Re: TACITUS' CHAMAVI; A GERMANIC TRIBE IN THE RHINE REGION

From: dgkilday57
Message: 67703
Date: 2011-06-06

--- In cybalist@yahoogroups.com, "Brian M. Scott" <bm.brian@...> wrote:
>
> At 9:07:34 PM on Monday, May 16, 2011, The Egyptian
> Chronicles wrote:
>
> [Quoting Wikipedia:]
>
> > Many settlements are named Hamm, including possibly a
> > modern city, Hamburg.
>
> Since the oldest form is <Hammaburg>, the first element
> seems likelier to be akin to OE <hamm> 'meadowland; enclosed
> plot of land; solid land in a marsh; land in the bend of a
> river' and OFris. <hamm> 'enclosed land'.
>
> > The name may have come from the Germanic equivalent of
> > Chamavi.
>
> [...]
>
> > The -AVI, an adjectival ending, later resulted in -au in
> > other place names, but was dropped in this one Chamavi in
> > this derivation would mean "men of the settlements" or
> > "settlers." When and in what sense they were so is lost in
> > prehistory."
>
> I'm skeptical of this 'adjectival ending' (except in
> originally Slavic names). I've not looked especially hard,
> but so far I've found only three tribal names in <-avi>:
> <Chamavi>, <Batavi>, and <Brisgavi>. In <Batavi> the <-avi>
> is probably *awjō 'island, thing on the water', and
> <Brisgavi> is <Bris-gavi>, not <Brisg-avi>, quite possibly
> with second element OHG <gawi> 'Gau, Flur, Gefilde, Land,
> Gegend, flaches Land, Provinz'. (<Bris-> may be Celtic, but
> this is uncertain.) <Passau> goes back to <Batavis>, from
> <Batavi>. In the handful of place-names that I've checked,
> <-au>, when not of Slavic origin, appears generally to be
> from *awjō when it isn't part of <-gau>.

The name <Bris(i)gavi> (Not. Imp.) evidently refers to the Alamanni who settled in the Breisgau, where we also find the town Breisach am Rhein and the mountain Breisach. The latter was anciently <Bri:sia:cus> (abl. <monte Brisiaco>, Itin. Ant.; root-vowel /i:/ required by the modern form). The town was presumably *Bri:sia:cum. The suffix is Latinized Gaulish, but the root is problematic, as others have noted.

Under *bHre:i-, *bHri:- 'mit scharfem Werkzeug schneiden' (IEW 166-7), J. Pokorny lists Gallo-Latin <bri:sa:re> 'frangere' as ancestral to French <briser> etc., but the other Celtic forms he cites require short /i/. W. Meyer-Lübke reconstructs the Romance protoform as *bri:sia:re 'brechen' (REW 1310), which is required by Old Italian <brigiare>. He has these comments on the verb:

"Ursprung unbekannt, <brisare>, das einmal in einem Scholion in der Bedeutung 'comprimere uvas' vorkommt, paßt begrifflich nicht; ir. <brissim> fällt mit <ss> auf. Auch die weite Verbreitung des Subst. neben der engen des Verbums ist merkwürdig."

The scholion in question is to Persius, and dated to the 7th or 8th century: "Brisaeus pater Liber cognominatus ... videtur ab uva, quia uvam invenerit et expresserit pedibus (brisare enim dicitur exprimere)." The scholiast's verb is clearly derived from <bri:sa> 'grape-residue, debris from wine-pressing' (Columella), which is reflected as <brisa> in Aragonese, Catalan, and Valencian (REW 1307). Derivation of <bri:sa> from Greek <brútea>, <brútia> (Ernout-Meillet) is probably incorrect, given the restricted distribution of the Romance forms and the fact that Columella was from Gades. More likely <bri:sa> is from Celtiberian *bri:sja:, earlier *bri:usja: 'debris', the IE root being *bHreh1us- or *bHreuh1s- 'to break into pieces' (in Pok. as *bHreus-(2), IEW 171). Old English <bry:san> 'to crush, pound' requires Germanic *bru:sjan, with */u:/ from earlier zero-grade */uh1/. Whether the laryngeal itself underwent metathesis in some Indo-European branches, or the resulting long diphthong */eu:/ underwent transfer of quantity to */e:u/, later Celtic */i:u/, I cannot say.

Gallo-Latin *bri:sia 'breaking to pieces, crumbling, debris', cognate with the Celtiberian word above, is likely the immediate source of *bri:sia:re. Why the Breisgau should have been called *Bri:sia: is not clear. Perhaps the Gauls recognized loess, the principal soil there, as the product of erosion, hence a sort of debris. (All soils can be regarded as debris, of course.) The Gauls appear to have been preceded in this area by the Volcae, but attributing *Bri:sia: to Volcan does not improve matters. (Evidence is slim. I suspect that Breisgauer Alemannic <Kinz(ig)> 'steep-sided hollow way in loess' is a Volcan relic cognate with Welsh <pant> 'valley, dent', <i bant> 'away'; Middle Irish <céte> '(hill)path, bare hilltop, meadow', from a zero-grade base *kWn.t-. I cannot accept the notion of dialectal Q-Gaulish effectively proposed by T. (Geiger) Bynon, "Concerning the Etymology of English <path>" (TrPhS 64:67-87, 1966), to explain this word as Celtic, but her paper is very useful otherwise.)

Speaking of <path>, some time ago I argued in favor of Germanic *paþa- as a pre-Grimm's Law borrowing from Gaulish *bato- 'passed (over), (easily) passable', from PIE *gW&2tó-. Pokorny assigns some Insular Celtic words pertaining to death, Old Irish <baîd> 'dies', etc., to PIE *gWa:- (i.e. *gWeh2-) 'to go, come', on the grounds that dying is a going forward from the realm of mortals. In English, <pass> is used in a similar sense. This Insular specialization of the word was not necessarily shared with Gaulish. If indeed *bato- meant 'passable' in Gaulish, and was typically substantivized with a noun 'way, road' understood, perhaps *Baton was locally 'the Way' for reaching the shrine of Lugus at Lug(u)dunon. Then, when Germans settled the area around the distributaries of the Rhine (obviously after Grimm's Law, since <Leyden> has not been shifted), they would have called the road *Batam and the surrounding area *Bata-awjo:, *Bata:wjo: 'the Wetland around the Way'. (I am presuming that in such compounds at this stage of Germanic, the stem-vowel /a/ would have combined with the initial /a/ of *awjo: to yield /a:/.) This being Latinized as <Bata:via>, the back-formation <Bata:vi:> for its inhabitants would follow.

If such an analysis as a compound is correct, perhaps <Chama:vi:> has a similar origin, but with a native Germanic first element. West Gmc. languages have reflexes of both *xama- and *xamma- in the sense 'angle of the knee, ham, hough'. The latter can continue IE *konh2mo- '(shin-)bone', but the former is likely a different root. If *xama- meant simply 'bent, bend' (IE *komo-), its specialized usage as 'bend of the knee' could have been folk-etymologically conflated with *xamma-. But as 'bend of a river', *xama- would have originally resisted confusion with purely anatomical terms. A locality called *Xama-awjo:, *Xama:wjo: 'the Wetland around the Bend' would have been Latinized as *Chama:via, and again the back-formation <Chama:vi:> would follow.

One of the Amber Islands was called <Austravia>, apparently another compound of *awjo:. Unfortunately I have no citation indicating its antepenultimate quantity, and no evidence that its inhabitants were called *Austravi.

DGK