Re: IE Roots

--- In cybalist@yahoogroups.com, Miguel Carrasquer <mcv@...> wrote:

> On Tue, 19 Aug 2003 23:00:46 +0000, etherman23 <etherman23@...>
> wrote:
>
> In the vast majority of cases, three laryngeals (*h1, *h2, *h3)
> suffice, and there is little disagreement about which roots contain
> which laryngeal.

Probably because these laryngeals are mechanically inserted into the
reconstructed forms.

> There are also some unresolved issues relating to the voicing and
> aspirating effects of laryngeals on neighbouring consonants, and the
> reflexes of *h3- in Hittite (sometimes h-, sometimes 0-), which I
> think do point to more than three laryngeals at some stage prior to
> PIE.

I'd be more likely to accept the laryngeal theory if a solid link
with PAA could be found (which to me is not a crazy idea).
Incidentally, are there any known languages which have syllabic
variants of the laryngeals?

> The labialized laryngeals **/hW/, **/?W/, **/xW/ and **/XW/ all
> give *h3 (o-colouring), but in Hittite we can perhaps distinguish
> between those that are reflected as h- (**xW, **XW(?)) and as 0-
> (**hW, **?W, **XW(?)).
> "Voicing *h3" would reflect earlier **/?W/ (but, alas, there seem
> to be no cases of "aspirating *h3" as we would expect in the case
> of **/hW/, **/xW/ or **/XW/).

It seems I've lost count somewhere. How many laryngeals (including
allophones) would you reconstruct? IMO, when we reconstruct a lot
we're starting to violate Occam's Razor to an absurd degree.