On Tue, 20 May 2003, Glen Gordon wrote:
>
> Your basis is patently opaque requiring a lengthy explanation
> as to what a vowel *o has to do with voicing, and what basis
> exists for a distinct phoneme **z, before any meat of the
> arguement can be discussed. You manage as well to completely
> ignore the etymology of the suffixes, automatically assuming
> that they MUST be ancient. (And lest I emphasize yet again
> that a rare phoneme in the commonest suffixes is weird, weird
> weird.)
Just this thing for the minute: How frequent a phoneme is /z/ in English
outside of endings where it is hyperfrequent? And what about /th/ (thorn)
which is the ending of ordinals (and used to mark the 3sg of the present)?
How well-founded is the principle that is being invoked here?
Even under the principle, you seem to have no objection to *-s as
something allowed to appear in IE endings. Now, as we go back in time, we
have no knowledge for most of the individual occurrences of sibilants
whether they were earlier voiceless or voiced. Therefore the belief that
an individual case of IE /s/ comes from a voiceless sibilant is just as
much an unfounded theory as is the belief that it used to be voiced. To
assess that question we need more relevant material; I have found some and
brought it into play, and its testimony is voiced for some, voiceless for
others, and non liquet for most.
Jens