On 28/03/06, Henrik Theiling <[log in to unmask]> wrote:
> Hi!
> "Mark J. Reed" <[log in to unmask]> writes:
> > OK, pardon my ignorance, but what is this "vocalized r" you keep
> > talking about?
> >...
> In standard pronunciation, closed syllables in -/r/ will use a vowel
> [6], thus forming a diphthong, instead of [R], which is only used
> prevocalically.  I don't know what exactly are the phonemes -- it
> probably depends on what you want to stress and how abstract you want
> to be.  Probably there is a common standard for phonemes.  However,
> /6/ is probably not a phoneme, but /r/ is and [6] is an allophone of
> /r/.  It makes interesting pairs when you add endings starting with
> vowels that make a closed syllable open, e.g. for my dialect:
>   _leer_    /le:r/   [lE:6]   'empty'  (predicative or citation form)
>   _leere_   /le:r@/  [le:R@]  'empty'  (attributive e.g.
> (I treat /@/ as a phoneme here, because it's more clear, but maybe you
> could treat it more abstractly as /e/ in unstressed position, but I usually
> write short [E] that way.)

Wikipedia writes /E/ for short e and suggests that [e] exists, at
least, as an allophone of /e:/. Would you consider that an
exaggeration? (In any case, it seems to me that if a language has both
/e:/ and /E:/, and a short equivalent of both that's pronounced [E],
the logical phonemic representation of it would be /E/. Is there any
particular reason why you don't?)

> > As far as we were taught, |r| always means /R/ (or whatever) in
> > German orthography.  Replacing it with [6] is mighty strange,
> > especially from the perspective of a native speaker of a decidedly
> > rhotic variety of English.
> As Roger says, it's similar to [@] allophone of /r/ in non-rhotic
> English, I think.

In many accents (tho not apparently for most non-rhotic speakers) it's
not an allophonic alternation, probably largely determined based on
whether the standard dialect of the area is rhotic or not. Most
Australians and Englishpeople, I get the impression, wouldn't know
whether a [@] was spelt with or without an <r>. (Obviously
morpheme-internally, e.g. in "sURprise", this is because the r never
resurfaces; morpheme-finally many dialects have historically
unjustified intrusive [r\] inserted.) Does it ever get to the stage in
German where you'd be thinking that it's actually not just an
allophone? Does [6] ever mean anything else except for /(@)r/?

On the other hand me and most Australians use [6] or something like it
as an allophone of /@/ (regardless of its origin) in word-final open
syllables, at least in certain parts of a phrase (predominately the
end of it). In fact, the most common diphthongal pronunciation I have
of /I@/ is [i:6]. (I also find [e:6] a lot easier to say than [e@],
but I say neither; it's always /e:/ for me.) I have no idea how one
would go about articulating either [a6] or [a@], though.

I've heard from somewhere that (some non-standard varieties of) German
also have an "intrusive r"---unhistorical [r] ([R], [4], [r\],
whatever) inserted after vowels that a phonetically the same or
similar as ones that have an etymologically-justified following /r/
re-inserted. Is that common? Do you know anything more of that?

I was about to ask also if German [6] and [R] ever alternated in a
single morpheme when affixes were added---but you've already answered
it :)