A support ticket that reads “the app is broken” is a request for help. A ticket that reads “the checkout service returns a 500 only on Tuesdays, only when the cart payload exceeds two megabytes, and only after the nightly cron job has run” is something else. The second is not a better-worded version of the first. It is most of a diagnosis. By the time you have stated the failure precisely enough to file it, there is almost nothing left to do — the bug, cornered by its own description, has nowhere left to hide.
That is the truth inside a tempting aphorism — that a question asked with enough precision answers itself. The aphorism is worth keeping, but not in that form, because in that form it is either empty or false. The work of rescuing it is where it gets interesting, because what survives turns out to be a claim not about precision but about what questions are.
Take “enough” at its word and the trap springs immediately. If “enough precision” just means the precision at which the answer becomes evident, the claim folds into an identity: a question precise enough to answer itself answers itself. It says nothing. And if “enough” is supposed to mean anything stronger — if precision is meant to deliver answers — the claim is plainly false. Fermat’s Last Theorem and the Riemann Hypothesis are stated with total precision. That precision did not dissolve them; it manufactured them. A perfectly sharp question can name a problem that stays open for three centuries. So the aphorism cannot be a law. Whatever is true in it is true of some questions and not others, and the real content is in the boundary between them.
The confusion lives in a single word doing two jobs. There is the precision needed to state a question well — clear terms, fixed constraints, no ambiguity about what would count as an answer. And there is the precision needed to resolve it. These are different thresholds, and the surprising fact is not that one implies the other — it doesn’t — but that for a large and practically important class of questions, they very nearly coincide. The clean bug report is one such case. A well-posed lemma, where the proof feels almost clerical once the statement is right, is another. A contract whose every term has been pinned down until no dispute remains to arbitrate is a third. In all of them, the labor of formulating and the labor of solving are not two tasks but one. The point is exact: a question answers itself not when it is merely precise, but when the work of making it precise is the same work as solving it.
Notice what this does to the ordinary picture of inquiry. We tend to imagine a question as an empty container and an answer as the content poured into it — the question points, the answer is fetched from somewhere out there. But a question precise enough to belong to this class is not empty at all. Its terms fix the admissible moves. Its constraints carve out the space of things that could possibly satisfy it. Its precision sets the bar for what would count as done. The question is a structure, and the answer is the shape that structure leaves for whatever fits. On this view, “answering” is less retrieval than reading — you read off what the well-built question has already determined. The content was in the question; the answer is its shadow.
Why should the two thresholds coincide at all? The most plausible reason is that the operations which make a question precise are often the same operations that solve it, wearing different clothes. To state a problem exactly, you must fix your definitions, surface the buried assumptions, and choose a representation. Those are not preliminaries to the solution. In many problems they are the solution, or all of it that was ever hard. Deciding to describe the crash by day, payload size, and preceding job is already the inferential act that isolates the cause; naming the cause afterward is entailed. This is why precision can feel as if it works for free. It doesn’t — but the work it does and the work of solving are the same work, which is why doing the one looks like getting the other thrown in.
A sharp reader will object that this is an elaborate renaming. A question is still a request for content; a precise question is just a request narrowed until little content would satisfy it. The vessel didn’t vanish — it shrank. Call it a structure if you like; nothing has been inverted.
That objection is right about more than it first appears, and the place where it’s right marks the boundary of the whole idea. Wherever a gap remains between full specification and answer, the container model is the correct one. The answer genuinely exists to be found, and the question merely points across the gap. This is exactly the predicament of the famous open problems. Fermat and Riemann fix their answer-spaces completely, but the rules of the system let you reach those answers only along a chain of permitted moves that may run arbitrarily long. Precision defines the destination; it does not pave the road. The inversion holds only where the road has no length — where, the instant the destination is specified, you are already standing on it. So the objection wins its case for one entire class of questions and loses it for another. The aphorism is not a universal truth about questions. It is the correct description of inquiry in the narrow-gap regime and the wrong description everywhere else, and the skill it quietly demands is telling the two regimes apart.
That distinction is not academic, because the comfortable version of the idea is most dangerous to the people it serves least. To someone whose binding constraint is the clarity of their own thinking — the engineer at a muddle, the analyst with tangled terms, the writer who cannot yet say what the piece is about — “ask it precisely” is the best advice there is, because their gap really is one of formulation and closing it really does close the problem. To someone whose question is already perfectly clear and whose obstacle lies elsewhere — who lacks the data, the access, or a proof technique no one has invented yet — the same advice is false and faintly insulting. Their precise question does not answer itself, and being told that it should is being told their wall is a failure of articulation. The aphorism is true exactly where the work is thinking. It has nothing to offer where the work is something else.
So keep the idea, but turn it the right way up. A precise question does not find its answer; in the cases that prove the rule, it casts it, and the answer is the shadow of a question built well enough to throw one. The discipline worth having is not learning to ask everything precisely. It is knowing which questions are the kind that can answer themselves — and refusing to tell anyone stranded at a real wall that they simply haven’t asked well enough.
cafebedouin.org 