Re: Toki Pona survey

Quoting "Thomas R. Wier" <trwier@...>:

> From:    Andreas Johansson <andjo@...>
> > I wouldn't be so pessimistic. Just combine Khoisan phonemic inventories
> > with Georgian syllable structure, and you're looking at millions of
> > distinctive _syllables_ already. Waitaminnit, and I'll get my virtual
> > envelope ...
> >
> > Lemme see; CCCCCCVCCCC syllable structure, 30 vowels and 100 consonants
> > make for some 30*100^10=3*10^21 syllables already. That gives us many
> > times more four-syllable words than there are electrons in the visible
> > universe. Should take a while yet till we need that many words, I
> > should think.
>
> Actually, Georgian allows up to *8* preceding consonants
> (e.g. gv-prckvn-i-s 'He is fleecing us'), but strongly resists
> complex codas.  I'm fairly sure they allow no more than about
> 2 or 3 in a complex coda.  Thus more like: CCCCCCCCVCC(C). What's
> more, if you count harmonic clusters as single units for purposes
> of syllabification, then the number of possible onset segments goes
> down some.
I didn't mean to imply that Georgian actually has CCCCCCVCCCC syllable
structure. Quite likely, there's no Khoisan language that's got exactly 30
vowels and 100 consonants either. I was using the names as vague labels for,
respectively, liberal syllable structures and inflated phonologies.

                                                     Andreas