Miguel Carrasquer wrote:
> The root, as I said, is *gW(e)ih3(w)- Latin vi:ta is regular from
> *gWih3-teh2. The developments are regular:
>
> Skt. Slav. Lat. Osc-U. P-Celtic Q-Celtic Grk
> *kW k, c k, c^ qu p p k(W) p,t,k
> *gW g, j g, z^ v b b b b,d,g
> *ghW gh, h g, z^ f, v f b g(W) ph,th,kh
>
>
>> We observe that the change is indeed gw > b.
>
> No we don't. We observe that the change is *gW > /b/ in Celtic and
> Osco-Umbrian only.
>
> For Latin, it's impossible that *gW > w went through a stage *b,
> because then this /b/ would have merged with /b/ from PIE *b (rare,
> but not inexistant) and /b/'s from other sources (*dw as in *dwenos >
> bonus). This doesn't happen: *b gives Latin /b/, and *gW gives Latin
> /w/. Otherwise there would have been no distinction btween bibo "I
> drink" and vivo "I live"
>
>
> =======================
> Miguel Carrasquer Vidal
> mcv@...
>
From the conntact area of Latin, from the geographic space it remains
interesting the Osc., Umbric, Celtic, Greek .
As we see from your draw, there is everywhere an gW > b, just Latin
makes here an exception where it is assumed PIE *gW > w. If indeed there
have been no stagium *b, it has to be searched. The omonimity shouldn't
be a problem here but I guess there is some work to do for showing the
transformations.It seems unlike that all the other languages in this
geographical space have had gW > b but just Latin did not .
If this idea is right then your analogy of "bibo" versus "vivo" will
imply that the "b" was still playing the alternance with "p" ( see the
form in iliryan and slavic), though there was no confusion . In fact for
"bibo" we have the root *pibo
which is selfexplaining about the alternance b/p and why it has been the
distinction between pibo= to drink and bibo= to live, later the one
becoming bibo and the another one vivo.
P.S.
I am sure you are convinced as well as me that Romanian "viaTa"
dialectal with shiboleth "g'iaTa" is exactly the form of PIE *gWiete (
see plural "vieTi" ) and this Root does not imply any difficultiy for
deriving from it the Rom. word.
In fact it is explained a lot trough this gW > b.
Latin "vesica", rom "bãSica", latin "lingWa", rom "*linba" > "limba" ,
etc.
for "venire" the root is too an *gWem- and the nord dacoromanian dialect
of Maramuresh ( I guess it was one with a predisposition for a stronger
satemisation lost the "W" in a very old stage and gave the satem form
with "z" gWem> g'en > zen = "a zâni"= to come because a change /vi/ >
/zi/ is too explain just this way I guess.