필사 모드: Is the Three-Act Structure a Law or a Convention — The Lineage of Story Structure and Its Limits

Opening — Why Films Tend to Turn at the Same Place

Here is an experiment worth trying. Think of a commercial film you saw recently and try to remember what happened at exactly the halfway mark of its running time. There is a high probability that something flipped somewhere near there. A goal changed, a fact nobody knew came out, or a person who had been running turned around and decided to face it.

That is not an accident. Nor is it a natural law of storytelling. It is a very specific convention manufactured by the twentieth-century film industry, and it became a convention for reasons of production as much as for reasons of story.

This piece sets out where the three-act structure came from, what it is good at, and where it does harm. To state the conclusion in advance: three acts is not a law to be obeyed, but neither is it an arbitrary convention that could just as easily be otherwise. It is a compression of empirical data about the patience of an audience.

Where the Three Acts Came From — From Aristotle to Syd Field

The pedigree of the three-act structure is often said to begin with Aristotle. That is only half true.

What Aristotle said in the Poetics is that a completed whole has a beginning, a middle and an end, that events must be linked by causality, and that a good tragedy contains a reversal in which the situation turns over and a recognition in which something unknown comes to be known. All of that still holds up. But he never prescribed a number of acts. The Greek tragedy he analyzed was divided in the first place not by acts but by the entrances and exits of the chorus.

The first to make the act into a norm was Horace of Rome, who wrote that a play ought to have five acts. In the nineteenth century Gustav Freytag of Germany presented, in Technique of the Drama (1863), a five-part pyramid of introduction, rise, climax, fall and catastrophe, and that was the standard in literary education for a long time. In other words, the default value of Western narrative theory was rather five.

The three acts hardened into their present shape much later. The decisive book is Syd Field's Screenplay (1979). Field offered one simple schema. One page of a screenplay is roughly one minute of film, and a standard screenplay is about 120 pages. Act one is setup at 30 pages, act two is confrontation at 60 pages, act three is resolution at 30 pages. And between acts one and two, and between acts two and three, sits a plot point, an event that bends the story past the point of no return, placed near page 25 and near page 85 respectively.

The influence of this book cannot be overstated. And the damage came out of exactly the same place. What Field did was not to explain structure but to convert structure into page numbers, and the moment numbers were attached, it turned from a tool of creation into a criterion of assessment. The story that a studio reader flips to page 25 to check whether the plot point is there is both an industry joke and a true one. The question of why this event is needed in this place was replaced by a check for whether an event is in this place.

What the Standard Beats Actually Do

Even so, there is a reason the standard beats survived. Each beat has a function that matters more than its position, and that function is needed regardless of names or pages. Memorize the function instead of the position and the diagram turns into a tool.

The two right-hand columns are the point of this table. Beats are not rules but prescriptions against symptoms. The real reason the midpoint matters is not the number 50 percent but that any story will feel like repetition in the middle if act two is one undivided lump. Human attention dulls fast when the same kind of event keeps coming. The midpoint is the device that cuts that habituation off.

The same goes for the function of the first act turn. What matters here is not the size of the event but whether the protagonist chose it themselves. A protagonist shoved into act two looks passive for the whole remaining running time, and the audience feels, without being able to explain why, that they cannot get invested. This is also why crossing the threshold comes after the refusal of the call in the hero's journey. The refusal exists in order to make a choice look like a choice.

The crisis beat is the slot most often skipped and most regrettably wasted. For an audience to believe a character has changed by the ending, there has to be an earlier scene in which they try the old way and fail. A change arrived at without failure reads as convenience, not as feeling.

Save the Cat — The Homogeneity the Beat Sheet Made

If Field converted three acts into pages, Blake Snyder turned that into a fifteen-slot checklist. The beat sheet in Save the Cat! (2005) runs, against a 110-page screenplay, from opening image on page 1, theme stated on page 5, catalyst on page 12, debate on pages 12 to 25, break into act two on page 25, B story on page 30, fun and games on pages 30 to 55, midpoint on page 55, all is lost on page 75, dark night of the soul on pages 75 to 85, and on to break into act three on page 85. The title comes from the first piece of advice: have the protagonist do something likeable early on, like saving a cat, so the audience is hooked.

As a practical manual the book is competent. The problem is that it went beyond a guide for writing and became the common language of investment review and development meetings. Peter Suderman's well-known 2013 critique in Slate pointed out that the major beats of several box-office hits sat at almost the same minute marks. That was the point at which the phenomenon of films with different material, different actors and different directors sharing a startlingly similar rhythm began to be noticeable to audiences as well.

It matters not to misread the point of the criticism. The problem is not that the beat sheet is wrong. It mostly works well. The problem is that when one solution that works well hardens into the only solution, the effect of that solution itself wears out. Audiences learn. To a viewer who knows that everything will be lost around minute 75, that scene is not despair but a milestone. Predictability is not a side effect of structure but a side effect of success.

For the same reason, masterpieces often break the beats on purpose. Pulp Fiction (1994) scatters the chronological order, and Memento (2000) simply runs from the back. Richard Linklater's Boyhood (2014) has no distinct plot points. Parasite (2019) changes the genre of the film itself near the midpoint, and the reason that is so powerful is that the audience has been trained to expect something else in that slot. A violation of convention becomes an event only when there is an audience that knows the convention.

