A revision is a second deadline, and the response letter is where many of them go wrong. Comments get skipped, responses drift from what the manuscript now says, and the editor cannot find where a change was made.
A good response letter is mostly structure. This post gives you a LaTeX template that enforces the structure, a workflow for the manuscript changes, and a short checklist. I compiled the template to make sure it works.
Journals differ in what they require, so check your venue's instructions for format and length. Treat this as a base to adapt.
The template
It numbers each reviewer's comments automatically, quotes the comment in a shaded box, and gives you a response block under each one. Changes you made in the manuscript are marked in blue so the editor can scan for them.
\documentclass[11pt]{article}
\usepackage[margin=1in]{geometry}
\usepackage{xcolor}
\usepackage{tcolorbox}
\newcounter{rc}
\newcommand{\reviewer}[1]{\section*{Reviewer #1}\setcounter{rc}{0}}
\newcommand{\rcomment}[1]{%
\stepcounter{rc}%
\begin{tcolorbox}[colback=gray!8, colframe=gray!50, title=Comment \arabic{rc}]#1\end{tcolorbox}}
\newenvironment{response}{\par\noindent\textbf{Response.}\ }{\par\bigskip}
\newcommand{\changed}[1]{{\color{blue}#1}}
\begin{document}
\section*{Response to Reviewers}
We thank the reviewers. Changes in the manuscript are marked in \changed{blue}.
\reviewer{1}
\rcomment{The evaluation uses only three benchmarks.}
\begin{response}
We added a fourth benchmark (Section 5.2, Table 3).
\changed{The result is consistent with the other three.}
\end{response}
\rcomment{Please state the learning rate.}
\begin{response}
The learning rate is now given in Section 4.1.
\end{response}
\reviewer{2}
\rcomment{The related work omits recent baselines.}
\begin{response}
We agree and added three baselines to Section 2.
\end{response}
\end{document}
It compiles on a standard TeX Live install with pdflatex. Each reviewer gets a heading, comments restart at 1 under each reviewer, and each response sits directly under the comment it answers.
The workflow
1. Copy every reviewer comment verbatim. Paste them in as \rcomment{...}, one per point. If a reviewer's paragraph holds three points, split it into three comments so each gets its own answer.
2. Make the manuscript changes first, then write the responses. A response should describe what the manuscript says now. Writing responses before the edits is how letters end up describing changes that never happened.
3. Point to a location every time. "We added this in Section 5.2, Table 3" is easy to check. "We addressed this" is not.
4. Mark changes in the manuscript. Define a macro in the manuscript preamble, and wrap new or changed text in it:
\newif\ifshowrev \showrevtrue % \showrevfalse for the final version
\newcommand{\rev}[1]{\ifshowrev{\color{blue}#1}\else #1\fi}
Unlike a comment macro, this keeps the text in both modes and only changes its color. For a full marked-up file, latexdiff between the submitted and revised versions is the usual route. I covered both in yesterday's track-changes post.
5. Cross-check at the end. Count the comments in the letter against the reviewers' reports. Every one should have a response.
Tone rules that save rounds
- Thank once, briefly. Spend your words on the answers.
- Answer the point, not the reviewer. Stay on the content.
- Disagree with evidence. If you did not make a change, say why, and point to a result, a reference or a limitation you now state in the paper.
- Never ignore a comment. If you cannot address one, say so and explain.
- Keep it short where you can. A one-line fix needs a one-line response.
A checklist for the day you submit
- [ ] Every reviewer comment appears in the letter, verbatim and numbered
- [ ] Every response gives a section, page or table reference
- [ ] The references match the revised manuscript (page numbers shift)
- [ ] Changed text is marked in the manuscript or in a diff file
- [ ] The letter compiles cleanly, with no
?? - [ ] Co-authors have read the final PDF of both the letter and the paper
Working on it together
A response letter is a shared document with a tight deadline, which is where live collaboration helps most: one co-author handles each reviewer, and everyone sees the letter and the manuscript change together. I work on LetX, a collaborative LaTeX editor with unlimited collaborators on its free plan, though the free plan limits compiles to 40 a day, which is worth knowing during a revision crunch. The template above works in any editor.
Next in this series: one shared bibliography file for co-authors, without merge pain. This post is part of the series on collaborative research writing.
What is the best response letter structure you have seen? Tell me in the comments.
Top comments (0)