?
Причастные относительные предложения в удинском языке по корпусным данным
The paper presents quantitative data on the modifying participial clauses in Udi (Lezgic, Nakh-Daghestanian), based on text corpora. There are two participles in Udi, a perfective and an imperfective one; modifying participial clauses precede nominal heads, and the participle is clause-final within its clause. Like in other Nakh-Daghestanian languages, modifying participial clauses in Udi are close equivalents of relative clauses proper. However, as they allow a wider range of possible associations between a head noun and a clause, they can be rather assigned to what is known as general noun-modifying clause constructions (GNMCCs). The main goal of the paper is the analysis of frequencies of different associations between participial clauses and head nouns in terms of arguments, adjuncts or otherwise. In total, about 1,000 occurrences of participial clauses in the Nizh dialect of Udi were taken into account, drawn from three corpora: one spoken, one written comprising a translated text of the Gospel of Luke, and another written corpus comprising two collections of original folklore. The Udi data were compared to the data on relativization frequencies available for a few other Nakh-Daghestanian languages, including Agul, Archi, Lezgian, etc. The main generalizations which can be made from the counts are as follows. In participial clauses, intransitive predicates (especially the verb ‘be’) turned out to be more frequent than transitive ones. Relativization of the three core arguments S, A and P accounts for the vast majority of all occurrences (more than 80%), with the intransitive subject S by far outnumbering agent and patient (the same is true for the other languages of the family). Unlike in some other related languages, relativization of the agent is more frequent than that of the patient in Udi. Relativizations frequencies for peripheral arguments and adjuncts is small compared to some other languages of the family (less than one fifth of all occurrences). Among the non-core relativizations, the locative and the temporal ones are the most common. Also, the Udi data confirms the impression that although “extended” uses (i.e. non-syntactic associations) typical of GNMCCs are indeed attested, their frequency is very low.