Re: A tripartite case system

Back from a mini-vacation...

Trebor wrote (in effect):

<<What are some natlangs with tripartite case systems?>>

All of the Austronesian languages "of the Philippine type" (and that's a direct
quote, from...Schachter?). These three cases are given names which vary based
on the language and author, so it's hard to give them names. The functions are
roughly as follows:

(1) The "ang" case: So-called because this is the case marker of this particular
case in Tagalog, this case marks what's traditionally been called the trigger
(or final 1, if you're doing RG, or, for the latest, the topic-like external
spec. of TP [Matt Pearson, to appear]). Basically, it marks the focus of the
sentence. The focus of the sentence can bear just about any semantic role,
including (but not limited to): Patient, Agent, Experiencer, Beneficiary, Goal,
or some locative argument.

(2) The "ng" case: This case marks everything that's not an "ang". It can bear,
again, any semantic role. However, the verb "agrees", in a sense, with the
semantic role of the "ang" case, so there's no confusion.

(3) The other case: This is generally called the genitive case, because it
basically marks possession. *However*, it also marks the agent (not necessarily
the trigger) in so-called "pag-complements", which are kind of like subordinate
clauses, and would be roughly equivalent, in spirit, to something like, "*His
hitting of the ball* is good".

So that's one tripartite case system. One might also argue, though, that English
is a tripartite case system (nominative, nonnominative, possessive). I wouldn't
agree, but one could.

-David