Proceedings of the Annual Meeting of the Cognitive Science Society

This study examines linguistic differences between truthful and false written narratives in Hindi using psycholinguistically motivated measures of language processing complexity. Adopting a corpus-based, processing-oriented approach, we investigate how the cognitive demands of producing falsehoods shape lexical and syntactic choices beyond surface-level cues. Participants produced written opinions that were either truthful or intentionally false, which were analyzed using metrics including surprisal, dependency distance, part-of-speech distributions, and type–token ratio. Our results show that lower mean surprisal, along with higher verb counts and reduced use of nouns and adverbs, significantly predicts false narratives. These patterns suggest that falsehood production in Hindi relies on linguistically efficient strategies characterized by predictable constructions and reduced informational specificity to ease planning and monitoring under cognitive load. By presenting evidence from an underrepresented South Asian language, this study advances cross-linguistic research on deception and highlights the value of processing-based metrics for understanding deceptive language production.