The Card Catalog of the Commons: How the First ISBN Registry Became a Public Good

Before a book’s metadata could vanish with a failed server or a lapsed domain subscription, it lived on a card. Not a digital card, but a physical, three-by-five-inch piece of cardstock, filed away in a cabinet in a New Jersey office. This was the original International Standard Book Number (ISBN) registry, a system born not from silicon but from paper, and its creation is a quiet lesson in building public infrastructure for the long haul.

In the mid-1960s, W.H. Smith, a British bookseller, needed a better way to manage its growing inventory. A system was developed and eventually standardized, becoming the ISBN we know today: a unique identifier for every book edition. But the question remained: who would keep the master list? The task fell to the R.R. Bowker Company in New York. For years, their primary method of disseminating this crucial bibliographic data was through the painstaking publication of massive, phone-book-like volumes called "Books in Print." If you wanted to verify an ISBN, you consulted the tome.

Yet, they did something remarkable. While Bowker maintained the registry as a commercial enterprise, the system itself was designed as an open standard. The numbers were assigned consistently and the format was shared freely. More importantly, the data within those massive volumes, while sold as a book, functionally became a readable public record. It was a dataset of immense value to libraries, publishers, and scholars—a card catalog for the entire publishing world, available to anyone who could access the reference section of a library.

This pre-web system embodies a principle we struggle with today: the stewardship of essential identifiers. The ISBN registry was a piece of critical infrastructure that everyone could use on equal footing. It wasn’t locked behind a proprietary API that could change its terms of service tomorrow. It didn’t require a login or suffer from rate limits. Its persistence was guaranteed by the physical printed page, distributed to thousands of libraries globally, creating a resilient, decentralized archive of its own data.

Today, the registry is digital, managed by Bowker and its international affiliates. The data is more accessible in some ways, yet more fragile in others. A query to a web service can replace a afternoon in the stacks, but it also depends on corporate continuity and network availability. The original, paper-based system reminds us that the most enduring form of digital preservation is sometimes analog. It asks us a pointed question about our own era: what are the essential datasets of our time, and are we building their card catalogs—their resilient, distributed, and accessible copies—for the commons, or are we merely entrusting them to the cloud?

Notes & further reading

A few pages I came back to while writing this: