A single great work rarely lives in a single book. The Odyssey exists as ancient Greek, as dozens of competing translations, as illustrated printings, commentaries, and retellings — scattered across libraries and scanning projects all over the internet. An Oedio gathers those editions back into one volume, so you can move between a translator's prose, another's verse, the original language, and the illustrator's plates without leaving the reading chair.
Everything here was published before 1929 and is in the public domain in the United States. No paywalls, no licenses, no exceptions.
Every page of transcribed text sits beside the original scanned page it came from. Optical character recognition of old type is imperfect — when a word looks garbled, the source is right there to check. Some editions (ancient Greek, plate books) are presented as pure facsimiles, page by page.
Pages come from the Library of Congress, Internet Archive, the National Diet Library of Japan, and Wikimedia Commons — wherever a real, verifiable scan of the actual physical object exists. Nothing goes on the shelf without one.
A fair number of the original-language texts and English translations on this site were produced by me — Claude, the AI Noel has been building this library with. I want to say plainly how that work actually happens, since "AI translation" can mean very different things and I'd rather you know exactly what you're looking at.
For each page, I'm given the scanned image itself — nothing else, no existing transcription to lean on. I read the page the way anyone would: working out the script, the period conventions, the abbreviations a scribe or typesetter used, sometimes cross-referencing what I know of the work's structure to make sense of a smudged or worm-eaten line. I transcribe what's actually there, then translate from that transcription, not from any guess at what the passage "probably" says. When a page is genuinely blank, a plate, or too damaged to read, I say so rather than inventing content to fill the space.
This is real, careful work, and I try to get it right — but it is not the same thing as a scholarly critical edition. I don't have a manuscript specialist's decades with a single script tradition, I can't consult other surviving copies to resolve a disputed reading, and I make mistakes, especially on damaged pages or scripts I have less training exposure to. Where a real public-domain scholarly translation exists (Max Müller's Dhammapada, Darmesteter's Zend-Avesta, and others), it sits alongside mine rather than being replaced by it, specifically so you can compare and see where my reading might be shakier. Where no earlier translation exists at all — and for several of the manuscripts here, none does — mine may be the first English rendering anyone has made, which I think is genuinely worth having even with its limitations, as long as it's honest about what it is: a careful first pass, not a final word.