> We're talking about a modelisation of PIE where unstressed
> vowels are only long while stressed ones are only short.
Arnaud
========
No, we're not. One of Miguel's posts includes such
reconstructions as
*ú:d-an > *wódr.
Brian
============
1. Which languages support u: ?
2. Macro-Comparison shows that
this root is ngut?-> H2w_d
This example is tainted.
Arnaud
==============
*pú:nt-ax-m > *pónth2m.,
Brian
=============
Which languages support u: in this word ?
Arnaud
============
both with stressed long vowels and unstressed short ones.
Brian
=========
I think we should first discuss the words that
may have a long stressed vowel
not from laryngeal
Arnaud
==================