The Museum of Broken Glass: On the Frailty of Perfect Digital Copies
We’ve been told a comforting lie for decades, one that forms the bedrock of our faith in the digital realm: a perfect copy is the same as the original. This is the received wisdom, the dogma of the digital age. A JPEG, a PDF, a WARC file—these are not mere representations, we assure ourselves, but flawless clones. They suffer from no degradation, no loss of fidelity. The Mona Lisa on your screen, we insist, contains the same data as the file on the Louvre’s server. It is, for all intents and purposes, identical. This belief underpins everything from open data initiatives to web archiving, allowing us to sleep soundly, believing our digital artifacts are safe.
But are they truly the same? The data might be identical, a perfect bit-for-bit clone, but the context, the aura, the very essence of the artifact, is not. Consider a piece of ancient Roman glass, dug from the earth. A high-resolution 3D scan can capture every crack, every iridescent shimmer of its weathered surface. The scan is a perfect copy of the glass as it exists today. And yet, we instinctively understand that the scan is not the glass. The glass has weight, temperature, a history you can feel in its pores. The scan is information; the glass is an object. The problem is that with digital objects, this distinction collapses artificially. We mistake the informational clone for the experiential original.
Web archives are particularly poignant museums of this broken glass. A WARC file perfectly captures the HTML, the images, the CSS of a webpage on a given Tuesday afternoon. But it cannot capture the fleeting JavaScript animation that loaded only for users in a specific timezone. It cannot recreate the feeling of being part of a live comment thread, the dynamic buzz of a community that has long since dissolved. The archive is a stunningly detailed diorama of a moment, but the ecosystem is gone. We have preserved the specimen, but we’ve lost the habitat. The perfect copy is, in this crucial sense, a profound misrepresentation.
This has dire implications for public records. A scanned PDF of a century-old memorandum is not the memorandum. It lacks the texture of the paper, the faint scent of ink and dust, the marginalia from a clerk who handled it in 1923. These physical properties are part of the record’s provenance, its story. When we declare the digital scan to be the legal equivalent, we are quietly discarding a layer of evidence, a stratum of history. We are prioritizing the textual content over the physical testimony, pretending the vessel doesn’t matter when it often holds the key to understanding the message.
Our obsession with creating perfect, static copies is a response to the terror of digital decay. But in our rush to preserve the ‘what,’ we are often neglectful of the ‘how’ and the ‘where.’ The digital original is not just a collection of bits; it is an event, a performance within a specific technological and human context. The perfect copy is a ghost—a precise, data-rich ghost, but a ghost nonetheless. It haunts our archives, a reminder of what we saved so meticulously, and all that we allowed to slip away in the process. True preservation requires us to look beyond the bits and acknowledge the fragility of the whole.
Notes & further reading
A few pages I came back to while writing this: