77

In a recent review, one reviewer said we must rework the paper in light of two other papers. One paper was a highly valuable suggestion, and we've enthusiastically taken it on board. However, the second paper seems to be an ultra-specific paper (one example among thousands) and not applicable to our work.

I'm thinking of writing something like:

There are many methods for XYZ (and an anonymous reviewer of this paper seemed fond of QRST [26]), and a survey was given in [15]. In this paper, we take the approach ABC for our specific problem.

(Here, I leave off "...since QRST does not actually apply to our problem".)

In this example, we (fairly politely) highlight that we're citing [26] at the reviewer's request. But I'm a bit worried it would be perceived as a passive-aggressive slight.

Question: Is it okay to mention we're citing an article only because a reviewer told us to?

I'm also considering two alternatives:

Alternative 1:

There are many methods for XYZ, such as QRST [26], LMNOP [4], and QRSTUV [5]; a survey is given in [15]. In this paper, we take the approach ABC.

But this seems like I'm adding even more virtually irrelevant citations.

Alternative 2:

Acknowledgment
A helpful anonymous reviewer suggested citing [26].

But while accurate, it might be considered provocative (and rude) and get the paper rejected.

This question is related to: How to deal with an unreasonable reviewer asking to cite irrelevant articles? But in our case, the reviewer is not being unreasonable, maybe just fond of this particular paper. And it's just one additional citation that's being requested.


Update: Thanks for everyone's help! To let you know what's going on (and maybe this will help someone who winds up in the same situation):

(TL;DR: The reviewer's suggested paper is related to an "in-between" paper which is related to our work; we feel it's not off-topic to cite it while discussing this in-between paper. It wasn't as off-topic as originally thought.)

  1. We generally agreed with the advice in the answers (don't cite it). We also found the editor's email specifically said not to cite irrelevant papers suggested by reviewers. But we dug deeper to ensure we're not perceived as being dismissive. What happened next:

  2. Upon more careful inspection, the reviewer described it as an "example". Eventually, we concluded that the reviewer was not giving us a list of papers to cite, but to generally say "look at papers like these".

  3. This turned out to be a gold mine: it gave many papers which we could use to put our work into a broader context; none of our paper's predecessors do this. In fact, we found a paper which almost identically states the motivation for our paper, but in a broader context (which empowers our justification for working on this topic).

  4. We looked into the suggested paper's references and papers citing that paper. We found another paper is a kind of "in-between" paper (related to both our paper, and the paper that was mentioned by the reviewer). We now discuss this in-between paper in detail, and explain how our method deviates from it. In the context of discussing this in-between paper, we feel it's not irrelevant to quickly cite the paper that was mentioned by the reviewer, without giving it undue weight.

(PS. In this particular instance, I feel this is not the reviewer's paper: other papers by the same authors (e.g. the in-between paper) are more natural to cite.)

9
  • 12
    How about citing the review paper, but not the specialized paper? Seems more useful for a reader, and one would hope that a reasonable reviewer would accept it, especially if the review describes the method used by the specialized paper.
    – Anyon
    Sep 17, 2018 at 11:55
  • 18
    How do you think readers of your paper will think about such remarks?
    – Mast
    Sep 17, 2018 at 12:44
  • 67
    When a seemingly random paper is proposed by a reviewer, there's a high chance that the reviewer is one its authors. Sep 17, 2018 at 13:00
  • 35
    "seemed fond" --> "pointed out". Maybe the anonymous reviewer has a point and you're still learning, so no need to be snarky. IMHO. :-)
    – PatrickT
    Sep 17, 2018 at 14:28
  • 4
    Relevant idea: janlo.de/wp/2010/04/04/scientific-citation-markup
    – Dirk
    Sep 18, 2018 at 4:36

6 Answers 6

183

As it turns out, not every comment or suggestion forwarded by a reviewer needs to be necessarily accommodated in your manuscript. Hence, as you seem to have a convincing argument against citing the suggested paper in your manuscript, in your rebuttal letter state that you have considered the suggested comment but you didn't find that paper to merit citation. (However, compose your argument politely, thanking the reviewer for the suggestion.)

2
  • 40
    Agree completely, reviewer comments are suggestions for improvement, not ironclad mandates. If the reviewer suggested that you to change your methodology to something you didn't agree with, or run a statistical test that you didn't think applied, including it anyway with the footnote "because the reviewer told me to" would be a poor argument indeed. Sep 17, 2018 at 15:41
  • 8
    And this may be particularly effective since you ended up (enthusiastically!) using the first paper the reviewer suggested
    – darthbith
    Sep 17, 2018 at 23:13
155

No, it's completely inappropriate to cite something while making snarky remarks about how you don't think you should. There is no upside whatsoever: it makes you look petty and obnoxious, and it looks like you're disparaging the authors of the paper being cited, even though you're actually disparaging the referee, which is hardly better. Either cite the paper respectfully or don't cite it at all.

If the suggested paper is not relevant, then you shouldn't cite it at all and your response to the referee should justify that decision. If the referee insists and the editor concurs, then just cite the paper in as neutral a way as possible.

Given that there seem to be multiple possible approaches to solve your problem, perhaps the best solution would be to write a short paragraph addressing them. Something along the lines of your "alternative 1" but with explanation of why you chose the method you did, instead of one of the others. Wouldn't that improve the paper?

