Re: [tied] Lat. Rivus

From: Piotr Gasiorowski
Message: 15190
Date: 2002-09-06

The "per-ri:vus" etymology was offered offhand (and I think not altogether seriously) by Eric Hamp in a paper where, among other things, he criticised Reichenkron for etymologising this word with the "Thracian prefix" pa-. Hamp merely wanted to show that a so-so etymology was no great feat. The Albanian connection is of course preferable by far, if only because we kill two birds with one stone.
As to the ancient Albanian form, it must be *per-ron-, with *o from pre-Albanian *e: or *a:. I wonder what the root might be. A direct reflex of *h1rei(-n)- 'flow' is out of the question: it could only yield Old Albanian *rin (*e- or nil-grade) or *ren (*o-grade).
----- Original Message -----
From: m_iacomi
Sent: Friday, September 06, 2002 12:55 PM
Subject: Re: [tied] Lat. Rivus

Piotr Gasiorowski <piotr.gasiorowski@......> wrote:

> I'm aware of two competing etymologies. <pârâu> is usually
> connected with Albanian përrua (përroni) < pre-Albanian per-re:n-
> (Jokl's etymology. I don't know how he explains *re:n- -- a loan
> from Celtic?). It might also derive from per-ri:vus (Latin
> elements).

Never heard of that.

> The Albanian connection is rather attractive and -re:n- > -rîu
> (-râu) is a possible Romanian development.

Most Romanian scholars give this word as from substrate. In his
Ph.D. thesis, Gr. Brancus reconstructs an ancient Albanian form
*përren and a proto-Romanian (that means before dialectal split
of Aromanian) one p@...@u, or *p@...@nu [the word is attested even
nowdays with [@] ([ã]) instead of [â] in both positions]. Under
this form, the connection seems very plausible.

             Marius Iacomi