Hi folks. I'm working with large size ideophone systems from natural
languages. I'd like to model some of the qualitative and qualitative
features of these systems to see how true they are to combinatorics.

Some languages have many thousands of these terms- Korean, Japanese,
Zulu, and Santali (a Munda language from N.E. India) have such high
numbers.

In many cases the terms are physicomechanical in connotation. For
instance /l/-initial ideophones in Santali seem usually to connote
smooth viscoelastic transfer of energy or masses capable of doing
this. Each phonological feature added to the mix delimits the meanings
so encoded. Add a terminal /k/ or /k'/ or /ng/, and a
vibrational/reflective sense is added. And intermediate materials
nuance the senses further.

The systematicity is diagrammatically iconic in nature, taking its cue
from the structure of the phonology itself.

In any case, without belaboring the specifics, what I want to know is
if any of you have used programs or languages that would allow one to
create a computational model of such relations- on both the
diagrammatical iconic side as well as the physicomechanical mapping.

Thanks much.
Jess Tauber
phonosemantics@...