Reviewer says this is already known

When a reviewer says your finding is already known, read the papers they cited before conceding anything — the figures, not the abstracts. The single most common outcome of that exercise is discovering that the prior work showed something adjacent to your result rather than your result, and that the difference between them is precisely what your paper contributes.

Reviewers work from memory of a literature they know well, and memory compresses. A study that showed an association in one cell line becomes, in recollection, a study that established the mechanism.

What the objection actually claims

“Already known” is three separate objections wearing one sentence, and they take three different responses. The first is a priority claim: a named paper reports your result, so yours is redundant. The second is an incrementality claim: your result follows so directly from published work that establishing it was not worth a paper — the substance of an incremental objection. The third is a framing claim: your introduction asserts a gap the reviewer does not believe exists, which is an objection to your positioning rather than to your data.

Read the review to work out which one you have. A priority claim names a paper. An incrementality claim names a field. A framing claim quotes your own sentence back at you, usually a “remains unknown” or “for the first time” in the introduction. Answering a framing objection with a defence of your data, or a priority objection with an argument about importance, is the most common way this exchange goes wrong, and it burns a revision round.

Where the comment names no paper at all, treat it as the third kind until the reviewer supplies one.

What to check in the cited work

The system. Was it the same organism, tissue, cell type, disease context? A finding in one system is not the same finding in another, and whether the difference matters is a scientific question you can answer. An effect in HEK293 cells overexpressing a tagged construct and an effect in primary cells carrying the endogenous allele are two results, and a phenotype in a constitutive whole-body knockout is not the phenotype in a tissue-restricted conditional — see knockdown and knockout differ.

The claim’s strength. Did they show correlation where you show causation? Necessity where you show sufficiency? Association where you show mechanism? These are different results, and the distinction is often invisible in an abstract. Reviewers use the verb “showed” for every rung of that ladder; the source paper’s figures usually do not. Place the prior claim on the same ladder your own reviewers apply to you — correlative rather than causal, mechanism not shown — and state in your response where each paper sits on it.

The conditions. Same concentration range, timescale, genetic background, disease stage? Write the numbers next to each other. A prior study at 10 µM for 24 hours and yours at 100 nM for 14 days are not the same experiment, and a reader looking at both figure legends cannot conclude otherwise.

The evidence. Was the prior claim supported by a figure, or asserted in a discussion? Discussion speculation cited later as established fact is common enough to be worth checking specifically. Check the replicate structure and the n behind the panel as well: a claim resting on one experiment with three technical wells is not the same evidentiary object as one resting on three independent biological replicates run on separate days, and that difference is itself a reviewable defect — see experiment not replicated.

If any of these differ materially, you have a response — and it is a factual one rather than an argument about importance.

How a claim becomes known without ever being shown

Citation practice manufactures certainty, and there is a measured account of how it happens. Steven Greenberg reconstructed the complete citation network for a single claim — that the β amyloid found in Alzheimer’s disease is produced in the skeletal muscle of patients with inclusion body myositis — across 242 papers and 675 citations addressing that belief, and published the analysis in the BMJ (2009;339:b2680). The network contained 220,553 citation paths supporting the claim.

Greenberg named three mechanisms, and each is something you can look for in the papers a reviewer sends. Citation bias: papers that refute the claim are cited far less often than papers that support it, so the supporting literature looks unanimous when it is not. Amplification: reviews and models that contain no primary data of their own are cited as though they did, and become the sources everyone quotes. Invention: a hypothesis stated as a hypothesis in one paper is cited by the next as an established finding, and the qualifier never comes back. Greenberg reported the same distortions in grant applications submitted to the National Institutes of Health.

The consequence for your response letter is narrow and useful. When a reviewer writes that something is known, ask which paper measured it, and trace each citation back to a primary experiment rather than to the review that asserts it. A claim whose entire support consists of two reviews and one discussion section is not established prior work, and saying so with the citation trail attached is a factual argument rather than a defensive one.

Worked example: tracing one “this was already shown”

Take a reviewer comment reading: “The dependence of Y on X was already shown by Smith et al., so figure 3 adds little.” Working the trace takes about an hour and produces a paragraph you can send.

Open Smith et al. at the figure, not the abstract. Suppose panel 4B is a 48-hour siRNA knockdown in one immortalised line, n = 3 wells from a single passage, with Y read out by a luciferase reporter and reported as a correlation across knockdown efficiency. Suppose there is no rescue arm and no second targeting sequence, and that the sentence “these data indicate that X drives Y” appears in the discussion rather than the results, after which two reviews cite Smith et al. as having established the dependence.

Now write your own side in the same terms. Your figure 3 is a conditional deletion in three tissues, with an orthogonal readout, an add-back rescue, and n = 6 animals per genotype across two independent cohorts. The difference is not enthusiasm; it is necessity demonstrated in vivo against association in a single line, plus the orthogonal validation the prior work did not have.

