Re: Romanian APA

You've given me so much to respond to I'll try to go over the main
points (but not all) and reorder your questions.

--- In cybalist@yahoogroups.com, "alexandru_mg3" <alexandru_mg3@...>
wrote:

>
> --- In cybalist@yahoogroups.com, "stlatos" <stlatos@> wrote:
> >
> > --- In cybalist@yahoogroups.com, "alexandru_mg3" <alexandru_mg3@>
> > wrote:
> >
> > > In Romanian is apa , in Baltic too, so you need to increase this
> > > areal: a huge part of Europe is affected
> > >
> > > Marius
> >
> > I think the Proto-Baltic form was *upYa: < *xx,pY+ (the 0-grade of
> > *xaxpY+ < *xaxkYw+). The Romanian seems to be directly from Latin
> > aqua < *xakYw+ with kw > p and so not directly related to the PIE
> > alternations.

> Regarding ap- in Proto-Baltic
> - The Old Prussian form is with a- too not with u- so the Proto-
> Baltic form was with a- too (the dialectal split between OldPrusian
> and Lithuanian is 'a new one' in relation with a supposed *h2h2 > u
> in Proto-Baltic)

Lith. can have either a or u before P, so a rule such as a>u/_p
couldn't exist. However, u>a/_P in OP seems possible.

Baltic and Slavic have most syllabic consonants changing to a form
with the vowels u or i. After a plain velar x(), > u & C, > uC.

> - the Sanskrit forms are with a- too to have a PIE time hint

But it's strong a: vs weak a. In a syllable containing another
plain velar x>a (instead of usual x>&>i):

xWoLk-xY-n.o+ ... kYor-xks ... kYor-x-k+
xWoLk-x-n.yo+ ... kYor-xks ... kYor-x-k+
xWoL-xk-n.yo+ ... kYor-xks ... kYor-x-k+
xWoL-ak-n.yo+ ... kYor-aks ... kYor-x-k+
xWoL-ak-n.yo+ ... kWor-akWs .. kWor-x-kW+
xWoL-ak-n.yo+ ... kWor-akWs .. kWor-x-w+
xWoL-at-n.yo+ ... kWor-akWs .. kWor-x-w+
xWoL-at-n.yo+ ... kWor-akWs .. kWor-a-w+
etc.
aratni- ...................... ka:rava-
elbow ........................ crow

xY,-n. > x,-n.y
metathesis
x>a / _K$ or $K_
contamination (with kWerfn.o+s)
kW>w / x,$_
k>t / _$n.
analogical mix of irregular forms

> We have Romanian Dunare with u- again, versus Danuvius and the dan-
> root

That's metathesis of a: and u.

> So when the above toponyms 'were already there' the Romans
> arrived in Balkans and the Balkan Romance > Romanian finally has
> today APA as water-word 'like' in the above toponyms and like in the
> ancient PIE root ap- too...Pure coincidence? I will not bet on this...

It doesn't matter what words came from PIE in the distant past;
Latin is the source of most current words and kw>p is a simple and
certain recent change.

> On the other hand, in Romanian a word like codru 'seems' (to use
> your terminology) to be from a supposed Latin *quodrum (unattested) <
> Latin quadrum 'in the same time'... so something isn't clear enough
> with all these and especially with the timeframe of Romanian kW>p gW>b

There was simply kwo>ko beforehand.

> The main point here: is that Proto-Romanian 'shared' almost all
> the phonetic rules with Proto-Albanian, including that ones of the
> pair codru / kod&r but their is no trace of kW>p gW>b for Proto-
> Albanian...

Albanian is not descended from Latin; the relatively late rules that
inlcude kw>p aren't shared and have no need or reason to be.