North by Northwest Didn't Always Have Direction

“There are 20 different ways of doing a movie, but if you know you’re doing a film for Hitchcock, the villains should all be suave, there’s very little violence, [and] there should be some wit, even when you’re killing somebody.”

Those are the words of screenwriter Ernest Lehman, reflecting on his experiences creating North by Northwest, as anthologized in Conversations With the Great Moviemakers of Hollywood’s Golden Age at the American Film Institute – an unfailingly satisfying read that I pick up off the bookshelf every few months to savor just a few pages at a time. Regular readers probably know this isn’t the first time I’ve quoted that anthology here at The Cooler, and it almost certainly won’t be the last. Page after page I find something enlightening: sometimes it’s a bit of trivia, sometimes it’s a well-told anecdote, but usually it’s a needed dose of perspective.

And that said, I turn it over to Ernest Lehman, circa 1976, discussing the creation of his only original screenplay, North by Northwest (emphasis mine):

“It was fun in a way, but it was extremely difficult. I recall having tried to quit that picture at least a dozen times, unknown to Mr. Hitchcock, who was off shooting Vertigo while I was writing the first 70-odd pages. I never knew what the hell I was going to write next. I used to go to my office and be scared to death because I just didn’t know what came next. I would write an opening scene and then the next scene. Some days I wouldn’t write anything. It was a lonesome and scary experience because I didn’t have anyone to talk to. Occasionally I would talk to Hitch, who was very helpful as we would bounce ideas off each other. Then I would go back to my office at MGM and call my agent and tell him, ‘I quit.’

“One day Hitchcock said, ‘I’ve always wanted to do a chase across the face of Mount Rushmore.’ Terrific, I thought, and wrote it down. He told me all the ideas he wanted to do, and I wrote them down. Lots and lots of ideas. He wanted to do a sequence with the longest dolly shot in history, taking place at the assembly line of the Ford Motor Company. It would starts at the beginning of the assembly line. The camera follows a car being put together before it’s driven off the assembly line and they discover there’s a body in the backseat. He always wanted to do a scene in the General Assembly of the United Nations. Somebody is giving a speech to the Assembly and refused to continue until the delegate from Peru wakes up. So someone taps the delegate from Peru, and he falls over dead. I wrote these ideas down, but the only idea that is actually in the picture is the chase across Mount Rushmore. I also wrote out a list of possible protagonists, like a Frank Sinatra-type singer, or a famous sports announcer, or a newspaper man or a Madison Avenue advertising executive. I decided the easiest for me to write was the Madison Avenue advertising executive, not because I knew any, but because I know how to stereotype him.

“Then we moved over to Hitchcock’s office at Paramount where he was working on Vertigo. We wasted a lot of time talking about the pleasantries of life. One day he told me something rather crucial, that a newspaper man in New York had told him that something like the CIA had once invented a decoy, a nonexistent agent, to throw the scent off a real agent. I thought that sounded good. Our hero, whoever he is, could be mistaken for this decoy. That I liked. I wrote the script page by page, scene by scene, never knowing what was coming next. The only thing I knew for sure was that we wanted to end up on Mount Rushmore, which is my least favorite part of the film, though most people remember it for that. I got only about one-tenth of the way through the story when Hitch gathered all the MGM executives together and told them everything we had. At the point where he ran out of story outline, he told them he had to go and that he’d see them at the preview. That was it.

“Then Hitchcock arrived at MGM and he signed Cary Grant and fixed a starting date. Here I am, sweating my way through the first draft, and I still didn’t know what the whole third act would be. I was so desperate I called Hitch and said I need to see him. I think he could sense how things were because he didn’t say, ‘Come down to my office.’ He said, ‘I’ll come down to your office.’ We met and I told him the truth. I said, ‘I’m totally stuck. I haven’t written a word in two weeks. It’s a disaster. What do we do?’ He said, ‘We’ll call in a novelist to sit with us and kick around some ideas.’ I said, ‘What will MGM say? I’m supposed to be writing this movie.’ He said, ‘I’ll tell them it’s my fault.’ I felt very guilty. We went down to his office and he talked about which novelist to get. It was a very gloomy scene.

“But I’m sitting there, and it’s not as if I wasn’t listening to him, but in the middle of this gloomy conversation I said, ‘Suppose he pulls out a gun and shoots him. Fake bullets. In a minute I’ll figure out why.’ My brain must have been working for months on this without my knowing it, because suddenly it became clear to me what the third act would be. … I left the office and never again was there any mention of bringing anyone else in.”

Incredible story, isn’t it? And what strikes me reading it is how much any screenwriter or filmmaker would be obliterated for saying anything similar today. Starting the screenplay without knowing what it’s about? Picking the dramatic hook after you’ve already started writing? Figuring out the ending long after you've begun? Writing an entire film to justify a director’s desire to film a single scene? Sounds like hackwork, doesn’t it? But who among us would call North by Northwest hackwork?

North by Northwest succeeds for so many reasons: because of Hitchcock’s tremendous eye and sense of dramatic tension, because of fun performances by Cary Grant, Eva Marie Saint, James Mason and Martin Landau, because of Bernard Herrmann’s awesome score, and because Lehman’s final screenplay was tremendous, no matter how he got there. The lesson is one we should already know but so easily forget: process is irrelevant. It’s what ends up on the screen that counts.