Software Tracks How the Russian Language Is Changing
Researchers at Murmansk Arctic University have built a program that analyzes hundreds of millions of words to measure how Russian evolves over time.
In Russia, a team at Murmansk Arctic University has introduced a computer program for diachronic lexical analysis. The system relies on automated lemmatization and quantitative text analysis. Its algorithm reduces words to their base form and calculates how their frequency shifts across historical periods.
The platform processes large text corpora and maps changes in word usage without manual annotation. The advance expands the scale of linguistic research, enabling scholars to work with lexical data at industrial volumes. As a result, researchers can track how the language changes over time.
The team tested the program using materials from the National Corpus of the Russian Language. The dataset included texts from the pre-revolutionary, Soviet, and post-Soviet periods. In total, the software analyzed more than 250 million words.
A Surprising Pattern
The findings point to a nuanced shift. Russian does not show an explosive surge in entirely new words. Instead, the total number of words actively used in everyday language continues to grow.
Numbers appear far more frequently in contemporary texts. Researchers attribute the shift to the influence of the digital environment and to the growing role of statistics, technology, and quantitative metrics in public life.
At the same time, some words gradually fall out of favor and appear less often in modern writing. The program detects these trends automatically.
The developers say the accumulated data could support more accurate forecasts of changes in both conversational and formal language. The large-scale datasets may also help train artificial intelligence systems to better understand lexical evolution and adapt to shifting usage patterns.
Because neural networks “communicate” with people through written language, the researchers argue, this kind of analysis is invaluable. A deeper understanding of how vocabulary evolves could ultimately improve how humans and machines understand one another.
