Not signed in (Sign In)

Vanilla 1.1.9 is a product of Lussumo. More Information: Documentation, Community Support.


    When I talk to people about MO, they sometimes ask what happens if somebody posts their thesis problem. See, for example, the "diploma thesis" meta thread. My usual answers are that (1) good thesis problems are very different from good MO questions, so it shouldn't come up much, and (2) there's a public "paper trail" so the possibility of academic dishonesty isn't any larger than anywhere else.

    Recently a user (who I'll call Q) posted a handful of questions. The questions got pretty substantive answers, but the answers got no more than 1 upvote, probably because the questions were somewhat technical, so Q was able to delete the questions. One of the users (who I'll call A) who answered one of the questions emailed a moderator to find out what happened to his nice answer, the moderator emailed Q saying that A would like the question and answer undeleted, and Q emailed back saying effectively, "Please don't undelete that. It's a problem I'm working on with my advisor and I'm not supposed to discuss it in public."

    In the spirit of open decision-making, I'm brining the issue to meta. If you're a 10k+ rep user, you can have a look at the questions and answers here, here, and here. If you're not a 10k+ rep user, you can email and one of us will direct you to some copies of the questions I've posted elsewhere.

    My opinions

    I'd like to undelete the questions and answers since it seems dishonest and unfair to the answerers to leave them deleted. More generally, I'm for the policy that a thread can be undeleted at the request of anybody who contributed, or anybody at all. To me, this is a natural extension of not allowing people to delete questions with substantive answers and of allowing 10k+ rep users to vote to undelete. I'd even be in favor of regularly scraping the database for deleted questions with substantive-looking answers to make sure this sort of abuse doesn't go unnoticed.

    • CommentAuthorWill Jagy
    • CommentTimeNov 5th 2010
    In a less dramatic example, I got nervous when

    seemed to be putting his entire education as MO questions. It has slowed down to a trickle, I guess as the semester has gotten underway in earnest.
    • CommentAuthorRyan Budney
    • CommentTimeNov 5th 2010 edited
    I agree the questions should be undeleted unless there was a compelling reason to delete them. From the information we have so far, it seems the worst possible outcome would be to keep them deleted.

    Small side-question: when does the OP lose the ability to delete the thread? I thought if there was one answer, the thread was undeleteable. But it appears it's more complicated than that.

    @Anton: for frequent viewers of MO and Meta (like yours truly), your anonymising efforts are, well, somewhat pathetic. :)

    Anyway, I am in favour of your point of view. While I sympathize with Q, he or she should learn a lesson from this: that if something should be kept private, putting it on the internet is a horribly bad idea.

    Also, may I propose that said policy not extend to deleted answers?


    @Will Jagy: it's a good thing that you don't follow math.SE then, else you'd develop a twitch within days.

    @WillieWong, IMO if there's a good reason to delete something, fine. Like if an answer is so wrong it's more confusing than a benefit to the thread. I think I have one or two like that and I'm happy that they're deleted.
    • CommentAuthorWill Jagy
    • CommentTimeNov 5th 2010

    Meg Ryan to Kevin Kline in "French Kiss"

    Inside. Fester and rot.

    I've seen it happen. You'll become one of those hunchbacked, lonely old men
    sitting in the corner of a crowded cafe, mumbling to yourself.

    My ass is twitching. You people make my ass twitch.

    Small side-question: when does the OP lose the ability to delete the thread? I thought if there was one answer, the thread was undeleteable. But it appears it's more complicated than that.

    The OP loses the ability to delete when there is an answer with at least two upvotes.

    @Anton: for frequent viewers of MO and Meta (like yours truly), your anonymising efforts are, well, somewhat pathetic. :)

    Are they more pathetic than I meant for them to be? I think the details are often important, so I want them to be accessible with a small amount of effort, but I want to erect enough of an energy barrier that the conversation doesn't stray into rubbernecking at some guy who made a mistake.

    Also, may I propose that said policy not extend to deleted answers?

    It's hard for me to imagine an answer where the same sorts of concerns apply.

