Re: [tied] Ablaut and accent

On Wed, 23 Apr 2003, Miguel Carrasquer wrote:

> On Wed, 23 Apr 2003 18:16:06 +0200 (MET DST), Jens Elmegaard Rasmussen
> <jer@...> wrote:
>
> >
> >
> >On Wed, 23 Apr 2003, Miguel Carrasquer wrote:
> >
> >> On Wed, 23 Apr 2003 00:19:50 +0200 (CEST), Jens Elmegaard Rasmussen
> >> <jer@...> wrote:
> >>
> >> >Outside of the main debate: Armenian has final stress, but still accepts
> >> >words like tasn, eLungn, ezr, astL. Surely such a language should not be
> >> >alarming to Indo-Europeanists. It does not follow from that, however, that
> >> >the possibility was utilized exactly in the case under discussion.
> >>
> >> No, but it follows from the fact that the accusative singular is a
> >> strong case. If the desinence had been syllabic, it would have
> >> attracted the accent. It didn't, so *-m was not syllabic.
> >
> >I'd say that's a fair way of presenting it. Another could be to say there
> >are degrees in this: Full vowels attract the accent, syllabic consonants
> >do not. Do we agree that strong paradigm forms have flexives without
> >original vowels, while those of weak paradigm forms do have vowels in
> >them? I believe that is exactly correct.
>
> Your initial accent rule works in such a degreed way, so agreed.
>
> However, in the relevant timeframe, when those weak and strong
> flectional forms arose, (Pre-)PIE was a rather different language than
> the one we're accustomed to. For one thing, zero grade had not
> happened yet, so there were no syllabic consonants at all in former
> unstressed syllables. The only place where syllabic consonants may
> have been present is precisely in the endings of the strong nominal
> and verbal forms, and the inventory we have is the following:
>
> nom.sg. *-s (perhaps *-z at the time)
> acc.sg. *-m
> collective *-h2
> 1sg. active *-m (perhaps *-mW at the time)
> 2sg. active *-s (perhaps *-sW at the time)
> 3sg. active *-t (and perhaps *-s(W) in the preterit)
> 1sg. stative *-h2
> 2sg. stative *-th2
>
> I think that's it. Assuming *h2 was still fully consonantal, that
> leaves only *-m as a possible candidate for a syllabic resonant, at
> the time.

Good point, I agree completely.

Jens