Re: [tied] Re: Latin barba in disaccord with Grimm's Law?

--- gprosti <gprosti@...> wrote:

> --- In cybalist@yahoogroups.com, Sean Whalen
> <stlatos@...> wrote:

> > This kind of reanalysis is always odd to the
> people
> > who haven't done it. If it was fairly late then
> > *treyeskWe but *trins, etc., wouldn't be as
> acceptable
> > as *kWetwores (since PIE changed the endings; the
> > beginning remained).

> I'm not sure I understand your second sentence. Are
> you saying that
> *kWe would have been analyzed as belonging to the
> preceding word in
> the case of *treyes kWe, but to the following word
> in the case of
> *trins kWe, due to differing syllable weights?

No, there's no certain way to say how reanalysis
will change a morpheme boundary.

The chronology I reconstruct (that would make it
more difficult to have *kWe connected to three) has
the numbers counted out to five (in bare stems), *kWe
added, the nom. plural forms of the numbers used in
counting (replacing bare stems), reanalysis.

With this series there's little chance an obvious
plural ending in *-es would get something added to the
end. Since there's no regular phrase or chant with
the accusatives, etc., there would be no *trinskWe.



__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com