Using the Framework to Fix What’s Broken

A draft problem lives at one of four levels, the same four dimensions that have organized this book from the start: structural, genre, arc and technique. Your first diagnostic act is finding which level a given problem originates at. "Something’s wrong" isn’t a diagnosis.

"The protagonist feels passive" is a symptom, and that symptom traces back to one of several distinct origins:

Apply a character-level fix to a structural absence, deepen the backstory, add interior monologue, make the protagonist more likable, and nothing gets better, because the cause never lived at the character level.

Which is why the four dimensions are a diagnostic sequence and not a menu.

At every level the underlying question stays the same: has the story earned its effect here, and if not, at what level did the earning fail? The four dimensions were a planning instrument in the opening chapters. In revision they become a diagnostic instrument, run in the same order.

The primary structural instrument is the reverse outline. An ordinary outline comes before writing. The reverse outline comes after, and it tells you what you actually wrote rather than what you meant to write. Go through the draft scene by scene and record four things for each:

The reverse outline makes invisible patterns visible at a glance: the sixty pages carrying no new information, the thirty consecutive scenes where the protagonist reacts rather than decides, the midpoint sitting at forty percent instead of the center. A scene that drives nothing, changes nothing and serves no function is a structural orphan, however well written.

The sagging middle that kills more long-form fiction than anything else is exactly this: protagonist agency draining away, things happening to the protagonist while other characters and coincidences make the decisions.

Subplots get diagnosed on the same map. Chart when each one appears, goes dormant and resolves, because a subplot that vanishes for a hundred pages and returns with no bridge is interrupting rather than integrating. A subplot earns its place only when it tests the main plot’s thematic question from another angle and leaves marks that reach the main plot. One that resolves cleanly offscreen, or contradicts the main plot’s emotional logic, is wallpaper rather than structure.

With the map in hand, an eleven-question checklist tests each major beat as a yes or a no:

Any answer that’s no, or uncertain, marks a structural origination point, and because structural problems show up downstream from where they start, the fix belongs upstream of the symptom.

The single most useful test among these is the wrong-strategy specificity check from Chapter 7: The Wrong Strategy and the B-Story. Complete the sentence: "my protagonist’s wrong strategy is to [specific method] because they believe [specific false belief]." Charles Foster Kane’s completes cleanly. He attempts to purchase love, to fill every relationship with things until the volume of things stands in for the thing he needs. That’s why Citizen Kane (Orson Welles, 1941) never wheels: every scene in the long middle either advances that strategy or shows its cost.

A sentence you can’t complete with specifics is an Act 2 that’s structurally underfunded, and no amount of line polish will fund it. The sequence framework from Chapter 2: The Structural Skeleton is what makes this locatable: a writer who can name the sections for the genre can locate any structural problem by position.

Once the structural skeleton holds, the genre diagnostic asks whether the story is honoring a specific genre contract, and which one. The most common genre-level failure is drift. A story that began as a thriller gradually loses its information-management discipline and reaches the climax with no clear epistemological resolution. A romance piles up pleasant relationship scenes but never makes the specific emotional stakes of the midpoint’s genuine connection legible.

The instrument here is the genre-specific naming tables built across the genre parts, used now as drift detectors. Name the eight sequences and twenty-four sections in the genre vocabulary the story has implicitly adopted, then ask, at each position, whether the scene delivers what its genre-specific name promises.

If the position that should be the Grand Gesture contains no gesture, if the scene that should be the Ordeal contains no cost, the genre contract at that position is empty. The naming table can’t be fooled by general story-ness, because it demands the specific vocabulary of the genre the story committed to. It’s the same instrument the cross-genre chapter used to detect collisions in hybrids, the dual name read at a single position, turned now on a single-genre draft to detect not collision but absence.

With structure and genre sound, the arc diagnostic asks which of the three arc types the story is attempting and whether it’s actually executing it. The most common arc failure is the flat-arc protagonist drifting into a positive arc by default.

A flat arc requires a protagonist who holds a truth the world resists. The protagonist doesn’t transform; the world does. But suppose you haven’t built enough structural resistance, haven’t organized the world specifically against the truth the protagonist holds. Then the protagonist gradually softens, accommodates, grows more open, and arrives at the ending changed rather than as the catalyst for the world’s change. Nobody intended that transformation. It happened because the pressure a flat arc requires was never built tightly enough.

Each arc has its own confirming question.

Vague arcs produce vague endings. An ending where the protagonist simply "grows" or "changes," with no nameable belief on either side of the change, is the downstream symptom of an arc nobody ever specified. The fix is at the arc level, not in the ending scene.

Technique comes last because it can only be evaluated once structure and arc are correct, and four specific technique diagnostics carry forward from earlier chapters.

What breaks when you run the sequence out of order is predictable, and it accounts for most wasted revision.

Each one is revision at the wrong level, and each leaves the actual cause untouched. So the diagnostic runs as a protocol, one distinct pass per level:

Any stage that surfaces a problem gets repaired before the next stage begins. Revision isn’t a single pass through the draft. It’s four passes, and they can’t be collapsed.

The hardest moment in any of them is the kill decision, the scene you love that isn’t earning its structural place. The test there is never whether the scene is well written but whether the story needs it, because attachment isn’t a structural argument. Nothing gets deleted in revision, only reclassified: the cut scene may belong to a different book, or it may have been your own necessary discovery that the reader doesn’t need to watch.

The deepest thing the four-level diagnostic gives you is that every draft problem has an address. "Something’s off" is a feeling. The framework converts the feeling into a location: which level, which sequence, which beat, which structural function. The moment the address is named, the revision turns specific rather than general.

Specific revision is recoverable. General revision compounds, because it changes things without knowing why, and a draft revised generally tends to pick up new problems faster than it sheds old ones.

This four-level diagnostic is the book’s whole synthesis turned into a working instrument. And it’s the same instrument, run in the same order, that builds a novel from a blank page in Chapter 91: From Blank Page to Complete Structural Map, where the four dimensions plan what has to go right rather than diagnose what went wrong. Planning and diagnosis are one tool used at two different moments in the process.