I just received the galley proofs for an article which has been accepted to a well-regarded math journal.

The copyediting is, quite frankly, terrible. The copyeditor has introduced dozens of mathematical, typographical, and bibliographic mistakes. The proofs were accompanied by a list of changes and questions, and some of the newly introduced mistakes were described in this list, but not all.

A few representative examples:

  • The editor has, in many places, replaced the mathematical notation << with "is much less than". It is universal within my subfield of mathematics that << means something very different from "is much less than".

  • The editor has typeset fractions in different ways, so that what was formerly in the denominator is now in the numerator.

  • The editor has monkeyed unacceptably with the bibliography. I used a software package, for which I looked up on their FAQ how to cite it. The editor replaced my correct bibliography entry with something different.

  • The tables now appear in wrong places. A cursory reading of the edited text reveals that it now makes no sense, and this would be obvious even to someone without specialized knowledge of my field.

I could go on and on, but I don't want to just rant. Instead I have two closely related questions:

  • First of all, is this typical? I find it difficult to imagine that a publisher who did this kind of shoddy work on a regular basis could stay in business. I wonder if I just had the bad luck to get the new employee this time around.

  • More importantly, is a strong response warranted? I am inclined to write to the journal staff, tell them that their work is unacceptably sloppy, explain in detail why, and ask them to start over from scratch, and to furnish a list of all changes made, no matter how minor.

    This is not just because I want to pick an argument. I have gone through my paper line-by-line several times in the past, making very sure that everything I said was correct. With this level of copyediting work, I am back to square one and I suspect I might accidentally miss several errors introduced by the copyeditors.

    Would such an e-mail be likely to produce the kind of results I'm looking for?

Thank you very much.

Update: Thanks to everyone who replied. I wrote a strong, but I hope polite, e-mail to my contact at the journal, listing several of the mistakes, and asking them to start over and to send me a complete list of changes. His first reply was a little bit ambiguous, appearing to perhaps misunderstand what I was asking for -- but he has since apologized and agreed to my requests.

One point of departure from Anonymous Mathematician's advice: I haven't said anything to the editorial board and the publisher, or discussed this issue (other than here, anonymously) with anyone but my coauthor and my contact at the journal -- happily, it looks like there won't be any reason to.

Update 2: As I requested, my contact at the journal started again from scratch, did a much more conservative job of copyediting, and provided me a copy of my file which was marked up in red and blue with every change they made. Needless to say this made my job quite easy and I thanked them for their good work.

  • 8
    There's a lot of bad copy-editing in math (though a few places, like MSP, have a reputation for doing it well). However, messing up numerators and denominators seems beyond the pale to me even by the usual low standards. Jul 27, 2013 at 17:09
  • 9
    My first article got expressions like $n/2$ replaced by $n^2$ in seven places; this was introduced after the galley proofs. I have found the mathematical editor very helpful and understanding when I raised the issue (ultimately, my paper was republished entirely). So, while this is not a complete answer, I would try to email the editor in charge, explaining in particular that you do it for the sake of the mathematical content. Jul 27, 2013 at 18:24

4 Answers 4


That does sound terrible. I haven't run into anything so bad myself, but I know people who have. It seems to vary a lot by publisher. My impression is that it happens occasionally when they try to save money on copyediting by outsourcing to a new, cheap copyediting company. If the results are terrible, then the publisher will switch to the next-cheapest company, and this will continue until the publisher finds an acceptable level of quality. Unfortunately, it leads to bad results for authors who get caught in the middle of this process. Hopefully the publisher will learn from this incident and the problem won't continue.

More importantly, is a strong response warranted?

Yes, certainly.

ask them to start over from scratch, and to furnish a list of all changes made, no matter how minor

I'd guess that getting them to start over from scratch will be difficult, but you should certainly insist on seeing a second set of proofs.

Asking for a list of all changes made could be useful, but it's not clear whether you can trust that they will compile a complete list (they may be sloppy about that too). Many copyeditors mark up a paper copy before changing the file, so you may have some luck in getting a photocopy/scan of the marked up copy.

I'd also recommend letting the editorial board know, and perhaps the publisher too. If you just resolve the issue directly with the copyeditors, it's possible that nobody else will find out what happened.

  • 10
    I think that asking them to restart from scratch is good policy, if only because crime must not pay. Also, if you ask them for an exhaustive list of change, then you will still have to read the galley proofs entirely, but you will be in a position to call their sloppiness again if you spot anything. Jul 27, 2013 at 18:27

This is perfectly normal. Copyediting is often of poor quality, even with reputable publishers.

In my experience, the following approach seems to work fairly well:

  • Be polite. Thank for the proofs. Do not accuse anyone, do not vent. Imagine how you would respond if you had found just one tiny typo, and follow the same pattern even if there are dozens of mistakes.

  • List every copyediting error, and explain carefully what is wrong in each instance. Try to be as thorough as possible. Most likely this is your last chance; do not assume that you will get another opportunity to review the proofs before the paper is printed.

  • Email your response ASAP, preferably before the deadline. Remember that copyeditors have their own deadlines to meet.

This way you can maximise the chances that the copyeditors will do their best to fix all mistakes in your paper — they will know what to fix and how, and they will also have time and motivation to do it.

Summary: Do not try to fix the whole world, just try to cooperate with the copyeditors to fix your paper.

  • 10
    Hi Jukka, +1, thank you very much for your reply. But I am afraid I am bothered by one thing you say. You recommend that I "maximise the chances" that the copyeditors "do their best" to make sure that the final version is no worse than what I gave them to begin with. Isn't this something we should expect and be able to insist upon? If not, in this day of personal homepages and the arXiv, then I seriously question why we (as a community) should bother with publishers at all.
    – Anonymous
    Jul 27, 2013 at 21:49
  • 11
    I think this response is too nice. If a copy editor is grossly negligent/unprofessional, they need more than just a calm list of errors. It is perfectly reasonable to ask the publisher (via the editor) to start over, while providing them both a clean set of source files and a clearly representative list of errors introduced by the copy editor. Yes, respond promptly and professionally, but be clear that you expect professionalism in return.
    – JeffE
    Jul 28, 2013 at 8:46
  • 4
    One problem with this course of action is that if one misses one error added during the first round of copy-editing, it is likely that the error will stay. In such a bad situation, asking the publisher to start the copy-editing from scratch is the only way I see to avoid that. Moreover, one should try not to take too much time to fix others mistake, unless it is one's job. Jul 28, 2013 at 9:50
  • 1
    @Anonymous: Well, commercial publishers try to provide some "added value" so that they can justify their existence. This is why the papers are mangled in the typesetting and copyediting processes. It is a strange game — your adversary tries to hide mistakes and you are trying to spot them quickly — but in my experience it is possible to win the game. Or at least not lose too badly. (Computer-aided tools help; for example, you can convert both publisher's proofs and your own version to plain text and automatically highlight the differences.) Jul 28, 2013 at 10:53
  • 1
    Fortunately, I've never had a reason to make such a request.
    – JeffE
    Jul 28, 2013 at 17:41

Tell them to revert everything to the original version, period. If they introduced 100 mistakes, it is not your job to find them all (besides, I doubt you'll find more than 70 anyway). Be polite but firm. Don't accuse anybody of anything but stand the ground even if you'll have to withdraw the paper. Your primary duty, as a writer, is to readers, not to publishers. Good luck!


If a copy-editor (CE) changes the meaning of your manuscript you obviously need to recorrect it. The CE should have received a set of "house rules" for how thing should be formatted but it is not likely the CE is an expert in evey field. In your case it sounds as if the CE has been "trigger happy". In any case, you should makethe necessary corrections and return the proof to the editor detailing what you find wrong with the CE edits. You should clearly state why the edits are unacceptable.

Concerning your point about location of tables, I am not sure what exactly you mean by "wrong place" but the journal will obviosuly place tables and figures where they make type-setting sense but should of course be located after where they are first referenced in the text.

So in some cases journals (CE) may change things to adhere to "house rules". These changes must be accepted but of course not when they change the meaning of the paper or introduce errors. I have experienced similar (but not as severe) edits to my own papers, I have also been in the position to enforce house rules in the journal I edit, but never to the point of changing the meaning. If in doubt I would have contacted the author or passed along a query about the particular issue.

  • 2
    Regarding "In any case, you should makethe necessary corrections and return the proof to the editor detailing what you find wrong with the CE edits." Thanks Peter, but this does not quite address my question. If the CE has introduced fifty errors I imagine that I will only find 47 of them. Would I be reasonable to insist that they start over? Thanks.
    – Anonymous
    Jul 27, 2013 at 16:52
  • 1
    @Anonymous Yes, if you feel the corrections have come completely out of control 8as it certainly seems) then the best would be for the CE to start over with the original file. you can make this point clear and that you cannot with certainty revert all the unnecessary corrections based on their proof. besides, it is the CE (=journal) that has caused the problem. Jul 27, 2013 at 17:36

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .