47

I am writing a paper presenting a mathematical theorem I recently proved. However, while reading one of the papers that I am citing, I found out that the same result (or something very similar) might be present in the author's Ph.D. thesis.

To be more precise, the author only alludes to the existence of such result, adding a reference to his thesis. The problem is that such thesis is not available online and is written in Russian; moreover, it appears that the author has left academia, given that his last publication dates back to year 2000.

In this context, I am not sure what to do. To be honest, I am tempted to simply ignore the author's claim, as I have no means of verifying it, omitting any mention of it in my paper. Could this be considered unethical or harm my reputation in any way?

  • 9
    Would this be your main result in the paper or not? – Buffy Feb 13 '22 at 19:57
  • 11
    "I am writing a paper" what is the function of paper, is it like an academic paper for journal or term paper? what is your job: phd, post-doc, faculty? – aaaaa says reinstate Monica Feb 14 '22 at 02:32
  • 1
    I had a similar problem with my thesis. When I tried to publish a paper, the main result was rejected on the basis that it was a well known old result. On examination of the literature I found that a related but incomplete result had been published. My supervisor said that I should not worry too much because everyone does things their own way. My paper as a whole was my own work. On the other hand, it took an international flight and visiting the other academics institute to chat with them for a week before they agreed to back off and even became a reviewer for my thesis. – Bruce Feb 14 '22 at 07:26
  • 22
    You can add a section at the end of your article entitled "Note added" and include an honest summary of the above descriptions that you wrote. I have seen some people receive positive feedback by doing this. – Robin Feb 14 '22 at 08:45
  • 5
    As the Tom Lehrer song goes: "And Nicolai Ivanovich Lobachevsky is his name! Oi!" – smci Feb 14 '22 at 17:42
  • Do you have a Russian speaking colleague in your Department? Can you ask them to translate a relevant part for you? – Dmitry Savostyanov Feb 15 '22 at 23:56
  • 1
    dude, of course you mention this other person's work, especially now since there is public record of the fact you know better. how is this even a struggle? do not ruin your reputation over an ethical violation. – RyRy the Fly Guy Jul 18 '23 at 15:07

8 Answers8

74

You do have means of verifying it - for example write to the author and ask him for more information or a copy of his thesis, contact any coauthors on the paper you have read, find someone who speaks Russian who can help you, contact the library of his university to get the thesis, ask your university or national library if they can help you.

If you try those things and still cannot find or verify his proof, then you can publish your paper, but you should mention that the other person alluded to an earlier proof of it and explain that you were unable to find or verify it.

gib
  • 2,537
  • 10
  • 17
  • 27
    You can also write to the department of the university in which the person worked. – Buffy Feb 13 '22 at 19:15
  • 67
    I'd add that some librarians are a) freakishly good at tracking things down and b) really enjoy doing it. If you know one, ask! – Matt Feb 13 '22 at 23:14
  • 1
    This does not answer if ignoring the earlier result "could this be considered unethical" so is really incomplete. – Terry Loring Feb 14 '22 at 15:46
  • 1
    @Terry Loring The 2nd paragraph hints at an answer to that. But for a straight answer and other very useful information, please see the answer by Kostya_I. – gib Feb 14 '22 at 21:49
  • 7
    I remember Andrew Wiles saying the first thing he did was to write to Pierre de Fermat and upon not receiving any answer, he went ahead and published his theorem. – PatrickT Feb 15 '22 at 05:47
  • @gib Indeed. My eye jumped over that. – Terry Loring Feb 15 '22 at 05:51
  • 4
    @Matt It always amazes me when academics have to point out librarians. It's like if people had free access to plumbers but still stubbornly tried to fix their own pipes with chewing gum and cello-tape, only to have an occasional lonely voice break through the miasma to remind us how freakishly good a professional plumber is at fixing a broken pipe.... why is this news to anyone conducting research?! – J... Feb 15 '22 at 13:58
36

What you are contemplating is a major ethical violation.

First, it's unclear why you mention that the thesis is in Russian; it's completely irrelevant. The obligation to acknowledge prior work and give due credit exists independently of its language. As an author, you must honestly figure out and explain what new you are adding to the literature, if any. If your result is not new, you cannot publish it in a journal of original research, at least not as the only/main content of the paper. In this case, you may consider putting your paper on arXiv, making it completely clear that it is an exposition of known results.

To be honest, in the age of Google translate, with a math paper, and being an expert on the topic (so more-or-less expecting what's there), it's not so difficult either.

Second, as soon as a published work is concerned, its availability is your problem - and most certainly, non-availability online does not nullify any of the above-mentioned obligations. But in practice, most Russian dissertations are not so hard to obtain. First, every dissertation has a 20p. summary, called "автореферат", that is published in about 100 copies and in many cases can be freely downloaded from the website of the Russian State Library. If that is not enough for your purpose, Russian state library also has a physical copy of every dissertation, and there are services that will scan it and send you a copy for about 10 euro. Disclaimer: the legality of this service seems to be OK, at least in practical terms, but I cannot vouch.

On top of that, there are obvious solutions, such as contacting the author or e.g. their department/collaborators in Russia for a copy of the dissertation or the "автореферат".

Kostya_I
  • 4,071
  • 1
  • 14
  • 21
  • In many cases, verifying that a certain theorem exists in another language is not so hard as symbols and formulas tend to look the same in any language. – Terry Loring Feb 14 '22 at 13:19
  • 4
    @TerryLoring It depends on the type of math you do. I would say that most theorems in my area tend to be written more often in plain words than in symbols and formulas... – Denis Nardin Feb 14 '22 at 15:25
  • 18
    This is a bit over the top. The fact that it's only appeared in Russian is relevant - it means it's much more useful for it to be published once more but this time in English. And the fact that the result has already appeared isn't a show-stopper. The original proof may be cumbersome, make slightly different assumptions, be wrong! etc etc or be in different setting, making the significance of the result less clear etc etc – innisfree Feb 15 '22 at 03:27
  • 2
    @innisfree, what you are arguing with? I agree with all your points. I suggested myself that it would be very useful of OP to put their paper on arXiv, even if its only added value is English. In my experience, a reputable journal would not consider such a paper for publication. As you suggest, there may be more of added value, in which case a journal will try to assess it, but that would be independent of the language of the original. That said, nothing is wrong with trying to publish as long as the contribution is accurately described. – Kostya_I Feb 15 '22 at 13:44
  • 1
    @innisfree It's relevant to those things, but it's not relevant to the question of whether the thesis potentially containing the result should be located, understood and if necessary, cited as the original source of the result. – kaya3 Feb 15 '22 at 17:12
  • Does anyone here who doesn't speak Russian have any experience with reading Russian math papers translated into English via google translate? Would you recommend that method? – Stef Feb 15 '22 at 18:40
  • @Stef, why not try yourself? mathnet.ru contains tons of articles in Russian free to download, for many of which a human translation into English exists, if you want to compare (e.d., "Алгебра и Анализ" is translated cover-to-cover as "St. Petersburg math journal".). – Kostya_I Feb 15 '22 at 19:15
  • @Kostya_I Because I already spend a large amount of my time reading research papers without google translate! – Stef Feb 15 '22 at 19:20
29

It was getting too big in the comments, so I recast this as an answer.

Mention the other claimed proof, and offer your own proof.

I had a similar problem with my thesis. When I tried to publish a paper, the main result was rejected on the basis that it was a well known old result. On examination of the literature I found that a related but incomplete result had been published.

My supervisor said that I should not worry too much because everyone does things their own way. If you honestly did it yourself, the chances are that you will have something new to say.

My paper as a whole was my own work. On the other hand, it took an international flight and visiting the other academics institute to chat with them for a week before they agreed to back off. It was a successful trip as they even became a reviewer for my thesis.

If you did not copy the result, if it was genuinely your own work, then I do not see that you are committing any moral outrage. If the result happens to be well published and you can add nothing new, then the paper should be rejected for lack of originality.

The moral problem would be if you read the person's work, copied the idea, and claimed it yourself.

But, you should definitely note that you became aware of a CLAIM that there is a proof - a claim that you tried to but failed to validate.

At the very least, the result appears not to be well known in the literature. And that is important. Bringing attention to something not well known is useful, as is (potentially) giving a new proof.

The moral problem here is that now that you know that someone claims an earlier proof - as much as you might be tempted - you should not ignore that you now know. You might wish that you never found out. But, that is not for us to decide. What is for us to decide is what shall we do with that information.

Bruce
  • 391
  • 2
  • 4
  • 10
    +1 Always be honest. If you suspect that there might be proof somewhere in literature, which is not readily available, just say so and proceed. There is no guarantee, that you work has not been done already by somebody else. – Dohn Joe Feb 14 '22 at 10:16
  • 6
    The final lines here sound an awful lot like Gandalf’s advice to Frodo. :) – Wildcard Feb 14 '22 at 20:03
  • @Wildcard – Bruce Feb 15 '22 at 01:03
13

Yes, concealing information from the reader in order to make your work appear more novel or to avoid doing more work is unethical. Whether it would harm your reputation depends on whether readers of your paper will be able to deduce that you did this on purpose, which is anyone's guess.

In any case, just because a proof of some result exists in some obscure, hard to access Russian source doesn't mean that you cannot publish the theorem. It merely means that you cannot claim priority.

Ideally, you should invest a reasonable effort into trying to find this proof (to establish your priority or lack thereof) and if it exists, to compare it to your own (to determine whether the merit of your proof lies only in being more accessible to the reader, or if the proof is substantially different).

A less ideal but somewhat understandable course of action would be to state that, in effect, a proof might exist but you didn't feel like investing the effort into digging up an obscure Russian thesis and having it translated into English.

Adam Přenosil
  • 10,304
  • 2
  • 32
  • 40
  • 4
    I think instead of "some obscure, inaccessible Russian source" it would be better to say "a Russian source that is not online" – gib Feb 13 '22 at 19:46
  • 4
    @gib Why would "online" be the decisive criterion? According to this, any thesis older than about 25 years would not be a source. Talk to librarian! - they might well be able to get you a copy of the thesis (not sure if this is the case for a thesis in Russia, but it certainly will work for a thesis in your country). The answer rightfully uses "inaccessible", and this is indeed the point. Suggesting to take non-availablility online, and the need to talk to a librarian to get hold of a copy, as a reason to ignore it is a rather lousy excuse! – user151413 Feb 13 '22 at 20:46
  • 2
    @gib I changed it to "hard to access", which seems like a fair summary of the situation to me . As you and Buffy point out, it sounds like one needs to either contact the author or the department. – Adam Přenosil Feb 13 '22 at 22:56
  • I don't think it's that easy. First, one cannot publish (in a journal) a proof that is already in the literature, at least as the main content of an original article. Second, publishing an exposition will likely lead to people citing the exposition in lieu of the original work. Therefore, if you really have no new content to add, an appropriate course of action is to publish a translation instead. A translator, justifiably, receives less credit, in particular, they are not listed as an author. – Kostya_I Feb 13 '22 at 23:25
  • 9
    Lots of things used to be hard to access, and we did our best using letters, librarians, and any other connections we could try. "I can't find it with Google in less than 5 minutes" is not a good excuse. – Jon Custer Feb 14 '22 at 02:21
  • 1
    @JonCuster I don't think we are in disagreement. I am, quite explicitly, not suggesting it as the prefered course of action, but simply as something that's infinitely preferable to lying by omission. Being a lazy but honest researcher (and hoping that the referees won't call me on it) is much less serious than being a dishonest researcher. – Adam Přenosil Feb 14 '22 at 11:19
  • 3
    @Kostya_I I disagree that, outside of the very top journals, one cannot publish as a main result a proof of a theorem that has already been proved by someone else. This depends entirely on the value that the paper brings to the mathematical community, which it can do in various ways. Here it's clear what the value would be, namely making a proof of the theorem easily accessible to the English-speaking reader. Unless this is done in a dishonest way, I fail to see the problem with this. – Adam Přenosil Feb 14 '22 at 11:22
  • @Kostya_I Re your second point, surely translating is not a realistic option here? The OP is hardly going to translate the thesis into English and the idea that they are going to convince someone else to do this strikes me as pure science fiction, unless the result is a major breakthrough. (And if they did, the result would be what – a pdf sitting on the OP's webpage somewhere?) Who benefits from the idea that publishing further proofs is now verboten? Certainly not the original author – their choices are between the result staying obscure or gaining further attention thanks to a new paper. – Adam Přenosil Feb 14 '22 at 11:26
  • @AdamPřenosil, I agree with the practical side, but still I've never seen a paper in a reputable math journal saying "our main result was proven in X, the goal of this paper is to provide an English source.". I feel it would be frowned upon, since if the goal of your paper can be achieved by translation, it sets a benchmark on how much credit you can claim. And if the result is not important enough for anyone to translate, why it would it be important enough to publish as an original contribution with a different author? – Kostya_I Feb 14 '22 at 12:07
  • 1
    @AdamPřenosil - we are in complete agreement, and I should have worded that better. – Jon Custer Feb 14 '22 at 13:52
  • @Kostya_l Fair enough about the first point – one would hope for some added value beyond that. On the other hand, very few researchers are going to spend their time translating a result which is (let us assume) perfectly publishable but average simply out of charity. That doesn't mean that they would not like to read about the result. I don't know about you, but my bar for papers which I'd be willing to spend my time translating is much, much higher than the bar for papers I'd recommend to accept as a referee (as evidenced by the fact that I haven't translated any). – Adam Přenosil Feb 14 '22 at 16:30
4

