Responding to Reviewers Without Conceding What Matters

A revise-and-resubmit is the most consequential document most researchers write badly. The paper is already good enough to be considered; what happens next depends largely on the response letter.

Two failures are common. Conceding everything, which produces a worse paper and frequently still gets rejected. And defending everything, which reads as unwillingness to engage. For a concrete operational example of how a workplace tool approaches a related coordination problem, see this guide.

Norms vary by field and by journal. Some expect a detailed point-by-point letter; some expect a short covering note. Check the journal's guidance and, where possible, an example from a colleague.

What the editor is reading it for

Not whether you agreed with everyone.

Whether the substantive concerns were addressed. Addressed can mean changed, or it can mean answered — the editor is deciding whether the objection still stands.

Whether you understood the objections. A response that answers a different question than the one asked suggests the revision will not converge.

Whether the paper improved. Editors want to publish good papers; a revision that is worse than the original is a problem for them.

Whether working with you will be manageable. This is unstated and it is real.

The structure that works

A short covering letter to the editor. What changed at a high level, and any point where you have disagreed with a reviewer and why. Flag disagreements here explicitly — an editor who discovers a disagreement buried on page eleven is less receptive than one who was told about it up front.

Then a point-by-point document. Every comment quoted, in order, each with your response, each with what changed and where.

Quote the comment. Do not paraphrase. Reviewers reading the response want to find their own words.

Say where the change is. Section, page, or the text itself. A reviewer should not have to search the manuscript to check.

Show the changed text where it is short. Quoting the new paragraph in the letter is faster for everyone than a cross-reference.

Answer every comment, including the ones you consider minor. A skipped comment reads as evasion, whatever the reason.

Disagreeing successfully

The part that decides the outcome.

You may disagree. Reviewers are sometimes wrong, sometimes read carelessly, and sometimes want a different paper. Conceding a point you believe is wrong produces a paper you cannot defend, and reviewers frequently notice the incoherence in the next round.

How to do it:

Acknowledge the concern before answering it. "The reviewer is right that this design cannot rule out X in general" — then the argument for why it does not threaten this particular result.

Give the reasoning, not the assertion. "We disagree" is not a response. The evidence, the logic, or the citation is.

Offer something. Even where you hold the substantive position, adding a paragraph acknowledging the limitation, a robustness check, or a clarification shows engagement. A disagreement plus a concession lands very differently from a flat refusal.

Do not disagree about everything. A letter that resists every point reads as defensive regardless of whether each individual response is correct. If you find yourself disagreeing throughout, reread — either the reviewers misunderstood the paper systematically, which is a writing problem, or you are being stubborn.

And never be rude about a reviewer. It will be read by them, and by the editor, and it costs you.

The concessions that damage the paper

Some changes make a paper worse, and accepting them because they were requested is the error.

Adding analysis you do not believe in. A robustness check you consider meaningless, included to satisfy a reviewer, sits in your paper permanently and you will be asked about it.

Hedging the finding into meaninglessness. Each requested qualification is reasonable; twelve of them produce a paper that claims nothing.

Expanding the literature review to include the reviewer's work where it is not relevant. This request is common and it is legitimate to decline politely.

Changing the question to the one the reviewer would have asked. The most serious version. If the paper becomes a different paper, you have lost the thing you were trying to publish.

Adding length. Most revisions grow. Ask what can come out for each thing that goes in.

Reviewers who contradict each other

Common, and it needs handling explicitly rather than by picking one.

Name it in the covering letter. "Reviewer 1 asks for X, Reviewer 2 for the opposite. We have done Y because..." The editor is the person who resolves this, and they can only do so if they know.

Do not silently satisfy one and ignore the other.

Ask the editor before revising, if the contradiction is fundamental. Most editors would rather answer a short email than referee a revision that went the wrong way.

Reading the comments before responding

Wait before writing. A first reading of a critical report produces a defensive letter. Two days changes what you write, reliably.

Separate the report into three: things that would change the conclusion, things that would strengthen the paper, and things that are presentational. Respond to all of them, and allocate your effort accordingly.

Look for the comment underneath the comment. A reviewer asking for an unnecessary robustness check is frequently signalling that they do not believe the identification. Answering the literal request may not address the concern, and the concern is what matters.

Where two reviewers raise the same thing in different words, that is the paper's real weakness, and it needs the substantive response rather than a partial one.

See reading a rejection for the related problem.

Practical points

Number everything, matching the reviewers' own numbering where they used one.

Use a consistent format — comment, response, change — so it can be read quickly.

Keep a version with changes marked, if the journal accepts one.

Do not exceed the resubmission deadline without asking. A short email requesting an extension is normal and is granted routinely; silence is not.

Have a co-author or colleague read the letter before you send it, specifically for tone. You will not be able to judge your own tone on a document you wrote while irritated.

Thank the reviewers once, at the start, briefly. Not on every point — that reads as performance.

When the decision is rejection with encouragement to resubmit

Different from a revise-and-resubmit and frequently misread as a rejection.

It usually means the editor is interested and the required work is substantial. Treat it as an invitation with a low probability rather than as a refusal.

Ask the editor whether a resubmission would be considered, if the letter is ambiguous. Editors answer this question.

The short version

A response letter is an argument. The editor is deciding whether the objections still stand.

Answer every comment, quote it, and say where the change is.

Disagree where you should, with reasoning, and pair it with a concession.

Wait two days before writing, and have someone else read it for tone.

And guard the paper. Some requested changes make it worse, and accepting them because they were requested is how a paper becomes something you cannot defend.

For an independent authoritative reference, consult the EQUATOR Network for reporting guidance for research writing.