Non-homogenizable Vocal Habitus: Podcast as Narrative Practice for Learning Subjectivity and Community Building

Abstract

Generative artificial intelligence (GAI) has transformed
the conditions of knowledge production in higher education,
intensifying concerns over homogenized learning outcomes.
This study employs a qualitative case study design drawing
on Bourdieu’s theory of practice. It analyzes 15 graduate
students’ podcast projects, supplemented by peer
questionnaire data, to examine how vocal narration reshapes
learning in the AI era. The findings reveal four
interrelated dimensions. First, podcast creation
constitutes a practical pathway for learners, allowing them
to actively engage with and navigate GAI-related
conditions. Second, through their embodied vocal habitus,
students delineate human boundaries that remain difficult
for GAI to replicate. Third, by reconnecting learning with
personal responsibility and lived experience, podcast
practices expand the conditions of possibility for learner
agency and individual empowerment. Finally, through
affective resonance within peer communities, students
realize the authentic value of learning beyond instrumental
or automated outcomes. This study argues that embodied and
situated vocal narration offers a critical means by which
learners can meaningfully respond to the world in the GAI
era.