Constructing a corpus of translated films: a corpus view of dubbing

Corpus linguistic research has in recent years entered the realm of Translation Studies by providing translation scholars and professionals alike with new insights into the processes and products of translation. Audiovisual translation is no exception, with corpus research being done on the linguistic and translational features of dubbing, as well as other modes of AVT, for example subtitling and audio description. The present paper explains the principles informing the design of the latest release of the PaviaCorpus of Film Dialogue, a parallel corpus of American and British film dialogues and their dubbed versions into Italian to which a comparable component of original Italian films has been added, with a view to discussing the main issues concerning the use ofcorpora in AVT. Bringing together some of the tenets of corpus linguistics, i.e. language patterning, routine and creativity, and AVT perspectives, insights gained from the Pavia Corpus are presented as evidence for a combined approach.