Alternative Structures — Kishotenketsu, the Story Circle, Slice of Life

The best evidence that three acts is not the only shape is that other shapes actually work.

Kishotenketsu is the four-part East Asian composition. It began as a craft term for composing the quatrain in classical Chinese poetry and later expanded into composition and narrative in general. The decisive difference from Western structure is the third stage. In the three-act structure the middle is the slot for escalating conflict, but the ten does not enlarge the conflict; it shifts the phase or the vantage point. Something that looks unrelated cuts in, and in the final ketsu the three preceding stages are bound together in a new meaning. The four-panel comic is the purest form of it, and stories that end satisfyingly without a distinct antagonist, such as My Neighbor Totoro (1988), are frequently cited as examples.

In fairness, though, it should be added that presenting kishotenketsu as the byword for a conflict-free narrative structure is closer to a re-reading that took hold in the West relatively recently. The usage in the original languages is broader and looser than that. Even so, the reason the frame is useful is clear: it makes you doubt the very premise that conflict is a mandatory component of story. That meaning can arise from juxtaposition and turn alone, without opposition, is a real option for a maker.

Dan Harmon's story circle is an alternative in the opposite direction. He compressed Joseph Campbell's hero's journey into eight stages. A character is in a zone of comfort, they want something, they enter an unfamiliar situation, they adapt to it, they get what they wanted, they pay a heavy price, they return to the familiar place, and they are changed. That it is a circle matters. The beginning and the end are the same spot with only the character altered, which makes the schema closer to a ledger of change than to a plot. Harmon used it repeatedly to design episodes of Community and Rick and Morty, and it has proven especially useful in a workplace that has to mass-produce twenty-two-minute stories.

Slice of life runs on an entirely different engine. Instead of events rising by causality, situation and repetition and minute differences push the story along. Jim Jarmusch's Paterson (2016) repeats nearly the same week, and in Sofia Coppola's Lost in Translation (2003) the two people do almost nothing special. That everyday-life manga and animation in Japan built an industry of their own rests on the same principle. What holds the audience here is not what will happen next but the sense of wanting to stay a while longer with this person and this space. Structure was not removed; it was swapped for something else.

What Matters More Than Structure — The Pixar Rule About Coincidence

In 2011, Emma Coats, then a story artist at Pixar, posted twenty-two rules of storytelling on Twitter. It was not an official company code but a personal list of things learned on the job, and the nineteenth of them is quoted especially often: coincidences that get characters into trouble are great, but coincidences that get them out of trouble are cheating.

That single line may matter more than the whole beat sheet. Here is why. Structure deals with where an event is placed, but this rule deals with whom an event comes from. And it is the latter that decides the satisfaction of a story.

Why is a coincidence that lands the protagonist in trouble acceptable? Because it is the setup of a situation. That it happened to rain that day, that they happened to meet that person, is a question the world throws at the character. When a coincidence rescues the character, by contrast, in that instant the character stops being the owner of their own story. The choices and the growth the audience watched for two hours become meaningless at the ending. Magic without rules wrecking the climax in fantasy and a case resolved by a clue that was never presented in detective fiction are exactly the same violation. Only the names differ: deus ex machina, soft magic, a breach of fair play.

So the real use of knowing structure does not lie in following it. It lies in diagnosing. Being able to put a name on the feeling when a story seems boring is the one certain benefit structural theory offers. If the middle sags, the midpoint is probably missing and the same kind of event is repeating. If the ending is flat, the character has probably never failed. If the protagonist is frustrating, they have probably never crossed the threshold on their own feet. The diagram is not an answer but a list of symptoms.

An Observation Exercise — Stopping at 25, 50 and 75 Percent

Checking is faster than reading. Pick any film at all. A favorite is even better.

• Stop at the 25 percent mark. Is the protagonist already inside a new situation? And was that move their own decision, or were they pushed into it? If they were pushed, watch how the director compensates for the passivity.
• Stop at the 50 percent mark. Is the protagonist's goal now the same as the goal at the start? If the goal is unchanged, the information has probably changed instead. If both are unchanged, that film was probably a little dull in the middle.
• Stop at the 75 percent mark. Is this the spot where the protagonist has lost the most? Try writing down what was lost in one sentence. If it will not go down on paper, it will be hard to explain where the feeling at the ending comes from either.

Do it three times and the table gets into your body. After that I would suggest going the other way and hunting for works that break with it. Running the same exercise on Boyhood, mentioned earlier, or on everyday-life works, or on stories that read as kishotenketsu, gives you a feel for how far the territory the diagram cannot explain extends. You learn more from the cases where the frame misses than from the cases where it fits.

Closing — Not a Law, but a Well-Organized Rule of Thumb

The three-act structure is not a law. It did not come straight down from Aristotle, and it is not a universal grammar of the human mind. It is a twentieth-century convention optimized under very specific conditions: a two-hour running time, one minute per page, and studio review.

At the same time it is not an arbitrary convention that could just as easily be otherwise. Compressed inside it is long observation of when an audience loses interest, when they come to believe in a character, and what makes a change look cheap. Mistake the convention for a law and you get homogeneous stories; ignore the convention and you generally get boring ones.

So the best attitude to take toward these diagrams is probably this. Structure is not a method for making a story but a map that tells you in advance where the story you made will lose its audience. Only someone who can read a map is able to choose to leave the road.