Writing dialogue realistic

WhyLabel: Dialogue is not a transcript of real life; it is compressed action where characters rarely say exactly what they mean.

HowLabel:

• Read the section on 'The Gap' between what is said and what is thought.
• Identify the 'Text' (the spoken words) and the 'Subtext' (the hidden desire or emotion).
• Practice writing a three-line exchange where a character asks for water but actually means 'I am terrified'.

DoneWhenLabel: You can identify the hidden motive in every line of a sample scene.

WhyLabel: Technical mastery of dialogue tags and beats prevents the reader from being distracted by 'clunky' writing.

HowLabel:

• Focus on the chapter 'Dialogue Mechanics'.
• Learn to use 'Action Beats' (e.g., 'He looked away') instead of adverbs (e.g., 'He said sadly').
• Internalize the rule: Use 'said' or 'asked' 90% of the time to keep tags invisible.

DoneWhenLabel: You have a list of 5 common dialogue mistakes to avoid in your first draft.

WhyLabel: Real speech is messy, repetitive, and full of interruptions, which adds flavor to fictional dialogue.

HowLabel:

• Go to a public space (cafe, park) and listen to a conversation for 15 minutes.
• Note how people use 'fillers' (um, like), interrupt each other, and leave sentences unfinished.
• Do not use these exactly as recorded, but use the 'rhythm' to break up perfect, grammatical sentences in your writing.

DoneWhenLabel: You have a one-page transcript of real-world speech patterns.

WhyLabel: Characters should be identifiable by their speech alone, without needing a dialogue tag.

HowLabel:

• Define their 'Vocabulary Level' (Academic, slang-heavy, minimalist).
• Choose a 'Speech Rhythm' (Short/staccato vs. long/meandering).
• Assign one 'Verbal Tic' or recurring phrase that reflects their background.

DoneWhenLabel: You have a profile for at least 3 characters detailing their unique speech style.

WhyLabel: Writing in a character's voice without the pressure of a plot helps you 'hear' them.

HowLabel:

• Let the character describe their favorite childhood memory.
• Focus entirely on their specific word choices and sentence structures.
• Read it back to see if it sounds like a real person or a generic narrator.

DoneWhenLabel: You have three distinct monologues that sound like three different people.

WhyLabel: Dialogue is most effective when characters want opposing things.

HowLabel:

• Set a timer for 45 minutes.
• Write a scene where two characters argue about something trivial (e.g., a lost key) while the real conflict is something deep (e.g., lack of trust).
• Avoid 'On-the-nose' dialogue where they state their feelings directly.

DoneWhenLabel: A 500-word draft of a high-tension dialogue scene.

WhyLabel: Showing a character's reaction through action is more immersive than telling the reader through an adverb.

HowLabel:

• Find every instance of 'he said [adverb]' (e.g., 'he said angrily').
• Replace it with a physical action (e.g., 'He slammed the glass onto the table').
• Ensure the action moves the plot or reveals character state.

DoneWhenLabel: All '-ly' adverbs are removed from your dialogue tags.

WhyLabel: Your ears catch rhythmic awkwardness that your eyes miss.

HowLabel:

• Read your dialogue scenes out loud at a normal speaking pace.
• Mark any spot where you stumble or run out of breath.
• Simplify sentences that feel 'written' rather than 'spoken'.

DoneWhenLabel: You have read the entire draft aloud and smoothed out all 'stumble points'.

WhyLabel: Real dialogue starts as late as possible and ends as early as possible to maintain narrative momentum.

HowLabel:

• Look at the beginning of every scene; if it starts with greetings, delete them.
• Look at the end of every scene; if it ends with 'See you later', delete it.
• Start the scene right when the conflict or the 'point' begins.

DoneWhenLabel: Every scene starts 'in media res' (in the middle of things).

WhyLabel: External feedback confirms if your characters actually sound distinct to a stranger.

HowLabel:

• Give a scene to a critique partner or friend.
• Remove all dialogue tags (he said/she said).
• Ask them if they can tell who is speaking based only on the words used.

DoneWhenLabel: A reader correctly identifies the speaker in at least 80% of the untagged lines.