Register as a predictor of the complement clause type in Russian: a corpus study

2019. № 1 (37), 147-183

MIKHAIL YU. KNYAZEV, Institute for Linguistic Studies of the Russian Academy of Sciences / National Research University Higher School of Economics, St. Petersburg branch / St. Petersburg State University (Russia, St. Petersburg),
EKATERINA A. RUDALEVA, National Research University Higher School of Economics, St. Petersburg branch (Russia, St. Petersburg, Moscow),


Register is a factor in linguistic variation and may affect the choice between competing constructions, along with grammatical, semantic, pragmatic and other factors. The paper investigates the effect of register on the realization of indicative complement clauses in Russian (as a čto-clause or as a special construction with the correlative to) with verbs selecting prepositional/oblique complements, cf. Ona nastaivaet (na tom), čto… ‘She insists (on the fact) that…’. The study is based on the data from the Russian National Corpus. The main hypothesis is that more formal registers tend to be associated with higher preference for the construction with the correlative. Five different registers are compared (academic, news, fiction, spontaneous and non-spontaneous speech). The influence of register was investigated in three ways: on the basis of the verb sample as a whole (26 verbs); individually for separate verbs; for attraction classes of verbs according to their choice of a certain strategy. The so-called “collostructional analysis” (for distinctive collexemes), developed by Stefan Th. Gries and Anatol Stefanowitsch, as well as the analysis of dispersion measured by the Deviation of Proportions (developed by Stefan Th. Gries) were used. The results showed a small increase in the proportion of the correlative construction in the academic and non-fiction registers (as opposed to news and fiction registers) both for the sample as a whole and for the two out of three attraction classes, which on the whole supports the main hypothesis. However, the effect of register on individual verbs (operationalized as change of attraction class accompanying change of register) proved negligible. In addition, the class of verbs neutral with respect to the attracted construction showed a profile different from the two other classes, suggesting that some property of register other than formality might influence their behavior.