Seventeen Students, Six Conferences, One Mentor. Rethinking Open Access Equity in South African Higher Education

Abstract

UNESCO's Recommendation on Open Science (UNESCO, 2021) envisions equitable participation in scientific knowledge production regardless of socioeconomic circumstance. Yet as Chan et al. (2020) argue, open science as currently operationalised risks reinforcing inequality in asymmetrical systems, privileging institutions already embedded in the knowledge ecosystem while leaving marginalised communities further behind.   This paper draws on direct mentorship of 17 undergraduate Mathematical Sciences students undertaking industry-based learning in South Africa to examine where this gap materialises in practice. Most arrived with passing grades masking shallow understanding: unable to interpret their own findings or sustain subject-matter conversations. Following structured mentorship and real-world exposure, students answered faculty and industry panel questions with a depth of understanding that visibly distinguished them from peers, a transformation directly attributed to real-world engagement. Karassellos et al. (2025) document precisely this disjuncture: formal university access does not equate to participation in the knowledge ecosystem, particularly for students whose backgrounds are not reflected in institutional cultures.   Financial precarity compounds the problem. Socioeconomic status significantly moderates academic adjustment and dropout intention among South African undergraduates (Mtshweni, 2021), with students frequently unable to access the conferences and professional networks that contextualise academic knowledge (Letseka, 2010). In 2024, South Africa's largest mathematical conference was co-chaired by a Vice Chancellor whose own mathematics department were not informed of free attendance opportunities; external mentorship intervened.   Through facilitation across six major events including the ACIAM Conference, Africa Tech Festival, and the Interledger Summit, all 17 students attended a minimum of three. A hackathon team peer-voted second place had never previously attended such an event. Drawing on Chan et al.'s (2020) framework for open science for and with communities, this paper proposes mentored open access as a replicable bridge and asks what structural obligations institutions must assume to make individual intervention unnecessary.