Retiring the word count

Back in the middle of August I gave up on getting Pantheon down to one hundred and fifty thousand words. Not missed the target, mind you. Retired it, on purpose, and swapped it for something nearer one hundred and sixty-five thousand, because those fifteen thousand words simply were not sitting anywhere I could get at them.

I did go looking. You look for fat, do not you: the bloated chapter, the scene that outstays its welcome. There was not one. All sixty-three chapters ran a little long, evenly, which is a far more annoying problem to have. It means the length is not waiting in a few places for you to lift it out. Cutting fifteen thousand words from a book shaped like that would have meant pulling settled prose apart so a number on a page read differently, which is not the same thing as making the book better. So the target went and the prose stayed.

The other thing that makes cutting hard here is that the plot is dense, and some of what is on the page is quietly doing work for what comes after this book. Pantheon is the first of three. I would rather leave a sentence standing than take out something you are going to need later, and a fair few have stayed for exactly that reason.

So the trimming had to happen a sentence at a time instead, and the ones that gave way easiest were the ones explaining what the book had already shown you. A clock reading handed over after the light in the scene had already told you the hour. A count stated all over again, fourteen lines after the first time. I have started calling it the gloss test: if a sentence only restates the one before it, out it goes. The manuscript has come down by two thousand and fifty-eight words since I said it was finished, and nearly all of that went one gloss at a time. It is slow going, and every single cut is really a decision to trust you to have got it the first time.

What keeps the rest of it honest is the map of when everything happens, which I wrote down before a word of the manuscript existed. I check the book against it one thing at a time and note down what I find before moving on. Dull, and the only method that has ever worked for me. A novel goes out of joint fastest exactly where it is reading best, once everything is flowing and you have stopped watching your feet. This week that turned up two spots where the sequence had drifted, and a run of stale pointers into the chapters, one lot of which had been printing a warning on every check for days with nobody reading it.