The response then writes itself. Name Smith et al., state their system and readout, state yours, name the difference in one sentence, and — where it is fair — note that the two intervening reviews are the route by which a discussion sentence became a settled result. Concede what is genuinely shared, which is that Smith et al. reported the association first, and concede it before you say anything else.

When the reviewer is right

Sometimes the work has been done. Two honest routes.

Reposition as confirmation. Independent replication is undervalued and increasingly publishable, particularly with a different system or method. Say plainly that this confirms and extends, and specify what the extension is. A paper that admits it replicates and says what it adds is more credible than one that pretends not to know. The case for replication is now quantitative rather than rhetorical: the Reproducibility Project: Cancer Biology repeated 50 experiments drawn from 23 papers, generating data on 158 effects, and found that for the positive effects the median replication effect size was 85% smaller than the median original, with 92% of replication effect sizes smaller than the effect they were repeating (Errington et al., eLife 2021;10:e71601). In psychology, the Open Science Collaboration replicated 100 studies and found 97% of the originals statistically significant against 36% of the replications (Science 2015;349:aac4716). A literature with those base rates has a standing use for a well-powered confirmation in a second system.

Find the part that is genuinely new. Rarely is an entire manuscript pre-empted. If three figures are known and two are not, the paper may be the two — smaller, faster to publish, and honest. Do the arithmetic explicitly before deciding: list every claim your abstract makes, mark each as pre-empted, extended or new, and count them. A manuscript whose new claims number two of nine is a different submission from one whose new claims number two of four, and the reviewer has effectively told you which one you are holding.

Where a confirmatory paper lands

Journal criteria decide whether “this confirms prior work” is fatal or routine, and the relevant policies are published rather than tacit. PLOS ONE states its standard directly: “We evaluate research on the basis of scientific validity, strong methodology, and high ethical standards—not perceived significance.” A sound confirmation is in scope there by stated policy, not by appeal to an editor’s goodwill.

PLOS Biology, a selective journal, publishes a window instead: “Manuscripts that confirm, replicate, extend, or are complementary to a recently published, significant advance are still eligible for consideration in PLOS Biology and may be submitted up to six months after the first article’s publication date”, provided the work is equally or more rigorous than the published study and meets the journal’s other criteria. The policy was set out in the editorial “The importance of being second” (PLOS Biology 2018;16(1):e2005203). If the paper that pre-empted you appeared inside that six-month window, give its publication date in your cover letter and name the policy.

eLife makes the same point through its vocabulary rather than its scope. Every eLife assessment rates significance of findings on a five-term scale — landmark, fundamental, important, valuable, useful, where “useful” is defined as “findings that have focused importance and scope” — and strength of evidence on a six-term scale running exceptional, compelling, convincing, solid, incomplete, inadequate. A paper assessed “useful” and “convincing” is still published; the model does not require a novelty verdict before publication happens at all.

Matching the venue’s stated criteria to what your paper actually establishes is a decision worth making before resubmission rather than after a second rejection — see journal selection.

Writing the response

Be specific and unemotional. Name the cited paper, state what it showed, state what yours shows, and name the difference in one sentence. Where you can, quote the prior work’s own limitation statement — authors are often clearer about the boundaries of their claims than later readers are.

Smith et al. reported that X correlates with Y in primary hepatocytes. Our data establish that X is required for Y in vivo, using conditional deletion in three tissues, and that the requirement is lost in the absence of Z. We have revised the introduction to make this relationship explicit.

That is a stronger answer than any assertion of novelty, because it lets the reviewer check it.

Four sentences do the work, in a fixed order: what the prior paper reported in its own terms, what your data establish, the one-sentence statement of the difference, and the change you made to the manuscript. The fourth is not optional. A response that argues without revising invites the identical comment on the next round, and editors read an unrevised introduction as a refusal. Where the objection was a framing claim, the revision belongs in the introduction and the abstract, and the response should quote the replacement sentence verbatim so the reviewer does not have to hunt for it.

Four phrases to cut before sending: “we respectfully disagree”, “as is well known in the field”, “the novelty of our work is self-evident”, and any sentence beginning “the reviewer appears to have misunderstood”. Each is read as evasion whatever argument follows it. Structure and worked wording for the letter as a whole are in how to write a response to reviewers and a response to reviewers sample.

When you cannot read the cited work

Unverifiable prior work is common, and each version of it has a different move. The one thing that never works is arguing around a citation you have not read.

Where the reviewer cites unpublished data of their own, you cannot check it and neither can the editor. Say exactly that, courteously: the manuscript cannot be evaluated against data that are not in the literature, and you would welcome the chance to cite the work when it appears. Put the point to the editor in the covering note as well as in the response, because it is the editor who decides how much weight an unverifiable claim carries.

Where the citation is a conference abstract, a preprint or a thesis, read it and treat it as the object it is. A meeting abstract with no figures does not establish a result and should not be answered as though it did. A preprint does establish priority of posting, and its posting date is public and machine-readable; if it pre-empts you, cite it and engage with it, because a manuscript that ignores a preprint the reviewer has already found reads far worse than one that addresses it.

Where the work is paywalled, or in a language you do not read, request it through interlibrary loan or write to the corresponding author, who will usually send a copy within days. If neither arrives in time, state in the response which paper you could not obtain and what you did to obtain it.

Where no citation is given at all, ask for one. “We would be glad to address this directly if the reviewer can point us to the specific report” is a legitimate request to make through the editor, and a novelty objection that does not survive that request was a framing objection all along.

When the objection comes from the editor instead

Editors phrase this objection differently from reviewers, and the phrasing tells you how much room you have. A letter saying the work is “not sufficiently novel for our readership” or “would be better suited to a more specialised journal” is a scope judgement reached without external review — the shape described in desk rejected without review — and it is usually final for that journal on that framing. Resubmitting the same framing to the same editor rarely changes the outcome; reframing it for a journal whose stated criteria fit does.

A letter that cites a specific paper as pre-empting yours is a factual claim, and factual claims can be corrected. If the cited paper does not show what the letter says it shows, a short appeal naming the figure and the difference is reasonable. If the cited paper appeared after your submission date, give both dates. The threshold for when an appeal is worth writing at all is set out in when to appeal a rejection.

Prevent it before submission

The already-known objection is usually a positioning failure discovered late. The standing review includes a missing-prior-work specialist that searches the literature for work you should be engaging with, and the study section lens tests whether your novelty claim rests on a genuine advance or on work that is merely unreported — with scholarly search enabled, it checks positioning claims against recent literature and reports assertions of unexplored territory in areas with recent activity.

Audit your own gap sentences before a reviewer does it for you. Find every sentence in the introduction and abstract of the form “remains unknown”, “has not been investigated”, “no study has examined” or “for the first time”, and require each one to survive a literal search of its own terms. A gap claim that a reviewer can break with a single query is the fastest available route to this objection, and an overclaim check is aimed at exactly that class of sentence.

Then state what is shared before you state what is new. An introduction that names the closest prior work, says what that work established, and only then says what remains open is much harder to answer with “already known”, because the reviewer’s candidate paper is already in your text with its scope correctly described.

More in this family: novelty and scope objections.

PerfectPaper reads a manuscript’s gap claims against the literature, names the specific papers a reviewer is likely to raise, and reports which of those papers actually establish the result the introduction says is unreported.

Review my manuscript

Frequently asked questions

How do I respond when a reviewer says my findings are already known?

Read the cited papers at the figure level before drafting anything, then answer factually: name the paper, state what it showed, state what yours shows, and name the difference in one sentence. Revise the introduction to reflect that comparison, and quote the revised sentence in your response.

What does it mean when a reviewer says a paper lacks novelty?

“Lacks novelty” usually means one of three things: a named paper pre-empts your result, the result follows too directly from published work, or your introduction claims a gap the reviewer does not accept. The remedy differs for each, so identify which one the comment describes before answering it.

What if the reviewer cites their own paper as prior work?

Read it on the merits and respond factually. If it genuinely pre-empts yours, say so. If it does not, state the difference precisely — a reviewer citing themselves is not automatically wrong, and treating the citation as a conflict rather than a claim weakens your position.

What if the reviewer is right and the work really has been done?

Reposition the paper as a confirmation rather than defending a novelty claim you cannot support. Describe it as a replication, say what it adds, and prefer a different system or a different method for the confirming work. PLOS Biology accepts confirmatory and complementary manuscripts submitted up to six months after the original article’s publication date, provided the work is equally or more rigorous.

How do I check whether prior work really showed what the reviewer says?

Read the figures rather than the abstract, and check the limitations section. Claims frequently strengthen as they pass from a paper’s results, through its discussion, into other people’s citations of it — the mechanism Greenberg documented in the BMJ (2009;339:b2680) as citation amplification and invention.

What if only part of my paper is pre-empted?

Publish the part that is not. A shorter paper making a claim nobody has made is worth more than a longer one where a reviewer has already flagged half as known. List every abstract claim, mark each pre-empted, extended or new, and count before deciding what the paper is.

Should I answer that my version is more rigorous than the work the reviewer cited?

Only if you can show it specifically — a larger cohort, a cleaner genetic model, an orthogonal method, a rescue arm the earlier study lacked. Name the difference in the prior paper’s own terms. Rigour claims without specifics read as defensiveness and invite the same comment on the next round.