From L2 Text-to-L1 Voice: AI Audio for Reading Equity and Inclusion among Multilingual Learners

• In Event: AERA Roundtable Session Thursday 2:15 pm - Gold 2
  In Roundtable Session: Access, Identity, and Equity in Technology-Mediated Language Learning

Abstract

This case study examines how AI-generated L2 text-to-L1 audio influences multilingual learners’ identity affirmation and reading engagement in a Canadian after-school reading program. Drawing on translanguaging and expansive literacy frameworks, the study focuses on two purposefully selected students, Anaya (Hindi-background) and Haile (Tigrinya-background), who differed in their L1 and L2 proficiency and immigration backgrounds. Over eight weeks, data were collected through observations, video recordings, volunteer reflection sheets, interviews, and reading comprehension checks. Findings suggest that AI-supported L1 audio affirmed students’ linguistic identities and fostered deeper engagement with L2 reading. By illustrating how emerging AI tools can support literacy development through learners’ full linguistic repertoires, the study offers insight into more inclusive and responsive AI-assisted language learning practices for multilingual students.

Authors

• Guofang Li, University of British Columbia

• Ziwen Mei, University of British Columbia

• Chinazam Faith Okeke, University of British Columbia

• Henny Yeung, Simon Fraser University