Barry skribis:

>
> Peter T. Daniels wrote:
> >
> > ? Since the vast majority of humans have never been literate, it would
> > be quite difficult for a writing system to have any but the slightest
> > effect on its language.
>
> Uh, that doesn't make sense on me. I don't see where the conclusion
> follows form the condition.
>
> For example, that Japanese is spoken in a manner that each syllable is
> distinct and of the same length, it seems likely that a writing system
> would evolve that reflects that. Other systems are possible of course,
> and Japanese sure has tried to paste several of them on top of the
> syllabic systems, but still - why wouldn't a language's writing system -
> either designed or evolved, take into account what the speakers notice
> as distinct and important about how they perceive their language, such
> as a limited number of syllables?
>
> Other languages may need to account for tonal differences....
>
> So why *wouldn't* a writing system that postdates the spoken one reflect
> the spoken language?

I think there's a misunderstanding here. Mr Daniels did not say that a
writing system does not reflect the spoken language. He said that the
spoken language does not change to reflect the writing system. For
example, suppose the speakers of Trotzil adopted a writing system
where the symbol 'C' was used for both /k/ and /g/. Since almost all
speakers of Trotzil remain illiterate, we would not find that all occurances
of /k/ had changed to /g/ in the spoken language, even though they
are represented by the same symbol in the writing.

--Ph. D.