British Standard ISO 999:1996 defines an index as “a systematic arrangement of entries designed to enable users to locate information in a document”. I find this interesting merely in the fact that there is an international standard definition.

When I was younger I made quite a few indexes. It was handy freelance work that could be done in the evenings. Of course making the index is often subject to schedule constraints, though in those days, when we used to give ourselves a year (or more) to get a book from raw manuscript to printed book, this didn’t often pose a huge problem. Mostly the indexes I made were for academic books, but I did hook up with an agency which involved me in a few trade titles. A trade book on naval battles was surprisingly dull work — probably the book was rather straightforward in structure. The interest of indexing lies largely in the discovering of the right approach to this particular book. Obviously names go into the index, and sometimes you’ll be asked to make separate indexes of names and of subjects. But it’s the subject indexing which contains such interest as there is in the indexing process. A side benefit is of course that you get to read a book while compiling its index, and often the book can be quite interesting.

The index is (almost always) the responsibility of the author of the book, something which will be specified in the contract. If the author doesn’t want to do the job however, the publisher will subcontract the work out to one of their team of freelance indexers, and debit the cost to the author’s royalty account. At Cambridge we used to rely on Mrs Anderson, a local indexing celebrity, who we kept pretty busy. Mrs Anderson wrote Book Indexing one of the “Cambridge Authors’ and Printers’ Guides” (C.U.P 1971) than which there is no better vade mecum for the indexer — though Mrs Anderson does dedicate her booklet to G. V. Carey author of Making an Index (C.U.P. 1951) a book with which I am not familiar. Both of them are, I am confident, more exhaustive than the first hit on a current Google search for “Making an index”, a more recent instruction sheet from Cambridge.

Mrs Anderson describes the function of indexing as “analytic, breaking down the contents of a book into small sub-divisions, and rearranging them alphabetically”. Nowadays there are fruitful short cuts offered by software as basic as Word, but back then you started out with a stack of blank index cards, or slips of paper, and with pencil in hand started reading the book. On page 2 perhaps you come upon a mention of the Paris sewage works. Do you need to index this? It may be the only time in the book it’s mentioned, being ultimately insignificant to the book’s theme. However by page 2 as you’ve got no idea what the probabilities of this are, you have to index it. So you take a card and write “Sewage works, Paris, 2” on it and read on. If you fail to make a card at this point and later find that say chapter six contains a vigorous description of the debate over the establishment of sewage works in France, you’ll have to go back to the start of the book (if you are lucky enough to remember that there was a previous mention of this subject) and start reading again to locate the earlier reference you missed. On the other hand, if the topic never comes up again, you’ll probably decide to discard that index card when you get to the end and “edit” the index. This is of course one reason the author is the ideal indexer — if they don’t know what comes later who does?

As you work your way into the book you will accumulate many cards which you store in alphabetical order in a cardboard shoebox. You may discover you need to expand  some of your cards, so that they contain sub-entries. Probably you’ll start to think about sub-entries when your card contains half a dozen page references. At that point you’ll need to go back over the pages you’ve referenced to see what an appropriate sub-entry or sub-entries might be. You might end up with “sewage works: construction; financing; planning; protests against;” each with its own clutch of page references. Maybe the Paris qualifier will end up redundant, or perhaps you’ll have extra cards all with the same subheadings, one for “Sewage works, Paris”, one for “Sewage works, Amiens” and so on. The point is you can’t know this ahead of time — by making the index, you discover what form the index should ideally take.

When you have completed the indexing, discarded any unnecessary single-entry cards, and created any new sub-entries that look like they might be desirable, you need to type out the index and submit it to the publisher. You probably won’t be called upon to read a proof of the index: there will probably not be time. Surprisingly perhaps there’s room for debate about how to alphabetize — is it “BBC, bagpipes, Bath, beacons” or “bagpipes, Bath, BBC, beacons”? — Depends on whether you are alphabetizing by word or by letter. In England those Scottish prefixes Mac, Mc and M’ are all alphabetized as if they were all spelled Mac, while in America they would be sorted as spelled. In a way it doesn’t matter where you come down on these vexed questions just as long as you are consistent; though one should aim to keep any system consistent with other parts of the book.

Some readers may have noticed that I recently created an index of all posts to this blog (see the bar at the top of the page above the title). This is a fairly crude alphabetical listing of the titles of the posts. I won’t promise to make the index more analytical, but at the same time I won’t rule out the possibility of doing so.