The Tyranny of the Timestamp: How Precise Dating Undermines Digital History
We have been taught to worship at the altar of the timestamp. In the realms of open data and digital preservation, the precise, machine-readable date is considered an unassailable good. It is the bedrock of version control, the anchor of chronology, the first principle of organizing any archive. To question its supremacy is heresy. And yet, I find myself heretical. Our slavish devotion to the exact moment—down to the second and millisecond—is, in many cases, actively sandblasting the nuance out of our digital history.
Consider a typical dataset from a city’s open data portal: a CSV file listing building permits. Each row is stamped with the exact second it was entered into the municipal database. This seems like a victory for precision. But what does that timestamp actually capture? It logs the moment a weary clerk finally keyed in the application, perhaps weeks after it was filed in a physical tray. It does not capture the date of the architect’s first sketch, the community board’s contentious meeting, or the first day of construction. The timestamp gives us a false sense of certainty, a deceptive pinpoint on a timeline that is, in reality, a foggy expanse of process and bureaucracy.
This obsession with pinpoint accuracy creates a distorted historical record. Future historians, armed with perfectly sorted datasets, will see a flurry of permit activity at 4:59 PM on a Friday—not because of a sudden city-wide urge to build, but because a departmental deadline forced a batch update. The messy, human reality of how things actually get done is erased by the clean, inhuman logic of the database entry. We are preserving the artifact of the system’s operation, not the story of the event itself.
This problem scales up to web archiving. A crawler captures a page at a specific millisecond. But a webpage is not a photograph; it is a living entity. A news article might be updated ten times in an hour with corrections, amplifications, and hot takes. Which timestamp represents the "true" article? The one that captured a typo-ridden first draft, or the one that captured the polished, but potentially less passionate, final version? By archiving each iteration with a hyper-precise marker, we fracture a fluid narrative into a thousand disjointed shards, making it incredibly difficult to reassemble the actual lived experience of reading the news as it unfolded that day.
Perhaps we need to embrace a fuzzier, more human timescale in our preservation efforts. Instead of, or in addition to, the ISO 8601 timestamp, we could benefit from descriptive metadata: "Early Draft," "Pre-Public Meeting," "Post-Implementation," or simply "Circa Summer 2023." This wouldn’t replace precision where it is truly needed, but it would supplement it with context. It is an argument for valuing the narrative truth of an event over the administrative truth of its record. By fetishizing the exact moment of digital birth, we risk preserving a beautifully precise, perfectly irrelevant history.
Notes & further reading
A few pages I came back to while writing this: