Hello,

Last week was a strange work week for me. I had two indexing tasks to complete, neither of which I typically do or enjoy. The week felt like a long grind, like I wasn’t really indexing even though I was.

My first and primary task was to update an index for a revised edition of a book I originally indexed in 2022. Instead of rewriting the index from scratch, I agreed to update the original index, which ended up being more work than I anticipated. I still think updating the index was the right call, but still not much fun. I’ll write about this next time.

My second task, which arrived in my inbox unexpectedly, was to trim an index by about forty percent. Which is…a lot. It is hard to find forty percent of fluff that can be easily cut, which means that difficult decisions need to be made. Again, not my favourite task.

Thankfully, I don’t often need to make such drastic cuts to an index, but knowing how to cut, as well as knowing how to index to a tight length limit, are important skills. It helps to have a plan.

Today I am going to walk you through the steps I took to cut the index down to size.

​

Some Context

This was an index I completed back in June, for a sports biography. It was for a trade publisher I regularly index for and for whom space is rarely an issue. In hindsight, that was perhaps the problem, in that I was busy and assumed that space wouldn’t be an issue either for this particular index, and so I did not follow up to confirm. For the publisher’s part, my contact was also busy and didn’t think to let me know.

A lesson for me: always confirm, even if I am confident of the answer. (Because I could be wrong!)

And so I received an email last week informing me that although the index filled twelve typeset pages, there was only room for seven pages. Could I please trim the index?

Trimming a finished index is not always easy, especially if the cuts need to be deep. All of those entries were chosen for a reason and structured in a particular way to make the index easier to navigate and access the information. Now I need to undo all of that? Or find a different, shorter way to achieve the same goals?

There is also the emotional gut punch of having to dismantle work that I have spent time and effort constructing. Kind of like writers having to kill their darlings.

But if the index is too long and the publisher is not able or willing to add more pages to the book, then the index does need to be shortened. And as much as I don’t want to do it, I also know the index the best and can hopefully minimize the damage, compared to someone else trimming haphazardly.

So I said yes, and began planning how best to cut.

​

Trimming Process

Step 1: Determine how much to cut. Thankfully, I had the proofs which contained the typeset index, so I could see exactly how many lines were available per page and how many lines needed to be trimmed. While helpful to know that five out of twelve pages needed to be cut, that is also vague, so if possible I prefer to know the number of lines. Examining the proofs, I determined that the index took up about 988 lines, and that it needed to be trimmed down to 564 lines. I also determined that each line contained a maximum of 35 characters.

Step 2: Set the margins in Cindex. To help me replicate the length in Cindex, the software I use for indexing, I find it helpful to try and replicate the line length. To do that, I create a dummy entry that is 35 characters long and set my margins to fit that. I can then use the Index Statistics function to see how many lines the index contains. I don’t expect this to be an exact correspondence to how the index will appear in the proofs, as designers have tricks for manipulating the layout on their end, but I find it usually works well enough.

Step 3: Decide what needs to stay. Now that I know how much needs to be cut and I have a way to track my progress, it is time to decide what I am NOT going to cut. This may seem counterintuitive, but I do have 564 lines to play with. I don’t need to cut everything, and deciding what to save helps me to preserve what I consider to be the most important aspects. This being a sports biography, the subject obviously needs to remain. I also decided to prioritize the team that this person was primarily associated with, as well as a few other key individuals and themes that played prominent roles.

Step 4: The first pass. I scan the index looking for any entries that seem like immediate candidates for trimming. This includes double-posts and cross-references that are nice to have but perhaps redundant. This also includes entries for information that are interesting but not central to the main subject of the book. I personally prefer larger indexes with lots of detail, but when space is tight, I need to narrow the focus. This first pass trimmed about ninety lines, so almost a quarter of the way there.

Step 5: Remove subheadings. Best practice is to include subheadings if there are more than 6-10 locators; the cutoff is somewhat a matter of personal preference and dependent on the audience, type of book, and space for the index. That said, if space is tight, my preference is to remove subheadings in order to squeeze in more arrays. So, I identify about a dozen arrays, mostly people, and remove the subheadings, resulting in some arrays with 10-18 locators each. Not ideal, but it does cut down on a few lines.

Step 6: Review major arrays. Though I have removed some subheadings, I haven’t removed all subheadings. Some arrays are too large and too central, and I don’t want arrays which contain 20+ undifferentiated locators. But for these major arrays, I still review the subheadings and see which I can remove and consolidate.

Step 7: The second pass. This, in my opinion, is the most difficult pass. The obvious cuts have largely been made and the index shrunk through removing and consolidating subheadings. But I still need to find another two hundred lines to cut. Being a sports biography, there are a lot of people and teams mentioned. My inclination is to pick these up, for the sports fan eager to find their favourite team or player, but there is no longer enough space for everyone. Going back through the index, I look for arrays with only one or two locators. I quickly search the proofs to see how each person or team to mentioned, and decide whether or not they should go. If you attended the ISC/SCI conference in Vancouver, this is an excellent time to apply Emma Warnken Johnson’s passing mentions flowchart. Having one through the entire index, I realize I’ve actually cut about twenty lines more than I need, and so I go back and add a few arrays back in.

​

Debrief

I have to admit, at the end, that the cutting starts to feel arbitrary. Is this player, with a minor mention, really more significant than that player? Which do I prioritize, this player or that coach? I honestly have no idea. I think it helps to be consistent in making and applying decisions, so that the resulting index still holds together, but at this late stage, counting lines, the simple truth is that some entries get to stay and others have to go. I need to accept that the index is good enough, so long as it fits.

And the index does fit. It is cut from about 988 lines down to 564 lines. The index is not as comprehensive or as user-friendly as I would like, but the essential elements are all still present and I’ve squeezed in as many other entries as I could.

It took me about two and a half hours to cut the index. I did not charge the publisher extra. I feel partly responsible because I should have asked, when I first received the proofs, about the space for the index. This is also a longtime client that I want to maintain a good relationship with and for most projects they don’t request revisions, so if the occasional index goes sideways I figure I can afford a few extra hours.

I genuinely hope you never need to cut an index almost by half. It’s not enjoyable. But it is easier with a plan. Even if I only needed to cut the index by ten percent, it still helps to have a plan, to have some definition in mind for what no longer makes the cut. I hope my example here gives you some ideas.

Yours in indexing,

Stephen