People with nine-to-five jobs often ask writers: `How do you manage to discipline yourself?' What they really want to know, is how do you manage to get out of bed in the morning. Fortunately, I went through that particular pain barrier many years ago when I left the British army. I found then that once its rigid external discipline, on which I had subconsciously relied, was suddenly removed, even my body clock went haywire for a considerable time. Self-discipline, I realised, was like a series of muscles, which in my case had atrophied in the ordered world of a military existence.
Today it is the priority of helping to get children off to school which sets my idle body in motion. But the working day, which starts as soon as they have departed, varies according to whether I am researching or writing. Stalingrad took two years of research and one year of writing.
I am fairly certain my current project, The Fall of Berlin, will follow a similar ratio. It is a frighteningly long investment in time. The secret fear of almost any writer is to lose all feeling and interest in the subject halfway through.
When working on non-fiction, one always has something to do. Any plea of writer's block is thus completely inadmissible. There is, however, one great similarity between writing novels and history. When a passage starts to go well you have to run with it. I have always been extremely dubious when veteran writers told newspaper interviewers about their 2,000 words written religiously each day before lunch. Rather like building a house, I find a book advances in spurts and slowdowns.
During critical sections, I often wake at three in the morning. Ideas soon start churning, and I know from experience that there is no point in continuing to lie there, wondering whether it is possible to jot notes in the dark without disturbing my wife.
It is far better to slip out of bed as silently as possible, edge down the creaking stairs and shut the door of my study quietly. There is a satisfaction, somehow both morally superior and illicit at the same time, in working when the rest of the household is sleeping.
My wife, Artemis Cooper, is also a writer. This prompts the less imaginative to think that we must be in competition the whole time, or that we get on each other's nerves living and working in the same house together. This may well be true for husband and wife novelists, but it is certainly not true for historians and biographers who are married. I find that loneliness is a writer's greatest enemy. One often needs an early opinion on a particular passage, an article or a chapter, so to have a trusted editor in the house - rather like that mad advertising idea of having a bank manager in the cupboard - is a huge advantage. It is also wonderful to have someone to talk to over coffee or lunch. Artemis and I discuss paradoxes in a character one of us may be writing about or how reliable a particular eyewitness to an event is likely to be. Just talking about something which puzzles you is often enough to sort out a tangle of logic or human motive.
The idea of a wife or husband working in another room within shouting distance is not a hindrance but a stimulus to concentration, just as one is encouraged to greater industry in a library when people all around you are working hard. I am convinced above all by one thing. Writing is definitely good for your marriage. Far from provoking a sense of claustrophobia, it means that the two of you are never short of conversation.
Antony Beevor's Stalingrad has just been published in paperback by Penguin.