In 1925, the Soviet director Sergei Eisenstein released Battleship Potemkin — a film about a 1905 mutiny aboard a tsarist warship — and demonstrated, with what he called the Odessa Steps sequence, that cinema's distinctive expressive power was not in the individual shot but in the cut between shots. A close-up of a screaming woman; a wide shot of soldiers' boots advancing; a baby carriage rolling down a vast staircase; a smashed pince-nez; a stone lion seeming to rise to its feet — no individual shot says what the sequence says. The meaning is generated in the audience's mind by the juxtaposition. Eisenstein called the technique montage and claimed, theoretically and grandiloquently, that it was the entire art of cinema. He was overstating; he was also, mostly, right.
The deep claim of montage theory is that the basic unit of cinema is not the shot but the cut between shots. Lev Kuleshov demonstrated this around 1918 with what is now called the Kuleshov effect: an unchanging close-up of an actor's face was intercut, in turn, with a bowl of soup, a corpse in a coffin, and a child playing, and audiences confidently read the same expression as hunger, then grief, then fondness. The face had not changed; the meaning had been generated entirely by the juxtaposition. Eisenstein took the result as the foundation of an entire art form, theorizing several gradations of cut — from rhythmic cuts on motion to intellectual montage that pairs unrelated images to produce a third idea (politician and peacock) — and arguing that the cinema's expressive power lay in this constructive activity of the audience's mind, working at twenty-four frames per second.
What happened in practice is that two great editorial traditions sit alongside each other. Continuity editing, codified by D. W. Griffith and the Hollywood studios through the 1930s, designs every cut to be invisible — eyeline matches, the 180-degree rule, matching on action, the shot-reverse-shot of dialogue — so that the diegetic world feels continuous and the technique disappears. Discontinuity editing — Soviet montage, the French New Wave's jump cuts in Breathless, the experimental cinema of Maya Deren and Stan Brakhage — does the opposite, foregrounding the cut as expressive material. The complement to both is mise-en-scène: everything in front of the camera, staged in long takes by directors like Mizoguchi, Ophuls, and the late Kubrick, where meaning accrues within the shot rather than between shots. Pacing, measured as average shot length, ranges from Tarkovsky's thirty-plus seconds to contemporary action cinema's under three; both can produce great work, and the contemporary mainstream uses both with considerable fluency.
Every screen — film, television, streaming, YouTube, TikTok, advertising, video games — uses montage as its primary expressive tool, and editing software like Premiere and DaVinci Resolve has put professional-grade tools on consumer laptops. Music video is a montage-first form, with the visual cutting on the beat and the image subordinate to the rhythm of the cuts; YouTube essay channels have taught montage theory to a generation that will never read Eisenstein. The hardest current problem is AI video generation: Sora, Veo, and Runway can now produce credible individual shots, but coherent multi-shot sequences remain difficult — because cinema is fundamentally the art of what happens between shots, and the cuts are exactly where current models struggle. The Soviet technique Eisenstein wrote books to defend in 1929 has, a century later, become the basic visual literacy of the developed world.