30

I submitted the preprint to arXiv and the manuscript to a Physical Review journal, and someone asked me to cite their article after seeing it from arXiv.

Their paper is a review. After reading that paper, I find that it has only a weak connection with our work.

My supervisor asks me to cite it. Should I refuse to cite?

uhoh
  • 3,152
  • 13
  • 31
lewton
  • 433
  • 4
  • 5
  • 10
    Maybe your finding is not certain, and the connection is stronger than you think? You can ask your supervisor for more details about how it should be discussed and cited. – usul Aug 21 '20 at 18:07
  • 8
    What reason did your supervisor give you to cite the paper? ("Because I said so" is not a reason.) – JeffE Aug 22 '20 at 22:41

5 Answers5

48

I can think of a number of reasons to cite a source:

  1. to acknowledge the work of others that you are building on (avoiding plagiarism)
  2. to give credit to the author of a particular phrase that you quote (avoiding plagiarism)
  3. to point readers to relevant previous literature
  4. to signal that your work is at the cutting edge of the field, and
  5. to situate your contribution in the context of related research
  6. to provide evidence for a claim

"Being asked" is not among them.

Unless you retrospectively find that one of the reasons above applies (esp. 3-5), politely decline to cite the paper in question. In particular, you should refuse to add a citation as a favor.

Addendum: I just saw your update to the question, stating that your supervisor asks you to cite the apparently irrelevant paper. This complicates the issue. I would ask them to explain why and where to add the citation. Perhaps they want you to slightly revise your paper, incorporating information from the new source, which would justify the citation. If not, it depends on your relationship whether you are in a position to push back.

Mutantoe
  • 111
  • 3
henning
  • 35,032
  • 10
  • 121
  • 151
  • 16
    If the supervisor comes with a reason that can be fit under the 6 items listed in this answer: cite it. It's not worth fighting against if the reason isn't absolute garbage, unfortunately. – Mast Aug 22 '20 at 12:15
  • 1
    You can always ask for clarification on how the paper in question is relevant for you to cite. – Eriks Klotins Aug 24 '20 at 12:47
42

I will just share my experiences (bad ones) to throw in some perspective.

I had submitted a paper once to a top conference in computer vision which was rejected for good reasons acceptable by me. But one of the reviewers, which I guessed was a top name in that area (also had a coursera course), asked me to cite 5 of his papers in the review. All were his papers.

A few years later, when I am doing my PhD, I got a paper accepted in a top conference. This time my supervisor sent me a list of 8-10 papers which I had to cite. Since there were constraints of space, he asked me to truncate a diagram and reduce the size of images to accommodate his list. Those papers had no relation whatsoever to my work. My supervisor is an IEEE fellow and has more than $1M in grants.

Citations improve H-index. H-index is crucial to get promotions and fundings. We had a very highly talented lecturer joining our university. He got 3 ICML papers accepted as first author (a top conference in machine learning) within a year, which is a fantastic achievement, but his grant application was rejected in favour of another lecturer who had 20 papers in mediocre conferences but a much better H-index. She works closely with my supervisor-- is a part of his nexus. The lecturer left my university and joined Google.

Academia is rotten with such practices. I regret that I agreed to add those citations - should have fought back. But then, fighting back is of little consequence when the system is fundamentally broken.

