--- In cybalist@..., "ikpeylough" <ikpeylough@...> wrote:
> --- In cybalist@..., Piotr Gasiorowski <piotr.gasiorowski@...>
> wrote:
> >
> > From: richardwordingham
> > To: cybalist@...
> > Sent: Tuesday, August 06, 2002 9:57 PM
> > Subject: [tied] Re: Retroflexes in Sanskrit
> >
> > > /h/ has simple realisation rules :)
> >
> > > (a) Syllable start, except after last stress [h]
> > > (b) Last consonant of intervocalic cluster [h]
> > > (c) Otherwise [N].
> >
> > ... except in the position after a tense vowel, where it turns
into
> > [Z], as in beige, prestige, rouge, ménage, mirage, badinage,
> > garage, etc. :) _That_'s the funny thing: [Z] and [N] occur after
> > different vowels -- also medially, where, as you noted, the vowel
> > before [Z] is typically /e/ (measure) or a tense nucleus
(seizure).
> > The only near-minimal pair I can think of is azure : hanger (for
> > those speakers who have the "ash" vowel in both). It could be a
> > true minimal pair in Cockney.
> >
> > Piotr
>
> You guys didn't quite mention that seizure and singer are a minimal
> pair for some Americans (near-minimal for many others), since they
> not only have a tense /i/, but a long one. I don't know what that
> means except that phonological theories seem to work best at the
> idiolect level.
>
> IKP
I, and I suspect also Piotr, did not consciously know this
pronunciation. Perhaps, though, it's why 'sizzure' felt so natural
to me.
Piotr has analysed the medial distribution of /Z/ in terms of
phonetics. I think it's purely historical, with no synchronic
tendencies. After all, there are plenty of words where /Z/ follows
tense /i/, e.g. 'vision'. However, in a word like 'incisure' (if it
truly exists - I had to do a dictionary search to find an example!)
the vowel is realised as /ai/. It's largely spelling that determines
the medial distribution of /Z/.
Similarly, I think the distribution of word final /Z/ is purely a
matter of recent sources, though there is a general tendency to
convert final /Z/ to /dZ/. If words like 'zho' (rhymes with 'hoe')
were widely used (outside scrabble, where the variant 'zo' is far
commoner), I believe the /Z/ there might be stable.
Final /N/ seems to be a different matter. It's noteworthy that the
name 'Laing' rhymes with 'hang' (i.e. has the "ash" vowel). However,
there is the onomatopoeic interjection 'boing', often associated with
the sound of a large spring. (Interestingly, the spell checker I use
objected to it!) I'm not sure whether such words count. Apart
from /oi/, /N/ only follows short vowels - I hesitate to say tense,
because of names like Hong Kong. The origin is historical, but the
pattern seems to be maintaining itself. The historical tendency is
odd - /mb/, /nd/ and /ld/ tended to lengthen vowels, but /ng/ kept
them short. Words such as 'Laing', which seem Scottish, suggest that
there have been regional variations.
Richard.