Re: R: Re: Greenberg's universals

Tim Smith wrote:

> This brings up a point that I've wondered about for some time.  Are there
> any natlangs that have indefinite articles but no definite ones?  In other
> words, where an unmarked NP is interpreted as definite, and has to be
> marked in some way to make it indefinite?  I've never heard of this, but
> intuitively it seems to make sense, since indefinite NPs are generally ones
> that are being introduced into the discourse for the first time, so you'd
> think that they might require some device to call attention to them.
Seems unlikely, because the definite category is the more "marked" one
semantically: "the house" (i.e., THIS PARTICULAR house) is a more
precise expression than "a house" (i.e., some unspecified house).  The
definite NP conveys more information than the indefinite NP: not only
that it is a house that is being referred to, but it is a particular
house.

But reversing this markedness order would be a nice idea for a conlang.
And it might be possible to evolve.  Imagine classical Arabic dropping
its definite article _al_ for some reason, retaining nunation
(indefiniteness is marked by a suffix _-n_; this is called "nunation" by
the name _Nun_ of the Arabic letter <n>).  Odd, certainly, but
impossible?

Jörg.