ИСТИНА

The preposition as a part of speech is quite difficult to determine due to its blur borders outside the central part: nobody knows, for example, how many prepositions there are in Russian (in Russian grammar 1980 210 prepositions are named, in Burtseva 2010 – 450, in Lepnev 2010 – 578, in Vsevolodova etc al – more than 1000). Of course, this mostly refers to complex prepositions which constantly evolve in language during grammaticalization (in its wide sense): poseredine × po seredin-e, po povodu × po povod-u, po tipu × po tip-u etc. These new prepositional units are still on their path to prepositions and thus are hard to define unambiguously as prepositions even although some indicators for distinguishing between complex prepositions and free noun-phrase sequences have been stated (Quirk et al 1985, Vsevolodova et al 2013), for example, are po milosti + Ngen, po mere + Ngen, po vole + Ngen, po sluchayu + Ngen prepositions or not yet? Therefore, it’s efficient to consider all units functioning as prepositions as spread in a functional-grammatical field of prepositions with iconic prepositions (po, v, na, mimo etc) in the center and marginal slightly grammaticalized constructions (po probleme + Ngen, po soobrazheniyam + Ngen, po kontrastu s + Ninstr). Nevertheless, the problem of objective description of the field is still relevant. The research basic idea is that frequency of prepositional units could be regarded as an effective instrument of prepositional field depiction. On the first step we collected prepositional units out of all sources available (dictionaries and published studies of Russian prepositions ). Every preposition of the obtained list (about 1.500 units) was scrutinized on its frequency in Russian National Corpora and then the list was ranked by the decrease of frequency, for example, po povodu + Ngen has frequency 68.13 ipm, po mere + Ngen – 32.03 ipm, po voprosam + Ngen – 8.43 ipm, po adresu + Ngen – 4.98 ipm, po obraztsu + Ngen – 1.83 ipm. Such ranging reflects the field structure from core to periphery prepositions. On the second step we aimed to reveal potential prepositions not mentioned in the available sources. In order to surmise them we analyzed lists of frequent collocations build on the most productive prepositional models in Russian: v + Nacc, v + Nprep, na + Nacc, na + Nprep, po + Ndat, po + Nprep and some others in Russian National Corpus. Thus around 150 preposition candidates were singled out, for instance, po slovam + Ngen (ipm 38.17), po dannym + Ngen (ipm 11.24), po delu + Ngen (ipm 5.77). Finally, both lists were combined and ranked by frequency decrease to show the continuum from the central to marginal prepositions. The approach selected enables the objective description of Russian prepositional field.