Орфографические соседи в русском языке: база данных и эксперимент, направленный на изучение морфологической декомпозиции

Abstract

The role of orthographic neighbors (e.g. trail – trial) in word processing has been discussed in many experimental studies. these studies address the problem of visual word recognition, lexical access in the mental lexicon and a number of other topics that are currently under debate in psycholinguistics. However, there are virtually no studies on the material of the Russian language – potentially, due to the fact that there is no resource that would allow searching for such words balanced according to different parameters. In the present paper, we present StimulStat, a database that solves this problem, and an experimental study based on it. a number of experiments studying transposition neighbors asked the following question: does it play a role whether the transposition takes place inside one morpheme or across the morphological boundary? this lets us understand whether multimorphemic words are decomposed in processing and, if yes, at what stage. unlike the majority of earlier studies, our experiment (using lexical decision task with priming) studies in ected forms rather than derived words. the results let us conclude that noun forms in oblique cases are not stored as a whole and are decomposed (at least, when they are not highly frequent), while the initial (nominative singular) form has a special status.