Resubmitting rejected manuscripts without changes?

I stumbled across a couple of social media conversations recently on the topic of authors resubmitting unchanged articles to a new journal immediately after rejection with referee comments. The first conversation asked whether it is permissible for authors to do this (viz. is it irresponsible or disrespectful to reviewers to simply ignore reviewer feedback?). The second conversation raised the issue from the perspective of referees and editors. Should a referee accept a request to review a paper at a second journal if the paper is unchanged? If so, is it permissible for them to simply resubmit the same referee report they did when they reviewed the paper previously? And, should editors be irritated with authors when this happens?

These are all interesting questions, and there was a great deal of disagreement about them in the conversations I was following. Still, I thought it might be worth discussing more openly here. What do you all think? Allow me to briefly share some thoughts, based on my own experience.

From an author's perspective, I think it all depends on the kind of feedback one gets. On the one hand, some referee feedback can be so vague that it's unclear why one should revise the paper. In some cases, all one might get from a reviewer is several sentences or a short paragraph stating, in very general terms, that they found the paper unpersuasive--without any detailed explanation. In other cases, referee feedback can be inconsistent: one or two reviewers can argue in their review that the paper accomplishes P successfully and is worthy of publication, whereas another reviewer judges things differently. In these cases, it is once again unclear to me why the author should have to revise the paper--unless the author judges for themselves that the reviewer's isolated concern(s) have merit. One other reason for this is that, if a paper has gone to multiple journals, different reviewers can have so many different concerns that if you revise the paper each time to address every single possible concern, the paper can become so long and unwieldy as to go far beyond journal length requirements. While reviewers having so many concerns can of course be grounds for "giving up" on a paper, they don't necessarily suggest one should do that--as some of the most famous articles and books in the discipline are ones that wide litanies of concerns have been raised against. The clearest case where an author should revise before resubmission, I think, is the case where multiple reviewers agree on a paper's demerits. If multiple reviewers think the paper is problematic in respect X, it would simply be unwise for the author not to do anything about it. Indeed, this seems to me the only clear case where not revising really would be "wasting everyone's time" (i.e. new reviewers, editors, etc.). Am I wrong?

From a referee's perspective, I certainly understand the temptation to accept and assignment to review an unchanged paper and resubmit an unchanged review. However, some of my experiences as a reviewer suggest to me that this can be a mistake. Sometimes, journals have a policy of forwarding the editor's decision email that goes out to an author. I think this is a good practice, as it can enable referees to see what other referees thought of the paper they reviewed. My experience is that in some cases, the contents of these emails can be surprising, indicating just how much different referees disagree about the merits of a given paper. In some cases, one referee might recommend outright acceptance, a second referee a revise-and-resubmit, and a third referee outright rejection. In cases like this, I'm inclined to think that the same reviewer accepting the paper as an assignment and submitting the same review as before gives their perspective as review undue bargaining power in the peer-review process. Part of what anonymized peer-review is for, as I've always understood it, is to counteract bias. Given that we all have epistemic biases, holding a paper's fate hostage to the perspective of one reviewer seems dangerous and unfair to me--unless, again, it is one of those cases where a reviewer knows that other reviewers felt the same way previously. 

In short, my general feeling is that the best way to think about these cases may be in terms of fair, collective deliberation. Authors should feel free to resubmit a paper without changes when an isolated reviewer has a concern that the author is unpersuaded by, but not when multiple reviewers raise the same issues. Similarly, referees should feel free to accept an assignment to review an unchanged paper, as well as resubmit an unchanged review, when they have reasons to believe that other reviewers judged the paper similarly--but should beware accepting such an assignment in other cases, on the grounds that it might hold the paper's fate hostage to their own reviewerly biases.

But again, these are just some of my thoughts. What are yours?

Comments

[RecentPhD]

Clearly if your wonderful paper gets rejected, the reviewer was a moron who didn't understand it.

Unfortunate as it may be, the world is full of morons, and it is reasonable to assume that many of them will fail to understand your wonderful paper in more or less the same way and for more or less the same reasons. So, you might as well revise so they don't have an excuse.

More seriously: I know that the thought of revising a paper that you've toiled on night after night can make you want to scream. I've been there. But most likely your reviewer won't be the only person with that complaint, so you might as well try to respond to it.

[Al]

On the referee side: There are lots of idiosyncratic reasons we have for liking or not liking a paper. Barring a serious error that sinks a paper, I now decline to referee for the second journal when I get a paper I got from another journal that looks to be unchanged (I do take a relatively close look). I think a paper should get another try with another reader - I'm most of us would wish that for ourselves when we get a frustrating rejection. But again, if there was a serious flaw that i felt very confident about I'd agree and point that out again (that hasn't happened come up yet).

[Amanda]

I don't know about that recent Phd. I have found that plenty of reviewers have idiosyncratic complaints. Not only that, but I think that no matter what complaints I fix, another reviewer will find different complaints. I think it is very rare these days to get a straight acceptance without a RandR. So I struggle knowing when to change a paper in response to a reviewer. It feels like a waste of time if a different reviewer is going to bring up other problems anyway. But idk, I guess I'm really not sure what to do in these circumstances.

[Skef]

To add a wrinkle ...

Suppose, as many people think, that there are norms that call for generally addressing feedback from one's reviewers before resubmitting. In light of those norms, what are one's responsibilities with respect to a desk rejection?

And, if it is acceptable to resubmit a desk-rejected paper, how does that not call into question the norm for generally addressing feedback, given that the (implicit?) message of a desk-rejection is that the paper is not of a quality that calls for review? Wouldn't a typical desk-rejected paper need more work than a typical reviewer-rejected paper? (Or does the lack of specific information allow one to always say "maybe it wasn't a good fit", even though that's a less common reason offered by editors when speaking generally.)

[been around a bit]

Provided there are comments from the referees, I think there is a prima facie obligation to address them in some form or other before an author submits the paper to another journal. As someone above already noted, if one person (mis-)read the paper in some way, others will as well, and the author needs to prevent this misreading. I can only think of one case in my publications where an attempt to address a reviewer's concerns made the paper worse. Generally, papers are improved, and I have published in good journals in my area (PhilSci, Synthese, SHPS, Nous, APQ).
As a referee I refuse to review a paper I have reviewed for another journal. I do this for two reasons. First, I do not want to reread the same (lousy!) paper. Second, I think the author deserves a fair shot, and if I was biased in anyway they deserve to have someone else read it. I have been asked now many times to review a paper I have reviewed before. Sometimes a crafty author will change the title and abstract, and I will agree to review it. But as soon as I realize it is the same paper I reviewed before I alert the editor and recuse myself.

[Pendaran]

Most of my papers had to be submitted a few times, some many times, before receiving an R&R and eventual acceptance. About half of the time comments would be useful, if only to help address areas easily misunderstood. If you're resubmitting a paper over and over making no changes, I think that is unwise. On the other hand, about half the time referees really are incompetent. Sometimes they don't even bother to read the paper. So, you shouldn't be afraid to ignore stupid comments. Anyway, that's my advice. It seemed to work for me.

[Untenured Ethics Professor]

In reply to Skef: Editors give desk rejections for a lot of reasons. Though the perceived quality of a paper is a common reason, lack of fit can be a reason (as you point out). So can a judgment that the journal has had too many recent articles on a given topic.

If you get an unexplained desk rejection, and you haven't yet gotten much feedback from others about your paper, receiving the rejection might be a good occasion to seek feedback. But if you've already received feedback (preferably including feedback from more-senior people, e.g. advisers or attendees at a conference not only for grad students), and the feedback you've gotten leads you believe that the paper is publication-ready, you are under no obligation to revise further before sending the paper out again.

[Sam Duncan]

I agree with Pendaran. The quality of comments varies wildly, and more fundamentally I think the spirit that referees approach the work varies too. Referees and the their comments can be bad in two basic ways. First they can be lazy and sloppy. On several occasions, I've gotten comments that were nothing more than three or four lines of unclear text rife with typos. In those cases it's not been clear to me what justification the referee had for rejecting much less how I could improve the paper in light of their comments. Second, they can be partisan to the extent that they simply are not going to give anyone who disagrees with them a fair hearing. If the comments amount to "Only an idiot could possibly believe this paper's claim especially in light of the brilliant work of X showing just how moronic it is" then I feel very little need to make any changes, especially if I've read X and been completely unimpressed. Of course in a lot of cases they can be both lazy and partisan.
If the referee put some level of effort into the comments and made any effort to give your work a fair hearing then I think you're obligated to at least seriously consider revisions and should have a good reason if you don't do make revisions in light of what they said (it's almost certainly in your self-interest to do so too). But it's shocking how often these basic conditions aren't met.
One other point I'd make is that given word limits it might very well be impossible to implement every good suggestion from referees. On more than one occasion I've had referees say, "Well this other related debate is important so it'd be great if the author could say something about it" and I've agreed. But the problem is that addressing that debate in any adequate way would take four or five pages and I was a page or two from the word limit set by the journal. In one case I actually even cut exactly the discussion they wanted so as to just squeak in under the limit.

[Recent Grad]

I have a question that's somewhat related to the topic of resubmitting rejected manuscripts without making changes. A little background: I'm a recent grad. I've had around 5 reviews for a couple of different papers (including one R&R and rejections for the other submissions). Every single time I've only had one referee report. You can imagine that only having one report complicates the question of whether to do revisions before resubmitting. Referees can be so divergent I find myself very hesitant to make changes based on only one report--unless the criticisms happen to strike me as significant and obviously correct.

I'm wondering whether only having one reviewer decide the fate of a paper is a regular occurrence or have I just been unlucky? I've also heard different explanations for why I'm only getting one report. Some tell me that means that the editor is having difficulty finding people to review the paper. Others tell me that the editor has two reports but is only forwarding me the report that recommends rejection since that's the one he/she agrees with. Does anyone have any insight about this?

[Amanda]

I think it could be any of those things. The possibilities, in order of my estimated likelihood:

1.The report you got was the first of two reviewers to turn it in. Because one rejection is enough for total rejection, there was no need to wait on the second report. (I have had emails regarding a paper I agreed to review. The email said I no longer had to do the review because the other reviewer's report was "sufficient".)

2.The editor could only find one reviewer

3. The second reviewer did not provide comments.

4. The second reviewer recommended acceptance and the editor doesn't want to deal with complaints.

In my experience, it is not uncommon to get one report. It happens maybe 1/4 to 1/3 of the time. (Of the times I actually get reports as opposed to desk rejections). So you have probably been a little unlucky so far.