Proactivity in Scaffolding Comprehension and Production in Real-Time Turn-Taking: A Case Study of Bridging Communication Gaps for Non-Native Speakers

Status: Published as a short Work-in-Progress at CSCW Companion 2025 (Bergen, Norway). Full manuscript in preparation.

This short paper introduces the proactive-scaffolding design space for AI tools that support non-native (L2) speakers during real-time multilingual conversation. Whereas prior CMC tools tend to react after a breakdown (e.g., translation, repair), this work argues for proactive scaffolds that anticipate cognitive trouble at specific stages of speech processing—before the speaker stalls or the listener disengages.

Using the XPLAIN system as a case study, the paper introduces three scaffold types—lexical clarifications, idea/content suggestions, and topic summaries—and reports preliminary findings from a Wizard-of-Oz study (N = 38) showing 30–80% perceived gains in communicative efficiency. The full manuscript expands these results with stage-aligned timing analyses and individual-difference moderators.

Recommended citation: He, W. P. & Fussell, S. R. (2025). "Proactivity in Scaffolding Comprehension and Production in Real-Time Turn-Taking: A Case Study of Bridging Communication Gaps for Non-Native Speakers." Companion Publication of the 28th ACM Conference on Computer-Supported Cooperative Work and Social Computing (CSCW Companion '25). Bergen, Norway. ACM.
Download Paper