Re: Comparative Notes on Hurro-Urartian, Northern Caucasian and Indo

----- Original Message -----
From: "Piotr Gasiorowski" <gpiotr@...>
To: <cybalist@yahoogroups.com>
Sent: Friday, September 26, 2008 12:14 AM
Subject: Re: [tied] Re: Comparative Notes on Hurro-Urartian, Northern
Caucasian and Indo-European


>
> On 2008-09-25 23:53, Arnaud Fournet wrote:
>
>> I don't understand how *h2ag^-tl-o could become a:l-a
>> Could you please detail the steps ?
>
> *h2ag^-tlah2 > *agsla: > *a(k)sla: > a:la. The changes here are regular.
> Cf. the diminutive axilla (like ve:lum/vexillum, aulla/auxilla or
> ma:lae/maxillae).
>
> Piotr
>
====================
I was expecting this chain.
I agree with "*agsla: > *a(k)sla: > a:la. The changes here are regular."

It's the first step that I disagree with.

To start with, I think such a thing like -tl- is impossible in PIE.
This cluster becomes -tr-.
Example number "three" *tel
Semitic tel > thel-eth
PIE tl- > tr-eyes
The root ter "three" is fictitious : there is about no word with the e or o
grade.

As regards the instrumental suffix ter/tel
the "alternation" tel/ter is in my opinion : *tel/*tr- (e grade/zero grade).
There is only one suffix with a phonological "problem".

The next point is (admitting tl can exist)
The syllabic cut in *h2ag^ # tlah2
I definitely can't see how t can become -s- in this _strong_ position.
Considering that tlaH2 (or tleH2 ?) is equivalent to CvvC (l being a weak
consonant)
A form like that should become a:ctra in Latin.

The traditionnal analysis as aks-lâ is better.

Arnaud