The art of preemptive ingenuity

Yesterday, my wife drew my attention to the latest episode of the podcast 99% Invisible, which irresistibly combines two of my favorite topics—film and graphic design. Its subject is Annie Atkins, who has designed props and visual materials for such works as The Tudors and The Grand Budapest Hotel. (Her account of how a misspelled word nearly made it onto a crucial prop in the latter film is both hilarious and horrifying.) But my favorite story that she shares is about a movie that isn’t exactly known for its flashy art direction:

The next job I went onto—it would have been Spielberg’s Bridge of Spies, which was a true story. We made a lot of newspapers for that film, and I remember us beginning to check the dates against the days, because I wanted to get it right. And then eventually the prop master said to me, “Do you know what, I think we’re just going to leave the dates off.” Because it wasn’t clear [what] sequence…these things were going to be shown in. And he said, you know, if you leave the dates off altogether, nobody will look for it. But if you put something there that’s wrong, then it might jump out. So we went with no dates in the end for those newspapers.

As far as filmmaking advice is concerned, this is cold, hard cash, even if I’ll never have the chance to put it into practice for myself. And I especially like the fact that it comes out of Bridge of Spies, a writerly movie with a screenplay by none other than the Coen Brothers, but which was still subject to decisions about its structure as late in the process as the editing stage.

Every movie, I expect, requires some degree of editorial reshuffling, and experienced directors will prepare for this during the production itself. The absence of dates on newspapers is one good example, and there’s an even better one in the book The Conversations, which the editor Walter Murch relates to the novelist Michael Ondaatje:

One thing that made it possible to [rearrange the order of scenes] in The Conversation was Francis [Coppola]’s belief that people should wear the same clothes most of the time. Harry is almost always wearing that transparent raincoat and his funny little crepe-soled shoes. This method of using costumes is something Francis had developed on other films, quite an accurate observation. He recognized that, first of all, people don’t change clothes in real life as often as they do in film. In film there’s a costume department interested in showing what it can do—which is only natural—so, on the smallest pretext, characters will change clothes. The problem is, that locks filmmakers into a more rigid scene structure. But if a character keeps the same clothes, you can put a scene in a different place and it doesn’t stand out.

Murch observes: “There’s a delicate balance between the timeline of a film’s story—which might take place over a series of days or weeks or months—and the fact that the film is only two hours long. You can stretch the amount of time somebody is in the same costume because the audience is subconsciously thinking, Well, I’ve only been here for two hours, so it’s not strange that he hasn’t changed clothes.”

The editor concludes: “It’s amazing how consistent you can make somebody’s costume and have it not stand out.” (Occasionally, a change of clothes will draw attention to editorial manipulation, as one scene is lifted out from its original place and slotted in elsewhere. One nice example is in Bullitt, where we see Steve McQueen in one scene at a grocery store in his iconic tweed coat and blue turtleneck, just before he goes home, showers, and changes into those clothes, which he wears for the rest of the movie.) The director Judd Apatow achieves the same result in another way, as his longtime editor Brent White notes: “[He’ll] have something he wants to say, but he doesn’t know exactly where it goes in the movie. Does it service the end? Does it go early? So he’ll shoot the same exact scene, the same exchange, with the actors in different wardrobes, so that I can slot it in at different points.” Like the newspapers in Bridge of Spies, this all assumes that changes to the plan will be necessary later on, and it prepares for them in advance. Presumably, you always hope to keep the order of scenes from the script when you cut the movie together, but the odds are that something won’t quite work when you sit down to watch the first assembly, so you build in safeguards to allow you to fix these issues when the time comes. If your budget is high enough, you can include reshoots in your shooting schedule, as Peter Jackson does, while the recent films of David Fincher indicate the range of problems that can be solved with digital tools in postproduction. But when you lack the resources for such expensive solutions, your only recourse is to be preemptively ingenious on the set, which forces you to think in terms of what you’ll want to see when you sit down to edit the footage many months from now.

