Re: [tied] Re: Short and long vowels; the explanation of Old Indian

From: Miguel Carrasquer
Message: 39274
Date: 2005-07-17

On Sun, 17 Jul 2005 15:05:31 -0500, Patrick Ryan
<proto-language@...> wrote:

>Pre-PIE had three vowels: *e, *a, and *o.
>
>These vowels entered into combination with the 'laryngeal', *H, to produce *e:, *a:, and *o: in PIE.
>
>All pre-PIE short vowels became PIE *A, the Ablaut vowel/segment, which eventually had the manifestations *e, *o, or *Ø; which form *A took was a function of stress-accentual conditions.
>
>PIE retained pre-PIE *e:, *a:, and *o: unchanged.
>
>In many PIE-derived languages, the zero-grade of *e: was *e; of *a:, *a; and of *o:, o.

"Many" here to be read as "one".

>Indo-Iranian changed all PIE *e, *a, and *o to <a>.
>
>In open syllables, *o was further modified to <a:>.
>
>Indo-Iranian changed all PIE *e:, *a:, and *o: to <a:>.
>
>The zero-grade of Indo-Iranian <a> was <Ø> where possible; although various devices were used to ameliorate difficult combinations.
>
>The zero-grade of Indo-Iranian <a:> was <i>.

Only if <a:> was derived from *eH, *oH.
Not for <a:> from */o/ in an open syllable, nor for <a:>
derived from PIE long */e:/ (the zero grade of both is <a>).

>Indo-Iranian <a> + <y> was <e:> (/ey/)

No it wasn't. It was /ay/: OP daiva-, Avest. daeva-, Ved.
de:va-.

Therefore, /a:/ can never have been /ay/, or it would have
merged with /ay/ < *ei, *oi.

And what about Indo-Iranian /a:y/ (Vedic <ai>)?

=======================
Miguel Carrasquer Vidal
mcv@...