Traduzione automatica e testualità: analisi della traduzione di un testo letterario dal russo all'italiano

Abstract

The aim of this paper is to compare, in line with the theoretical framework developed by the School of Basel, Human Translation (HT) and Neural Machine Translation (NMT) on the basis of a Russian literary text that was translated into Italian. Data show how NMT presents fewer interventions at the level of text structure than HT, conforming more to the textual characteristics of the source language. HT, in comparison to NMT, shows a lower degree of segmentation, a tendency to avoid short-distance repetitions of the same lexeme, and predominantly uses subordination. These findings are in line with the existing literature on Italian-Russian contrastive text linguistics. The analysis demonstrates the effectiveness of the Basel Model in studying text structure translation for the Russian-Italian language pair. The results of this study can be applied for a more conscious use of NMT, accurate post-editing and the teaching of Russian-Italian translation.