A question whose answer doesn't exist yet can't be collected, only framed

Some questions have answers that exist right now, somewhere, hidden. How many missiles the Soviets had was that kind of question: the number was real, written down, guarded. Steal enough of it and the argument ends. Spending billions on satellites against a question like that is rational, because the answer is a thing you can go get.

Other questions have no answer anywhere. Whether the Soviet Union would collapse wasn't a secret Gorbachev was keeping; he didn't know either. The answer was contingent, waiting on interactions that hadn't happened yet. Gregory Treverton called the first kind puzzles and the second mysteries, and his point was that they differ in what kind of work can touch them, not in difficulty. Collection solves puzzles. A mystery can only be framed: name the driving factors, watch how they've interacted before, hold the estimate loosely.

The nasty part is the asymmetry. More information reliably helps a puzzle and routinely worsens a mystery, burying the frame under [competing true signals](Intelligence noise is made of competing true signals, not random static). Iraq's WMD got worked as a puzzle, find the stockpiles, when the load-bearing question was a mystery: did Saddam fear Washington more than he feared his own neighbors?

Source claim: A question whose answer does not yet exist anywhere cannot be resolved by gathering more information, only framed by judgment about how its driving factors interact.