Tomasz Korybski and Elena Davitti (University of Surrey)
In this talk, we will focus on effective editing skills in interlingual respeaking, i.e., the abilities which allow the respeaker to correct subtitles in real time while continuing to perform this very complex and multitasking form of human-AI interaction and which, as of now, appear to be uniquely available to humans. According to Romero-Fresco and Pöchhacker (2017: 159), effective editions (EEs) in respeaking are deliberate “deviations from the source text that do not involve a loss of information or that even enhance the communicative effectiveness of the subtitles”. EEs are not factored into the NTR formula to calculate accuracy, in which predominant error types drive overall scores, but still form part of the overall assessment. However, they are captured as part of the analytical process, and can be looked at in context.
Our discussion will be grounded in empirical findings from the SMART project (Shaping Multilingual Access through Respeaking Technology, Economic and Social Research Council UK, ES/T002530/1, 2020-2023), which has explored the skillset necessary for interlingual respeaking. Through an experimental, multi-method and exploratory approach, SMART has collected an extensive database from language professionals (n=51) to refine our understanding of the process and product of respeaking, namely what human variables underly this practice, as well as what can contribute to output accuracy. In the SMART project, we applied the NTR model (Romero-Fresco and Pöchhacker 2017) as a basis for accuracy analysis, complemented with data from other (cognitive, self-reflective, survey-based) methods.
As effective editing is an underexplored area, we used our database of recordings, transcripts and NTR analyses of language professionals’ respeaking performance as a starting point to identify and categorise recurring strategic interventions that can fall under the umbrella term of effective editions. In this talk we will show examples of content processing and streamlining through live (effective) editing implemented by our participants after undertaking substantial respeaking training, with a view to expanding Romero-Fresco and Pöchhacker’s general definition of this key, yet complex, human skill.
To further highlight the importance of human agency in live captioning, we will present a comparison of fragments delivered through a fully automated live captioning workflow (with no humans involved) with the human-centric workflow investigated in the SMART project. Finally, we will discuss how our findings on ‘what the human does well’ via effective editing can contribute to addressing questions related to the added value of the human in increasingly technologized working environments, and how our data can be used to formulate recommendations to maximise the effectiveness of upskilling for language professionals.
References
Davitti, E. (2021) ‘Interlingual Respeaking: SMART opportunities at the interface of human interpreting and technology’, in ‘Negotiating a new path. Trends in Translation and Interpreting 2021, ITI (Institute of Translation and Interpreting) Trends e- book.
Davitti, E. and A. Sandrelli (2020) ‘Embracing the complexity: a pilot study on interlingual respeaking’, Journal of Audiovisual Translation 3(2): 103-139, European Association for Studies in Screen Translation.
Romero-Fresco, P., and Pöchhacker, F. (2017). Quality assessment in interlingual live subtitling: The NTR Model. Linguistica Antverpiensia,16,149–167.
Tomasz Korybski is Research Fellow at the Centre for Translation Studies, University of Surrey, Guildford, UK. Building on his 15+ years of experience as conference interpreter and interpreter trainer, he focuses on investigating the impact of new technologies (in particular speech recognition technologies) on the performance of interpreters as well as on the advent of new, hybrid modalities in cross-lingual communication afforded by machine translation, automatic speech recognition, and respeaking technologies. He is currently working on the ESRC-funded SMART project and exploring alternative, semi-automated workflows for real-time interlingual speech-to-text communication via SR in the parallel project MATRIC (Machine Translation and Respeaking in Interlingual Communication) together with other CTS members (Davitti, Orasan, Braun).
Elena Davitti is Associate Professor at the Centre for Translation Studies, University of Surrey (UK). Her research interests include hybrid modalities of spoken language transfer, methods for real-time interlingual speech-to-text and how increasing automation of these processes would modify human- led workflows. Elena is leading the ‘SMART’ project (Shaping Multilingual Access with Respeaking Technology, 2020-2023, ESRC UK, ES/T002530/1) on interlingual respeaking with an international consortium of collaborators and advisors from academia (UNINT Rome, University of Vigo, University of Roehampton, University of Vienna, University of Antwerp, Macquarie University) and from the industry (Ai-Media, SUB-TI, Sky). Elena has also published on interactional and multimodal dynamics of interpreter-mediated interaction, and she has been co-investigator on several EU-funded projects on video-mediated interpreting (AVIDICUS 3, SHIFT in Orality) and innovations in interpreter education (EVIVA, WEB-PSI). Elena has served on the boards of projects and organisations in her fields of research (e.g. ILSA Advisory Board, GALMA, IATIS).