Mark J. Reed wrote:
> On Fri, Nov 12, 2004 at 04:24:04PM -0500, Pascal A. Kramm wrote:
> > On Wed, 10 Nov 2004 11:58:24 -0500, Roger Mills <[log in to unmask]> wrote:
> >
> > >Fronting of velars before front V is also seen in the German ach-laut
> > >vs.
> > >ich-laut; both sounds presumably represent the same _phoneme_ /x/
> > >though
> > >they are phonetically quite distinct.
> >
> > No, they don't represent the same. Exclusively "ch" in "Bach" represents
> > /x/, while "ch" in "ich" always represents /C/.
> You're misunderstanding.  No one is arguing that the  *pronunciation*
> varies
> between [x] or [C] depending on environment.  Roger's statement was that
> he believes the underlying phoneme to be the same.

> I don't know enough about German to know if he's correct.
I don't either, but that's what I've always heard from people who do, or
ought to, know.

 But what
> would be required for him to be *incorrect* is more than the fact that
> they sound different.  The most direct indication that they're separate
> phonemes would be examples where the two pronunciations aren't in
> complementary distribution.

Yes. Are there any German words of the shape [FRONT vowel i,e,,, etc]
followed by the VELAR [x]?.  Conversely, are they any words of the shape
[BACK vowel u,o,a] followed by the PALATAL [C]? I don't think so, aside from
the one or two exceptional forms cited in the recent discussion of the
diminutive suffix "-chen" [-C@n], which, I gather, does occur exceptionally
after a few words with back vowel. As I understand it, back vowels are
supposed to umlaut (> front) before this suffix; but one or two
counterexamples involving a single grammatical particle, and a morpheme
boundary, do not prove that the sound is contrastive elsewhere in the

It seems most likely IMO that "-chen" simply has adopted a stereotyped
pronunciation with [C] regardless of what precedes it.