When you’re stuck in rewrite hell — version after version, none of them actually better — the problem usually isn’t your writing skill. It’s something quieter that most writers never name.
Most revisions fight themselves. Here’s why.
Every working writer has felt the difference between a revision session that compounded — the draft genuinely got better, structurally, in ways that held — and a session that felt productive in the moment but produced a draft that wasn’t actually any stronger. Same writer. Same craft skill. Two sessions that looked identical from the outside. One worked. One didn’t. Most of us file the difference under mood or fatigue and move on.
It isn’t skill, and it isn’t mood. It’s a missing step in the cognitive operation, and once you can name the step, you can stop skipping it.
This dispatch will do two things. First, show what the conventional “passes” model — dev, line, copy — is hiding from you about what revision actually is. Second, give you a four-stage loop, borrowed from somewhere outside writing entirely, that names the move where craft thinking actually happens — the move most writers cut out of their revision sessions without realizing they’ve cut anything at all.
The Origin: Why Passes Made Sense
The pass model wasn’t a mistake. It was a fix for a real problem.
Writers who revise without structure tend to polish surface prose while structural issues sit underneath, untouched. They spend an hour rewriting a paragraph that should have been cut, miss the chapter break that’s in the wrong place, and emerge from the session with a manuscript that reads slightly better at the sentence level and is no closer to working at the book level. The pass model — developmental first, then line, then copy — solves that. It forces big issues to the front, prevents wasted polish on prose that’s about to be deleted, and gives the writer a sequence to follow when instinct won’t say what to fix next.
That’s a real win, and the argument here depends on conceding it. The pass model is genuinely useful, especially for writers who haven’t internalized story structure deeply enough to revise without scaffolding.
But the pass model is a scheduling metaphor. It tells you the order to do things in. It doesn’t tell you what you’re actually doing when you revise — what the cognitive operation looks like underneath the schedule. Writers who stop at the schedule have only modeled half the work, and the half they’ve left unmodeled is exactly where revisions either compound or burn.
The Loop: What’s Actually Happening
The framework I want to borrow comes from a fighter pilot. John Boyd, working on air-combat doctrine in the 1950s, noticed that the pilot whose decision cycle is more accurately oriented — not faster, more accurately oriented — wins, because they consistently make moves that address the actual situation while their opponent is making moves that address an imagined one. He called the cycle the OODA loop: Observe, Orient, Decide, Act. It migrated out of cockpits into business strategy, security operations, anywhere decisions happen under pressure with imperfect information.
Revision is exactly that. Imperfect information. Decisions under pressure. A cognitive cycle the writer runs, often unconsciously, every time they touch a draft.
Walk the four stages, applied to editing:
Observe — Read what’s actually on the page. Not what you intended. Not what you remember writing six months ago. What’s there, on this read, today.
Orient — Frame what you observed against your structural model. What is this scene trying to do? What does the chapter owe the larger arc? What does the genre contract ask of this beat? What does the character’s voice in this passage require? This is the step that does the craft work. This is where structural and prose knowledge gets applied to the raw observation.
Decide — Choose the specific revision move. Once Orient has happened correctly, Decide is usually fast and obvious. The fix is implied by the framing.
Act — Make the change.
Then the loop closes. Observe again. The new prose is new evidence, and the loop runs on the changed material. Orient against it. Decide again. Act again.
The key insight is that Orient is where structural and craft thinking actually happens. It’s where you reach into your knowledge of story, genre, voice, character, and pacing and use it to interpret what you’ve observed. Drop Orient, and revision becomes raw observation feeding directly into raw action — which is exactly what panicked, unstructured revision feels like from the inside. You read a passage, feel friction, change something, read again, feel friction, change something else. No frame. No accumulating model. Just twitching at the prose.
The loop’s other quiet strength is that it delivers its own feedback. You don’t need an editorial letter or a critique partner to tell you what comes next — the change you just made becomes the next thing you observe, and re-orienting against it surfaces the next move. Every turn of the loop generates the input for the turn after it. The feedback is built in.
That self-feedback is also how the loop tells you when to stop. You run it on the scene, then on the chapter, then on the whole book, and you keep running it until a full turn produces no change — until you Observe, Orient, and find nothing the framing asks you to Decide. That’s a real stopping condition, not just fatigue. A finished scene isn’t one you’ve gotten tired of looking at. It’s one where the loop runs clean.