kosmos
  • 2,394
  • 8
  • 15
  • 9
    It sounds to me that someone needs to talk to the people issuing the grants. They have obviously adopted this procedure as a way of saving themselves effort. It is bad for academia as a whole and the problem needs to be rooted out. Maybe a prestigious general magazine would investigate and run the story. That would wake people up. Corruption in academia is topical at the moment - at least as far as a well-known individual is concerned. – chasly - supports Monica Aug 22 '20 at 09:26
  • 1
    @chasly-reinstateMonica The nexus is deep-rooted. The academicians form networks even within the gran allocation committee. They form networks of reviewers and citations and reap rewards. For example, my supervisor connected to the editor of topmost IEEE transactions ensured that a paper is reviewed in a week to ensure the promotion of a colleague in my department. One could write a book on how rotten this whole system is. – kosmos Aug 22 '20 at 10:14
  • 3
    Maybe these are the people to talk to. https://www.interpol.int/en/Crimes/Corruption
    • Corruption in academia, especially in technical subjects could cost billions by promoting substandard work.
     – chasly - supports Monica Aug 22 '20 at 10:20
  • 2
    @chasly-reinstateMonica Please define substandard work and also clarify who will make this assessment? The substandard work of one is perfectly fine for another. Yes there is rot in academia, and yes granting agencies - sometimes for fear of being sued - avoid making qualitative judgments but on balance the system works reasonably well. It is far from perfect and exceptions like the ones illustrated above happen all too frequently, but this kind of alliances always occur when too many people fight for decreasing resources. – ZeroTheHero Aug 22 '20 at 12:57
  • 1
    @ ZeroTheHero - I was merely echoing what I inferred from the following passage in the above answer, "We had a very highly talented lecturer joining our university. He got 3 ICML papers accepted (a top conference in machine learning) within a year, which is a fantastic achievement, but his grant application was rejected in favour of another lecturer who had 20 papers in mediocre conference" - I took mediocre to mean substandard. – chasly - supports Monica Aug 22 '20 at 13:04
  • 1
    @chasly-reinstateMonica The problem of intellectual circularity is one that cannot be so easily solved: a good researcher is one recognized as good by other good researchers. Thus the research landscape fragments into cliques (largely but not completely mutually exclusive) so that the work done by members of one clique is appreciated inside but not outside the clique. – ZeroTheHero Aug 22 '20 at 13:45
  • 1
    @chasly-reinstateMonica one needs to think long and hard before hitting a person in the career. Such action must not be done casually. Either you accuse somebody of academic misconduct, or you do not. There is no middle ground, and serious ramifications. – Scott Seidman Aug 22 '20 at 17:18
  • @chasly-reinstateMonica I agree with you and I also endorse the use of the term substandard. at--ZeroTheHero A good researcher recognized as good by other good researcher is fundamentally a poor model and not suited for modern research. My university clearly sees these malpractices but continues to endorse it because it also brings it the money. I wish I could elaborate more but for the character limit. – kosmos Aug 23 '20 at 01:10
  • 2
    I'm honestly surprised, and quite sad, to see this as the last answer so far. I expected the first answer to be this. I'm just in the initial phase of my PhD, and pretty much from the very beginning have been seeing such disappointing practices. As a PhD student, my supervisor is not very happy with the fact that unlike many others, I don't write papers flooded with self citations of the works of our group. Unfortunately I have academic siblings here whose papers have been cited only by other papers of their own (many times) and those refereed by our supervisor, and it's deeply disappointing. – nara Aug 23 '20 at 15:34
  • @nra I understand your predicament. It is a difficult trap to wade through. – kosmos Aug 23 '20 at 15:39
  • 4
    So, basically Goodhart's law. When a certain metric starts to get used to decide policy, it will soon stop being a useful metric (as people start gaming it). – vsz Aug 24 '20 at 06:01
  • 1
    Please don't generalize so much. You might find the case of Artemi Cerdà interesting. –  Aug 24 '20 at 07:08
  • 1
    That's a horrible story. I guess Academia has too many people who game the system, but we don't know, from your anecdote, how prevalent these opportunistic strategies are. What is your suggestion for OP, by the way? Play along and cite, or don't? – henning Aug 24 '20 at 08:15
  • @chasly-reinstateMonica citation boosting is an academic sin but neither illegal nor technically corruption. I don't think the police would be interested. – henning Aug 24 '20 at 08:18
  • 1
    @henning--reinstateMonica I played along as it did not hurt me directly. The paper was already accepted and I was asked to make alterations for camera ready. After a few months, I felt annoyed about something similar. I raised this issue with the student affairs dean. And nothing happened. Confrontation is of little use. One has to wait for the opportune moment. – kosmos Aug 24 '20 at 08:30
  • As terrible as this is, you make the same mistake by saying "He got 3 ICML papers accepted as first author". The prestige of a venue you publish in is an even worst measure than H-index. I'd take a well cited paper in a medicore venue over an uncited paper in a top venue every time. Neither is a substitute for directly examining the work itself though. – Ian Sudbery Aug 24 '20 at 11:22
  • @IanSudbery Getting a paper accepted in a prestigious venue is one thing but 3 in the same year facing immense competition from across the world is a statement. And any university should strive to retain such a talent. As a PhD student, I am the one who lost most due to this. Getting to work with good colleagues is probably the most important thing in my career. Though at a high level your point makes sense. – kosmos Aug 24 '20 at 11:29
  • Getting papers in three papers in a year is impressive, but it is the content of those papers that should be important, not the prestige of the venue. Otherwise this is just the same as judging people by the impact factors of the journals they publish in. Being accepted to a good conference/journal involves the judgement or 3, maybe 4 people. Being ctied 100 times involves the judgement of 100 people. Yes citations can be gamed - so can any system; so can journal/conference acceptance (we all know who our mates review for). – Ian Sudbery Aug 24 '20 at 11:39
  • @IanSudbery Well, it is much easier to game the system to get citations than game three separate sets of reviewers for the 3 papers, in a double-blind reviewing system in the best conference in the field. Citations, many a time is inflated because a work, though mediocre, is easy to replicate, especially in CS. Master's students from not so good universities do that for their projects to churn out another mediocre paper while citing the original one. I would give more weight to a verdict of reviewers of the top conference than citations from such sources. No doubt, conferences are also gamed. – kosmos Aug 24 '20 at 16:06
18

