SourceLibrary: A Vision for AI-Assisted Translation
Half a million Renaissance Latin texts await translation. We're building tools to make that possible—not by replacing scholars, but by empowering them with expert-driven, AI-assisted workflows.
Exploring the hidden libraries of Renaissance Latin—data, methodology, and discoveries.
See data visualizations →FEATURED
Half a million Renaissance Latin texts await translation. We're building tools to make that possible—not by replacing scholars, but by empowering them with expert-driven, AI-assisted workflows.
The Renaissance produced half a million Latin works. 97% have never been translated. What are we missing?
A systematic approach to translation priorities. We score works by historical impact, translation gap, feasibility, and audience to identify what to translate first.
ALL ESSAYS (18)
Half a million Renaissance Latin texts await translation. We're building tools to make that possible—not by replacing scholars, but by empowering them with expert-driven, AI-assisted workflows.
How we evolved from prefix matching (2%) to fuzzy strings (18.6%) to semantic embeddings (65%) to multi-signal matching (26%). The story of matching the BPH catalog against the Internet Archive.
We matched 10,683 Latin works from Amsterdam's Hermetic library against the Internet Archive. Fuzzy matching found 18.6%—with dramatic variation by century (65% for 15th c., 11% for 20th c.).
We cross-referenced Amsterdam's esoteric collection with standard bibliographies. 35% of their 15th-century holdings don't appear in ISTC or USTC. These aren't marginal texts—they include Ficino, Pico, and Hermes Trismegistus.
German overtook Latin in the 1670s. We analyzed the complete USTC database to pinpoint exactly when Europe's lingua franca lost its dominance.
We count 500,000 surviving Latin works. But estimates suggest 25-80% of editions were lost entirely. What does this mean for our understanding of the Renaissance?
Theology is the largest category in the Latin corpus—and the most misunderstood. Why we don't lead with it, and why it still matters.
Kircher, Sennert, Weyer, Liceti. 324,690 Latin works from the Scientific Revolution—and the scholars behind them who remain untranslated.
Zabarella, Cardano, della Porta, Telesio. The century of the Scientific Revolution—and the Latin authors who shaped it but remain inaccessible.
Giorgio Valla, Giovanni Pontano, Paul of Venice. The 15th century gave us the Renaissance—and hundreds of Latin thinkers whose works remain untranslated.
Jakob Martini wrote 836 works. Johann Gerhard wrote 697. You've never read a word they wrote—because almost none of it has been translated.
You'd think the famous Renaissance humanists would be fully available. They're not. Ficino, Pico, Valla—vast bodies of work remain untranslated.
Some Latin books went through 100+ editions. They shaped European thought for centuries. Today, they're completely inaccessible.
Only 416 Latin works ever appeared in Latin-English bilingual editions. The numbers reveal a staggering accessibility crisis.
How many Latin works have been translated into English? I catalogued 3,232 translation volumes across 45+ sources to find out. The coverage rates surprised me.
How do you count something that's never been counted? Building a comprehensive database of Latin-to-English translations—and discovering how much we don't know.
Applying Forlong's 'Rivers of Life' methodology to trace how Hermetica, alchemy, Kabbalah, and Rosicrucianism flowed through Renaissance publishing. Draft for discussion.
Documenting the sources and methods behind our accessibility estimates. How we derived the 18% digitized, 8% OCR, and 3% translated figures.