The Library of Babel’s Ghost: On Applying Conservation Science to Digital Decay
In the hushed, temperature-controlled rooms of a museum conservation lab, a specialist works not to restore an artifact to its original glory, but to stabilize its decay. The goal isn't a spectacle of newness; it's a managed deceleration of time. Cracks in a ceramic vessel are filled not with matching glaze, but with a reversible, distinguishable material. The fading pigments of a portrait are protected from further light exposure, their current state documented with spectral imaging. This philosophy—an intricate dance between preservation and acceptance—holds a profound, untapped lesson for those of us trying to keep digital objects from dissolving into the static of obsolescence.
We often treat digital preservation as a problem of brute-force copying, a belief that if we just have enough redundant backups, the thing is saved. But conservation science teaches us that preservation is not about stopping time, but about understanding and mediating its effects. A 19th-century photograph will inevitably silver; the question is how to slow the chemical reaction and document the change. Similarly, a PDF from the early web will inevitably become unreadable as its proprietary code is abandoned; the question is how to migrate its significant properties—its content, its layout, its intellectual form—into a new, more sustainable vessel, while meticulously documenting what was lost or altered in the transfer.
One of the most critical concepts borrowed from the conservator’s bench is the idea of "reversibility." A conservator’s intervention should, in theory, be undoable by future experts with better techniques. In the digital realm, this translates to the principle of non-destructive normalization. When we convert a fragile, obsolete file format into a standard, open one like PDF/A or TIFF, we must keep the original bitstream alongside the new version. This creates a lineage, an archaeological record of the object’s preservation journey. The converted file is the accessible reading copy; the original is the untouched artifact, waiting for a future technology that might understand its nuances better than we can today.
Furthermore, conservation is deeply contextual. A conservator doesn't just analyze the paint; they study the canvas, the frame, the environment. Digital preservation, when seen through this lens, becomes about more than the file itself. It demands we also archive the software that renders it, the operating system that supports it, and crucially, the documentation that explains its purpose and creation. The digital object is not a solitary entity; it is an ecosystem. Preserving a Microsoft WordStar document from the 1980s is an academic exercise without also preserving, or at least exhaustively documenting, the experience of using WordStar.
Ultimately, the conservator’s approach steers us away from the fantasy of digital immortality and toward the more sober, sustainable practice of digital hospice. It is about making the inevitable decay manageable, meaningful, and well-documented. It asks us to be not just technicians making copies, but curators of a process. We are not building a pyramid to last forever, but tending a garden that will constantly require care, judgment, and a humble acceptance that some losses are inevitable. Our task is to ensure that when a digital object does fade, it does so with a clear and whispered explanation of what it once was.
Notes & further reading
A few pages I came back to while writing this: