentioned earlier. With these belong the few Continental names : Isch (a tributary of the Saar : Isca in the eighth century), Ijssche (Brabant; Isca in 822), Ischel (near Traunstein; Iscala in 984), Ischl (a tributary of the Traun; Iscala in 890), and Ischer (Alsace ; < *Iscara). To these Holder Ii. 77 adds a further assumed *Iscara and an Ambiscara.
The problem to which Jackson drew attention in 1970 is in no way diminished by all this. On the one hand, all our sources for Isca-names in Britian (ten, or eleven if we include Iscalis) and the Continental examples from medieval documents all show I-; on the other hand, the derivatives in the Celtic languages demand original Esca, from what-ever origins (all the etyma proposed have original *ei-, which became e in Common Celtic and continued as e in British : LHEB 330). Given the certain identity of ancient Isca 1 and Isca 3 rivers with modem Exe (which has, like Axe, an Anglo-Saxon metathesis) and Usk (without metathesis, like Esk) rivers respectively we may not postulate some exceptional factor such as a Celtic *Esca wrongly recorded as Romano-British Isca in the first-century period when native names were becoming established in Latin form. By 1970 Jackson had evidently dropped the explanation which he tentatively gave in LHEB 259 : that short i in a British penultimate became e in Late British if the following syllable was final and contained a, and that, if Exe, Axe and Esk are from Isca ('which is uncertain'), this explains the e in these; the reason no doubt being that the resulting e is not after all the e which is demanded by medieval and modem Celtic derivatives in versions of these names. There remains one tenuous possibility, which Jackson in LHEB dismissed. It is that Isca became *Esca in the spoken Latin of Britain, just as stressed i regularly became close ê in the speech of most parts of the Empire by the third century. In LHEB 259 there is the uncompromising statement that 'the change to ç...emphatically did not occur in British Latin as we know it, though it could presumably have done so in the low-class V.L. which may have been current in the cities'. The basis of this statement is, of course, the fact that most (not all) words having stressed i that were borrowed from Latin retain this i in Welsh (e.g. fïde(m) > ffydd, cippu(m) > cyff; but as suggested by Smith in ANRW (forthcoming), there may be much learned influence in the arresting of an expected process, and on the contrary, insciptions show examples of i > e (e.g. ella for ïlla in RIB 154, Bath; stepibus for stipibus in the Lydney pavement, CIL VII, 137; baselicam in RIB 978, Netherby, A.D. 222). We then have to suppose that this Esca of spoken Latin influenced the local British pronunciation of the place-names and river-names in question. Since both places are westerly one could argue that Latin speech continued longer in them than in most other regions, and in the area of the legionary fortress at Caerleon the influence of Latin must have been considerable.
This is certainly special pleading, but is at least a possible explanation of what remains a puzzling situation. The reverse would have to apply to the Gallo-Latin Isca- names of the Continent, which against expectation have preserved I- in their modem forms (despite a form Esca recorded for the river Isch of the Saar in the eighth century : Holder II 77)
IDENTI FI CATION. The river Exe, Devon. 

-----------------------------

Petit problème d'ordre technique : l'impossibilité de faire apparaître les accentuations.

Pour ceux qui le désirent, je peux faire une photocopie .

Me donner les adresses sur messagerie privée.

Le texte ci-dessus a été intégré à l'encyclopédie :

http://marikavel.org/lieux/lieux-i.htm

JCE