+48 22 5526033
Paraphrasing in respeakingcomparing linguistic competence of interpreters, translators and bilingualsRespeaking is a method of producing subtitling for live events and TV programmes.Respeakers repeat speakers' utterances so that they may be changed by speech recognition software into subtitles for the deaf and hard of hearing. Respeakers need to paraphrase the text so that it conforms with temporal and spatial constraints of subtitling. Due to the similarities between respeaking, interpreting and translation, we tested interpreters, translators and bilingual controls on a paraphrasing task to see if interpreters or translators would manifest any advantage thanks to experience.Following a respeaking training, the participants were asked to paraphrase sentences with semantic redundancies, oral discourse markers and false starts in a simultaneous and delayed condition. Contrary to our predictions, we found that experience did not modulate paraphrasing quality or speed in general, but interpreters did outperform other groups when eliminating semantic redundancies, which were also the most difficult reformulations to tackle for all participants. The data suggest that while interpreters and translators are not better predisposed to become respeakers than regular bilinguals, at least as regards the paraphrasing performance, certain aspects of the interpreting experience (the need to express meaning concisely within time constraints) may offer slight advantage in producing well-formed respoken subtitles.