PYCCKOE B/HA
Danko Sipka
sipkadan at hum.amu.edu.pl
Wed May 29 06:22:27 UTC 1996
I would also be very interested in getting information
about available literature on v/na in Slavic languages
(not only in Russian). A student of mine is writing a
MA paper on how to formalize Polish <-> Serbo-Croatian
transfer of prepositions, and one of the things that
has to be done is to describe and possibly formalize
the differences in how these two languages use w/na:u/na.
Not only that Slavic native speakers have problems when they
learn other Slavic language, but they will also have
doubts in their own mother tongue. For example, S-Cr native
speaker will use either u po{tu/na po{tu when institution is
meant, or recently Jan Miodek, one of the leading Polish
normative linguists, has discussed the usage of na Uniwersytecie
vs. w Uniwersytecie, saying that Polish native speakers
frequently correct themselves saying w, and he prescribes na.
Serbo-Croatian usage can be described as follows.
u/na (spatial and pseudospacial)
|
|
----------------------------------
place institution/situation
| |
| |
-------------- ----------------
open enclosed regular exception
|
| U NA U
| ku}i poslu {etnji
----------- manastiru Fakultetu {koli
regular exception sobi Katedri ...
dvorani misi
NA U crkvi predstavi
trgu parku zgradi ...
ulici gradu ...
stadionu ...
livadi
predgra|u
selu
...
It seems to me that differences between Slavic languages are in
the regular/exception branching.
The place vs. situation/institution difference can be illustrated
using the following examples:
biti u crkvi na misi
biti u pozori{tu na predstavi
Using narrative convention, and including historical or foreign
language influence date, one could speculate on the fact that some
places are perceived differently from the other, like 'grad' (city)
is perceived more as enclosed than 'selo' village, or 'park' (park),
particularly the French one is perceived as closed object... But
all this will help neither the process of teaching nor formalization
for the purposes of NLP. This above would take only two yes/no decisions,
and two lists of exceptions.
Danko Sipka
More information about the SEELANG
mailing list