The Unforgetting Machine: Why Digital Preservation Needs More Deliberate Amnesia

We live in an age of perfect, terrifying memory. Our professional mantra in digital preservation is ‘save everything.’ We build vast repositories, crawl the entire web, and hoard petabytes of data with a collector’s zeal, believing that our duty is to create an unbroken, comprehensive record for the future. But what if this instinct, however well-intentioned, is creating an intellectual and ethical quagmire? What if the most responsible act of preservation is sometimes a deliberate, curated act of amnesia?

The common advice is to err on the side of inclusion. Disk space is cheap, the reasoning goes, so why not keep it all? This logic ignores the immense, non-financial cost of context. A petabyte of poorly described, context-starved data is not a treasure trove; it’s a landfill. Future historians will not thank us for gifting them a billion unlabeled photographs, a trillion social media posts stripped of the social fabric that gave them meaning, or the raw log files of a thousand defunct municipal websites. Without the crucial metadata—the ‘why’ and the ‘for whom’—this data becomes noise, a burden of proof with nothing left to prove.

The Weight of Everything

More dangerously, saving everything can be an act of historical whitewashing. It pretends that the digital record is a neutral, objective artifact. It is not. What gets crawled, what gets saved, and what gets funding is a product of human bias, technical limitations, and institutional priorities. By claiming comprehensiveness, we risk presenting a skewed reality as a complete one. We preserve the shouts from the town square but lose the hushed conversations in the alleyways, all while asserting we have the whole story.

This is where deliberate amnesia becomes a radical tool for honesty. It forces us to make conscious, documented decisions about what truly holds value. It is the difference between a hoarder’s basement and a curator’s gallery. The act of discarding is an act of interpretation, a statement of value that is itself a critical piece of metadata for future scholars. It says, “We saved this because…” and, just as importantly, “We did not save that because…”. This transparency about our selection criteria is far more valuable than a silent, indiscriminate hoard.

This isn’t a call for destruction but for conscious curation. We must shift our focus from building The Unforgetting Machine to crafting a critical ledger of remembrance. Our goal should not be to save all data, but to save the right data with the right context, and to be brutally honest about the gaps we leave behind. Sometimes, the most profound act of preservation is to carefully, thoughtfully, let go.

Notes & further reading

A few pages I came back to while writing this: