Writer's block is usually a decision problem
· The StoryProp team
The stuck day has a familiar shape. You sit down with coffee and reasonable intentions, reread yesterday's pages, adjust a comma, check a fact that didn't need checking, and forty minutes later you are tidying the kitchen. Nothing is wrong, exactly. The file is open, the story is where you left it, and there is no next sentence.
Most advice on how to overcome writer's block treats this as a mood to be managed: take a walk, change rooms, wait for the weather to pass. Sometimes that works, because sometimes it is a mood. More often, though, the block is an unmade decision wearing a mood as a disguise. Either you can't write the scene because you haven't decided what the scene is for, or the next sentence would commit you to something and a quiet, protective part of you doesn't want to be committed yet. Neither is a failure of talent, and both can be fixed at the desk.
When you are stuck, look for the missing decision
The question of what to do when you are stuck writing has a boring, reliable answer: stop pushing at the sentence and start interrogating the stall. Ask what you would have to know to write the next line — not to write it beautifully, just to write it at all. The answer is usually startlingly specific, and it is almost never about words.
Consider a scene from *The Quiet House*, the sample manuscript this blog keeps returning to. Mara has come back to her hometown. Her brother Daniel is making coffee, and the chipped cup is out on the counter — the one their mother refused to throw away. The scene stalls halfway down the page, and not because the prose is difficult. It stalls because the writer hasn't decided whether Daniel knows that Mara read the letter. If he knows, the scene is a duel, and every pleasantry in it is loaded. If he doesn't, it is a scene about Mara's guilt, and the cup means something else entirely. No sentence can be right until that question is settled, which is exactly why every sentence feels wrong.
So write the question down, in plain words, away from the manuscript. “I don't know what happens next in my story” is almost never the real problem; it is the fog that covers a smaller, sharper question. Does Daniel know? Whose scene is this? What does Mara want to carry out of that kitchen? A block you can phrase as a question is already half dissolved, because questions can be answered, and moods can only be waited out.
Decide small: the exit, not the meaning
The next trap is deciding at the wrong altitude. Stuck on a kitchen scene, a writer will convene a referendum on the entire book — what it is about, whether the ending can work, whether the premise was ever any good. That referendum cannot be won at nine in the morning, and it does not need to be. You do not have to know what the book means to write this scene. You have to know how the scene ends: who leaves the room first, and carrying what.
The scene's exit is a usefully small decision. “Mara leaves without saying it, and takes the cup” is one sentence of commitment, and it unlocks the whole page above it — now you know what the dialogue is avoiding, what the silences are doing, where to let the moment sit. Decide it provisionally if that is all you can afford. A small decision buys you an afternoon of writing; a grand one buys you another week of staring.
It also helps to write the decision down somewhere it will still exist tomorrow — a notebook, an index card, a running list of things you have settled. An open question drains a little energy every time you brush against it; a settled one, even provisionally settled, stops costing anything. A surprising amount of what feels like lost inspiration is just the fatigue of re-litigating choices you never wrote down.
Options beat a blank page
Blank freedom is the most expensive thing on a writer's desk. “Anything could happen next” sounds like liberty and functions like paralysis, because the mind is poor at choosing from an infinite, unwritten menu and very good at judging a short, concrete one. Editors know this. A good one never says “make it better”; she says “you could cut the aunt entirely, or move the reveal to the porch scene — I'd cut the aunt.” Options with consequences give your taste something to grip.
You can run the same trick on yourself. When a scene stalls, list three ways it could go, including one you suspect is bad. Daniel confronts her. Daniel says nothing and lets her twist. Daniel makes a joke that lands wrong. Then pick the least wrong. The picking matters less than the rejecting: the moment you can say why an option fails, you have learned something true about the scene you actually want.
This is also the honest case for an AI tool to overcome writer's block — not as a machine that writes the book for you, but as a generator of options for your judgment. A project-aware agent like StoryProp starts instead from the decisions you have already made — the records of canon, characters, and story shape it reads before drafting — so when you ask for three ways out of the kitchen, the exits belong to your story rather than to a generic one. And a proposal remains a proposal: what you agree to gets written down, and what you don't stays a suggestion you were never committed to.
Make being wrong cheap
The second kind of block — the protecting instinct — does not yield to decisiveness, because it is not confusion. Part of you understands that the next sentence is a commitment. Write “Daniel knew all along” and you have bent ten chapters around it; that is a real cost, and reluctance to pay it is not laziness. The fix is not courage. The fix is making the commitment reversible, so that courage gets cheap.
In practice, that means drafting both ways when the stakes feel high. Write the confrontation version of the kitchen scene, then the silent version, and read them side by side. The wrong one will announce itself on the page in a way it never does in your head. This is easiest when your tools keep every draft instead of overwriting it: in StoryProp, passes sit next to each other for comparison, a whole pass can be undone, and any earlier draft can be restored — so chasing an alternate never risks the original. However you manage it — copies of the file, dated snapshots — the principle holds: the writer who can retreat will advance.
One caution about scale. A stuck morning is a decision problem; a manuscript that has been stalled for a season is usually a different animal with different causes, and that pattern deserves its own diagnosis — it is the subject of why long projects die.
How to write when you don't feel inspired
Do not wait to feel like it. Inspiration is mostly what momentum feels like from the inside: on days when the next decision is already made, the writing feels effortless, and we name that feeling inspired. The causality runs from motion to mood, not the other way around. Which means the reliable route to an inspired Thursday is a small decision made on Wednesday.
Hence the most useful session habit I know: stop at a doorway, not at a wall. Before you close the file, decide the first move of the next session — the exit of the scene you will open with, the question the next chapter has to answer — and write it down. The stuck day is usually manufactured the day before, by stopping exactly where the decisions ran out. Leave yourself a decided edge, and tomorrow's first sentence is waiting for you, half written.
So the next time the fog rolls in, skip the referendum on your talent and run the smaller audit. What have I not decided? What is the smallest version of that decision — an exit, a knowledge state, a want? What are three ways it could go, and which is least wrong? What would make being wrong cheap enough to try? Answer those, and the mood tends to take care of itself. A block is rarely a sign that the book is broken. It is usually a sign that the book is waiting for you to make up your mind.
A stuck day usually ends the moment one small decision gets made and recorded somewhere it will survive the night. That is the motion StoryProp is built around: you ask for options, choose one, the choice becomes a record the agent reads before it drafts again, and every draft is kept — so trying the wrong exit costs nothing but the time it takes to compare.