Re: THEORY: Why more than two grammatical relations?

On Fri, 19 Oct 2007 09:35:57 -0400, J. 'Mach' Wust
<j_mach_wust@...> wrote:

>Thanks for the references.
You're welcome.


>So the terms seem to be simpler than what I searched for:
They may be.


>The difference between "core" and "oblique" denominates the difference
>between unmarked and prepositional phrases in English.
In English core arguments (or direct arguments) are expressed without
adpositions (unless the "to" form of the Indirect Object is a core argument),
and oblique arguments are expressed with adpositions.
But that doesn't hold for every language.
The lower on Keenan&Comrie's "Noun-Phrase Accessibility Hierarchy" an
argument is, the likelier each particular language is to express it by an
adposition.
Many languages have "accusative adpositions", yet their Direct Object is still a
core argument.  Some languages even have adpositions for their Subjects;
the "ang" in Tagalog may be both a definite article and a "nominative
adposition".


>I also get the impression that the term "grammatical relation" is better suited
>for a language such as English that doesn't have case marks.
I do not know whether or not that is the case; but languages like Greek and
Latin and German and many other European languages have usually been
analyzed as having grammatical relations.  On the other hand Sanskrit was
originally analyzed (by Panini) without them, though now, if my impression is
correct, most people analyze it with them; just as some linguists have
proposed analyses of Classical Greek without GRs.

>---
>grüess
>mach
Thanks.