Re: Renfrew's theory renamed as Vasco-Caucasian

From: afyangh
Message: 50227
Date: 2007-10-08

--- In, "Patrick Ryan" <proto-language@...>
> ----- Original Message -----
> From: etherman23<mailto:etherman23@...>
> To:<>
> Sent: Friday, October 05, 2007 11:52 PM
> Subject: Re: [tied] Renfrew's theory renamed as Vasco-Caucasian
> --- In<>, "Patrick
Ryan" <proto-language@>
> wrote:
> > I put the following proposal to you. An early form of PIE did
have the
> > vowels *i and *u. However, when stressed these vowels
diphthongized to
> > *ei and *eu. Thus the zero grades represent the original
forms. In
> > this light verbal roots in *u and *i are easy to find since the
> > zero-grade represents the actual vocalic structure.
> >

I don't think this is a phonetic process.
It is a morphological process.

And it is much older that PIE.
At least the following languages share the same reinterpretation of
inherited vowels *i and *u as a-vocalic *y and *w :
- Semitic, Touareg, Egyptian.
In PIE, only (rare) isolated words not obviously linked with a
verbal root survived this reinterpretation.


> > ***
> >
> > In the verbal form of *CVC, *V is a stress-accented *é. *CéC is
> the commonest form of verbal root formation in PIE.
> Indeed.
> > Presumably, then, as a verbal root, *CVC should really be
> *C'VC.
> Yes.
> > If we assume that *V = *i as well as *e/*o/*ø, since the
> (only) verbal pattern calls for *'V, we should expect **CíC not
> which is better explained as *Cey + root extension.
> My proposal is **'CiC > *'CeiC.
> > Your proposal initially necessitates changing the verbal
> of *C'VC to *CVC(') whenever *V is *i or *u then secondarily
> all traces of this _verbal_ anomaly by emending *CiC(') and *CuC
(') to
> *CéiC and *CéuC.
> There's no need to reverse stress like this. Stressed **i > *ei,
> stressed **u > *eu (possibly through an intermediate stage,
> ***i > **ai > *ei, and sim.). This is somewhat like the Great
> Shift in English.
> ***
> Your proposal is possible, of course; one objection is, though,
that several of the derived languages have <í> which has NOT changed
to <éi> alongside actual <éi>.
> But, I was taught that the simplest explanation that will
suffice should be preferred.
> I think everyone would agree that <y> can occur as both <i> and
> My position is that *Céi is simply a positional variant of *Céy
before consonants or finally.

Your proposal necessitates the introduction of a new process for
data that can be explained without it.

A.F :

There is plenty of data, whether within PIE or outside PIE that show
that the assertion is wrong.
*i or *u never stop to be vowels during all stages of (Pre-/Post-)
PIE, even though a clear morphological pressure nearly pushed them
out of the vocalic system in most languages.

You cannot explain Latin i:du:s within the framework of standard PIE.
Nor Gothic trudan = English to tread
Nor Greek words : ikhthu:s, i-kn-us etc.

Clear PIE data obviously necessitate a new process to be explained.
The standard framework fails to explain these words.

> Patrick Ryan
> ***
> ******