English participial conditional construction “if untreated”: Intelligent vs anthropogenic translation

The aim of the research is to develop a methodology for post-editing the translation of the English participial conditional clause, aimed at improving the quality of translation and eliminating errors characteristic of intelligent translation systems. To achieve this goal, a comparative analysis was made of anthropogenic translation performed by professional translators and translation performed by DeepL, Yandex Translate, and the ChatGPT-4о neural network. The scientific novelty of the research is as follows: 1) for the first time, the patterns of translation of the construction under consideration are revealed, consisting in the transformation of the English participle into a personal form of the Russian verb, a verbal noun, or a nominal group; 2) for the first time, typical errors of intelligent translation are systematized and described, mainly related to inaccuracies in the culturally significant adaptation of the text, violation of the Russian language usage, and interpretation of the semantics and syntax of the participial conditional construction. As a result of the research, practical recommendations are proposed aimed at improving the quality of intelligent translation: eliminating typical errors, restoring logical connections in the sentence, and observing the usage of the Russian language.