Every scene in a screenplay is on trial. The charge: existing. The verdict: does the story need you, or have you been loitering?
Phase 3 · Revision & Professionalization · Week 24 of 32Last week you fixed the shape — the macro pass identified what stays, what moves, what goes, and what's missing. This week you drop one level, from arc to scene, and interrogate every scene in the draft individually. The scene pass is triage. You're walking through a field hospital after a battle, assessing each patient: this one is healthy, leave it alone; this one needs treatment, it can be saved; this one is gone, tag it and move on. The diagnostic question for every scene is the same question you learned in Week 2: what happens if I cut this? If the answer is "the story breaks" — information is lost, the causal chain snaps, the audience would be confused — the scene is essential. If the answer is "the story continues without interruption" — nothing is lost that isn't communicated elsewhere, the chain holds, the audience wouldn't notice — the scene is a candidate for cutting. A first draft always contains scenes that felt necessary during drafting but, viewed from the vantage of the complete draft, are redundant, misplaced, or structurally inert. The scene pass finds them.
The scene list. Before you cut anything, you need to see everything. The scene list is an inventory of every scene in your current draft — post–macro pass, with the CUTS and MOVES from Week 23 already applied. For each scene, you record four pieces of information: a scene number, the scene heading (location + time of day), the scene's purpose tag (its primary structural function, using the taxonomy from Week 2: establish, complicate, reveal, decide, release — or a more specific function like "deliver promise-map item #4" or "activate escalation Level 3"), and the scene's page count (how many pages it occupies). The scene list is a spreadsheet view of your screenplay — the same document viewed not as a narrative but as a collection of components, each one subject to evaluation.
Four tests for scene-level triage. Run each scene through these four tests. A scene that fails two or more tests is a strong cut or merge candidate.
Test 1: The purpose test. Can you name the scene's purpose in one sentence — its specific structural job? Not "it moves the story forward" (that's too vague to be useful) and not "it establishes character" (every scene should do that as a secondary function). The purpose must be concrete: "This scene delivers the contradiction complication — Ray's son needs the air filtration the donor funds" or "This scene is the first beat of escalation Level 2 — Graham imposes a cost." If you can't name a specific purpose, the scene may be doing atmospheric work rather than structural work. Atmospheric scenes are legitimate — but a screenplay can usually afford only three or four of them. If you have twelve scenes whose only identifiable purpose is "establishes mood" or "shows the world," nine of them need to go or be absorbed into scenes that do structural work simultaneously.
Test 2: The redundancy test. Does another scene in the screenplay do the same job? Two scenes that both establish the same relationship, deliver the same information, or create the same emotional register are redundant — one must go, and the survivor inherits whatever essential detail the cut scene contained. Redundancy is the most common scene-level problem in first drafts because writers instinctively reinforce important points: if the antagonist's threat is crucial, the writer writes three scenes showing the threat instead of one scene that makes the threat undeniable. The fix isn't to write less. The fix is to write one scene so well that the other two become unnecessary.
Test 3: The pace test. Is the scene's length proportional to its dramatic importance? A scene that delivers a minor piece of information in three pages is too long. A scene that delivers the midpoint revelation in one page is too short. The pace test isn't about absolute length — it's about relative length. The most important scenes in the screenplay should be the longest. The transitional and informational scenes should be the shortest. If your scene list shows a two-page scene at the midpoint and a four-page scene delivering a minor complication, the proportions are inverted. Either the midpoint needs expansion or the complication needs compression.
Test 4: The engine test. Does the scene have all three engine parts running — goal, friction, turn? You learned this in Week 13, and every scene you wrote during drafting should have the engine operating. But during the push to complete the draft, some scenes may have lost parts of the engine — particularly the turn. Scenes where a character enters, talks, and leaves without anything changing are the most common engine failure in first drafts. They exist because the writer needed to get from point A to point B and wrote a bridging scene that relays information without advancing the drama. The fix: either add a turn (something changes by the scene's end that wasn't true at its beginning) or cut the scene and relay its information through a line of dialogue in an adjacent scene that already has a working engine.
