For me, almost everything that is called mise en scène is a big joke. In the cinema, there are very few people who are really metteurs-en-scène; there are very few who have ever had the opportunity to direct. The only mise en scène of real importance is practiced in the editing. I needed nine months to edit Citizen Kane, six days a week. Yes, I edited [The Magnificent] Ambersons, despite the fact that there were scenes not by me, but my editing was modified. The basic editing is mine and, when a scene of the film holds together, it is because I edited it. In other words, everything happens as if a man painted a picture: he finishes it and someone comes to do the touch up, but he cannot of course add paint all over the surface of the canvas. I worked months and months on the editing of Ambersons before it was taken away from me: all this work is thus there, on the screen. But for my style, for my vision of cinema, the editing is not one aspect, it is the aspect. Directing is an invention of people like you; it is not an art, or at most an art for a minute a day. This minute is terribly crucial, but it happens only very rarely. The only moment where one can exercise any control over a film is in the editing. But in the editing room, I work very slowly, which always unleashes the temper of the producers who snatch the film from my hands. I don’t know why it takes me so much time: I could work forever on the editing of a film. For me, the strip of celluloid is put together like a musical score, and this execution is determined by the editing; just like a conductor interprets a piece of music in rubato, another will play it in a very dry and academic manner and a third will be very romantic, and so on. The images themselves are not sufficient: they are very important, but are only images. The essential is the length of each image, what follows each image: it is the very eloquence of the cinema that is constructed in the editing room.
I don’t believe that the sum of the editing work is a function of the brevity of the shots. It is an error to think that the Russians worked a lot on editing because they shot in short scenes. You can spend a lot of time on the editing of a film in long scenes, because you are not content to just glue them one scene to the next.
I can’t believe that editing is not essential for the director, the only moment where he completely controls the form of his film. When I shoot, the sun determines something against which I can’t fight, the actor makes his intervention to which I must adapt myself and the story; I only manage to dominate what I can. The only place where I exercise an absolute control is in the editing room: consequently, that is when the director is, in power, a real artist, because I believe that a film is only good to the extent that the director manages to control his different materials and is not content to simply finish the film.
Unfortunately, Welles lost the battle on that one. Ah well, here's an example from The Magnificent Ambersons where he uses what Sergei Eisenstein referred to as parallel montage (editing between 2 events to create a sense of their existing simultaneously):
Welles was pragmatic about any cinematic technique he used. If he thought he needed a long take, he'd use that, too -- usually in a tracking shot, such as the one at the beginning of Touch of Evil that followed the planting of a bomb in a car to its effects down the road. For a more stridently ideological approach, you need a Russian, namely the aforementioned Eisenstein. His dialectical approach preferred a fixed camera with editing taking the brunt of interpretation, such as in his most famous scene, the Odessa Steps from Battleship Potemkin:
For Eisenstein, the third image created in the mind's eye from the synthesis of two filmed events being shown subsequently in time was the essence of cinema. The filmed events become akin to moving photographs. The logical extension of this view is Marker's La Jetée, where all movement through narrative time is by way of editing together static images. Andrei Tarkovsky, another ideologically inclined Russian, rejected Eisenstein's imperious use of montage:
The idea of "montage cinema"—that editing brings together two concepts and thus engenders a new, third one—again seems to me to be incompatible with the nature of cinema. Art can never have the interplay of concepts as its ultimate goal. The image is tied to the concrete and the material, yet reaches out along mysterious paths to regions beyond the spirit... -- from Tarkovsky's Sculpting in Time
I'm not sure what regions there are beyond the spirit -- is that meta-metaphysical? -- but Tarkovsky's point is clear enough. As Benjamin Halligan explains in this essay (from which I grabbed the preceding quote):
Eisenstein had shaped cinema in a very literary way. The associative montage that came to be termed "Eisensteinian" was, for Tarkovsky, that of the novel: the cut juxtaposes images and creates meaning, sometimes termed a "third image," although today we might consider it more along "conceptual" lines. The Eisensteinian cut was the novelist's full-stop.
When scientific understanding is lacking, spiritual interpretations tend to hold sway. And I think Halligan is correct injecting the cognitive scientific term 'concept' into the debate. What was once the Cartesian soul is now what we -- well, we atheists -- call the mind. Anyone who's read J.G. Ballard's The Overloaded Man knows what happens when concepts aren't involved in perception. In his story, a fellow manages to divorce all the objects in his environment from any conceptual or functional understanding, leaving him trapped in a phenomenal hell of meaningless forms.
This was Ballard's way of testing what would happen if you cut off half of the old Kantian adage: "Concepts without percepts are empty; percepts without concepts are blind." And it's my roundabout way of saying that Tarkovsky's exploration of time in his slow-to-non-moving camera is just as dependent on the conceptual as Eisenstein's montage. Consider the torturous candle-lighting scene from Nostalghia:
For the time-image in this scene to not be reduced to what some feral beast would encounter on a tv set, any meaning -- mental, spiritual, or "meta-spiritual" -- has to pass through (be derived from) cognitive processing. What is this man doing? Why is he doing it? How does this scene argue against Eisensteinian cinema? Etc.. All of that depends on concepts. It's not that the beauty of the Tarkovsky's material images don't create a different effect from Eisenstein's (or Welles'), only they do so by effecting a different understanding via cognition and its concepts. The scene is torturous precisely because you have to stay with man every step of the way. Contrary to Bazin, the long take is no more ontologically tied up with reality than montage -- both depend on our conceptual understanding of what reality is.
In a more favorable view of rapid editing that, similar to Tarkovsky's view, relates it to a literary style, Bryan Boyd suggests the multiplex variety of today is traceable back to Vladimir Nabokov (yet another Russian):
The average shot length in Hollywood movies has been shrinking as viewers have learned to assimilate film faster and to cope with the information rush of the modern world. Nabokov has influenced writers from acclaimed oldsters (Italo Calvino, W. G. Sebald, Salman Rushdie, Martin Amis, Orhan Pamuk) to feisty youngsters (Zadie Smith, Marisha Pessl) by introducing into fiction something akin to modern film’s reduction in shot length, its rapidity of changes of subject or perspective. I suspect that storytelling in general has speeded up our capacity to shift attention from one perspective to another. Homer generally moves from subject to subject slowly compared with modern storytelling, let alone Nabokov, but even Homer can swiftly shift level and focus when he suddenly backgrounds a warrior dying on the battlefield.
I'm not sure I like the idea of Michael Bay being a more suitable director for Lolita than Stanley Kubrick, but Boyd does give a good example of how conceptual editing is just as readily usable for understanding reality as long takes.
I leave you with an example from another master of cinematic time, Bela Tarr, wherein he demonstrates how Eastern Europe can be conceptualized with a long take over the in-camera editing of a laborious tracking shot: