Re: a new project of conlang

Christophe Grandsire wrote:

> GENITIVES, ATTRIBUTIVES and ADJECTIVES:
>
>         These are three different constructions whose purpose is to
> determine a noun with another noun, or the meaning of a root.
I am currently trying to disentangle adjectives from stative verbs,
without much success. As for genitives, I just run them together
as normal nouns "John horse shoe nail" like a Germanic compound
but with spaces to breate.

> GENDER:
>
>         Gender is neither the unproductive system that we find in French,
> nor the classes system of Swahili. It's rather a productive system of
> derivations (roots have no gender in themselves, roj meaning for example
> 'human' in a very broad sense) that can seem a little like Esperanto, but
> with a "gender-like" flavour. As gender is used for agreement between nouns
> and their determinants (pronominal complexes with antecedents, genitives,
> attributives and adjectives), I kept the name 'gender' to describe that
> system. A gender can have subgenders, and subgenders can have subsubgenders.
> The tree of this system is:
>
> - animated: letter k.   - human: k'a.   (- masculine: k'a-n.)
>                                         (- feminine: k'a-ti.)
>                                         (- group: k'a-se.)
>                         - animal: ki.   (- masculine: ki-n.)
>                                         (- feminine: ki-ti.)
>                                         (- group: ki-se.)
>                         - other (gods, extraterrestrials...): k.
> - inanimated: letter m. - plants: me.
>                         - part of animated: mi-k.
>                         - object: m'aj.
>                         - 'material' (used with ingredients in cooking): m.
>                         - pseudo-animated (fire, planets, earthquakes): m'a-k.
> - conceptual: letter j. - idea, art, doctrine (everything in -ism): jer.
>                         - abstraction (of something concrete): j'a.
>                         - quality (in a broad sense): j.
>
>         I borrowed this idea from Carlos Thompson and adapted it in my way.
> He picked my curiosity with his idea of subgenders.
Using genders that way is cool. What do you do to distinguish
a habitual (not cases) "employ-er" from an "employ-ee" etc.?
I might use noun classes (genders) for that sort of thing.