The Pristine Deception: Why Perfect Digital Preservation May Be the Enemy of Access

It’s an article of faith in our field: we must strive to preserve digital objects in their most pristine, original state. We speak of bit-for-bit fidelity, of cryptographic checksums that must never change, of archival masters kept in digital cold storage. This is the gold standard, the holy grail. To alter the file is to commit a kind of archival heresy. But what if this unyielding pursuit of perfect preservation is, in some critical ways, actively working against our ultimate goal: access and understanding across time?

Consider a simple text file from 1994, saved in a proprietary word processor format. Our orthodox preservationist triumphantly secures the original .DOC file, verifies its hash, and stores it in three geographically dispersed repositories. The bits are safe. The record is preserved. And yet, it is functionally dead. Without the specific, long-extinct software to interpret it, the file is an inert digital fossil. The perfectly preserved object is a perfectly sealed one.

My counterintuitive proposal is this: sometimes, the most ethical act of preservation is not to guard the original, but to conscientiously and transparently transform it into a form that future beings—human or machine—can actually use. This isn't about carelessness; it's about prioritizing utility over purity. It’s the archival equivalent of translating a dead language. We lose the precise typography of the cuneiform tablet by translating it to English, but we gain the meaning, the story, the history. The translation itself becomes a new, vital preservation layer.

The Living Copy as Archive

This challenges the core tenet that the ‘original’ is sacrosanct. In a digital realm where the ‘original’ is just a pattern of magnetic platters or electrical charges, what is the true object? The pattern, or the human-readable information it represents? By treating the migration to open, sustainable formats not as a necessary evil but as a core preservation act, we shift the paradigm. We accept that a well-documented, openly readable PDF/A, or a plain-text file, derived from that 1994 .DOC, is more preserved than the locked original. It has been given a new, longer life in a lingua franca.

This principle extends beyond file formats. Think of a massive, complex database released as open data. The orthodox approach demands the entire, structurally perfect dump. But what if the sheer, unwieldy complexity of that ‘pristine’ dataset makes it impenetrable to journalists, community activists, or curious citizens? A thoughtfully simplified extract, or a version pre-joined with key geographic data, might be the thing that actually gets used, interrogated, and kept alive in the public discourse. The ‘imperfect’ version becomes the conduit for understanding, while the ‘perfect’ master sits silently in the vault.

This is not an argument for sloppiness. Every transformation must be meticulously documented—a process often called preserving the ‘provenance’ and ‘chain of custody.’ The goal is to create a lineage, not a single, frozen point. By fetishizing the original bitstream above all else, we risk building cemeteries of flawless, unreadable data. Perhaps it’s time we valued the living, breathing, usable copy just as highly. After all, an archive no one can understand is merely a tomb.

Notes & further reading

A few pages I came back to while writing this: