37
$\begingroup$

There has been a lot of discussion about questions that do not get answered on MSE, but are not considered 'research level' on MO and thus get closed. The problem seems to be that certain questions are too easy for MO, but get lost on MSE between the hundreds of easy-to-answer questions. (Even if you stick to a single tag (algebraic geometry in my case), many questions are easy exercises.)

One could argue that the target audience of MO is primarily faculty, and that MSE seems to be used mostly by undergraduates (this is certainly the impression one gets by looking at the active questions on the main page). If this analysis is correct, could it be useful to create a new 'in between' SE page aimed mostly at graduate students and postdocs?

The advantage of such a website would be that it would constitute a single place where graduate students can post and answer questions; where no question is too hard; but where one can still filter out elementary questions so that the interesting ones stick to the front page. This might also encourage more of the senior MO members to engage with the not-quite-research-level questions.

Graduate school is a weird time, and the types of questions one asks and is capable of answering are a big step away from both undergraduate work and research-level mathematics.

Remark. The reason I post this question is that I think there is a real problem with intermediate level questions, experienced by many people. This is one suggestion for dealing with it, but really I want to stimulate discussion rather than propose a single solution.

$\endgroup$
  • 35
    $\begingroup$ My opinion is that questions by graduate students are fine on MO. I conceive of the difference between MO and MSE as essentially the difference between serious graduate students and undergraduates. $\endgroup$ – Joel David Hamkins May 28 '16 at 0:01
  • 20
    $\begingroup$ @Joel, where should humorous graduate students post? $\endgroup$ – Gerry Myerson May 28 '16 at 0:25
  • 14
    $\begingroup$ @Joel: unfortunately, the community often seems to disagree. This may not be the view of any individual, or even of many individuals, but collectively we have decided that most questions a graduate student has are not appropriate for MO. And even if they are, graduate students often feel discouraged from asking them. $\endgroup$ – R. van Dobben de Bruyn May 28 '16 at 0:46
  • 23
    $\begingroup$ I don't think we've collectively decided any such thing. There are a huge number of very successful questions asked here on MO by graduate students, perhaps thousands of such questions. I encourage graduate students everywhere to ask serious well-thought-out questions here on MO that arise naturally in their graduate studies. We shall all get to consider some interesting mathematics this way. $\endgroup$ – Joel David Hamkins May 28 '16 at 1:16
  • 18
    $\begingroup$ @GerryMyerson I guess they shoud post the snarky second comments... $\endgroup$ – Joel David Hamkins May 28 '16 at 1:17
  • 3
    $\begingroup$ Some related discussion at meta.math.SE: Postgrad Mathematics (which mentions a short-lived area51 proposal), Concern about lesser attention towards relatively advanced questions, Would splitting the site into more elementary and more advanced questions help?. $\endgroup$ – Martin Sleziak May 28 '16 at 2:07
  • 2
    $\begingroup$ @Martin: thanks, those discussions are very relevant indeed. My [selection biased?] impression is that the issue is acknowledged, but perhaps my proposal (which apparently is not a new idea) is not the correct solution. $\endgroup$ – R. van Dobben de Bruyn May 28 '16 at 3:02
  • 2
    $\begingroup$ I feel that the issue has not been addressed appropriately. However, this solution has been proposed in the past, often with the same conclusion. This is cause for me to vote to close the question at this time. (Technical note: it is not a duplicate, because all occurrences of this proposal are on meta.MSE) $\endgroup$ – R. van Dobben de Bruyn May 28 '16 at 3:47
  • 6
    $\begingroup$ As far as I have seen, if you are asking questions not to get solutions, and you've shown your efforts and explained where you got stuck, then your question will be well received. If you don't, then you run the risk of having it closed and/or downvoted and not answered. $\endgroup$ – Asaf Karagila May 28 '16 at 4:54
  • 23
    $\begingroup$ I don't see a need for a site in between MO and Math.SE for graduate students. -- Interesting questions by graduate students are definitely on-topic on MO, and for the 'rest' (e.g. textbook exercises, questions regarding the understanding of some standard definitions, etc.) there is still Math.SE. $\endgroup$ – Stefan Kohl May 28 '16 at 21:57
  • 8
    $\begingroup$ I've always thought (and have repeatedly advocated that) the threshold for MO should be "could a strong second-year graduate student ask this question?" This is the most generous interpretation of "research-level" that I think is possible, and I think (not coincidentally) the most beneficial interpretation for the MO community to take. $\endgroup$ – Steve Huntsman May 30 '16 at 14:14
  • 6
    $\begingroup$ Graduate students should use mathoverflow so that they can get to know the larger mathematical community outside of their own university. Furthermore, there are plenty of good graduate students who have asked and answered many questions here on mathoverflow. Not only can graduate students ask interesting questions but sometimes the best person to answer a certain MO question is a graduate student because that graduate student may have some specialized knowledge or may simply be more interested in the question than other mathoverflow users. $\endgroup$ – Joseph Van Name Jun 2 '16 at 5:44
  • 4
    $\begingroup$ I though would also encourage graduate students to both ask and answer questions here on mathoverflow instead of just asking questions. I think we can all agree that it is not good for someone to only ask questions here on mathoverflow but to attempt to answer questions at a similar rate as one asks questions. One estimate between MSE and MO level graduate students is if the graduate student is able to answer at least some good mathoverflow questions. $\endgroup$ – Joseph Van Name Jun 2 '16 at 5:45
  • $\begingroup$ there are declared charters/ scope, and then there is voting/ emergent community feedback/ dynamics, and these are not always the same thing, and it seems sometimes a lot of dialog/ discussion on this tends to miss that. there also exists a kind of "SE culture". here is a vaguely similar meta discussion on cstheory that mentions (eg in comments) too much rigor/ intimidation for graduate students where it might be affecting overall community involvement/ engagement. is interesting activity on cstheory declining? $\endgroup$ – vzn Jun 5 '16 at 17:03
50
$\begingroup$

I believe it is impossible in general to distinguish questions asked by a graduate student and those asked by a mathematician in a domain in which they are not specialist. Since the latter are allowed on MathOverflow, it would make little sense to disallow the former or relegate them elsewhither. It is hard to argue that MathOverflow receives too many questions, or that there it is at a serious risk of being overflooded by questions by graduate students.

There are also a number of questions being asked (and to avoid pointing the finger, I plead guilty myself), and often well-received, which are obviously motivated not by actual mathematical research, but just "general intellectual curiosity" (as in "I need to know this": I hope we can all agree that this is healthy). Again, it would make little sense to allow those and not questions asked by graduate students.

I think the main criteria should be something like: (1) the question's answer is not easy to find in standard textbooks on the subject, and (2) it has mathematical interest (either on its own or in order to solve a problem that does). This should exclude most cases of homework. I believe graduate students should be encouraged to take the time to think on their own before asking a question, and be sure to frame it carefully (and make sure it's not homework), but so long as they do so, and show their efforts, they should feel perfectly welcome to post on MathOverflow, and this should be made clear.

In a practical sense, apart from posters who will blatantly disregard guidelines and rules because they don't even read them, and a few cranks, I suspect there's more of a tendency to err on the side of self-censorship than in the other direction. I've mentioned MathOverflow to a number of colleagues, and surprisingly many of them are apparently too shy to join, even to discuss their domain of research, because they feel intimidated by the level of the existing discussions, or for fear of seeming foolish (he who never hesitated to raise his hand and ask a question at a seminar, let him be the first to throw a stone ☺). So I would find most welcome a change that could make the site seem just a little less "elitist".

$\endgroup$
  • 3
    $\begingroup$ What if I don't raise my hand, but just wait for a lull, then cut in with my question? $\endgroup$ – Asaf Karagila May 30 '16 at 4:01
  • $\begingroup$ (3) the experts of the domain are happy to share their knowledge with the students $\endgroup$ – reuns Nov 5 '16 at 6:06
14
$\begingroup$

It is doubtful that SE would allow another "mathematical" mathematics site to open.

In fact some time ago (2012, so even prior to MO's move to the SE network) there was an Area 51 proposal for a "Postgrad Mathematics" site. The proposal itself has been deleted, but a discussion or two live on. It is highly likely that an SE employee closed it after deeming that it would "tend to drain audience from an existing Stack Exchange site".

That math.se has grown immensely since 2012 probably wouldn't sway the SE folk. Frequently proposals for sites centred on specific programming languages/technologies are closed because they would "tend to drain audience from" Stack Overflow.

$\endgroup$
  • 2
    $\begingroup$ The split into two (MO and M.se) would itself not be possible nowadays, so a third site is pretty clearly impossible. $\endgroup$ – Noah Snyder Jun 3 '16 at 5:35
12
$\begingroup$

Personal opinion: I think that the current system of having several of separate sites for different math-related topics (apart from MO and MSE, there are also related ones like scientific computing, math educators, history of science...) is not the best one, and adding new sites doesn't help; it just makes the community more fragmented.

Maybe we need tags like / on a common site, and a better automated system to filter out questions based on the preferences of the individual users. After all, Stack Overflow is a very high-traffic site that covers all levels of programming questions, from helloworld.c to monads and variadic templates, and no one bats an eye.

