Live and Let Interpret: Real-Time Subtitling and the Physiology of Interpreter Empowerment

This study presents a timely and empirically grounded response to the epistemic challenge of augmenting human cognition in the era of AI-integrated interpreter training. While biocognitive evidence has shown that Automatic Speech Recognition (ASR) subtitles can enhance interpreting performance and mitigate mental load (Li & Chmiel, 2024), there is as yet no study that evaluates these effects using physiological data—real-time, embodied indicators of cognitive effort. This project advances a new methodology that brings together biometric measurement, behavioral analysis, and qualitative feedback to examine the impact of ASR subtitling as a form of Human-Centered, Augmented Machine Translation (HCAMT). It directly answers the call for new methodologies to evaluate HCAMT experience and to develop empowering tools and workflows in interpreting.