Ask a casual viewer of Hindi cinema what the music in a film is "for," and you'll usually get one answer: it's the songs. The dance numbers, the romantic duets, the item track that gets released as a single months before the film itself. This answer isn't wrong. It's just incomplete in a way that hides an entire craft underneath it, because a film's songs and a film's score are not the same job, performed by the same instinct, toward the same end. They're two different tools, built to do two different things, and a music director or composer who confuses them produces a film that feels either over-scored or strangely mute in exactly the moments it most needs to speak.

The underscore is the music beneath a scene — the cue that shapes how a moment feels without ever asking to be noticed, present in countless scenes across world cinema that carry no song at all. A song, by contrast, can do something an underscore structurally cannot: it can carry a scene's entire narrative weight, advance a relationship, externalize a character's interior life so completely that, as this curriculum's Choreography department has already established, the song stops accompanying the story and becomes the story for as long as it plays. Confusing these two functions — treating the underscore like a song waiting to happen, or treating a song like background texture — is the first mistake this department exists to correct.

Start with the underscore, because the principle governing it is the same principle that has governed every department in this curriculum since its founding post: music exists to serve theme, not to decorate a scene. Sidney Lumet states this as a kind of professional confession, almost an apology on behalf of an entire art form pressed into service of someone else's story: "Music, clearly one of our greatest art forms, must be subjugated to the needs of the picture. That's the nature of moviemaking. Even though it may take over completely at certain points, its function is primarily supportive." A score that exists to showcase the composer's talent, or to manipulate an audience's emotions independent of what the story actually needs at that moment, has already failed at the job Lumet describes — and he's just as clear about the opposite failure, the one he calls "mickey-mousing": "When the score is predictable, when it duplicates in melody and arrangement the action up there on the screen, we call it 'mickey-mousing.' The reference is obviously to cartoon music, which duplicates everything down to Jerry kicking Tom's teeth in." A score that simply illustrates what's already visible on screen isn't adding meaning. It's adding redundancy, and redundancy is the opposite of what music in a film should be doing.

What should it be doing instead? Lumet's answer is the single most useful sentence in this entire post: "I don't want to 'mickey-mouse.' I want the score to say something that nothing else in the picture is saying." This is scoring to theme rather than scoring to plot, and Lumet's own account of The Verdict shows exactly what that distinction produces in practice. Paul Newman's character, Frank Galvin, has almost no explored backstory in the film — a hint of a rough divorce, a suggestion he was made the fall guy in his father-in-law's law firm, nothing more. And yet Lumet told his composer, Johnny Mandel, something specific and, on paper, completely unjustified by anything visible in the script: "I wanted the deep, buried sound of a religious childhood: parochial school, children's church choir. He was possibly an acolyte." Why invent a backstory the film never shows? Because, as Lumet explains, "the picture was about this man's resurrection, he had to have been brought up religiously, so he would have somewhere to fall from. The picture could then be about his return to faith. The score's function was to provide the state of grace from which to fall." Nothing in the dialogue tells you this. Nothing in the picture shows it. The score alone carries that information, and it carries it precisely because Lumet wasn't scoring what happens in the plot — he was scoring what the film is about, the same founding distinction this curriculum's Direction department established as the riverbed for every other decision on a film.

The Pawnbroker makes the same point even more explicitly, because Lumet describes building the entire score architecture around theme before a single note was written. He'd already established that the film is about "how and why we establish our own prisons" — Sol Nazerman, a Holocaust survivor turned pawnbroker in Harlem, having sealed himself inside a self-protective coldness so total it amounts to a kind of living death. "The concept of the score was 'Harlem triumphant!'" Lumet writes, "that the life, pain, and energy of his life there forced him to feel again." He built the entire score around two competing themes — one for Europe, one for Harlem — a musical architecture mapping the psychological war the film's theme was actually about, not a soundtrack illustrating pawnshop transactions or street scenes as they occurred. This is what scoring to theme actually looks like in practice: a composer and director deciding, before almost anything else, what emotional argument the music needs to be making underneath the story, and then building every cue to serve that argument rather than simply narrating events as they unfold on screen.

No composer in the history of cinema demonstrates the full power of this principle more completely than Bernard Herrmann, whose decades-long collaboration with Alfred Hitchcock produced some of the clearest evidence in film history that a score can function as psychological architecture rather than emotional accompaniment.

Herrmann's score for Psycho is the example every film student eventually encounters, and it's worth understanding precisely why it works rather than simply admiring that it does. The shrieking, stabbing string figures during the shower murder aren't illustrating violence the way mickey-mousing would — they're not literally timed to each thrust of the knife the way cartoon music might sync to a punch. They're rendering, in pure sound, the psychological violence of the moment: panic, fragmentation, the literal sensation of a mind and a body being torn apart simultaneously. Herrmann is reported to have originally scored the shower scene without music at all, at Hitchcock's request, before insisting the director listen to what he'd composed instead — and the now-famous strings won the argument because they did something the bare sound of water and screaming alone couldn't do: they made the audience feel the attack from inside Marion Crane's collapsing perception of what was happening to her, rather than simply watching it happen to someone else.

