LANGUAGE LEARNING & TECHNOLOGY
A refereed journal for L2 researchers and educators interested in the role of technology in advancing language learning and teaching.

Adapting to imperfect technology: A time-course eye-tracking study of students’ cognitive engagement with ASR captions in simultaneous interpreting

Nov. 14, 2026, 1:02 p.m.
Nov. 17, 2026, 12:18 a.m.
Nov. 17, 2026, 12:18 a.m.
open.access
[['https://scholarspace.manoa.hawaii.edu/bitstreams/5332eeab-e942-4eac-91b1-fc33e8e2c8fd/download', '30_01_10125-73718.pdf']]
[['https://scholarspace.manoa.hawaii.edu/bitstreams/cb411c7b-df1f-43bf-9431-889fbf46d47f/download', 'full_text']]
Volume 30 Number 1, 2026
Ma, Xingcheng Zhong, Linping
2026-11-14T02:26:05Z
2026
2026-11-16
Automatic speech recognition (ASR) is increasingly used in interpreter education, yet how student interpreters cognitively engage with imperfect ASR captions remains underexplored. Drawing on cognitive engagement and Cognitive Load Theory, this study examined how task-intrinsic load, operationalised as syntactic complexity, and technology-induced extraneous load, operationalised as ASR caption errors, jointly shaped students’ real-time engagement during simultaneous interpreting. Twenty-five postgraduate interpreting students completed ASR-supported interpreting tasks while their eye movements and item-level interpreting accuracy were recorded. A time-course eye-tracking approach was used to capture moment-by-moment responses around manipulated caption events. Results showed that engagement with ASR captions was selective and multidimensional. Students increased visual attention to captions under high syntactic complexity only when captions were correct, whereas this complexity-driven reliance disappeared when captions contained errors. Blink suppression was mainly associated with high syntactic complexity, suggesting sustained attentional engagement. Pupil-size patterns suggested fluctuating effortful processing, with minor errors eliciting greater temporal variation than severe errors under high complexity. Interpreting accuracy declined when captions contained errors, especially severe errors. These findings show that ASR-supported interpreting involves selective attention allocation and effort regulation under competing intrinsic and extraneous demands, highlighting the need for training in strategic caption use and cognitive flexibility.
20
20
Ma, X., & Zhong, L. (2026). Adapting to imperfect technology: A time-course eye-tracking study of students’ cognitive engagement with ASR captions in simultaneous interpreting. Language Learning & Technology, 30(1), 1–20. https://doi.org/10.64152/10125/73718
https://doi.org/10.64152/10125/73718
1094-3501
https://hdl.handle.net/10125/73718
en
1
Language Learning & Technology
University of Hawaii National Foreign Language Resource Center Center for Language & Technology
http://rightsstatements.org/vocab/InC/1.0/
Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License
https://creativecommons.org/licenses/by-nc-nd/4.0/
/item/10125-73718/
1
automatic speech recognition (ASR), simultaneous interpreting, engagement, eye-tracking, time-course analysis
Adapting to imperfect technology: A time-course eye-tracking study of students’ cognitive engagement with ASR captions in simultaneous interpreting
Article Text
30