The Myth of the Immortal Bit: Why Data Doesn't Want to Be Preserved
There’s a comforting, almost sacred mantra in our field: that digital preservation is a battle against time, fought with checksums and migration paths. We speak of data as if it possesses an innate, Platonic desire for immortality, a will to persist that we, as archivists and custodians, are honor-bound to fulfill. We’ve convinced ourselves that a bit, once correctly written, is eternal. This is the myth of the immortal bit, and it’s a dangerous simplification that obscures the true, messy nature of our work.
The myth suggests that preservation is a purely technical problem. If we can perfectly copy a sequence of ones and zeros from one medium to the next, forever, the job is done. But data is not just a sequence; it is a message, and a message requires a context to be understood. We can perfectly preserve the bits of a Microsoft Word 97 document, but without the software, the operating system, the fonts, and the cultural context to interpret its formatting and intent, those bits are a fossilized skeleton. We have the bone structure, but not the flesh, the breath, or the life.
This fallacy leads us to build pristine, climate-controlled digital mausoleums filled with perfect, un-changing copies of data that is effectively already dead. We preserve the vessel but lose the wine. The real work of preservation isn't just about preventing bit rot; it's about preventing meaning rot. It’s an act of continuous, active interpretation. It requires emulators, documentation, and a deep understanding of the original ecosystem that gave the data its purpose. A spreadsheet from 1985 isn’t just a collection of numbers; it’s a artifact of specific business logic, software behavior, and user intention that must be excavated and explained.
The Inevitable Decay of Context
This is where the received wisdom truly fails us. It implies that with enough effort, we can achieve a state of perfect, static preservation. But context is inherently fluid and decays at an alarming rate. The hyperlinks in a early web document don’t just break; the entire network of meaning they pointed to vanishes. The joke in a newsgroup post from 1993 loses its punchline without the specific social dynamics of that group at that moment. We can save the text, but the subtext evaporates.
By clinging to the myth of the immortal bit, we risk focusing our resources on the easiest part of the problem—the physical integrity of the data—while the much harder problem, the preservation of its significance, quietly slips through our fingers. Perhaps it’s time to shift our metaphor. Instead of immortal guardians, we are more like translators or curators. Our goal isn’t to stop time, but to build bridges across its gaps, providing the commentary, the tools, and the frameworks that allow future generations to not just read the bits, but to understand the world they came from. The data doesn’t want to be preserved; it simply is. It is we who want to remember, and that is a far more human, and more complicated, endeavor.
Notes & further reading
A few pages I came back to while writing this: