On Wed, 10 Nov 2004 20:23:25 +0100, Andreas Johansson <[log in to unmask]> wrote:

>Quoting Andreas Johansson <[log in to unmask]>:
>> But I'm slightly worried to (re-)discover that the X-SAMPA document
>> claims that Polish _czy_ and _trzy_ are a minimal pair for [tS1] and
>> [t-S1]! CXS could write [tS)1] and [t-S1] to be entirely unambiguous. How
>> anyone can tell those apart in connected speech rather boggles my mind,
>> however ...
>Lest someone accuses me of not being exhaustive, I guess I should add that
>X-SAMPA is also able to denote the affricate pronunciation unambiguously -
>by [t_S1] - but that this usage is not recommended, the affricate
>interpretation as default being prefered.

Because of this reason, it'd be better to mark the non-affricate. This'd
require a sign for "audible release", kind of the opposite of <_}> (sign for
"no audible release").

The minus-sign might just work for this purpose, though it's no standard

In any case, this is rather a flaw of IPA than of CXS.

j. 'mach' wust