British Journal of Sociology of Education

In this article, we share results of our feminist co/autoethnography, through which we studied our own (failed) experimentation with generative AI as a 'problematic research partner'. We orchestrated and participated in a series of feminist co-generative dialogues, spanning two phases, centred on the tensions and possibilities of AI in education (AIEd). We view our (failed) experiment as productive, as an unexpected glitch that opened up new pathways and directions, resulting in the identification of four feminist lenses for addressing the wicked problem of AIEd: Glitch feminism, Marxist-socialist feminism, decolonial feminism, and data feminism. We consider the possibilities that each perspective offers for educators who are grappling with the ever-expanding and overwhelming presence of AI.

This article explores the often-ignored embodied dimension of rurality and its role in shaping students’ transitions into higher education. Drawing on a ‘figured worlds’ framework, this narrative study examines how rural-background students in China navigate urban universities through embodied practice. Their encounters with university spaces reveal how rural-urban divides are inscribed through material inequalities, sensory hierarchies, and classed subjectivities. The metaphor ‘dancing with bonds’ captures how students negotiate structural constraints through bodily improvisations, reconfiguring their positionalities and identifies within elite university spaces. Their narratives illustrate how the emotional imprints of rurality, as enduring structures, are actively renegotiated through embodied agency, fostering new ways of belonging and becoming. By positioning rurality as a distinct analytical lens, this study underscores the need for greater attention to embodiment in research on rural students’ lived experiences, highlighting the complex interplay of positionality, agency, and identity formation.

Class-based parenting approaches assume that children in working-class environments participate less in extracurricular activities than those in affluent areas. However, there is a mediating role played by parents navigating out-of-school educational opportunities and neighbourhood risks, also shaped by families’ ethnic and migration ties. This article explores the diverse patterns of participation in organised leisure activities among primary school children in disadvantaged areas. We use latent profile analysis on survey data from three Barcelona neighbourhoods (N = 731). To identify patterns, we measure the intensity and breadth of children’s participation across public, private, community, and home settings, both at school and outside it. The analysis reveals seven participation profiles in organised leisure activities, with migration status shaping the differentiated use of educational spaces. These patterns indicate disparities in access to the breadth and quality of activities and help to identify the challenges of school-based programmes in mitigating segregation during children’s free time.

This article applies Foucault’s theories of surveillance, disciplinary power, and normalisation to examine the shifting power dynamics in AI-mediated education. Drawing on qualitative responses from 27 English and Chinese-speaking stakeholders, including students, teachers, administrators and parents with no geographical restrictions, this study investigates how AI surveillance reshapes behaviours and perceptions and identifies emergent norms in education. Thematic analysis revealed four key themes: normalising ubiquitous surveillance and behavioural control, prioritising efficiency over autonomy, reaffirming the importance of human elements in AI-assisted education, and ensuring human-AI collaboration. Machine learning is not neutral but an active agent of algorithmic control, reflecting a post-panoptic power structure. It introduces new forms of disciplinary power, encouraging behaviours aligned with efficiency at the expense of autonomy and privacy. Despite these shifts, the findings underscore the vital role of human engagement, pointing to a future where human agency remains central in regulating and complementing AI in education.

In the last decades, the field of applied linguistics has called for an expanded knowledge on language learning in different learner groups and contexts, including forced migrants and low-literate learners. This article focuses on an under-studied learner group in an under-studied context: low literate adult refugees (LESLLA) subjected to language requirements to ensure their legal status in their host country. We combine the concepts of liminality and emotion to analyze the migrants’ narrated experiences of second language learning and linguistic integration. The study draws on interviews with 14 low literate refugees in Norway. The analyses reveal how the learners position themselves towards, and are being positioned by, language learning, education for integration, and formal migration requirements. We argue that within liminal spaces arising in the process towards legal integration, learners are placed in a prolonged state of unsafety and liminality that hamper the conditions and motivation for second language learning.