I am back to thinking about how one could best represent the plethora of concepts from the human perspective. Whether we should invent new words (which are basically arbitrary sound sequences, using the conlang phonology), or we should combine multiple simple words into terms?
The example that quickly comes to mind is human names, contrasting English names (John, Jim, Jane, etc.) with stereotyped Native American names (Black Horse, White Cloud, etc.). English names (from a layman perspective) are meaningless, while these Native American names are composed of meaningful words (though they lose their meaning when being used for identifiers like this).
What is easier to grok, or easier to use in conversation when you are talking to someone and they don't know the term set you are going to be using (like when you are teaching someone something new)? On one hand it seems easier to compose simple words into longer phrases (like the Native American example), or like we have "magnetic resonance imaging" instead of calling it "foobarbaz". In English we have a mixture, we invent new words like "tylenol" or "proton", which aren't part of everyday speech. But then we have like "dwarf star" or "computer science" instead of "xiamakan" and "isholang" sort of thing.
In English, animal names (bear, coyote, etc.), human names, place names (Africa, Asia, etc.), rock names (granite, lapis lazuli etc.), religions (judaism, islam, taoism, etc.), and many other things use words which aren't part of the basic lexicon. But other things like the examples I mentioned above are multi-word terms from simpler concepts. Often times the word in the native language means something simple though, like "jew" comes from "yehuda" which means "praised". So we have the "praised people" and the "submitting people" (islam).
What are the pros and cons of both approaches?
Like I mentioned, to me it seems easier to use simpler words composed into longer terms than to make up arbitrary words (even if the word technically has a useful etymology, for layman's sake it is a meaningless set of sounds). But then on the other hand, calling a "wolf" a "large tooth animal" and a cat a "pounce animal" might not be descriptive/identifying enough for practical communication. I'm not sure if you can go completely one of these approaches over another, or you need a mix.
Having arbitrary words/sound sequences like "proton" and "muon", not to mention the name of Chinese and Arabic named cities coming from English, means the size of the vocabulary grows to hundreds of thousands of words, if not millions. Which makes me desire using a small set of say 10k base words, and combining them into terms, of which you can get to millions of terms as well, but using simpler words as a foundation.
This is a big choice to make for a conlang, should you limit the size of your vocabulary to a smallish set of words (closed set, like 10k words, not talking toki-pona level minimalism), or should you allow arbitrary new words to be formed (open set)? If it's not an easy answer, how would you go about conceptualizing the choices to be made here?
One thought I had is to use arbitrary new words for the most basic things that you can observe (dog, cat), name, or recognize as a valid pattern (for example, mathematical "groups", calling them groups is confusing, but calling them "groupon" or something might be better). Then you can add more arbitrary words as necessary, but keep it to a minimum if possible, and instead opt for combining words into multi-word terms.
That would work for animals. So some animals get their own word (horse, dog, cat, deer, etc.), but others get built off those (hippo = river horse, giraffe = long neck deer, which are two Chinese examples). Things which it would help to think of them as their own pattern not based on anything else, you give a fresh new word (proton, electron, etc., but those could be called "positive particle", "negative particle" too, instead). Other concepts you map to multi-word terms. But I just am confused on how I can better think about what approach to take in the grand scheme.
With math jargon in particular, I find it particularly hard to understand new concepts because they invent new words. "Abelian group" for example, when they could just say "commutative composition groupon" or something. But commutative is also a complex concept, so it's like you would have an expression "groupon where you can combine the two elements in any order", or "order independent element combining groupon" :). That's where it gets tough, what would lead to better understanding in the end?
It seems that having a short term ("abelian group") would be best in some cases for learning, as long as there is a way to easily look it up and see "groupon where you can combine the two elements in any order" or some more primitive definition quickly. Not sure if there are other ways of looking at / solving this conundrum. Clearly, when you become versed in the topics, having a short reference is most helpful (abelian, for example). But it's definitely hard to get over the hump when you hear new words, that's for sure too.
Could you get away with calling an abelian group a "cloud nine stack"? Meaningless basic words used to reference it, would that be easier than a made up sound sequence?