This is an edited version of a recent Twitter thread, which will auto-delete after a year (the thread, not this post).
Responding to reviewer comments is my favorite part of academic writing. I have an excellent revision accept rate and am also a journal editor and reviewer, so I thought I’d share my reviewer response letter techniques, for what it’s worth. The following is what I do and advise and has never caused me difficulty.
First, I write the response to reviewer letter, and get input on it from my coauthors, before I revise the manuscript. The letter is my revision roadmap. Second, I always write a letter, not create a table. I think it is easier to read. As an editor and reviewer, I do not like a table layout for the revision letter, acknowledging that others may love it (I dunno).
Here’s how I write that letter:
1. Paraphrase (or Selective Quote), Part 1: I paraphrase or partially quote reviewer comments and acknowledge that I am doing so. This gives me the freedom to (a) cut out useless stuff and (b) focus the editor and reviewers on action items. I have reasons for doing this (below), and am in some good company in doing so. But I want to point out that others take or advise or have been advised a “full quote” approach (see this tweet and this one). Essentially what I end up doing the vast majority of the time is directly quoting the actionable parts of review comments and leaving out the throat clearing and stuff that doesn’t suggest what to do.
Why paraphrase or at least quote only actionable parts? A lot of what reviewers write you don’t want to remind them of and isn’t actionable anyway. E.g., “I had a hard time reading this paper.” Or, “I was initially troubled with the general approach.” Or, “While this topic doesn’t interest me, it may still be suitable for the journal.” Or, “[Insert very long paragraph of background with no action item for you.]”
Whatever. Not actionable. Not helpful to remind them of their prior, negative or irrelevant thoughts. Cut it.
Also, my main revision strategy is to make the letter super easy to read, but also self-contained — more on that later. Paraphrasing (or quoting only action items) goes a long way to tightening it up and reducing editor/reviewer (and coauthor) burden. It’s good for everyone, in my view. If you acknowledge you’ve done this up front, the editor can always ask you to put in full quotes. This has never, ever happened to me, but YMMV.
2. Paraphrase (or Quote), Part 2: I rarely directly quote passages of my revised manuscript in my letter. Why? It’s a pain to keep the quote in the letter the same as the revised paper as you go through revisions with coauthors. Instead, I paraphrase the changes in my letter. Again, others take the “full quote” approach here, and that can be just fine.
Another reason I rarely fully, directly quote my revision: there’s a bit of artistry here. If a reviewer makes a comment that is reasonable enough to warrant an edit, but I feel really can be done in a more terse manner or slightly differently than the reviewer suggested, or if there’s just a lot of background I’d like to add to the letter but not to the specific change in the manuscript, I can do that and then say, “We have incorporated these ideas [meaning the ones the reviewer just stated or I just described] by augmenting the language in the [say where].”
See what I did there? The reviewer gets the full load of context, and I reassure the reviewer that the ideas are incorporated, but the actual change may be modest. If I were to quote it in the letter, it could look insubstantial. I don’t want to give that impression.
But also, I always at least paraphrase or describe changes for every comment in the letter because I want the letter to be self-contained. I do not want the editor/reviewers to scrutinize the manuscript looking for how I changed it in response to a comment. Every second of additional scrutiny can raise new concerns. I don’t want that.
(The argument in favor of fully quoting all changes in the letter is that reviewers won’t need to look at the paper to see the exact wording, reducing unwanted scrutiny. Maybe. On the other hand, sometimes quoting the paper passage in the letter causes it to be out of context — and adding the full context could require a lot more writing and additional burden for the reviewer. If the quote is out of context, that may encourage a reviewer to scrutinize the paper. My feeling and experience, and again, others may differ, is that a good paraphrase that provides all the context efficiently is effective. Also, sometimes the change is super long and really can be tighter for the purposes of the letter, relieving a lot of reading. A middle way is to quote when it is helpful and brief and paraphrase when it isn’t.)
At the end of the day, what I want is for editors/reviewers to read the letter and think, “This looks good, I’ll sign off!” A good letter can do that. One that doesn’t say how the manuscript was changed in response to every comment (whether via quote or paraphrase) is effectively telling the editor/reviewer to go do more work. You don’t want to do that. It is annoying to them, which is not in your interest.
Further pro tip: I do not, in the letter, reference changes in the manuscript by page number. Why? The page numbers won’t always be the same in the package reviewers get and/or the number of the PDF can be different than the numbers you put on the pages (maybe you didn’t number the title page or something) — it’s confusing. Plus, stuff can shift as I revise and I don’t want to get the page number wrong. I just say where the change is in the structure (Intro, Methods, etc. or 3rd paragraph of Results or whatever). Simple. Fool proof.
3. Unique Numbers: I uniquely number every (actionable) reviewer comment. It doesn’t matter how they numbered them, if at all, or organized them, if at all. Some give three actions in one comment. I split them up and uniquely number. I use a system like E-1, E-2, … for the editor’s comments; 1-1, 1-2, 1-3 for reviewer 1; 2-1 2-2, 2-3 for reviewer 2, etc. (Exception: If a reviewer gave a long list of minor edits, I bundle them and just say I addressed them all.)
Why unique numbers? Many comments are duplicative. With unique numbers you can avoid duplicating yourself. Just refer back or forward by number. “See our response to 2-3” etc. Never type the same response twice. This is a burden for you, to keep them consistent, and a burden on the reader. You don’t want to burden the reader (editor/reviewer).
Another use of unique review comment numbers that I’ve employed is to add comments with the numbers in the revised manuscripts where I’ve addressed each one. That is, where I added text to address comment 2-3, I make a comment, “Response to 2-3.” This is useful to make sure I addressed every comment with some kind of change to the manuscript. When I’m iterating between letter and manuscript it helps me keep them synced up. Admittedly, I don’t always use the comment numbers this way or always insist that advisees do so. But it can be helpful. (I remove these comments when I’m done, before submitting, though there really isn’t anything bad that is likely to happen if they were left in. Perhaps reviewers would like them.)
4. Fully Responsive: I am responsive to every comment. What is a response? This is an art. It’s everything I can think of within reason and within scope of the paper and that isn’t likely to raise new objections. Mostly it’s doing substantially more than just saying, “No we won’t.” (I call that giving a reviewer the “stiff arm.” It doesn’t feel good to them, so I don’t do it because I want them to like me and my paper.) That doesn’t mean do stupid stuff. At the least I explain why I considered the suggestion and decided against it, or refer the reviewer elsewhere for support for my view, etc.
5. Be Nice: I’m often angry at reviewers initially. That’s OK! I get through it by letting it out in a mean letter first. I write what I really think of their ideas or how they clearly didn’t read the paper. But, that’s just for me. I then revise it taking the more appropriate and professional high road. I express gratitude for every comment (with different language): “Thank you.” “Good idea.” “We concur.” “We apologize for the oversight.” “What a thoughtful suggestion.” “We appreciate the consideration you’ve given this.” “That initially occurred to us too. [Good for when you need to explain why it won’t work.]” “That’s a reasonable idea.” “Yours is an innovative suggestion.” …
Bottom line: Make the letter as concise, yet thorough, as possible, while ensuring it is self-contained, easy to follow, and kind/generous. That’s about it. (Even if you feel you need to fully or partially exactly quote the review comments, you can do all the rest and it will likely help.)
