The prankster principle

In an interview with McKinsey Quarterly, Ed Catmull of Pixar was recently asked: “How do you, as the leader of a company, simultaneously create a culture of doubt—of being open to careful, systematic introspection—and inspire confidence?” He replied:

The fundamental tension [at Pixar] is that people want clear leadership, but what we’re doing is inherently messy. We know, intellectually, that if we want to do something new, there will be some unpredictable problems. But if it gets too messy, it actually does fall apart. And adhering to the pure, original plan falls apart, too, because it doesn’t represent reality. So you are always in this balance between clear leadership and chaos; in fact that’s where you’re supposed to be. Rather than thinking, “Okay, my job is to prevent or avoid all the messes,” I just try to say, “well, let’s make sure it doesn’t get too messy.”

Which sounds a lot like the observation from the scientist Max Delbrück that I never tire of quoting: “If you’re too sloppy, then you never get reproducible results, and then you never can draw any conclusions; but if you are just a little sloppy, then when you see something startling, you [can] nail it down…I called it the ‘Principle of Limited Sloppiness.’”

Most artists are aware that creativity requires a certain degree of controlled messiness, and scientists—or artists who work in fields where science and technology play a central role, as they do at Pixar—seem to be particularly conscious of this. As the zoologist John Zachary Young said:

Each individual uses the store of randomness, with which he was born, to build during his life rules which are useful and can be passed on…We might therefore take as our general picture of the universe a system of continuity in which there are two elements, randomness and organization, disorder and order, if you like, alternating with one another in such a fashion as to maintain continuity.

I suspect that scientists feel compelled to articulate this point so explicitly because there are so many other factors that discourage it in the pursuit of ordinary research. Order, cleanliness, and control are regarded as scientific virtues, and for good reason, which makes it all the more important to introduce a few elements of disorder in a systematic way. Or, failing that, to acknowledge the usefulness of disorder and to tolerate it to a certain extent.

When you’re working by yourself, you find that both your headspace and your workspace tend to arrive at whatever level of messiness works best for you. On any given day, the degree of clutter in my office is more or less the same, with occasional deviations toward greater or lesser neatness: it’s a nest that I’ve feathered into a comfortable setting for productivity—or inactivity, which often amounts to the same thing. It’s tricker when different personalities have to work together. What sets Pixar apart is its ability to preserve that healthy alternation between order and disorder, while still releasing a blockbuster movie every year. It does this, in part, by limiting the number of feature films that it has in production at any one time, and by building in systems for feedback and deconstruction, with an environment that encourages artists to start again from scratch. There’s also a tradition of prankishness that the company has tried to preserve. As Catmull says:

For example, when we were building Pixar, the people at the time played a lot of practical jokes on each other, and they loved that. They think it’s awesome when there are practical jokes and people do things that are wild and crazy…Without intending to, the culture slowly shifts. How do you keep the shift from happening? I can’t go out and say, “Okay, we’re going to organize some wild and crazy activities.” Top-down organizing of spontaneous activities isn’t a good idea.

It’s hard to scale up a culture of practical jokes, and Pixar has faced the same challenges here as elsewhere. The mixed outcomes of Brave and, to some extent, The Good Dinosaur show that the studio isn’t infallible, and a creative process that depends on a movie sucking for three out of four years can run into trouble when you shift that timeline. But the fact that Pixar places so much importance on this kind of prankishness is revealing in itself. It arises in large part from its roots in the movies, which have been faced with the problem of maintaining messiness in the face of big industrial pressures almost from the beginning. (Orson Welles spoke of “the orderly disorder” that emerges from the need to make quick decisions while moving large amounts of people and equipment, and Stanley Kubrick was constantly on the lookout for collaborators like Ken Adam who would allow him to be similarly spontaneous.) There’s a long tradition of pranks on movie sets, shading imperceptibly from the gags we associate with the likes of George Clooney to the borderline insane tactics that Werner Herzog uses to keep that sense of danger alive. The danger, as Herzog is careful to assure us, is more apparent than real, and it’s more a way of fruitfully disordering what might otherwise become safe and predictable. But just by the right amount. As the artist Frank Stella has said of his own work: “I disorder it a little bit or, I should say, I reorder it. I wouldn’t be so presumptuous to claim that I had the ability to disorder it. I wish I did.”

A few tips on faking it

Note: Since I’m taking a deserved break for Thanksgiving, I’m reposting a few popular posts this week from earlier in this blog’s run. This post was originally published, in a slightly different form, on December 28, 2012.

It’s always satisfying when a story comes full circle, or when a moment near the end of the narrative reveals a pattern of symbols or themes that was only dimly visible before. This kind of structure requires both careful planning and some degree of luck: a story that is too obviously structured can seem artificial or contrived, while the best kind of deep structure can take even the author by surprise. More often, however, a writer will reach the end of a project only to find that its structure is shapeless or absent, with a story that seems like nothing but a series of loosely connected events. The smart thing to do at this point would be to throw out the whole thing and start again—something that few of us have the courage to attempt. The alternative is, well, to fake it: to look for a few quick fixes that will make the story look more structured than it really is, in hopes of fooling the casual reader or critic. Is it cheating? Sure. But it’s a form of cheating of which nearly every artist has been guilty at one time or another, and once you’re aware of it, you start to see it everywhere you look. With just a few simple tricks, soon you, too, will be faking it with the best:

1. If you can’t find a theme, pretend it’s there anyway. Ideally, theme ought to arise organically from the events of the story itself, rather than being conceived beforehand or imposed after the fact. Sometimes, though, you wind up a theme that seems thin or nonexistent. The answer, if you’re determined to fake it, is to pick a theme that seems appropriate and mention it on the slightest pretext. The great recent example is Pixar’s Brave, which repeats the word “fate” so insistently that it clearly hopes that nobody notices that it doesn’t have much to do with fate at all, or at least has little of interest to say on the subject. I’m not above this kind of thing myself: when the title of my second novel was changed at the last minute to City of Exiles, which I selected more or less because it sounded good, I went back and tweaked the draft in places to tease out the theme of exile wherever possible. Hopefully, this kind of retouching should be invisible, and if you’re lucky, you’ll find a real theme lurking there after all. In storytelling, as in jazz, sometimes you just need to fake it till you make it.

2. Don’t be afraid to repeat yourself. History, as Mark Twain says, doesn’t repeat itself, but it rhymes. So, too, does a good novel: elements that occur early in the story can, and should, come back to play a larger role. As before, we’d like to believe that this is the result of serendipity or good planning, but I’ve found that it doesn’t hurt to go back, when you’re nearing the end of a writing project, to see if there are elements that could be profitably reintroduced. A character who appears only once and never returns, or a detail introduced in the book’s early pages that doesn’t play a part later on, is an annoying loose end; bring them back again at an unexpected time, and you start to look pretty smart. In City of Exiles, for instance, an unscrupulous solicitor named Owen Dancy appears early in the book, only to never be mentioned again. This struck me as an oversight, so not only did I bring him back, but I had him play a crucial part in the epilogue. As soon as something occurs twice, it starts to look like structure, and three times is even better. This kind of systematic mining of one’s work for meaningful repetitions is something that every writer should do. Like the Plains Indians, we try to use every part of the animal.

3. When in doubt, go back to where you started. When we see the NO TRESPASSING sign at Xanadu for the last time at the end of Citizen Kane, it feels like a circle has closed; the same is true of the picket fence and red roses in the opening and closing shots of Blue Velvet. At its best, this kind of bookending reflects a ring or circular structure that has been part of the work from the beginning, but sometimes only the illusion of symmetry is required. You see this in movies, like the original Spider-Man, that repeat the opening narration again at the end: it feels like a recurrence of deeper themes, when it may just be a simple editing trick. (At a higher level, you have a movie like Raging Bull, which reportedly didn’t work at all in test screenings until a snippet of the closing scene was appended to the beginning.) A true ring composition demands detailed planning, while mechanically opening and closing on the same phrase or image requires no skill at all—but if you aren’t sure how to end a story, even the fake version will often get you ninety percent of the way there. Because it’s always satisfying when a story, or a blog post, comes full circle. Isn’t it?

Blinn’s Law and the paradox of efficiency

Note: I seem to have come down with a touch of whatever stomach bug my daughter had this week, so I’m taking the day off. This post was originally published, in a somewhat different form, on August 9, 2011. 

As technology advances, rendering time remains constant.
—Blinn’s Law

Why isn’t writing easier? Looking at the resources that contemporary authors have at their disposal, it’s easy to conclude that we should all be perfect writing machines. Word processing software, from WordStar to Scrivener, has made the physical process of writing more streamlined than ever; Google and Amazon have given us access to a world of information that would have been inconceivable even fifteen years ago; and research, editing, and revision have been made immeasurably more efficient. And yet writing itself doesn’t seem all that much less difficult than before. The amount of time a professional novelist needs to spend writing each day—let’s say three or four hours—hasn’t decreased much since Trollope, and for most of us, it still takes a year or two to write a decent novel.

