Re: feminine, masculine and... ?

H. S. Teoh wrote:
>
> Hey, I like this idea of distinguishing between noun categories,
> more than just physical gender. I especially like the idea of
> having a separate grammatical gender for conceptual objects.
> Perhaps I should think about incorporating more genders into my
> conlang, since currently, it's based on physical gender, so most
> nouns are neuter -- it might be better to split up this rather
> large category of neuter nouns so that at least abstract nouns
> are distinguished from genderless physical nouns.
I guess when you mean physical gender you mean sexual gender...

Well, even if there is a relationship between gramatical gender and
sex in indueuropean languages, this relationship is more like: words
for females are usually femenine and words for males are usually
masculine, but femenine and masculine can be used to non-sexual
concepts or even for epicene concepts: "la persona", "la girafa", "el
guepardo".  IIRC, German has three genders: masculine, femenine and
neuter and many non-sexual concepts are either masculine or femenine.
Swedish developed a four gender: real, including most non-sexual
concepts that was originally masculine or femenine (an masculine,
femenine and real nouns are usually called en-words, against neutral
nouns which are ett-words: if you don't care about pronouns or
adjectives in -e, there are only two genders: en-words and ett-words).

When I was designing my last YANC (yet another new conlang), I was
planing to make a three gender system: animate, real and neuter, with
animate further divided in masculine, femenine, child, and epiceme.
Adjectives would agree in the higher division of genders.  All animate
nouns would be either declinable: horse-MASC: stallion, horse-FEM:
mare, horse-CHILD: colt, horse-EPI: horse; or be either masculine,
femenine or child.

Even non-sexual animates are either masculine, femenine or child: this
includes: low order animals, organizations, spirits, nations, computer
programs, things like fire, weather and running water, etc.

Inanimates would be divided in real or neutral.  Real would share some
case declessions with animates, and some rules would help
distinguishing between real and neuter in many cases, but most of the
cases it would be an arbitrary classification.

When I actually used the idea for Protoandinean, I simplified things a
bit (I guess a protolanguage could be as complex as any other natlang,
but I thing that a reconstructed protolanguage shouldn't be too
complex, unless you make too many arbitrary decisions).  First: there
is no further divisions in animates.  Second: inanimates are divided
in more semantical cathegories: count and mass nouns (all animates are
countable).

In conclusion: gender doesn't have to be semantical, but semantics
could help in classification (like masculine/femenine in IE).  and:
sex is not the only possible classification.

-- Carlos Th