Re: [tied] Re: Syncope

Me:

> In our discussion of *wertmn, you wanted me to explain
> its derivation with the assumption that because it
> operates under normal quantitative ablaut that it must
> date to preSyncope times. With the examples of
> strive/strove and mouse/mice, we see that this is an
> empty assumption and not necessarily true.

Jens:

> No, we were speaking of *wert-mn as a type.

Type of what? Type of ablaut pattern?

> I do not see any reason to doubt the authenticity of
> *wert-mn.

Neither do I really with QAR at my side now. However I'm
just saying that *wertmn isn't necessarily ancient even if
it shows a regular ablaut.

> And I see very strong reasons to suspect the language had
> at least *some* words of that structure since it allowed
> this one to be created.

Bingo.

> If you mean lengthening when you write Quantitative Ablaut,
> I'd say yes, the nominative lengthening applied to an
> intermediate stage in the vanishing process e > o > zero,
> so that from unaccented /e/ we get the result /o:/.
> Did you see a need to tell me that?

Erh, wait. Nominative Lengthening (Szemernyi) has nothing
to do with quantitative ablaut but even so, I thought you
said that nominative *-s was actually *-z in the past and
that's the reason for lengthening, not ablaut.


= gLeN