Writing good responses to reviewer comments?

In our most recent "how can we help you?" thread, a reader asks:

I'd love to see some advice on how to write a good response to reviewer comments. In particular, I'm wondering whether it's appropriate to explain a response to Reviewer B, for example, in part by explaining what Reviewer A requested. For example, Reviewer B suggests changes to section 1. Reviewer A suggests a change which indicates that scrapping section 1 altogether was the way to go. In responding to Reviewer B, would it ever be appropriate to explain that section 1 was scrapped because of both reviewers criticisms? Or is one supposed to restrict their responses only to what each reviewer has suggested and make no reference to the other reviewer's comments. Thanks!

Good question! One reader submitted the following reply:

Write the response for the editor - so: "because Ref1 suggested changes X and Y in Sect 1, and Ref2 suggested that Sect 1 was unnecessary, I have decided eliminate Sect 1. ..."

I wonder whether this is enough, though. If you scrap section 1 of a paper because of Reviewer A suggested it but you don't explain why to Reviewer B, couldn't Reviewer B end up puzzled? What if they don't understand why you've scrapped it and wish that you'd kept it? For these reasons, I'm inclined to think it may be best to explain changes like these to both reviewers, not just to the editor (since the editor might not forward those comments to either reviewer).

What do you all think?

Comments

[Jamin Asay]

When I submit revised versions of papers, I reproduce the reviewer reports in full, responding point by point. That way each reviewer sees what the other reviewer(s) said, and what my responses are. And the editors and referees can see that everything has been addressed without having to dig up the original reports.

As a referee, it's useful to see what the other referees said, and if I have access to those reports, I will generally share with the editor whether I think the author has adequately responded to the other referees' points. (And if I think the other referees' points are off base, I will share that, too.)

[collaboration not negotiation]

Do editors sometimes not forward response-to-reviewer letters to reviewers? That never occurred to me! Seems bad.

In response letters, I do my best to make it clear that I have listened carefully to all of the suggestions, and done my best to use them to improve the paper. The core of the letter usually takes the form of a list: here is a problem with the initial submission that I needed to solve; here is how I solved it. Solving it might mean rejecting one reviewer's suggestion, but if you can explain a good reason why you're rejecting it, most reviewers will accept that. In OP's case, I would think "Reviewers identified X problem with section 1; here is why I think deleting section 1 is the best way to address the problem" would work for both reviewers.

Instead of treating the revision like individual negotiations with isolated reviewers, this approach highlights that everyone involved in an R&R is collaborating on improving the manuscript. Sometimes reviewers in hardcore gatekeeper mode do need to be reminded of that, but in my experience so far it is very rare for a reviewer to throw a tantrum if I explain why I have opted to solve a problem they helped identify in a way other than what they suggested.

[whatever]

In my experience, doing this well is really important to whether the paper gets passed R&R. I wonder, though, if this is just how it looks from my (the author's) perspective. What do editors think?

In any case, I write my reply to reviewers in a more essayistic format. I give an introduction which states what I took to be the driving criticism in each report (ideally, there will be overlap), and then briefly say what I've done to address it. Then I usually divide the reply into a couple sections. One dealing with the main criticisms from each reviewer, and then a second section detailing other minor fixes. I always bold reviewer names so they know when I am talking to them (e.g. R1 and R2 in bold), and I also make reference to page numbers in the document where the fixes occur in case they want to check. But yeah, in general, I write these in a more essay, almost a letter-writing fashion. And I don't often set my replies up as if I am going to move point by point. What do referees like to see?

[academic migrant]

Would just to like to highlight that sometimes the jobs and likelihood of quitting academics are in the hands of reviewers, so there is an obvious power relation regarding revisions. I basically do anything and everything reviewers say, even if I strongly believes that sometimes this would compromise the quality of papers. I hope reviewers can also keep this in mind (as I try my best to do when I review): reviewers never have good evidence that the authors are rationally persuaded by the comments and suggestions.

[Recently tenured]

Here is what has worked well for me (I had around 7 R&Rs as a junior philosopher, and successfully converted all of them to acceptances):

I copy and paste all referees' comments into a single document and then respond inline (usually paragraph by paragraph, or point by point if the referee made a numbered list). I generally assume that referees will pretty much just read my responses to their own comments, so when how I responded to Reviewer #1's comment is affected by Reviewer #2's comment and vice versa, I inform each referee about that in my response to each of their comments. If you are responding to a referee's comment in a way that might surprise them, you should absolutely tell them why (whether it's because of something another referee said, or something you realized yourself in the editing process, etc.)

As for more general advice, as a junior person who really needed publications I tried to just make the referees feel smart and like they did a great job refereeing the paper, without being obvious that I was trying to do this. This doesn't mean that I took *every* suggestion, but I did respond thoughtfully to every suggestion in my remarks to the referees, and offered an explanation for why I didn't take the suggestion that wouldn't make the referee feel like I though it was a dumb suggestion.

As a referee myself, what annoys me most is when an author doesn't make a change I suggested and doesn't say anything about why. I will basically never recommend Accept when an author does that, because I only include suggestions in my referee report that are (in my judgment) necessary. So I'm open to being convinced that it's not necessary, but I'm not open to just being ignored. This sometimes happens when people write their replies to referees in more of the essay format others have mentioned.

[Louis]

I think Jamin Asay and Recently tenured are giving good advice. Three things I would add:

1. It is often a good idea to break the resubmission letter into two parts, with the first part specifying major revisions occasioned by the guidance of referees and editors. This is where you would explain why you deleted an entire section. The second part then goes as Recently Tenured suggests, with specific responses. This provides a convenient reference for the referees.

2. It is a very good idea to mark new or revised text in a revised submission. (Some journals instruct you to do so.) I do this using color: altered text is typeset in another color. (It's kind to choose a non-obnoxious color.) Obviously, deletions are not so marked, so you need to note them specifically in the resubmission letter.

3. It is extremely important in the detailed responses to referee comments to be *very specific* about where to find the revised text that responds to any given comment. Don't make the referees hunt or guess.