Paul Bennett skrev:
 > ----- Original Message ----- From: Benct Philip Jonsson
 > <[log in to unmask]>
 >> [log in to unmask] skrev:
 >>> No 'p' of 'f' in an IE lang for 'father'?
 >> There are a number of ways *p can be lost
 >> *p\ > xv) doesn't strike me as very improbable.
 > The path was
 > . *p?\='te:r
 > . --> pX='te:r (voicing assim)
 > . --> pX@'te:r (epenethetic @)
 > . --> px@'te:r (fricative collapse)
 > . --> px@'ter (V[+long] > V[-long] / _ {R,N} #)
 > . --> xp@'ter (TSV > STV / # _)
 > . --> xpU'ter (@ > U / V[+high])
 > . --> xpU'te (weakening of finals)
 > . --> xfU'se (frikhathivization)
 > . --> xfu'se (vowel collapse)
 > . --> xvu'se ("start darkening"[*])
 > . --> xvy'se (fronting harmony (or is that umlaut?))
 > . --> ' (initial stress)

So I wasn't totally off in assuming that the _v_ of _'xvyse_
to _*p_! :-)

Might you treat me to the reason why you think H2 was [?\],
and what, then, are h1 and h3? I may have asked before, but
in that case don't remember the answer. My own rather bland
hypothesis is h1 = /h/ h2 = /G/ and h3 = /G_w/. Mind you I'm
not fixed on the allophones at all, there might well have
been pharyngeal and/or uvular allophones, h1, or some
instances of it, might earlier have been [x] or [?], but at
a certain point Ch2 was able to give C_h but Ch1 wasn't. I
even have a hunch there might have been a whole set / ? h x
x_w G G_w/ but some of the distinctions were lost
everywhere, and some of them may have merged with phonemes
outside this group: *x_w > *[W] > *w. Note that none of
these necessarily corresponds to the h4 of Beekes et alii!
I'm a non-believer in epenthtic vowels, so at some point all
the H's in Greek need to have been something that could
actually become a vowel. I'm hard put to believe in any
synchronic system, except a very ephemeral one, having even
allophonic aspiration but lacking /h/, and h1 seems to me
the best candidate. H- dropping varieties of English may be
a counter-example, but even they have a faculative [h] in
every V-initial word, which even looks like a likely
situation for late PIE. The history of French illustrates
that [?], [h] and [@] may in fact all be allophones of a
single phoneme.

Last but not least may you treat us to the etymology of
_zörseňi_? Is it related to _*dyaus ph2ter_ at all save in
meaning? (not that I can't see _*dy_ > _z_, but I can't see
_xvyse_ in there!) And what's the matter with your
Arabization? It seems to begin in Waaw...

/BP 8^)
   B.Philip Jonsson mailto:[log in to unmask] (delete X)
"Truth, Sir, is a cow which will give [skeptics] no more milk,
and so they are gone to milk the bull."
                                     -- Sam. Johnson (no rel. ;)