pronunciation of laryngeals; connected text in PIE

--- In cybalist@..., Miguel Carrasquer Vidal <mcv@...> wrote
(Subject: Re: Greek laryngeals):-

> PIE had, as far as we can tell by what they did to the vowels, three
> laryngeals: *h1, *h2 and *h3. It is not known exactly how they were
> pronounced, but some of the options can be narrowed down. *h1 must
> have been /?/ or /h/, a weak sound that disappeared everywhere.

Some people talk about two different H1's. If so, one was /?/ and the
other was /h/.

> *h2 was maintained in Anatolian as velar /x/ (or uvular /X/), and
> it's reasonable to suppose this was also its original value
> (pharyngeal/epiglottal /H/ is another possibility).

I think it was epiglottal /H/ as Arabic in e.g. the name /Muh2ammad/.
PIE epiglottal /H/ could easily have presented in Anatolian as the
German ach-laut, or similar, as I know from scuba diving in the Red
Sea and in the process having the (pleasure?) of trying to pronounce
Arabic laryngeals in the course of learning some of the native
language, sounds not in the previous languages of either me then, or
of the Anatolians before PIE-speakers invaded from the steppes.

Is uvular /X/ a fricativised version of /q/?

> *h3 is the hardest to pin down, and opinions vary between that it
> didn't exist at
> all (as defended here by Piotr), through that it was a labialized
> variant of any of the above (/?w/, /hw/, /xw/), or that it was a
> voiced sound (velar /G/, uvular /R/ or pharyngeal/epiglottal /3/).

I think it was /3/, the ayin sound. In my mouth at least, it tends to
o-flavor adjacent front vowels. If it was any other of the suggestions
here, it would likely have presented in recorded languages as /w/ or a
guttural more often. Compare the equation Latin `odi' = Greek
`odussomai' = "I hate", Arabic /3adu:w/ = "enemy", all from root
/3-d-w/, and Greek `a(w)e:mai' = "of a wind, to blow". Arabic /Hawa:?/
= "air", all from root /H2-w-?/, if you believe in an old connection
between PIE and Semitic.

What is progress so far in people writing connected text in PIE? I
have heard of Schleichter's fable about the sheep and the horses, but
any hope of legally discussion of it here is grossly handcuffed by
copyright rules.

> The most interesting thing about the Iberian script (not alphabet!)
> is ... <http://rabbitmoon.home.mindspring.com/asw/iberian.html>):

I tried to look at that page just now, twice, and it has gone 404.