From: fournet.arnaud
Message: 58597
Date: 2008-05-18
----- Original Message -----
From: "Richard Wordingham" <richard.wordingham@...>
> In other words, /a/ /e/ /i/ /o/ /u/ /V/ (but) and /aj/ /ow/ /aw/ are
> phonemes in English because they appear in any context including C_nC.
>But some are rather lacking before a velar nasal in English. The
>sequence /eN/ is rather artificial in English (I take issue with J.D.
>O'Connor here, while using his evidence) - it occurs in _dreng_, as a
>borrowing from Old English, but the regularly derived form is dialect
>_dring_. Long vowels and clear diphthongs are rather lacking before
>/N/ - the only examples I can think of are the interjection _boing_,
>the mispronunciation of _Laing_ as /leinN/, and the _Oink! Oink!_ of
>pigs.
========
English /N/ is itself a very rare phoneme
with distribution holes.
So It's certainly not the right choice.
Arnaud
==========
>Would you argue the Classical Arabic lacked phonemes /i:/ and /u:/
>because they do not occur before consonant clusters, thus giving the
>untidy vowel phoneme set /a/, /a:/, /i/ and /u/?
>In short, I do not find your argument convincing.
> Richard.
=============
I don't understand your objection.
It seems to me that (classical) Arabic never admits a sequence
like v + C + C when the vowel is long, be it a: u: or i:
in other words
baHr is possible ba:Hr is not.
ibn is possible i:bn is not
Even when the sequence is v1 C C v2
v1 cannot be long.
I could not find any counterexamples.
MAybe your can explain what you have in mind.
What examples have you got ?
Arnaud
============