Common sense is one of what I call the “Ten Principles for Creating Better Indexes.” Those principles (or characteristics) are accuracy, audience and access, clarity, common sense, comprehensiveness, conciseness, consistency, metatopic and structure, readability, and reflexivity. While all of these principles are finely interwoven, it is possible to disentangle them a bit, examine what each means, and apply them to our daily indexing work. This is the second essay in a series in which I’ll briefly consider each principle, explore some related tools for the indexer’s toolbox, and provide some sources for further consideration.
What is common sense?
Webster’s Collegiate Dictionary (10th ed.) defines common sense as (1) “the unreflective opinions of ordinary people”; and (2) “sound and prudent but often unsophisticated judgment.” I take umbrage at that definition, to be honest, and I suspect Thomas Paine (Common Sense) might as well. I would define common sense as that sound, everyday smart thinking applied in any situation. It’s what we apply when we have something that’s not working (or working too well, in the case of a baby’s lungs), and we check the simple stuff first: Is the television plugged in? Does the baby have a wet diaper?
We must also, I argue, apply common sense to our indexing. But the common sense we should apply to indexing is balanced with the other principles noted above. Then our indexes will be examples of complex usability and simple elegance.
Term choices and natural language
One of the topics indexers like to discuss is bias in indexing, whether applied by a publisher or an indexer. Bias has to do with what terms are utilized in the index, particularly for main headings. Can the user find Controversial Issue A in the index? Or is it buried under some innocuous concept?
There’s a great story related to this idea in Nudge (Thaler and Sunstein): The director of food services for a large school district discovered that students’ choices in the cafeteria line were influenced by the placement of food. That is, if fruits and veggies were placed in front and closer to the students, while French fries and less healthy options were farther back and less reachable, students were more likely to choose healthier options. Thaler and Sunstein call the food services director a “choice architect.” They also discuss the wider implications of such choice architecture for individual choice, which makes fascinating reading.
What struck me about the concept was this: Indexers are choice architects!
We make choices in every word that goes into an index. We decide how something obscure might be made more accessible. Or not. It’s important to stop and think about the choices we make. Are they the choices that users might make as they search for something?
This all indicates to me that we should use natural, everyday (commonsense) language in our indexes. That is not to say that we should “dumb” our language down. Rather, we must be aware of jargon and translate it as much as possible, so our indexes are usable across a variety of audiences. That might mean using cross-references to get users to the jargon term (or brand-new term, in some cases) and/or providing some meaning in a parenthetical qualifier. Thus, as choice architects, we construct entry arrays and the encompassing index structure in a way that makes the information findable via a variety of routes.
How do we manage to do that? We digest (or analyze) the text. A simple regurgitation of the text into alphabetical order generally isn’t all that useful. I am not the first person to use the digestion analogy. In The Case for Books, Robert Darnton quotes William Drake (an avid reader from the mid-1600s): “The meat which we have taken, so long as it swimmeth whole in our stomachs, is a burden, but when it changeth from that which it was, then at length it turns into strength and nourishment. The same let us do in our reading books. . . . let us endeavor to digest and concoct them.” Thus, Drake argued for extracting the essence from books. As indexers, we digest the text and concoct the index.
While we need to use the terms of the text as much as possible, we must often pull from outside the text for a useful main heading to capture something from within the text. Natural language is key to maintaining the connection between the outside world of the user and the inside world of the text.
I’ve heard a few indexers insist that we must limit ourselves to using only the terms of the text. That is poppycock, to use a polite phrase. Index users, as noted above, come to the index from outside the text. That is, they are not likely to be steeped in the thought processes of the author. In other words, the index is the go-between to get users into the text.
In order to create the best possible main headings, it is also helpful to have a good general grasp of the main subject of the text (or metatopic). It’s not that experienced indexers can’t create a reasonable index for a subject that’s unfamiliar, but it makes a substantive difference for the quality of the index overall if the indexer has that basic knowledge.
Many techniques are applicable to creating subheadings that make sense, including the following three. First, the relationship between the subheading and the main heading must be absolutely clear. Consider this: “animals: hunting, xx.” Is this animals hunting prey? Or humans hunting animals? Obviously, it makes a big difference! That’s an actual example from a recent index review I did. And the meaning of the subheading wasn’t made any more clear from other subheadings, which included references to pets. To repeat, the relationship between the main heading and the subheading must be clear and instantly obvious to a wide variety of audiences. My sense is that if something isn’t clear, users will pass on to something else or give up altogether.
Second, subheadings should, as far as possible, have the most important word first, so that it sorts on that (aside from prepositions, etc.). Here’s another example from that index review: “Century Magazine: publication of Grant’s war articles” (there were other subs, too). The subheading introduces redundancy and buries the important name in the middle!
Third, indexers may also apply logic to the sort order of subheadings, whereby, for example, arrest and conviction come before execution. It doesn’t always work: birth, career, death, finances, illness. But it’s fun to play around with the wording and see how much the alphabet can work along with logic.
Double-posting acronym-type entries
Common sense is aptly applied to double-posting as well. Consider this entry: “EULEX (European Union’s rule-of-law mission), 246.” Best practice is to double-post all acronym-type entries under both the acronym and the spelled out name, unless there are subheadings, in which case a cross-reference points from one to other. In the EULEX case, though, double-posting was not necessary because it sorted right before European Union. So it’s common sense not to create an unnecessary entry. But the best practice is also based on common sense, to avoid sending users in search of an alternative entry when the handful of locators could simply be placed in both locations.
When to break the rules
Common sense applies to breaking the rules, and the EULEX entry is an example of this. However, indexers must have really good arguments for breaking the rules. In an index on cancer trial studies, I used acronyms as preferred terms (rather than spelled-out names). My argument was that the text was so laden with acronyms that people would very likely look first for the acronym in the index. In addition, the acronym as preferred term allowed diseases and institutional entities to sort together in a more logical fashion. But that’s not typically my practice, which is to use the spelled-out name as the preferred term.
For a book on ethnographic writing, the index was fairly straightforward with entries for memoirs, person, place, structure, voice, and so on. The text included a section on the writing process. For that one entry array, I force sorted the subheadings to maintain the process flow. Thus, I broke the rule for one entry array in the index in order to maintain the logical order.
Common sense, though deprecated by its dictionary definition, is a key tool for indexers. It serves particularly as a balance to reflexivity and as an aid to clarity and readability. Common sense can also tip the scale when considering the best construction for main or subheadings and whether or not to create a double-post versus a cross-reference. We should apply common sense to create better, more usable indexes.
Darnton, Robert. The Case for Books: Past, Present, and Future. New York: PublicAffairs, 2009.
Thaler, Richard H., and Cass R. Sunstein. Nudge: Improving Decisions about Health, Wealth, and Happiness. Updated ed. New York, Penguin, 2009.