Decoding the relationship between language, entropy, and cognitive functioning: insights from a statistical analysis of Romanian literature

Abstract

The capacity for effective communication and sharing of ideas has been instrumental in driving the human evolution. The focus of this paper is on exploring the fundamental mathematical framework that underlies natural language, and demonstrating that practical applications can readily uncover theoretical principles like conditional probabilities and informational entropy. This study utilizes a comprehensive Romanian language corpus to conduct a statistical analysis of individual and conditional word probabilities. Additionally, an experimental survey was designed to evaluate the respondents’ subconscious selection of follower words. By comparing the results of the survey with the reference results obtained from the statistical analysis of the representative corpus, significant correlations were discovered between the conditional entropy and the survey participants’ chosen options. The study considers demographic variables such as age, gender, and education level and evaluates their contribution to the results. The paper also proposes several possible directions for future research.