Lars:
> But supposing glottalization or other strengthening
> of articulation (p:) having developed from a
> previous state of the series, and, thereby
> eliminating the sound that would have been
> initial /b/, would this have left an empty gap in
> the series?
Yes, this has been Bomhard's position however he also
has published the idea that IE itself had ejectives.
I can't agree with that last part. I prefer to think
that if IE really had 'aspirate voiced' stops, it
had developed them out of formerly 'fully-voiced'
stops. So if *dH is fully-voiced (as with French 'd'),
IE *d on the other hand would be only half-voiced (as
with English 'd'). The lack of *b would be a lack of
a half-voiced labial stop. The fully-voiced labial
stop would be seen in *bH.
So that's a gap that kinda makes more sense if you
think of 'half-voiced' as a more marked feature
than 'fully-voiced'. Ultimately, I still think that
this all comes from a former system with ejectives
that lacked *p? or in which *p? was very rare.
> Are other candidates to be considered instead of h3?
Yes. We can simply have *p? merging with plain *p, or
it could become voiced with *b. I prefer the former
view.
> h1 = glottal stop?
This is the general view. I also think that mediofinal
IE *h1 was loosened to /h/ while initial *h1 was /?/.
> Has this previous history of the medial series been
> taken into account by those trying to link PIE to
> Kartvelian, Inuit, Etruscan?
Yes. My view is largely based on that of Bomhard. I
consider *d and *g to correspond with Proto-Tyrrhenian
*t and *k respectively (hence Etruscan /t/ and /c/).
I reconstruct Indo-Tyrrhenian 'fortis' *t: and *k: as
the antecedents for both branches.
Indo-Tyrrhenian *t: and *k: in turn would come from
Proto-Steppe *t? and *k?, situated in the Central
Asian plains circa 9000 BCE. Proto-Boreal merges the
Proto-Steppe ejective and voiced stops with the
voiceless plain stops, so *t? and *k? generally
become *t and *k (at least in initial position) and
end up as such in Uralic and EskimoAleut.
Proto-Kartvelian is, afaic, a branch that seperated
early from Nostratic. It keeps *t? and *k? unchanged
for the most part. AfroAsiatic also appears to
preserve the original 3-way stop contrast.
Summary:
Nostratic Kartvelian Ur EA IE Tyr
*p *p *p *p *p *f
*t *t *t *t *t *tH
*k *k *k *k *k *kH
*t? *t? *t *t *d *t
*k? *k? *k *k *g *k
*b *b *p *p *bH *p
*d *d *t *t *dH *tH
*g *g *k *k *gH *kH
= gLeN
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com