Answers are grounded in this paper and cite their sources.
Compares neural network translations of Newspeak with professional translations, highlighting neural network limitations.
The subject of the study is the peculiarities of the translation of artificial languages by neural networks based on the material of Newspeak, an artificial language from the novel by G. Orwell's "1984". The purpose of the work is to identify the limitations and capabilities of modern artificial intelligence models in the translation of artificial languages in comparison with professional translations using the example of Newspeak. The research material includes fragments of an appendix from the dystopian novel "Principles of Newspeak" and the main Newspeak lexemes extracted from the literary text, as well as their interpretations in translations by V. Golyshev, D. Tselovalnikova and two neural network systems — BotHub and GigaChat. The study used comparative translation analysis, semantic and structural comparison of occasionalisms, and linguistic stylistic assessment of the degree of preservation of artificiality, morphological regularity, and ideological orientation of Newspeak. Special attention is paid to the compliance of translations with pre-formulated criteria reflecting the specifics of Newspeak. It has been established that professional translators demonstrate different strategies: Golyshev is focused on artistic adaptation, whereas Tselovalnikova strives for structural accuracy. Neural network systems can reproduce individual elements of Newspeak word-making, but they tend to normalize artificial language. GigaChat has shown the most stable results, but it also requires clear restrictions and subsequent post-editing by a professional translator. The scientific novelty of the research consists in a comprehensive assessment of the possibilities of neural network translation in relation to artificial languages and in identifying factors that hinder the algorithmic interpretation of Newspeak. Thus, the use of neural networks in the translation of artificial languages is possible only as an auxiliary tool under the supervision of a professional translator.
No takes yet. Share an insight, caveat, or question.
Darya Aleksandrovna Sushkova (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: