In my field (chemistry), the authors are known to the peer reviewers. A critical point is vague about a manuscript I am reviewing; it can be rejected or accepted based on this point.

Can I directly contact the authors to ask for clarification or I must conduct all correspondence via the editor? I know that the normal procedure is to contact the editor, but I thought it is faster to directly contact the authors, as it may need multiple communications.

  • 1
    Also, you stumbled upon the major flaw of the classical peer-review system in times of electronic communication.
    – Wrzlprmft
    Commented Dec 11, 2015 at 7:57
  • 2
    @Wrzlprmft I don't see a "major flaw of the classical peer-review system" in the mild annoyance of having to write an email to a different address.
    – Cape Code
    Commented Dec 11, 2015 at 8:26
  • 1
    @CapeCode: That’s not the problem. The problem is that communication via the editor takes a considerable overhead of time, if it happens beyond the “official” reviews at all. Nowadays, when an undelayed and anonymised communication beween authors and referees is possible, a lot of time is wasted by not doing it. As far as I know, only few journals try to move away from the classical submission–review–resubmission–review–… system, e.g., Frontiers.
    – Wrzlprmft
    Commented Dec 11, 2015 at 9:18
  • 3
    @Wrzlprmft you may think it'd be better if reviewers and authors could have a free and unfettered electronic dialogue, but honestly I'm not so sure. I can see that scenario leading to a gradual decrease in authors' discipline to submit polished, publication-ready papers, and instead putting more and more burden on reviewers to engage in dialogue with the authors, giving them advice about improving their papers (since it's so easy and quick), etc. Ultimately that could create more work for reviewers, make it harder to get good reviewers, with the final papers being no better than before. ...
    – Dan Romik
    Commented Dec 11, 2015 at 9:57
  • 2
    ... From that point of view, making it more difficult (but not impossible) for authors and reviewers to communicate, as with the current "classical submission–review–resubmission–review–… system", may actually ultimately result in higher quality papers.
    – Dan Romik
    Commented Dec 11, 2015 at 10:01

2 Answers 2


Can I directly contact the authors to ask for clarification or I must conduct all correspondence via the editor?

No, you really shouldn't contact the authors directly. I can see two options regarding the possibility of doing that: you can do it either from your normal email account (or in a similar way that discloses your identity, e.g. a phone call); or you can do it from a throwaway account that preserves your anonymity. Both of those approaches seem very wrong to me, but the first is much worse than the second.

In the first approach, you would be spoiling your anonymity as a reviewer. That would make your entire review useless. The whole point of anonymous peer review is that you can be honest in your review without having to worry about the authors knowing who you are (and without others having to worry if you were influenced by the knowledge that the authors know your identity). You may think that this is a problem only if you intend to write bad things about the paper, but it is just as much of a problem if you want to write good things, since the journal and its readers want to know that you wrote those good things out of an honest motivation and not to personally ingratiate yourself with the authors.

In the second approach, the problem of giving up your anonymity may not exist, but it's still unethical to go behind the back of the editor. The editor is responsible for the quality and the integrity of the peer review process, and needs to be informed of any communication between you and the authors. You may trust your own judgment that you would make sensible use of the information you would be getting from the authors, but the editor might disagree, and it is not your place to appropriate his or her jurisdiction over the review process. As an example, perhaps the journal or editor take the approach that the paper must be judged solely based on its submitted form with no room for clarifying questions. Or, perhaps you would reach a sensible decision to accept the paper based on additional details the authors emailed you, but you would then neglect to require them to add those details to the paper, leaving the paper flawed from the point of view of the journal and its readers - and so on; one can imagine many ways in which a reviewer, no matter how sensible and well-intentioned, can screw up this process in some subtle way due to not having the perspective and experience (and knowledge of the journal's policies) that the editor has. That is why the editor needs to be involved.

I know that the normal procedure is to contact the editor, but I thought it is faster to directly contact the authors, as it may need multiple communications.

First of all, why should you care which way is faster? You are the reviewer. It is not your problem if the process goes slowly, it is the authors' problem, and they should have thought of including the missing details when they wrote the paper. Second of all, if multiple communications are needed for the paper to have a chance at being acceptable, then quite possibly you shouldn't be bothering to request clarification in the first place. To me this state of affairs suggests a pretty strong likelihood that the paper is simply not acceptable in its present form, so you should either flat out reject it or make a "reject and resubmit" (or "major revision required") decision, and write back with a report explaining what details need to be added or clarified to make a more in-depth evaluation of the paper possible. It would then be the authors' responsibility to comply and submit a better version if they want the paper accepted.

To summarize, it is understandable (and noble) that you want to be efficient and help the authors improve their paper. However, as a reviewer you are not required to go out of your way to help the authors with an elaborate back and forth exchange of information and critique, and you most certainly should not do so without the involvement of the editor. If you really want to have a detailed communication with the authors, I don't see anything wrong with that as long as it is done through the editor and with their approval.

  • +1 for "First of all, why should you care which way is faster? You are the reviewer". This points out that there is something likely influencing the motivations of the OP...E.g., knowing the authors personally. This is why I favor double blind peer review.
    – neuronet
    Commented Dec 11, 2015 at 15:09
  • 2
    @neuronet This might obviously be the case, but it might just as easily be the case that he's a decent human being and the paper is in general good and interesting. To be perfectly honest saying "Why should you care which way is faster?" strikes me as incredibly callous. Even if the paper is not good someone probably wrote it and they might have some possible feelings.
    – DRF
    Commented Dec 11, 2015 at 15:13
  • @DRF:fair enough: it could be as simple as the OP being new to review process (indeed, that's actually pretty clear).The key is that whatever helpful comment the reviewer has to say to authors can and should be said through the editor. Anything else smacks of special treatment, and the reviewer needs to remain impartial. It's the apparent compromising of impartiality of the reviewer that was my worry. Going outside the accepted communication routes to speed along the paper does not seem impartial. But your point is well-taken: there are many reasons they could want to do this.
    – neuronet
    Commented Dec 11, 2015 at 15:33
  • @neuronet Thank you. I didn't mean to imply that it is a good idea to contact the authors directly. It certainly casts doubt on the process of review and is generally a bad idea. I also agree that double blind peer review would is a good idea but unfortunately in many disciplines and particularly for papers by senior researchers it's just not realistic. You can either have the reviewer know enough to be able to conduct a thorough review or you can have the authors be anonymous, but you can't have both due to specialization.
    – DRF
    Commented Dec 11, 2015 at 15:51
  • 1
    "although if you want to, I don't see anything wrong with that" seems a little out of place when you just spent several paragraphs explaining exactly what you do see as wrong with that (well, sort of). I agree with the main points of your answer, I just found that phrase a little unexpected.
    – David Z
    Commented Dec 11, 2015 at 16:01

Best done via the editor. Either channel all correspondence via the editor - even address your concerns to the editor, as in, could you find out for me etc. or ask the editor: "I am OK with my identity being revealed to the authors. Please do so and ask on my behalf if I may approach them directly with my queries."

Of courses the review process itself does not have to be confrontational and is eminently suited for just such a query. "I think very much hinges on the meaning of the following passage: [...] It is not clear to me whether the flootnums referred to here are chizbotz or actually chuzbitz, and I believe this is a vital point because..."

If something along those lines still does not elicit a satisfactory response, then maybe you should take the heavy decision.

You must log in to answer this question.

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