Discovery!

Experimenting with the phonetic development of Watakassí, I discovered
something surprising.  Long ago, I'd developed a rule that eliminated an
unstressed /a/ before a glide.  Well, turns out this allows any
consonant to be followed by /j/, including /s/, /z/, /t/, /d/, /k/ -
these traditionally became /S/, /Z/, /tS/, /dZ/, and /C/ when followed
by /j/.  Also, /g/ could not have been followed by /j/.  I'm undecided
whether to make the palatization a productive rule, thus [saja] -> [sja]
-> [Sa], or to allow [sj], etc., to exist, making /S/, etc. undeniably
phonemes, even if with odd distributions.  There'd be the problem of how
to represent them in orthography, perhaps by s'ya (sya = /Sa/, showing
the (traditionally) underlying /sja/).  The native syllabry would
definitely fail to distinguish.

--
Dievas dave dantis; Dievas duos duonos
God gave teeth; God will give bread - Lithuanian proverb
ICQ: 18656696
AIM Screen-Name: NikTailor