Merge before you cut. When two scenes do similar work, the first instinct is to cut one. But merging is often the better option — combining the essential elements of both scenes into a single, denser scene that does double duty. A scene where Nora gets information from David and a separate scene where Nora gets information from Ray can be merged into a single scene where Nora gets both pieces of information in the same conversation — or where she gets one piece from David and the other arrives by text during the same scene, creating a complication within a complication. Merging is harder than cutting because it requires rewriting rather than just deleting. But the result is a denser screenplay where every scene carries multiple payloads.
The five-scene minimum. This week's deliverable requires you to cut or merge at least five scenes. That number isn't arbitrary. A feature screenplay draft of ninety to one hundred twenty pages typically contains fifty to seventy scenes. Experience suggests that roughly ten percent of those scenes — five to seven — will fail two or more of the four tests. If you can't find five scenes to cut or merge, you're being too protective. Read the scene list again with the redundancy test dialed up. Check for scenes that deliver information the audience already has. Look for scenes whose purpose could be absorbed by the scene immediately before or after them. Five is the floor, not the ceiling.
In Week 2, you learned the 4-line rule: no action block longer than four lines. During drafting, that discipline may have loosened — under the pressure of generating pages, blocks of five, six, or seven lines creep in. The scene pass is where you tighten them. The revision-stage standard is stricter: the 3-line discipline. During the scene pass, no action block should exceed three lines unless the content genuinely requires a fourth.
The 3-line discipline isn't just about visual density — it's a diagnostic tool. A block that can't be compressed below four lines is often trying to describe too much in one breath. Breaking it into two blocks forces you to identify the two visual beats it contains and give each one its own moment on the page. The result: better pacing, clearer visual storytelling, and a page that reads faster without losing content.
The content is identical. The revision splits one block into three beats: the exit and the sedan (establishing detail), the stop-and-look (protagonist's attention shift), and the light (the discovery that something is wrong). Each beat gets its own moment. The reader's eye moves through three distinct perceptions instead of absorbing one dense paragraph. The page reads faster. The sedan and the light both land harder because they're isolated.
The trimming process often reveals unnecessary words. "Crosses the parking lot toward her car" — does the reader need to know she's heading for her car? If the car isn't relevant to the next beat, cut the destination. "Just the janitor's truck and a sedan" — the janitor's truck is set dressing. If it doesn't matter later, cut it. Trimming action lines is simultaneously a formatting exercise and an editing exercise: you're removing words that don't earn their place on the page, the same way you're removing scenes that don't earn their place in the screenplay.
10-minute drill: Open your draft to any five consecutive pages. For every action block longer than three lines, rewrite it as two or three shorter blocks — each one a distinct visual beat. Count the words you remove in the process. A typical five-page revision should cut 50–100 words through trimming alone, without removing any narrative content. Time yourself.
Assignment: Read 2 produced feature screenplays, chosen for lean, efficient scene work.
Script Pair Brief: Select one screenplay known for extreme economy — a film where every scene is ruthlessly pared, no moment is wasted, and the screenplay runs under a hundred pages while delivering a full dramatic experience. Then select one that's celebrated despite running long — a screenplay of 120+ pages where the extended length is justified by scene density, character complexity, or sustained dramatic intensity. The first teaches you what maximum compression looks like at the scene level — how much work a single scene can do when no page is wasted. The second teaches you that length isn't the enemy — redundancy is. A long screenplay where every scene earns its place is better than a short screenplay where half the scenes are empty.
Where to find scripts legally: IMSDb, The Script Lab, SimplyScripts, studio FYC releases, and public library systems.
Why this pairing: The scene pass can produce two failure modes: over-cutting (removing scenes that were doing essential emotional or atmospheric work, leaving the screenplay skeletal) and under-cutting (protecting scenes out of attachment rather than necessity, leaving the screenplay bloated). Reading both extremes calibrates your instinct: lean enough that nothing loiters, generous enough that nothing essential is missing.
Reading Lens (track these while reading):
1. For the lean screenplay: make a scene list for ten consecutive scenes. Tag each with its purpose. Can you find any scene that does only one job, or does every scene serve multiple functions simultaneously? 2. For the long screenplay: find the three longest scenes. Do they earn their length — does each page within the scene advance the drama or deepen the character? Where (if anywhere) do the long scenes sag? 3. In both screenplays, identify one scene that could be cut without breaking the story. Is it really dispensable, or is it doing invisible work — atmospheric, rhythmic, relational — that the story would miss? 4. Find a moment in either screenplay where two scenes are doing the same work (redundancy). Does the writer get away with it — and if so, what makes the repetition feel intentional rather than wasteful? 5. Look at the action-line density in both screenplays. Which one uses shorter action blocks? Does the visual density correlate with the felt pace?
Journal Prompts:
1. For the lean screenplay: choose one scene that does triple duty — advancing plot, revealing character, and testing the theme simultaneously. Analyze how the writer layers these functions without the scene feeling overstuffed. Could you restructure one of your own scenes to carry a similar payload? 2. For the long screenplay: if you had to cut fifteen pages without changing the story, which scenes or scene-sections would you remove? What's the difference between "this scene is too long" and "this scene needs every page it has"? 3. Compare the average scene lengths in both screenplays. Does the lean script simply have shorter scenes, or does it also have fewer scenes? Which strategy (shorter or fewer) is more effective for your own draft? 4. Find a scene in either screenplay that enters late and exits early — where the writer has stripped the preamble and the postscript. How late can you enter? How early can you exit? Write the entry and exit points for one of your own scenes, applying this technique. 5. After reading both screenplays, look at your scene list. Has your sense of what "earns its place" changed? Identify one additional scene in your draft that might not survive the four tests after this week's reading.
Deliverable: Scene list with purpose tags; cut or merge at least 5 scenes.
Constraints: Produce two artifacts and execute the scene-pass revisions:
(a) Scene list (2–4 typed pages). Every scene in your current draft, numbered sequentially. For each scene: scene heading, purpose tag (one sentence naming the specific structural function), page count, and a four-test score (purpose ✓/✗, redundancy ✓/✗, pace ✓/✗, engine ✓/✗). Flag every scene that fails two or more tests as a cut/merge candidate.
(b) Cut/merge execution. Cut or merge at least five scenes flagged by the four tests. For each cut: note the scene number, state what was lost (if anything), and describe where the lost element was redistributed (to which surviving scene). For each merge: note which scenes were combined, describe the merged version in 2–3 sentences, and verify that the merged scene has a working engine (goal, friction, turn). After executing the cuts and merges, update your draft — remove the cut scenes, insert the merged scenes, and verify that the causal chain still holds across the transitions.
(c) Apply the 3-line discipline across the entire revised draft. No action block longer than three lines (with rare exceptions for four-line blocks where content demands it). This is a full-manuscript pass — page by page, block by block. It's tedious. It's essential. The trimmed draft will be visibly leaner and faster on the page.
Quality bar: The five cut/merge minimum must include at least two outright cuts (scenes removed entirely, not just merged) and at least one merge (two scenes combined into one). The scene list must account for every scene in the draft — no scenes skipped or ignored. Every surviving scene must pass all four tests. If a scene fails the engine test (missing goal, friction, or turn), either fix the engine or cut the scene — do not leave engineless scenes in the draft. The 3-line discipline must be applied consistently across all pages, not just the first twenty.
Estimated time: 10–14 hours (scene list: 2–3 hours; triage and cut/merge decisions: 2–3 hours; executing cuts and merges in the draft: 2–3 hours; 3-line discipline pass: 2–3 hours; reading: remaining time).
Human Draft Reminder: You make every cut. You write every merge. AI can help you evaluate — but the knife is in your hand, and you decide where it falls.
This week's narrowed reader roles: Reader A asks "Which scenes could be cut without losing story?" — evaluating structural redundancy and necessity. Reader B asks "Where did I start skimming?" — evaluating the felt experience of pace and engagement. Submit your scene list (or a section of it — twenty to thirty scenes) to both readers along with your cut/merge plan.
Add this week's entry to your Disagreement Log. The scene-pass divergence is the sharpest of Phase 3: Reader A identifies scenes to cut based on structural redundancy (these two scenes do the same work, remove one). Reader B identifies scenes to save based on emotional necessity (that scene carries the only moment of warmth in a twenty-page stretch — cutting it makes the screenplay inhospitable). The resolution requires asking a third question that neither reader asks directly: can the emotional work be absorbed by a surviving scene? If the warm scene is cut, can the warmth be added to an adjacent scene that already has a structural purpose? If yes, cut and redistribute. If no — if the warmth depends on the specific characters, setting, or rhythm of the cut scene and can't be transplanted — keep it, and find a different scene to cut. Log the disagreement. Note which scenes survived because of emotional necessity and which were cut despite emotional attachment. The difference between those two lists is your developing editorial judgment.
