On Proofreading and Copy Edits

Among writers, it’s generally held that “Where do you get your ideas?” is the most annoying question that one can be asked. Actually, I beg to differ. That question seems quite natural to me, as it centers on the essential mystery of creativity—but I don’t have an answer to it for precisely that reason, and most other writers don’t either, which is why the inquiry sometimes rankles.

No, for me the most irritating question that can be asked of a writer is, “Does nobody proofread your books?” That, friends, really gets my goat, and here is why.

I’ve just finished proofreading the British galleys for A Book of Bones, and it nearly brought me to the edge of a breakdown. In case you’re unfamiliar with the processes through which a book comes into being, they are basically this: the author delivers a book—in my case, at somewhere between the seventh and twelfth draft, aided by the sometimes dubious merits of Microsoft Word’s spelling and grammar check; the editor may or may not suggest changes; the author may or may not take some or all of these changes on board, and alter the manuscript accordingly; the book is then copy-edited, which means a specialist editor goes through the manuscript to look for errors and inconsistencies, as well as to mark it up for the printer; and the copy edit is returned to the author, who then deals with the copy editor’s queries, and signs off on any changes. A lot of this is now done on screen, but actually it should be carried out on paper, which is what I prefer my publishers and editors to do. The eye tends to scan when it reads on a screen, and scanning is the enemy of error-spotting. If you’re serious about looking for mistakes, it should be hard copy all the way.

Okay, so that’s the end of part one, as it were. Part two involves the copy-edited manuscript being typeset, so that it now resembles the finished pages of a book. Those loose pages, referred to as “galleys,” or “proofs,” are then sent back to the author (and often to another copy editor) for checking. Again, I prefer to do this on paper.

For A Book of Bones, I also arranged for copies of the proofs to be sent to Minion Clair, who has a good eye, and my first ex-girlfriend Clio (I know that sounds odd, but we really are on good terms, and she’s an academic with a self-professed geek streak). Meanwhile, Swati, my British editor’s assistant, went through her set to crosscheck it with the original manuscript, in order to make sure that the changes made by the original copy editor, and me, went through correctly.

And finally, I went through the proofs myself, line by line and page by page, very slowly. As I completed a section, I would pass it to my significant other—and fellow author—Jennie Ridyard, who would also go through it line by line, page by page, checking not only my corrections—to ensure that in rectifying one error, I didn’t inadvertently create another—but also reading the book for the first time, so that she was performing a dual role as both editor and reader.

Meanwhile, I was going slowly mad, as I always do at this stage. The cause of the madness is this: as the author, I’m not only looking for errors, repetitions, misplaced commas, and the like—not only my own, but those that have resulted from the typesetting process, because that introduces new mistakes into the text in the form of transcribed words, missed lines, etc.—but I’m also trying to retain all the elements of plot and characters in my head to ensure that the novel is consistent not only grammatically, but thematically. Checking for textual errors and looking for structural ones are two very different tasks, but there simply isn’t time to do one followed by the other. The author has to undertake both simultaneously. It’s exhausting, and really one can only examine about thirty pages or so a day while maintaining full concentration. It also has a curiously debilitating effect on a writer’s confidence, on both the micro and macro levels: as every comma, every choice of word, becomes potentially open to question, the writer begins to doubt whether the book itself has any worth.

And remember: all this is usually going on after the writer has already started his or her next book, which means putting that on hold and thinking oneself back into a book that one might have completed many months before, while also either continuing to write the work in progress or, as is more likely, putting it on hold for a few weeks and hoping that it won’t have slipped away upon one’s return.

Then there’s this: when the book is typeset, one begins to examine it in a different manner. Because it is now set in the manner of finished text, one starts to spot mistakes that slipped through during the copy-edit, when the book was presented in its original manuscript form. In my experience, I make more changes to the proofs than I do to the copy-edited manuscript—sometimes multiple alterations per page.

And that costs money. The equation used to be that changing a line at proof stage cost about $1, so rewriting a page would cost $300, give or take. (It’s probably less now, given the changes in technology, but significant alterations to a manuscript will still be expensive.) Ideally, then, if the author changes one word for another, the new word should have the same number of letters as the old, so that only a single line needs to be reset. Longer changes may affect the next line, which could mean resetting a paragraph, or page—or God forbid, an entire chapter. Publishers generally agree to absorb the cost of changes to a total of about ten percent of the proofs. After that, it comes out of the author’s pocket. Alterations, therefore, are often as much a question of mathematics as anything else, and the thesaurus becomes a much-loved friend.

When I had finished my readthrough of A Book of Bones, I followed it by going through the errors spotted, and changes suggested, by Jennie, Clio, and Clair, and adjusted the by now heavily annotated proofs still further. Eventually—and very close to deadline, as is typical given the timescale involved—the proofs were returned to the publisher, and a new set will now be made. At the end of this month I’ll have four days to compare my photocopied set of the original proofs with the new set, in order to make sure that my notes were clear, and have gone through without a hiccup, which is unlikely. (In the US, that second set of proofs will usually be read by another proofreader employed by the publisher, but I’ll also have a chance to look at them again.)

A Book of Bones is 700 pages long, equating to more than 200,000 words. With each word averaging five characters, that’s over 1,000,000 characters, not including punctuation marks. In other words, that’s significantly more than 1,000,000 opportunities for errors to creep in; and creep in they will, despite the best efforts of everyone involved—although even if the casual reader were still to find a handful of mistakes in A Book of Bones, it would represent a margin of error with which many scientists would be content, let alone many writers.

It’s the fallible nature of all human endeavour, and none of us is immune. Even the late John Updike, who was meticulous to a fault, once remarked that the first thing he spotted upon opening a random page in any new book of his was the error. Consider it the finger of God tapping gently on one’s shoulder as a reminder that perfection is for Him alone.

So if you do spot an error in a writer’s book, be kind. (The Internet makes it easier than ever to be unkind, and we should resist that urge.) Drop a discreet email to the writer, and the mistake will almost certainly be rectified in the next edition. But perhaps add a few words of encouragement to your message, because I guarantee you that the writer concerned will be sitting with head in hands, overcome by a sadness beyond all knowing at this latest flaw.

Because we tyr, we really do.

Oh, hang on…