So what happened? In some ways, it’s an example of the paradox of labor-saving devices: instead of taking advantage of our additional leisure time, we create more work for ourselves. It also reflects the fact that the real work of writing a novel is rarely connected to how fast you can type. But I prefer to think of it as a variation on Blinn’s Law. As graphics pioneer James Blinn first pointed out, in animation, rendering time remains constant, even as computers get faster. An artist gets accustomed to waiting a certain number of hours for an image to render, so as hardware improves, instead of using it to save time, he employs it to render more complex graphics. This is why rendering time at Pixar has remained essentially constant over the past fifteen years. (Although the difference between Toy Story and Cars 2, or even Brave, is a reminder that rendering isn’t everything.)

Similarly, whatever time I save by writing on a laptop rather than a manual typewriter is canceled out by the hours I spend making additional small changes and edits along the way. The Icon Thief probably went through eighteen substantial drafts before the final version was delivered to my publisher, an amount of revision and rewriting that would have been unthinkable without Word. Is the novel better as a result? On a purely technical level, yes. Is the underlying story more interesting than if I’d written it by hand? Probably not. Blinn’s Law tells us that the leaves and grass in the background of a shot will look increasingly great, but it says nothing about the quality of storytelling. Which seems to imply that the countless tiny changes that a writer like me makes to each draft are only a waste of effort.

And yet here’s the thing: I still needed all that time. No matter how efficient the physical side of the process becomes, it’s still desirable for a writer to live with a novel, or a studio to live with a movie, for at least a year or so. (In the case of a film like Frozen, that gestational period can amount to a decade or more.) For most of us, there seems to be a fixed developmental period for decent art, a minimum amount of time that a story needs to simmer and evolve. The endless small revisions aren’t the point: the point is that while you’re altering a word or shifting a paragraph here or there, the story is growing in your head in unexpected ways. Even as you fiddle with the punctuation, seismic changes are taking place. But for this to happen, you need to be at your desk for a certain number of hours. So what do we do in the meantime? We do what Pixar does: we render. That’s the wonderful paradox of efficiency: it allows us, as artists, to get the inefficiency we all need.

A few tips on faking it

It’s always satisfying when a story comes full circle, or when a moment near the end of the narrative reveals a pattern of symbols or themes that was only dimly visible before. This kind of structure requires both careful planning and some degree of luck: a story that is too obviously structured can seem artificial or contrived, while the best kind of deep structure can take even the author by surprise. More often, however, a writer will reach the end of a project only to find that its structure is shapeless or absent, with a story that seems like nothing but a series of loosely connected events. The smart thing to do at this point would be to throw out the whole thing and start again—something that few of us have the courage to attempt. The alternative is, well, to fake it: to look for a few quick fixes that will make the story look more structured than it really is, in hopes of fooling the casual reader or critic. Is it cheating? Sure. But it’s a form of cheating of which nearly every artist has been guilty at one time or another, and once you’re aware of it, you start to see it everywhere you look. With just a few simple tricks, soon you, too, will be faking it with the best:

1. If you can’t find a theme, pretend it’s there anyway. Ideally, theme ought to arise organically from the events of the story itself, rather than being conceived beforehand or imposed after the fact. Sometimes, though, you wind up a theme that seems thin or nonexistent. The answer, if you’re determined to fake it, is to pick a theme that seems appropriate and mention it on the slightest pretext. The great recent example is Pixar’s Brave, which repeats the word “fate” so insistently that it clearly hopes that nobody notices that it doesn’t have much to do with fate at all, or at least has little of interest to say on the subject. I’m not above this kind of thing myself: when the title of my second novel was changed at the last minute to City of Exiles, which I selected more or less because it sounded good, I went back and tweaked the draft in places to tease out the theme of exile wherever possible. Hopefully, this kind of retouching should be invisible, and if you’re lucky, you’ll find a real theme lurking there after all. In storytelling, as in jazz, sometimes you just need to fake it till you make it.

2. Don’t be afraid to repeat yourself. History, as Mark Twain says, doesn’t repeat itself, but it rhymes. So, too, does a good novel: elements that occur early in the story can, and should, come back to play a larger role. As before, we’d like to believe that this is the result of serendipity or good planning, but I’ve found that it doesn’t hurt to go back, when you’re nearing the end of a writing project, to see if there are elements that could be profitably reintroduced. A character who appears only once and never returns, or a detail introduced in the book’s early pages that doesn’t play a part later on, is an annoying loose end; bring them back again at an unexpected time, and you start to look pretty smart. In City of Exiles, for instance, an unscrupulous solicitor named Owen Dancy appears early in the book, only to never be mentioned again. This struck me as an oversight, so not only did I bring him back, but I had him play a crucial part in the epilogue. As soon as something occurs twice, it starts to look like structure, and three times is even better. This kind of systematic mining of one’s work for meaningful repetitions is something that every writer should do. Like the Plains Indians, we try to use every part of the animal.