This is the principle behind one of my favorite pieces of directorial advice ever, which David Mamet provides in the otherwise flawed Bambi vs. Godzilla:

Always get an exit and an entrance. More wisdom for the director in the cutting room. The scene involves the hero sitting in a café. Dialogue scene, blah blah blah. Well and good, but when you shoot it, shoot the hero coming in and sitting down. And then, at the end, shoot him getting up and leaving. Why? Because the film is going to tell you various things about itself, and many of your most cherished preconceptions will prove false. The scene that works great on paper will prove a disaster. An interchange of twenty perfect lines will be found to require only two, the scene will go too long, you will discover another scene is needed, and you can’t get the hero there if he doesn’t get up from the table, et cetera. Shoot an entrance and an exit. It’s free.

I learned a corollary from John Sayles: at the end of the take, in a close-up or one-shot, have the speaker look left, right, up, and down. Why? Because you might just find you can get out of the scene if you can have the speaker throw the focus. To what? To an actor or insert to be shot later, or to be found in (stolen from) another scene. It’s free. Shoot it, ’cause you just might need it.

This kind of preemptive ingenuity, in matters both large and small, is what really separates professionals from amateurs. Something always goes wrong, and the plan that we had in mind never quite matches what we have in the end. Professionals don’t always get it right the first time, either—but they know this, and they’re ready for it.

Shoot the piano player

In his flawed but occasionally fascinating book Bambi vs. Godzilla, the playwright and director David Mamet spends a chapter discussing the concept of aesthetic distance, which is violated whenever viewers remember that they’re simply watching a movie. Mamet provides a memorable example:

An actor portrays a pianist. The actor sits down to play, and the camera moves, without a cut, to his hands, to assure us, the audience, that he is actually playing. The filmmakers, we see, have taken pains to show the viewers that no trickery has occurred, but in so doing, they have taught us only that the actor portraying the part can actually play the piano. This addresses a concern that we did not have. We never wondered if the actor could actually play the piano. We accepted the storyteller’s assurances that the character could play the piano, as we found such acceptance naturally essential to our understanding of the story.

Mamet imagines a hypothetical dialogue between the director and the audience: “I’m going to tell you a story about a pianist.” “Oh, good: I wonder what happens to her!” “But first, before I do, I will take pains to reassure you that the actor you see portraying the hero can actually play the piano.” And he concludes:

We didn’t care till the filmmaker brought it up, at which point we realized that, rather than being told a story, we were being shown a demonstration. We took off our “audience” hat and put on our “judge” hat. We judged the demonstration conclusive but, in so doing, got yanked right out of the drama. The aesthetic distance had been violated.

Let’s table this for now, and turn to a recent article in The Atlantic titled “The Remarkable Laziness of Woody Allen.” To prosecute the case laid out in the headline, the film critic Christopher Orr draws on Eric Lax’s new book Start to Finish: Woody Allen and the Art of Moviemaking, which describes the making of Irrational Man—a movie that nobody saw, which doesn’t make the book sound any less interesting. For Orr, however, it’s “an indictment framed as an encomium,” and he lists what he evidently sees as devastating charges:

Allen’s editor sometimes has to live with technical imperfections in the footage because he hasn’t shot enough takes for her to choose from…As for the shoot itself, Allen has confessed, “I don’t do any preparation. I don’t do any rehearsals. Most of the times I don’t even know what we’re going to shoot.” Indeed, Allen rarely has any conversations whatsoever with his actors before they show up on set…In addition to limiting the number of takes on any given shot, he strongly prefers “master shots”—those that capture an entire scene from one angle—over multiple shots that would subsequently need to be edited together.

For another filmmaker, all of these qualities might be seen as strengths, but that’s beside the point. Here’s the relevant passage:

The minimal commitment that appearing in an Allen film entails is a highly relevant consideration for a time-strapped actor. Lax himself notes the contrast with Mike Leigh—another director of small, art-house films—who rehearses his actors for weeks before shooting even starts. For Damien Chazelle’s La La Land, Stone and her co-star, Ryan Gosling, rehearsed for four months before the cameras rolled. Among other chores, they practiced singing, dancing, and, in Gosling’s case, piano. The fact that Stone’s Irrational Man character plays piano is less central to that movie’s plot, but Allen didn’t expect her even to fake it. He simply shot her recital with the piano blocking her hands.