37

To complement the other answers, most of which seem to take as granted the assumption that the reviewer's request is indeed pointless and made by mistake (or possibly to pad their own citation count, as some answers allude to), let me take the charitable view towards the reviewer and assume that they actually think, correctly or not, that the method QRST described in the paper they suggested you cite is appropriate for your problem, and possibly even better suited for it than the method ABC that you're currently using.

If so, it's plausible that this particular reviewer may not be the only reader who will think so.

Perhaps method QRST is considered by many to be "the state of the art" for solving similar problems, while method ABC is older and often considered to be less efficient and/or less accurate. Thus, other readers might plausibly also look at your paper and think "why did they use ABC, when QRST would surely have worked better?" In the worst case, they might assume that you're only using ABC because your knowledge of the field is years or decades out of date and you're not aware of any newer and better methods, making the overall quality of your research suspect, or even that any unexpected results you've obtained are probably just errors caused by the inaccuracy of the ABC method.

If that's indeed the case (or if you have even the slightest reason to suspect that it might be the case for some readers), you ought to demonstrate that you're indeed aware of the current state of the art in your field, and explain why method ABC is, in fact, better suited for your problem than QRST or other more recent methods. In some cases, this might indeed mean citing the suggested QRST reference just to show that you're aware of it and have considered that method before choosing not to use it.

Depending on how you choose to phrase this, the result might end up looking somewhat similar to your first alternative suggestion, e.g.:

There are many methods for XYZ, of which a recent survey is given in [15]. Unfortunately, some of the more advanced methods, such as QRST [26], LMNOP [4], and QRSTUV [5], are not well suited to this particular problem, because <insert explanation why>. In this paper, we instead take the classic approach ABC [2], which does not suffer from the limitations described above, and can still provide accurate results in a reasonable time, provided that [...].


More generally, the point I'm trying to argue here is that review comments may be mistaken, but they're rarely if ever completely without cause. If a reviewer thinks there is an issue with some part of your paper, then it's likely that other readers will also think so. That means that you should do something to address the perceived issue, even if it's only to clarify your writing so that others are less likely to misunderstand it the same way as the reviewer did.

(Of course, there are occasional exceptions to this rule. Sometimes reviewers are poorly chosen, and sometimes you may just have to accept that a particular reviewer does not actually belong to the target audience of your paper, or that their suggestions might have an inappropriate ulterior motive. But in general, it's still best to start from an assumption of competence and good faith, until and unless there is clear and unescapable evidence to the contrary.)

2
  • 5
    +1 This. Very true that if a reviewer thinks that a paper should be cited, other readers will think so as well and might wonder why it's being ignored in the paper. Cite within context.
    – Gimelist
    Sep 18, 2018 at 7:15
  • 2
    I agree with this anwer, though the OP stating QRST “seems to be an ultra-specific paper (one example among thousands)” makes it unlikely that it would be “considered by many to be ‘the state of the art’ ”. Sep 18, 2018 at 11:50
29

There is no reason or regulation that you have to include the citation at all in a case like this. The reviewer's comments are just that: comments. The editor may take them as requirements or not, but the paper is still yours and you should use your judgement about what to include.

But to avoid problems that may arise with the editor, include your reasons in a note to him/her. We didn't include x because y. The editor may send it back, but I doubt it. The editor may also send it out for additional review but likely to a different set of reviewers. Even if it is reviewed again by the same person, as long as they don't have a particular motive for the suggestion, they should evaluate your paper as a whole. But your paper contains nothing provocative if you just omit the citation.

Let the editor work for you and explain why you haven't taken a particular suggestion among many that you did find helpful.

24

There is a tension between two motivations, which seem implicit in your question and approach:

  1. You want to accommodate the request, although it is perhaps unreasonable, to make sure your paper gets published.
  2. You want to include only what's relevant and reasonable, so as not to compromise the paper's quality for the sake of "review politics".

You propose to resolve the tension by including the requested citation but distancing yourself from it at the same time. However, this comes across as either rude and petty or as perfunctory. Chances are, this won't get the job done: The reviewer may be less inclined to recommend acceptance of your paper if your tone is rude or his request is served perfunctorily; and you don't really stand your ground either.

My suggestion would be to treat the superfluous reference as a mere suggestion and to not include it. I would point out the reasons in the accompanying letter to the editor, in which you also detail the other changes you made to the draft. This helps both of your goals while being upfront: It maintains your integrity and the quality of the paper, but it also improves your chance of acceptance, since in the last instance, the editor makes the decision, and your argument for not including the reference sounds persuasive.

3

ALWAYS put the reader first. What would you like to read as a reader? Comment "method QRST [16] is irrelevant here but reviewer wanted it included" or nothing at all? Out of your suggestions, only 2nd one is viable even though you don't like it.

If such statement doesn't already rub the editor the wrong way and he demands removal of the statement or simply rejects your paper, reviewer is likely to be offended too - he will try hard to find reasons to reject the paper or at least demand this statement removal. (Well, unless he is only after that citation, but I doubt it - as you said, he recommended a valuable paper too, so he seems to know the field).

Finally, considering there are multiple approaches to solve the problem - enough of them to warrant a review article - there is absolutely no harm in adding yet another reference to yet another method solving the problem. After all, you surely have a section of introduction dedicated to other approaches and showing why you are better. It will be just another citation in the section mainly used to pad the reference list.

You must log in to answer this question.

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