I wonder how reviewers would react if authors of a submitted paper refuse to cite paywalled papers. Is it a valid reason to refuse citing some papers?

I am mostly interested in the field of computer science, and English-speaking venues.

  • 39
    Well it certainly won't go down well if you're submitting to a paywall'd journal ;) Commented Feb 8, 2016 at 19:56
  • 98
    IMO by failing to cite someone's work, I think you'd be doing the authors a disservice rather than the publisher.
    – CKM
    Commented Feb 8, 2016 at 21:37
  • 15
    In my own field (physics), there is almost no such thing anymore as a paper that is only in paper form, or a paper that is only available digitally behind a paywall. Essentially all new papers go on arxiv.org. If your field is not like this, then is the problem (a) that certain authors are too clueless to post preprints, (b) that your field doesn't have the equivalent of arxiv.org, or (c) that the journal has a contract that prohibits posting preprints?
    – user1482
    Commented Feb 9, 2016 at 1:04
  • 54
    This may be a novel idea to the young folks, but the traditional way one could go and read a journal article was to get out of your chair, go to the library, and get the journal. If it's not there, you can request the library borrow it from another that does. Even if you're not affiliated with a local university, members of the public can usually get library access for either free or close to free. What you're paying for is convenience more than access - the paywall is rather less daunting than you're making it out to be. Protesting inconvenience is not likely going to get you very far.
    – J...
    Commented Feb 9, 2016 at 10:53
  • 18
    @BenCrowell ... In my own field (mathematics) we frequently refer to papers written before the invention of the Internet. Some of them have been digitized by now, but far from all.
    – GEdgar
    Commented Feb 9, 2016 at 12:56

13 Answers 13


If I was reviewing a paper, and the authors failed to cite important literature, I'd recommend rejection until the authors provided correct citations. It's the authors' responsibility to provide appropriate references.

I assume that many other reviewers would feel the same.

  • 52
    Indeed. While I understand the OP's argument, you can't just ignore relevant literature.. Commented Feb 8, 2016 at 20:21
  • 1
    I don't think this answers the question because the two don't have to be mutually exclusive. If the authors refuse to cite paywalled papers but still cite appropriate sources (however unlikely this is in your field), should the refs reject the paper on principle? I think this is what OP is asking. Commented Feb 9, 2016 at 15:22
  • 21
    @user1717828 I'm not sure I understand what you're saying. Are you talking about a situation where an author makes some statement citing paper X (non-paywalled), but the referee believes the author should have used paper Y (paywalled) to justify that statement? If you're citing just to justify a statement, that could be fine, but citations are often used to give credit, and in that case there is no flexibility in which paper you need to cite.
    – David Z
    Commented Feb 9, 2016 at 16:18

I wonder how reviewers would react if authors of a submitted paper refuse to cite paywalled papers. Is it a valid reason to refuse citing some papers?

Certainly not. There are a few areas where this could be OK; for example, if you are citing expository material for background (rather than to assign credit), then you can choose whichever sources you feel are best.

However, in many cases you have a scholarly obligation to cite papers, for example to give credit to people whose work you are building on, and there are no acceptable grounds for refusing to do so. It's a serious form of academic misconduct, even if it is done for idealistic reasons.

If I ran across an author who refused to make necessary citations, I would be extremely displeased, and I would recommend that the paper not be published until the citations were included. I would not fully trust that person's judgment in the future, and I would be suspicious that other papers might be missing important citations.

Instead of omitting citations, you could add some brief commentary about the lack of open access. (Reviewers or readers might dislike it, but it's in no way academic misconduct.) You should be very careful with that, since you could really offend an author who has made the paper available, just not where you looked. For example, it might be in an institutional repository. If you want to avoid giving offense but still encourage open access and help readers, you could give suitable arXiv or repository links to each paper for which you can find them.

