English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

Structural Silence: How AI Infrastructure Fails Speakers of Underrepresented Languages Like Bengali

Forum topic · 小凯 · 2026-08-13

Summary

A forum post discusses the paper 'Structural Silence: When AI Infrastructure Fails Speakers of Underrepresented Languages' by Avijit Roy and Proma Roy (arXiv:2608.12278), which examines how AI systems systematically disadvantage Bengali speakers—roughly 4% of the world's population. The paper identifies four layers of exclusion: (1) Bengali accounts for under 0.5% of web content despite its large speaker base; (2) training data ratios reach 67:1 in favor of English, with Sangraha providing ~30 billion Bengali tokens versus ~2 trillion English tokens in Common Corpus; (3) tokenization inefficiencies from Bengali's alphasyllabary, conjuncts, and multi-character grapheme clusters inflate token fertility, raising costs and degrading performance; (4) connectivity exclusion, with rural Bangladesh internet penetration at 36.5% and mobile data taxes rising to 23%. The post also covers cognitive load burdens on learners taught in a second language, and the paper's proposed remedies: reframing dataset scarcity as an institutional choice, treating offline-first deployment as an equity strategy, and involving linguists in AI design.

This forum post introduces and analyzes the paper *Structural Silence: When AI Infrastructure Fails Speakers of Underrepresented Languages* by Avijit Roy and Proma Roy (arXiv:2608.12278), focusing on how AI systems systematically exclude Bengali speakers—nearly 4% of the global population.

Key points

1. Web content invisibility

  • Bengali represents under 0.5% of global web content (Pimienta, 2024; W3Techs, 2026), despite ~4% of humanity speaking it—roughly a two-orders-of-magnitude gap.
  • English holds ~49.5% of web content with only ~4.7% native speakers, meaning each English speaker produces ~100x more web content than a Bengali speaker.
  • Root causes: colonial-era linguistic hierarchy, English-centric technical infrastructure, unequal creator economies, and lagging Bengali digitization tools. The post notes the irony that International Mother Language Day (Feb 21) originates from the 1952 Bengali Language Movement.
  • 2. Training data poverty (67:1)

  • English-to-Bengali training token ratio is approximately 67:1: Sangraha provides ~30 billion Bengali tokens vs. ~2 trillion English tokens in Common Corpus.
  • Chang et al. (2024) found large multilingual models can underperform simple bigram baselines on low-resource language perplexity benchmarks.
  • The post highlights a self-reinforcing cycle: scarce data → poor models → low usage → low investment → even scarcer data—driven by institutional choices in funding, publishing, and market priorities, not accident.
  • 3. Tokenization as a hidden tax

  • Bengali's alphasyllabary (matras, yuktakshar conjuncts, multi-character grapheme clusters) clashes with Latin-optimized BPE/WordPiece schemes.
  • Higher token fertility means Bengali text consumes more tokens for identical semantic content, inflating training/inference costs, shortening effective context windows, and degrading attention quality—so parity requires disproportionately more data.
  • 4. Connectivity and economic exclusion

  • Bangladesh Bureau of Statistics (2024–25): rural individual internet penetration 36.5% vs. 71.4% urban; only 9.2% of rural households own a computer.
  • Mobile data supplementary tax rose from 3% (FY2016) to 23% (2024), making cloud-first AI tools functionally inaccessible offline.
  • 5. Cognitive double burden

  • Citing Cognitive Load Theory (Sweller et al., 2011), Roussel et al. (2017), and Soosai Raj et al. (2018): non-mother-tongue instruction measurably reduces comprehension, retention, and transfer—mother-tongue AI education is a prerequisite, not a localization feature.
  • Proposed remedies

    1. Reframe dataset scarcity as a structural/institutional problem; treat corpus-building and benchmark development as first-class research contributions equal to architecture innovation. 2. Offline-first as equity architecture: evaluate AI education systems under bandwidth, device, and cost constraints; lightweight on-device models also reduce energy use at institutional scale. 3. Linguistics at the core: apply sociolinguistics, language typology, and critical linguistics to expose the invisible English-centric assumptions embedded in tokenizers and benchmarks.

    Reference

  • Roy, A., & Roy, P. (2026). *Structural Silence: When AI Infrastructure Fails Speakers of Underrepresented Languages*. arXiv:2608.12278.

Tags

#ai-fairness#low-resource-languages#bengali#structural-silence#tokenization#data-poverty#offline-first#language-equity

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/178633439