I have the impression, that there is no concensus among MO moderators about sharing knowledge on MO; the impression is based on quite contradictory reactions to offering to share such a knowledge, ranging from encouraging to share it, despite some skepticism about the correctness or practical applicability to discouriging doing so, based on speculations on the mathematical value of such an answer or, on speculations about the personal motivation for sharing knowledge.

I had read the discussion related to Answering my own question and Appropriate Reaction to a Failed Reference Request on MO and thus felt encouraged to share my knowlegde, but wanted to present in a way that would be appropriate for MO (Preferred Conciseness of Contributions), which, besides one useful comment, only brought up severe objections against sharing my knowledge.

My question is therefore, whether there is consensus at least among MO moderators about what is appropriate and in what form, when "pressing the button" for answering own questions or sharing knowledge?
What should happen with speculative comments and answers that aim at discouraging answering one's own question or sharing knowledge?

  • 4
    $\begingroup$ MO philospohy is a bit different from general SE philospohy regarding this matter. Various things you might find encouraging posting q+a and alike are not due to MO moderators but SE moderators/staff. I suspect the lack of consensus you perceive is not so much among MO moderators but among MO and SE. $\endgroup$
    – user9072
    Dec 31, 2013 at 15:24
  • $\begingroup$ I'm not sure how you are using the term 'moderators', but if you are including experienced high-rep users like Andy Putman, then I see no contradiction whatever between our responses at your third link. I think Stefan has summed it up for you very neatly in his answer. I do see now that giving a description of your approach to the material that you seek a reference to in the literature is not actually supplying a reference to the literature, and could thus be construed as self-promotion, which is frowned upon. You could however submit a preprint to ArXiv and then refer to that in an answer. $\endgroup$
    – Todd Trimble Mod
    Dec 31, 2013 at 16:14
  • 1
    $\begingroup$ So, we have certain possibilities. You proved a new theorem, and you put it here instead of publishing. Or you put it on arXiv, but then you also put it here as publicity. Probably both of these are to be discouraged... $\endgroup$ Dec 31, 2013 at 16:14
  • 1
    $\begingroup$ I think what we have here are matters of good taste and judgment, rather than what is allowed in "the rules". I personally would not howl in protest if you put something on the ArXiv and then write a discreet answer that says something like, "if anyone is at all interested, I have written about this here [link]". But it could smack of self-promotion to others. At this point, use your judgment -- as far as "permission" goes, refer to what Stefan wrote. $\endgroup$
    – Todd Trimble Mod
    Dec 31, 2013 at 16:19

2 Answers 2


I cannot speak for the moderators, but as far as I understand, you should only ask a question on MO if you do not already know the answer.

However it may still happen that some time later (be it days, weeks, months or years -- this does not matter) you find an answer to your question yourself. If nobody else has provided a satisfactory answer to your question in the meantime, then it is considered good practice that you answer your own question, to avoid other people wasting their time on telling you what you already have found out by yourself.

Another situation where "answering" your own question is considered good practice is when other people have answered your question in the comments. In this case you provide a CW (community wiki) answer summarizing the relevant comments, in order to make it visible for everybody that your question is already answered and to get it off the unanswered questions list.

  • $\begingroup$ Thank you for the clarification, Stefan. So, in order to adhere to MO rules, I can't supply an answer to my failed reference request. Should someone be interested in my solution, I would be willing to share it outside of MO on request. $\endgroup$ Dec 31, 2013 at 14:19
  • 4
    $\begingroup$ The details matter here. If a reference request is for your entire result, then it looks too much like self promotion. If the reference request is for a lemma you wish you didn't have to prove, I think it's fine to subsequently add a link saying something like "I gave a proof of this fact as Lemma 3.17 in ..." $\endgroup$ Dec 31, 2013 at 20:32
  • 1
    $\begingroup$ I think this is a useful thread; such distinctions help clarify my own thinking on this topic, and I have to say that what I have written earlier is not necessarily my final thoughts on this. $\endgroup$
    – Todd Trimble Mod
    Dec 31, 2013 at 20:47
  • 2
    $\begingroup$ I think it is helpful to look at case studies here. Two well-received "self-answers" that involved references to the OP's papers are Misha's answer to mathoverflow.net/questions/136944/quasi-homomorphisms-of-groups and my answer to mathoverflow.net/questions/56103/…. My answer is of the kind that Scott Morrison mentioned, while Misha's answer was literally a reference to a new paper of his which answered the question. I think that there were two things about these answers that made them acceptable. First, it was clear in both cases $\endgroup$ Dec 31, 2013 at 22:29
  • 3
    $\begingroup$ It's happened to me a few times that I asked a question to which I did not know the answer, but in the hours after posting the question, it naturally being on my mind, I came to an answer myself (or a partial answer) and so posted an answer. For example: mathoverflow.net/questions/14830/which-graphs-are-cayley-graphs. $\endgroup$ Dec 31, 2013 at 22:36
  • 2
    $\begingroup$ (continued) that the OP did not know the answer when they posted the question. They really wanted to know the answer and did not intend to advertise their work. Second (and this is more controversial but I think plays a role in the community's judgement), both Misha and I are long-time posters here who answer many questions of other people. We are also active researchers and many people on MO know us in person. This leads people to cut us a bit more slack (though I think we would still be harshly judged if it looked like we were engaged in self-promotion). $\endgroup$ Dec 31, 2013 at 22:41
  • $\begingroup$ It is worth noting that JDH's example also fits both of my criteria. $\endgroup$ Dec 31, 2013 at 22:42
  • $\begingroup$ Thank you all for the valuable feedback; as I am fairly novice on MO and not associated to a research community, I still have a steep learning curve concerning the culture of research communities. $\endgroup$ Jan 1, 2014 at 12:00
  • 2
    $\begingroup$ I also had a fine reception of a self-answered question mathoverflow.net/questions/33770/… (the type of question is somewhat different, but at least this shows that you do not need to have answered a lot of questions here to get a good reception of such as this. It does still satisfy the first part, as I did not know the answer at the time of asking). $\endgroup$ Jan 1, 2014 at 19:25
  • $\begingroup$ "I cannot speak for the moderators, but as far as I understand, you should only ask a question on MO if you do not already know the answer." What's so different between finding an answer before posting the question or after posting it? $\endgroup$ Dec 1, 2015 at 21:14

I personally don't mind people posting questions and answers at the same time. I don't think there is a big difference between finding an answer just before posting the question on MO or right after posting the question on MO.

What is more important IMO is the motivation for posting. If the real goal is sharing an answer to a question that might be useful also for others then I don't see a problem. On the other hand if the motivation is promoting then I think it is probably problematic to post it on MO.

You should think carefully:

  1. do you want an answer to the question? or
  2. do you want to ask a reference request question about the current state of a topic and what people have tried? or
  3. do you want to check if your solution is correct/valuable/publishable?

If the first one then IMHO it is OK to post your solution as long as the question is on-topic on MO (though as you can see others think it is problematic).

If the second one (reference request) then your new solution is not an answer to the question. An answer to the question would be references, not your solution/idea/etc.

If the third one (peer review) then I think it is likely to be problematic to even post the question on MO.

  • $\begingroup$ "What is more important IMO is the motivation for posting." I think there are at least two problems in this view. The first problem is that you never know the OP's real motivation. The second is that a mathematical problem should be evaluated only by its intrinsic value. The OP's motivation is irrelevant to the value. $\endgroup$ Dec 1, 2015 at 21:05

You must log in to answer this question.

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