I wouldn't say I'm exactly making a naturalistic language, making a 
language that I like is more important to me than realism - that being 
said, though, I do want to pay attention to some things that someone 
making a naturalistic language would need to pay attention to. I think 
my plan for dealing with this so-called problem is this:

I will allocate a certain number of each of the possible syllables for 
various purposes. I'll assign the syllables to the different purposes as 
I go, but I'll try to roughly stick to the numbers I decided on. These 
would be the different purposes:

- A purely grammatical word, (e.g. a preposition or a conjunction)
- A suffix or prefix that has some grammatical purpose (e.g. something 
that makes a word plural, or denotes gender)
- A syllable that has no specific meaning, but can be used either for 
self-segregating morphology, or as the "shift" or "alt" mentioned in 
Larry's post
- A word (usually a root word) with a non-grammatical purpose that has 
some meaning, like a noun, verb, adjective, etc.

I don't know exactly what numbers I would want to assign to each of 
these, and I don't think it's too important to stick strictly to them, 
but they can just be a rough guideline.

I will also probably end up with a lot of multi-syllable root words like 
David mentioned. I think I do want to have some compounding, but not too 

On 2/4/2018 11:29 AM, Alex Fink wrote:
> On Sun, 4 Feb 2018 05:54:20 -0800, Jim Thain <[log in to unmask]> wrote:
>> One of the things I found interesting about the Polynesian Languages,
>> is that the 'one' word usually has many meanings.
>> My dictionary of Hawaiian is full of words like:
>> Kele 1 NVS Watery, muddy, wet, swampy, greasy, fat, lush. Figuratively:
>> impurity....
>> 2 NVI To sail; reached by sailing; flight, sailing....
>> 3 VT To scrape cooked taro with 'opihi shell or spoon after peeling is
>> removed,
>> to clean it in preparation for pounding.
>> 4 Variant of kela
>> 5 N. A fresh water weed
>> 6 N. Any kind of wild taro
>> 7 (rare) N. Jelly jam
>> That is quite a lot of strangely different meanings for one little word.
>> There are others with even longer blurbs.
>> Context usually makes which meaning is meant clear.
> Linguists like to differentiate between _polysemy_ and _homonymy_.  A word with multiple translations is polysemous if there's really just one *meaning* there, but it's a broad meaning, broad enough that to translate it we'd use multiple different words; it is homonymous if the two meanings are unrelated and belong to the same phonological form by pure accident.  If you see a word glossed "1. hand, 2. arm", it'll be polysemous: the speakers think of the hand-plus-arm as a single body part.  If you see a word glossed "1. hand, 2. cod" it'll probably be homonymous.  It looks like some of each is going on in Jim's Hawaiian dictionary entry, but I don't know the cultural background well enough to diagnose it.  (Like everything in language there are edge cases, but never mind those for now.)
> If you, ATMunn, are making a naturalistic language -- see Logan's message about that -- with lots of very short roots, then there may or may not be sweeping polysemy but it's virtually certain that there will be a good measure of homonymy.  As Jim's saying, it doesn't cause the problems one might fear it would: the different uses can be told apart, especially when they are words of different parts of speech or otherwise different syntax (like a transitive vs. an intransitive verb, if your language sharply distinguishes those).  And you can pile a few more homonyms in there if they are (mostly) restricted to fixed collocations.  Think of English "keep at bay" or "bay horse": those two "bay"s are homonymous with (i.e. ultimately different from) the already existing three homonymous words "bay" (a plant, a water feature, an architectural feature) and don't really cause added confusion.
> And that's another thing.  If a naturalistic language has lots of short words, they won't all be primitive ultra-basic easy-to-derive-from meanings.  Some of them will just be random words which sound change has reduced to that length and that simply haven't gotten replaced.
> Alex