THEORY: Deriving adjectives from nouns

Marcos Franco wrote:

> I have four kind of adjectives from a noun in UTL:
> 1. direct one:                  menta
> 2. direct agglutinated one:     mentoa / mentea (-o and -e are
> sing&plural noun endings respectively)
> 3. indirect (suffixed) one:     ment-al-a
> 4. preposition-compounded one: mento-pora, mento-kuna, mento-pera...
>
> As for their meanings:
> 1m. which is... mento
> 2m. related with... mento/mente
> 3m. of (de)... mento/e
> 4m. por, kun, per, (etc)... mento
>
> As you see, 1. and 4. are precise-meaning, though 2. and 3. are not,
> and I have still to decide some things, like:
> a. What's the exact meaning of preposition "de" (of) (which would
> affect 3.)
> b. Whether it should be better to have 2m. for 3. and and 3m. for 2.
> c. Whether it would be worthy to have a 5th adjective: genitive one,
> expressed by -y added to the noun, or to attach genitive's meaning to
> 2 (after assigning to it 2m to 3).
That seems like a lot of fine distinctions for what is usually
quite vague in my own speech. I don't know if many natlangs
make such distinctions. Even genitive (for me) is only usefully
distinguished from normal adjective by being able to take
more adjectives itself, as in "pretty little girl(s)(') school",
though many langs do distinguish different types of possession.
Also, how will you distinguish "painting crew from "painted bridge"?

Last time I tried this, I ended up with 6 kinds:
-a, plain vague adjective
-o, genitive, really a noun used as adjective
-ia before noun, active participle without object, "painting crew"
-ua before noun, passive/inverse, "painted bridge"
-ia after noun, takes object, "the crew painting the bridge"
-ua after noun, "the bridge painted by the crew"