Book Indexes: Multivolume Indexes

Ælfwine Mischler

Last year, I had the pleasure of indexing the third and final volume of a history of Egyptology while creating a combined index of volumes 1–3. (I confess that I have a bit of a soft spot for this book. Volume 1 was my first paid index — and a complicated text for a first-timer — and I was thrilled that the author included me in the acknowledgments.)

When I indexed volumes 1 and 2, the publisher had not thought of having a combined index in the last volume, so I did nothing out of the ordinary in indexing the first two books. When the publisher asked for a combined index, I asked colleagues for any tips or tricks, and they alerted me that it would be a lot more work than just merging the first two files into the third. They were not joking! (Fortunately, I was able to negotiate a higher per-page price.)

More Editing

The publisher gave me PDFs of the final indexes for volumes 1 and 2, and I compared these carefully with the indexes I had written. I wanted to see any changes the publisher had made and refresh my memory of both the subjects I had indexed and their organization.

In my indexing software (I use Sky Indexing), I made a copy of each volume’s index and entered the publisher’s edits, and then increased the locator numbers (a locator is a page or a range of pages) by 1000 in volume 1 and by 2000 in volume 2. Thus, for example, page 35 in volume 1 became 1035 and page 35 in volume 2 became 2035. I then merged these into a new file, in which I indexed volume 3. I changed the page numbers to the correct forms with volume numbers as a final step so I would not have to type 3: before every locator for the new items.

The real extra work came in creating and organizing subentries. Many entries in volumes 1 and 2 had only a few locators without subentries. When the indexes were combined, these entries had too many locators and I had to make subentries. This required going back into the PDFs for those volumes and rereading those pages.

Other entries in a single volume had subentries, but there were so many in the combined index that they became unwieldy. I reworded some subentries to combine them, but more often, I put the subentries into broad categories and split them into nested entries.

Two of the great names in Egyptology illustrate this editing process.

Howard Carter, who discovered Tutankhamun’s tomb in 1922, had only five locators in the index of volume 1 (covering from antiquity to 1881), with no subentries. In volume 2 (covering from 1881–1941), he had 14 subentries. In volume 3, the discovery of Tutankhamun covered 40 pages in two chapters, and in the combined index, Carter was in two nested entries:

Sir Flinders Petrie also appeared in all three volumes. In the volume 1 index, he had seven locators with no subentries. (In volume 1, which covered a much longer span of time, the managing editor and I agreed to use longer strings of locators to save space.) In volume 2, there were 26 subentries for Petrie. In volume 3, the Petrie entry was nested to break the subentries into broad categories:

I made some other changes in the combined index. Many of the big names had a subentry “career” or “early career” or “legacy.” These were all force-sorted as the first subentry under the name. Volume 1 discussed many books. I reviewed these entries and removed some from the index that were mentioned with little or no discussion. This was relatively easy to do in the indexing software because I could group all the records that had italics in the main entry. If a book had only one locator, I reread the page in the PDF. Sometimes there was sufficient discussion to keep the title in the index. In addition to these smaller edits, I reorganized some of the large entries.

When I was finished with the editing, I changed the locator numbers to volume and correct page number — an easy task in the software.

This long, complicated index needed a final check. For this, I generated a page proof in numbered order. (This option may not be available in all indexing software.) I went through the page proof line by line. This allowed me to check that double-posted items were correct; for example, that 2:17–20 appeared in both “Petrie, Sir William Flinders, methods and techniques of: excavation” and “methods and techniques of archaeologists: of Petrie.”

With a Heads-up

In this situation, I did not know that a combined index would be required in the last volume of the series when I worked on the first two volumes. What if I had a heads-up on another project? What would I do differently in indexing the early volumes?

I would create subentries for anything that was likely to appear in the following volumes, even if it did not require subentries in the current volume. When I was finished editing the index with the extraneous subentries, I would suppress them in the current index, saving them for the later combined index.

This could be done in one of two ways. I could save the index with a different name, and then in the new one, consume the extraneous subentries, that is, remove the subentries but retain the locators, which my software can easily do. When I made the combined index, I would merge the file with the subentries into the new file.

Or I could duplicate each of the entries with extraneous subentries in one file, label them with a color code and filter them out, and then consume the subentries in the unfiltered records. To make the combined index, I would unfilter the records with subentries.

Either way, the combined index would still be more work than a single-volume index, and I would charge a higher rate. A combined index is more than the sum of its parts. Be aware of this if you are either of the parties negotiating for such an index.

Ælfwine Mischler is an American copyeditor and indexer in Cairo, Egypt, who has been the head copyeditor at a large Islamic website and a senior editor for an EFL textbook publisher. She often edits and indexes books on Islamic studies, Middle East studies, and Egyptology.

