The Cathedral Clockmaker's Lesson: On Designing for Unknowable Futures
Deep in the vaulted stone of a European cathedral, a timepiece ticks. It is not a modern quartz movement, sealed and precise, but a sprawling contraption of iron and brass, its gears and levers open to the air. The cathedral clockmaker who built it centuries ago faced a problem not unlike our own in digital preservation: how to create something that would endure long after their own understanding of time had become obsolete. Their solution, born from a craft wholly unrelated to servers and datasets, offers a surprisingly potent model for our work today: the principle of the open mechanism.
These ancient clocks were not built for eternal, perfect function. The makers knew the weights would need replacing, the escapement would wear down, the very bell that struck the hour might one day crack. Instead of striving for an unchangeable perfection, they designed for maintainability. Every part was accessible. The logic of the movement was not hidden in a sealed case but laid bare in the interaction of its components. A future craftsperson, who might have no original schematics and a completely different set of tools, could still observe the mechanism, diagnose its failing part, and craft a replacement. The clock could evolve, its original timekeeping soul preserved even as its physical body was renewed piece by piece.
We face a similar challenge with digital preservation. Our instinct is often to create a perfect, self-contained archive—a digital vault that seals a file format, an application, and its operating environment in a kind of technological amber. We try to future-proof by predicting what software will be available in a hundred years. But this is the equivalent of the clockmaker trying to build a clock that never needs maintenance, an impossible task. Formats become unreadable, emulators become obsolete, and the very context in which data was created fades from memory. The sealed archive becomes a black box, its contents inaccessible not because they are corrupted, but because the key to their logic has been lost to time.
The cathedral clockmaker’s lesson is to design for the unknown maintainer. In our realm, this means prioritizing readable, well-documented, and simple formats over complex, proprietary ones. It means embracing open standards and creating thorough metadata that explains not just what a file is, but why it was created and how its component parts relate to one another. It is about building archives that are, like the open clock mechanism, transparent in their construction.
The goal is not to build a time capsule to be unearthed by future archaeologists who must painstakingly decipher it. The goal is to build a living clock tower where future stewards, equipped with tools we cannot imagine, can walk in, look at the workings, and understand the original intent well enough to keep it ticking. It is an act of profound humility. It admits that we cannot foresee the future, but we can build with a generosity that allows the future to understand us. The true measure of our preservation efforts may not be in the bits we save today, but in the clarity of the instructions we leave etched, not in stone, but in the very structure of the data itself.
Notes & further reading
A few pages I came back to while writing this: