New Database Maps the Emotional Pull of 6,708 Russian Words

A new resource for studying how emotion shapes language has been unveiled: ENRuN-2, the first large database of emotional ratings for Russian words. The database, described in Behavior Research Methods, contains 6,708 nouns evaluated by thousands of native speakers on both continuous emotional dimensions and discrete emotion categories, a combination its developers note is rare among word-rating resources in any language. Researchers have long described the emotional content of words in two complementary ways.

The categorical approach instead asks which discrete emotions a word evokes, such as happiness, sadness, fear, anger, or disgust. ENRuN-2 closes that gap for Russian. Each of the 6,708 nouns was rated for valence and arousal as well as for its relatedness to five emotion categories: happiness, sadness, fear, anger, and disgust.

The ratings came from 3,902 native speakers, and data quality procedures refined this into a final analytic sample of 2,681 participants, a substantial pool by the standards of affective norming projects. The authors report strong reliability. Inter-rater consistency across the emotional parameters was high, with mean intraclass correlations ranging from .84 to .96, and individual raters showed strong stability of their ratings over time.

These figures suggest that the emotional meanings of the words were perceived consistently across a large group of speakers. Convergent validity also held up. Correlations between the new ratings and existing Russian affective norms ranged from rs = .50 to .98, indicating that ENRuN-2 aligns well with earlier datasets while extending their coverage substantially.

Such agreement matters because it shows the new norms are measuring the same underlying emotional properties rather than idiosyncratic features of the collection process. The structure of the ratings themselves offered a further check. Analysis of how the emotional parameters relate to one another revealed patterns that are mostly nonlinear, and these relationships were consistent with findings from similar databases in other languages, which the authors take as further support for the validity of the ratings.

Gender analyses produced a nuanced picture. Overall, differences between male and female raters were negligible, and their ratings correlated strongly with one another. Yet for some individual words the ratings of men and women diverged substantially, suggesting that aggregate agreement can conceal meaningful item-level variation that researchers may want to examine when selecting stimuli.

The practical reach of the database is broad. Databases with emotional ratings of words are essential for studying emotional information processing in psychology, psycholinguistics, neuroscience, and related disciplines. By offering both dimensional and categorical descriptors for thousands of Russian nouns, ENRuN-2 gives Russian-language researchers, and those conducting cross-linguistic comparisons, a shared instrument comparable to norms available for other languages.

The database is freely available on the Open Science Framework, lowering the barrier for laboratories worldwide to adopt Russian emotional stimuli. Still, with its scale, dual-framework design, and documented reliability and validity, ENRuN-2 stands as a notable addition to the international family of affective word databases.

Subject of Research: Dimensional and categorical emotional ratings of Russian nouns

Article Title: Dimensional and categorical emotional ratings of Russian nouns: The ENRuN-2 database

Article References: Lyusin, D., & Sysoeva, T. (2026). Dimensional and categorical emotional ratings of Russian nouns: The ENRuN-2 database. Behavior Research Methods, 58(12), Article 335. https://doi.org/10.3758/s13428-026-03218-3

Image Credits: AI Generated

DOI: 10.3758/s13428-026-03218-3

Keywords: ENRuN-2, Russian language, emotional ratings, valence, arousal, discrete emotions, affective norms, psycholinguistics, word database, reliability, open science, Behavior Research Methods