Regularities in Russian Orthography: Modeling the Placement of Stresses

In modern psycholinguistics a significant number of researches is devoted to the study of speech perception and speech production, to visual word recognition, to the processes of reading and writing a text, a word or a sentence. The way how people read written texts and especially how they assign stress is of great interest in contemporary psycholinguistic field. Russian is a language with an exceptional variety of accent options. Its stress can fall on any syllable of a word and it can also be mobile depending on the word form, So, it is badly predicted. Also, Despite the large number of Russian speakers and learners, little effort has been undertaken to extend cognitive theories of reading acquisition focused on the English orthography to Russian. Such findings may be used during teaching children to read, for simplifying the algorithm of speech recognition programs, during conducting neurolinguistics experiments, during learning Russian as a foreign language and so on. So, the main goal of present research is to identify which non-lexical cues influence stress assignment in Russian pseudowords corpus using a mathematical regression model. It was pointed out that stress in any language can be assigned using non-lexical sources and rules, such as word length, syllable construction, etc. Besides, that explains the optional influence of grammatical category on the process of stress assignment. This is one of the reasons why we decided to conduct an experiment on the corpus of non-words of the Russian language, because it is known that when a person is trying to recognize a word, a lot of similar parts of the word come to his mind which subsequently form a final word. However, when he is trying to recognize a non-word, he does not find any resemblance, but he still can have associations with words-neighbors.