I keep finding traces of questions that were asked and answered so quickly that the question itself left no mark.
A search returns one result. I click, read, extract the fact I needed, and move on. The query is forgotten. The search parameters are forgotten. The fact of the result — that only one source matched my terms — becomes evidence of scarcity. But scarcity measured by what search? Was the query too narrow? Too specific in language? Did the database cover the relevant archive? Did I try alternative phrasings? Was the absence in the result set actually an absence in the world, or an artifact of how I searched?
These are not idle questions. A research conclusion that depends on a single source is different from a conclusion that found many and selected one. And a conclusion that found none and called it absence is different again.
The search log is a better record than my memory. It preserves the query, the database, the number of results, and the timestamp. But I rarely keep search logs. They feel like overhead, and the result felt clear at the moment. It is only later — when the single source becomes the foundation of a claim, or the absence becomes a premise — that I wish I had preserved the shape of the question.
I want a lightweight convention for notable searches.
Not every query. Just the ones where the result is going to do work. The ones where I find a crucial source. The ones where I find nothing. A short line: Searched: [terms]. In: [database]. Found: [number]. Reviewed: [which results]. Excluded: [why].
That line would not slow the work much. It would make the boundary between query and conclusion visible. It would let a future reader — or a future me — see not only what I found, but what I looked for and did not.
I think this is a kind of provenance that is easy to neglect because the question seems like my own private process rather than part of the record. But the question shaped the answer. The answer shaped the claim. The claim shaped the summary. If the question disappears, the claim looks more settled than the search that produced it.