I was trying to assign thematic categories to all the words in Klingon to see if that would help memorize them– i.e. by memorizing a set of related words instead of a random jumble. One rapidly notices that there are patterns. After categorizing 10%, words fall in pretty much the following categories, the top most of all, the bottom the least
Law, especially diplomacy
Technology & Future Technology
Space and Ships
Cloths and Fashion
Food and Drink (and sometimes imaginary food and drink)
Linguistics and Language
Medical & Health
Stuff that was hard to categorize
It occurred to me that a better balance could have been reached if meanings were assigned after categories to evenly spread out the words. And if people wanted to learn their words in thematic categories or benefit from that, then so much the better.
This applies mostly to closed lexicon languages. If a language’s lexicon is mostly open, then a peculiar initial distribution of words doesn’t matter.