adpositions and typological generalizations

[Andrew J. Koontz-Garboden]

Is anyone aware of work that has been done on typological
generalizations regarding the overt expression of semantic categories by
way of adpositions?  Specifically, I'm wondering if it's the case that
if you have an overt lexical item for adpositional category x (i.e.
locative), then you will always have one for category y (i.e.
directional), z (i.e. comitative), etc.  I am led to this question by
the observation that many creole languages are said to have
"all-purpose" adpositions.  So, in short, I wonder if there is some sort
of implicational hierarchy that might be at work, or has been claimed to
be at work.

In a similar, although slightly different way of looking at the
question, Clancy Clements has suggested that maybe the use of overt
adpositions may simply have to do with what info. is recoverable from
the verb itself.

e.g.    (1)  We go 0 home.
        (2)  We drive to home. (?-perhaps there is something with the noun
'home'          too; you can get We go TO the store, We drive TO the store, but
these           are strange with home)
        (3)  We are 0 home.
        (4)  We eat at home.

In these cases (where info is recoverable as in 1 and 3), there might be
a preponderance of zero marking.  From this perspective, what I would be
asking is in which cases is the PP NEVER zero, sometimes zero, most
always zero, always zero.

Andrew Koontz-Garboden

************************************
Andrew J. Koontz-Garboden
Department of Linguistics and
Department of Spanish and Portuguese
Indiana University / BH848
Bloomington, IN 47405   U.S.A.

agarbode at indiana.edu
http://php.indiana.edu/~agarbode/
************************************