Alex Fink wrote:
> I was stumbling around SIL's webpage recently when I came upon this:
> It's part of the DDP, a suite of tools and resources for developing
> dictionaries for minority languages, so there's probably other good stuff
> there too.  But the list of semantic domains available at that particular
> page is what got my attention: it's a big comprehensive taxonomically
> organized picture of semantic space, with a list of questions for eliciting
> words in each semantic domain.  The whole thing has 1793 nodes in the
> taxonomy and 41889 English lexemes as examples.

Thanks for the link; that could be useful.

> Building your lexicon systematically according to some sort of taxonomy
> strikes me as a good idea, in that it makes explicit the various divisions
> of semantic space that your lang's lexemes are creating, especially if
> you're intending to have these divisions be coherently different to the way
> familiar languages do it.  But I've never done this myself, except in a few
> small domains, like colors.  Has anyone?
> Alex

The majority of the Minza vocabulary is arranged in categories. Some of 
the categories also have subcategories; the overall category of "space" 
for instance has subcategories of "directions", "dimensions", 
"measurement", "shapes". Many of the verbs fall into the "actions" 
category, which has subcategories of "planning", "causation", 
"creation", "motion", "transfer", "using", and "alteration". The 
category of "musical instruments" (itself under "music", which is under 
"arts") has a number of main categories such as "aerophones" and 
"idiophones", and some of these are further subdivided into smaller 
groups such as "double reeds".

It's tempting to use scientific classification for animals and plants, 
but I've been wondering if there's a better way for words of a language. 
It might be nice to divide "plants" into "trees", "bushes", "small 
flowering plants", and so on. But where do you go from there? You could 
divide trees into conifers and the rest, classify them by the bark 
texture or leaf shape, and so on. But I don't know much about trees; I 
hear a word like "elm" and know that it's some kind of tree, but if you 
show me a bunch of pictures of trees and ask me to point out the elm, 
I'd have to guess. So I fall back on the scientific classification, even 
though that puts trees into a bunch of different groups and mixes them 
up with smaller plants.

A while back I took the Minza category list and adapted it for the 
Lindiga vocabulary; since an earlier version of Lindiga was the initial 
vocabulary source for Minza, this approach worked well, and I've since 
put more development into the Lindiga category list that hasn't been 
reflected in the Minza vocabulary. Nimrina is also using a category 
system based on the Minza vocabulary list, although I don't yet have 
many actual Nimrina words. Other languages such as Yasaro have their own 
category lists. I think probably Tirelat or Ludireo were among the 
earliest of my languages to use any kind of classification system for 
vocabulary. I didn't do that back in the pencil and paper days.