Re: [tied] More on the crummy sanguis/asrk connection

From: Miguel Carrasquer Vidal
Message: 5071
Date: 2000-12-15

On Fri, 15 Dec 2000 01:41:57 +0100, Miguel Carrasquer Vidal
<mcv@...> wrote:

>>>>/asrk/. Whatever. Point is, the -k/-j- (which would be IE -*g^)
>>>
>>>Actually, it could be *k^, *g^, *g^h, *k, *g, *gh, *kw, *gw or *ghw.

Taking a closer look, I think the palatals can be eliminated (they
should have given *<ásrt.>). There are no voiced aspirates ever in
the auslaut in Vedic (or PIE), so the options are really limited to
*k, *kw, *g and *gw. We recently discussed the infrequency of *k and
*g in general, so the only realistic alternatives are *kw and *gw.

[For me personally, the only alternative is *-gw, given that in my
view **-kw had already developed to *-h3, and *-k^ to *-h2, in
pre-PIE. That's what forced the idea on me. I don't know why Pokorny
also reconstructs *esr(gw)].

=======================
Miguel Carrasquer Vidal
mcv@...