Re: Case

OK.  A canonical "case system" exists when a language has:

1.  A closed group of categories, each one of which defines one or a
bundle of semantic roles (often very general) that a nominal can fill,
e.g. "thing-moved-towards," "topic," "beneficiary," "theme," "agent,"
"patient," "possessor," "thing associated with another nominal in an
unspecified way," and whatnot.  The bundle of roles that a particular
category indicates often form a radial category a la George Lakoff,
with a central member and numerous peripheral members.

2.  A set of markers that indicate which of that closed group of
categories a particular nominal word fits into.  This set of markers
may vary from nominal to nominal, and it may not always be sufficient
to nail down one of the categories in particular.  Suppletion to mark
case, esp. in pronouns, is a very canonical case phenomenon.

3.  Some form of agreement system, so that
adjectives/demonstratives/whatnot have the same case as the nouns they
modify and therefore display appropriate markers.

(I am here disregarding the neo-Chomskian notion of Case Filters and
all that rot because I personally have no use for Chomsky of any kind,
neo or otherwise, but YMMV of course.)

Now, I can't think of any way to establish the existence of the
categories mentioned in #1 without some form of marker, as mentioned
in #2.

In English we do indeed have markers/suppletion for pronouns (#2)
which fit into a small number of categories (#1), but nothing for any
other kind of nominal, and we have no detectable agreement phenomena
as in #3, due to the lack of inflecting adjectives or demonstratives.

Therefore, while it's possible to maintain that we have a case
system,  we don't have the full range of phenomena that a canonical
case system displays -- especially agreement.

All this is IMHO. :)

Ed Heil ------------------------------- edheil@postmark.net
"Facts are meaningless! You can use facts to prove anything
   that's even _remotely_ true!"           -- Homer Simpson

Padraic Brown wrote:

> On Wed, 14 Jul 1999, Patrick Dunn wrote:
>
> > On Wed, 14 Jul 1999, Nik Taylor wrote:
> >
> > > Kristian Jensen wrote:
> > > > Boreanesian gets away with just three cases, which is perhaps the
> > > > absolute minimum in any language.
> > >
> > > What about English's two?
> >
> > English has three by my count.  The accusative usually isn't marked,
> > except for pronouns.  So we have nom (I), genitive (my, or bob's), and
> > accusative (me).
> >
>
> It has four by mine.  I recall a few discussions with my Germanic Studies
> prof. several years ago, now, in which we concluded that English could be
> described as having the four standard Germanic cases (Nom, Gen, Dat, Acc).
> Only we don't have endings for 3 of them.  Granted, this is pushing things
> things a bit, but there it is.
>
> Jack likes Johns book, but instead gives Sue a dollar for hers.
> Nom        Gen   Acc                     Dat
>
> On the other hand, when learning English in grammar school, the word
> "case" never reared its head once.  Except as boxes into which we could
> store musical instuments, or things heard in court.
>
> I'm not sure what the minimum no. of cases can be.  I suppose one could
> argue that, for ex., Spanish nouns don't have case: when the need arises
> for specifying the nouns function within the sentence, you simply pick the
> correct preposition.
>
> Padraic.
>