Re: [tied] Nominative: A hybrid view

On Tue, 3 Jun 2003, Glen Gordon wrote:

>
> Jens:
> >Well, how *do* you explain that the singular of the stem of the numeral
> >'eight' exists in Kartvelian in the meaning 'four', if the IE dual form is
> >to have been a *plural* instead? Did 'eight' originally just mean "several
> >fours, be it eight, twelve, sixteen or higher"? How can sheer nonsense be
> >avoided here?
>
> It would seem that the "nonsense" here is the preconceived notions voiced
> in the question itself.
>
> The word for "eight" in IE is *okto:u while the Kartvelian form for "four"
> is *otxo-. They happen to look similar but we know that similarity on its
> own
> is fool's gold.

Why would a dual-looking word meaning 'eight' correspond to a singular
candidate-for-borrowing meaning exactly 'four' by accident? This is a fact
much stronger than the ones we do base our rules on. You may believe it to
be accidental, but you cannot demand that I follow you in that.

> If the Kartvelian form has somehow been borrowed from IE and that it
> shows that the IE form at that stage (whatever stage that may be, probably
> a late one) is being treated as a dual does not in any way prove that the
> oldest stage of Common IE, that common to both Anatolian and the other
> IE languages, had much of a dual number system.

If the borrowing means 'four' it must have been taken before the creation
of the numeral *kWet-wor-es, fem. *kWete-sr-es which is as old as anything
we can even begin to analyze (and then some). And if the borrowing has
been taken from a dual word meaning 'eight' as we know it in PIE, then
that must have been analyzable as a dual form, and so the dual category
must be that old.

> The dual system is not ancient. I'd estimate about a thousand years old by
> the time of Reconstructed IE at most.

On what grounds? How can the "bh-case" take the shape -bhya:m in Sanskrit
(corresponding with OIr. -b, a palatalized labial spirant followed by
nasalization of the next word), if the only analogical models are those of
the sg. or the pl. which are IE dat. *-ey, abl. *-es, instr. *-eh1,
dat./ab.pl. *-bhyos, instr.pl *-bhis? How did the gen.dual get the form
*-oHs, and the loc.dual *-oHu, and why did they take the variants
*-oyH-oHs, *-oyH-oHu if they only had the role models of the sg. and the
pl. to take after? (In the verb this gets even worse.) I am trying to keep
an unbiased mind, but the dual as an innovation just will not look
sensible to me.

Jens