Sometimes I find my self reviewing papers with authors who are not very capable English writers. I am a non-native speaker myself, and I know how difficult it can be to proof-read a paper in English, in particular if you don't have access to any native English speaker. If the scientific content of the paper is otherwise good, I therefore try to do what I can to correct grammar, poorly constructed sentences, spelling mistakes etc.

It often results in very long referee reports, which take ages for me to type. I was wondering if it would be proper form to simply scan a version annotated by hand, and attach it to the referee report? If you received such a thing, would you be grateful for the time spent, or would you consider it lazy that I did not type in all the suggestions in a proper report?

  • 11
    Correcting spelling and grammar is not a reviewer's responsibility. – Anonymous Physicist Jul 2 '20 at 11:00
  • 13
    I know. But if the content is good, then it is a shame if an otherwise sensible paper is illegible because of poor grammar and spelling. – nabla Jul 2 '20 at 11:02
  • 3
    Journals have, or should have, copyeditors who do that. – Anonymous Physicist Jul 2 '20 at 11:07
  • 17
    I agree that they should (otherwise - what are they paid for?), but unless you publish in a high profile journal, it is my experience that they don't. – nabla Jul 2 '20 at 11:09
  • 3
    If you receive the paper as a pdf, then it is possible to annotate the pdf directly so there is no confusion about what and where the corrections should be. – Solar Mike Jul 2 '20 at 12:44

I think it really comes down to the editor. You should first ask the editor of the journal if there is any issue and then think if your handwriting is clear enough. Hoewver, please keep in mind that scanned copies cannot be copy - pasted. You could return a pdf or Word version of the paper with annotated comments, either on the side or baloons (Adobe Acrobat and Adobe Reader allow for that and I find it very useful when finalising a draft). This sounds to me like a more convenient solution that a scanned copy.

I would like to stress that, despite your good intentions, the full responsibility of presentation and language falls on the authors, not the referee. A paper is not just about content but also sound presentation and communication of ideas.


A scanned referee report is fine, so long as it is legible. This requires good handwriting, patience, and a good scanner.

  • Completely agreed with this answer, except re “a good scanner”: these days, there are plenty of scanner apps available that produce good-quality “scans” from a phone camera. In iOS, for instance, you can use the default Notes app — start a note, click the “insert image” button, and it offers a scan as one of the options. (In my online teaching last semester, I encouraged students to submit handwritten homeworks scanned with apps, and the results were good enough quality that there was no extra friction during grading.) – PLL Jul 3 '20 at 14:22

Please don't feel obligated to proofread manuscripts as a part of review process. This is not a reviewer's responsibility. Journals are supposed to have proofreading services and/or to encourage authors to send their manuscripts for proofreading / editing. It is sufficient to comment on whether or not language check is required for the manuscript to be accepted.

Proofreading takes a lot of your time, which you are not compensated for. Unless you are also a professional proofreader, you can inadvertently introduce some occasional mistakes on top of correcting others. Moreover, if reviewers keep doing a proofreader's job, it will create an unjustified expectation on both journal's and author's side that proofreading is a reviewer's job. As a reviewer, I am frustrated with authors sending their manuscripts in a very poor state (not even spellchecked using a computer), which slows down the review process. It is important to remember that journals and authors benefit from publication directly, while reviewers do their service on a voluntary basis and for the good of the whole community.

  • 4
    Journal paid proofreaders, aren't, I think, domain experts. Some of their advice changes the meaning in unfortunate ways. – Buffy Jul 2 '20 at 12:57
  • @Buffy Ultimately, it's down to each journal whom they hire. If paid proofreaders introduce a small number of errors, they can be corrected by authors. If they introduce a lot of errors, authors will complain and journals will be forced to hire better ones. Reviewers should not involve themselves in this process. – Dmitry Savostyanov Jul 2 '20 at 13:00
  • 14
    I disagree that discouraging the OP is a good thing. They are a volunteer, they can go beyond the minimum if they like. I applaud it, actually. If you don't like the state of a manuscript sent to you for review, you are also free to reject it. – Buffy Jul 2 '20 at 13:07
  • 4
    @Buffy It depends. In academia, we often fail to impostor syndrome and spend enormous amount of personal time and effort doing things "beyond minimum". This is easily exploited by for-profit publishers and/or Universities, capitalising on unpaid academic overtime or voluntary review process. It is important to alert early career colleagues that some areas of work are actually someone else's responsibility (and they are paid to do it). – Dmitry Savostyanov Jul 2 '20 at 13:12
  • Just as @Buffy, I'm skeptical of the "doing too much will create an unjustified expectation" line of argument wherever it is being made. It leads to janteloven, which is hardly what we need more of in science. And you can prevent the expectations from rising by openly declaring that what you are doing is unusual and you're doing it because the paper is really worth it or because the author is new or something like this. (It also gives you a new reason to decline referee requests if you don't think the paper is worth it.) – darij grinberg Jul 4 '20 at 10:03

I'd like to add a concern: Adding scanned documents or annotated PDFs might unwillingly disclose your identity. It is good practice to keep reviewer anonymous for authors. If you are recognized by your handwriting, by PDF metadata, or just the practice of adding scans (which might be unique in your narrow sub-field), that would be a bad side effect.

  • The remark about PDF metadata applies to all referee reports submitted as PDF (e.g., most reports in my subject), not just scanned ones. Actually, scanned PDFs are less likely to contain identifying metadata. Good editors will probably scrub metadata from PDFs automatically via script (e.g., "qpdf --empty --pages [original PDF file] 1-z -- [output PDF file]"). As a referee you are not responsible for doing your darndest to get rid of metadata at all cost. – darij grinberg Jul 4 '20 at 9:58

You can do this as long as you get your meaning across. I can't speak for other editors, but I'm happy to receive any reviewer report at all (since I'm after all taking up your time), and won't begrudge you for not typing it out.

However, I'm skeptical it is the most efficient way. Today's copyeditors have a lot of shorthand symbols (example) that often makes writing more efficient than typing (this depends on the copyeditor; I know some who prefer to type anyway). Still, even if you are familiar with copyeditor symbols, there's no guarantee the authors are.

If you can't use the symbols, then the advantage of writing is you can clearly mark where the changes are and don't have to write "on page 3 second paragraph, change X to Y". But you can do that electronically too, e.g. here's how to do it in Adobe Acrobat. At that point the only advantage of writing is if you're dealing with heavy mathematical symbols that aren't easily written electronically. If you are making lots of these corrections and typing is inconvenient, then sure, although I suspect if these are necessary then the manuscript has more flaws than a proofreader can solve.

Bottom line: if it really is the best way for you, feel free to write & scan; however, there's a good chance there exist tools that make doing it electronically more efficient, to save nothing of the paper/trees saved.


This is not part of what you should be doing. People are not asked to review a scientific article for their proofreading abilities. You can and should certainly include in your review a remark that "there are many spelling and grammar mistakes that should be addressed before the paper is suitable for publication," but really you should only specifically address parts of the paper that are hard to understand or have typos.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.