$\endgroup$
  • 4
    $\begingroup$ A filtering system would only be as good as the tagging, and it is not clear to me that users will tag their own question correctly (or indeed tag other people's questions correctly). $\endgroup$ – Yemon Choi Jun 2 '16 at 22:46
  • 5
    $\begingroup$ Also (as discussed before), SE websites discourage the use of meta-tags. See for example the help centre. $\endgroup$ – R. van Dobben de Bruyn Jun 3 '16 at 2:01
  • 4
    $\begingroup$ @YemonChoi I don't see it as a specific downside of my proposal, but as a more general issue: we'd have the same problem if you replace "tag" with "SE site". Often users don't post their question to the correct site, and we rely on high-rep users and moderators to correct them. $\endgroup$ – Federico Poloni Jun 3 '16 at 7:03
  • 3
    $\begingroup$ What possible benefit would this bring to MO? It certainly has downsides (e.g. a much lower signal to noise ratio). There is a reason I don't participate on math.se... $\endgroup$ – Andy Putman Jun 3 '16 at 14:55
  • 3
    $\begingroup$ @AndyPutman "What benefits would this bring to MO" seems a strange point of view to me: I could not care less about MO as an abstract entity. Better question: what benefits would it bring to you and me as users? I'd love an automatic system that aggregates the most interesting questions (based on my favorite tags and activity) that are currently scattered along five separate sites. It would let me find the good stuff more easily. It would eliminate the reason why you don't participate on math.se, for instance. $\endgroup$ – Federico Poloni Jun 3 '16 at 21:01
  • 3
    $\begingroup$ @FedericoPoloni: I very much doubt that I would be interested in questions at math.se or the other sites. Given how poorly people use tags, combining the sites would definitely force me to wade through many more poor questions than I do now (though probably I would just get fed up and leave). What is more, I really do not want eg the math.se people voting/commenting on the questions and answers at MO. There are so many of them that they would quickly overwhelm the knowledgeable users. $\endgroup$ – Andy Putman Jun 3 '16 at 22:30
  • 2
    $\begingroup$ @Andy, given the 101 association bonus for users of multiple SE sites, "the math.se people voting/commenting on the questions and answers at MO" is already possible now. (Certainly, I'm one such user.) $\endgroup$ – J. M. is not a mathematician Jun 4 '16 at 0:31
  • $\begingroup$ Setting aside a bit unclear phrase math.SE people, the issues pointed out in @AndyPutman's comment are more closely related to some older discussions at meta.MO, such as The Association Bonus and Measures to separate math overflow from the rest of the stack exchange network. $\endgroup$ – Martin Sleziak Jun 4 '16 at 6:37
  • 2
    $\begingroup$ "I'd love an automatic system that aggregates the most interesting questions (based on my favorite tags and activity) that are currently scattered along five separate sites." There is a cross-site filtering system. Here is one I set up for myself focussing on geometry: link. Here is some explanation of Tag Sets. $\endgroup$ – Joseph O'Rourke Jun 4 '16 at 12:52
  • $\begingroup$ @AndyPutman I don't find most MSE questions too interesting either, and rarely use it (not that I'm a heavy MO user either), but occasionally there are questions I like. So I wouldn't say the site is completely useless to "senior" mathematicians, though I agree there is too much "noise" to make me want to browse MSE. $\endgroup$ – Kimball Jun 4 '16 at 19:58
  • $\begingroup$ @Kimball There were some suggestions on how to find interesting questions on Math.SE. This is not exactly related to this discussion, so I have posted some that I am aware of in chat. (We can continue this discussion there, if needed.) $\endgroup$ – Martin Sleziak Jun 5 '16 at 9:15
6
$\begingroup$

After carefully reading all your thoughts, I conclude that MO is a perfectly acceptable place for graduate students to post well-thought through questions. Yet, many graduate students (and even professors, cf. for example Gro-Tsen's comment on this question) are intimidated by how harshly the community votes.

What this leads me to conclude is that there is a serious disconnect between the type of question that is theoretically acceptable and the collective demand of the community. As I said in a comment to my original question: this may not result from the opinion of any single individual, or even of many individuals, but should be viewed rather as a collective behaviour. (Come to think of it, I probably have been guilty of downvoting perfectly reasonable questions myself.)

As for a solution, I think one thing we can do is try to be more understanding of each other's background (and potential lack thereof), and vote more with the community guidelines in mind.

$\endgroup$
  • 5
    $\begingroup$ I think the "level" of a question is not the right parameter here. -- The required minimum "level" of a question on MO is not that high, and at times there are also well-received soft questions with not much concrete mathematical content at all. What is probably rather the issue here is how a question is formulated, and I think it is not without reason that the community usually doesn't like questions which are written down like on a personal piece of scrap paper (regardless of "level"). $\endgroup$ – Stefan Kohl Jun 13 '16 at 20:27
  • 1
    $\begingroup$ I second Stefan. I personally haven't seen any technical question above undergraduate level which is well written and shows at least some effort on the part of the author to find an answer that is not well treated. The issue I think many new comers don't know how to write a clear concise question demonstrating at least a bit of personal effort to find an answer. (Soft questions from new comers are treated more harshly and have to satisfy higher expectations, but that I think is intentional.) $\endgroup$ – Kaveh Jun 28 '16 at 7:49
  • $\begingroup$ @Kaveh I feel like I have seen some in the recent years (though "above undergraduate" is hard to define; in terms of triviality or in terms of it being a part of the standard curriculum). $\endgroup$ – user138661 May 20 at 7:40

You must log in to answer this question.

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