KANONARA

Three-act structure

Setup, confrontation, resolution, with turning points at the act breaks: the beats of the three-act model, and how to use it without being flattened by it.

Published 24 Sept 2026 · 3 min read

Three-act structure is the foundational Western dramatic model that nearly every later framework builds on: Act I sets up, Act II confronts, Act III resolves. Syd Field's Screenplay (1979) fixed the modern version for screenwriters, and novelists use the same skeleton with manuscript percentages in place of screen time.

Its usual beats, in order: exposition, inciting incident, plot point one (closing Act I), rising action, midpoint, plot point two (closing Act II), pre-climax, climax, and resolution or denouement.

The three acts, in plain terms

Act I: setup. The ordinary world, the character within it, and the inciting incident: the event that starts the story. The act ends at plot point one, the decision or shove that commits the protagonist to the story's central problem. Rule of thumb: about the first quarter of the book.

Act II: confrontation. The longest act, and the one that sags when unplanned. Rising action complicates the problem; the midpoint is the act's hinge, where the protagonist shifts from reacting to acting (or learns something that reframes the problem); plot point two is the reverse or deepening that makes the final push necessary. Roughly the middle half.

Act III: resolution. The climax, where the central problem is met directly, and the denouement, which shows the new ordinary. About the final quarter.

What the acts are actually for

The three acts are not boxes scenes go into; they are promises about change. Act I promises a problem worth a book. Act II promises complication rather than repetition. Act III promises a confrontation the middle has earned. When a draft "feels long in the middle", the usual diagnosis is an Act II of events rather than escalations: things happen, but the problem is not getting harder.

Three tools fix most three-act problems:

  1. Move the midpoint. If the middle sags, the midpoint is usually missing: nothing reframes the problem halfway. A false victory or false defeat that changes the protagonist's understanding is the strongest middle engine there is.
  2. Earn the act breaks. A plot point is a door that closes: after it, going back to the ordinary world must be impossible or unacceptable. If your protagonist could plausibly just quit, the break is a chapter heading, not a turning point.
  3. Size the climax to the middle. The climax should spend what Act II banked: relationships strained, knowledge gained, prices paid. A climax that solves the problem with something the middle never established is the definition of an unearned ending.

Using it without being used by it

  • Percentages are habits, not laws. A first plot point at 30% instead of 25% is invisible; one at 50% is a choice you should know you are making.
  • Compressed forms are the same skeleton. Save the Cat itemizes three acts into fifteen beats; the Story Circle compresses the arc to eight steps; the Hero's Journey mythologizes it. Choosing among them is choosing resolution, not choosing a different story.
  • Some stories want another shape. Quiet, episodic and slice-of-life stories fit kishotenketsu, which builds shape from a twist instead of an escalating confrontation.

Structure and continuity are different jobs

A structure framework tells you what should happen next. It will not tell you whether chapter eighteen contradicts chapter seven; that is a continuity problem, and long stories need both an outline and a story bible for exactly that reason.

A ten-minute diagnosis

Map the draft you have to the beats above, one line each, with chapter numbers. Then read only your lines. The pattern finds most structural problems on its own: an inciting incident and then four lines of nothing new (the opening is throat-clearing), two plot points three pages apart (the second act collapsed), a midpoint indistinguishable from the beats around it (the middle is a corridor, not a hinge), a climax that introduces its solution (an unearned ending). Whatever line looks wrong when said plainly is the line worth rewriting, and the fix is almost always to move or strengthen one beat, not to re-outline the book.

Sources

  1. Wikipedia: Three-act structure
  2. Reedsy: three-act structure

Craft claims above trace to these sources. Where a point is practitioner consensus rather than settled fact, the text says so.

Keep reading