2
$\begingroup$

I don't have much experience with "history of math" questions or the ho.history-overview tag. I tried to distill an answer to my question from looking at existing questions, but failed to find the pattern. Here are a few partial answers:

  1. If the question is about the history of a mathematical definition or concept, with a concrete answer, like "who was the first to do X?" then it seems to be non-CW. Example. Example2
  2. If the question is going to elicit a big list, or have answers that are primarily opinions (so that votes represent agreement/disagreement with the opinions being given) then it seems it should be CW. Example. But even this is not a hard-and-fast rule. Here's an example where the OP writes "I would like to see a longer such list!" suggesting that he wants a "big list" but the question is still not CW.

Then there are cases where I see conflicting evidence and so I don't know the standard.

  1. If the question is trying to figure out the identity of a name from the research literature ("who is Y?") then it seems the community lacks a consensus. Example that's not CW vs Example that is CW

  2. If the question is about math-adjacent history but not about math research/concepts/definitions etc, then it seems to stay open if and only if the community finds it interesting. I can't find any consensus about whether or not it should be CW. Example where it's not CW. Another example not CW. Another example not CW (but barely survived closing). An example where it was made CW by moderators. And lastly, an example where it was closed.

Hence my questions:

(1) When, precisely, should a history question be closed?

In the examples above, I see an increasing trend towards trying to send the question to the History of Science and Mathematics Stackexchange. But, I can't find a thread on meta where the consensus of the community (by answers or by voting patterns) suggested that we want to send all history questions away.

(2) When should a history question be CW?

It's interesting to note that a question can be closed and still not CW, so still generate reputation gain/loss for the OP and answers. So, I think it's worth it to answer both questions.

$\endgroup$
5
  • 8
    $\begingroup$ MathOverflow is not an axiomatic system. $\endgroup$
    – Gerry Myerson
    Jul 5, 2022 at 6:52
  • 1
    $\begingroup$ @GerryMyerson, perhaps MO is an inconsistent axiomatic system? $\endgroup$
    – LSpice
    Jul 13, 2022 at 18:02
  • 3
    $\begingroup$ I agree with Gerry Myerson. There's no way to lay down the law in advance to cover every conceivable situation. And even if one tried, not enough people would see The Law for it to have a credible claim to represent the community consensus. This is inherently a case-by-case type of question. We have general CW guidelines; just try to follow them, and if a specific case seems off, then cross that bridge when you get to it. $\endgroup$
    – Timothy Chow
    Jul 14, 2022 at 13:42
  • 1
    $\begingroup$ What does CW mean? $\endgroup$
    – Mathemphetamine
    Aug 11, 2022 at 0:59
  • 1
    $\begingroup$ The "precisely" in this question makes it not worth answering -- since MO policies are not and should not be algorithms, there is no precise answer. I recommend deleting that word. $\endgroup$
    – user44143
    Sep 1, 2022 at 1:29

2 Answers 2

Reset to default
3
$\begingroup$

Unlike the three meta threads about whether history questions should be closed, I have not found a similar discussion about CW status. The following is a combination of what I distilled from the examples in my original post, and my own opinion.

(2) A history question should be CW if either:

  • It's asking for a big list
  • It's likely to generate answers that are primarily opinions, so that up/down votes are meant to express agreement with the opinions and not an assessment of their correctness
  • It does not have a concrete answer, so answers will be speculative

A history question should be non-CW if none of the above apply, and:

  • It has a concrete answer and a way to verify if any given answer is correct or not.

For example, questions where the answer is "the list of members of the 1962 Fields committee" or "a quote showing that concept X was first introduced in paper Y" fall under the last bullet point.

$\endgroup$
4
  • 1
    $\begingroup$ Nothing in this criterion is history-specific, though. It seems just the regular criteria for CW. $\endgroup$
    – Federico Poloni
    Jul 7, 2022 at 13:51
  • 1
    $\begingroup$ @FedericoPoloni If only it were so simple. But the conflicting decisions I linked to in my question suggest some disagreement in the community. Several that "should be CW" according to this answer are not, and some that "should be non-CW" are made CW by moderators. I find it strange. $\endgroup$
    – David White
    Jul 7, 2022 at 18:19
  • 1
    $\begingroup$ Although hopefully "the list of members of the 1962 Fields committee" would be closed; to the doubtful extent that that is appropriate for SO at all, it surely belongs on HSMSE. $\endgroup$
    – LSpice
    Jul 13, 2022 at 18:05
  • $\begingroup$ I would remove verifiability — see Jim Humphrey’s comment (mathoverflow.net/questions/54513/…): Anecdotes about mathematicians are fun but often impossible to check decades later, unlike proofs. $\endgroup$
    – user44143
    Sep 25, 2022 at 6:24
0
$\begingroup$

I want to give my opinion on the two questions, so people can express agreement or disagreement via up/down votes. Since there are two questions, I'll write two answers, to make it easier to disentangle what people agree/disagree with. If you disagree, it would be good to write a comment as well, since a vote alone doesn't tell precisely what you disagree with, and the whole point of this was to stimulate a needed discussion.

We have had several previous meta discussions about when history questions should be closed or migrated. Discussion1, Discussion2, Discussion3 (in reverse chronological order). I read these and here's what I distilled.

(1) A history question should remain open and remain on MathOverflow if any of the following apply:

  • it's about research into the history of mathematics
  • it's about the history of a mathematical concept
  • the "key points are largely mathematical in nature"
  • the question appears to be of interest to the MathOverflow community

This is not a complete list, but it's the best I could distill. If you think other items should be on the list, please leave a comment.

$\endgroup$
1
  • $\begingroup$ I’d restrict the first three items to topics for MathOverflow rather than MathStackExchange. For historical questions, I find that often coincides with topics that require graduate mathematical education, and topics from after the mid-1800s. $\endgroup$
    – user44143
    Sep 25, 2022 at 6:28

You must log in to answer this question.

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