Graham Harman has been posting a series of enlightening thoughts on writing as he races toward a book deadline, taking only two months from start to finish. The book in question has a word limit (a character limit, really) because it is destined for immediate translation, and the translation has to be done on a budget.

The whole series is worth a read, both for experienced writers and new ones. I’d like to draw attention to today’s post in particular, though, because in it Harman draws a connection between Nick and my book Racing the Beam, which Graham read just before I met up with him in Cairo in early July, and his own work.

With this book [Harman’s ms in progress] Iâ??m in a similar position to Atari programmers or weapons miniaturizers. Never have I packed this many of my ideas into a mere 70 double-spaced pages (the final 70 pages will be largely new material that none of my readers have ever seen before). Many of the cuts have been painful, of the order of: â??my God, will the argument even make sense if I remove that step?â?

But more often than not, compression and allusiveness can be made to work. This book is certainly going to have a very rapid pace, like a high-speed film of a horse running.

Harman’s right that Racing the Beam is about constraint in general as much as (perhaps even more than?) it’s about the Atari VCS. Indeed, I’ve been outlining a new project entirely devoted to the concept of constraint, and it’s motivating to hear that the idea is helpful.

In academic writing, we usually don’t have many external constraints. In fact, it’s not uncommon to hear scholars complain that university presses have become less willing to publish books over 300 pages or so. But now that I’ve written a few books, I welcome length constraints, and even impose them on myself.

I think Unit Operations is about 80,000 words. Persuasive Games is around 150,000 words. Racing the Beam weighed in right at 60,000, I believe, and it felt like the right size for a book, both physically and conceptually. I don’t know that I’d never strive to write a long and involved book again, but right now, I’m very interested in the 40,000 – 60,000 word work. Incidentally, that’s about the size of a typical trade non-fiction book. Maybe there’s a reason for that.

published July 26, 2009


  1. Mark J. Nelson

    As a reader, an interesting experience is when there are multiple versions, of what could more or less be called the same work, available at varying lengths. The most common case is where a really long work was written initially with no constraints, and it was subsequently cut down for publication; book-ified theses are a common example. I sometimes find the long version better, and sometimes the short, depending on the writer and what I want out of the work.

    Of course, constraint works differently there: cutting a 150,000-word book down to 60,000 isn’t constrained composition, but rather editing of a work that was composed without those constraints. (Is there anything interesting to say about the difference?)