Re: Self-segregating morphology again - in simpler terms, with list of methods

> > 5. Require the last segment of each morpheme to code the length of the
> > morpheme.
>
>Maybe theoretically possible, but seems rather perverse.
>And maybe pointless.  It seems like you would have to
>combine it with some other scheme that limits which
>segment can occur where, or else you wouldn't know
>where the last segment is so you can count backwards
>from it to find the last whole morpheme....
Well, you could start parsing from the end of the complete utterance. (Which
could lend itself to some very nice "garden-path" sentences, but it wouldn't
be otherwise very usable.)

Yet another idea could be to mark the *center* of each unit. This would
probably work better on the word level than the morpheme level... Two
marking systems are probably needed, one for words with an even number of
syllabes, and another for words with an odd number of syllabes.

Marking some other arbitrary part of each word, like the penultimate
syllabe, would work too, except short words would turn out problematic then.

Also, I don't think restricting the number of syllabes has been mentioned
yet? Gives a bit more space to work with than restricting the number of
phonemes. There's the issue of consonant clusters, but I imagine it wouldn't
be too hard to work around this - just ban clusters than can occur as a
combination of valid initials and valid finals. Of course, if there's a
limited selection of valid initials and valid finals in the first place,
that would also lend itself to the much more wieldy end/beginning marking.

And to further make sure all possibilities have been brought up already - it
might be included in your definition of "phonological segment", but
allophony's always a good marker in case the phoneme inventory starts
looking a bit empty after enforcing all the self-segregation rules. It's a
bit hard to think of good allophones for sonorants, however. Altering
phonation gets old eventually.

John Vertical