This article is devoted to the theoretical substantiation and practical design of a bilingual Russian-Polish dictionary model of artistic substandard language based on empirical data from a parallel corpus. The study is based on Victor Pelevin’s novels Empire V, Batman Apollo, S.N.U.F.F. and their Polish translations made by Ewa Rojewska-Olejarczuk and Aleksander Janowski. The paper critically examines the limitations of traditional academic lexicography in capturing the dynamic layers of low colloquial vocabulary, jargon, and authorial neologisms that perform crucial ontological and text-forming functions in postmodern prose.
As a key methodological tool, the authors apply a corpus-oriented approach combined with the calculation
of the statistical idiolect specificity coefficient (weirdness). The comparative analysis revealed significant variability in translation strategies, as well as the inevitability of register shifts and pragmatic transformations driven by the translators’ individual styles. The main result of the study is a five-zone invariant structure of the dictionary entry (lexeme passport), including the lemma, corpus frequency index, distribution of equivalents, aligned mini-concordance of contexts, and commentary on translation. The developed resource functions as a verified “dictionary of translation suggestions”, presenting equivalence as a dynamic distributive variable. The implementation of corpus-oriented resources opens new perspectives for describing the translation process and reliably selecting functional equivalents in contemporary literature.