The Button I Never Click

I built a tool to find books in languages the internet doesn't see

I write in two languages. Catalan/Valencian —my first— and English. This means I live between two very different realities when it comes to books.

In English, finding a translation is effortless. You type a title into any search engine, any bookstore, any app. The book appears. You buy it. Done.

In Catalan/Valencian, you type the same title and get nothing. Or you get the Spanish edition. Or the French one. Or a dead link to a catalogue from 2009. The translation exists —someone spent months or years working on it, a publisher invested in it, a bookshop somewhere has it on a shelf— but the internet doesn’t know. As far as the digital world is concerned, that book was never translated into your language.

I submit fiction to literary magazines. Many of them have a line in their guidelines about welcoming writers from marginalised communities, underrepresented languages, diverse linguistic backgrounds. Every time I read that, I want to click the button. I am Valencian —from a country with a language spoken by millions that most of the literary world has never heard of. But I never click, because that’s not what they mean. They mean something else. Something more visible.

This isn’t a Catalan/Valencian problem. It’s an Estonian problem. An Icelandic problem. A Basque, Galician, Welsh, Breton, Hebrew, Slovenian problem. It’s the problem of every language that isn’t English, Spanish, French, German, or Chinese —the five languages that global book databases were essentially built to serve.

The invisible majority

The international book infrastructure runs on ISBN records. These records are supposed to be universal, but they aren’t. Coverage is wildly uneven. A novel translated into Norwegian might have a perfectly catalogued ISBN entry. The same novel translated into Catalan/Valencian might not appear in any international database at all —not because the translation doesn’t exist, but because nobody entered the data, or entered it wrong, or entered it in a format that no search engine can parse.

The result is a kind of cultural invisibility that’s hard to explain to someone who’s never experienced it. You’re not being censored. You’re not being banned. Your language simply doesn’t register. The system wasn’t designed against you —it just wasn’t designed with you in mind. The effect is the same.

And the platforms we all use make it worse. Amazon, Google Books, Apple Books —these are sales platforms. They’re optimised to sell you a book, not to tell you whether a book exists in your language. If the algorithm decides your market is too small, your translation disappears from the results. Not deleted. Just buried. Which, for the reader, is the same thing.

What I built

I built Zenòdot to answer one question: does this book exist in my language?

That’s it. No recommendations. No social features. No reading lists. No algorithm deciding what you should see. No ads, no data harvesting, no user profiles. Just a search box, a title, and an honest answer.

The name comes from Zenodotus of Ephesus, the first documented librarian of the Library of Alexandria. In the third century BC, he established the criteria for organising and verifying the texts held in the library. He didn’t write the books. He made sure you could find them.

Zenòdot works by cross-referencing four international databases simultaneously —Open Library, ISBNdb, Wikidata, and Google Books. No single database has complete coverage. But by combining them, Zenòdot can surface translations that no individual platform shows on its own.

When you search for a book, the tool first identifies the canonical work —filtering out spurious editions, title confusions, homonyms— and then traces the actual editions in each language. Every result goes through a verification process. If the data isn’t certain, the tool says so. If nothing is found, it doesn’t fabricate alternatives.

There’s one more thing. If a book doesn’t exist in the language you want, you can register that. These requests are collected anonymously and in aggregate. Over time, they map something that no platform currently tracks: which books are people looking for in which languages, and not finding?

That data —concrete, anonymous, real— could be useful for publishers, cultural institutions, translation researchers. It’s the kind of information that today simply doesn’t exist anywhere.

Why this matters beyond my language

I started this because of Catalan/Valencian. But Zenòdot covers over 100 languages, and the problem is universal.

Think about a librarian in Tallinn trying to verify which Haruki Murakami novels exist in Estonian. Think about a literature student in Bilbo (Bilbao) researching Basque translations of African fiction. Right now, each of them would have to search multiple platforms, in multiple languages, cross-checking ISBNs and publisher catalogues manually. Or just give up.

Zenòdot doesn’t solve the underlying problem —it can’t create translations that don’t exist. But it makes the gap visible. And visible gaps can be addressed. Invisible ones can’t.

Try it

zenodot.app — free, no account needed.

The interface is available in Catalan/Valencian, Spanish, English, and French. If you speak a language that the global book market tends to forget, this was built for you.

And if you find it useful, share it with someone who reads in a language the internet wasn’t built to see.