Linguists Presented Research Findings on Disinformation Texts at CILC 2026 in Madrid

From 10 to 12 June 2026, Xinying Chen and Miroslav Kubát presented findings of the linguistic research at the 17^th International Conference on Corpus Linguistics, CILC 2026. The conference was held at the Autonomous University of Madrid and focused on current approaches, methods, and future perspectives in corpus linguistics.

The project team members delivered two presentations examining the linguistic characteristics of Czech online news. The first paper, Beyond Topics: Morphological Profiling of Credible and Manipulative Czech News, focused on differences in the morphological structure of credible and manipulative texts. The analysis showed that the individual news categories differ, for example, in the distribution of parts of speech, grammatical case forms, pronouns, and other grammatical features. The second paper, Corpus-Based Lexical Complexity Indexes for Czech: Design and Application, introduced a new set of measures for assessing the lexical complexity of Czech texts. These measures were subsequently applied to compare credible and manipulative news articles.

The findings of both studies confirm that manipulative texts differ systematically from credible reporting in both their grammatical and lexical structure. They often contain more complex linguistic structures and may place greater cognitive demands on readers when processing information. The research thus contributes to a more precise description of the linguistic mechanisms employed in disinformation-oriented media content.

Updated: 24. 06. 2026