In addition to what has been mentioned in other answers:

You may find that your proof is substantially different from the existing proof. If so, this is still a potentially valuable contribution to the literature. This is especially true if your proof contains a "new idea" which could be applied in other situations. Comparing and contrasting your approach with the other approach should be part of your paper.

Steven Gubkin
  • 2,166
  • 1
  • 12
  • 12
3

Other answers describe how you can find more information on the thesis in question. I agree with others that you should mention the possible predecessor. However, Russian theses are considered unpublished (at least that was the situation in the past), although the requirement was that main results of a Russian thesis had to be published. On the other hand, if the predecessor quotes his thesis and nothing else, maybe the result was only in the thesis and thus unpublished. This may be a basis for you to publish the result and its proof.

I was in your situation some time ago. I derived a pretty important result and found a similar claim in a book. I issued an arxiv preprint quoting their claim and mentioning that I had not been able to find the proof. After that I contacted the authors of the book. It was not easy to sort out, but eventually they explained that they had in mind something quite different from what I derived.

EDIT (Feb 14, 2022): By the way, I find it amusing that, technically, Wiles was in a situation similar to that of the OP, as Fermat had claimed having a proof of his Last Theorem :-)

akhmeteli
  • 1,315
  • 7
  • 12
  • 1
    But let's not forget that that plagiarism is shoeing someone else's ideas as your own. It does not matter where exactly they were published. It can happen in published research, it can hapoen in homeworks, it can happen in belletry. If you dfound this idea in an existing thesis, you have to cite it and clearly say that it already appeared there. – Vladimir F Героям слава Feb 14 '22 at 07:59
  • 1
    @VladimirF Finding out that someone else had the same idea before is not plagarism, nor to be honest are "ideas" worth much. The proof matters the notion that a proof exists does not. – Jack Aidley Feb 14 '22 at 09:36
  • @JackAidley What is plagiarism is taking someone else's ideas and saying they are yours. I never said that independent discovery was plagiarism. But finding an "unpublished" proof in a thesis and publishing it as your proof is 100% plagiarism. Publishing a proof that you re-discovered but staying silent on the issue that you did find out it has been discovered before is unethical in my books. It is not straight-out plagiarism, but it is deliberate priority usurping. – Vladimir F Героям слава Feb 14 '22 at 09:49
  • @VladimirF Then we are in agreement. – Jack Aidley Feb 14 '22 at 11:35
  • @VladimirF : But I specifically noted that the OP should mention the possible predecessor and thus acknowledge his claim. If the OP finds the thesis and the proof, they should acknowledge the proof. But until the OP finds the earlier proof, publishing the OP's proof is not plagiarism. See also my EDIT. – akhmeteli Feb 14 '22 at 13:30
3

Talk to your library. They usually have a service to scout things which are difficult to find (long distance lending), also some of the better libraries have a translation service; but usually a professor needs to sign off on that request.

lalala
  • 1,104
  • 6
  • 9
0

It’s a rough situation to be in. My honest opinion: Is it Unethical? Yes. Very unethical to ignore the Russian author’s claims. Think if the situation was reversed and, you were the first author.. Plus, you were able to find the truth. Count on others being capable of finding out. You don’t want to risk ruining your career for it. I agree with others, try your best to contact the author or Department via a library and on your own. If not possible to contact, make a note explaining that you found x and y information tried to get a hold of it via ABCD methods, but you couldn’t verify or review it in depth. For this reason you decided to continue your research project. Good luck!

M R
  • 1
  • 1