Budget tier this week: Your chosen tier, directly improved by the scene pass.
Top 3 production benefits of scene-level cutting: 1. Location elimination — every scene you cut that takes place in a unique location removes that location from the production plan. If the teachers' lounge scene is cut and it was the only scene in the teachers' lounge, you've eliminated a location. Count the locations in your post-pass draft and compare to the pre-pass count. 2. Cast reduction — scenes that introduce single-appearance characters are prime cut candidates. Every cut eliminates a casting need, a contract, and scheduling complexity. Count speaking roles after the pass. 3. Shooting day efficiency — merging two short scenes in different locations into one scene in a single location means the crew stays put instead of moving. At every budget tier, crew moves are dead time — expensive hours where no footage is captured. Every merge that consolidates locations is a direct time savings on the shooting schedule.
Cheaper equivalent: A scene where Nora meets an environmental lawyer at the lawyer's office (new location, new actor, single appearance) can be replaced by a phone call in Nora's car — one-sided, her reactions carrying the drama, the legal information delivered through what she says and how she responds. No new location. No new actor. Same information. The production footprint drops to zero marginal cost.
Worth-it spend: The scenes that survived the triage. Every scene that passed all four tests has earned its place — it does structural work, it's not redundant, its length matches its importance, and its engine runs. These scenes are the screenplay. They deserve the production resources to be shot well: the right location, enough coverage, enough takes for the performance to land. The scene pass earns the right to spend wisely by ensuring that nothing is spent on scenes that don't matter.
A professional script reader has an internal page-count tolerance that adjusts by genre. A thriller that runs 115 pages feels indulgent — the genre promises economy. A character-driven drama at 115 pages gets more patience — the genre signals depth. But every reader, regardless of genre tolerance, has a skim reflex that activates when consecutive scenes feel like they're doing the same work. Two scenes in a row where the protagonist gathers information. Two scenes in a row where allies express concern. Two scenes in a row where the antagonist applies pressure without escalation. The reader doesn't think "these are redundant." The reader thinks "I know what's going to happen" — and they start skimming. The scene pass protects against the skim reflex by ensuring that consecutive scenes vary in function, register, and dramatic priority. A reader who never finds a reason to skim is a reader who reaches your ending with full engagement.
You've cut at least five scenes — pieces of writing you spent hours producing, scenes you thought were necessary during drafting. Write about one of the cuts that hurt. Not the one that was obviously dispensable — the one where you felt the loss. What did that scene mean to you as the writer? Was it a scene you were proud of technically, a scene that touched something personal, or a scene where you felt the characters most vividly? Now ask: was it cut because it was doing dispensable work, or because it was doing work the screenplay didn't need? There's a difference. A scene can be excellent and unnecessary. That's not a failure of the scene. It's a success of the triage — you found a scene good enough to miss, which means you're cutting well.
By the end of this week you should have:
• Read 2 produced screenplays with a focus on lean, efficient scene construction
• Produced a scene list with purpose tags, page counts, and four-test scores for every scene
• Cut or merged at least 5 scenes (minimum 2 outright cuts, minimum 1 merge)
• Redistributed essential elements from cut scenes to surviving scenes
• Applied the 3-line discipline to action blocks across the entire draft
• Verified the causal chain holds after all cuts and merges
• Completed the action-line trimming drill (5 pages trimmed, 10 minutes)
• Run the scene list through both Revision Specialist prompts
• Updated your Disagreement Log with this week's entry
Week 25: Rewrite Dialogue Pass (Voice + Compression). The shape is right. The scenes are trimmed. Now you drop another level — from scene to line — and work on the dialogue. The dialogue pass has two objectives: voice (every character sounds different — distinct in word choice, rhythm, avoidance patterns, and relationship to silence) and compression (every line is shorter than it needs to be, because characters in high-stakes situations don't explain themselves). You'll select twenty pages of your most dialogue-dense scenes and rewrite them with a specific test: can you identify who's speaking without character names? If every character sounds the same — if you could swap their dialogue blocks without the reader noticing — the voice work begins.