A well designed index can help instrument smarter ebooks, making it easier for readers to find & retrieve info more efficiently

Indexers are the offensive linemen of book publishing. No one notices their work until they screw up. The index has been on my mind lately, and not just because I cursed a (print) book for omitting a key word in its lookup list. (Tip: use Amazon’s Search Inside tool as a makeshift index.) I’ve also been having some enormously instructive exchanges with folks who do serious, in-the-trenches indexing work—special shout out to Jan Wright, Joshua Tallent, and Nancy Humphreys—and who grapple with today’s question: why has the ebook index gone AWOL?

I’ll get to some reasons in a moment, but first let’s consider why it is that people use indexes. Looking up a specific term, of course, is the biggie. You’re reading a book on illuminated manuscripts, say, and forget what the term “gloss” means. So you pop open the back pages to track it down. Or that zucchini that just rolled to the front of the fridge looks like it’s got one more day in it and you need a simple recipe, stat. The index in your favorite cookbook is your best bet. But beyond these simple retrieval tasks—which, after all, a good search tool is adequate for—don’t forget all those other reader services an index provides. It:

  • Includes concepts rather than just words. In Henry Aaron: The Last Hero you can, for example, find sub-entries under the Media listing on “racially biased coverage and stereotypes perpetuated in” (try searching for that!). For students, essay writers, and other serious readers, the ability to undertake thematic and concept-specific explorations of a book is hugely valuable.
  • Provides guided discovery. Consider the zucchini scenario I just mentioned. A well done print index is a perfect place to explore a cluster of related topics. By organizing zucchini recipes into different kinds (fried, broiled, steamed, etc.) a cookbook’s index helps recipe searchers make some high level decisions (fun or healthy?) before following the choices that await. The see also pointers provide similar help.
  • Helps when you know what you want, but aren’t sure how to describe it. For example, say you want to create multi-level bullet lists in Word. By heading to the entries on, say, outlines and lists readers can usually home in on the answer.
  • Signals depth of coverage. For example, readers know that the first entry in this listing—St. Cloud, 84-92, 172—contains more info than the second entry.
  • Provides a handy one-stop tally of coverage points throughout a book. Again, for students and scholars looking to review all mentions of a particular item, this can be a big help.
  • Gives tire-kickers a sense of the book’s coverage. Sure, the table of contents—not to mention plain page flipping—helps prospective buyers evaluate a book, but serious readers will sniff through an index to get a sense of what awaits.

In sum, an index is a kind of a collection of pre-made searches: rather than diving headlong and unawares into a search oval’s do-it-yourself void, an index presents would-be searchers with an already assembled, alphabetized list of the 500 or so most common query items. (Microsoft’s effort to brand Bing as a “decision engine” offers an apt analogy; index is to search as a decision engine is to a search engine.) Speaking of search: of course, the standard ebook search oval has its role in the world of digital books. But for the kinds of guided lookup missions listed above, it’s a poor substitute for an index.

In an age of info abundance, books that give readers more efficient ways to access what they need are books that are going to sell better.

So why, then, do the vast majority of ebooks today come without indexes? First, implementation challenges. A reader’s ability to adjust font size plays havoc with page references. Where, for example, in an ebook should these index entries…

Castro, Fidel

bear wrestling stories, 98

beard grooming tips, 62

…point to? At certain font sizes, those hyperlinked numbers might lead exactly where you want to go. Or you might have to page ahead—or back—a click or three to find it. Which direction? That way: <=>. Factor in varying screen sizes and you’re halfway up the indexer’s version of the Tower of Babel. Compounding matters, Adobe’s popular InDesign software strips out all the index markers when it creates an ePub file. HTML anchor points can be manually inserted in the ePub file but that’s time consuming. Given the extra time and money required to build an index that may frustrate as often as it satisfies, no wonder most publishers, when asked about indexes, point to search as a decent substitute.

It’s not. And worse: the absence of an index deepens reader prejudice regarding the value of ebooks. I don’t think it’s unfair to say that books that do less are worth less. It doesn’t have to be that way.

Not only can an ebook’s index match the utility of its print counterpart, it can improve upon shortcomings we’ve all learned to deal with. (Things like unindexed terms; having to traipse back and forth between index entry and target; and disorientation—that confusion you feel when landing amidst a long list of sub-entries, and you’re left trying to figure out where you are in the overall A-Z path.)

A well designed digital index, in sum, can be a key part of instrumenting smarter books, ones that help readers find and retrieve information more efficiently.

