Pojasnjevalni izrazi v zakonodajnih dokumentih Evropske unije

This study examines code glosses, i.e. a category of metadiscourse that clarifies or exemplifies content, in the EU’s legislative corpus, using the English and Croatian DGT Translation Memory parallel corpus (Steinberger et al., 2012). Code glosses enhance coherence and reader comprehension by reformulating or elaborating arguments, which is especially useful in legal texts, yet their use in such texts remains understudied compared to the academic genre. Employing Sketch Engine for automated extraction and manual analysis, the research reveals nuanced differences in the frequency and form of code glosses between English and Croatian legislative texts (in Croatian code glosses are used slightly less, but certain code glosses, such as colons, are used more frequently), reflecting the tendency to keep the structure of the original text in translations. These findings contribute to a better understanding of cross-linguistic textual adaptation in legislative contexts. Future work will explore other metadiscourse categories and their impact on reader comprehension in legal contexts.