Why you should build your own goddamned book index

If you write a non-fiction book, you’re not done until you’ve proofed the galley pages and arranged for an index to be composed. Unless you have lucked into some kind of contractual arrangement I’ve never heard of that leaves indexing to the publisher, you’re either going to have to create the index yourself or hire someone to do it for you. I think it’s safe to say that most authors don’t enjoy indexing, which is altogether different from appreciating indexing. After publishing more than twenty works of non-fiction, I’m here to tell you that indexing is good for you. Doing it is time-consuming, unpaid labour, but doing it well will improve the book.

Publishers routinely build about three weeks into the production process for the author to proof the galleys and create an index. That’s not a lot of time in which to re-read a copyedited, typeset manuscript and create an index for it. You may find yourself too pressed for time in other duties to pull off both, and end up turning to a professional indexer. In my long publishing career (I published my first book in 1981), I think I’ve hired an indexer twice. One was so-so. The other, a fellow graduate student, was superb. (Said fellow student then refused to be paid. You know who you are, you goofball.) 

I’m not sure how much indexers demand these days, although I was told recently that a budget of $4 per galley page would be ballpark for a thorough one. For my most recent book, Jackson’s Wars, I debated hiring an indexer because I was concerned about how much time I would have available. In the end, I built the index myself, and I am so, so glad I did, and not (only) because I saved a whack of money. 

Even before indexing Jackson’s Wars, I knew that indexing does double-duty as a last-gasp fact/quality-control check. Index-building is the one sure-fire way to drill down into a completed manuscript and find problems that have eluded you, the editor, and the copyeditor. Hopefully, you won’t find many, but you’ll be grateful for the ones that you do find, and that you have a chance to fix them.

My chief nemesis in writing—in life, in fact— is proper names. It’s like a mental stutter. I routinely suffer a brain cramp when trying to summon one in everyday conversation, so much so that when I have to do a media interview on something I’ve written, I write out a list of names that I think I’m going to have to recall, for quick reference. (No, it’s not early onset dementia. I’ve been having this problem since at least my teens, when I stood up at a regatta awards ceremony to accept a trophy and could not recall the name of one of my crew.) In writing, I do crazy things with names. I’ll turn a Harold into a Howard in some part of the manuscript, a Susan into a Sarah. Indexing is when I save myself from these errors. Typically they involve a very secondary figure that only pops up a few times, and the instances might be separated by several chapters. Louis Grier, the brother of E. Wyly Grier, only appears twice in Jackson’s Wars, in two widely separated places. On first reference, I called him Julian Grier. I have no idea why, but he did attend the Académie Julian in Paris. I took another minor character, Percy Robinson, for a spin for one paragraph as Percy Robertson.

But wait: shouldn’t copyediting have save me from these errors? Well, yes and no. In a perfect world, a copyeditor would build a style sheet with every proper name referenced, to ensure accuracy and consistency. But that’s a huge job, and as an author you shouldn’t count on it happening in a time-pressed publishing process. As the author, you could build such a style sheet yourself, and that’s actually a good first step anyway in composing an index, which is why you should try to build your own.

The operative assumption when building an index is that the indexer should find mistakes, because they are indexing the same set of uncorrected pages that you are proofing, and no book emerges from proofing without corrections. The problem with using a paid indexer is that, depending on how thorough they are, they may not catch these errors. The quick-and-dirty way to build an index is to start at page 1 of a pdf proof and build a list of names and places as you scroll through, then use the search function to find every example. If an indexer doesn’t read the book first, and become familiar with what it’s about, they won’t have any cause to wonder if the person with the last name Grier who is named Julian at the start of the book and pops up much later as Louis in fact might be the same person. They’ll just assign page numbers to each reference.

How to build a truly useful index is another subject altogether, but I usually give myself two or three weeks. Doing it properly is a big job. The main reason is that a proper index isn’t simply a catalogue of people and things. And even if it were, the rule of thumb of good indexing is that there shouldn’t be more than six page references for an entry; after that, you need sub-entries, or different entries altogether. An index also ought to capture ideas and themes, and deciding what to index, and how, is an artisanal process. People have said to me, “I don’t know why someone hasn’t designed an algorithm that can do that job.” So far, they haven’t, and I don’t know how they could. Every book index I have produced has been a different sort of beast from the last one. Indexing Jackson’s Wars, which is both lengthy and revolves around the life of one person, was particularly tricky, and it took me a while to figure out how to structure the entries without having the index become mostly an entry for A.Y. Jackson with an enormous string of unnavigable sub-entries. Can everything associated with “Jackson, A.Y.” possibly be covered under one entry, no matter how many sub-entries are used? Should what A.Y. Jackson thought about Tom Thomson be a sub-entry under Jackson, or Thomson? Both would seem a waste of space. Should there be an index entry dedicated just to his artwork? And is there so much artwork that there really should be two entries, one for his works in peacetime, the other for his works as a war artist? Would a chronological entry under his name be a useful guide to the book’s contents? If I have an entry on “conscription,” should British conscription be separate from Canadian conscription? And should Jackson’s attitudes to conscription be a sub-entry under conscription, or under Jackson? (I’ll leave you to examine the published index to find the answer to these scintillating questions.)

Which leads me to address the basic question of why a book has an index. I dislike the idea of an index being a substitute for not having to read a book in its entirety, if your intent on writing it was to have people actually read it, and not just dip in and out of it for handy facts. But the most basic use of an index is probably someone (like you or me) wanting to see if a book has anything useful to them inside it. That’s a good reason for having your index address themes and ideas, rather than just names, because your net readership may depend on how well your index reflects what’s inside the book. But indexes are also of great assistance after someone has read a book. “I’m sure I read something about this person or that subject in that book,” you will think, and the index will confirm that for you and lead you to it. Frankly, I rely on the indexes of my own books to find things I wrote.

No publisher wants you making significant changes at the page proofing stage (and changes that cause reflow play havoc with indexing). But indexing your book should be embraced as an opportunity to do a last, deep dive into its contents. It’s your final chance to make sure not only that you have spelled some minor figure’s name consistently, but also that something hasn’t slipped past everyone involved in the complex process of turning a manuscript into a book. Indexing has a way of flushing out those errors in a way that proofreading never does. And for me, indexing has a way of stitching together themes, observations, and arguments in a concrete way that reassures me I have addressed them in the proper places.

Update: people are already wondering how I can devote two to three weeks to building an index when a publisher only allows about three weeks for both page proofing and indexing. Basically, by doing a lot of advance prep. I build a list of proper names and thematic stuff before the page proofs arrive, to give me a healthy head start on the indexing. Depending on the manuscript size, that can take a few days. For proofing, I reread the edited manuscript several times, flagging anything I want to fix in the proofs, before they arrive. I probably then devote about a week to proofreading and correcting the pages and two weeks to building the index. But the indexing process, as I have noted, will turn up things I need to fix, above and beyond the page proofing.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Create a website or blog at WordPress.com

Up ↑

%d bloggers like this: