A CORPUS-BASED APPROACH TO LEGAL TRANSLATION QUALITY ASSESSMENT (GERMAN-UKRAINIAN LANGUAGE PAIR)

Abstract

Background. Legal translation is one of the most sought-after areas of specialised translation in today's globalised world. The demand for high-quality translation of legal texts is constantly growing. Despite this, German-Ukrainian legal terminology and translation studies are still characterised by limited resources and a lack of systematisation of existing works. Translation in the field of constitutional law has so far been overlooked by researchers. The research material is the text of the Basic Law of the Federal Republic of Germany and its first official translation into Ukrainian (2024).

Methods. The research is based on a systemic-prescriptive paradigm and combines quantitative (corpus) and qualitative approaches. A translation corpus of the Grundgesetz (1286 segments) was created in memoQ for the study. The data collection method is parameterised corpus search. The methods of analysis are systematisation (identification of translation decisions relevant to the research, organisation of data into logical structures), analytics (comparative analysis of recurrent patterns of variability in translation decisions), statistics (descriptive statistical analysis of the object) and reflection (critical analysis of research assumptions and methodological limitations).

Results. The research confirmed two heuristic hypotheses put forward by the authors. In particular, it was proven that, firstly, Ukrainian translators demonstrate an unsystematic approach to preserving the consistency of the original text, and secondly, that commonly used vocabulary demonstrates a higher level of variability in reproduction compared to legal vocabulary itself, which can be explained by the corresponding focus of the translator's attention and neglect of conventional background vocabulary.

Conclusions. The cases of terminological and syntactic inconsistency identified by means of parameterised corpus search demonstrate both the heuristic potential of the chosen methodology and its certain limitations. The advantages of the corpus approach include the possibility of quantitative verification of hypotheses, systematic identification of recurring patterns, and objectification of the assessment of the translation decision's consistency. At the same time, critical limitations have been identified: the inability to automatically distinguish between functionally motivated and unmotivated inconsistencies without additional qualitative analysis, the indirect nature of the correlation between quantitative indicators of consistency and the overall quality of the translation, as well as the technical impossibility of automated detection of patterns of different cases. Therefore, to assess the quality of specialised translation, it is best to combine quantitative and qualitative methods, which complement each other's limitations. The conclusions of the research highlight the need to institutionalise quality control of legal translations.