Celtic Germanic relationship

[CONNOLLY at LATTE.MEMPHIS.EDU]

>From:	IN%"Indo-European at xkl.com"  3-DEC-2000 17:36:54.83
>Subj:	RE: Re[2]: Celtic Germanic relationship

>On Fri, 1 Dec 2000 01:52:23 -0500, "Brian M. Scott"
><BMScott at stratos.net> wrote:

>>>> Eg., the PrtGrmc word for "iron" is a Celtic loan, as are several
>>>> other terms (ruler and servant, for instance) and the form of these
>>>> loans indicates that they were borrowed before the first Germanic
>>>> sound-shift, since it underwent that change.

>>BJE> Do these loans provide evidence for all three consonant grades?
>>BJE> The only words that come to mind are *i:sarn- (iron) and *ri:k- (power),
>>BJE> which only provide evidence for PIE *g -> PGmc *k.

Miguel Carrasquer Vidal replied:

>>What about Goth. <andbahts>, OHG <ampahti>, ON <amba'tt> (specialized
>>to 'bondswoman'), etc., from Celt. <ambactus> 'Dienstmann' (here with
>>Latin ending)? Or OHG <Walh> from the Celtic tribal name <Volcae>?
>>They appear to offer evidence for PIE *k -> PGmc *h.

The former could as well be attributed to a long-standing synchronic rule of
Germanic: /-kt- -gt-/ -> [xt].

>On the other hand, a word like Celt. (*bhle:wos >) *bli:wos > Gmc.
>*bli:waz "lead" (German <Blei>) offers no evidence for *b -> *p (nor
>does *ambaktos -> *ambahtaz).

This one is thorny for several reasons: Celtic /b/ typically reflects PIE /bh/
rather than rare /b/, and it is far from clear whether (traditional) PIE voiced
stops and voiced aspirates would already have merged in (pre)-Celtic at the
time of the borrowing.  Nor is it clear that _Blei_ was actually borrowed from
Celtic: PIE _*bhleHiwo-_ (e-coloring laryngeal) would produce exactly the right
outcome and relate neatly with _Blech_ 'sheetmetal' < _*bhlHi-go-_, which
cannot be Celtic.

Leo A. Connolly                         Foreign Languages & Literatures
connolly at memphis.edu                    University of Memphis