Orthographic Processing of Russian Case Forms in Sentential Context

The aim of the study is to compare the processing of Russian case forms in sentential contexts. In most contexts, only one form is correct, so the question is which case errors are more costly and why. Russian has a rich inflectional morphology: Russian nouns are inflected for six cases and two numbers and belong to several inflectional classes with diverse patterns of syncretism. Many studies of case processing in morphologically rich languages investigated differences between noun forms presented in isolation, case hierarchies were revealed according to the reaction times in a lexical decision task. Frequency, syncretism and other factors were found to affect reaction times. However, morphological processing in isolation and in context may be influenced by different factors, and it is not clear which factors are relevant in sentential context and which are not. In the authors' self-paced reading study (96 participants, Latin square design), Russian sentences with prepositions requiring different cases (Genitive, Dative, Accusative, Instrumental) were used in four conditions, with correct and incorrect case forms: e.g., Desert iz maliny-GEN/ *maline-DAT/ *malinu-ACC/ *malinoy-INS soderzhit mnogo vitaminov [A Raspberry dessert contains many vitamins]. All target nouns were feminine singular of the same inflectional class, this allowed avoiding undesired syncretism patterns. The authors used linear mixed-effects models for statistical analysis. All errors were read longer than correct case forms. Among incorrect forms, Genitive forms after prepositions requiring Accusative and Accusative forms after prepositions requiring Dative were processed significantly faster than other errors (p<0.01). This can be explained by the syncretism of case affixes across different inflection types in Russian: the Gen.Sg affix in the first declension coincides with the Acc.Pl affix in many inanimate nouns; the Acc.Sg affix in the first declension coincides with the Dat.Sg one in the second declension. The data show that after a preposition the reader not only predicts a particular form (hence correct forms are always processed faster independently from their relative frequency and place in the case hierarchy, which are usually called upon to explain RTs to isolated forms), but also expects a particular set of inflections, which are associated with the relevant case in one or another inflectional class. The second finding is novel and provides evidence for the morphological decomposition and independent representation of inflectional affixes in the mental lexicon.

Keywords

Authors

References