Active case marking (was Re: draqa syntax - help please?)

Marcus Smith wrote:
>
> Jörg Rhiemeier wrote:
>
> >"The man" is an experiencer here, the perception is something that
> >occurs to him.  This use of the dative is also used to express that
> >something did (or failed to do) erroneously, not intending it, as
> >in the famous _Na Turin dagnent Veleg mjeln_ example you probably
> >remember.
>
> Certainly.  I can see why you chose to do that.  I was just noting that it
> is unusual.  I regard that as a good thing. :)
>
> >I don't really do so.  It is just that the preposition _an_ governs
> >the agentive case.  Most Nur-ellen prepositions govern the objective,
> >but there are a few prepositions marking semantic roles which require
> >an animate entity, that govern the agentive.
>
> That makes sense.  It is quite common for languages not to allow inanimate
> datives (goals/benefactives/malefactives/etc).  It's true of most active
> languages I've looked at.  Sometimes they aren't completely banned, but
> speakers definitely frown on them.
The idea behind it is that those prepositions which involve some kind
of psychological experience in the entity governed by them (and thus
cannot be reasonably used with inanimate nouns anyway) govern the
agentive case.  The cases are also used to distinguish between the
two genitives.  Both genitives are marked by the same preposition,
_e_.  The possessive genitive is _e_ + agentive; it marks alienable
possession and also serves as a subjective genitive with verbal nouns.
The partitive genitive is _e_ + objective; it marks inalienable
possession or partial affection of an object, and also serves as an
objective genitive with verbal nouns.  The idea behind this is that
inanimate things have parts, but they cannot "own" other items
the way people own things.  The usage of the different cases for
subjective and objective genitives with verbal nouns should be obvious.

> > >       and the fact that the marking is on the noun.  I would
> > > certainly classify it as active.
> >
> >Is there any theoretical explanation why active languages are usually
> >head-marking?
>
> In a nutshell (email me privately for a longer story), the semantics that
> determine active marking is a property of the verb,
Certainly it is, and the distinction between active and non-active
verbs clearly exists in Nur-ellen, even if it primarily manifests
not in the choice of different agreement markers on the verb,
but in the choice of different cases on the noun.

Active verb: _linn_ "sing"

I   hin       linn.
The AGT.child sing
"The child sings."

Non-active verb: _lant_ "fall"

I   jin       lant.
The OBJ.child fall
"The child falls."

>       but case generally is a
> property of the structure of the sentence, or at least the combination of
> several individual elements besides the verb and nouns.
Yes.  My explanation of the active case marking system (when I designed
it, I didn't know that "all active languages are head-marking";
I also didn't understand the concept of head-marking back then,
and just did what I thought was a neat idea) is that animacy and
volition are properties of the noun.  For example, a computer
(inanimate) and a human (animate) might perform the same steps
in a calculation.  Both do the same thing, but only one of them,
the human, is aware of what he is doing, and (possibly) does this
volitionally.  The computer is programmed to do it, and deosn't perform
the calculation on its own behalf, but on the behalf of someone else.
Hence the usage of different cases.

>       (Note: active
> marking is not the same as case.  Active languages can also be accusative
> or ergative on their nouns!)
Yes; an example of ergative case marking in active language is
Amman-iar;
isn't case marking in Chickasaw accusative?

Active case marking seems to be extremely rare if it exists at all
in natlangs; hence I could not use generally accepted terms
for the cases, but had to find my own.  First, I called them
"nominative" and "accusative", then "ergative" and "absolutive",
even later "active" and "stative", until I settled on
"agentive" and "objective", terms I found in Fillmore's case grammar
where, however, they do not refer to morphological cases, but to
"case relations" which exist at a deeper level.

Jörg.