From: fournet.arnaud
Message: 55917
Date: 2008-03-25
----- Original Message -----
From: Patrick Ryan
In my opinion, the earliest PIE word for 'bone' was *(n)ka(:)s-,
PR
========
I suppose you are aware a proto-form is not
a phonological shopping-list.
It can't be with ou without two things.
Arnaud
==========
corresponding to Egyptian <qs> (no 'laryngeal' - naturally long vowel).
The form *kost(h)- is a result of the addition of *-dh, a formant of bodily
parts, which has become *t(h) by contact eith the final *-s.
This is a case where semantic differentiation was _not_ affected by
shortening of the originally long vowel (-> *kast(h)-); and for the same
reason as above, became the Ablautvokal, *A: *kAst(h)-.
Accordingly, it could under circumstances appear as *kést(h)- or
*kòst(h)'V-. and even *kØst(h)'V-, which was simplified to *ost(h)-.
An extreme minority opinion but I though one or two readers might be
interested.
Patrick
===========
It's fairly hard to give some balance to PIE data
and it's not easier to deal with PAA data.
Anatolian hastai- suggests a velar spirant as initial.
So one root is *Hos-t-
Others PIE words suggest #kos-t-
PAA is k._ts.
This suggests the -s- in s-t-
is phonotactically unvoiced by -t-
in fact this is a "hidden" H2 *ts?-
This idea is conforted by
konH2-m
Greek knH2m-ê
H2 is here : ts?-
Now the initial in Host-
can't be voiced if we accept that
kost is k + Host > kost
The word is unfortunately not present in Skrt.
Hence we have *(k)-H1-o-ts?-t-
Skrt asth- is strange
it suggests a metathesis
H1ets?-t > ?est-H1- > asth-
Egyptian is not so clear.
q_s : in fact *k?is < *k?its-
it suggests the -?- was transfered
from kH1_ts?- > k_ts? > k?_ts > k?_s
written q_s
Strange. Requires more cases of that
kind of transfer.
Arnaud
==========