LIMBAJUL CREDINȚEI ÎN LUMINA METODELOR COMPUTAȚIONALE

Abstract

The study proposes a computational reading of Orthodox religious discourse during
the communist regime, analysing how the language of faith was adapted, encoded and
maintained in a restrictive ideological context. Based on a corpus of pastoral texts published
in the magazine Mitropolia Moldovei și Sucevei during 1948-1989, our research combines
traditional linguistic analysis with automatic language processing tools, by means of the
Orange Data Mining platform. Through techniques such as frequency analysis, lexical
viewing and topic modelling, prevailing topics are identified, such as divinity, resurrection,
peace, light, unity, which reveal both teological continuity and subtle forms of discourse
adaptation to the ideological vocabulary of the regime.