Introduction. The article presents the results of an analytical review of scientific papers reflecting the stages of formation and development of cognitive poetics as an interdisciplinary research area in Russian linguistics. The purpose of the article is to systematize and critically analyze approaches, concepts and methodological principles of the study of concepts that constitute the cognitive basis of literary texts, authors’ idiostyles and, by extension, of poetic sociolects and literary genres.Methodology and sources. The paper reviews theoretical concepts and approaches (developed by R. Tsur, P. Stockwell, I.A. Tarasova, N.S. Bolotnova, L.O. Butakova, etc.), which served as the foundations of cognitive poetics. A comparative analysis of works focusing on the structure of literary concepts, their textual representation aspect and categorization criteria is carried out. The studies which concentrate on the application of corpus methods in order to analyze literary texts, including the works by B. Fischer-Starcke, D. Siepmann, A. Stefanowitsch, M. Mahlberg, etc., also formed the methodological basis for the analytical review.Results and discussion. As a result of the review study, the evolution of theoretical and methodological approaches to literary concepts and literary texts’ ideospheres is demonstrated. The main concepts that influenced the formation of cognitive poetics are identified; the relevant problems of cognitive studies of poetic and prose texts are analyzed. It’s shown that it is necessary to further consider the problem of identifying dominant concepts of authors’ idiostyles describing cognitive mechanisms that underlie the semantic variability of individual concepts. The importance of using corpus methodology in the study of literary discourse in its synchronic and diachronic perspective is discussed.Conclusion. It is shown that, despite the fact that the author's creative consciousness has been studied intensively and for a long time in Russian cognitive poetics, there are still problems that need further consideration. The following aspects must be analyzed: the criteria for choosing the method of modeling a literary concept, taking into consideration its type; the elaboration of uniform terminology in existing approaches; the development of the algorithms for describing the structure of literary concepts and identifying conceptual dominants of authors’ ideospheres. A conclusion was made regarding the potential of applying corpus research methods in the field of cognitive poetics. A range of the tasks related to the cognitive research of literary texts that could be solved by using statistical methods of corpus linguistics was determined.