Advertisement

Book Indexes: Part 8 — More Lessons Learned in Using DEXembed for the First Time

By Ælfwine Mischler

In Part 7 of this series on book indexes, I wrote about some of the things I learned about embedded indexing when I used DEXembed, an add-in for Word available from Editorium.com, for the first time. In this post, I continue with lessons learned.

You Did What?

I was using paragraph numbers for locators. (DEXembed can also use word number or top of the page, but these are generally less satisfactory.) Before I started indexing, I told the author repeatedly that the paragraphing had to remain stable and he could not add new text, so I was none too pleased when he sent a new paragraph to be added in the middle after I was already three-quarters through the book. I asked DEXembed developer Jack Lyon whether I could just remove the paragraph numbers, add the new text, renumber the paragraph numbers, and move on, and what would happen to the bookmarks I had made.

He did not know the answers but suggested I do a test run on a copy of the file. I removed the paragraph numbers, added the new text, and told DEXembed to number the paragraphs again. It ground away for a minute or two and then told me they were numbered, but I saw no red numbers to the left of the paragraphs — until I scrolled down and saw that the numbering started with 1 a few paragraphs after the new text.

I removed the numbers, saved the file as another name, and tried again to number the paragraphs. This time it worked. Then I had to go into my Sky indexing file to correct the locator numbers after the inserted text.

We Do Not All Count Alike

DEXembed allows you to go to a particular paragraph using Ctrl + Alt + P. However, whenever I entered a paragraph number, DEXembed would select a paragraph before the one I wanted: If I entered Ctrl + Alt + 100, it would take me to 99.

To create a range, I put the cursor in the first paragraph of the topic, clicked Ctrl + Alt + M (to mark the beginning of the topic), put the cursor in the last paragraph, and clicked Ctrl + Alt + L (to get the locator). DEXembed created a bookmark and, because I had opted so, put a comment with the beginning and end paragraph numbers — but they were always off. If I selected paragraphs 225–245, the comment would say 226–246. I was not certain which range I should put in the indexing software — the one I had selected or the one in the comment. I used the actual selected range, which later proved to be the correct decision. What was causing this discrepancy, and would it cause a problem later on?

Yes, it did cause a problem when I embedded the entries. The XE field codes went into the chapter section headings instead of into the paragraphs where they belonged. Back to the developer. He thought a paragraph had been deleted and told me how to find the problem. If Ctrl + Alt + P takes you to the paragraph numbered 10 when you specify 10, but it takes you to the paragraph numbered 499 when you specify 500, you know the problem is between paragraphs 10 and 500. Keep selecting paragraphs every 100 or so (Ctrl + Alt + 100, Ctrl + Alt + 200, etc.) until you find the source of the problem.

In fact, I learned that an extra Enter, not a deleted one, was causing the faulty numbering, which began at the start of the indexable material, after the preface. Following the preface was an Enter and a Page Break. When I deleted the extra Enter, the DEXembed took me to the correct paragraphs.

To be safe, I checked every 50 paragraphs throughout the text and found another place where I had accidentally added an Enter in the middle of a word. Once that was deleted, the numbering problem was solved and the XE field codes went into the correct places.

Work, Work, Work

Embedded indexes are a lot more work for the indexer, who will typically charge more for this version than for a back-of-the-book index. The most trying time is when the actual embedding takes place. At this point, everything is up to Word and there is nothing the indexer can do but hope the embedding works. I kept getting error messages half-way through the embedding, until the developer told me to reboot. I found it also useful to close every other application and the browser while embedding.

If I found an error in the generated index, I had to correct it in Sky, generate the index from Sky as a text file, embed it in another copy of the book (perhaps rebooting first, and praying all the while), generate the index in Word, copy that to another Word file, and make any manual corrections required by Word. (See my December 2017 post.)

Because I was new to using DEXembed, I had to repeat this process several times. Fortunately, I do learn from my mistakes, so the next such project should be easier.

Generally, DEXembed was easy to use, and the developer was helpful in answering my questions. I liked using paragraph numbers for locators and limiting the use of ranges. However, because of Word’s limitations, DEXembed requires an extra step to index footnotes or endnotes, which I did not need to do for this project and did not test. Editorium’s NoteStripper will strip the notes out as regular text. You then number the paragraphs and write the index. After embedding the entries, you use NoteStripper to turn the notes back into regular Word notes. (Note that this is not the same version of NoteStripper that came with Editor’s ToolKit Plus.) One more thing for me to test drive!

Ælfwine Mischler is an American copyeditor and indexer in Cairo, Egypt, who has been the head copyeditor at a large Islamic website and a senior editor for an EFL textbook publisher. She often edits and indexes books on Islamic studies, Middle East studies, and Egyptology.