Yes, there's always revision. I confess to having been the world's worst reviser. I couldn't bear to change a single one of my hard-won episodes. Which isn't the best approach to take – not with publishers, not with readers either. After all, the novel only exists in cooperation with the reader, you create it together – and if the reader feels there's not much suspense in a particular episode, it's no use to say, But there must be, it's there because I put it there!

Nowadays, I go out of my way to get feedback – from at least a dozen sample readers for any first draft, plus editors of course. I think of editors as readers who have a special talent for explaining their responses, articulating why something isn't working. I wouldn't rely on any single reader, but if several readers are saying the same thing, I know I have to revise.

I tell myself that revision is a challenge to lateral thinking. Could events happen in a different way? If I lose something (and theres's always a loss in revision), can I gain something else in a new and unexpected way? Can I make some other possibility as wanted as my first possibility was? Can I make it as real and colourful – or even more so?

That's what I tell myself, but it's still hard knacker. The difficulty is that every part of a novel impacts on every other part. Put one thing right, and you're almost certainly pushing something else out of shape. For serious structural revisions, I don't believe you can ever confine changes to one isolated chunk, I believe you have to rethink all the way through. I'm find that the time I spend on revisions is about a third to a half of the time I spend on producing the first draft.