Кодирование прямого дополнения в мокшанском языке
The work is devoted to the differential object marjing in Moksha lnaguage (Mordvin, Finno-Ugric). The DO in Moksha can be encoded in different ways: it can have a special definite genitive or possessive genitive marker on the corresponding NP or this NP does not have a special marker. The verb can have an affix triggered by both Subject and Object or only by Subject. The first (definite) mode is opposed to the second (unmarked) mode. The factor that is in focus is the referential properties of an NP in the DO position. The analysis of data has shown that it is not "the definiteness" od the DO that triggers the choice of DO encoding. The more complicated referential features come into play. I suggest the new rule for definite mode choice: there should be a set of entities that are discourse anchored (present in the world of discourse); the DO with definite encoding must refer to a certain subset of the first set.