The pass model collapses into this loop, not the other way around. A “dev pass” is a series of OODA loops where Orient is structural. A “line pass” is a series of OODA loops where Orient is prose-level. They’re the same operation at different scales. Pretending they’re separate is what makes the boundaries leak — every line edit can surface a structural issue, every structural fix forces line-level rewriting, and writers who insist on the clean pass boundary end up either ignoring the problems they surface or restarting their pass from scratch.
The pass model schedules revision. The loop describes it. Most writers know the schedule and have never named the operation.
The Diagnostic: Where Writers Skip Orient
If the operation is the loop, then the failure mode is the missing step. And in my experience, almost every revision that didn’t compound is a revision that collapsed Observe straight into Act and skipped Orient entirely.
Here’s the diagnostic question. Between reading a passage and deciding to change it, can you name what structural model you used to interpret what you read? If the honest answer is I just felt it was off — you’ve skipped Orient. The friction was real. The interpretation was missing.
There are three common shapes the skip takes.
The first is the polish reflex. You read a passage, feel friction, and start tightening sentences. The friction may be prose-level — but it may equally be a structural issue showing up as prose friction. The scene’s job is unclear, the character’s POV obligation isn’t being met, a setup is missing. The friction is real; the location of the cause is somewhere else. Polishing the surface buries the structural signal under cleaner prose, and the next time you read the chapter, the friction will be back, harder to find, because now the surface is smooth.
The second is the tighten reflex. Cutting words feels productive. Word counts go down, red ink goes up, the session feels like it’s accomplishing something. But cutting words before Orient means you may be cutting the very prose that’s load-bearing for a structural beat you haven’t framed yet. The clearest version of this failure is the writer who tightens a long passage to a clean, sharp paragraph and then realizes three chapters later that the cut prose was carrying a setup the payoff now lacks. The tightening was pristine. The structural function was erased.
The third is instinct trust. Experienced writers have internalized Orient enough that it sometimes runs unconsciously — and sometimes the unconscious version is reliable. The trouble is that the writer can’t always tell the difference between I’ve Oriented automatically and I’ve skipped to Decide and called it Orient. Long drafts make this worse. By chapter twenty-two, your unconscious model of what an earlier scene was doing has been corrupted by everything that happened in chapters one through twenty-one, and the unconscious Orient is running on stale framing. The revision feels right. The result contradicts chapter four.
The forced-Orient move is simple and a little annoying. After reading a passage, before changing anything, write one sentence that names what the passage is trying to do at the scene, chapter, and arc level. If you can’t write the sentence, you haven’t Oriented. Don’t decide yet.
In practice, this turns margin notes from fix this into this passage’s job is X; the prose isn’t doing X because Y. The first half forces Orient. The second half feeds Decide. The single discipline — articulating purpose before authorizing change — is the move that turns a revision session from word-shuffling into structural work.
The slow-down rule, if you want a one-liner: if a revision session changed prose without changing your understanding of any scene’s job, you weren’t running a loop. You were polishing.
Where This Gets Complicated
The honest concession: some writers have internalized story structure so deeply that they Orient unconsciously and reliably, and for them the forced sentence is friction with no added signal.
But most of us — even experienced writers — have patchy Orient, and the forced loop is most valuable exactly when you’re sure you don’t need it. The move is to notice the difference between unconscious Orient that’s reliable and unconscious Orient that you’ve stopped checking. The forced sentence is how you check: twenty seconds on the chapters where your instinct was already right, and on the chapters where it wasn’t, the cost of skipping it is the next three drafts.
Externalizing Orient — putting the framing in language outside your head, in margin notes or a working document or whatever scaffolding helps — is one way to keep the discipline honest. What matters is that it exists somewhere before the first change gets made; the form it takes is up to you.
The Takeaway
The pass model schedules revision. The OODA loop describes what revision actually is. Writers who learn only the schedule keep skipping the step where craft happens.
The replacement is the loop. Watch yourself for the Observe-to-Act collapse. Force a one-sentence Orient before any change. Slow down by exactly the amount required to make every change structurally informed — no more, no less. The compound interest pays out across the draft because every change you make is now framed against the same accumulating model of the work, instead of being a series of disconnected reactions to whatever friction the last paragraph happened to produce.
Try one experiment in your next revision session. After reading a passage, before changing anything, write a single sentence naming what the passage is supposed to do. Notice how often you can’t. That gap — between what you’ve read and what you can articulate as its purpose — is the loop you’ve been failing to close. Closing it is the entire game. And when you’ve closed it everywhere — every scene, then every chapter — until a full loop turns up nothing left to change, the loop runs clean. That’s the closest thing you’ll get to the draft telling you it’s done.
Until next time, thanks for listening.
Leave a Reply