The Burden of Bearing Witness: Digital Practices of Marginalized Social Media Users in High-Stakes Contexts

Paper Title

The Burden of Bearing Witness: Digital Practices of Marginalized Social Media Users in High-Stakes Contexts

Publication Info

• Topic area: Social media use by marginalized populations in high-stakes contexts.
• Keywords: Social media, marginalized populations, content moderation, high-stakes contexts, digital labor, platform governance, digital colonialism, Venezuela, Nigeria, transgender.

Background and Problem

• Problem / challenge: Social media platforms fail to address the urgent infrastructural and contextual needs of marginalized users in high-stakes contexts, such as humanitarian crises, violence, and systemic oppression. Platform governance and design often exacerbate these users' precarity.
• Significance: Marginalized populations rely on social media as critical infrastructure for survival, activism, and community support. However, platform failures hinder their ability to document experiences, share resources, and advocate for their communities.
• Motivation and related work: Prior research highlights how marginalized populations use social media for care and archival purposes during crises. However, gaps remain in understanding how platform design and governance intersect with identity, marginality, and labor, particularly in high-stakes contexts. This paper aims to address these gaps.

Solution

• Proposed approach: The study examines how marginalized populations in Venezuela, Nigeria, and the U.S. use social media in non-normative ways to navigate high-stakes contexts, revealing platform design flaws and proposing alternative infrastructural features.
• Novelty:
  1. Identifies how platform failures exacerbate precarity for marginalized users in high-stakes contexts.
  2. Highlights the invisible sociotechnical repair work these users perform to navigate platform limitations.
  3. Demonstrates how platform governance replicates physical-world power asymmetries and digital colonialism.
  4. Proposes design recommendations, including emergency response tools and exit mechanisms for wellbeing.
• Procedure and key techniques:
  • Conducted three qualitative and ethnographic case studies in Venezuela, Nigeria, and the U.S.
  • Methods included interviews, autoethnography, and digital ethnography.
  • Analyzed participants’ use of social media to document crises, engage in activism, and navigate platform governance.

Results

• Concrete findings:
  • Venezuelan participants used social media to document the humanitarian crisis but faced content moderation challenges, self-censorship, and burnout.
  • Nigerian participants relied on social media to document ethnic violence and human rights violations but encountered moderation failures, especially in low-resource languages.
  • U.S. trans participants faced over-moderation of trans-specific content and harassment, highlighting the erasure of trans bodies and experiences.
• Advantage over baselines:
  • Demonstrates how marginalized users develop innovative workarounds and alternative infrastructures to address platform failures.
  • Provides a nuanced understanding of how platform governance disproportionately affects marginalized populations in diverse contexts.
• Experiments / evaluation:
  • Case studies involved 13 interviews and 8 months of participant observation in the Venezuelan context, 9 interviews and digital ethnography in Nigeria, and 104 interviews with trans technology creators in the U.S.
  • Data were analyzed thematically to identify commonalities and differences across contexts.
• Limitations and future work:
  • Platforms’ lack of financial incentives to implement inclusive design poses a significant barrier.
  • Future research should explore how structural and systemic challenges hinder the adoption of design recommendations and further impact marginalized users.

Summary

This study examines how marginalized populations in Venezuela, Nigeria, and the U.S. use social media in high-stakes contexts to document crises, advocate for their communities, and navigate platform governance. The findings reveal that platform failures exacerbate users’ precarity and force them to engage in invisible sociotechnical repair work. The paper highlights the structural flaws in platform design, rooted in digital colonialism, and proposes alternative infrastructural features, such as emergency response tools and exit mechanisms for wellbeing. However, the authors acknowledge the challenges of implementing these recommendations due to platforms’ prevailing logics of efficiency and power asymmetries. Future work should address these systemic barriers to create more inclusive social media platforms.