3. When in doubt, go back to where you started. When we see the NO TRESPASSING sign at Xanadu for the last time at the end of Citizen Kane, it feels like a circle has closed; the same is true of the picket fence and red roses in the opening and closing shots of Blue Velvet. At its best, this kind of bookending reflects a ring or circular structure that has been part of the work from the beginning, but sometimes only the illusion of symmetry is required. You see this in movies, like the original Spider-Man, that repeat the opening narration again at the end: it feels like a recurrence of deeper themes, when it may just be a simple editing trick. (At a higher level, you have a movie like Raging Bull, which reportedly didn’t work at all in test screenings until a snippet of the closing scene was appended to the beginning.) A true ring composition demands detailed planning, while mechanically opening and closing on the same phrase or image requires no skill at all—but if you aren’t sure how to end a story, even the fake version will often get you ninety percent of the way there. Because it’s always satisfying when a story, or a blog post, comes full circle. Isn’t it?

Brave and the fate of Pixar

Note: Spoilers follow for Brave.

It pains me to say this, but there’s no other way: I no longer fully trust Pixar. While I’m aware that this may not be a popular opinion, Brave strikes me as their weakest movie of any kind, weaker even than Cars 2. As I said at the time, Cars 2 had big problems, but it was only a rewrite or two away from being a entertaining movie. Brave, by contrast, comes off as fundamentally misconceived, and in ways that raise troubling questions about Pixar’s vaunted storytelling skills. There’s no doubt Pixar takes its storytelling very seriously, and as we saw with the recent list of narrative tips shared by artist Emma Coats, it’s developed a formidable bag of tricks. But in the case of a movie like Brave, such tricks amount to smart tactics in the service of no strategy whatsoever. Much of Brave works fine on its own terms—it’s consistently beautiful, ambitious, and rendered with a lot of love. But the more I think about it, the more it looks like a story that could only be fixed by being thrown out and radically reconceived.

At its heart, Brave‘s story is startlingly simple: a teenage princess, Merida, annoyed by her mother, Queen Elinor, casts a spell that turns her mother into a bear. This isn’t a bad premise in itself, but as handled by Brave, it suffers from three major problems: 1. Neither Merida nor her mother are strongly developed enough as characters to make the latter’s transformation meaningful. We don’t really know Queen Elinor before she’s transformed and can no longer speak, so the long sequences with Merida and Elinor as a bear can’t build on anything that came before. 2. Elinor’s metamorphosis is supposed to bring mother and daughter closer together, but there’s nothing in the situation that reveals anything new about their relationship. It’s just a generic crisis that doesn’t cast any light on the central conflict, which is that Merida is smarting under her mother’s expectations. 3. The movie’s treatment of magic is casual at best, with Merida essentially getting her spell from a witch who dispenses plot points, and the rules are never really explained, which undermines any narrative tension, especially near the end.

It isn’t hard to think of a version of this story that would have worked better than the one we’ve been given. We could make Elinor, not Merida, the central character, which automatically makes her transformation more interesting. We could turn Elinor’s father, the king, into a bear, and have mother and daughter work together to save him. We could have Merida take a rebellious interest in magic, and be drawn to a witch—not the witch we see here, but perhaps someone more like Maleficent—as an alternative mother figure in place of the queen, with disastrous consequences. Or we could even keep the story we have and approach it with a lighter touch, as Miyazaki might have done. Totoro barely has any plot at all, yet the grace of its conception makes it seem elegant rather than half-baked. Brave‘s technical splendor actually works against it here: it’s so visually compelling that it takes us a long time to realize that we’ve been given a rather simpleminded children’s movie, and that the studio gave less effort to exploring Merida’s motivations than it did to developing her hair.

In the end, we’re left with a deeply muddled movie whose constant harping on themes of destiny only makes its confusions all the more clear. Merida, for all her talk about fate, doesn’t seem to have any particular sense of what she wants out of life, and neither does the movie around her. (Just repeating the word “fate” over and over won’t convince us that you have anything interesting to say on the subject.) And the result is a film that seems less like an ordinary misfire than a tragic waste of resources. It’s possible that the change of directors was to blame, or the fact that, contrary to what the filmmakers have said, the studio was so intent on making a movie with a female protagonist and a fairy tale setting that it forgot to make either distinctive—or to see that Tangled had already done a better job. If I’m being hard on Pixar, it’s because it’s capable of far more, and I’m afraid it may see Brave as the best it can do. But it isn’t: it’s the work of a great studio that has lost its way. And only time will tell if Pixar can manage to change its fate.