
The useful question is not whether the writing is good enough. It is which component has failed, because components fail in recognizable ways.
The diagnostic approach to how to write a novel works like this: describe what the manuscript is doing wrong in mechanical terms rather than emotional ones, locate the earliest point where that behavior begins, and fix there. Symptoms surface downstream of causes, so repairs applied where the pain is felt tend to fail.
When You Stop Opening the File
Avoidance is data. A writer who was working steadily and has now not opened the manuscript in three weeks is usually avoiding a specific thing, not writing in general.
Find out what. Open the file and read the last scene you wrote. Often there is a decision you did not want to make: a character who has to do something unlikeable, a subplot you know is dead, a chapter that will have to go. The avoidance lifts once the decision is made, and it does not lift before.
Timelines That Stop Adding Up
Days go missing, seasons repeat, a pregnancy resolves in four months. Most drafts develop this quietly, and it surfaces late, usually in a beta reader's note about something feeling off.
Build a simple table: chapter number, day or date, location, who is present. An hour of this catches nearly everything. It also tends to expose pacing faults that had nothing to do with chronology — three chapters covering one afternoon, then a leap over the six months where the character actually changed.
Likability Is Usually the Wrong Diagnosis
When readers say they did not connect with a protagonist, writers tend to soften the character. This almost never works, and it often makes the problem worse by removing the friction that made them interesting.
The actual fault is usually opacity or passivity. Readers stay with unpleasant characters who want something visibly and act on it. They abandon pleasant characters who drift. Before softening anything, check whether the protagonist has made a decision with a cost in the last forty pages. Nine times out of ten, they have not.
Rushed Usually Means Missing Scenes
The opposite note — it moved too fast, I could not keep up — is almost never about sentence speed. It means significant events were summarized, so the reader was informed of a change instead of watching it happen.
Find the turning points and check whether each is staged. A relationship that ends in narration, a decision reported after the fact, a confrontation covered in a sentence: those are the gaps. Adding scenes at three or four of them fixes more pacing complaints than any amount of trimming elsewhere.
A Draft That Only Sees
Search the manuscript for sound, smell, temperature, and texture. Many drafts are almost entirely visual, which makes them feel oddly remote without anyone being able to say why.
The repair is not to sprinkle sensory detail evenly. It is to pick the two or three moments that most need physical presence and ground those specifically, through a perceiving character in a particular mood. What someone notices characterizes them. What they fail to notice characterizes them more.
Revising Yourself Into a Worse Book
It is possible to address every note and end up with a weaker manuscript. This happens when a writer implements prescriptions rather than diagnosing the experiences behind them.
Symptoms are data; solutions are opinions. A reader who was confused was genuinely confused, but their suggested fix comes from a book they would have written rather than yours. Before changing anything, ask what the note implies about the manuscript's behavior. Then design your own repair, and check it against the reason the scene existed.
Showing the Draft Too Early
Handing out an incomplete manuscript is a common stall. The feedback arrives before the book knows what it is, and it freezes the draft in a shape the writer had not committed to.
If the impulse is for reassurance rather than diagnosis, name that and ask for the reassurance directly, from someone who is not reading critically. Substantive notes belong after a complete draft. A half-manuscript can only be assessed as a fragment, and every reader will invent a different completion.
A One-Hour Manuscript Audit
1. Chapter, day, location, who is present — build the table
2. Mark every scene as staged or summarized
3. Note which chapters could be removed without consequence
4. Circle each point where the protagonist chooses something
5. Read the list, not the manuscript, and look for the gap
Diagnostic Questions
Why does my novel feel boring in the middle?
Almost always because the protagonist has stopped causing events and started receiving them. Check the last several scenes: if they open with news arriving or decisions made elsewhere, the engine has stalled. A choice with a real cost restarts it faster than adding incident.
How do I know if the problem is structure or prose?
Apply a line-level fix to one troubled chapter and reread the surrounding material. If the same discomfort reappears a few pages later in slightly different form, the fault is structural. Prose problems stay local; structural ones recur wherever you look.
Should I fix a stalled draft or start over?
Fix it, in most cases. Starting over usually reproduces the same fault at the same point, because the fault came from a decision made before page one. Diagnose first — if the cause is a premise that never resolved, that is repairable without discarding the pages.
Pick one symptom and run one test on it before the week ends. A single mechanical observation about your real pages beats another month of general advice on how to write a novel.
소설쓰는법