Herrmann's score for Vertigo works through an entirely different but equally theme-driven logic. The film is about obsession, about a man's psychological collapse into a fantasy he cannot distinguish from reality, and Herrmann's score — built around a swirling, circular motif that never resolves the way a conventional theme would — embodies that obsession structurally rather than simply commenting on it. The music doesn't accompany Scottie's vertigo. It performs vertigo, harmonically, refusing the resolution an audience's ear naturally expects, exactly the way Scottie's own mind refuses to let him resolve his fixation on Madeleine. This is scoring to theme taken to its furthest, most architectural extreme: not "what does this scene need to feel like" but "what does this entire film's psychological condition sound like, structurally, if you built it in music instead of image." Herrmann and Hitchcock's collaboration proves, across film after film, that a score's highest function isn't emotional support. It's psychological revelation — finding, in pure sound, the shape of something a script and a camera can gesture toward but never fully render on their own.

Now turn to a tradition that solved an entirely different problem — not "how does score reveal psychology" but "how does a song become powerful enough to function as a complete scene in itself" — and built, across decades, one of the most distinctive music systems in world cinema to answer it.

The Indian music director tradition occupies a structural position with no exact equivalent in Hollywood film scoring, and understanding why requires understanding the role itself. A music director in the Hindi film industry has historically been responsible not just for an underscore in the Hollywood sense, but for composing the songs that carry enormous narrative weight throughout a film — meaning the line between "composer" and "songwriter," kept largely separate in the West, collapses into a single creative authority responsible for both functions simultaneously. S.D. Burman, working across decades of Hindi cinema, built a body of work where songs functioned with the same structural seriousness Lumet describes for an underscore — but pushed further, often becoming entire scenes in their own right rather than commentary beneath one. A Burman composition wasn't simply pleasant melody layered under a romantic moment. It was frequently constructed with the same dramatic intentionality Lumet describes building into The Pawnbroker's two competing themes — a song's structure tracking a relationship's actual emotional arc across its verses, functioning less like a music cue and more like a complete dramatic unit unto itself.

His son, R.D. Burman, extended this tradition into a register of his own, building compositions that absorbed influences far beyond traditional Indian film music — jazz, funk, Latin rhythms, Western orchestration — while still operating within the same fundamental principle this entire post has been establishing: music in service of what a scene, or an entire film, actually needs to communicate. What made R.D. Burman's songs function as scenes in themselves wasn't simply their musical inventiveness. It was that each composition was built with a specific narrative or emotional job to do — a song built to carry the exact texture of a particular courtship, or a particular character's specific energy, in a way generic, interchangeable melody never could. This is the same discipline Herrmann brings to Vertigo, translated into an entirely different musical and industrial system: composition as dramatic architecture, not as decoration laid over a finished scene.

None of this would have been possible at the scale Indian cinema demanded without the playback singing system, and it's worth understanding precisely what that system changed, because it's a technical and industrial fact with direct creative consequences. Playback singing separates the performer's voice from the performer's body: a trained vocalist records a song in a studio, and the on-screen actor lip-syncs to that pre-recorded track during the picturization. This curriculum's Choreography department has already established what this technical separation made possible on the movement side — choreography unconstrained by what a human voice could sustain while simultaneously executing complex physical performance. The same separation transformed what a music director could ask a song to do dramatically. Once the singing voice and the acting body were no longer the same physical instrument, a music director could write music of far greater vocal complexity and range than most actors could have performed live, while the actor's job shifted entirely toward embodying the song's emotional content through performance and movement rather than through vocal technique. This is precisely why a Burman composition could carry such dramatic weight: the actual vocal performance, delivered by a specialist singer with full technical command of the composition's demands, could be as emotionally and technically ambitious as the scene required, completely decoupled from whatever vocal limitations the on-screen actor might have brought to a live performance.

This brings the whole department back to the question this post opened with, and it's worth stating as the central diagnostic tool you'll use throughout your work in this craft: when is a song doing a scene's job, rather than simply decorating one? The test isn't whether the song is beautiful, or catchy, or technically accomplished — Herrmann's underscores are all of those things too, and they function entirely differently from a song built to replace a scene's dialogue. The test is whether removing the song would mean removing actual narrative or emotional information the audience needs, the same way removing one of Lumet's carefully placed score cues in The Pawnbroker would remove the audience's access to Sol Nazerman's psychological war between Europe and Harlem. A song functioning as score-in-disguise — present mainly for spectacle, removable without genuine narrative loss — is doing the job an underscore should be doing instead, dressed in the costume of a song because the industry's commercial expectations demand musical numbers regardless of whether the story actually needs one at that moment. A song doing real scene-work, the way the best Burman compositions do, could not be cut without losing something the rest of the film genuinely depends on.

This is the discipline that separates a music director functioning at the level Lumet describes — asking, before any note is written, "what function should the score serve? How can it contribute to the basic question of 'what is the picture about?'" — from one simply filling contractually expected musical slots in a runtime. The tools differ enormously between Hollywood underscoring and Hindi playback composition. The underlying question, asked by Herrmann scoring Hitchcock's psychological collapse and by Burman scoring a courtship's full emotional arc alike, never changes: what does this story need its music to say that nothing else in the film is saying?

The modules ahead in this department will take apart the practical craft this post has only introduced the philosophy of — how a spotting session actually works and what a composer and director decide together before a single cue is written, how the music director system functioned as a creative and industrial structure across Indian film history, how a song actually gets composed, arranged, and recorded for picturization, and how the broader history of playback singers shaped what a song could demand of a voice once that voice was freed from the constraints of live, on-camera performance. None of that craft will matter, though, without the foundation this post has tried to establish first: a score and a song are different instruments, built to do different jobs, and a film that confuses one for the other has already lost the thread connecting its music to whatever the story actually needs said.