There were many people in France writing realistic novels in the nineteenth century—Balzac comes to mind—but Flaubert was the most conscious of what he was doing, and agonized about it the most. Closely observed reality, for its own sake, had not really been a part of the tradition of literature in the eighteenth century. Flaubert will spend a whole page evoking tiny sounds and motes of dust in an empty room because he's getting at something. He's saying there's meaning to be got out of the very closely observed events of ordinary reality. In literary, scientific and photographic terms—the invention of photography happened when he was in his teens—the nineteenth century, to a much greater degree than the eighteenth, was concerned with the close observation of reality. All of science in the nineteenth century was about very close observation of small things…. The nineteenth century focussed and greatly expanded these concepts. It made them central to the novel, to the symphony, to painting.

O: There's a line of Saul Bellow's—“I write to discover the next room of my fate.” In this way, I think, many novels are self-portraits—or future self-portraits, self-explorations, even if the story is set in an alien situation. You can try on this costume, that costume.

M: Somebody once asked W. H. Auden, “Is it true that you can write only what you know?”And he said,“Yes it is. But you don't know what you know until you write it.” Writing is a process of discovery of what you really do know. You can't limit yourself in advance to what you know, because you don't know everything you know.

As I began to eliminate things, I would have the feeling that I couldn't remove a certain scene, because it so clearly expressed what we were after. But after hesitating, I'd cut it anyway … forced to because of the length of the film. Then I'd have this paradoxical feeling that by taking away something I now had even more of it. It was almost biblical in its idea of abundance. How can you take away something and wind up with more of it?

The analogy I came up with was the image of a room illuminated by a bare blue lightbulb. Let's say the intention is to have “blueness” in this room, so when you walk in you see a bulb casting a blue light. And you think, This is the source of the blue, the source of all blueness. On the other hand, the lightbulb is so intense, so unshaded, that you squint. It's a harsh light. It's blue, but it's so much what it is that you have to shield yourself from it.

There are frequently scenes that are the metaphorical equivalent of that bulb. The scene is making the point so directly that you have to mentally squint. And when you think, What would happen if we got rid of that blue lightbulb, you wonder, But then where will the blue come from? Let's take it out and see. That's always the key: Let's just take it out and find out what happens.

So you unscrew the lightbulb … there are other sources of light in the room. And once that glaring source of light is gone, your eyes open up. The wonderful thing about vision is that when something is too intense, your irises close down to protect against it—as when you look at the sun. But when there is less light, your eye opens up and makes more of the light that is there.

M: One of the side benefits of moving Zoetrope up to San Francisco in 1969 was to take the filmmaking out of a self-contained film universe. In Los Angeles it's very easy, if you get to a certain level in your profession, to live, breathe, eat, think, sleep film. And to have so many offers that you're working all the time—without that ability, given by time and space, to reinvent.

In San Francisco there's not that much work. There aren't that many producers up here. There's a kind of enforced idleness between projects that allows you to develop other interests and then, in the best sense, to plow the results of those interests back into the next film.

It has to be said—both systems have their risks. The risk of the Hitchcock-ian system is that you may stifle the creative force of the people who are collaborating with you. The film that results—even if it's a perfect vision of what somebody had in his head—can be lifeless: it seems to exist on its own, without the necessary collaboration either of the people who made the film or even, ultimately, the audience. It says: I am what I am whether you like it or not.

On the other hand, the risk with the process-driven film is that it can collapse into chaos. Somehow the central organizing vision can be so eaten away and compromised by all the various contributors that it collapses under its own weight.

You remember you told me how much you liked the line breaks in my translations of Malaparte? The decision where to cut film is very similar to the decision, in writing poetry, of where to end each line. On which word? That end point has little if anything to do with the grammar of the sentence. It's just that the line is full and ripe at that point, full of meaning and ripe with rhythm. By ending it where he does, the poet exposes that last word to the blankness of the page, which is a way of emphasizing the word. If he adds two words after it, he immerses that word within the line, and it has less visibility, less significance. We do very much the same in film: the end of a shot gives the image of that last frame an added significance, which we exploit.

The advantage of writing and editing is that at any time you can stop what you're doing and walk around the block, or have lunch, or take a phone call, or go dig in the garden and think. At any particular moment, the editor has the freedom to interact or not interact with the material. You can always step away and let the subconscious do some work.

When I write a script, I lie down—because that's the opposite of standing up. I stand up to edit, so I lie down to write. I take a little tape recorder and, without being aware of it, go into a light hypnotic trance. I pretend the film is finished and I'm simply describing what was happening. I start out chronologically but then skip around. Anything that occurs to me, I say into the recorder. Because I'm lying down, because my eyes are closed, because I'm not looking at anything, and the ideas are being captured only by this silent scribe—the tape recorder—there's nothing for me to criticize. It's just coming out.

That is my way of disarming the editorial side. Putting myself in a situation that is as opposite as possible to how I edit—both physically and mentally. To encourage those ideas to come out of the woods like little animals and drink at the pool safely, without feeling that the falcon is going to come down and tear them apart.

In any film scene there's a petitioner and a grantor, a weaker character and a stronger. Otherwise you don't have a scene. It may be obvious what the power relationship is, or it may be hidden from one or even both of the characters themselves. So at each moment in every scene there is a dynamic between two people that can be expressed in many ways.

In making a film you're trying to get the most interesting orchestration of all these elements, which, like music, need to be harmonic yet contradictory. If they're completely contradictory, then there's chaos. It's like when instruments are tuning up before a performance, you can't make anything coherent out of it. It's a fascinating, evocative sound, but only for about fifteen seconds. If, on the other hand, all the instruments play the same notes—if they're too harmonic, in other words—yes, there's coherence, but I'm bored after a few minutes. Just as bored as with the chaos of tuning up.