The pickup artists

Over the last few decades, we’ve seen a series of mostly unheralded technological and cultural developments that have allowed movies to be shaped more like novels—that is, as works of art that remain malleable and open to revision almost up to the last minute. Digital editing tools allow for cuts and rearrangements to be made relatively quickly, and they open up the possibility of even more sophisticated adjustments. The Girl With the Dragon Tattoo, for instance, includes shots in which an actor’s performance from one take was invisibly combined with another, while the use of high-definition digital video made it possible to crop the frame, recenter images, and even create camera movement where none was there before. In the old days, the addition of new material in postproduction was mostly restricted to voiceovers that played over an existing shot to clarify a plot point, or to pickup shots, which are usually inserts that can filmed on the cheap. (It would be amusing to make a list of closeup shots of hands in movies that actually belong to the editor or director. I can think of two examples off the top of my head: The Conversation and The Usual Suspects.) In some cases, you can get the main cast back for new scenes, and directors like Peter Jackson, who learned how useful it could be to have an actor constantly available during the shooting of The Lord of the Rings, have begun to allocate a few weeks into their schedule explicitly for reshoots.

As I was writing this post, my eye was caught by an article in the New York Times that notes that the famous subway grate scene in The Seven Year Itch was actually a reshoot, which reminds us that this isn’t anything new. But it feels more like a standard part of the blockbuster toolbox than it ever did before, and along with the resources provided by digital editing, it means that movies can be continually refined almost up to the release date. (It’s worth noting, of course, that the full range of such tools are available only to big tentpole pictures, which means that millions of dollars are required to recreate the kind of creative freedom that every writer possesses while sitting alone at his or her desk.) But we still tend to associate reshoots with a troubled production. Reports of new footage being shot dogged Rogue One for most of the summer, and the obvious rejoinder, which was made at the time, is to argue that such reshoots are routine. In fact, the truth was a bit more complicated. As the Hollywood Reporter pointed out, the screenwriter Tony Gilroy was initially brought in for a rewrite, but his role quickly expanded:

Tony Gilroy…will pocket north of $5 million for his efforts…[He] first was brought in to help write dialogue and scenes for Rogue’s reshoots and was being paid $200,000 a week, according to several sources. That figure is fairly normal for a top-tier writer on a big-budget studio film. But as the workload (and the reshoots) expanded, so did Gilroy’s time and paycheck.

The article continued: “Gilroy started on Rogue One in June, and by August, he was taking a leading role with Edwards in postproduction, which lasted well into the fall. The reshoots are said to have tackled several issues in the film, including the ending.” This is fairly unprecedented, at least in the way it’s being presented here. You’ll occasionally hear about one director taking over for another, but not about one helming reshoots for a comparable salary and a writer’s credit alone. In part, this may be a matter of optics: Disney wouldn’t have wanted to openly replace the director for such an important release. It may also reflect Tony Gilroy’s peculiar position in Hollywood. Gilroy never seemed particularly content as a screenwriter, but his track record as a director has been mixed, so he responded by turning himself into a professional fixer, like a Robert Towne upgraded for the digital age. And the reshoots appear to have been both unusually extensive and conceived with a writerly touch. As editor John Gilroy—Tony’s brother—told Yahoo Movies UK:

[The reshoots] gave you the film that you see today. I think they were incredibly helpful. The story was reconceptualized to some degree, there were scenes that were added at the beginning and fleshed out. We wanted to make more of the other characters, like Cassian’s character, and Bodhi’s character…The scene with Cassian’s introduction with the spy, Bodhi traipsing through Jedha on his way to see Saw, these are things that were added. Also Jyn, how we set her up and her escape from the transporter, that was all done to set up the story better.

The editor Colin Goudie added: “The point with the opening scenes that John was just describing was that the introductions in the opening scene, in the prologue, [were] always the same. Jyn’s just a little girl, so when you see her as an adult, what you saw initially was her in a meeting. That’s not a nice introduction. So having her in prison and then a prison breakout, with Cassian on a mission…everybody was a bit more ballsy, or a bit more exciting, and a bit more interesting.” In other words, the new scenes didn’t just clarify what was already there, but brought out character points that didn’t exist at all, which is exactly the sort of thing that a writer does in a rewrite. And it worked. Rogue One can feel a little mechanical at times, but all of the pieces come together in a satisfying way, and it has a cleaner and more coherent narrative line than The Force Awakens. The strategies that it used to get there, from the story reel to the reshoot, were on a larger scale than usual, but that was almost certainly due to the tyranny of the calendar. Even more than its predecessor, Rogue One had to come out on schedule and live up to expectations: it’s the film that sets the pattern for an annual Star Wars movie between now and the end of time. The editorial team’s objective was to deliver it in the window available, and they succeeded. (Goudie notes that the first assembly was just ten minutes longer than the final cut, thanks largely to the insights that the story reel provided—it bought them time at the beginning that they could cash in at the end.) Every film requires some combination of time, money, and ingenuity, and as Rogue One demonstrates, any two of the three can be used to make for a lack of the third. As Goudie concludes: “It was like life imitating art. Let’s get a band of people and put them together on this secret mission.”

