Print

Print


Raymond Brown wrote:
>At 8:09 pm +0000 29/4/02, Andreas Johansson wrote:
> >Raymond Brown wrote:
>[snip]
> >>
> >>Not necessarily.  In fact both modern Greek and Gaelic have /x/ with
> >>allophone of /C/ before front vowels, and a separate phoneme [G] which
>is
> >>/j/ before front vowels.
>
>OOPS! I got the delimiters the wrong way round in the second part of the
>sentence.  I should have written:
>"and a separate phoneme /G/ which is [j] before front vowels".
>
> >Yes, but the Greeks at least keep the voicing intact in these variations.
>
>So, indeed, do the Gaels.
>
> >[x] and [j] as allophones of the same sound would be pretty extreme I do
> >think.
>
>I wasn't aware that this had been suggested.  I think there must have been
>some misunderstanding.

Well, it all began with you mentioning the trouble of finding IAL friendly
values of |c q x|. I suggested (among other things) that |x|=/x/ would be a
good idea; [x] being quite common around the globe. Furthermore, I suggested
that if [X] and [G] were allowed as free allophones, even fewer people would
have trouble with it. To this you responded that [G] might be bad cause it's
likely to >[j] or even >zero in some postions. I replied that depending on
what other phonemes are around, neither is necessarily a problem, altho' [x]
and [j] and zero would be pretty weird as allophones of the same phoneme,
and likely to cause a split.

All this could, of course, be avoided by decreeing that /x/ always must be
voiceless. An allophone [C] before front vowels might be trouble if you have
a sound like /S/ or /C/ (which I suggested for |c|), but is naturalistic
enough.

                                        Andreas

_________________________________________________________________
Join the world’s largest e-mail service with MSN Hotmail.
http://www.hotmail.com