Automatic live captioning in simultaneous interpreting: A comparative analysis of quality

The increasing availability of automatic live captioning (ALC) technology offers new possibilities for its application in simultaneous interpreting (SI). However, its effectiveness in SI contexts requires further investigation, particularly compared to conventional SI modes. This study addresses this gap by conducting a comparative analysis of four interpreting modes: SI without text, SI with text, sight translation and SI with ALC. We evaluated the performance of 27 interpreter trainees from three perspectives: human raters’ assessment of interpreting quality, interpreters’ self-perceived performance and local interpreting quality in relation to specific problem triggers and automatic speech recognition (ASR) errors. Results indicate that interpreters in SI with ALC perform at an intermediate quality level, surpassing SI without text, but not matching the quality achieved in SI with text or sight translation. Our analysis of problem triggers and ASR errors reveals a nuanced relationship between technological assistance and interpreter performance, highlighting both the benefits and challenges of SI with ALC. This research contributes to the growing body of knowledge on SI with ALC by examining the specific role and impact of ALC in SI, offering insights for computer-assisted interpreting development and suggesting potential applications in interpreter training programs.