The Language of Science Journalism in the Digital Media Environment: According to Georgian Multimedia Publications

Abstract

This study provides a corpus-based linguistic analysis of science journalism within the Georgian digital media environment, focusing on publications from Marketer.ge and 1tv.ge – Science. Analyzing a corpus of 70 articles containing 27,046 tokens and 9,299 unique types, the research evaluates thematic distribution, multimedia integration, hypertexuality, lexical diversity (TTR), lexical density, and readability. Thematic analysis reveals a dominant focus on natural sciences (51%) and medicine (30%), driven by societal demand. The structural analysis shows high hypertextual integration, with 87% of articles utilizing hyperlinks. While the overall corpus TTR is 34% due to functional word frequency, individual articles exhibit exceptionally high lexical diversity (60%–87%) and stable lexical density (0.65–0.85), proving high informational capacity. A key finding highlights a distinct disciplinary divergence: humanities-based media texts display lower readability indexes, longer sentence structures (averaging 24.5 words), and higher lexical density compared to hard sciences. This indicates that tech-journalists adapt technical terms more thoroughly, whereas humanities discourse remains syntactically complex and heavily reliant on unadapted abstract concepts. Ultimately, digital science journalism effectively transforms dense academic data into engaging, semi-popularized narratives.