On Thu, 06 Dec 2001 22:56:35, "Glen Gordon" <glengordon01@...>

>Miguel states that *N can't occur word-finally...

Not me, but the Uralicist Pekka Sammallahti.

>but if we really
>think about it, one is hard-pressed to find words with final *-t,
>yet it exists as a plural marker.

Nouns, adjectives and verbal roots almost always end in a vowel (among
the list of polysyllabic PU words given by Sammallahti, only käläw,
natiw, weNiw (wäNiw) [various in-laws] and täwiw (tewiw) "lung", end
in a consonant). So when Sammallahti refers to the fact that of the
PU consonans only /r/, /x/ and /N/ do not occur word-finally (or
word-initially), he's not talking just about lexical words, he is
including the evidence from grammatical suffixes.

Hungarian 1sg. -k can come from PU *-g or *-k, I suppose. Eskimo has
1sg. -ka, Aleut has -N (< *-k), PIE has *-h2 (< -k) (stative),
Afro-Asiatic has *-k (stative).

Miguel Carrasquer Vidal