Should I refuse to cite?

Do nothing. You've read the paper and found it "has only a weak connection with ours," so there's no need to cite. You needn't refuse; just do nothing. If you must respond, then say something like:

Thank you for bringing your work to my attention. I have taken a look, but I don't see a strong connection and as such I believe it would be inappropriate to cite. If I've missed something, do let me know.

You can add a comma to the opening sentence and follow with I found it interesting, if appropriate.

user2768
  • 40,637
  • 9
  • 93
  • 144
  • 6
    I would limit myself to the first sentence. The rest is just starting a debate that the OP will not win with a person so aggressively seeking citations and a supervisor suggesting to go along with the request. – ZeroTheHero Aug 22 '20 at 13:00
12

Politics! Politics! Politics!

  1. Citations are now (for better or worse) the coin of the realm.
  2. It is inappropriate for someone to contact you asking for a citation. This person can rightly bring his work to your attention but asking to cite is a spoonful too much. You should be allowed to decide for yourself is this review is relevant or not to your work.
  3. You and your co-authors should decide if this additional work is worth citing. Those in favour should have an argument as to why this work was overlooked in the first place.
  4. If possible, hold firm. After all, you have (or should have) the final say as first author.
ZeroTheHero
  • 27,311
  • 2
  • 48
  • 111
  • "It is inappropriate for someone to contact you asking for a citation." I disagree. A citation to a review yes, but I would absolutely not hesitate to contact an author if they failed to cite my previous work if they had done the same thing. – Ian Sudbery Aug 24 '20 at 11:34
  • @ZeroTheHero I'd say it's fine to ask/tell if citing your work is neccessary for the "avoiding plagiarism" reasons (i.e. really should have been cited in the first place). Otherwise I agree with you. – DavidW Aug 24 '20 at 11:57
  • My comment contained an error. It should have read: "It is inappropriate for someone to contact you asking for a citation." I disagree. A citation to a review yes, but I would absolutely not hesitate to contact an author if they failed to cite my previous work if I had done the same thing. That is, if their study was a address the same question as mine, particularly if it was using the same methods, then it should cite my study. – Ian Sudbery Aug 24 '20 at 12:15
4

If you didn't read the review paper and went direct to the original paper(s), I can't see why you should cite a secondary source. If you did that exhaustively, you would include everyone's textbook about the subject plus the Wikipedia entry!

Alternatively, if there is pressure, look for one or or more specifics where the reviewer has thrown light where you would or could have missed it. Or maybe give a cursory acknowledgement by simply saying, "There have been a number of useful review papers in this area including X, Y and Z." and say no more.

chasly - supports Monica
  • 4,715
  • 12
  • 27
  • On the other hand, if only read the Wikipedia summary of a paper, do cite the Wikipedia article rather than the paper itself! https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4232290 – Nemo Aug 23 '20 at 08:01
  • This answer lacks some of the points that you made in your comments. – Solar Mike Aug 24 '20 at 06:32