"Show don't tell" means letting readers infer meaning from action, detail, and dialogue instead of receiving the writer's explanation at a distance. It is one of the most repeated rules in writing, and one of the least useful as a slogan. This guide treats it as what it actually is: a revision skill. You learn to spot telling in a finished draft, decide whether it belongs there, and rewrite only the passages that need more scene, specificity, or inference.
The phrase gets tossed around in workshops and writing forums as if the goal is to eliminate every declarative sentence. It is not. Telling has legitimate uses: compressing time, moving between scenes, delivering factual context a reader needs before the action starts. The real question is whether a given passage earns its distance from the reader or whether it shortcuts an experience the reader should have felt directly.
What follows is a diagnostic framework: five types of telling worth catching in revision, a serious argument for when telling is the right choice, applications beyond fiction, and exercises that build the skill.
Key takeaways
- "Show don't tell" is a revision technique, not a rule to follow while drafting. Write freely first, then diagnose.
- Telling shows up in five patterns: emotion labels, thought verbs, filter words, summarized scenes, and generic abstraction. Each has a different fix.
- Not all telling is a problem. Time jumps, transitions, backstory, and routine action are better told than shown.
- The principle applies beyond fiction: personal essays, case studies, business copy, and emails all benefit from concrete detail over abstract claims.
- AI-generated prose defaults to summary and generalization, making "show don't tell" a practical AI revision skill.
- Fixing every sentence is unnecessary. Identify the passages that should carry emotional weight and focus your revision there.
Why this paragraph feels flat even when it tells the truth
A paragraph can be factually correct and structurally sound and still feel like nothing. The reader finishes it and retains nothing because the writing explained the meaning instead of letting the reader arrive at it.
That gap between explanation and experience is the core of "show don't tell." The concept traces back at least to Chekhov's 1886 letter to his brother Alexander, where he advised against stating that a character felt sad and instead suggested rendering the details that would let the reader see the sadness. The often-quoted "Don't tell me the moon is shining; show me the glint of light on broken glass" is a later condensation of that advice, not a direct quote.
As a practical matter, telling tends to appear in five recognizable patterns. The table below maps each one to its typical symptom and the section where it gets addressed.
| What you notice | The telling pattern | Where to fix it |
|---|---|---|
| "She was furious," "He felt relieved" | Emotion labels | Named the emotion |
| "He realized," "She knew," "I felt" | Thought verbs and filter words | Explained the thought |
| A whole event compressed into one paragraph | Summary where a scene belongs | Summarized the scene |
| Detail on routine action that does not matter | Overwriting, showing too much | Showed everything |
| "It was beautiful," "The meeting was productive" | Generic abstraction | All sections above |
The fix is not "show more." It is "show the right things." The sections below work through each pattern with a single before/after pair, then the guide turns to when telling is the better tool.
You named the emotion instead of making the reader feel it
Emotion labels ("She was angry," "He felt lonely," "They were excited") hand the reader a conclusion instead of the evidence. The reader is told what to feel rather than placed inside the moment where the feeling lives.
The fix is physical. Replace the label with what a camera would capture: body language, dialogue, environment, or action that implies the emotion without naming it.
This does not mean emotion words are banned. "She was tired" works fine in a transition paragraph that compresses an uneventful afternoon. The label becomes a problem when it sits inside a passage that is supposed to carry weight, the scene where the reader should feel something alongside the character.
She was devastated by the news.
She set the phone on the table, screen down, and stared at the wall behind it.
The label "devastated" tells the reader what to feel. The rewrite shows behavior and lets the reader infer the emotion from the stillness.
When you are editing and a single abstract word carries the entire emotional load of a passage, try selecting it in Inki's editor and comparing alternative phrasings. Sometimes a more specific word is enough; sometimes the sentence needs restructuring around action instead.
You explained the thought instead of showing the pressure around it
Thought verbs ("He realized," "She understood," "They decided") and filter words ("She noticed," "He saw," "I heard") put the narrator between the reader and the experience. The reader watches the character process information instead of processing it alongside them.
Chuck Palahniuk's advice on "thought verbs" is useful here: instead of writing "He knew the door was locked," put the character in front of the locked door and let the reader feel the resistance.
Filter words are a subtler version of the same problem. "She noticed the coffee had gone cold" adds a perceptual layer that distances the reader. "The coffee had gone cold" puts the reader directly in the scene.
Not every filter needs cutting. Filters are useful when the act of perception matters: "She heard the footsteps stop" draws attention to the hearing itself, which might be the point in a suspense scene. Cut them when the perception is routine.
He realized that the meeting had been a waste of time.
Forty-five minutes, no decisions, and three people checking their phones under the table.
"He realized" hands the reader a judgment. The rewrite puts the reader in the room to reach that judgment independently.
You summarized the moment that should have been a scene
Summary compresses time. Scene expands it. Both are necessary, but they serve different purposes. Summary moves the narrative forward; scene makes the reader experience a specific moment.
The revision question is: does this passage describe something the reader should feel happening in real time, or is it connective tissue between scenes? If the answer is real time, it needs scene treatment: action, dialogue, setting, and sensory detail unfolding moment by moment.
A useful test from Renni Browne and Dave King's Self-Editing for Fiction Writers: if a passage begins with "After" or "Over the next few weeks," it is summary. If it begins with a character doing something specific in a specific place, it is scene. The judgment call is whether the material deserves scene treatment.
The job interview went badly. She answered the questions poorly and left feeling embarrassed.
She blanked on the second question, something about conflict resolution, and filled the silence by straightening her notebook on the table. The interviewer waited. She started a sentence, stopped, and said, "Can I come back to that one?"
