Recently, I submitted a paper to a maths journal (which is a decent journal with a solid editorial board). It was eventually accepted after making some revisions suggested by the referees, at which point I uploaded my final submission. This final submission was then edited by a copy editor at the journal; as well as the usual cosmetic changes in ensuring that my papered adhered to the cosmetic style of the journal, the copy editor also made several grammatical changes. I then was given an opportunity to read through this edited version and suggest any final corrections before my paper is uploaded to the journal's website.
This, of course, is all fairly standard practice. However, many of the grammatical changes made by the copy editor were incorrect. Of course, I noted this in my comments, so I hope these changes will be reverted before the paper is published.
It is appropriate for me to take further action in notifying the journal (e.g. a member of the editorial board) that a copy editor is repeatedly introducing errors?
This of course seems a little petty. On the other hand, these same grammatical errors occur repeatedly throughout papers in this journal (at least in recent papers published online), and I'm sure that I'm not the only reader who finds these mistakes irritating.
In case anyone is curious about the errors, the most common mistake is that the copy editor repeatedly replaced with a comma my usage of a semicolon before an independent clause, especially such a clause in the imperative mood: an example would be something like "This can be proved via the method of Gauss, see [1]". Other such comma splices were introduced - all by replacing semicolons with commas.
Additional grammatical mistakes were introduced that were clearly incorrect: for example, I perhaps overuse the phrase "Note that", and this was pruned on a couple of occasions by the copy editor, but in more than one case, the rest of the sentence wasn't edited to ensure that it still made sense.