*G^EN-

[Rich Alderson]

On 17 Jun 2001, Pat Ryan wrote:

> Let me clarify what I was saying. Even if we could trace all attested IE
> forms back to *g^enH1-, and no attested form could be derived from **g^en-
> (which I do not believe to be the case), I would still maintain that the
> non-attestation of **g^en- is a historical accident, and that **g^en- still
> must be reconstructed for some earlier date in order to provide the basis for
> *g^enH1-.

That is of course one possibility, but let's examine one other that is at least
as likely:

Let us suppose the existence of a stem extension **-n-, added to roots to form
a verbal stem, and a rule of "laryngeal metathesis" which requires that the
sequence **-HN- => **-NH- (where <N> here represents any nasal).  In that case,
the expected pre-form of reconstructed *g'enH1- must ultimately be **gVH1-n-,
with a root **gVH1.

So without further evidence for the reduced shape of the root, I propose that
we stick with the reconstructible form.

								Rich Alderson