The Immaculate Conception of Data: A Critique of the 'Born-Digital' Ideal
In the world of digital preservation, we often speak of records as being ‘born-digital.’ The term is meant to distinguish a file created on a computer from one that has been digitized from an analog source, like a scanned photograph or a typed manuscript. It carries a certain purity, an implication of an immaculate digital conception. A born-digital document is seen as the original, the authentic, the unsullied source from which all copies flow. It is the ideal we are told to strive for in our archives. But this received wisdom, I’ve come to believe, is a myth that obscures a more complex and interesting truth.
The myth suggests that a Microsoft Word document written in 1995 and saved to a floppy disk is a pristine, self-contained object. We imagine it as a perfect digital embryo, waiting to be discovered and rendered by some future digital midwife. The reality is far messier. That .DOC file is not an island. It is a nexus of dependencies. It was born not in a vacuum, but within the specific ecosystem of Windows 3.1, a particular version of Word, a specific printer driver that might have influenced its page layout, and a set of system fonts that may no longer exist. Its very existence is contingent on a vast, invisible scaffolding of software and hardware that we never think to preserve alongside it.
This framing of ‘born-digital’ as a pure state leads us to make a critical error. We focus our preservation efforts on the file itself—the bitstream—believing that if we can perfectly preserve those ones and zeros, we have saved the record. But we often fail to capture the context, the environment, the experience of the data. A preserved .DOC file is not the document. The document is the text as it was intended to be rendered, with its formatting, its pagination, its typographical choices. Without the original environment, we are preserving a ghost, a potentiality that we may never be able to fully reconstruct.
This critique isn’t just academic pedantry; it has real consequences for how we build archives. By fetishizing the ‘born-digital’ object, we risk creating collections of beautiful, perfectly preserved corpses—files that are technically intact but contextually dead. The true unit of preservation should not be the file, but the computational environment required to make it legible. We need to be archiving not just data, but the operating systems, the applications, the emulators, and the documentation that give that data meaning.
Let’s abandon the misleading purity of ‘born-digital.’ Instead, let’s speak of ‘system-born’ records. This acknowledges that no digital file is ever truly born alone. It is always a product of a complex system, and its preservation is only complete when we acknowledge and attempt to preserve, or at least document, that system of birth. Our goal should not be to capture a mythical digital original, but to preserve the conditions for its continued life.
Notes & further reading
A few pages I came back to while writing this: