Echoes of Accountability: Computational Analysis of Post-Junta Parliamentary Questions in Greece (1974-1977)

This paper examines the parliamentary discourse of the Hellenic Parliament during the first post-junta legislative period (1974-1977) by analyzing a digitized corpus of 1,338 written parliamentary questions. These questions, written in the formal Katharevousa Greek language, represent a significant moment in Greece's parliamentary history and linguistic evolution. Digitized using machine-learning-enhanced Optical Character Recognition (OCR) methods, this corpus offers an opportunity to explore the socio-political concerns of the immediate post-dictatorship era while showcasing the challenges of processing historical documents (Fitsilis et al., 2024). Our methodology combines advanced computational stylistics and natural language processing (NLP) techniques to examine the corpus. Tools such as topic modeling, semantic similarity analysis and sentiment analysis are applied to uncover thematic trends, political priorities and the discourse devices employed by legislators (Nay, 2021). These methods allow for an in-depth understanding of parliamentary oversight and representation during a period of political reconstruction. Key findings are anticipated to reveal how written questions served as instruments of accountability, signaling legislative priorities and engaging with public concerns during a transformative era in Greece. The study also investigates linguistic characteristics unique to Katharevousa, contrasting them with modern Greek to evaluate changes in parliamentary discourse over time. By employing comparative stylometric analysis, including measures of vocabulary diversity, sentence complexity and thematic distribution, the research identifies continuities and divergences between past and present parliamentary practices. Beyond its historical significance, the findings contribute to broader discussions on the role of language in political research (Dunmire, 2012). Moreover, by utilizing state-of-the-art computational methods, this research establishes a foundation for integrating AI-driven tools into parliamentary studies, enhancing the analysis of legislative behavior and linguistic evolution. The study concludes by offering recommendations for applying these methods in future research, with implications for preserving and analyzing historical archives in similar parliamentary contexts worldwide. References Dunmire, P. L. (2012). Political discourse analysis: Exploring the language of politics and the politics of language. Language and Linguistics Compass, 6(11), 735-751. https://doi.org/10.1002/lnc3.365 Fitsilis, F. et al. (2024). Digitization of Written Parliamentary Questions from the Historical Archive (1974–1977) of the Hellenic Parliament. In: Mouchère, H., Zhu, A. (eds) Document Analysis and Recognition – ICDAR 2024 Workshops. ICDAR 2024. Lecture Notes in Computer Science, vol 14935. Springer, Cham. https://doi.org/10.1007/978-3-031-70645-5_8 Nay, J. (2021). Natural language processing for legal texts. In D. M. Katz, R. Dolin, & M. Bommarito (Eds.), Legal informatics. Cambridge University Press. Available at http://dx.doi.org/10.2139/ssrn.3438276