  • 16
    you could really offend an author who has made the paper available - and those who haven't as well.
    – Dan Romik
    Commented Feb 8, 2016 at 20:19
  • 11
    At what point is an author ethically justified in refusing to go through the hassle and expense of acquiring a paper that is not made accessible by the authors?
    – DQdlM
    Commented Feb 9, 2016 at 1:44
  • 14
    @Kenny this is not an ethics question. Authors are free to acquire or not acquire anything they wish/don't wish to, and make various other decisions. Eventually, as a result of those decisions the author will either produce high quality, useful research that other scientists will appreciate, or he/she will produce low quality, useless research. There is nothing unethical about producing useless research (assuming it's not done with fraudulent intent), but it's not good for your career. (I guess one can argue that sabotaging your own career is unethical, but I would just call it incompetence.)
    – Dan Romik
    Commented Feb 9, 2016 at 2:20
  • 23
    @KennyPeanuts: The scholarly obligation is to cite, not to acquire. If you are aware of a paper, you can't pretend you don't know about it just because you don't want to track down a copy or can't. Of course citing a paper you haven't seen is awkward and can look bad. In practice it's almost always worth the effort of tracking it down. But the difference is that omitting the paper entirely is unethical, while citing it with an explanation that sources say it contains something but you do not have access to it is unusual but ethical. Commented Feb 9, 2016 at 5:06
  • 24
    (Incidentally, I think authors do have a scholarly obligation to make their work as widely available as reasonably possible. But much of the world apparently disagrees, as judged by their publishing behavior.) Commented Feb 9, 2016 at 5:08

If you would like to boycott paywalled journals, go right ahead - that is your right. As a reviewer I would reject your paper for not citing relevant sources. As a reader of your paper and/or author of a paper you didn't cite, I would be severely antagonized as well.

The bottom line is, your ideological battles should not be waged on the backs of honest readers and authors who dedicate their lives to producing and disseminating good science. But if you want to commit career suicide, be my guest - no one will stop you.

Edit: the saying "Be the change that you wish to see in the world" also comes to mind. In that vein, if you don't like paywalled journals, the honorable course of action would be to simply not publish in them yourself. This would be vastly superior from a moral, ethical, and philosophical point of view to waging some kind of take-no-prisoners, collateral-damage-be-damned nuclear warfare against them, which is effectively what your question is proposing.

  • Said better than my comment said it, comment withdrawn and upvoted.
    – keshlam
    Commented Feb 9, 2016 at 3:38
  • 1
    Regarding your edit, I don't think it's a good idea to encourage early-career academics to "boycott" the reputable journals of their field because they read on the internet that they have a bad karma.
    – Cape Code
    Commented Feb 9, 2016 at 9:03
  • 2
    @CapeCode thanks for your comment, but to clarify, I wasn't (and am not) encouraging anyone to boycott anyone else - quite the contrary I should think, considering my use of terms like "career suicide". In my edit I was expressing an opinion that if OP was going to boycott paywalled journals, then one particular kind of boycott is more legitimate than another. This is not the same as encouraging a boycott though.
    – Dan Romik
    Commented Feb 9, 2016 at 15:31
  • It's up to the reader to decide whether they want to pay out for a paper that is behind a paywall or not (they might decide against it out of poverty, lack of interest, or because they are boycotting such papers; that's entirely up to them). It's not up to the author to decide whether the reader should be informed about the existence of such a paper or not. On the other hand that implies the reader has no right to complain if a paper is only available behind a paywall.
    – gnasher729
    Commented Feb 15, 2016 at 10:50

I am in agreement with a number of the prior answers that state that it is an author's responsibility to ensure that appropriate literature is cited.

It is often the case, however, that multiple possible citations are reasonable. For example:

  • A single extended work can often be effectively cited from any of a family of related publications, e.g., the original idea, a refined and well-formalized version, a practical demonstration, a review paper with a good discussion, etc. Typically one needs to cite just one or a few out of the set.

  • Some works are cited to give examples of a large class of related work, rather than for that individual work per se.

In cases like these, where there are many reasonable alternatives in citation, it seems entirely reasonable to me for an author to choose to favor more open publications over less open publications. It will not allow ideological purity, as there are of course those many cases where you do need to cite something that is not open access. Still, rewarding those who choose openness may be a good compromise position to promote openness without compromising other key ethical principles of science.

