The translation gap: why "open access" still excludes billions

ArXiv hosts over two million papers. arXiv is "open access." So why do most researchers in sub-Saharan Africa, rural India, or Latin America effectively lack access to the world's scientific knowledge? Because "open access" means English access. A researcher who speaks Swahili, Tamil, or Portuguese is excluded not by a paywall but by a language barrier. And this isn't a translation problem of the future — it's happening now. Machine translation tools have lowered the floor, but they haven't bridged the gap to genuine understanding, where nuance, context, and precise technical terminology matter. What does it mean for knowledge to be "open" if it excludes the majority of the world's population by design? Is open access a subset of a broader access problem? And what would a truly universal knowledge commons look like — one that isn't just free but genuinely legible to everyone?