Why do we bother to copy‐edit?

I would guess that many readers of this journal started their publishing careers, as I did, in copy-editing. We were well-trained and assiduous – we never doubted for a moment that what we did was necessary and important. But is it? ‘The copy-editor’s task is to ensure that the material which goes to the typesetter is clear, consistent, unambiguous and well organised; to make it easy for the reader to follow and understand, while changing as little as possible of the author’s text.’1 In addition to making the text clear to the reader, copy-editors also check completeness of text and illustrations (including verifying that permissions have been sought and obtained where necessary), and ensure consistency with standards within the discipline (e.g. of nomenclature) as well as with their own journal’s house style. In today’s e-journal world, it is more important than ever that they check the citations for accuracy and completeness, and add DOIs in order to enable citation linking (e.g. via CrossRef2). Copy-editors refer to a number of ‘bibles’, in particular Butcher’s Copyediting3 and the Chicago Manual of Style;4 publishers also send their copy-editors on training courses such as those offered in the UK by the Publishers’ Training Centre.5 There have been various estimates of the cost of copy-editing a journal article. Donald King6 recently updated all the estimates he had been able to find to current values: he shows figures from Marks7 for ‘editorial mechanics’ of $254– 444 per article, from Holmes8 for ‘editing, pre-marking, tagging’ of $903 and from Dryburgh9 for ‘rewriting, copy-editing’ of c. $60–610. So are we spending all this effort and money in vain? Of course, effective copy-editing is invisible to the reader – if it shows, it has been done badly! This is, perhaps, its Achilles’ heel; only the author knows that anything has been changed. In 2002 ALPSP carried out a study10 in which respondents were asked, as authors and again as readers, which elements of the current journals system they would wish to see preserved in future. In the former role, 60% felt that ‘content editing and improvement of articles’ should definitely be preserved, and 50% felt the same about copy-editing; however, with their readers’ hat on, these figures came down to 39% and 34% respectively. As authors, 45% valued ‘checking of citations/adding citation links’ – again, as authors only 28%. Fewer and fewer publishers do their copy-editing in-house; instead they rely on expert freelance copyeditors (e.g. those certified by the Society for Editors and Proofreaders11). In an attempt to cut costs, however, we now see a trend towards either outsourcing copy-editing overseas – almost inevitably, therefore, to non-native English speakers12 – or even cutting back or dropping it altogether.13 Does it matter? Two recent studies of the potential effects of author self-archiving on future journal subscriptions, though consistent on many points, came to different conclusions on the importance of the final publisher’s version after copyediting. The respondents to Ware’s survey14 said that an author’s postprint version (i.e. including changes from peer review but not copyedited) would not be an acceptable substitute for the published version; however, Beckett and Inger’s15 respondents felt that the post-peer review version was adequate, irrespective of whether it had been edited by the publisher. If the latter survey is correct, and librarians are anxious to have the post-peer-review version but do not care whether it is copy-edited or not, we have to ask ourselves why we should spend time and money doing it at all. Learned Publishing has recently published two studies on the difference made by publishers’ copy-editing (Wates and Campbell16 in the previous issue, and Goodman et al.17 in this one). It is noticeable that both found a very small percentage of differences that actually affected the meaning of the article – in Goodman’s case, some actually introduced by the publisher; however, the former identified a significant percentage of editorial queries (43%) about incomplete or inaccurate references, as well as a smaller percentage (14%) about missing data. Without complete, accurate, and correctly Editorial 163