How to answer a reviewer who says your English needs improvement.

When a reviewer says your English needs improvement, do three things in order: work out whether the complaint is really about grammar, clarity, or structure; run a separate revision pass for each; then describe that work in your response letter in terms the editor can verify. The line reads like a judgment on you, which is why it stings. It is also the cheapest problem in most reports to close. A questioned method can mean months of new experiments. Language costs you one careful week.

What does the comment usually mean, and what does it rarely mean?

Most of the time it means the reviewer had to read your sentences twice to find the point. It rarely means your grammar is broken, and it almost never means the paper was rejected on language alone.

"The English needs improvement" is a stock phrase, and reviewers stretch it over everything from a missing article in the abstract to a Discussion with no visible spine. Some journals set a floor in their guidelines. Frontiers in Neurology states: "Frontiers requires manuscripts submitted to meet international English language standards to be considered for publication." That is a floor, not a demand that you sound like a native speaker. It asks that a native speaker be able to read you without effort.

For an author working between Arabic and English, the comment lands harder than the report intends, because it arrives attached to your name rather than to a section number. The text is what the reviewer actually read, and text is the one thing you can change before the deadline. The reviewer said the English needs improvement sorts the comment into its variants; this post is about what to do next.

How do you tell grammar, clarity, and structure apart?

Read the reviewer's other comments and any marked-up pages before you touch the file. Grammar problems show up as line edits, clarity problems as misreadings of what you actually wrote, and structure problems as requests for things the paper already contains.

Each leaves a different fingerprint. An annotated PDF full of small corrections points at grammar and usage, the quickest kind to clear. A reviewer who restates one of your results incorrectly is telling you the sentence carrying that result is unclear, however grammatical it looks. When the comments keep asking for material that already exists, a limitation you stated in section 4.3 or a control described in the Methods, the paper has a structure problem: the information is there, but not where a reader reaches for it.

Bilingual writers should check a fourth case: sentences that are correct in English yet keep the shape of Arabic, with the verb landing late or clauses strung together the way Arabic strings them. Seven signs your paper reads like a translation names the recurring patterns and the fix for each.

Which revision passes matter most?

One pass per problem, in this order: structure, then clarity, then grammar. Polishing commas in a paragraph you may later delete wastes the week you have.

The structure pass is paragraph-level. Give each paragraph a single claim, put that claim in the opening sentence, then read the opening sentences in sequence and check that they summarize the section on their own. Move buried limitations and definitions to where the reader needs them. This pass changes the most while costing the least, because you are cutting and moving rather than rewriting.

The clarity pass runs sentence by sentence through the sections the reviewer stumbled on. Bring the subject and verb near the front. Split any sentence that carries two claims. Trade inflated phrasing for the plain version. Grammar comes last, once you know which sentences will survive the first two passes. For Arabic-English writers the repeat offenders are articles, prepositions tied to particular verbs, and tense drifting between Methods and Results. Keep a running list of your own habitual errors and hunt them one at a time; a targeted search for article mistakes beats another vague reread of the whole draft.

How do you document the language revision in the response letter?

State what you changed, how widely, and where the editor can confirm it. Two or three sentences in the general reply, plus pointers to rewritten passages inside the point-by-point entries.

A bare "the English has been improved" gives the editor nothing to check and reads as a wish rather than work. Compare it with this:

We revised the full manuscript for language. The Methods and Results were edited sentence by sentence for grammar, and the Discussion was reorganized so each paragraph opens with its claim. All edits are visible in the tracked-changes file, and the most heavily revised passages are quoted in our replies to comments 2 and 5.

That version makes three statements the editor can verify by opening one file. If the surrounding letter still feels shapeless, how to write a response to reviewers letter covers structure and tone, and the point-by-point format editors expect shows where a language comment sits among the rest.

Do you have to disclose editing services or AI tools?

Name a professional editing service if you used one, since a named service is a fact the editor can check. For AI tools, follow your journal's disclosure policy to the letter, because journals now split on this.

No certificate is needed to make the mention land. The certificate is optional support; the revised text carries the argument. Disclosure is the part to get exactly right. The Frontiers in Neurology guidelines state: "If generative AI technology was used to produce written or visual content, this must be acknowledged in the acknowledgements section (and methods section if applicable), listing the name, version, model, and source of the AI technology." Read the clause closely. The acknowledgement belongs in the manuscript itself, not only in your letter, and it names four items: tool, version, model, source. If your target journal publishes no AI policy at all, a short acknowledgement is still the safer route. One disclosed sentence costs nothing; an undisclosed tool found later costs the editor's trust.

How do you stop the comment before it happens next time?

Run the same three passes before you submit, and check the draft against the journal's stated language and disclosure rules while you can still act on them.

The passes do not change; only the timing does. Keep the personal error list running between papers, because your grammar habits travel with you from one manuscript to the next. Swap structure reads with a colleague, where each of you checks only whether every paragraph announces its claim. Read the abstract and the topic sentences aloud, since the ear catches translationese the eye forgives. A pre-submission review that measures the manuscript against the journal's own rules, which is the work Waraq does, surfaces the language floor and the AI disclosure clause above before a reviewer ever sees the draft.

Reviewers who write this comment rarely reopen it once the resubmission reads cleanly and the letter shows the work. Fix it in the right order, describe it plainly, and the thread closes.