Re: a case-free language?

I find this answer quite interesting, because its something obvious I
overlooked when I read the original post. :) From this point of view
then something which merges with the noun its attached to can still be
an adposition if it has scope over more than one NP (and isn't repeated)
in coordinated structures? For instance, if I had

-a- neutral
-i- subject
-u- object

and these were inserted inside the word then -i- and -u- would still be
adpositions if they follow your criteria (non-repetition)? Example:

t-n = man
f-r = woman
sa = and

tan sa fir (man and woman - subject)
tan sa fur (man and woman - object)
tin (man - subject)
tun (man - object)

This system satisfies the criteria you suggested for -i- and -u- to be
adpositions, but they're not what I would think of as adpositions, since
they're completely merged with the noun and can't be separated.

>
>The test of whether a language uses postpositions or cases is
>whether, in a coordinated structure, both conjuncts are obliged
>to show the form.  Japanese and Korea are said not to have cases
>because the postposition has scope over both conjuncts.  Some
>languages have both case and postpositions.  In Georgian, e.g.,
>postpositions subcategorize for certain case endings, even when
>phonological rules obliterate these endings:
>
>  (1) propesor-is-tvis [propesoristwis]
>      professor-GEN-for
>      'for the professor'
>  (2) kalak-s-ši [kalakSi]
>      city-DAT-in/to
>      'in the city'
>
>When we coordinate these, cases show up on the coordinand:
>
>  (3) ekim-is    da  propesor-is-tvis
>      doctor-GEN and professor-GEN-for
>      'for the doctor and professor'
>  (4) saxl-s    da  kalak-s-ši
>      house-DAT and city-DAT-in/to
>      'in the house and the city'
>
>Georgian is a particularly good example, since the forms are
>clearly cliticized to the NP, and yet equally clearly case form
>is being selected.  Incidentally, these kinds of tests show that
>one Northeast Caucasian language I studied last year, Lak, has
>46 cases, which is supposedly second in the world only to Tabassaran,
>also NE-Cauc, which has IIRC 50 cases.
>
>
>
>>Why is it that Finnish is described as having cases, while the related
>>Hungarian is described as not having any? Is it because Hungarian
>>affixes are added to an invariable word stem (except for -val/-vel
>>IIRC) while Finnish endings change the word stem?
>>
>>
>
>Most of the literature I've seen on Hungarian suggests that it
>has both case and postpositions. I'm not an expert on Hungarian,
>but in a book for which I just wrote a review on H. phonology,
>the author refers lists 17 cases (e.g. bokorban 'in the bush')
>and postpositions (a ház mögött 'behind the house'), although
>it's not clear to me that postpositions select cased-NPs like in
>Georgian (I simply don't know enough about Hungarian).
>
>==========================================================================
>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
>
>
>

Replies