So do we shoot the piano player’s hands or not? The boring answer, unfortunately, is that it depends—but perhaps we can dig a little deeper. It seems safe to say that it would be impossible to make The Pianist with Adrian Brody’s hands conveniently blocked from view for the whole movie. But I’m equally confident that it doesn’t matter the slightest bit in Irrational Man, which I haven’t seen, whether or not Emma Stone is really playing the piano. La La Land is a slightly trickier case. It would be hard to envision it without at least a few shots of Ryan Gosling playing the piano, and Damien Chazelle isn’t above indulging in exactly the camera move that Mamet decries, in which it tilts down to reassure us that it’s really Gosling playing. Yet the fact that we’re even talking about this gets down to a fundamental problem with the movie, which I mostly like and admire. Its characters are archetypes who draw much of their energy from the auras of the actors who play them, and in the case of Stone, who is luminous and moving as an aspiring actress suffering through an endless series of auditions, the film gets a lot of mileage from our knowledge that she’s been in the same situation. Gosling, to put it mildly, has never been an aspiring jazz pianist. This shouldn’t even matter, but every time we see him playing the piano, he briefly ceases to be a struggling artist and becomes a handsome movie star who has spent three months learning to fake it. And I suspect that the movie would have been elevated immensely by casting a real musician. (This ties into another issue with La La Land, which is that it resorts to telling us that its characters deserve to be stars, rather than showing it to us in overwhelming terms through Gosling and Stone’s singing and dancing, which is merely passable. It’s in sharp contrast to Martin Scorsese’s New York, New York, one of its clear spiritual predecessors, in which it’s impossible to watch Liza Minnelli without becoming convinced that she ought to be the biggest star in the world. And when you think of how quirky, repellent, and individual Minnelli and Robert De Niro are allowed to be in that film, La La Land starts to look a little schematic.)

And I don’t think I’m overstating it when I argue that the seemingly minor dilemma of whether to show the piano player’s hands shades into the larger problem of how much we expect our actors to really be what they pretend that they are. I don’t think any less of Bill Murray because he had to employ Terry Fryer as a “hand double” for his piano solo in Groundhog Day, and I don’t mind that the most famous movie piano player of them all—Dooley Wilson in Casablanca—was faking it. And there’s no question that you’re taken out of the movie a little when you see Richard Chamberlain playing Tchaikovsky’s Piano Concerto No. 1 in The Music Lovers, however impressive it might be. (I’m willing to forgive De Niro learning to mime the saxophone for New York, New York, if only because it’s hard to imagine how it would look otherwise. The piano is just about the only instrument in which it can plausibly be left at the director’s discretion. And in his article, revealingly, Orr fails to mention that none other than Woody Allen was insistent that Sean Penn learn the guitar for Sweet and Lowdown. As Allen himself might say, it depends.) On some level, we respond to an actor playing the piano much like the fans of Doctor Zhivago, whom Pauline Kael devastatingly called “the same sort of people who are delighted when a stage set has running water or a painted horse looks real enough to ride.” But it can serve the story as much as it can detract from it, and the hard part is knowing how and when. As one director notes:

Anybody can learn how to play the piano. For some people it will be very, very difficult—but they can learn it. There’s almost no one who can’t learn to play the piano. There’s a wide range in the middle, of people who can play the piano with various degrees of skill; a very, very narrow band at the top, of people who can play brilliantly and build upon a technical skill to create great art. The same thing is true of cinematography and sound mixing. Just technical skills. Directing is just a technical skill.

This is Mamet writing in On Directing Film, which is possibly the single best work on storytelling I know. You might not believe him when he says that directing is “just a technical skill,” but if you do, there’s a simple way to test if you have it. Do you show the piano player’s hands? If you know the right answer for every scene, you just might be a director.

