Or rather, questions where the answer is, "there is no answer".

Inspired by the comments of this recent question. Some members of the community wants to close the question, others think it is better to give an answer stating it is unknown. I can see merits with both sides, although that question has now been closed.

Do we have a consensus about what to do with these kind of questions? It seems to me we do get these kind of questions fairly regularly.

Also, if the community consensus is for closing, then could we get a new close reason? The example was closed for "too broad" which isn't a very helpful message. Something like, for example, "This question cannot be answered by the tools available to history" might be more informative and incorporates the answer "it isn't known" into the closing message.

  • 4
    Excellent and timely meta contribution. – Samuel Russell Oct 16 '14 at 20:54

I'm personally on the side of answering that there is (or perhaps can be) no known answer, with an explanation of why. The theory here is that if its otherwise a legit question, someone else is liable to come here wondering about it too. It would be nice to have an answer for them.

Not that I'm casting aspersions on the closers of the linked question. I can see an argument in this case for at least needing to define "kill" better (eg: personally, or by units under his direct orders, or by people nominally under his leadership, or indirectly through starvation or disease, or by other peoples fighting for land after being displaced by Huns, or...)

  • 1
    We should also add the (historiography) tag to questions where the answer is "unanswerable". – Samuel Russell Oct 16 '14 at 20:52
  • 2
    We should also close answering by newer stackexchangers to avoid "me too"s. (listed separately as it is a separate concept) – Samuel Russell Oct 16 '14 at 20:52
  • This seems to be at the moment the site consensus. Yet the OP's question that raised the issue (Atilla the Hun) remains closed :( – DVK Oct 25 '14 at 17:31
  • @DVK I've voted to undelete the question that the community bot auto-deleted; in line with the newly presumed consensus. – LateralFractal Oct 30 '14 at 10:43

I think at the heart of this debate is whether questions on potentially unanswerable topics add value to the History Stack Exchange.

On the one hand the question itself might be historically incoherent to those in the know.

For instance: "Which German composer invented the piano?"

Such questions can only be answered by critiquing the question directly. An exhausting and sometimes fruitless exercise as you may not be able to clean them up even with the OP's input; as you end up semantically replacing both the handle and the head of the axe - so is it even the same axe?

On the other hand the question might coherent but unanswerable with the knowledge and resources at the disposal of the current stack exchange membership. Perhaps - in so far as we possess the arrogance to speak for all historians - not answerable by any historian with current sources, tools and absence of a time machine.

For instance: "How much of the annual budget of Pharaoh Khufu's regime was allocated to tomb building?"

We can answer these as unanswerable or close them as unanswerable; but we run up against problems either way1. If we answer them as unanswerable, then the first person to post gets votes for proving a negative; hardly fair, and hardly pleasing to most OPs and people accustomed to negation being in comments only. If we close them as unanswerable; we don't allow for someone knowing the answer, partial answers, duplicate unanswerables with the better content2, or incentivising people to explain authoritatively why it's unanswerable3.

Then you have what I call the "zero or many" question:answer paradox. These are questions where a quantifiable answer exists but it's either zero or many answers.

For instance: "Was there a king that had seven legitimate sons?"

It's a paradox in so far that instead providing 0 (no answer - implicit), 1 (no answer - explicit) or N answers (subset of known or knowable answers), the question is closed as too broad or opinion based because it didn't resolve to precisely one answer. When in practice, if a question has multiple valid answers then the question should have multiple answers4; not none to frustration of the OP.

So. To return to the basic criteria of: Does it add value?

  • If the question isn't historically coherent and you can't fix it: Close
  • If question is unanswerable and the OP wants the question addressed as-is5 and this is the first time we've seen this question: Answer why the question is unanswerable at the current time. Should include the Why so as not to cover for personal ignorance.
  • If the question is unanswerable and we've see it before, more or less as worded: Close with appropriate duplicate link
  • If the question, if answerable, allows for multiple equally valid answers: Allow any answer and applicable voting on quality of said answers until the high threshold is met for putting the question on hold / freezing out extra answers.

1. Courtesy of Stack Exchange's origin as a computing programming Q&A with clear falsifiability.
2. Unless you didn't close the first instance of an unanswerable question.
3. Comments don't accrue reputation; can not be cleaned or refined after 5 minutes; and not guaranteed to be persistent in the underlying database.
4. Rare cases of a super-abundance of answers can be treated the same as other stacks do through use of the On Hold action for curtailing Me Too! answers.
5. Occasionally the OP meant to ask a question that is answerable, but more often the OP really does want some reply to the question as asked. For knowing what we don't know is vital in any branch of human knowledge.


I make the argument for closing such questions unanswered:

  • an answer is not possible within the boundaries of history
  • we are meant to curate the site for the boundaries of history
  • (feel free to expand and improve this position by editing) – Samuel Russell Oct 16 '14 at 20:54

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .