Re[2]: [tied] Re: *a/*a: ablaut

At 2:07:13 PM on Saturday, February 16, 2008, fournet.arnaud
wrote:

> From: "Sergejus Tarasovas" <S.Tarasovas@...>

>> fournet.arnaud

>>>Sent: Saturday, February 16, 2008 1:49 PM

>>>It's anti-phonological to suppose that unstressed vowels
>>>can be long while stressed are short.

>> Lithuanian speakers have probably forgotten to ask :)

> You also have long stressed vowels. And short unstressed
> ones.

> We're talking about a modelisation of PIE where unstressed
> vowels are only long while stressed ones are only short.

No, we're not. One of Miguel's posts includes such
reconstructions as

*ú:d-an > *wódr.

and

*pú:nt-ax-m > *pónth2m.,

both with stressed long vowels and unstressed short ones.

Brian