What can I do if I discover a reparable error in an article that I submitted to a journal of mathematics? I want to know the possible solutions so that the article is not rejected. And what happens if I submit the revised article to another journal in order to remedy this problem? Thanks for your help.

So I add some elements to clarify:

  • The mistake is in the proof of one of the main results of the work.

  • The correction is not complexe and it not affect the schema of the work, but I don't know if the referee when he correct it affect or not the schema of this work.

  • In the case where the referee will not signal the mistake, but give another remarks in his report, can I in this case signal the mistake to him, or I must follow their instructions strictly in order to have a favorable decision after I sent the revised version? The same question is when the referee corrects the error but affect the schema of the work while we can do it without affecting the structure of the paper.

  • For the option: send the journal a revision and explain the reason for it, does the journal accept the revised version? and how it class it?(because every submited article has only one code under the same journal!).

  • For the option:retract the article, is it a good solution? and if I will submit the revised article to the same journal, the later accept it?

I can not decide if I wait the answer of the referee, send a revised version, or I withdraw totaly the aricle and submit it once again.

Thunks again.

  • $\begingroup$ I think you can not submit a paper to two different journals before the decision. If it is not a serious error, then the reviewer would point it out in your report and it will not affect the decision. $\endgroup$ Oct 31, 2013 at 18:33
  • 1
    $\begingroup$ How would submitting the revised article to another journal remedy the problem? $\endgroup$ Oct 31, 2013 at 18:33

1 Answer 1


The best solution is to be honest: send the journal a revision and explain the reason for it. If the error is local, the referee will be easy on you (easier than if she/he discovers the error). Sending a revision to another journal is only an option after your withdraw the original submission, this is just the basic ethics of things.

If you discover the error after the paper has been published, it is not a tragedy but a different situation that happened to many of us (including myself). Then again the best solution is to be honest: publish an erratum on your webpage, or on the arXiv, or ideally in the journal. You can do these things in this order: first you park (or collect) the errata on your webpage, then after a while you put it on the arXiv, finally when it matured enough you submit it to the journal explaining the reason for it.

  • $\begingroup$ I don't completely agree with the first paragraph: I think you should either retract the article (but this seems a little odd if you are sure to be able to fix the mistake) or wait until the refeereing is complete and then signal the mistake and supply a (correct) revised version of the paper in case the paper is accepted without the referee noticing the fault (otherwise he will probably have supplied a solution of his own)-if it is rejected because of the mistake you can always submit another version later. $\endgroup$ Oct 31, 2013 at 18:39
  • 13
    $\begingroup$ @Jean: I understand your reasons, but I cannot agree with them. Not letting the referee know about a serious error is really misleading her/him. The error will be disclosed anyways, so I don't see the point of waiting. $\endgroup$
    – GH from MO
    Oct 31, 2013 at 18:45
  • $\begingroup$ I think that if the error is really serious (if it affects one of the main results, for example) then you should definitely retract the paper and submit a new version (to the same journal or another). If a person has begun refereeing something and is given something significantly different instead it will not be good for the process. The 'letting it be' approach I suggested was more in the case of a mistake in a proof which you can fix without consequences on the scheme of the paper (for example, a wrong argument in the proof of an intermediary result which holds nevertheless). $\endgroup$ Oct 31, 2013 at 18:52
  • 1
    $\begingroup$ @JeanRaimbault: I disagree. If I were to referee a paper, I would much rather the author be up front about things like this so I don't waste time reading a manuscript that he or she knows is flawed. If the author submits a revised manuscript correcting the error, I as the referee will have to read it all again anyway. The part I've looked at before the correction made it to my mailbox is a sunk cost; why make it any bigger? $\endgroup$ Nov 1, 2013 at 14:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.