From: Glen Gordon
Message: 17029
Date: 2002-12-05
>Vowels do not erode away just like that. [...] In these personalThe zero-grade of *gV is not **g. Are you being funny or are you
>pronouns, the particle was simply univerbated at an earlier stage,
>and became subject to zero grade. That's all.
>The demonstrative *to- remained productive and was even affixedThis has nothing to do with *ge because *ego: is fully attested.
>anew to nouns and verbs (Arm. article -d, Slavic 3sg. -tU?), but
>surely that doesn't mean that the earlier 3sg. verbal ending *-t,
>or, as you claim, the n.sg. ending *-d cannot have the same origin?
>Latvian and Armenian <es> show that *g^ stood in word-final position.... because the vowel eroded away in some stage, however early,