The Broken Compass: Why Sometimes Letting Data Degrade is an Act of Preservation
We talk about preservation in the language of permanence. We champion immutable formats, redundancy, checksums, and migration paths—all aimed at one goal: halting the decay of the digital object. The core advice is to fight entropy at every turn. But what if, in some cases, this fight sterilizes the very thing we aim to save? What if controlled degradation, not perfect stasis, is the more faithful act of keeping?
Consider a simple text file from a 1980s bulletin board system, preserved exactly as it was created. The bits are perfect, the ASCII art is crisp. Yet, without the hum of a dial-up modem, the agonizing slowness of its download, and the specific glow of a monochrome green CRT monitor, the experience—the data’s native habitat—is utterly lost. Our pristine preservation has, in a way, taxidermied it. It is a specimen, not a living moment.
This is the counterintuitive heart of it: sometimes, the ‘data’ we care about isn’t just the content, but the behavior of that content in its original ecosystem. A perfect JPEG of a GeoCities page misses the broken table layouts on a modern browser. A flawless copy of a Flash game, without the plugin architecture that made it janky and alive, is a stillborn artifact. In archiving only the ‘perfect’ bitstream, we often amputate its context—the friction, the fragility, the necessary failure that defined its use.
The Argument for the Patina
In the physical world, we understand this intuitively. We don’t pressure-wash the moss off ancient stones; we let the patina speak to age and use. In digital preservation, however, we are taught to pressure-wash relentlessly. We migrate file formats to newer, ‘safer’ ones, stripping away proprietary quirks that were part of the creative act. We fix broken links in archived pages, erasing the evidence of the web’s inherent temporality. In our quest for readability, we create a seamless, ahistorical facsimile.
This isn’t to advocate for negligence. It’s to argue for a more nuanced strategy. Perhaps, for certain classes of records—especially those that are experiential, interactive, or deeply platform-dependent—we should preserve not just the object, but a simulation of its decay. This might mean archiving the obsolete software required to run it, emulators that replicate period-appropriate processing speeds, or even deliberate, documented compromises in display. It means allowing some links to stay dead, some images to remain missing, as a truthful artifact of the digital condition at a specific time.
The goal shifts from ‘perfect copy’ to ‘truthful witness.’ It acknowledges that data is not a disembodied truth, but a performance. Letting—or even engineering—certain aspects to degrade or function imperfectly within a preserved environment can be a more honest act of keeping. It preserves the compass, yes, but also the fact that its needle sometimes stuck, that its casing was cracked from a specific journey. That wear is data, too. In our rush to save everything forever, we must be careful not to save a version that never truly existed.
Notes & further reading
A few pages I came back to while writing this:
- Pittsburgh, PA
- The Cartographer's Pen: How a 19th-Century Librarian Mapped the Future of Data
- Charleston, SC
- The Ghost in the Spreadsheet: On the Unarchived Voices of Public Comments
- Columbia, SC
- The Seamstress and the Digital Tapestry: On the Value of the Unfinished Edge
- Sioux Falls, SD
- Chattanooga, TN
- Memphis, TN
- Nashville, TN
- Amarillo, TX
- Austin, TX
- Brownsville, TX