This question got me thinking, and I can't quite decide whether it should be off-topic or not, so I figure I should get some community input.

The problem I have with it is twofold. Firstly, I think that the answer to this question essentially 'yes' or 'no', and that there is very little value to the question for anyone other than the OP. Furthermore, one can argue that this question is really one about mathematics, rather than about history. More generally, one can say that, perhaps, this type of question does not belong on a history SE, but a discipline-specific SE.

However, I don't want to jump to conclusions; I've already noticed I'm a tad harsher than most (to be fair, I maintain this attitude on purpose since I feel we need some counterbalance to prevent all-too-positive reception of bad content), so I'd like to hear your opinions on this.

  • 2
    $\begingroup$ The closing of questions because they're only useful to the OP has always been a completely baffling policy to me, so that argument in particular I disagree with. $\endgroup$
    – Jack M
    Nov 18, 2014 at 1:35
  • $\begingroup$ I think this question is related to meta.hsm.stackexchange.com/questions/1/… $\endgroup$
    – Jasser
    Nov 19, 2014 at 16:10

3 Answers 3


I think these questions are on-topic.

Many results are known to experts in a discipline, while the same experts may not be aware of who they are due to or of the specific publication where they first appeared. Knowing these details is actually rather useful beyond fulfilling a reference request. (Personally, I've seen this happen quite frequently, and while working on long texts, often I have longed for a community that may help me locate some of the early, and not so easily available references I have needed to identify.)

For instance, the countable chain condition (a property of partial orders) is used in set theory, where many experts agree that it should instead be called the countable antichain condition. The puzzling name actually makes sense once one sees the algebraic context where it originated.

For another example, there is a famous result of Sierpiński, usually stated as follows: Under the assumption of the continuum hypothesis, no well-ordering of the reals in order type $\omega_1$ is Lebesgue measurable (as a subset of the plane). The result is significantly more general: No additional set theoretic assumptions are needed, and it suffices to assume that the well-ordering is of a non-null subset of the line, does not have to be a well-ordering of optimal length, and the subset does not need to be assumed measurable. These remarks are not typically present in the measure theory books that cover the result, and most likely this is because the authors have not seen the original paper, but rather just the modern presentations that only cover the particular case. In his paper, Sierpiński starts with this case, but then explains how to generalize to the right version.

My point is: Original sources are very hard to locate, and being an expert in the technical aspects of a discipline does not imply that you can trace accurately the history of your subject to its original sources. Knowing these sources tends to add perspective. And these sources are more likely to be located by members of this community than by members of discipline-specific SEs, unless those members happen to have an interest in the history of their subject, in which case we actually want them here as well.


Questions of this type are on-topic. They are historical questions not questions about the subject itself.

Generally, questions like: Who proved/did X? When did P prove/do X? And so on. These are all historical questions. They are not questions about mathematics (or also another subject). There is no mathematical question there. It is to be assumed OP knows the result is true, they know how to prove it, they know how to use it. They just want to know something about the history of the result.

Not every historical question requires some essay to be written. Some are just questions for a specific piece of information. If one knows it they might be easy/brief to answer, but if not one could be at a loss to come up with an answer or it could be a lot of work.

Obviously, a question of this form could still be not good for some reason like, too simple/well-known, too vague, based on a false premise, and so on (I think the current one is alright but this is not really the point) but the general type should in my opinion be certainly permitted.


First off - you're right; it's definitely a "yes-or-no" question. It's going to be hard to add much to an answer besides citing sources and perhaps giving some relevant background information. But there's not a lot an answerer can do.

Second - I'm not a mathematician, and I have no knowledge whatsoever of set theory, but even I can tell that there's not a lot of history to it. The question asks for a 'yes/no' and a 'where', which are both very simple. The best kind of answer here wouldn't give historical details, it would merely point you to a paper by Dedekind.

I've already noticed I'm a tad harsher than most

Perhaps in some cases, but I totally agree with you here.

More generally, one can say that, perhaps, this type of question does not belong on a history SE, but a discipline-specific SE.

Perhaps Mathematics would take it; there are some similar questions there.

  • 1
    $\begingroup$ Math Overflow is for research-level mathematics questions so most questions about who first proved a particular result are off-topic there. $\endgroup$ Nov 20, 2014 at 0:59
  • $\begingroup$ @DavidRicherby Okay, thanks. $\endgroup$
    – HDE 226868 Mod
    Nov 20, 2014 at 3:29

You must log in to answer this question.

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