  • 2
    I disagree. You should cite the *most relevant" previous work(s). Doing otherwise is to do your readers a disservice. You can cite an alternative, open access one, too (and I'd appreciate that, if I can't get the paywalled one).
    – vonbrand
    Commented Feb 9, 2016 at 0:20
  • 4
    @vonbrand And how do you define "most relevant"? In my experience there is often a lot of flexibility and subjectivity in the choice.
    – jakebeal
    Commented Feb 9, 2016 at 1:05
  • 1
    @jakebeal I suggest the following as an operational definition of "most relevant": "a paper I'd be most inclined to cite if I did not know or care whether the journal it was published in was paywalled". Subjective or not, this is well-defined, and I agree with vonbrand that not citing the papers that in your opinion as an author are most relevant is doing a disservice to your readers. Whether you can somehow hide behind subjectivity to get away with it and not get caught is beside the point, since it seems pretty clear that the OP wants to wage his war on paywalled journals publicly.
    – Dan Romik
    Commented Feb 9, 2016 at 2:08
  • 14
    @DanRomik That seems a bit silly. If I am citing a lemma which is well known and where the original source isn't known, or if I am citing expository material which is in many places, then I should choose the source most useful to my readers. That means things like well written, in a major journal, using modern vocabulary, written in English (if I am publishing in English) and so forth. I have no problem adding "publicly available, on the arXiv or in an open journal" to that list; it is another thing which will make the citation more useful to my readers. Commented Feb 9, 2016 at 16:09
  • 5
    @DavidSpeyer good point, and I see my operational definition is flawed. I guess the correct definition of most relevant is that it best serves the interests of the readers, and you've convinced me that accessibility of the paper is a legitimate factor to consider - as long as this is done from the point of view of considering what serves your actual readers (in contrast to the OP's idea of trying to fight a crusade that, while well-intentioned and having the goal of ultimately helping the community of all readers, will do a disservice to the specific readers of his specific papers).
    – Dan Romik
    Commented Feb 9, 2016 at 16:47

Many of the answers have talked about fairness of citing other papers and that is hugely important. You need to cite others for both academic honesty and to give credit to the work that you've built off of. But another (possibly more important) reason of including related works is proof that you're knowledgeable in the area.

If you're publishing on a topic and skip half the relevant papers, how are readers and reviewers able to know that you've done your due diligence and understand the area? It's an important indication that you actually know what you're talking about and are knowledgeable enough for someone to take you seriously.


Instead of refusing to cite paywalled articles, what about contacting their authors, encouraging them to upload their articles to an open repository? John Dove has proposed this idea recently.

He imagines that the (open access) publisher itself could do this work on behalf of the author, checking each reference against a database of open access articles and emailing the author if it was not found.


"I am mostly interested in the field of computer science, and English-speaking venues."

You do not really talk about journal standards and reviewer quality. In an epoch of "publish or perish" perils, there is a lot of computer science conferences and journals, with different levels. Some won't care about the quality of your paper, as long as you pay. This is a first option.

If the intent is to address publications with higher standards, and one has a strong religion against paywalls, computer science is a world where you have options: indeed, many CS people publish online, preprints, extended versions, or open access journnals. Then, if you really want to play, you can cite the online etc. versions, and put the "paywall ref" as a note:

Alan Turing, "On Computable Numbers, with an Application to the Entscheidungsproblem", http://turing.sci/comp-numb.pdf (also appeared in Proceedings of the London Mathematical Society s2-42, 230–265, 1937)

Thus, you show you know the relevant literature to the reviewers. Some may like or not that you do not put the references in the standard way, but you might have more troubles with the editors or the publisher. But it might be ok with open access CS journals, because the times are changing.

  • 3
    What is the most important, with respect to the reviewers, is to show and prove that you know what you are writing about. I disagree. What is the most important is to serve the readers you are ostensibly trying to inform by writing your paper. Intentionally hiding relevant information from the readers is unacceptable from a reviewer's point of view and would be immediate cause for rejection, regardless of whether the author has shown himself to be knowledgeable about the subject he was writing about or not.
    – Dan Romik
    Commented Feb 8, 2016 at 22:16
  • You do not disagree. The question was about reviewers. Publishing means making something public, and thus should serve the readers. Reviewers and publishers are filters. I do not see (yet) where I promote hiding relevant information. Commented Feb 8, 2016 at 22:35
  • Additional comments: a handful lot of the papers published are not read (apart from the authors and the reviewers). A lot of papers are published only because academia requires publishing. It would be nice if only worthy papers were written and published, but only in a beautiful world Commented Feb 8, 2016 at 22:40
  • 1
    LaurentDuval I do disagree. Specifically I disagree with your statement about what's most important to reviewers (and since I have a lot of experience as a reviewer, I dare say my opinion about what's important to reviewers counts for something). Are you seriously saying that you know better than me if I disagree or not? Maybe it's an English language thing, but you should know your statement that I don't disagree comes across as pretty bizarre.
    – Dan Romik
    Commented Feb 8, 2016 at 22:59
  • I do respect your opinion and expertise. I do apologize if my formulation sounded offending. I believe that we were not talking at the same level of the review process. Let me rephrase: as a reviewer, one thing important to me is that the authors prove they know the relevant litterature, since science should be a cumulative process. I am picky about accurate references, but that's a personal think. Some editing could be done at the editor or publisher side. I removed the sentence because I do not think my opinion is a general rule Commented Feb 9, 2016 at 18:47

Franck, you're forgetting that not very long ago most sources were only available in physical form anyway, so paywalling-or-not was irrelevant. Whether an article is behind a paywall or not is irrelevant w.r.t. citing it - in my opinion. If you used it, you need to cite it (choice of versions notwithstanding).