Rogue One and the logic of the story reel

Last week, I came across a conversation on Yahoo Movies UK with John Gilroy and Colin Goudie, two of the editors who worked on Rogue One. I’ve never read an interview with a movie editor that wasn’t loaded with insights into storytelling, and this one is no exception. Here’s my favorite tidbit, in which Goudie describes cutting together a story reel early in the production process:

There was no screenplay, there was just a story breakdown at that point, scene by scene. [Director Gareth Edwards] got me to rip hundreds of movies and basically make Rogue One using other films so that they could work out how much dialogue they actually needed in the film.

It’s very simple to have a line [in the script] that reads “Krennic’s shuttle descends to the planet.” Now that takes maybe two to three seconds in other films, but if you look at any other Star Wars film you realize that takes forty-five seconds or a minute of screen time. So by making the whole film that way—I used a lot of the Star Wars films—but also hundreds of other films, too, it gave us a good idea of the timing.

This is a striking observation in itself. If Rogue One does an excellent job of recreating the feel of its source material, and I think it does, it’s because it honors its rhythms—which differ in subtle respects from those of other films—to an extent that the recent Star Trek movies mostly don’t. Goudie continues:

For example, the sequence of them breaking into the vault, I was ripping the big door closing in WarGames to work out how long does a vault door take to close.

So that’s what I did, and that was three months work to do that, and that had captions at the bottom which explained the action that was going to be taking place, and two thirds of the screen was filled with the concept art that had already been done and one quarter, the bottom corner, was the little movie clip to give you how long that scene would actually take.

Then I used dialogue from other movies to give you a sense of how long it would take in other films for someone to be interrogated. So for instance, when Jyn gets interrogated at the beginning of the film by the Rebel council, I used the scene where Ripley gets interrogated in Aliens.

This might seem like little more than interesting trivia, but there’s actually a lot to unpack. You could argue that the ability to construct an entire Star Wars movie out of analogous scenes from other films only points to how derivative the series has always been: it’s hard to imagine doing this for, say, Manchester By the Sea, or even Inception. But that’s also a big part of the franchise’s appeal. Umberto Eco famously said that Casablanca was made up of the memories of other movies, and he suggested that a cult movie—which we can revisit in our imagination from different angles, rather than recalling it as a seamless whole—is necessarily “unhinged”:

Only an unhinged movie survives as a disconnected series of images, of peaks, of visual icebergs. It should display not one central idea but many. It should not reveal a coherent philosophy of composition. It must live on, and because of, its glorious ricketiness.

After reminding us of the uncertain circumstances under which Casablanca was written and filmed, Eco then suggests: “When you don’t know how to deal with a story, you put stereotyped situations in it because you know that they, at least, have already worked elsewhere…My guess is that…[director Michael Curtiz] was simply quoting, unconsciously, similar situations in other movies and trying to provide a reasonably complete repetition of them.”

What interests me the most is Eco’s conclusion: “What Casablanca does unconsciously, other movies will do with extreme intertextual awareness, assuming also that the addressee is equally aware of their purposes.” He cites Raiders of the Lost Ark and E.T. as two examples, and he easily could have named Star Wars as well, which is explicitly made up of such references. (In fact, George Lucas was putting together story reels before there was even a word for it: “Every time there was a war movie on television, like The Bridges at Toko-Ri, I would watch it—and if there was a dogfight sequence, I would videotape it. Then we would transfer that to 16mm film, and I’d just edit it according to my story of Star Wars. It was really my way of getting a sense of the movement of the spaceships.”) What Eco doesn’t mention—perhaps because he was writing a generation ago—is how such films can pass through intertextuality and end up on the other side. They create memories for viewers who aren’t familiar with the originals, and they end up being quoted in turn by filmmakers who only know Star Wars. They become texts in themselves. In assembling a story reel from hundreds of other movies, Edwards and Goudie were only doing in a literal fashion what most storytellers do in their heads. They figure out how a story should “look” at its highest level, in a rough sketch of the whole, and fill in the details later. The difference here is that Rogue One had the budget and resources to pay someone to do it for real, in a form that could be timed down to the second and reviewed by others, on the assumption that it would save money and effort down the line. Did it work? I’ll be talking about this more tomorrow.