Re: XSAMPA for /ts/?

En réponse à Tristan McLeay :


>At the end of the English word 'cats', it's just /ts/ (and probably
>[ts], though someone out there probably says [?s]), simply being a
>sequence of /t/ and /s/. In the German word 'zehn', the t-s ligature
>might be appropriate for a phonemic transcription, but I'm not sure that
>there's a phonetic difference between [ts] (sequence) and [t_s]
>(affricative).
If I've understood my phonetics correctly (and since I've never studied
phonetics, I may well be wrong ;)) ), the difference between [ts] and [t_s]
would be in the length of the transition part between [t] and [s], as well
as in the length of the whole cluster. In [ts], we have two phones
following each other, so the transition between each other must be short.
Moreover, being two phones, the length of the cluster will be about twice
the length of a single short phone (approximately of course). In [t_s], on
the other hand, we have a single phone which has just a starting point
different from its ending point. The transition part is long, actually the
same length as the phone itself. And the phone's length will be closer to
the one of a single short phone than of a cluster of two phones.

Of course, all this is theoretical. In practice, the error bars can be
large ;))) . But I'm curious to know how much time a German 'z' lasts
compared to simple German phones. I'd guess it's about as long as a simple
phone, unlike the 'ts' in 'cats'.

Christophe Grandsire.

http://rainbow.conlang.free.fr

You need a straight mind to invent a twisted conlang.