Evaluating the effects of computer-assisted simultaneous interpreting: evidence from interpreting quality, cognitive load, and visual attention

This study investigates computer-assisted simultaneous interpreting (CASI), examining how automatic speech recognition (ASR) and machine translation (MT) affect interpreting quality and perceived cognitive load, and how interpreters visually attend to these supports. Twelve novice interpreters in China and seven in Australia completed L2–L1 (English –Chinese) and L1–L2 (Chinese –English) tasks of varying technicality in conventional simultaneous interpreting (SI) and CASI modes. Across both cohorts, CASI improved overall quality and source -target correspondence, particularly in more technical tasks, while effects on fluency and cognitive load were mixed. Eye-tracking in the Australian cohort showed greater attention to ASR than MT in L2–L1, but more balanced attention in L1–L2. More frequent switching between ASR and MT was associated with lower L1–L2 quality. Greater ASR attention corresponded to higher fluency in L2–L1 but lower fluency in L1–L2, with the reverse pattern for MT. Greater ASR attention was also related to lower perceived cognitive load in more technical tasks, whereas greater MT attention tended to coincide with a higher load. Overall, the effects of CASI varied with task technicality, interpreting direction, and visual attention patterns.