Alas, for the immediate future, indexes won’t likely be appearing in most ebooks on the main e-reading systems—Kindle, Nook, iBooks and so on. That’s because the companies behind these platforms haven’t programmed their reading software in a way that would make indexes easy—or at least easier—to implement. Amazon comes closest. Those page numbers you now see on some Kindle books appear thanks to the so-called “page list” data that many publishers provide—essentially lists correlating ebook location with pages in a print book. That data could, in theory, be used to automate the implementation of indexes and make them more accurate.

And yet even if the wonkery gods cooperate and, say, the Nook app sprouts an index button, victory, at least in the eye of index enthusiasts, won’t be at hand. The real opportunity lies in moving beyond thinking how to program ebooks to recreate page-based approaches to indexing. That is, book designers, indexers, publishers, and e-reading system manufacturers need to think deeply about that fundamental question I raised earlier—why do readers turn to the index?—and then pair those answers with the ways in which screen-based publications can service that need.

The answer surely won’t lie in simply taking a digital snapshot of a print index and turning each entry into a hyperlink. Instead, I suspect we’ll soon discover how search tools and indexes can work in partnership, passing queries from one to the other as different readers arrive with different questions. Some of these may be quite precise—where’s the first place that mitosis is defined?—and thus lend themselves well to a keyboard-based (or even spoken) lookup. Other searches that begin with broader questions—what does Tom Friedman’s book Hot, Flat, and Crowded have to say about the topic of agriculture—may benefit from the suggestive guidance that an index often provides.

Let me wrap up with a quick tour of my own vision for Index 2.0. The first question I tackled is: where do you put it? I think two places make sense. First, in the Go To menu which many readers consult when they want to navigate to specific sections in a book.

An ebook app's Go To menu with an option to visit the index

And I also think anyone who’s tapped, say, the Kindle app’s search icon (the magnifier glass shown in the previous drawing) is also someone who potentially might want the help of an index. But it depends; so why not let them decide?

The pop-up search box, with an option to head over to the index

Here the reader can either enter a search phrase, as usual, or tap “Browse Index instead?” to head over to the full index.

So what would would greet the reader who opted for the index browsing option? I played around with a bunch of different layout designs and ended up going with something that looks pretty familiar.

An ebook index in full browsing mode

Swipe left or right to page through the full collection of index entries; for quick navigation use the letter picker at bottom. Tap on any term to see the entry in full, shown in the next drawing.

The reason I ended up going with a more or less familiar three-column listing is that part of an index’s value lies in the awareness that all those adjacent terms provoke. Especially for those not completely sure of what they’re looking for, being able to see other entries can help you discover new and different paths into the book.

When the user taps an entry, a preview screen appears.

The index entry for the term "millet"

Here, she can scroll in the right column through a list of every occurence in the book and decide whether to save for later reading or view the full passage immediately. On the left side are some “See also” entries; these handy pointers can help readers in need of a gentle reminder that their term is part of a larger constellation of knowledge.

Finally, what about that partnership between the index and the search box that I mentioned earlier? Why not give people who are about to perform a search, a quick and simple option to head over to the index? That’s what this last figure demonstrates.

A pop-up search box, with results and an option to view the found term's index entry

Here’s a regular pop-up search window. The first entry offers an easy exit from Searchland to Indexville.

Via A New Kind of Book

6 COMMENTS

  1. Great thoughts, indeed, on an important topic, and one on which conclusions or accepted practice are yet to be established. I must admit, though, my first thought, on seeing the word “millet” is the confessional organization and administrative system of the Ottoman empire!

  2. Great post, thank you.

    We can actually produce pretty smart indexes now. For those publishers following a true XML book production workflow, rather than converting direct from InDesign to Epub, there are options to prepare embedded indexes. This in theory enables the index list to be properly hyperlinked, i.e. the reader is taken to the precise content location, not just the vague page area. It’s a new way of working but it has some great benefits in the production process too and helps in content reporpusing.

    I like your ideas about unifying search and the index too – with some smart thinking we could unleash some really user friendly navigation aids.

  3. Your idea for ebook indexing is interesting but you are thinking in terms of accessing the words in the text not the ideas. The index is more than a word search function. A good indexer points to the topics being discussed that are not explicitly named within the text. This is the problem with embedded indexing. One solution, as is done in legal documents, is to number each paragraph and use that as the locator. In this system, either the discussion of that topic begins here, or the term you are looking for, is found in this paragraph.

  4. Peter,
    These are some great thoughts on eBook indexing.
    My only fear is that it is going to take forever (if at all) for the various eBook device manufacturers to add the proper functionality in order to support these ideas.

    But I think you are really on to something here. And this discussion should absolutely continue, but on a much larger scale.

The TeleRead community values your civil and thoughtful comments. We use a cache, so expect a delay. Problems? E-mail newteleread@gmail.com.