Haruki Murakami: How I write my novels

Just as the content of my novels differs from one to the next, so do the manner of their composition, where they were written, and how long they took to complete. Nevertheless, as far as I can tell, the overall pattern – the basic sequence of steps, the “rules” I follow and so forth – doesn’t vary that much. This formula, if it can be called that, pushes me to establish a fixed routine within my life and work – then and only then does writing a full-length novel become possible. Since a novel is a long-term project requiring an inordinate amount of energy, creating this solid base is absolutely crucial. If I screw that up, my strength may give out partway through.

The first step in my novel-writing process is, metaphorically, to clean off my desk. My stance is that I will work on nothing but the novel until it is completed, so I need to prepare. If I happen to be writing a series of essays, for example, I have to break it off, at least for the time being. Unless something really extraordinary comes along, all new projects are turned down. I’m the sort of person who when I throw myself into one thing, can’t do anything else. It’s true that I often work on translations while writing a novel, but those are done at my own pace and without any deadline, and I use them to give me a break from my writing. Translation is a technical process, so it uses a different part of the brain than creative writing. Rather than hindering the progress of a novel, therefore, working simultaneously on a translation can actually aid in the process by helping me keep my mental balance, a bit like stretching before exercising.

When writing a novel, my rule is to produce roughly ten Japanese manuscript pages (the equivalent of sixteen hundred English words) every day. This works out to about two and a half pages on my computer, but I base my calculations on the old system out of habit. On days where I want to write more, I still stop after ten pages; when I don’t feel like writing, I force myself to somehow fulfil my quota. Why do I do it this way? Because it is especially important to maintain a steady pace when tackling a big project. That can’t work if you write a lot one day and nothing at all the next. So I punch in, write my ten pages, and then punch out, as if I’m working on a timecard.

That’s not how an artist should go about his art, some may say. It sounds more like working in a factory. And I concur – that’s not how artists work. But why must a novelist be an artist? Who made that rule? No one, right? So why not write in whatever way is most natural to you? Moreover, refusing to think of oneself as an artist removes a lot of pressure. More than being artists, novelists should think of themselves as “free”– “free” meaning that we are able to do what we like, when we like, in a way we like without worrying about how the world sees us. This is far better than wearing the stiff and formal robes of the artist.

Isak Dinesen once said, “I write a little every day, without hope and without despair.” I write my ten pages the same way. Cool and detached. “Without hope and without despair” says it perfectly. I wake early each morning, brew a fresh pot of coffee, and work for four or five hours straight. Ten pages a day means three hundred pages a month. That works out to eighteen hundred pages in six months.

No sooner is the first draft finished than a new chapter begins: that of rewriting. No time is better spent than the time I spend rewriting, and nothing is more fun.

I take a short break (it depends on the situation, but usually about one week) before undertaking the first rewrite. Then I start at the beginning and plough straight through to the end. At this stage, I make sweeping changes, leaving nothing untouched. No matter how long the novel is, or how complex its structure, I will have composed it without any fixed outline, not knowing how it will unfold or end, letting things take their course and improvising as I go along. This is by far the most fun way to write. As a result, though, the story is riddled with all sorts of contradictions and inconsistencies. Characters may radically change partway through. The timeline may become tangled. These glitches must be fixed if the novel is to flow smoothly in a comprehensible manner. In the process, some lengthy sections may have to be cut back, while other sections may have to be expanded. Entirely new episodes may have to be added.

In the case of The Wind-Up Bird Chronicle, I decided that a large chunk of what I had written didn’t fit into the whole, so I excised it and used it as the base for a subsequent novel, South of the Border, West of the Sun. That’s an extreme example, though—in most cases, the sections I cut out are gone forever.

This rewrite usually takes a month or two. When I finish, I break for another week or so and then begin the second rewrite. As with the first rewrite, I start at the head and work my way through. The difference is that now I focus on the details of the manuscript, fine-tuning passages of natural description, for example, and adjusting the tone of the dialogues. I check to ensure that nothing in the plot is out of place, that hard-to-read sections are made easier, and that the story flows smoothly and naturally. No major surgery takes place during this stage, just lots of nips and tucks. Once I have finished, I take another break and then plunge into the third rewrite. This time no cutting is involved. Instead, I tweak the novel, tightening a screw here, loosening a screw there, making sure that all is in place.

Novels are, by definition, longer works, which means the reader can be stifled if the screws are too tight. Correspondingly, there are spots where I leave them loose, to allow the reader room to breathe. There must also be a balance between the novel as a whole and its parts. All these things require careful adjustment. Some critics like to extract one section of text and castigate the writer for its sloppiness, a practice that strikes me as quite unfair. After all, a novel— like a living, breathing human being—needs to have its loose and sloppy parts. Only that way can the tightly constructed sections achieve their full effect.

At this stage of the game, I take a longer break. For two weeks to a month, if possible, I stick the manuscript in my desk drawer and forget it. At least I try to. In the interim I may take a trip, or concentrate on my translating. The time spent working on a long novel is important, to be sure, but time spent doing nothing is no less so. The same principle applies to a factory or a construction site: manufactured goods are left to settle before being shipped, and concrete is cured in the open air before being built upon. It is through this process that materials are allowed to set or dry out. The same thing holds true with novels. If you fail to let a novel sit for a certain length of time, the parts won’t adhere, or will fail to dry and therefore be weakened.

Once the novel has fully settled, it is time for another detailed and exhaustive run-through. Thanks to my time away, my impressions of the work will have changed quite a lot. Weaknesses I haven’t noticed before jump out at me. I can sense what has depth and what doesn’t. Just as the work has settled, so too has my state of mind.

Once the settling period is over and the subsequent rewrite completed, I move on to the next step. By this point, the novel has assumed what will be more or less its final form, so I can show it to a first reader — namely, my wife.

Photo at top: Desiree Adams / Penguin
Novelist as a Vocation is out now.