Miguel:
>I certainly didn't invent it. *-oh3 is a standard reconstruction
>of the PIE dual.
If there is no evidence for *h3 attested, then it can't be a valid
reconstruction, whether you deem it "standard" or not.
>As I fully expected, the point I was making has not been addressed.
That's because there is nothing to address. We are talking about
Pre-IE. The Common IE dual is in *-o:- (*h3 is not attested). There
is nothing about a Pre-IE *h3 that explains the presence of the
later *-u any better than a long vowel, or hell, even *u itself!
Your antilogical longing for unmotivated complexity is transparent
and very vexing.
Reject if you will (and you no doubt will because it's not your idea
and you like to reject ideas that aren't your own) but I personally
explain the *u as the by-product of final *-o: -- That's it. Very
simple. Ultimately, yes, I strangely agree with you that the *-o:u
ending in *okto:u and *dwo:u was once *-a-xe (with a laryngeal,
although technically *h2). Here, *-xe is a collective (otherwise
becoming inanimate collective *-x and animate, later feminine,
*-ax) attached to a stem ending in *a.
Hence the following scenario starting with Mid IE:
*t:Wa-xe/*kWetWa-xe "two/eight"
> *dWax/*aktWax (unstressed vowel loss)
> *dwa:/*aktwa: (uvularized *x > NULL)
> *dwo:/*okto: (vowel shift: *a > *o)
> *dwo:u/*okto:u (final *-o: > *-o:u)
By the way, *h2, while pronounced [h.] in Common IE, was
originally [h] with a uvular allophone [h.] next to *a. On the
other hand, *h1 was originally solely [?], only later becoming
[h] mediofinally. In the end, there are still only three laryngeals
that we need concern ourselves in IE.
>One more point: why was the PIE dual *ok^toh3 borrowed
>into Kartvelian as *os^txw?
The Kartvelian form is the word for "four", which is thought
to be borrowed from IE *okto:u "eight", although we are then
forced to assume that Kartvelians knew the IE word was a
dual, while strangely treating it like a singular. The sibilantization
of plain *k is a typical feature of satem languages and suggests
that, if it is borrowed, it is a late post-IE loan. While you'd
leap to the conclusion that is certain evidence of a laryngeal
in IE, there is the nagging concrete possibility that it was merely
IE *t that was interpreted as *tx in Kartvelian, without there ever
being a laryngeal present. Since the jury's out here then, you
can't use Kartvelian as firm substantiation of your otherwise
empty claim.
= gLeN
_________________________________________________________________
MSN 8 with e-mail virus protection service: 2 months FREE*
http://join.msn.com/?page=features/virus