Re: Cases, again

Quoting Ray Brown <ray.brown@...>:

> On Wednesday, March 17, 2004, at 06:30 AM, David Peterson wrote:
>
> > Mad Martin wrote (I mean no disrespect--I think it's a cool accident
> > based on your e-mail!):
> >
> >> I've got another question regarding noun cases. In a situation where a
> >> language has Nominative, Accusative, Dative and Genitive, but no other
> >> cases,
>
> Like Modern German or, if we disregard the vocative, like ancient Greek or
> like Volapük (which enjoyed a spectacular tho brief success as an IAL
> before being eclipsed by the two-case Esperanto). I'm certain there are
> other examples.
I think Common Germanic had NADG? Icelandic has at any rate, and the other
North Germanic languages used to. Old English and Gothic too, IIRC.


> >> Or should there be a
> >> separate fifth case that is used in these situations, a sort of
> >> catch-all, miscellaneous case?
>
> If the 5th case is a catch-all to be used without adpositions in all
> situations where the other fours cases are not deemed appropriate, it will
> be too ambiguous.
Trebor's example of a one-adposition language nonewithstanding, a language
simply has to be able to distinguish between things like "he sat in the house"
and "he sat outside the house". If not by adpositions and/or case, you'll have
to come up with something pretty clever.

Possible solution for a one-adposition language:

he.NOM sat ADP interior.ACC house.GEN

he.NOM sat ADP area-immediately-outside.ACC house.GEN

The risk that the whole ADP-NOUN.ACC sequence will collapse into a single
specific adposition hardly needs pointing out.

Let's be a bit more creative:

he.NOM in-sat ADP house.ACC

he.NOM outside-sat ADP house.ACC

That is, verbal prefixes clarify. You wouldn't need any more prefixes than
English has prepositions - indeed, you could probably get away with
considerably fewer if some ambiguity is tolerated. You can also combine this
with German-style indication of directionality by choice of case for the noun
of the prepositional phrase, allowing you to get by without sep'rate prefixes
for "in", "into" and "out of".

> Ancient Greek, with Mad Martin's four cases, was a tad more interesting;
> it tended to do the following:
> - if the adposition* denoted "motion towards" it governed the accusative
> case;
> - if the adposition* denoted "motion from" it governed the genitive case;
> - if no motion was denoted, the adposition governed the dative case.
>
> For example:
> para + ACC. = to (the side of)
> para + GEN. = from (the side)
> para + DAT. = at (the side of), beside, near
>
> pros + ACC. = towards
> pros + GEN. = from
> pros + DAT. = at
>
> hypo + ACC. = (to a place) under [e.g. he ran under a tree]
> hypo + GEN. = from under
> hypo + DAT. = under, beneath [no motion]
>
> *usually placed prepostionally but sometimes, especially but not
> exclusively in verse, placed postpositionally. With disyllabic adposition,
>   the pitch accent was on the second syllable if prepositioned, but on the
> first syllable id postpositioned.
I did something similar in Yargish (originally inspired by some Caucasian
language, IIRC). Taking the noun _yuran_ "troll" and the postpositions _-
ja_ "at" and _-zay_ "behind" we get:

yuranu-ja  "to(wards) a troll" (dative)
yuraniz-ja "at a troll" (locative)
yurana-ja  "from a troll" (ergative)

yuranu-zay  "to behind a troll"
yuraniz-zay "(stationary position) behind a troll"
yurana-zay  "from behind a troll"

The locative case only ever occurs with a postposition - the dative and
ergative also with the independent functions you'd expect.

                                                       Andreas

Replies