Let's say you ask a question that asks about the existence of some type of object, like a result, a document or a historical event; such as:
- What group theoretic results were known for several special cases before the general definition of a group was established?
- Has any broad scientific consensus ever been based on corruption rather than evidence?
- Was Occam’s razor ever wrong?
Whenever you ask this kind of question, there's always the possibility that the answer is going to be "what you're asking for doesn't exist". Unfortunately, if an answerer says this, it's hard to tell how credible the response is. This is a problem for both the answerer and the asker:
- The asker is hesitant to accept the answer in case it's wrong, and they don't want to reduce the chances somebody who might actually have an example from coming forward and answering the question (e.g. accepting the answer would take it off the Unanswered list).
- The answerer is hesitant to answer in case they're wrong, and there is actually an example that they're not aware of.
Do we need a policy to handle these cases? I can see this happening a lot. The only really satisfying way to answer such a question would be to do exhaustive research, and write an answer detailing all the places you looked, and arguing why if you didn't find it there, it probably doesn't exist. But we can't be expecting everyone to put that much time and work into every single answer or we'll never answer anything.