> On Wed, 11 Feb 2004 14:43:01 +0000, tgpedersen <tgpedersen@...>plain
> >The /a/ that follows in Latin is the main argument why the PIE
> >velars must be different from the other two velar seriessounds
> No. The main argument is that we have the correspondences:
> "satem" "centum"
> c' k
> k k
> k kW,
> which point to three different sounds in the proto-language.
> This is no different from what we have in e.g. Romance:
> Latin Sardo rest
> i i e
> e: e e
> e e E
> Since Latin is attested, we know there were originally three vowel
> to account for the attested variation. We're not in a positionwhere we
> can conjure up a scheme where "/e/ is an artifact caused by theloan of
> closed /e/'s into a language where front vowels alternated *i/*e andI am glad that you agree with me.
> *e-/*E- in paradigms", etc.