The summary tells the reader the interview went badly. The scene puts the reader in the chair, watching it happen.
You showed everything, and now the draft drags
The overcorrection is real. Writers who internalize "show don't tell" sometimes expand every moment into a full scene, including the ones that do not matter. The result is a draft that drags because the reader cannot tell which moments carry weight.
Telling is the right tool for:
- Time jumps ("Three months passed," "By spring")
- Routine action that establishes pattern but does not change anything
- Backstory the reader needs before a scene makes sense
- Offstage events the reader needs to know about but does not need to witness
- Transitions between scenes
The principle is pacing. Showing slows time and intensifies focus. Telling compresses time and keeps the narrative moving. A draft that shows everything moves at one speed, and that speed is slow.
The practical test: if you removed a fully rendered scene and replaced it with a single summary sentence, would the reader lose anything they cared about? If not, the summary is better.
She walked to the door, reached for the handle, turned it clockwise, pulled the door open, stepped through the threshold, and walked down the hallway toward the elevator.
She left the office.
Unless the act of leaving carries symbolic or emotional weight, three words do the job. The expanded version wastes the reader's attention on movement that means nothing.
The advice sounds fictional, but it also fixes essays, case studies, and copy
"Show don't tell" originated in fiction craft, but the underlying principle, concrete detail over abstract assertion, applies anywhere a writer wants a reader to believe something.
In personal essays and admissions writing, Vanderbilt's writing center recommends the same approach: replace generalizations with specific scenes and observed details. "I learned a lot from the experience" is telling. A paragraph that puts the reader inside the experience and lets them see what changed is showing.
In business writing and case studies, the equivalent is the difference between claiming a result and demonstrating it. "Our onboarding process improved customer satisfaction" is a claim. "The new onboarding flow cut support tickets by [X]% in the first month" is evidence.
In copywriting, the rule adapts rather than ports directly. Landing pages need explicit claims because the reader is scanning, not reading. The showing happens in proof sections: customer quotes, specific numbers, screenshots, and scenario descriptions. The fiction version of "show don't tell" does not map 1
onto sales copy, but the instinct, use concrete evidence instead of abstract claims, does.AI gave you a clean paragraph, but it still sounds generic
AI-generated prose tends toward summary, generalization, and abstraction. This is not a quirk of a specific tool; it reflects how language models work. Models optimize for plausible next-token predictions, which converge on safe, general statements rather than specific, surprising ones.
The result is prose that tells by default. "The team worked hard to deliver results" instead of a specific scene. "Communication is key" instead of an example of what communication looked like in practice. The output reads like a summary of an article rather than the article itself.
Revising AI prose for "show don't tell" follows the same diagnostic patterns: look for emotion labels, thought verbs, summary where scene belongs, and generic abstraction. The difference is frequency. In a human draft, these patterns appear in patches. In an AI draft, they are the default mode.
Effective leadership requires strong communication skills and the ability to inspire teams toward achieving organizational goals.
The director pulled two engineers into a hallway conversation, sketched the timeline on a napkin, and asked what they would cut to hit the date.
The AI version could describe any leader in any organization. The revision puts a specific person in a specific moment doing something observable.
When you have a full draft, Inki's Review can surface passages that read as overly explanatory or flat, so you can focus your revision on the sections that need scene treatment or concrete detail. For more on editing AI output for professional contexts, see How to Edit ChatGPT Output for Business.
Three quick exercises to train your revision eye
Reading about "show don't tell" builds awareness. Practicing it builds the reflex. These exercises target the three most common telling patterns.
Exercise 1: Emotion label to scene. Take this sentence: "Mark was nervous about the presentation." Write a 2-3 sentence passage that shows Mark's nervousness without using the word "nervous" or any synonym. Aim for physical behavior, dialogue, or environmental detail. A strong answer puts the reader in the room with Mark and lets them feel the nerves without being told.
Exercise 2: Cut the filter words. Rewrite this paragraph by removing every filter word (noticed, saw, felt, heard, realized) and adjusting the sentences so the sensory information reaches the reader directly:
"She noticed the lights were off. She saw a note on the counter. She felt a draft from the window and realized it had been left open."
A strong revision delivers the same information in three direct statements with no perceptual middleman.
Exercise 3: Summary to scene. Expand this summary into a 3-5 sentence scene: "The dinner was awkward. Nobody spoke about the argument from earlier, and they left early."
Put the reader at the table: specific actions, specific silence, a specific moment where someone decides to leave. The challenge is adding scene texture without inventing unnecessary detail.
A show-don't-tell checklist before you publish
Run through this list on any draft where you suspect flat, overly explanatory passages:
- Emotionally important passages contain action, dialogue, or sensory detail rather than labels like "angry," "sad," or "excited."
- Thought verbs ("realized," "understood," "knew") appear only where the cognitive act itself matters, not as a shortcut for scene.
- Filter words ("noticed," "saw," "felt") are cut except where the act of perception is the point.
- Passages that deserve real-time treatment are rendered as scenes, not compressed into summary.
- Routine action, time jumps, and transitions are told, not shown. Not everything earns a scene.
- Abstract claims ("effective," "productive," "significant") are replaced with specific evidence or observable detail.
- Each before/after revision targets a single telling pattern, not multiple problems at once.
- You have read the draft aloud and confirmed that the emotionally important passages land with weight, not explanation.
When telling belongs, leave it. The goal is not maximum showing. It is putting the camera on the moments that matter and summarizing everything else.
For a broader editing workflow that covers rhythm, hedging, and voice alongside these patterns, see How to Self-Edit Your Writing. If your draft started as AI output, How to Make AI Writing Sound Natural covers the specific vocabulary and structural patterns that make AI prose feel flat.