Skip to main content

Free guide

When the Process Goes Wrong: appeals, delays, coercion, and getting an error off your record

11 min read · verified August 21, 2026

Every process in this guide was read off the journal or publisher page that defines it, on the date above, and every one is linked at the end.

The single most useful thing here is that an appeal and a complaint are different instruments. They have different grounds, they go to different people, and they are decided in different ways. Most authors send one long email that mixes the two, and it fails on both counts. Once you know which one you need, the rest of this is procedure.

Yours to keep, free

The whole guide is on this page. The PDF is the same thing, formatted to print or keep on your phone.

Download the PDF (PDF file)

01Appeal or complain, because they are not the same thing

This is the fork everything else hangs off. An appeal contests a decision about your manuscript. A complaint contests how the journal behaved. Sending the wrong one to the wrong person is the most common own goal in this whole area.

  • An appeal asks for a rejection to be reconsidered. At Nature Portfolio it starts with the Handling Editor, following the instructions on that journal’s own site, and the final decision rests with the editorial team.
  • A complaint is a separate track and it covers two different things: failures of process, such as long delays, and matters of publication ethics, such as misconduct. Those go to the Chief Editor or Editor in Chief rather than through the appeal route.
  • Nature Portfolio states the boundary directly: to complain about an editorial decision to reject, use the appeal policy instead. The complaint procedure is not a second appeal with a stronger tone.
  • This matters because the two are judged against completely different standards. An appeal has to show the decision was wrong on narrow grounds. A complaint has to show a process or a person fell short. Evidence that would win one is irrelevant to the other.
  • So before you write anything, answer one question: are you saying the decision was wrong, or are you saying the journal behaved badly? If the honest answer is both, they are two messages, not one.

02What an appeal is actually allowed to argue

Appeals are far narrower than authors assume. PLOS publishes its grounds, and they are worth reading closely, because most appeals fail for arguing something that was never admissible.

  • PLOS will consider an appeal on essentially two grounds: that a reviewer or editor made a significant factual error, or that their objectivity may be compromised by a documented competing interest.
  • There is a third condition attached to both, and it is the one people miss: a reversal on those grounds would have to change the original decision. A real error that would not have changed the outcome is not a basis for an appeal.
  • "The reviewers did not appreciate the contribution" is not on that list. Nor is disagreement with their judgement, however well argued. Merit is what the original decision was about, and re-arguing it is asking the same question again rather than raising a new one.
  • Note the word documented. A suspicion of a competing interest is not the same as one you can point to, and the appeal has to point to it.
  • The formality is real. PLOS requires an appeal in writing to a named address with a specific subject line and a completed Appeal Request Form, and states it will not consider appeals made by phone or other informal means.
  • PLOS also says decisions on appeals are final without exception, and may decline to review an appeal at all where integrity concerns contributed to the rejection. You get one attempt, so aim it at an admissible ground.

Which instrument do you need

  • The decision was wrong because a reviewer got a fact wrong, and fixing it changes the outcome. That is an APPEAL, to the handling editor, citing the specific error.
  • The decision may have been made by someone with a competing interest you can document. That is an APPEAL, and name the documented conflict rather than implying it.
  • It has been months with no word. That is a COMPLAINT about process, to the Chief Editor or Editor in Chief, and it is a defined route rather than a favour you are asking for.
  • You were asked to add citations that serve the journal rather than the paper. That is an ETHICS matter, not an appeal, and it has its own escalation path above the handling editor.
  • You disagree with the reviewers about the importance of your work. That is none of these. Revise and submit elsewhere, which is faster than any appeal and usually right.

03When it has gone quiet for months

Silence is the most common failure and the one authors tolerate longest, usually because it feels rude to chase. There is a named process for it, which means chasing is expected rather than presumptuous.

  • Know the normal clock before you decide something is wrong. PLOS states that external peer reviewers typically have 10 days to submit their review, and that the journal office follows up with late reviewers and keeps the author informed of delays.
  • That is the benchmark that makes silence legible. Reviewers running weeks late is ordinary. A journal not telling you is the part that has actually failed, and it is the part worth raising.
  • A long delay is a process complaint, and at Nature Portfolio it is handled by the Chief Editor or Editor in Chief together with the handling editor, who investigate and give the complainant feedback.
  • Write it as a process complaint rather than a plea. Give the manuscript number, the submission date, the date of the last substantive update, and ask for a status and an expected date. No apology for asking.
  • Keep it short and unemotional. You are asking a journal to account for its own stated timeline, which is a reasonable request that gets a better response than frustration does.

04Coercive citation, defined precisely enough to act on

An editor asks you to add citations and they all point at their own journal. There is a rigorous definition of when that crosses the line, and it comes from the study that measured the practice rather than from opinion.

  • The original definition has three parts, and a request is coercive when all of them hold: it gives no indication that the manuscript was lacking in attribution, it makes no suggestion as to specific articles, authors, or a body of work requiring review, and it only guides the author to add citations from the editor’s journal.
  • That is an operational test, not a vibe. Run the request you received against those three and you have your answer without having to decide whether anyone acted in bad faith.
  • The wording is often courteous, which is what makes it hard to name. The published example reads, with the journal redacted here, that the author cites the journal once in 42 references and is therefore kindly asked to add references of articles published in it. Polite, and it names no finding.
  • It is common. In a survey of 6,672 researchers and data from 832 journals, 86 percent viewed coercion as inappropriate and 81 percent agreed it reduces a journal’s prestige, yet 57 percent still said they would add superfluous citations before submitting to a journal known to coerce. The norm is understood and widely broken anyway.
  • It is not evenly distributed. Being an assistant professor rather than a full professor added 5.5 percentage points to the chance of being coerced, and each additional coauthor subtracted 2.0 points. Junior and alone is the profile, which reads less like malice than like a prediction about who will push back.
  • Higher ranked journals in that data were more likely to coerce, not less, and the authors note they cannot tell whether coercion helped build the rank or protects it.
  • Scope, stated honestly: this is 2012 data from economics, psychology, sociology and business disciplines. The authors say explicitly that further work is needed to know how far the pattern extends into the physical and biological sciences, so treat it as well-evidenced there and plausible elsewhere.

05The move that calls it, without torching the acceptance

You do not have to accuse anyone, and you should not. You ask one question that a legitimate request can answer easily and a coercive one cannot answer at all, because the definition says so.

  • The publisher’s own standard is the lever. Elsevier states that an editor shall not require that references to that or any other journal’s articles be included except for genuine scholarly reasons, and that reviewers should not suggest citations to their own or their associates’ work unless the suggested citations are for genuine scientific reasons.
  • So ask which specific claim in your manuscript each suggested reference supports. That is a normal editorial question, phrased as cooperation rather than accusation.
  • A genuine request answers it in one line, because a real gap in attribution points at a specific finding. That is the good outcome and it happens often. You cite the papers, the manuscript is better, and nothing was wrong.
  • A coercive request cannot answer it, by its own definition, because making no suggestion as to specific articles or a body of work requiring review is one of the three criteria. You have not accused anyone and you have still established the thing that matters.
  • Either way you now have it in writing, which is the part that counts if you ever need the ethics route below.
  • Do not simply refuse and do not simply comply. Refusing without asking looks obstructive and gives you nothing on the record. Complying quietly is how 57 percent of researchers keep the practice profitable.

The reply that asks the only question that matters

  • Answer the substantive review points first and separately. This is one line in a cooperative reply, not a standalone confrontation.
  • Ask which claim each suggested reference supports. Specific, answerable, and not an accusation.
  • Say yes in advance to whatever is genuine. It costs nothing and it makes the question obviously good faith.
  • Keep it in the submission system rather than a private email, so it sits in the record by default.
  • If the answer never comes and the request is repeated, stop treating it as an editorial matter. It is now an ethics complaint.

This works because of how coercive citation is defined: a request that makes no suggestion as to specific articles or a body of work requiring review is one of the three criteria. A legitimate request names a finding in one line. Asking costs you nothing and reads as diligence either way.

06When the editor is the problem, there is a route above the editor

Authors assume the handling editor is the end of the line, so a problem with that editor feels unappealable. It is not. The escalation path is published and it explicitly contemplates the person at the top being the subject.

  • Ethics complaints are handled in the first instance by the Chief Editor or Editor in Chief, or the handling editor, following the guidelines published by the Committee on Publication Ethics.
  • If the conduct of the Chief Editor or Editor in Chief is itself the subject of the complaint, Nature Portfolio directs the complainant to the Springer Nature Research Integrity Group instead. The obvious dead end is anticipated and routed around.
  • If you remain dissatisfied with how a complaint was handled, it is escalated to the journal’s editorial management team, and where no publishing contact can be identified the Research Integrity Group is again the address.
  • The practical lesson is to use their vocabulary. A message that says it is a complaint about publication ethics enters a defined process. The same facts in an angry email to the editor who caused the problem enter nothing.
  • Bring documents, not adjectives. Dates, the manuscript number, and the exact wording of what was asked of you. The process is designed to act on a record.

07Getting an error off your record after publication

Something is wrong in a published paper, possibly yours. The category the journal chooses decides what your record looks like for the rest of your career, and the categories are not interchangeable.

  • An Author Correction covers an important error made by the authors. A Publisher Correction covers an important error made by the journal. Same visible fix, completely different attribution, and typesetting and production errors belong in the second one.
  • That distinction is worth insisting on. If the journal introduced the error, ask for a Publisher Correction by name rather than accepting an Author Correction, because the notice is indexed and permanent.
  • An Addendum is the route when significant additional information crucial to understanding the article emerges later. It is not an admission of error.
  • An Editor’s Note signals that the journal has opened an inquiry. It is online only, HTML only, and is not indexed. An Editorial Expression of Concern is the serious version: it gets a DOI and is indexed in major databases. The gap between those two is large and worth understanding before you agree to either.
  • A Retraction is issued when the integrity of the work is substantially undermined by errors in conduct, analysis or reporting, or by an ethics violation. The original stays available as a PDF and is marked retracted.
  • The detail almost nobody knows: retraction statements typically include a statement of assent or dissent from the authors. If you disagree with a retraction, your dissent goes on the record beside it. Ask for it in writing rather than assuming a retraction has to be silent.

Where to check

The databases this guide sends you to, all free and open without an account.

  • Committee on Publication Ethics (COPE) · The guidelines journal editors are expected to follow, including flowcharts for specific situations. Nature Portfolio’s own complaints policy points its editors here, which is why naming it in a complaint carries weight.
  • Retraction Watch database · Search whether a journal or a paper already has a history. Useful before you submit, and useful context if you are raising a concern about published work.
  • PLOS appeals policy · The clearest published statement of what an appeal may and may not argue. Worth reading even if you are submitting elsewhere, because the grounds are broadly typical.

Where these numbers come from

  1. 01Appeals are considered on a significant factual error or a documented competing interest, only where a reversal would change the decision; appeals must be in writing with a request form; decisions on appeals are final without exception; reviewers typically have 10 days PLOS, Editorial and Peer Review Process
  2. 02Appeals start with the handling editor and are deprioritised behind active submissions; complaints about process and about publication ethics are a separate track to the Chief Editor; conduct of the Chief Editor escalates to the Springer Nature Research Integrity Group Nature Portfolio, Appeals and Complaints
  3. 03Author Correction versus Publisher Correction, Addendum, Editor’s Note (not indexed) versus Editorial Expression of Concern (indexed, with a DOI), and retraction statements typically recording author assent or dissent Nature Portfolio, Corrections, Retractions and Matters Arising
  4. 04Editors shall not require citations to their own or any journal except for genuine scholarly reasons, and reviewers shall not suggest their own work except for genuine scientific reasons Elsevier, Publishing Ethics
  5. 05The three-part definition of coercive self-citation, the survey of 6,672 researchers and 832 journals, and the regression showing assistant professors at +5.5 percentage points and each additional coauthor at -2.0 Wilhite and Fong, Coercive Citation in Academic Publishing, Science, 2012

This guide will go out of date

I will tell you when it does

One email a week on what changed in academic publishing and whether it actually matters for your paper. Every number links its source.

One email a week, Mondays. Leave whenever you like.

The next step

Every process here is slower than the thing that avoids it.

An appeal that succeeds still costs months, and PLOS says appeal decisions are final while Nature Portfolio says they are deprioritised behind live submissions. The leverage is earlier, in a manuscript whose claims are specific enough that a reviewer has less to push back on and an editor has less room to ask you for something that is not about your science. The Academy works on that part, on your own draft, one task at a time. Keep this guide for the day you need it, and hope you do not.

Get published, one task at a time.

See what the Academy is