On 2008-02-19 18:36, Patrick Ryan wrote:
> Unless I am misunderstanding your response, it seems beside my point.
>
>
> In the words above, I would consider the roots to be:
>
> *Ha(:)w- *Ha(;)l- *mey-
I try to base my opinions on observable evidence. There _is_ evidence
that the final *-s in *h2wegs- (*g, not *g^ -- sorry, I misspelt the
root in my previous posting) is an old suffix. Namely, *h2aug- is more
or less synonymous with *h2wegs-, and the same pattern (CeCC- : CCeCs-)
recurs in a few similar instances. By contrast, to justify the
segmentation of *h2aug- into *h2au-g- you'd have to show at the very
least that there are other roots of the shape *h2au-C- or *h2w-eC- with
a meaning similar to 'grow, increase'. Without such evidence there is no
case.
Piotr