Ask a writer how a book gets made and you will usually get a sequence. First you plan it. Then you write it. Then you edit it. Three phases, in order, each one finished before the next begins.
Now watch an ordinary hour at the desk.
You open the chapter. Before you type a word you decide what the scene is actually for, which is planning. You write four hundred words. Halfway down, your character says something that only works if she already knows about the letter, so you decide that she does — planning again, and it has just changed chapter nine. You read the four hundred words back and cut sixty because the paragraph is explaining a thing the scene already shows, which is editing. Then you keep going.
Three jobs. One hour. No order.
The sequence describes storage, not writing
The three-phase model is not a description of how books are written. It is a description of how books used to be kept.
When the plan was index cards, the draft was a stack of paper and the revision was a marked-up photocopy, the three jobs lived in three physical objects. You could not consult the outline while your hands were on the keys, so you did not. The separation was a fact about desks and paper, and the sequence grew up around it as a coping strategy. Do all the planning first, because going back is expensive. Do all the revising last, for the same reason.
Paper is gone and the sequence stayed. Most book writing software still encodes it, and the setups writers assemble around it encode it harder: the plan in one place, the draft in another, the notes for revision in a third. The tools changed and the assumption underneath them did not.
It is worth being precise about what the assumption is. It is not that planning, drafting and editing are different jobs — they are, and treating them as the same job is its own mistake. It is that you finish one before entering the next. That the plan is settled by the time you draft. That the draft is complete by the time you revise. Every setup that puts the three jobs in three separate places is quietly betting on that, because the cost of moving between them is only acceptable if you rarely have to.
Writers make moves, not stages
Here is the thing the sequence gets backwards. Planning, drafting and revising are not stages of a book. They are moves, and a book is written by making all three of them over and over at small scale.
The moves have different sizes. Sometimes planning is a six-week structural rethink. More often it is the eleven seconds in which you decide that this scene ends before the argument resolves. Sometimes revision is a full pass with a printout. More often it is deleting a clause. The big versions of these moves are rare and memorable, which is why they end up in the model. The small versions are constant and forgettable, which is why they do not — and the small ones are most of the work.
Count them honestly for one session and see what the number is. Whatever it comes to, almost none of them will have felt like entering a phase. They felt like writing. They were writing. That is what writing is made of.
- Write for forty-five minutes as you normally would. Do not change anything about how you work.
- Keep a scrap of paper beside you. Every time you decide something about the book that is not the sentence in front of you — a character now knows a thing, a chapter has to move, this scene is doing the job you thought the next one would — put a tick on it.
- Put a different mark every time you go back and change something you wrote in this session.
- At the end, count. Then compare it to the number you would have guessed before you started — the gap is the finding.
The ticks are the point. Every one of them is a decision that now exists — and it exists somewhere. In the draft, where it will be hard to find again. In your head, where it will last about a day. Or in the plan, if the plan was close enough to reach.
You do not notice the move you did not make
This is the part that makes the false sequence durable. A setup that separates the three jobs does not stop you from moving between them. It just makes each move cost something — a window, a search, a reorientation, the small tax of remembering where that file lives. And costs of that size do not produce a decision to stop. They produce a drift.
You do not think "updating the outline would take ninety seconds, so I will skip it." You think nothing at all, and you keep writing, and the outline does not get updated. The move you did not make leaves no trace. There is no error message for a decision that was never written down, no gap in the draft where the structural thought should have been. The session still felt productive. It was productive. Something is simply missing from it, and the thing about that particular something is that it is invisible by construction.
That is why this is worth taking seriously rather than filing under minor inconvenience. Friction that produces a visible failure gets fixed. Friction that produces an absence gets absorbed, for years, and gets called being disciplined.
- You have an outline you have not opened in a fortnight, and you would describe it as "roughly right."
- You keep a running list somewhere of things to deal with in revision, and the list is where good intentions go to die.
- You can name a decision you made while drafting that you later had to reconstruct by rereading your own chapters.
- When you do open the plan, your first task is working out which parts of it are still true.
None of those is a discipline problem. They are all the same structural fact seen from different angles: the three jobs a book needs are not equally reachable from where you actually sit, which is inside the draft.
What this arc is about
The obvious next question is what the drift actually costs. It is tempting to answer "time," and that answer is both true and far too small — the seconds are trivial and nobody ever finished a book because they saved them.
The real cost lands somewhere more specific, and it lands on revision. When the plan is not reachable from the draft, the revision you do is the revision you can do: the sentence-level kind, the kind that makes prose better while leaving the structure exactly as wrong as it was. That is the subject of part two. Part three is about what "one place" has to mean before it helps, and how to actually plan, write and edit a book in one place.
For now the claim is only this. Your book does not move in three phases. It moves in three jobs, constantly, in whatever order the work demands — and the question worth asking of any writing setup is not which of the three it does well, but how far apart it keeps them.
