Language editing, and whether English can sink a paper

Can a paper be rejected for its English?

Yes, and PLOS ONE says so plainly. It requires manuscripts to be submitted in English, and states that submissions failing to meet its publication criterion for language standards may be rejected.¹

That is a criterion rather than a preference, and it sits alongside the journal's other conditions of consideration. A paper can be scientifically sound and still fail it, and the failure happens before review rather than during it.

The same journal also states that it does not copyedit submissions before publication, and encourages authors to use professional editing services before initial submission.¹

Journals differ and most say less than this. Where a journal's instructions are silent on language, the safe assumption is that a reviewer's patience is finite rather than that the standard does not exist.

Does editing improve my chances of acceptance?

It removes one specific ground for rejection and it does not do more than that. Where a journal can reject on language standards, a paper that meets them is not rejected on that ground.

What no editor can promise is acceptance. Novelty, fit with the journal, methods, and the reviewers' judgment are decided by people an editor never speaks to, and any claim to influence them should be treated as a warning sign.

The honest framing is negative rather than positive. Editing does not make a paper more likely to be accepted for its science. What it does is stop the language from being the reason the science is not read.

That is worth paying for on its own terms. A desk rejection costs months, and the months are the expensive part rather than the fee.

Should I edit before or after peer review?

Before, if the journal can reject on language, because a paper rejected at the desk is never reviewed. The whole benefit of a language edit is realized at the point the paper is first read.

Editing again after review is a different and smaller job. Revisions written under time pressure, in response to reviewers, are frequently the roughest prose in the final paper, and they arrive in a manuscript where everything else has been edited.

A workable arrangement is a full pass before submission, then a short pass on the revisions and the response letter afterward. Agreeing both at the start costs less than commissioning the second one in the week it is needed, and it also secures the time.

What is a language-editing certificate?

A statement from an editing service that a manuscript has been edited by a native or professional English editor, which some journals invite authors to supply. It is a record of what was done rather than a guarantee of anything.

No journal we found treats a certificate as a substitute for meeting its language criterion, and none we found promises acceptance to a paper that carries one. Where a journal asks for evidence of editing, a certificate answers the request.

What it does not do is transfer responsibility. The authors remain accountable for the paper under the ICMJE criteria, whatever any certificate says.²

It is also worth checking whether the journal wants one before buying one. Some ask, many do not, and a certificate nobody requested changes nothing about the submission.

How much does editing a paper cost?

The Editorial Freelancers Association's 2026 rate chart reports academic work separately for faculty publication. Copyediting academic humanities material appears at 4.0 to 5.0 cents a word, or $50.00 to $60.00 an hour, and the STEM equivalent at 3.8 to 5.0 cents a word, or $50.00 to $65.00 an hour.³

Proofreading academic humanities work for publication appears at 2.8 to 4.0 cents a word, and developmental editing at 5.0 to 7.3 cents.³ The chart is based on a survey of more than 1,100 respondents, and the EFA states its figures are for guidance only and are not mandated or recommended rates.⁴

A journal article is short enough that the per-word arithmetic is manageable. A 6,000-word paper at the copyediting rates above is a few hundred dollars rather than a few thousand, which changes the calculation against a desk rejection.

What can an editor not fix?

Four things, and it is worth knowing them before paying. An editor cannot make the study novel, and cannot make the methods adequate. Nor can an editor make the results support a stronger claim, or make the paper a fit for a journal that does not publish this kind of work.

Nor can an editor supply missing content. Where a paper does not report something a reporting guideline requires, the finding comes back as a gap for the authors to fill rather than as text an editor has written.

What an editor can do is make the paper say what its authors meant, consistently, in language a reviewer reads without friction. That is a real contribution and a bounded one.

Is my paper too rough to be worth editing?

Rarely, and the useful question is which level of work it needs rather than whether it is beyond help. A paper whose argument is sound and whose sentences are hard going is exactly what language editing is for.

The case where editing is the wrong purchase is a paper whose structure does not yet work. Polishing sentences inside sections that will be reorganized wastes the pass, and a structural read first is both cheaper and more useful.

An editor reading a sample can say which of the two you have. That reading is part of quoting rather than a separate service, and it costs nothing to ask for.

Where the answer is that the structure needs work first, that is not a refusal. It is a cheaper piece of work offered ahead of a more expensive one.

English is not my first language. Does that change anything?

It changes what the editing concentrates on and it does not change the standard. Constructions that are grammatical but not idiomatic appear throughout rather than in patches, article and preposition use is the commonest pattern, and technical vocabulary is usually already correct.

One thing worth saying plainly: writing in an additional language is not a defect to be apologized for, and detectors marketed as identifying machine-written text produce false positives on exactly this kind of prose. An editor who treats unfamiliar phrasing as suspicious is the wrong editor.

We edit for clarity and correctness while keeping the author's own voice, which the professional standards treat as part of the work rather than a courtesy.⁵ Every document a client sends us is treated in confidence, including manuscripts not yet submitted anywhere.