Re: Wofir aka The Whorf-Sapir Hypothesis

H.S.Teoh wrote:
>On a related note, I've decided that my current conlang will have NO
>diphthongs. There is a (somewhat contrived?) mechanism of splitting a long
>vowel into two short vowels during inflection, and if the inflection rules
>happen to make both short vowels identical, they are "merged" back into a
>long vowel. But if the vowels are different, they are pronounced with a
>distinct glottal stop between them. Not sure how realistic this is,
>though... these odd rules came about mostly due to aesthetic
>considerations.
>
>Some of the descendant langs will, of course, lose the glottal stop and
>start acquiring diphthongs and the like. (In fact, the "merging back"
>phenomenon above could be a start of this process of dropping the glottal
>stop between vowels...)

Interestingly, when Kash was born, it had diphthongs /ay, aw/ at least in
word-final, and the syllabary had special symbols for them.  Then it changed
to alphabetic writing, with a rule forbidding two vowel symbols in a row, so
/ay, aw/  had to be written "ayi, awu", and for a while were pronounced as
two syllables.  But I LIKE diphthongs, so they've been creeping back in, and
now we need a spelling reform.......Simple enough to eliminate all y's
before/after /i/, since that's entirely automatic; but some cases of
/#wu.../ or /...uwV.../ arise from *b or maybe *v.  So for the moment, the
Committee is still in a stew.

In the Austronesian languages, it's far more common to separate _identical_
vowels with automatic glottal stop, and, like Kash, separate "iV" or "uV"
with  homorganic y or w glides, respectively.  But elision/lengthening of
like vowels ~ /?/ between unlike is certainly a realistic option too.