Miguel:
>> (The matter may be relevant to the topic of the labiovelars, given
>> that it's always tempting to link ziv(a)- "live, alive?" to PIE
>> *gWih3w-).
Me:
> I think that's a red herring. The expected reflex of IE *gWihWw- in
> Tyrrhenian would presumably be *kixwe in the 3ps, becoming Etruscan
> *cicHva.
Alright, let's hold the presses... Perhaps I'm being too hasty by
just dismissing this. Let's explore it. Perhaps /ziva/ can mean
'he/she lives' in the contexts it is found in and there is a Lemnian
reflex showing that it is a Tyrrhenian word. I believe we've discussed
this before. Ugh.
If let's say *gWihWw- relates to a Tyrrhenian form like *ciwa then
(based on Etr /ziva/ and Lem /zivai/), we would have to show that
labialized velars, like their labialized dental counterparts, regularly
change to *c- and in certain specific contexts. It doesn't work with the
IE *kWer- = Tyr *ker- equation so why then would *gW- be palatalized
here? We have to wonder what happened to *hW here too when it apparently
survives in the plural ending *-xo, unless perhaps medial
non-intervocalic laryngeals disappear. It apparently survives
intervocalically in Tyrrhenian *proxa 'brother' (Etr /ruva/) < ITyr
*bar-axWa 'little brother' (MIE *baraxWe > IE *braxter-; development
affected by *maxter- and *pxter-).
The only way I can see this working smoothly is if somehow after the
merger of *kW and *k:W to *kW but before the change of *kW to *k, some
instances of *kW became *tW (similar to *kWi- > Greek /ti-/ actually).
Would it be *kW before *i becomes *tW then? But it would be nice if
there were other examples of Tyrrhenian *c before *i equating with IE
*kW and *gW before *ei. Maybe this is worthy of closer examination.
= gLeN