Re: [tied] Re: Pretonic laryngeals in roots

On 2007-03-16 17:19, Sean Whalen wrote:

> Wouldn't *o > o: > a:? Greek has analogy with *
> pYròxW > pró (I'd say many branches had final a(:)x >
> a(:) and o(:)xW > o(:)).

A valid point, but wouldn't *proh3- have gone to *pra:-, had the
comparative been *proh3-tero-? Besides, Brugmann's Law regularly fails
before the contrastive suffix, see Ved. katara- : Gk. poteros <
*kWo-tero-. There must be a reason for that, but it's likelier to have
something to do with the suffix itself (e.g. *-t-h1ero- rather than just
*-t-ero-?).

>>> However, -idus is not from *-e-h1- + -to-.
>> There's
>>> no reason for the h1 to disappear
>> Why not?
>
> Well, it didn't in your other examples. If you're
> bringing analogy into play I can't make a certain case
> against it based only on that.

There are Greek words short vowels possibly reflecting older forms:
lútHron 'defilement' from *l(e)uh1-, áetHlon 'prize of contest' from
*hweh1-, báthron 'base, pedestal' from *gWeh2-. Compare the restoration
of the vowel of the lexical base (<star>) in Eng. starry, whereas the
regular development is that seen in <carry> or <baron>.

>> vi:vidus : ji:vatHa-
>
> I've seen no evidence that *exYt > etH > atH in
> Sanskrit. They are based on the same root, but based
> on other changes in Indic I'd say it's from
> * gWixWwa+xto+ 'given / possessed of (long) life' so
> the *axt > atH is regular.
>
> I'm not sure what you would predict for denta:tus,
> cornu:tus (that is, you didn't give any examples of
> xtV) so I'll wait before saying more.

The preaspiration rule is an old change, heavily layered over by the
results of later, productive processes (like e.g. Verner's Law or
i-umlaut in Modern English). It's the odd residue that gives it away,
not the productive derivatives.

> I'd say stative verbs in *-èxY+ had adj. in
> *-(e)xYdó+ meaning 'being'. So *xYruudhexYó+ >
> *xYruudheyó+ 'red', *xYruudh(e)xYdó+ 'being red,
> blushing' >
> rubeus, OCS ryz^dI; and Early Irish ruidiud 'blush'

In Latin they systematically correspond to adjestives in -idus (albeo: :
albidus etc.). Olsen treats them as plain deverbal adjectives in *-to-,
which seems to me an entirely reasonable and natural analysis.

Piotr