complexity of scripts

Nik Taylor wrote:
> . . . they also showed an Atlantean city and the script seemed
> complex enough for a logographic system.  I wonder if those were
> just squiggles, or if there was anything to it.
I always wondered whether Chinese fantasy writers invent characters ...

I've heard that if you know the shape and sequence of the strokes, you
can uniquely(?) identify a kanji, because of the sequence rules; that
implies that you could thus encode a corpus of kanji, and use a Markov
process to generate new ones.  Has it been done?


--
Anton Sherwood  --  http://www.ogre.nu/