On Purity in Heists

Fred Astaire and Jackie Chan agree: if you know what you're doing, in a filmed dance or fight, you don't want the camera to move, and they expressed this agreement in pretty similar ways, Astaire saying "either the camera dances or I do", and Chan "when the camera angle moves, that means the actors, they don't know how to fight … I never move my camera. Always steady". You don't have to show all the action with a relatively static, wide shot just because you can, just because the action you'd be showing will be well executed, but if you're making a dance movie, or a martial arts movie, not doing so would be kind of perverse. On the other hand, if your actors are not good at dancing or movie fighting, you certainly still have the option of shooting them in the wide without a lot of cuts, but the result will be more obviously bad than if you make it flashier.

There's an analogy to this in heist films, escape films, the kind of film, in general, in which the overcoming of obstacles in service of a goal is a large part of the scaffolding of the plot. Consider the difference between Rififi and (the prompt for this post, though plenty of other examples can be furnished) the first episode of the Lupin TV series. Rififi shows the heist straight through. I admit I can't remember how they gain access in the first place via the adjacent storefront; I believe this too is explained if not depicted in detail, but certainly from that point on—the actual break-in, the disabling of security, the theft, the escape—we see it happen, over the course of half an hour (a quarter of the movie's runtime). Not only are there no surprises in the execution of the robbery revealed to the viewer after the fact, there aren't even any surprises as it happens. The viewer already knows how it's going to go, because we've seen the robbers forming the plan, figuring out how they're going to do it. A key moment comes when they discover how to deal with the new alarm the store has recently installed. These men are essentially engineers and we watch them overcome a difficult technical problem in a believable way. (Not only was the process of figuring it out depicted believably, the solution is not just believable but actually functional, leading to copycat robberies in the real world. This, perhaps, is superogatory.) The actual execution of the plan is still remarkably tense—the fact that we know the plan doesn't mean that carrying it out will automatically succeed, after all—but it's notable that we seem them do pretty much exactly what they told us they'd do.

How different is the syuzhet of Lupin! The plot concerns the theft of a necklace, and neither the theft itself nor the planning thereof is depicted straightforwardly. (Assane, the thief, is planning everything himself, so there's no real reason for it to be articulated aloud, but he does have help, and obviously they could have arranged for it to be joint work if they'd wanted to.) Everything comes out of order, each mysterious tidbit explained later in the order of depiction by a flashback that situates and, theoretically, explains it.

Why would you do it like this? It's not quite parallel to the fight scene case, where we can imagine that a director is working with a fully-formed script and firm set of casting choices, and must now figure out how to depict a fight with these performers. This structure—I assume—is as written, not a subsequent reworking of a plot initially written linearly. But this structure has a very clear advantage, which is that it enables you to obscure the fact that Assane's plan doesn't make any fucking sense. If there had been a scene in which Assane explained to his jeweler friend what his scheme was, in detail similar to that which we see in Rififi, the audience would rightly scoff. It is dependent on coincidence, luck, and the unwitting participation of third parties to an absurd degree. The flashback structure allows us to see something happen which we don't understand, or which we see only partially, and then to see it more fully, with the same event getting more layers in subsequent flashbacks. Each little bit moving forward in time poses a question: why is he doing this? How is this going to work? How's Assane going to writtle out of this one? Each flashback gives us a bit of insight: Ah! That's how it happened! That's how he did it! The ensuing back-and-forth of puzzle and answer impresses us with how neatly the complex scheme came together, because it did happen thus and so, and keeps us feeling satisfied, because each immediate question is resolved, and then we're on to another one, and the resolution, being retrospective, makes matters feel settled.

A plan, however, is not retrospective but prospective. The primary service the mixed-up temporality of the syuzhet serves is that it replaces a plan, which must take into account future possibilities and contingencies, with a record of events, which only did happen one way. But it's the intelligence embodied in the prospective plan that, to me at least, you want in a heist movie. You want the mastermind to actually have masterminded something, not to be playing hope chess. Rififi's explicit depiction of the planning itself (Le Cercle Rouge shares this feature, but it is clearly a Rififi homage) is both reality effect and discipline: it is not impossible to simply depict the heist and then explain how the plan worked retrospectively (I wouldn't call Inside Man an unqualified success, but the hostage/robber intermingling seemed effectively done in this regard to me), but having someone say "we will do this, and then that will happen" invites the audience, and hopefully the writer, to say: "will it, though? What if it doesn't?". (Obviously, this is no panacea. Several of the elements of the plan in Ocean's 8 that are discussed in advance are completely ridiculous, but it just barrels ahead anyway. But then, both it and the similarly afflicted Ocean's 11 belong firmly to the "beautiful face, huge" school of filmmaking.)

There may be other modes of presentation which would also serve Lupin, but the linear style of Rififi could never work. When the temporal structure of the heist moves, that means that the plan is nonsense.