Cat got your tongue

On Friday, I spent most of the morning trying to remember a cat. I wanted to write a blog post about animals in film, and I vaguely recalled an anecdote about a movie scene involving a cat that got its head caught in something. A box? A can? A mailing tube? I wasn’t sure. All I knew was that the sequence involved a single uninterrupted shot of a cat squeezing its head into and out of a tight space, and given the notorious difficulty of training cats to do anything on camera, it was all but impossible to figure out how it was done. But I couldn’t remember anything else. It couldn’t have been just a cute scene; it had to be tied into the plot, so there had to be an element of danger or suspense. Maybe there was a bomb in the box? Or it took place during a heist where the smallest sound would give the thief away? I tried searching for every combination I could imagine of the words “cat,” “head,” and “stuck,” and I even ended up studying a list of famous movie cats. No dice. And although I knew that the story, whatever it was, would only end up being a tiny part of the post I was writing, it seemed crucial to track it down, or it would bug me forever.

Yet there was something about that prolonged inventory of my mental archives that was oddly satisfying. These days, we’re used to having instant access to all human knowledge, as long as we know the right search terms, to the point where we’re likely to forget how profoundly strange it all is. As recently as a decade ago, when I wrote my senior thesis, the idea of being able to search the text of printed books to track down every last mention of a subject was unthinkable. Scholars had to either absorb a huge body of material or trust to indexes and luck. I spent hours browsing at random in the stacks at the Smyth Classical Library, and I vividly remember my advisor’s response when I came up with an obscure passage in Statius that gave me the exact bit of evidence my argument needed: “How did you find this?” I wasn’t entirely sure then, and I certainly don’t remember now. These days, it wouldn’t be impressive at all: it’s just a matter of typing in the right keywords. But there’s a loss here, too, and it has less to do with the moment of discovery than with the long process of preparation, consolidation, and patience that made it possible.

When it came to my cat story, I ended up reliving a condensed version of what like was like before Google, complete with a few false breakthroughs. (At one point, I exclaimed aloud: “Rififi!” But nope, that wasn’t it.) Ultimately, I finally remembered where I’d read the anecdote, and it came to me, appropriately enough, as I was doing the dishes: it was in David Mamet’s Bambi vs. Godzilla, a book I read years ago and no longer own. In case you didn’t click through last week, here’s Mamet’s account of the scene, which occurs in George Stevens’s film of The Diary of Anne Frank:

[Otto] Frank retreats to the attic and all wait breathlessly, while the Nazis scout the house below. Now comes the cat. She pads along a kitchen ledge in the hidden attic, she puts her head into a funnel resting on the ledge, she pushes the funnel toward the edge. Now everyone in the world holds his or her breath. Now the funnel goes over the ledge. But wait, the cat’s head is still stuck in the funnel.

Should the funnel drop off, the Nazis will hear, and discover the hidden attic and kill all the inhabitants. But continue to wait—the cat now pulls its head, the funnel still on it, back onto the ledge, and now draws its head off. What a great sequence. But how did they do it?

Mamet goes through a bunch of possibilities—they filled the funnel with tuna fish and some glue, or used monofilament line, or magnets, or shot the scene backwards—only to conclude that none of his solutions make any sense. At last, he calls up George Stevens, Jr., who gives him the answer: the director “just turned on the cameras and shot an unbelievable amount of film, waiting for some cat to do something ‘uncatlike.'” It’s a great story. But it’s only interesting after you’ve followed Mamet through all his wrong turns: it wouldn’t be nearly as compelling if he had called up the director’s son at the start. And I feel the same way about my own little search. I misremembered almost everything. I certainly wasn’t thinking of a funnel, or Anne Frank, and when I finally got it, it was based on a hunch that the source was probably Mamet, to whom I owe many of my favorite stories about the movies. But the search was worth it in itself. It reminded me of Marlon Brando’s cat in The Godfather, and of the cat in Alien, and many others, and it gave me an excuse to stare at my library shelves and rack my brain for a while. And I liked it. Because as nice as it can be to have access to a universe of information, it’s good to be reminded that there’s more than one way to skin a cat.

David Mamet and the limits of craft

If you’ve been reading this blog for any period of time, you’ve no doubt gathered that I like David Mamet. While I generally agree with Lawrence Weschler that Walter Murch is the smartest person in America, there was a time in my life when I would have ranked Mamet at least a close second. I’ve learned a tremendous amount about craft from his essays, interviews, and commentary tracks, and in particular, his little book On Directing Film is the most useful guide to storytelling I’ve ever seen. As I’ve mentioned before, I discovered it at a point when I thought I’d figured out the writing process to my own satisfaction, so reading it was a little like having an efficiency expert visit your business for a day and set you straight regarding best practices. I encountered it too late for it to have any real influence on The Icon Thief, but it was a major reason I was able to get City of Exiles from conception to finished draft in under a year, and it’s since become an indispensable part of my approach to writing. I try to read it again every six months or so, especially when I’m starting a new project, and I’m still amazed by its level of insight and practicality.

