About 4 years ago, before I started my PhD, I published a paper which used a method to improve an existing result. There were 2 papers on the existing result and method which I used for my work and appropriately cited. A couple of days ago I discovered another paper on arXiV by the same authors, which dealt with the same problem, which I did not cite for reasons I can't remember. Also, I realize that they address the same problem that I worked on, in a small sub-section of their arXiV paper.

Based on the references that I used, I gathered that the problem I solved had not been looked into in the works of theirs that I cited and made a statement about this fact. Although, the method that I have used is different from theirs, it still makes my statement erroneous. I feel that I am thoroughly screwed and the damage is irreparable. How do I go about resolving this issue? I am willing to let everybody involved on my paper know about it, including the authors whose papers that I have not cited. Is there anything more to be done?

  • 3
    FYI, this situation occurs in math more than outsiders probably realize. I referenced a rather extreme example in a comment to this mathoverflow question, leading the OP to remark "I don't offhand recall any topic, at least in analysis, where basic examples and results have been rediscovered, reproved, and republished so often ...". See also Alexander Woo's answer to Finding related result in mathematics, although in this case the discussion is about results found prior to publication. Feb 20, 2023 at 17:39
  • 4
    What exactly is the damage, in your view, and what statement is "erroneous"? Feb 20, 2023 at 17:42
  • 9
    ... the false assumption that such a thing had not been done -- My gut reaction is that this is not all that serious, but I imagine in the future you'll be more inclined to say things like "seems not to have been done" or "we have not found this result in the literature"! Feb 20, 2023 at 18:12
  • 2
    You might also want to have a look at this answer of mine to a related question. Feb 20, 2023 at 18:21
  • 1
    Regarding my earlier comment "... I imagine in the future ...", the examples given in this answer are worth reading over. Feb 28, 2023 at 8:58

3 Answers 3


It's difficult to tell precisely as a bystander, but this just doesn't sound like a particularly serious inaccuracy. From your description, your previous publication make no claims about the other authors' paper you've more recently found.

I realize that they address the same problem that I worked on, in a small sub-section of their arXiV paper.

So you missed a small sub-section of a preprint in your literature search. This type of thing happens all the time, at least in my part of math. Yes, you could wish that you had spotted it, but it sounds like your method was still a novel approach to the problem. You're certainly not "screwed" with "irreparable damage" to the paper.

If you like, in the future you can add a bit of hedging to novelty statements going forward. I sometimes include this kind of language:

To the best of the authors' knowledge, the method we will discuss here is the first to solve the problem of fizzing buzzes.

In the grand scheme of things it actually doesn't matter, except in acknowledging that literature searches are hard and that you haven't intentionally neglected something a reader or reviewer happens to know about.

  • There are two things that I'm really beating myself up over: 1) Not having discovered the arXiV version, or having discovered it but missing the small part which addresses my paper, and 2) using my lack of knowledge of it in some way to justify my novelty by saying that this was not addressed in their work. I see your point that since I did not make a claim about their arXiV version at all and also haven't cited it, it wouldn't be considered majorly careless or dishonest. The method that I have used remains different from their work, all though the goal that is achieved happens to be the same. Feb 20, 2023 at 18:03
  • 9
    @burnedstudent I'm pretty sure in hindsight that I misunderstood a whole genre/branch of work in the area of my MSc thesis and essentially missed it in both the thesis and paper's lit review. My approach was new, but I could've probably fit it into that genre's existing framework gracefully. C'est la vie, I don't lose sleep over it. The main idea was still new and interesting.
    – user137975
    Feb 20, 2023 at 18:11
  • 1
    This makes me want to cry. The number of times that I have discarded potential problems because a method to solve it already exists is really large! This is something that I will definitely keep in mind. Thanks for your answers! Feb 20, 2023 at 18:29

I am assuming that when you say published, you mean published in a peer-reviewed journal. On the other hand, the paper you found was not published. It was on ArXiv. I argue that because of this, you should not feed bad at all. You weren't scooped, since that paper wasn't published.

ArXiv is a non-archival venue. There is no peer-review process, and some academics (myself included) are extremely wary of citing ArXiv papers that haven't been published elsewhere. If they were published elsewhere, I cite that version, not the ArXiv version. The main purpose of ArXiv in my opinion (which is how I use it), is to make your work accessible to a broad audience. Some folks treat their ArXiv papers as published, and expect to receive academic credit for them. I disagree with this viewpoint on a moral/ethical basis, as it undermines the entire purpose of the academic publishing process.

Now, let's pretend that the paper was published and that you simply missed it. Is it truly identical? It seems very unlikely that the result, method, techniques, everything is the same.

I wouldn't worry about it. Let me tell you a story that happened to me not too long ago. We had a result (call it A) that we wanted to publish. When we were in the process of submitting it, a paper with a result B was "published" on ArXiv. Result B generalized our result. However, given the argument I made above, we went ahead and submitted our work, and it got accepted. Later that year, another result C, was published in another venue, which was a special case of our work. They were all valuable contributions in their own way, even though the only one that "mattered" was Result B (the most general one). Joke's on them though - we published an even more general result earlier this year!

This is how science works. We build on each other's work, and if a topic is interesting, you'll have a lot of duplications. As long as you're operating in good faith, citing sources appropriately and acknowledging effort, you're doing it right.

  • 2
    I find your statements about ArXiv somewhat bothersome. Yes, documents on ArXiv do not go through a peer-review process, and thus do not have the same kind of "quality checking" that makes the claims in academically published papers somewhat more trustworthy than a random statement. But in the described case, it does not sound as if the paper could have been cited for a claim, something that may or may not be true. It would have been cited for what it is or contains. I certainly do not blame the OP for not finding the paper, but had they been aware of its contents, a statement along ... Feb 20, 2023 at 23:49
  • ... the lines of "We are the first to look at this problem." would be disingenuous. They were not, and the very existence of an earlier document that talks about the same problem is proof of that, irrespective of whether that earlier document is a peer-reviewed publication, an ArXiv preprint, or a or a torn out notebook page discarded and found in some rubbish bin. Feb 20, 2023 at 23:49
  • 1
    Perhaps it’s an artifact of my field (CS), but encouraging the practice of people putting very preliminary stuff on ArXiv just to be first results in worse academic practices. Computer science has problematic review practices (to put things mildly), and rushing to put things on ArXiv make it worse IMO
    – Spark
    Feb 21, 2023 at 1:11
  • 1
    @burnedstudent: Ah, "had not been looked into in the works of theirs" did not make that quite clear, but then it's not much of an issue. (In any case, if you genuinely didn't find the paragraph that tackles the issue I really don't see a problem, anyway.) Feb 21, 2023 at 7:56
  • 1
    @O.R.Mapper I would have been happy in hindsight if I had found their paper and method so that I could compare my approach with theirs. Feb 21, 2023 at 15:56

Personally, while not a huge issue, I do feel like other answers are downplaying this a little more than it should be. In some sense it has become a chronic issue, to the point where I feel more than half of the results I see have been done before and few people in the field seem to have any idea that it was done before. Reviewers don't notice, colleagues are excitedly talking about the novelty of the approach, etc... and I'm like, "but xxxx did this back in the 30s." The harm isn't in this happening once, it is that over time entire fields seem to suffer a collective loss of memory.

So while I wouldn't beat yourself up about it, it might be worth publishing an erratum as a service to those that come after you trying to determine the novelty of their own work (they might be reading your work and not have seen the other one yet, and then may never read the other work). Of course with the caveats that 1) the journal you published in is amenable to such minor corrections and 2) your field and coauthors are similarly amenable to the idea. Personally, I respect authors willing to correct their works more than those that do not, but I admit that I know others that view any sort of correction, no matter how minor, as a black mark on the work.

  • I am a little anxious about how my co-authors would respond to it. I'm not sure if I would receive the same benefit of doubt that people seem to provide on here. I am definitely inclined towards doing what you suggested, as I indicated in my post. Feb 21, 2023 at 15:18

You must log in to answer this question.

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