The Ghost in the Spreadsheet: A Single Typo and the Weight of Public Data
I remember the exact moment I realized open data wasn’t just about ones and zeroes, but about the stories they accidentally tell. It was a Tuesday afternoon, the kind where the light from the window hits your monitor just wrong, and I was deep in a CSV file. This particular dataset was a goldmine: decades of municipal park maintenance records, freshly released by our city’s new open data portal. My goal was simple—map the history of a specific oak tree’s care in my local park.
Scrolling through thousands of rows, the entries blurred into a rhythmic monotony of dates, work order numbers, and arborist codes. Then, my finger paused on the trackpad. There, amidst the clinical precision, was an entry that didn’t fit. The date was correct, the work order number sequential, but in the ‘Tree Species’ column, someone had typed ‘OAK,’ and then, as if in a moment of pure distraction, immediately after it: ‘IS TALL.’
The full cell read: ‘OAK IS TALL.’
I laughed out loud. Some city worker, perhaps on a mobile device in the dappled light under that very tree, had finished their official entry and then, almost poetically, appended their own casual observation. It was a tiny, human fingerprint on the cold glass of bureaucracy. For a second, I could almost see them there, looking up, noting the fact before moving on to the next task. That typo was a fossilized sigh, a moment of personal reflection caught in a system designed solely for objective fact.
But my amusement quickly curdled into a profound unease. This data wasn’t just a historical record; it was live. It fed into budget allocation models, informed future planting initiatives, and was cited in reports. What if an algorithm somewhere, devoid of context, began categorizing a new species called ‘OAK IS TALL’? What if a researcher years from now took that entry as gospel, a bizarre arborical term lost to time? This tiny, human error had escaped the sanitization process and was now living, rent-free, in the official record.
We talk about digital preservation in grand terms—migrating formats, fighting bit rot, ensuring accessibility for centuries. We rarely talk about preserving the humanity encoded within the data, the mistakes and asides that reveal its creation. That typo is more than an error; it’s a testament. It’s proof that a person, not just a system, was involved. In our rush to clean, standardize, and make data ‘usable,’ we often scrub away these traces. We create a perfect, sterile, and ultimately false history.
That ‘OAK IS TALL’ is my ghost in the machine. It’s a reminder that public data is never purely objective. It’s collected by people, with all our flaws, distractions, and momentary lapses into poetry. Preserving data isn’t just about ensuring the files open; it’s about preserving the truth of its origin, quirks and all. Because sometimes, the most important part of the record isn’t the data itself, but the utterly human story of how it came to be.
Notes & further reading
A few pages I came back to while writing this: