Helping a colleague with an English manuscript is genuinely useful work. It is also where a helpful edit quietly becomes a takeover — and the person whose paper it is rarely says so.
The line is between fixing how something is said and changing what is being said.
What is fair to change
- Grammar, article use, tense consistency.
- Sentence length and structure for readability.
- Word choice where a term is technically wrong in English usage.
- Paragraph order where the logic is genuinely hard to follow.
What should be raised, not changed
- The framing of the contribution.
- Which results are emphasised.
- Strength of claims in the discussion.
- Which literature is engaged with.
- Raise these as questions in comments, and let the author decide.
How to give the feedback
- Use tracked changes so the author sees every edit.
- Explain the reason for a pattern of edits once, rather than repeating silently.
- Separate "this is wrong in English" from "I would phrase it differently".
- Return the file rather than a rewritten version.
What not to do
- Do not rewrite wholesale and return it as finished.
- Do not treat non-native phrasing as evidence of weak thinking.
- Do not use the editing role to claim authorship.
- Language editing alone is not an authorship contribution.
If you are the author receiving edits
- Accept language corrections; they are usually right.
- Push back on changes to meaning; you know your data.
- Ask for the reason behind edits you do not understand.
- Your voice in the paper is legitimate even if it is not native-sounding.
One thing worth remembering
Separate "this is incorrect English" from "I would have written it differently" and label them differently.
The first is a service. The second is a preference — and presenting preference as correction is how a paper slowly stops sounding like the person who did the research, without anyone intending it.
Câu hỏi thường gặp
What is fair to change when editing?
Grammar and article use, tense consistency, sentence length and structure for readability, word choice where a term is technically wrong, and paragraph order where logic is hard to follow.
What should be raised rather than changed?
The framing of the contribution, which results are emphasised, strength of claims in the discussion, and which literature is engaged with — raised as questions for the author to decide.
How should the feedback be given?
Use tracked changes so every edit is visible, explain a pattern of edits once rather than repeating silently, and return the file rather than a rewritten version.
Does language editing earn authorship?
No — language editing alone is not an authorship contribution, and the editing role should not be used to claim one.
What distinction matters most?
Between this is incorrect English and I would have written it differently — presenting preference as correction is how a paper stops sounding like the person who did the research.
