PrevHist: A Dataset of Old and Middle Hungarian Preverb Constructions

This data paper presents a dataset of Old and Middle Hungarian preverb-verb constructions, covering a time span from the late twelfth century to 1772. The resource consists of 68,458 records. It was extracted automatically from two corpora: the Old Hungarian Corpus and the Old and Middle Hungarian corpus of informal language. Each record is annotated for a wide range of morphosyntactic features and metadata, notably the year of writing and the register. The dataset can be used to explore the diachronic trajectory of preverb-verb constructions and, eventually, to gain a better understanding of preverbs’ grammaticalization process in Hungarian.