    If an answer has multiple contributors, I think any one of them should have the power to undelete it. I certainly agree that there's no sense in undeleting confusing or just flat wrong answers that the owner deleted, but the hypothetical situation is that somebody honestly feels that there is some value in the deleted post and has gone out of their way to say so.


    @Anton and Ryan: my hypothetical scenario for answers is the following: say somebody (maybe yours truly) made a completely boneheaded answer. Someone points out a trivial mistake in a comment, and then proceeds to give a rather technical description of the subtlety of the question and answer (disclaimer: I haven't seen it happen, just fibbing at the moment). Then the poster decides that the non-answer is embarassingly wrong and deletes it. Should the commenter have a say in whether the answer show be resurrected? (Because of his rather insightful comment.)

    My vote is no. The commenter should be able to re-post whatever he wrote as a new answer. (Possibly with the help from one of the Mods to copy over his comment text so he doesn't have to re-type all of it.) So the wrong answer can stay safely deleted with no loss of information. But this is something that potentially need discussing if we decide that a deleted question can be un-deleted if someone else finds it useful.


    @WillieWong: I agree with you about that hypothetical, so let's spice it up a bit. Suppose the wonderful and insightful comment doesn't make sense as an answer, but adds real value in the context of the boneheaded error. Should the commenter then post a separate answer saying something strange like, "One could make the following error, in which case ..."? In this situation, if I were the bonehead, I would opt to preserve the interaction as it happened. If it were somebody else who really wanted the answer deleted, I'd probably try to convince them to do the same, but I don't think I could justify absolutely insisting on it.

    @Anton and Willie: I'm happy with Anton's gray-area hypothetical remaining gray and hypothetical until we ever come across a situation where there is enough emotional investment among the community to try and come to a decision.

    @Ryan: that sounds good to me.


    Returning to the original issue, out of curiosity, has it been pointed out to Q that regardless of whether the question is undeleted, it is already viewable by at least 20 people and that number will most likely grow substantially?

    I mentioned this to Scott M. in person yesterday, but my opinion is that you should tell Q "Please have your advisor email us and explain the situation. If we do not get an email from your advisor within a week we will undelete the questions."

    Any legitimate reason for these deletions (e.g. advisor was worried the problem would get stolen and asked Q to delete, etc.) would allow the student to talk to the advisor. But if Q is trying to hide something from their advisor then there's no reason we should be helping Q cheat.

    For what it's worth, I agree completely with Noah.

    (I'm sitting with Anton in LAX right now.) I also strongly agree with undeletion for the reasons others have stated. I would be bothered if they stayed deleted. If Q makes a case that they should stay deleted, that convinces the community, then I'd be satisfied.

    @Noah: I don't really see why the opinions of Q's advisor should make much difference. Even if the advisor wanted the questions to remain deleted, I'd still be firmly in favor of undeleting them. Regardless of whether Q is cheating with his advisor by deleting the questions, he is cheating the people who answered his questions. Getting his advior's help in doing so doesn't make the situation any better.

    Regarding the possibility that somebody else steals Q's thesis problem, Ravi made a very nice point when we were chatting earlier: there's a definite social cost in the mathematical community to stealing a student's thesis problem. One could argue that having a public record does as much to help Q in his claim to the problems as it does to harm him by making the problems more public. That might be overstating the case a bit, but the point is that in the presence of substantive answers, I don't see a good argument in favor of leaving the questions deleted.


    Okay, I've gone ahead and undeleted the questions!

    I'll write to the OP too, to let him know about this meta thread and its progress.


    The links in Anton's original post seem to have disappeared. Are the questions still deleted? I would like to read them if they were reopened, thank you.


    They've been removed to avoid making a spectacle. Any who wants pointers to the links enough to make us send you a email should contact me or directly.

    • CommentAuthorMariano
    • CommentTimeNov 9th 2010

    So you undeleted and then removed them? I guess there were undisclosed events! :)

    • CommentAuthorHailong Dao
    • CommentTimeNov 9th 2010 edited

    @Mariano: I think the links in Anton's post were removed, but the posts are still there.


    Hailong is correct. The posts are now undeleted on MO, but I edited out the links in the original post of this meta thread. I guess 10k+ rep users can still see which questions they were by looking at "recently undeleted questions" under the delete tool in the tools menu.