Re: Bader's article on *-os(y)o

--- In cybalist@yahoogroups.com, enlil@... wrote:

> What you need to address is how a z-allophone of *s is actually
> "weaker" and why a _new phoneme_ is the only answer. So far,
> everything that you say can be replaced by an allophone of *s
> in all those positions and still yield the same results. It's a
> small change but one that doesn't present the problems that your
> theory automatically presents, such as the fact that **z would
> evidently be rare yet too ironically common as it would be the most
> commonly used suffix (ie: nominative) in IE! That simply makes no
> sense. With the z-allophone, we merely see *s in all stages, and
> it frankly makes sense why *-s, one of the most common phonemes to
> begin with in IE, would be used to create the most common suffix.

And how, pray, does this state of affairs differ from English? As
English /z/ mostly arises from a split from /s/, we don't necessarily
have an unsurpassable problem in deriving nominative **-z from a
demonstrative *se. (One can argue that what we see as **-s opposed
to **-z was in fact [c] and defer the argument as to whether it
derives from *tW.)

Incidentally, where is the evidence that **z was rare? With a
merger /z/ > /s/, what evidence can we expect to see?

>
>
> > I see no grounds on which this could be questioned.
>
> My solution with the more optimal z-allophone is the basis for the
> question. It solves questions that your theory doesn't address
> properly like that of the above while still acknowledging the
> alternation of *e/*o as indicative of original voicing of *s.
>
>
> > So thematic nominative *-o-s and thematic 2sg *-e-s form a minimal
> > pair quite some time back,
>
> Yes, but that doesn't guarantee that it is so in IE itself nor
> does this guarantee that *-s in both cases is anything other than
> *-s with allophonic variety.
>
> Since you're so keen on minimal pairs, you should present one that
> WOULD be of relevance to this debate. One with two instances of
> _*-os_ where one must in reality be *-oz and the other *-os.
>
> Ironically, it's the fact that you use the thematic vowel

differences

> as evidence for voicing that makes it impossible, based on that

alone,

> that this is anything more than allophonic voicing of *-s! For all

we

> know, we may have *-os [-oz] and *-es [-es] in the final stage

before

> fracturing. If the **z is predictable in these environments as you
> seem to contradictively suggest, it sounds like an allophone to me.
> It looks like the z-allophone of *-s can be expected after *o 100%
> of the time.
>
> So your minimal pair proves allophony, not a new phoneme.

No! If the vowels were always different, this minimal pair provides
no evidence of allophony! It simple shows that thematic /o/ and /e/
are different!

Richard.