Annotated Corpus Dataset of Clausal Complementation with Serbian Modal Verbs

This dataset contains annotations of 2,122 instances of clausal complementation with the Serbian modal verbs moći (‘can’), morati (‘must’), and trebati (‘should’), extracted from the CLASSLA-web.rs corpus. Each instance is annotated for complement type, modal flavor, voice, subject properties, and object properties, yielding eleven annotation categories (twelve for trebati). The dataset was created to facilitate research on the alternation between infinitival and da-present complements in Serbian and related BCMS varieties. A detailed description of the annotation scheme, category definitions, examples, and references is provided in the accompanying documentation file (Dataset description.pdf).