The Clock in the Cathedral: What Old Stone Can Teach Us About Digital Permanence

I was thinking recently about a clock. Not a smartwatch or a server timestamp, but a centuries-old clock carved into the stone of a European cathedral. It tells the time, of course, but it also tracks the phases of the moon and the movements of the stars. It’s a complex, beautiful machine designed to be understood by the community that built it and to last far beyond their lifetimes. It’s still there, still working, still readable. It struck me that this clock is a perfect, ancient analog for what we’re trying to achieve with digital preservation: creating a public record meant to endure.

Our modern approach to preserving public data often feels like the opposite of that cathedral clock. We build intricate digital systems—formats, protocols, repositories—that are incredibly powerful but also incredibly fragile. They are not carved in stone; they are etched in sand, vulnerable to the next tide of technological change. A PDF becomes unreadable, a database schema is forgotten, a proprietary software license expires. The data might still exist as bits on a drive, but its meaning, its context, its readability has evaporated. It’s a clock without hands.

The cathedral builders operated on a principle we’ve largely abandoned: the separation of the mechanism from the message. The clock’s functionality—telling time—is inseparable from its physical, human-readable form. You don’t need a specific, licensed software program from a now-defunct company to interpret its gears and dials. The knowledge required to read it is built into our shared human experience of the sky and the passage of days.

This is the profound challenge and opportunity of open data and digital preservation. It’s not enough to just save the ones and zeros. We must also save the means to understand them. This means championing open formats, detailed and persistent metadata, and emulation strategies that are themselves open and documented. It means thinking like a stonemason, not a software engineer. We must ask: if someone stumbles upon this data in a hundred years, with no manual and no living memory of its origin, will they be able to make sense of it? Will our digital clock still tell time?

The goal isn’t to rebuild the internet in stone, but to imbue our digital public squares with some of that cathedral’s resilience. It’s about designing for permanence and clarity, ensuring that the records we fight to preserve and make open today don’t become indecipherable relics tomorrow. They should remain what they were always meant to be: a working clock for the community, for as long as we need to tell the time.

Notes & further reading

A few pages I came back to while writing this: