Another note: (Since the first one wasn’t heeded...) OK, so the question was closed again. My appeal below was to both sides of the issue, so even though five people sought fit to do otherwise, I’d ask those who think the question should be open not to reopen it before this discussion has proceeded. This shouldn’t be a war of votes but an exchange of opinions and arguments.

Note: While it was certainly one possible outcome of this discussion that the question would eventually be reopened, I was hoping that this would take place after a thorough discussion and hopefully some form of consensus on the criteria. I see now that (likely in connection with this post) the question has already been reopened. That was not my immediate objective (otherwise I would have voted to reopen it instead of posting here). I hope that we can keep this discussion here focussed on the general questions that this case brought up and not let it turn into an edit war about whether this particular question stays open or closed. I see there are already two new close votes. I think it would be most conducive to a principled discussion if everyone on either side of the issue would refrain from voting in either direction until the required discussion has taken place.

Proof of some Riemann identity was closed as “off-topic”: “This question is not about mathematics, within the scope defined in the help center”. As is unfortunately often the case, the notice is wrong and unhelpful, since that help center page gives no indication at all why this question would be considered off-topic.

I assume that the basis for the closure is Is it ok to ask about the correctness of preprints of crank friendly topics?, and in particular the by far most upvoted answer to that question. (I couldn’t find any other basis; if there is one, please point it out.)

I don’t want to go into the question whether that’s an accepted policy (it says “Proposed policy” at the top, and not everything with lots of upvotes is a policy), because I mostly agree with it. Rather, I’d like to know whether people think that it applies and/or should apply to this particular sort of question. I don’t find it very clear or helpful in that regard. It gives several examples, but most of these are more or less obvious; for instance, “have you read X” is obviously not a question on mathematics.

The present question summarizes the argument of the preprint and points out specific ideas that seem interesting: “What I like about it is his analytic approach using the behaviour of general series at singular points. [...] In particular I want to point out his main result [...] What do you think?” That seems to be exactly what the proposed policy says how questions about such preprints should be asked: “In summary, focus your question on specific technical aspects of the work that you'd like clarification on. ”

At the time of closure, the question had $8$ upvotes and $2$ downvotes, so a majority of voters (even when including the close voters) thought that it was a good question. It doesn’t fall under any of the “BAD” examples given in the proposed policy:

  • Despite the introductory sentence “I was wondering if anyone is aware of this proof”, it doesn’t ask “have you read it”.

  • It doesn’t ask “is the proof correct” – to the contrary, it says the OP couldn’t find a flaw but presumes there is one.

  • It doesn’t say “I don't understand the paper and don't want to spend time reading it, can someone verify that the author’s claim is correct?” – quite the opposite, it shows that the OP read the paper and in fact summarizes it for us so that we can comment on the approach without having to read it ourselves.

So my questions are:

  • Does this question fall under this proposed policy? (My view would be that it doesn’t.)
  • Should this question fall under this proposed policy? (My view would be that it shouldn’t.)
  • If so, why?
  • If so, should the proposed policy be updated to provide a clearer basis for deciding which posts about preprints are off-topic?

(Obviously, if the answer to the second question is negative, the question should be reopened.)

  • 3
    $\begingroup$ It's true that the author of the question made some effort, but at the end of the day it still seems like a basic "correctness" question, which is "bad" according to the policy you linked to. I don't get your point with 'It doesn’t ask “is the proof correct” – to the contrary, it says the OP couldn’t find a flaw but presumes there is one.' On the substance that's the same. Granted, it does show some maturity on the side of the person that asked, but I don't think that alone is enough. $\endgroup$ – quid Apr 5 '20 at 14:06
  • 3
    $\begingroup$ @Diger as far as I am concerned there is no problem in the formulation, it is a problem with the type of question itself. The point of my comment was to understand what joriki wants to express. Ultimately I think "Is this correct?" is the same as "Is there an error?" $\endgroup$ – quid Apr 5 '20 at 14:34
  • 4
    $\begingroup$ @Diger there we isolated the problem. "I just want to discuss this work" is not an appropriate type of question. This site is not for discussing preprints. One can ask to the point questions about preprints. General calls for discussions are off-topic. $\endgroup$ – quid Apr 5 '20 at 14:44
  • 4
    $\begingroup$ @Diger: This is not a discussion board, it's a question-and-answer site. $\endgroup$ – joriki Apr 5 '20 at 14:47
  • 4
    $\begingroup$ IMHO, "Do you find this approach promising" is an opinion-based question, and both questions you ask about, @joriki, puts math.se in the role of evaluating the merits of a pre-print, which I suspect is correctly off-topic. However, I suspect that if there is one step, or two, that an OP doesn't understand, it would be appropriate to ask: I don't understand how the author went from foo to fie on page 2, tag (3)...". $\endgroup$ – amWhy Apr 5 '20 at 14:57
  • 8
    $\begingroup$ The sort of questions related to preprints that are suitable for this site are precise questions about specific steps (possibly about the correctness of the step). The question "what do you think?" is basically never a suitable question. Questions about a preprint in the GM category that has not been published after this long are unlikely to lead anywhere interesting. $\endgroup$ – Tobias Kildetoft Apr 5 '20 at 15:03
  • 4
    $\begingroup$ @TobiasKildetoft: If "unlikely to lead anywhere interesting" were a criterion, many if not most of the questions on this site would have to be closed as off-topic. It's not the purpose of this site to advance research. The bulk of the questions are about well-known things that are certain not to lead to anything interesting. People should be allowed to ask about things, and if someone is not interested, they don't have to answer. The discussion should focus on the purpose of the proposed policy, which is not to ensure that questions on this site lead to something interesting. $\endgroup$ – joriki Apr 5 '20 at 15:16
  • 5
    $\begingroup$ @joriki Something like "In the proof of Theorem 2.3 the author A invokes Proposition 5.2 from Published Paper P by B. However, in PP by B the at the start of section 5 the assumption is made that all operator are self-adjoint. However, in the proof of Theorem 2.3 it is not mentioned that or why operator D to which this result is applied is self-adjoint. Is it true that D is self-adjoint (and if so why?) or is there a problem in the proof?" $\endgroup$ – quid Apr 5 '20 at 15:22
  • 6
    $\begingroup$ @Diger: I think the suggested criterion here is not whether such things would come up in the discussion you're seeking, but whether your original question was this specific. $\endgroup$ – joriki Apr 5 '20 at 15:29
  • 6
    $\begingroup$ @Diger I gave an example of the type of question I would consider as legitimate. Indeed potential answers include: "The operator D is in general not self-adjoint (consider for example this case). The proof seems incomplete." "The operator D is self-adjoint. Here is why..." "The operator D is not self-adjoint. But in fact Prop 5.2 holds in general. See Remark 5.7 in PP." Or whatever might be the case. The point is that the question I sketch is clearly and narrowly scoped. (Of course the question could still be poor in other regards, also as actual Q it should contain more context.) $\endgroup$ – quid Apr 5 '20 at 15:33
  • 4
    $\begingroup$ @Diger it is off-topic. You said you want to discuss the paper. This is exactly what the site is not intended for: it is not for discussion. This is one of the original cornerstones of SE: not for discussion. Sorry, for the repetition. But if you want to make the point that we should allow discussion that's a non-starter. (This does not mean it is without merit in principle, but it's a non-starter on the main site; create a chat room if you want.) $\endgroup$ – quid Apr 5 '20 at 16:14
  • 6
    $\begingroup$ @amWhy: I was sort of waiting for you to say that :-) Please note that $5$ people voted to reopen the post. Not everyone enjoys having discussions in this style, but their votes count nonetheless. By the way, two people not agreeing with the consensus does mean that there's no consensus, but I don't have much hope that we'll reach a consensus in this way anyway. (And as above I mean both "sides" by that.) $\endgroup$ – joriki Apr 5 '20 at 20:45
  • 6
    $\begingroup$ @amWhy: I'm not sure what "still" refers to. Your argument was that there was a consensus. It was already flawed if there had been only two against it. Clearly there's no consensus when there's $\frac13$ against it. So there's no "still". $\endgroup$ – joriki Apr 5 '20 at 20:49
  • 6
    $\begingroup$ @amWhy: It's good to know that you don't take my concerns seriously. I'll take that into account. I take your concerns seriously despite everything I dislike about your style. (For the benefit of readers who don't know the circumstances: I gave an embarrassingly flawed answer to that question and soon deleted it when the error was pointed out. So I have no reputation to gain from the post being reopened.) $\endgroup$ – joriki Apr 5 '20 at 20:55
  • 5
    $\begingroup$ I left this comment on the question: On Mathoverflow, there is a long-established policy of not vetting papers, and this question would, accordingly, be closed. On math.stackexchange, I don't think there is any such policy, and the question makes a welcome change from the two thousand variations on "how do you differentiate $x^2$?" There is real mathematics in this question (even though the paper may well be worthless) and I'd like to see someone take a crack at it. $\endgroup$ – Gerry Myerson Apr 5 '20 at 23:10

You must log in to answer this question.

Browse other questions tagged .