A Note on Editing and Revision

I write short novels (and sometimes short non-fiction). My three published books are: 1) (non-fiction) 60,000 words (plus illustrations and extensive endnotes); 2) (novel) 60,000 words (plus extensive illustrations); and 3) (forthcoming novel) 66,000 words.

I sometimes write short non-fiction pieces (and I’ve published many articles in academic journals in the past), but I never ever write short stories. I just don’t think in ways that suit that form. But I’ll also never write a novel (or, likely, a book of any kind) over 75,000 words. I’ll never write a sequel either (the novel I’m just starting now has thematic and situational links with my last one, but no continuities of plot).

All of this makes me a very marginal writer in commercial terms, irrespective of the subject or quality of my writing. Many agents or publishers won’t look at manuscripts under 75,000 words. Nonetheless, I have no intention of padding stories out to meet artificial targets if that feels wrong to me. And interestingly, my editors generally agree – generally I get pushed to make things shorter.

For my forthcoming novel The Angels of L19 I can track word counts very precisely. The longest complete draft appears to be v59, which is 76,000 words (though I think I had already removed a couple of surplus scenes from this, so there could theoretically have been an ur version of about 80,000 words). I submitted v70 for my PhD in creative writing, and that was 73,000 words. There was some discussion around this as departmental guidelines at the University of Kent suggested it should be at least 90,000 words. My position was: that’s what I’m submitting, I’m not changing it, so feel free to fail it if you have a problem with that. In the event, I passed with no corrections. I submitted v75 to Weatherglass Books, and that was 71,000 words. The published version will be 66,000.

In other words, the editorial interventions made what was already a short book into an even shorter one. The trajectory was not to bulk it up, or fill gaps, but empty it out where it was too crowded, too stuffed with complications and explanations. Make its shape clearer and simpler, but also make it more mysterious and open – fill it with silences. I think this was right – it’s the direction my writing wants to move in. For The Angels of L19 in particular, I had the cautionary example of Donnie Darko in mind (my elevator pitch is 'Donnie Darko but all the characters are evangelical Christians'). The initial release of that film is close to perfect, even with its ellipses and lacunae. The Director's Cut can't resist explaining everything, thereby reducing mystery to banality. The differences are small, but the effect is catastrophic.

A note also on ‘versions’ or ‘drafts’: 75 plus seems a lot, and probably is. But a draft isn’t what it used to be. A new draft used to require writing or typing the entire manuscript out from scratch again, and incorporating corrections added (usually by hand) to the typescript of the previous draft. A few holdouts may still do precisely this, but for most of us, word processing means we can make a potentially infinite number of alterations to any given version of a document. I work with Word, and I move the draft number up one whenever I make what feels like a significant cut or change: whenever, in other words, I feel I might conceivably need to retain or refer back to the prior version. I write individual chapters in separate files, revising these several times before I incorporate them into the master file compiling all chapters completed thus far. This explains why I can get up to v59 before I have a proper workable complete draft. Then, for The Angels of L19, there were approx. twenty more versions after that before it went to the typesetter.

How many ‘drafts’ in old-style money is that equivalent to? I don’t really know – eight to ten total? But it’s hard to say.

I wish I didn’t feel compelled to endlessly revise and dicker about. Partly this is a function of trying to figure out an original form (first two books, which are both quite experimental formally, in very different ways), and then of trying to change my working method and master a more traditional form, but dig deeper emotionally (The Angels of L19). It’s also a function of how long it took to find a publisher for all three books: I spent five years submitting the first; the second similarly took four years (I mean after completing the first draft); the third only three – which I suppose is progress of a sort. Under these circumstances, it’s hard not to constantly second-guess what the perceived problems might be and try to make it better for the next round. But if the perceived problem is word count, then I guess I'm shit out of luck for the rest of my career.