Semantic Content of Grammatical Gender?

Hello all,

About 5 years ago, I started experimenting with grammatical gender in my
conlang, Angosey.  I decided that it would be neat to have a particle that
could change the entire meaning of a noun.  For example, one particle would
designate the noun as a physical object, another as an emotion, situation,
etc.  For example, the noun root "zirath" becomes:

                au zirath			eye
		al zirath			suspicion, appraisal
		ay zirath			stillness, watchfulness
		sa zirath			observation
		in zirath			vantage point
		tha zirath		        a far-reaching idea

The idea was to have a purely semantic change that would allow me to coin up
to six different meanings from one root.  This was meant to oppose gender
particles in French and Spanish, which have absolutely no semantic
relationship with their nouns unless the noun is an animal or a person ("la
table" is semantically no different than "le table").

Then, in 2006, I went to Tanzania and learned Kiswahili.  Kiswahili's gender
system is superficially similar to Angosey's, but I got really frustrated by
the way it was being taught.  For example, we would learn one gender
particle that referred to "fruits and liquids" only to learn that "car" was
in the same category, and so on.  I think it would have been a lot less
confusing to have just taught it as a grammatical (rather than semantic)
distinction, like "le/la" is in French.

I'm running into the same problem with Angosey.  When I coin a new word, I
have to decide what category a certain meaning should belong to, and it's
getting to the point where distinguishing via semantics is akin to splitting
hairs.

Has anyone else attempted a grammar that made strict semantic distinctions?
 Did you run into similar problems, and if so, how did you solve them?

My mind wanders further, into Sapir-Whorf territory.  If I learn a language
(or create a language) with strict semantic categories, does it affect how I
see the world?

Replies