Of course, if you wrote your own paper and it really doesn't relate to paywalled work, then you don't have to cite any of it - but remember it might be relevant and you didn't know it.

Now, it would be interesting if you could make the argument "I did not cite paper X because I did not have the money to get the copy and read it" - but that's unlikely to be true unless you're, say, an undergrad from a poor country or something. And even then it probably won't be accepted.


It's a reason, it might even be a good moral reason, but how do you intend to communicate it at first? I'd hate to have to go through a round of reject/resub. in order to let the reviewers know that I didn't cite a critical and well-known reference on purpose. Also, it hardly seems fair not to cite an article that is relatively ancient which was written, submitted, and published when authors had no or few alternatives for submission and this issue was not so hot. The Elsevier math journal boycott seems to have petered out, do you expect that another boycott is going to pick up steam?


If an author studies certain papers and they contribute to the work, such that the citations would actually show where certain ideas came from, then by omitting them, the author behaving like a hypocrite at best and plagiarist at worst.

If there is a paywalled paper that you didn't actually read, you shouldn't cite it, because then you're only citing it to stuff your list of citations. If someone were to interrogate you about that cited paper, you would be immediately exposed as not knowing anything about it beyond the superficial summary of its results given in the non-paywalled abstract.

Useless citations that exist just for the sake of inflating the list of citations (to make the paper appear more important and more thoroughly researched) should be trimmed, regardless of whether they are paywalled.

If you did read the paywalled paper, then it behooves your readers to know that you read that paper, and to make up their own minds whether they want to chase the citation through the paywall. You're not actually yourself conforming to the ideology of eschewing paywalled papers, so don't foist that ideology on your readers.


If, by some marvelous coincidence, all the most relevant papers are non-paywalled it would be mostly okay. You absolutely must cite any paper whose work you're building on, regardless of the paywall. If you use any information/data/ideas from a paper, you have to cite it. To do otherwise is academic dishonesty. I wouldn't say that you can "refuse" to cite any paper for any reason. If it so happens that your paper doesn't need to cite any paywalled papers, then you're in the clear. This happens a lot in physics, where the most relevant papers are often free. If you have read a paywalled paper and it helped you in any way with the paper you're working on: cite it.

On a side note, if you're at a university, corporation, or research facility, they usually provide free access to a majority of the reputable journals out there.


It would be fine not to cite paywalled articles. You just have to make sure that for readers to understand the context of your work the paywalled articles are all completely and totally irrelevant.

If on the other hand the work is relevant and you don't cite it, then I (and most reviewers) would advise the editor to reject your article.

Translation - if being paywalled is the basis for not citing the paper then most reviewers would recommend rejection.

  • Note - since I've gotten a surprising number of downvotes, I'm curious for anyone else downvoting - do you disagree with my answer that you have to cite all relevant work whether paywalled or not? Or do you think I've somehow failed to answer the question?
    – Joel
    Commented Feb 11, 2016 at 22:56
  • I did not downvote, but downvoting signifies "This answer is not useful." If people find this answer not useful for whatever reason, downvoting is warranted. If an "answer" fails to answer the question, then flagging is warranted instead. Commented Feb 11, 2016 at 23:06
  • @MichaelHoffman - I recognize what downvoting is supposed to signify (I've got 5000ish reputation across stack exchange). But I recognize that people might find this "not useful" because they disagree with my answer or they feel I didn't answer the question. Whether/how I should edit the answer is very different depending on why they feel it is not useful. So I am asking anyone else downvoting if they would explain which it is.
    – Joel
    Commented Feb 11, 2016 at 23:17
  • 1
    I mostly only get downvoted when I don't give knee-jerk support to charging authors for writing. Commented Feb 25, 2016 at 2:17

Are you aware that the decision of the international community not to cite research done on Jewish prisoners by Nazis was considered highly controversial? There was no question that the knowledge had come by unethical means, but the question was whether it is right to flush away any human knowledge? Given the arguments over that, it's pretty unlikely that an academic community would back you shunning those who prefer the traditional strategy of charging readers over the new strategy of charging the authors.

You must log in to answer this question.

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