Alec Nevala-Lee

Thoughts on art, creativity, and the writing life.

Character counts

with 3 comments

Novel spreadsheet with chapter lengths

A couple of years ago, I published a post called “Writing by Numbers,” in which I described the unconventional approach I took to editing my novel Eternal Empire. I’d already made several passes through the manuscript with an eye to cuts, but it was still nowhere close to 100,000 words, which was the maximum I’d contracted to deliver. What I did, in the end, was make a spreadsheet in which I recorded the length of each chapter and how much it fell short or exceeded the average. Chapters that ran long for no good reason became targets for pruning. (I was especially hard on transitional chapters, which included material that was necessary to advance the story, but which were relatively sedate compared to the central set pieces and action scenes.) And it all sort of worked. The manuscript shed several thousand words, a line or two at a time, and I still credit this quantitative approach with guiding my scalpel to the right places. In all likelihood, if I’d tackled the edit more intuitively, I would have cut most of it anyway, but the numbers gave me the push I needed.

In retrospect, though, I’ve concluded that the numbers weren’t the key factor here. The spreadsheet was less important in itself than as a kind of conceptual screen, a way of regarding a familiar manuscript from a new angle. Revision hinges on the ability to read your own work as it if had been written by a stranger, or, as Zadie Smith says, even an enemy, and nearly every relevant strategy has this end result in mind. The simplest way to get some distance is to take some time off—ideally four to six weeks, and it’s even better if you’ve been working on an unrelated project in the meantime. Changing the typeface, font size, or margins has a similarly alienating effect, although I’m rarely brave enough to go that far. The same is true of reading the work in a different setting, out loud, or on paper rather than on your laptop. And the quantitative approach has an analogous effect: it directs your attention to areas of the story you might never have noticed if you were merely reading through it with an author’s eye. Any target word count is inherently arbitrary, but that’s exactly why it works.

A page from my rough draft

I got to thinking about this again after a friend recommended that I read the thoughts of ecologist Stephen Heard on the subject of revision. Heard is speaking to an audience of academic writers, and his advice has more to do with submitting journal articles for review than with writing a novel, but many of his points are still valuable. He covers some of the same tips that I mention above—reviewing your work in a different font, in a different location, or even at a time of day when you’re tired—and he has a particularly interesting take on revising for length: “Manuscript lengths are most often expressed as word counts, but I suggest you work with character counts instead, because replacing long words with short ones is just as helpful to the reader as reducing the number of words.” This takes the quantitative approach to the extreme: the story is no longer a series of pages, or even individual words, but a string of characters, each of which has equal weight when it comes to reducing the length. I haven’t tried it yet, but I have a feeling that it might turn out to be a useful tool, especially in nonfiction, even if its effects are close to subliminal. 

Obviously, an approach like this might be more practical in reworking a 2,000-word essay than a 100,000-word novel, for which it feels a little like digging a hole through a cliff with a needle, as the Humbug does in The Phantom Tollbooth. But while it might not be entirely feasible for longer works, it can’t help but make you more aware about the average length of your words, which can only be a good thing. It reminds me of an analogy given by the great Christopher Alexander, who describes how we can check whether a metal surface is smooth by inking a standard block and rubbing it against the face we’re testing:

If our metal face is not quite level, ink marks appear on it at those points which are higher than the rest. We grind these high spots, and try to fit it against the block again. The face is level when it fits the block perfectly, so that there are no high spots which stand out any more.

What’s nice about this approach is that the evidence is unambiguous—the marks show you exactly where the surface needs grinding. In fiction, the result doesn’t need to be a uniform face: all good novels have peaks and valleys, alternating rhythms, and variety of pacing. But the first step is to figure out what you have. And as in most things in life, the numbers can reveal patterns that the eye alone never could.

Written by nevalalee

January 28, 2015 at 9:21 am

3 Responses

Subscribe to comments with RSS.

  1. I recall one writer explaining that she used her word processor to calculate the Flesch reading ease for each chapter, then tackled the ones that were high. (She was a young adult author, so reading level was a key metric for her, but it could still have some value in a novel for grown ups…) She said this nearly always resulted in fewer, punchier words remaining.

    I guess these are partly means of find a way into a problem. Somewhere to start.

    As far as reviewing in a different font, and talking about academic work: I often use LaTeX for academic work ( which is (sort of) a mark up language. That means form and content are largely separated and you can edit it in any text editor or word processor, and this sometimes is useful in getting a fresh view of the text.


    January 28, 2015 at 7:06 pm

  2. I’ve always been intrigued by TeX/LaTeX, mostly because some of the books written in it—like Knuth’s Art of Computer Programming—are so beautiful. And the Flesch approach might be even more practical than character counts for longer works. I’ll have to try it one of these days.


    January 29, 2015 at 12:58 pm

  3. I like how TeX is incredibly cross-platform and backwards compatible (it’s a plain text file), how it lets you author macros so easily, and (given it is an algorithm) how good a job it does of laying out documents. For writing scientific papers, it lets you see how it will look when typeset, which for any work with figures and symbols and equations is of enormous value. Used properly it can generate indices, glossaries, tables of contents, etc. I have envisaged a sort of LaTeX class file designed to aid in writing a long fiction, with commands that (for example) let you automagically create a database for each character (hair colour etc) so that when you process the document you get a manuscript plus a sort of index to tell you where a character shows up in the book, and automatically collated data on what they look like, and so forth. Probably not practicable, but definitely possible; but creating this would be procrastination, I suspect.


    January 31, 2015 at 1:07 am

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: