Re: inverse constructions

>From: Charles <catty@...>
>Subject:      Re: inverse constructions
>Gerald Koenig wrote:
>> I find it interesting to speculate on what is the next corresponding
>> level of grammatical complexity for this model. It takes 6 more
>> grapefruit to form a supporting layer for the structure on my table,
>> giving a new pyramid of 10 objects total. On that theory, 10 is the
>> next natural number of cases. As we know, there is no shortage of cases
>> in conlangs. Is there a natural set of 10? One for each finger?
                        tok
Charles mu ja ki tok:  @====>*
                            ki
>If cases are equivalent to prepositions, why do the IE languages
>have so many of them, yet look with wonder on Finnish with its bunch?
I didn't know about this equivalence. I can see where genitive=of;
ablative=from; locative=at; dative=to; what are nom and acc?

>Anyway, continuing the idea of 3 core cases for, approximately,
>nominative accusative dative, there are 3 more common obliques,
>genitive ablative locative.
OK, good, that would give 6 naturals, 3 more would be required to fit a
packing theory. My error above, one of the 10 objects is the verb,
presumably the top of the tetrahedron. I'm not so sure the the model
scales up. Perhaps there is a shift to another structure giving a
series 4, 7 instead of 4,10. I'll take a look at other polyhedra. Or
perhaps the theory is correct and there are 3 more natural cases, I
don't know.  Does anyone have candidates? I notice that the obliques
can be construed as usual as spatials, perhaps the missing three also
can.

>
>But then they start splitting into infinite sub-types,
>or until exhausting patience. There one can start using verbs
>like "concerning". At this point, since my conlang had only,
>effectively, 2 core cases plus a genitive, I left it:
>no prepositions at all, just serial verbs and participles.
>I think it really does "work", but I feel uneasy about it.
Jerry mu ja ka tok:
                          tok
Jerry is about to say:  *@=====>
                         ka
It seems to me that if you are representing the semantics of cases,
or case roles, by other means, nothing is lost, and it's a matter
of taste what the number of grammatical forms that specifically identify
case are. Nik made the point as did you that English still actually
has cases although they are not so labelled. As far as cases are
classifiers I personally like an infinite set; but here I have been
speculating on what are the common "natural kinds" as used in
philosophy, if there are any such in case grammar.

Jerry