EXPERIMENTAL STUDY OF THE INFLUENCE OF REGISTER FORMALITY ON THE PREFERRED REALIZATION OF THE COMPLEMENT CLAUSE IN RUSSIAN

Abstract

In Russian, sentential complements (that-clauses) are optionally preceded by the demonstrative to ‘that’ (sometimes called ‘correlative’), with the required case marking or preposition selected by the verb. Several factors have been suggested in the literature as influencing the choice of the construction with to, including the matrix predicate as well as information structure. Th is work investigates the influence of register formality on the preferred realization of the complement. It is hypothesized that the construction with to, being the more explicit variant, is associated with more formal registers. Th is hypothesis has been previously tested on the basis of corpus data and received limited support. Here it is tested in a (two-alternative) forced-choice experiment where the register (news vs. spoken) is manipulated within items and subjects. 18 alternating predicates were tested, with two matched pairs of sentences with the contrasted registers constructed for each predicate. In addition, the experiment tested the influence of lexical biases, as measured by the ‘collostructional analysis’ or by relative frequency of the alternating constructions in the corpus. Th e results showed a reliable effect of lexical biases, as well as a much weaker but significant effect of formality suggesting that in the news register the construction with to is preferred more oft en compared to the spoken register, in line with the experimental hypothesis, whereas register did not interact with the lexical bias. Interestingly, a substantial amount of variation in the results was due to lexical biases as well as individual speaker preferences for one or the other construction.