Glenn:
> It's also been raised that uvulars aren't preserved
> in later IE languages but then neither are laryngeals.
> Next!
Not really the same. I gather by preservation of laryngeals you mean
direct, consonantal, reflexes of it.
We have no direct reflex of the "uvular" - the only arguments for it are
frequent a-vocalism with it and the markedness. The markedness could be
explained in a different fashion (cf. for instance the classic
Illich-Svitychian explanation) and the a-vocalism is also inconclusive,
although quite possible. The problem is of course with the roots like
*kes- etc. (but that can be resolved in different ways of course).
With the laryngeals, the case is however very different and by equilizing
the laryngeals and the "uvulars" Glenn is seriously mistaken.
Laryngeals have, unlike the "uvulars", as we all know, left many many
indirect traces - beginning with the prothetic vowels in Greek, the
Balto-Slavic accentuation, the vowel colouring (*h2e > *h2a has no
exceptions, unlike the supposed *qe > *qa) etc. However, there are also
direct traces - _h_ in Anatolian from *h2 and probably *h3, #h- in
Armenian from *h2- and *h3- and, according to Beekes, te glottal stop in
Gatha Avestan. I'll leave the unclear cases of Albanian #h- out of this.
Thus, the case of the laryngeals and the case of the "uvulars" is not the
same even by a long shot.
Mate