The Myth of the Immortal PDF: On the Perishable Nature of 'Permanent' Records
We have been sold a bill of goods on permanence. In our rush to digitize the world’s paper, we placed an almost theological faith in a single format: the PDF. It is the default, the standard, the go-to for any institution wishing to present a document as finished, official, and above all, permanent. We click “Save As PDF” with the quiet satisfaction of someone etching words into stone. The deed is done. The record is saved for eternity.
But this is a comforting illusion. The PDF, for all its locked-in-place aesthetics, is not a monolith. It is a complex wrapper, a digital Russian doll that can contain text, images, fonts, and JavaScript, all held together by a specification that evolves. And like any complex system, its future readability is not guaranteed. The very thing that makes a PDF seem so concrete—its fidelity to a specific visual layout—is its greatest vulnerability. It prioritizes the appearance of permanence over the reality of accessibility.
The Brittle Blueprint
Consider a PDF created today using the latest software. It might rely on proprietary fonts embedded in a specific way. It might contain interactive elements built on current web standards. It might be secured with encryption that, while robust now, may one day be obsolete or unsupported. Now, fast-forward fifty years. Will software still exist that can perfectly unpack this specific combination of elements? Or will we be left with a digital artifact that is visually intact but functionally crippled—a beautiful, unreadable tombstone?
This is the core of the critique: the PDF is often a record of how a document *looked*, not what it *means*. True preservation requires a focus on the underlying data, the plain text, the semantic meaning that can be migrated to new formats as technology changes. The PDF, in its quest for perfect visual replication, often buries that data in a presentation layer that is destined to become a relic.
We see this already. Older PDFs, particularly those built from scans, are often mere images of text, completely opaque to search engines and screen readers. They are digital photocopies, not digital records. They satisfy the letter of public records law—the document is “available”—while utterly failing its spirit of accessibility and utility.
The received wisdom we must challenge is that creating a PDF is an act of preservation. It is not. It is an act of packaging. The real work of preservation comes later, in the continual effort to ensure that the contents of that package can still be unpacked and understood. It requires a mindset that sees the PDF not as a final destination, but as one potential snapshot in a document’s long life. To mistake the snapshot for the subject is to confuse the map for the territory. In our quest to build a lasting record, we may have instead built a vast, fragile museum of yesterday’s software, waiting for its lights to flicker out.
Notes & further reading
A few pages I came back to while writing this: