[tied] Re: PIE Word Formation Q&A (1)

--- In cybalist@yahoogroups.com, Mate Kapoviæ <mkapovic@...> wrote:

>
> > Such forms could also easily be from earlier e.g. *wédors, cf.
> > *kwóns > *kwó:n. The point is, I don't see how we have all of the
> > facts just yet.
>
> Now, why would a collective have an *-s as an ending? Come on...

Then again, who's to say that they're necessarily collectives?

> > IIRC, Sihler also mentions that such Sanskrit forms are rather
> > enigmatic and difficult to reconstruct for IE, which is my point.
> > Furthermore, how does the *s come to intrude between the *n and
> > the *i? That's what I'm asking here.
>
> It's a mix of the old *vaca:s and the younger ending -a:ni (instead
> of -a:, where -ni is itself a younger ending based on the new
> *-nh2). The -a:- is already there, but it's hard to reconcile the
> -ni with the original *vaca:s (**vaca:sni would hardly work). Thus
> the -s- stays, the -n- appears as the anusva:ra after the -a:- and
> the -i is at the end. Phonologically, vaca:m.si is really vaca:nsi
> which and this -a:nsi is a mix of old -a:s and newer ending -(a:)ni.

You did not explain to me *how* the presumably inherited *s split the
apparent *-ni ending in two. How do you think it happened (assuming
for the moment, that it did)? Metathesis? Analogy? What?

- Rob