Yet there’s a shadow side to Mamet’s intelligence and mastery. It’s taken me a long time to figure out what it is, and I’ve been thinking about it a lot recently, ever since seeing Mamet’s latest movie, Phil Spector, which aired over the weekend on HBO. Like all of his films, it’s watchable, full of good dialogue, and admirably streamlined: it clocks in at just over ninety minutes, and there isn’t an ounce of fat on the screenplay. All the same, it feels weirdly like half a movie, or a brilliant sketch of something better, which is true of nearly all of Mamet’s work as a director. I haven’t seen a movie of his I didn’t like—I even enjoyed Redbelt—but there’s something clinical and detached about his style that leaves even his best films feeling a little thin. And the more I think about it, the more it seems like an inevitable consequence of his approach to craft. Mamet’s method is as rigorous as mathematics: you figure out the sequence of objectives for each character, then craft the scene and the individual shots to convey this information as simply as possible. Hence his beloved story about Stanislavsky:

Stanislavsky was once having dinner with a steamboat captain on the Volga River and Stanislavsky said, “How is it that among all the major and minor paths of the Volga River, which are so many and so dangerous, you manage to always steer the boat safely?” And the captain said, “I stick to the channel; it’s marked.”

If nothing else, Mamet’s movies stick to the channel, and his philosophy as a director has always been that you shouldn’t stray much to either side. Most famously, he believes that if a script has been properly written, the actors just need to say the lines clearly and without inflection, and the words themselves will do the work—although if Phil Spector is any indication, even Mamet can’t always get this from Pacino. This approach to storytelling is unimpeachably correct, and if you’re going to imitate any director, you can’t go too far wrong by following Mamet: at worst, you’ll end up with a first draft that is mechanical but basically efficient, which is far from the worst that can happen. (As T.S. Eliot says in one of his essays, a poet who imitates Dante will wind up with a boring poem, but someone who imitates Shakespeare is likely to make a fool of himself.) But Mamet has essentially transformed himself into a director who delivers brilliant, clean, unimpeachable first drafts. And it’s no accident that the best movies based on his work—which I’d argue are The Untouchables and Glengarry Glen Ross—were made by other directors.

And we’ve seen much the same progression in Mamet’s prose, which has devolved from the wit and lucidity of On Directing Film to something crabbed, aphoristic, and airless. Bambi Vs. Godzilla contains five or six pages that include some of the best storytelling advice imaginable—if you’re curious, it’s in the chapter “The Wisdom of the Ancients”—surrounded by material so tight and hermetic that reading it becomes physically enervating. The same is true, sadly, of Three Uses of the Knife, and I’m too discouraged to even try The Secret Knowledge. Which is just a reminder, as if we needed one, of the pitfalls of genius. Mamet remains the most intelligent living writer I know, and when it comes to the nuts and bolts of craft, he’s right about almost everything. But being consistently right for forty years can be dangerous in itself. Mamet is very good at what he does, and unlike a lot of artists, he knows the reasons why. But there’s a point where logic and craft take you only so far, at least not without being willing to embrace the possibility of failure or foolishness. Mamet, like most smart men, simply can’t take that risk. And although he’s still the best there is at sticking to the channel, there’s a chance that a lot of viewers will simply decide to change it.

Searching for the slate piece

Yesterday, with some trepidation, I read through the partial manuscript of the sequel to The Icon Thief for the first time, going through the entire draft in close to one sitting. At this point, I’ve finished the first two sections of the novel, totaling about 100,000 words—although this will be cut considerably—and still need to write the conclusion, which will be much shorter. It seemed like a good time, then, to pause and take stock of what I have. For the past few months, I’ve been writing with my head down, doing a chapter a day and then moving on immediately to the next. The upshot is that I’ve got much of the story on paper, but still don’t know, precisely, what this novel is about, which is something you discover only after much brooding and rereading. And this is also how you discover the ending.

Ending any story is hard. When you write a novel, in particular, you generally have at least some idea of where the plot is headed, but ultimately the story creates its own momentum, accumulating episodes and images until it tells you, quite forcefully, where it wants to go. That’s why rereading the manuscript at this point is particularly important: images or characters that seem incidental at first glance may suddenly take on new life, until it becomes clear that no ending will be satisfying that doesn’t include some or all of these elements. In The Art of Fiction, John Gardner discusses this closing return of images at great length, going so far as to invent a sort of algebraic notation: let a represent a pair of bloody shoes, b a willow tree, c an orphan home, and so on, each of which conjures up associates of its own, until, at the end of the novel, we get something like this:

I have to admit, though, that I’ve been reading The Art of Fiction for most of my life and still have no idea what this diagram means—which only indicates that the return of images at the end of a novel is a complicated matter indeed. More useful, perhaps, is David Mamet’s concept of the slate piece. I’ve mentioned it before, but basically, in filmmaking, the slate piece is the fragment of film captured at the beginning of a take, just after the slate board has been clapped and withdrawn. Normally this piece of film, taken before the scene starts, is discarded, but sometimes it can be mined for useful footage during the editing process—say, if the director desperately needs a shot of the actor glancing to his left, in order to intercut it with a previously unrelated shot to pace up a scene. In Bambi vs. Godzilla, Mamet writes:

This accidental, extra, hidden piece of information is called the slate piece. And most of moviemaking, as a writer, a director, a designer, is the attempt not to invent but to discover that hidden information—the slate piece—that is already lurking in the film.

To a surprising extent, this is true for a novelist as well, especially as one approaches the end. So I definitely had Mamet’s slate piece in mind as I read over my first draft—which went pretty well, or at least as well as this sort of thing ever can. There’s barely a sentence or paragraph that doesn’t contain something cringeworthy, but as far as I can tell, the overall structure is sound, and it doesn’t seem to require any wholesale changes. The major issue, at the moment, is that it needs to be cut, which is only what I expected—I tend to write my drafts about twenty percent too long, which gives me a lot of leeway when the time comes to revise. I see scenes that need to be compressed, whole pages that need to be removed, chapters that need to be cut by half, which, fortunately, is something I know how to do. The hard part, then, will be finding the ending. This is what I’m going to be focusing on for the next four weeks.

Mamet’s unsubtle knife

Regular readers of this blog will know that one of my favorite books on storytelling is David Mamet’s On Directing Film. Since I’m preparing to outline my second novel over the next couple of weeks, I read it over again the other day—it’s only a hundred pages or so—and if anything, my admiration for it has only increased: it’s probably the single most useful guide to creating a plot I’ve ever seen. I’ve spoken about his advice here before, but today, I thought I’d highlight the three questions that Mamet reminds authors to ask themselves while writing:

1. What is the scene about?
2. What is the protagonist’s objective?
3. How do we know when we’re done?

It sounds simple, I know, but these three questions are the key to all good fiction. I was so invigorated by reading On Directing Film again, in fact, that I ran to the library to borrow a copy of Mamet’s other major book on writing, Three Uses of the Knife. Unfortunately, the second book isn’t nearly as good: instead of the chatty, inviting style of On Directing Film, which is structured around a series of classes that Mamet taught at Columbia University, Three Uses comes off as compressed and crabby, which was also true of his more recent Bambi vs. Godzilla.

It’s a shame, because Mamet is still just about the smartest man in movies—his intelligence shines through in all of his commentary tracks—and, judging from his earlier book, he has more useful information to impart about the craft of storytelling than anyone else I can name. In his recent work, sadly, the desire to teach seems to have given way to the temptation to lecture, a tendency which probably isn’t unrelated to his simultaneous slide into cultural conservatism. The title of Three Uses of the Knife, for instance, comes from the blues musician Leadbelly, but you need to go back to On Directing Film to figure out what Mamet means by it:

As Leadbelly says about the blues, he says in the first verse use a knife to cut bread, and in the second verse use a knife to shave, and in the third verse use it to kill your unfaithful girlfriend. It’s the same knife, but the stakes change, which is exactly the way a play or movie is structured. You don’t want to use the knife in the first verse to cut bread and in the second verse use it to cut cheese. We already know it can cut bread. What else can it do?

