SubCom: A parallel corpus of English-Arabic comedy shows subtitles

Abstract

Translating humor is challenging as it is often tied to niche linguistic, interactional, and psychological contexts. This paper presents a new English-Arabic parallel corpus of stand-up comedy shows subtitles as a pedagogic tool for translating authentic examples of humor. Furthermore, the researchers experiment with the corpus by designing customized translator training tasks for checking the accuracy of translating verbal irony in a humor translation workshop. The tasks are designed based on bottom-up and top-down scaffolding of learning objectives based on a revised model of Bloom's taxonomy (2001). Moreover, the researchers propose Mateo's model of translating irony (1995) as a method of translating verbal irony in the context of stand-up comedy shows. Finally, the researchers recommend using parallel corpora as a pedagogic tool in humor translation training workshops and classrooms to bring the elements of trainees' creativity to overcome controversial translation issues.