The Fundamentals of Copyediting
To Change or Not to Change
Ah, isn't this the question? The answer requires time, patience, and practice--and it is very dependent on the individual Copy Editor, journal, and author. Put simply, your first task as a Copy Editor is to ensure that a manuscript is formatted properly so that the printer can translate the electronic file into a readable proof; mainly, this is the proper placement of codes within the text. Your second task is to make certain that the article is understandable. This means that all sentences should be complete sentences, that all abbreviations have been defined, and that the concepts and experiments presented are clear and can be followed. Your final task is to ensure that all the papers you edit follow APS style, so that all APS articles have a consistent approach to the material and are edited similarly.
Sometimes, however, there are stumbling blocks in your tasks. You may have an article where the author insists upon a strange layout or complex coding. Or the author may insist in wording sentences in such a way that the flow is abysmal, incomplete, and could never, under any circumstances, make sense. They may also state, under no uncertain terms, that an abbreviation used so frequently that it should not be defined anywhere in the text. It may also be that APS style, the goal of which is to be consistent, fails in this regard: your use of grammar and punctuation may run counter to the author's, or even another Copy Editor's, use of style.
This is when it comes down to a matter of individual Copy Editor philosophy. We all agree that, in the whole, less editing and fewer changes are for the better; the fewer alterations that you make to the text, the lower the chance that the author will balk at the changes made or that you could introduce scientifically relevant errors to the results. While many of the formatting codes that we use are nonnegotiable with the authors and are required for the printer to properly construct the layout for the article, they are nonetheless very versatile and can accommodate author peculiarities. In regard to grammar and style concerns, we save changing the text to those instances where the author has committed gross infractions against the English language. This is, again, a completely subjective decision!
The answer to the conundrum is that what you change and how you change it depends a great deal on the stage of production of the article. At manuscript stage, you have a greater number of options and can therefore feel more free to make greater numbers of changes. Each time you change something that could have an effect of the readability of the sentence or alter the data, however, it is your responsibility to query the author and request that they review and accept the revised sentence. Feel absolutely free to query for anything you think is important; for very complex (or eccentric) papers, there could be over 20 queries--it's not like you're forced to limit yourself. If you want, you can also e-mail the author, tell them about any global changes you want to make, and ask for their approval. You can also freely consult with other Copy Editors or your Journal Supervisor, or you can use the Internet or use other resources to review subject matter about any proposed changes.
At proof stage and beyond, though, your options become more limited. The author has now seen the paper, reviewed it, and made note of any changes that they feel are necessary. And the author is, of course, the final say on the paper barring some unusual event. When you make corrections to a proof, you run a greater risk of changing something important--and that may be something that shouldn't have been changed. This does not mean that things do not get changed at proof stage, because sometimes these changes are scientifically necessary and have been overlooked by the author or the Copy Editor who handled the manuscript. It is just necessary for these changes to be considered more carefully and made more sparingly. At proof stage, the task of making the paper readable is more important than making sure that it conforms to APS style. If there is anything else to be said, it's to, once again, use your best judgment!