This is invaluable advice—you don’t want to hit the same beat twice in a row, unless you’re writing for television, in which case you’ll probably need to hit the same beat over and over again. On Directing Film contains dozens of similarly lucid pieces of guidance; Three Uses of the Knife yields perhaps two or three, because Mamet’s obsessively tight style has squeezed out the rest. Which tells me that what the world really needs is a great book of conversations with Mamet. And guess what? As luck would have it, it exists.

Tree of Codes and the slate piece

Yesterday’s post about Ralph Rosenblum’s excavation of Annie Hall from the brilliant mess of Anhedonia reminds me, not quite randomly, of Jonathan Safran Foer’s new book, Twee of Codes. (Excuse me, I mean Tree of Codes—and I can’t believe that I’m the first to make that particular joke.) It certainly has an astonishing premise: Foer has literally carved a new story out of a larger novel, Bruno Schulz’s Street of Crocodiles, by selectively removing most of the text, leaving only a few scattered words on each page. The result, which uses elaborate die cuts to remove the deleted passages, looks amazing, even if Foer sounds oddly dissatisfied by the outcome. (“French flaps would have been nice.”) If nothing else, it’s a reminder of how important a book’s physical body can be.

I haven’t read Tree of Codes yet—among other things, I don’t have forty bucks to spend on it—and while I don’t think it absolves Foer from his obligation, as a major novelist, to write more conventional fiction, I’m impressed by its ambition. It’s worth noting, though, that this particular experiment isn’t so different from what every film editor is called upon to do: extracting a useful narrative out of an existing mass of intractable material. Annie Hall is only the most extreme example. Nearly every movie, for the sake of pacing and coherence, especially if it can’t afford reshoots, is forced to combine scenes that originally appeared in different parts of the film, say, or use a reaction shot from one scene to bridge an unrelated gap. David Mamet, in his otherwise disappointing Bambi vs. Godzilla, has a terrific passage about this, which is worth quoting in its entirety:

Stuck in a scene, in the editing room, sometimes the roof falls in: an actor has not picked up his cue, and the scene stops dead—there is no cutaway (no other actor to cut to, to “pace up” the sequence), and the movie grinds to a halt.

“If only,” the director or editor says, “if only the actor sitting there like a sphinx had looked to his left: if he’d looked to his left, instead of his right, I could intercut his close-up with a shot of the other actor and pace up the scene.”

But no, the actor never looked to his left, and the scene is doomed to death. But perhaps there is one hope.

The director says, “Check the slate piece.”

What is the slate piece?

Here’s how it goes: When the shot is set up, the actors are called in and placed. The sound guy calls “rolling,” the camera is turned on, the operator tells the camera assistant to “mark it,” the assistant puts the slate board (the once actual slate with chalk markings, now electronic) in front of the lens to record on film the shot’s number and take. The shot is thus “slated,” the director calls “action,” and the take begins.

But, we may note, there was a moment, when the camera was filming, before the shot was slated, when the actor was waiting for action to be called. In this moment he may have looked to his left, his right, up or down, frowned, or smiled or yawned or done any number of things that just might magically come to the aid of a stalled or otherwise doomed shot.

This accidental, extra, hidden piece of information is called the slate piece. And most of moviemaking, as a writer, a director, a designer, is the attempt not to invent but to discover that hidden information—the slate piece—that is already lurking in the film.

Which strikes me as very similar to what Foer has done here. His task was hugely complicated, of course, by the fact that not just the words themselves but their order was fixed and immovable. To create a workable narrative from this material, his ingenuity must have been pushed to its limits—which, in itself, is admirable, especially in a writer whose work can otherwise seem, er, a little undisciplined.

It’s unclear how useful his example is to the rest of us (I can’t imagine Tree of Codes as anything but an interesting experiment) but it does raise the question of how writers can selectively use constraints as a spur to creativity. I’ll be writing more about this later today.