Today we launch another semi-regular feature: "From the Short Stack," which consists of offerings from the three dozen or so film and TV-related books that I never tire of reading.
Today's short stack selection is David Mamet's On Directing Film. Originally published in 1991, when Mamet had only three credits as a movie director, it's a concise, forceful but not totally closed-off work, at once philosophical and technical. Essentially, it's kind of a notebook by a filmmaker who's still struggling with a new medium, and who wishes to construct a set of general principles that will help him get out of his own way and make reasonably intelligent, watchable films—films that honor Mamet the screenwriter without necessarily being a slave to him.
Mamet's introduction begins with some caveats: the book, he says, came out of a Columbia University course in film direction that he undertook after having directed two pictures, House of Games and Things Change. "Like a pilot with two hundred hours of flight time, I was the most dangerous thing around," he admits. "I had unquestionably progressed beyond the neophyte stage but was not experienced enough to realize the extent of my ignorance." Having thus framed the book as a manifesto-in-progress, Mamet then proceeds to work up a set of rules that will enable a person with scant experience to direct a halfway decent movie that serves the narrative and its characters, and that makes simplicity a virtue.
I'll leave it up to House Next Door readers to say whether Mamet the filmmaker practices what Mamet the manifesto-writer preaches, or if Mamet the filmmaker ever became the equal of Mamet the playwright, or if his plays were ever such great shakes to start with. For now, I'll concentrate on one aspect of the book that continues to obsess me: Mamet's feelings on the Steadicam.
As readers of this thoroughly geeked-out blog probably know already, the Steadicam is a stabilized handheld camera rig created by cinematographer and inventor Garrett Brown. It effectively fuses a camera operator to the camera via an elaborate body harness system that displaces the camera's weight across the armature and the operator's body, allowing for smooth shots in virtually any sort of terrain, without having to install dolly tracks, a crane or anything else.
To name just a few examples, the Steadicam was employed on the Copacabana tracking shot in Goodfellas, Rocky Balboa's jog up the Philadelphia Art Museum steps in Rocky, the opening nightclub scene in Boogie Nights, and almost all of The Shining. It's become an industry workhorse—a miracle of modern science, no doubt. But in 1991, Mamet distrusted it—not because he was any sort of Luddite, but because he thought the Steadicam encouraged lazy filmmaking.
""What should film schools teach?" Mamet asks, then answers, "An understanding of the technique of the juxtaposition of uninflected images to create in the mind of the viewer the progression of the story."
Then he launches into the following screed: "The Steadicam, like many another technological miracle, has done injury; it has injured American movies, because it makes it so easy to follow the protagonist around, one no longer has to think, 'What is the shot?' or "Where should I put the camera this morning?' but if you love that morning's work at dailies, you'll hate it when you're in the editing room. Because what you're seeing in dailies is not for your amusement; it should not be 'little plays.' it should be uninflected shots that can eventually cut, one to the other, to tell the story."
Elsewhere, in a transcript of a Columbia film school workshop session where the students try to come up with a shot list that will work through a short dramatic scene, Mamet warns his class to "…tell the story in cuts. We're going to adopt this as our motto."
He goes on to say, "Obviously, there are some times when when you are going to need to follow the protagonist around for a bit; but only when it is the best way to tell the story; which, if we are dedicated in the happy application of these criteria, we will find is very seldom the case. See, while we have the luxury of time, here in class or at home making up the storyboard, we have the capacity to tell the story the best way. We can then go on the set and film it. … When we're on the set, we don't have the luxury. Then we HAVE to follow the protagonist around, and we'd better have ourselves a Steadicam."
The word "Steadicam" is asterisked. The footnote bashes on the Steadicam further: "The Steadicam is no more capable of aiding in the creation of a good movie than the computer is in the writing of a good novel—both are labor-saving devices which simplify and make more attractive the mindless aspects of a creative endeavor."
Since the publication of On Directing Film, Mamet seems to have softened his anti-Steadicam stance, as evidenced by some elaborate and (to my eye) aesthetically justifiable Steadicam work in his films Heist and Spartan. But he hasn't softened that much. He is still no Oliver Stone or Martin Scorsese or Paul Thomas Anderson (all Steadicam fiends), and he's certainly not as Steadicam-addicted as, say, TV producer John Wells (ER, The West Wing) or the directors of the CSI franchise. Mamet seems to view the Steadicam as the equivalent of a fire axe hanging in a glass case; he only busts it out when he feels he really, really needs it.
All in all, Mamet's judicious, even (cough, cough) Spartan use of the Steadicam still sets him apart from 90% of film and TV directors with budgets, who often seem to treat Steadicams as a device of first rather than last resort, and do in fact seem to view the camera as a means of letting them "cover" action, then piece it together later in the editing room, as opposed to Mamet's old school fondness for the carefully storyboarded shot list.
Although I've never had the chance to meet Mamet and query him on Steadicam theory, I suspect he'd agree that certain Steadicam shots are both spectacular and dramatically justifiable. I think when he warns that easy access to a Steadicam breeds lazy filmmaking, he's not talking about Punch-Drunk Love orRussian Ark, where the Steadicam is used heavily, to achieve very specific effects. Rather, he's saying that all things considered, it's better to decide on a particular shot, put the camera on a tripod and roll film—the Filmmaking 101 approach—rather than covering the action with a Steadicam from multiple angles (or moving through an ensemble in wide shot with a Steadicam).
I've discussed Mamet's Steadicam theory with everyone from directors and cinematographers to regular film fans to Garrett Brown himself. There doesn't seem to be much middleground; either people think the camera is a tool like any other—no more or less likely to be misused than, say, crane shots or slow motion or CGI—or else they agree with Mamet and think Steadicams are a godsend for lazy filmmakers and are used way too much.
My feelings on the subject change from day to day, so I open the floor to you. Does Mamet have a point? Or is he just old fashioned?
Matt Zoller Seitz is the creator of The House Next Door.