Print Page   |   Contact Us   |   Sign In   |   Join or Create a Guest Account
Raters’ sensitivity to visual cues during communication breakdowns
Share |


Just as various linguistic and nonlinguistic factors (intonation, stress, eye-gaze) help L2 speakers notice interactional feedback (Philp, 1998; McDonough et al., 2016), they may also facilitate recognition of communication breakdowns. Previous L1 studies have shown that eye-gaze, for example, is associated with requests for repair and listener responses (Bavelas et al., 2002). To explore the role of visual cues in L2 communication breakdowns, we examined whether raters are sensitive to nonverbal information (e.g., eyebrow lift, head tilt) provided by interlocutors in episodes of nonunderstanding, defined as instances when listeners realize that they cannot make sense of an interlocutor’s utterance (Bremer, 1996).
Using data from a larger study (McDonough et al., in review), conversations between a bilingual French-English research assistant (RA) and English L2 university students (N = 21) were audio-recorded while both interlocutors’ eye-gaze was tracked using the faceLab 5.0 system. Transcripts were examined for episodes of nonunderstanding (RA requested clarification) and comparable episodes of understanding (RA asked follow-up questions). One understanding and one nonunderstanding episode with comparable question lengths (within 100 milliseconds) were selected from each participant (42 episodes).

The episodes were rated by bilingual undergraduate university students (N = 45) under three conditions: audio of the participant’s initial utterance, visual image of the RA’s face when asking the clarification or follow-up question, and the combination of audio and video. When rating audio and audio/video episodes, raters transcribed the utterance (intelligibility) and rated ease of understanding (comprehensibility) using 1,000-point scales. When evaluating the RA’s image, they rated the extent of the RA’s comprehension using 1,000-point scales and completed a verbal report protocol. Raters’ eye-gaze while rating the RA’s images was also tracked. Preliminary results indicate that raters are sensitive to visual cues when viewing instances of nonunderstanding. Methodological issues for studies of eye-tracking during face-to-face interaction are discussed.


This study examined whether raters are sensitive to the nonverbal information provided by interlocutors during communication breakdowns. Both understanding and non-understanding episodes were rated using visual and audio stimuli, and raters’ eye gaze to the interlocutor’s face was tracked. Methodological issues for eye-tracking involving face-to-face interaction are discussed.

Membership Software Powered by YourMembership  ::  Legal