Structure, character, and conflict are the architecture. Craft is the material — the quality of every moment, every scene, every sentence. A story can have a perfect structure and still be inert if the craft is weak.
Show Don't Tell
The most repeated writing advice and the most misunderstood.
"Show don't tell" doesn't mean you can never tell. It means: don't only tell. Don't describe a character's emotional state when you can depict the action or choice that reveals it. Don't narrate what happened when you can dramatize it.
Telling: "John was angry." Showing: "John set down the cup very carefully, as though it might break."
The second version does more work: it shows the anger, implies self-control, and reveals something about who John is. It takes more words but creates more experience.
The rule has a corollary: tell when showing would be inefficient. Not every minor transition needs dramatization. The craft is knowing when to dramatize and when to summarize.
Subtext is the deeper version of show-don't-tell. Characters almost never say what they actually mean — they say something that gestures toward it, deflects it, or denies it. What they say is the text. What they mean is the subtext.
"It's fine." (It's not fine.) "You don't have to come." (Please come.) "I'm not angry." (I'm furious.)
Real conversation is almost entirely subtext. When characters say exactly what they feel, it's because the situation demands it — a confession, a breaking point, a death. The rest of the time, they're circling. Drama lives in that gap between what is said and what is meant.
Scene Construction
Every scene is a mini-story. It has a beginning, middle, and end. It should start in one place — emotionally, situationally — and end somewhere different.
McKee's test: what changes in this scene? If nothing changes, the scene doesn't need to exist.
A scene that starts with the protagonist feeling confident and ends with them feeling threatened has turned. A scene that starts and ends with both characters exactly where they were is a non-scene.
The scene should do multiple things:
- Advance the plot (something changes in the external situation)
- Reveal character (we learn something new about who someone is)
- Develop theme (the scene's events connect to the story's larger argument)
A scene that only does one of these is weak. A scene that does all three earns its place.
Scene structure:
- Entry — enter the scene as late as possible (start in the middle of what's happening, not before)
- Goal — a character wants something specific in this scene
- Conflict — they can't easily get it
- Turning point — something unexpected forces a change of direction
- Exit — leave the scene as early as possible (cut before the resolution is explained)
"Enter late, leave early" is one of the most practically useful rules. The boring parts of scenes are usually the arrivals and departures. Cut them.
The Pixar Method
Pixar story artist Emma Coats distilled 22 rules of storytelling. Several are particularly relevant to scene construction and craft:
-
Rule 4: "Once upon a time there was ___. Every day, ___. One day ___. Because of that, ___. Because of that, ___. Until finally ___." The "because of that" connective tissue is what separates story from sequence. Events should cause each other.
-
Rule 6: "What is your character good at? What do they take comfort in? Now throw the polar opposite at them. Challenge them."
-
Rule 19: "Coincidences to get characters into trouble are great; coincidences to get them out of it are cheating." This is the rule against convenient writing.
Convenient Writing
The weakest form of screenplay or prose failure: characters behave in ways their established nature makes impossible, purely because the plot needs them to.
An intelligent character does something inexplicably stupid. A decisive character freezes. A careful character becomes reckless — not because the story has taken them there, but because the writer needed the next scene.
The audience notices this viscerally even when they can't articulate it. It feels wrong. It breaks the fictional dream.
The problem is almost always backwards causation: the writer knows what needs to happen (the plot beat) and works backwards to get the character there, regardless of whether the character would actually go there. The solution is to work forwards: if your character genuinely would not do the thing you need them to do, either change the character or change the plot beat.
Convenient writing is disrespectful to the audience. It assumes they won't notice or won't care. They notice. They care. Every time a character does something unmotivated for the writer's convenience, the audience is reminded they're watching a construction — and the story loses them.
Good dialogue is not realistic conversation. Realistic conversation is full of interruptions, non-sequiturs, repetition, and nothing happening. Good dialogue is compressed and purposeful — it feels real while doing far more work than real speech.
Every line of dialogue should accomplish at least one of:
- Reveal character
- Advance plot
- Create conflict
- Deliver subtext
Dialogue that only delivers information is the most common mistake. Characters explaining things to each other that both characters already know ("As you know, Bob, the virus was created in...") is called an exposition dump. It's the writer talking through the character. The audience hears the mechanism.
Better: find a reason the characters need to actually communicate — surprise, disagreement, discovery. Let the information come out because the scene demands it, not because the plot needs it delivered.
Voice is the other dimension of dialogue. Characters should sound different from each other and from the narrator. A character's vocabulary, rhythm, and what they choose not to say should be distinctively theirs.
Theme
Theme is not a message. It is not a moral. It is a question, or at most, a statement about the human condition that the story explores.
Theme is almost never stated outright by the author — it's embodied by the story. If you can extract the theme by reading the first and last pages without the middle, the story hasn't earned it.
The theme should be a live question at the start — genuinely unresolved. The story runs its test. By the end, the story has taken a position — not by telling the audience what to think, but by showing them what happened.
The easiest way to identify your story's theme: what is the protagonist's central problem? What is it really about, underneath the plot? That question, and the story's implicit answer to it, is the theme.
The First Draft
Stephen King's On Writing has the most useful advice on this: write the first draft with the door closed. Don't stop to fix. Don't re-read. Get the story out. The first draft is you telling yourself the story. The second draft is where you make it work for the reader.
Anne Lamott calls it the "shitty first draft." The point is permission: you are not allowed to write a good first draft, so stop trying. Write a complete one. A complete bad draft is infinitely more useful than an incomplete refined one.
Revision is where craft actually lives. Most published writing has been rewritten many times. The first draft provides the material; revision is where you find the story inside it.