Re: Def. of Case WAS: Cases, again

Date:    Thu, 18 Mar 2004 11:41:04 -0500
From:    "Mark J. Reed" <markjreed@...>

> This is clearly true since different languages have different cases, and
> often use different names even for cases that serve the same semantic
> role.  As you noted, English is generally regarded as not having an
> accusative (or dative or oblique) case; instead, it has a single case
> which is called the "objective" that is used where those other cases are
> used in other languages.
Somewhat of an overgeneralization.  In most English dialects, you
can say something like "John and me went downtown", where an
explicitly "objective" marking is used for the conjoined *agent*.
But this is entirely unlike closely related German:  *"Johann und
mich sind in das Stadtzentrum gegangen".

> Note also that there does not have to be a visible difference for
> two words to be in different cases.  In general, if *any* noun makes
> the distinction visible, then the distinction is considered to be there
> for all nouns, even when not visible.
This is the antistructuralist view.  Structuralists would say you
the distinction between the two cases is neutralized for almost all
items in the lexicon but the pronouns.

> Most languages group two of these roles together syntactically by using
> the same case for them, while using a separate case for the third role.
The reality is probably that all languages are sensitive to different
thematic role assignment in some subparts of their grammar.  Many
languages will do as you say, and group the S and A arguments for purposes
of case assignment of finite verbs, but behave differently for, say,
thematic role assignment in nominalizations, or for subordinated
clauses.  My recent research on Georgian suggests that it has an S/A
pivot in the present/future and aorist series, but something like a
Split-S pivot in the perfect.

> Some languages have a built-in hierarchy into which nouns fall, and the
> default roles are determined by their relative ranks.  In an animacy
> hierarchy, for instance, assuming "boy" is more animate than "cat" is
> more animate than "ball", "The cat hits the ball" and "The ball hits the cat"
> both mean that the cat is doing the hitting, while "The cat hits the
> boy" and "The boy hits the cat" both mean that the boy is doing the
> hitting and the cat is getting hit.
That's not quite the way it works.  Real hierarchical languages may
or may not have relatively fixed word-order, but they typically have
some kind of overt morphological realization of the atypical animacy
relation.  In Meskwaki (Algonkian), discourse also plays an important
role:

  wa:pam-e:-w-a       John-a            Mary-ani
  see-DIR-3Sg-Anim.Sg John-Anim.Prox.Sg Mary-Anim.Obv.Sg
  "John (prox) sees Mary (obv)"

  wa:pam-ekw-w-a      John-a            Mary-ani
  see-INV-3Sg-Anim.Sg John-Anim.Prox.Sg Mary-Anim.Obv.Sg
  "Mary (obv) sees John (prox)."

Here, the hierarchy goes {1, 2} > 3 Anim Prox > 3 Anim. Obv >
3 Anim Further Obv (if any, and only realized on verbs) >
Inanimate. Here, the going top to bottom you get the direct
series of thematic markers;  from bottom to top you get the
inverse marker.

 =========================================================================
Thomas Wier	       "I find it useful to meet my subjects personally,
Dept. of Linguistics    because our secret police don't get it right
University of Chicago   half the time." -- octogenarian Sheikh Zayed of
1010 E. 59th Street     Abu Dhabi, to a French reporter.
Chicago, IL 60637