On Tue, 6 Feb 2001, jesse stephen bangs wrote:

> A while ago, someone mentioned that prepositions do not ever govern the
> nominative case in languages that mark case.  Unfortunately, my conlang
> Yivríndil does just that, and so I says to myself, "This won't do.  I
> don't mind breaking a language universal every now and then, since they
> all have *some* exceptions, but this one was claimed to have *no*
> exceptions!  And I don't want to be the only exception out there, since I
> strive for naturalness in my lang."  So I did a little syntactic
> slight-of-hand and decided that prepositions govern the accusative case,
> which is cheating since *the accusative case is never marked*.  There was
> an accusative ending that survived in pronouns until a few hundred years
> ago (con-timeline), but it's fallen out of use.

Interesting.  :-)  Actually, I've been wondering how, if the vocative is
only marked on singular 2nd-declension masculine nouns (and even then not
always as in "puer" and "ager" type nouns?) in Latin, how you can call
all the other vocatives a case when they look just like the nominative.
But maybe they *were* marked and dropped out.  =^)  I really should ask
Prof. Nussbaum.  He apparently teaches a scary/cool variety of languages,
like Sanskrit and stuff.

And then again, to be fair, I should probably also be quibbling over
things like neuters being the same in accusative, nominative *and*
vocative.  :-p  (Any Latin pedants out there, I've only been in 3 weeks
of class so my knowledge is definitely incomplete!)

> Is this cheating?  And can anybody come up with a natlang counterexample
> to this language universal?  If not, I claim first dibs on the
> self-referential Jesse's Language Universal: "All language universals have
> exceptions."

Darned if I know.  I just bet someone's claimed that language universal
already, though.  =^)