Dialogue Under Pressure: A Practical Revision Workbook

A hands-on companion to the blog post of the same name | Ford Mountain Publishing, LLC


This workbook doesn’t recapitulate theory. It puts the tools in your hands.

Each section targets a specific structural failure. Come to each one with a scene from your current manuscript — or use the sample prompts provided. Work through them in order the first time. After that, drop into whatever section matches where you’re stuck.

One rule before you begin: Don’t polish lines until you’ve worked through Sections 1–3. Word-level revision is the last step, not the first.


Section 1 — Scene Pressure Audit

Most flat dialogue starts here: the scene has no structural pressure. Before you touch a single line, you have to diagnose the architecture underneath it.

The Test: If the characters spoke nothing but complete, honest truths to each other — would anything in the scene actually change? If the answer is no, the scene isn’t built yet.

Choose a scene from your current manuscript. Fill in:

Character A wants from Character B in this specific scene: (Not their arc goal. Not their thematic purpose. What do they want from this person, right now?)

Character B wants from Character A:

What happens if the truth is fully spoken in this scene?

What does each character most need to avoid revealing — and why?

Diagnostic Verdict

If your answers were…Then you need to…
Vague or interchangeable (“to connect,” “to understand”)Rebuild the scene’s specific stakes before writing any dialogue
Identical for both charactersGive each character a competing want — collision is the structure
“Nothing changes if the truth is spoken”Add cost to honesty: social, emotional, or informational

Section 2 — Power Tracking

Dialogue is a power struggle. That doesn’t mean your characters need to be hostile — it means every exchange shifts something. Who knows more? Who needs more? Who is performing, and who is watching the performance?

Take the same scene from Section 1. Read through it and answer:

Who controls the conversation at the opening? (Who sets the terms, the topic, the tone?)

Where does control shift — and what causes the shift? (A revelation, a deflection, a silence, a piece of information suddenly shared or withheld?)

Who holds power at the scene’s end? Same person or different?

What did the losing character lose? What did the gaining character win?

If no shift occurred: The scene is exposition wearing a costume. Return to Section 1 and identify what each character is protecting — then let that thing be threatened.

Power Shift Checklist

  • ☐ The power dynamic is different at the scene’s end than at its beginning
  • ☐ I can identify the specific line or beat where control shifted
  • ☐ The character who loses power is changed by that loss — not just inconvenienced
  • ☐ Something has been revealed, decided, or discovered that cannot be taken back

Section 3 — Before / After: Controlled Rewrite

This section trains one skill: getting emotion off the page and into the architecture of the scene. When characters name what they feel, the reader has nothing left to discover. Discovery is what keeps people reading.

The Rule: No emotion labels. No direct confessions. The feeling must be present — but it can only exist in behavior, in what is avoided, in what is said instead of the real thing.

Prompt 1

“I was scared you wouldn’t come back.”

Rewrite this line — or the exchange that contains it — without any emotion words, without any direct confession of feeling. The fear must be present. It cannot be named.

Your rewrite:

Prompt 2 — From Your Own Manuscript

Find a line in your current project where a character states an emotion directly. Write it here, then rewrite it below.

Original line:

Rewrite (no emotion labels, no direct confession):

Ask yourself after each rewrite:

  • Is the emotional truth still present — just not stated?
  • What does the character do instead of confessing?
  • What does the other character hear — which may not be what was intended?

Section 4 — Resistance Injection Drill

Resistance is the engine under subtext. When characters answer directly, honestly, and completely — the scene has nowhere to go. This drill builds the habit of indirect response.

The Drill: Write 6–10 lines of dialogue between two characters. At least 3 responses must avoid answering directly. One character must change the topic under pressure.

Use this scenario, or substitute one from your manuscript:

One character has come back after a long absence. The other has been waiting. Neither of them will say what they actually need.

Your dialogue (6–10 lines):

After writing, go back and mark:

  • Circle each line where the responding character deflects rather than answers
  • Underline the line where the topic changes — note what caused the change
  • Put a star next to the line that carries the most weight without saying the most
  • Identify: what does each character want that they never ask for directly?

Section 5 — Genre Transformation Drill

The same line carries completely different pressure depending on genre context. This drill develops your ability to load a simple line with the expectations, costs, and power dynamics specific to each form.

Base line: “You shouldn’t be here.”

Rewrite this line — and the one response that follows — in each genre below. Change as much as needed, but preserve the core meaning: presence is forbidden; the speaker has the authority to forbid it.


Romance Version Think: forbidden presence, desire that can’t be named, something to lose.

Thriller Version Think: surveillance, danger, what gets said in a room where the walls have ears.

Fantasy Version Think: power, law, consequence — what does it mean to be where you are not permitted?

Reflection: What changed between versions? What did you add — specificity, consequence, physical detail — to make the line carry weight in its new context? That added element is what your original scene may be missing.


Section 6 — Compression Exercise

Most dialogue scenes are too long — not because the writer said too much, but because not enough of what was said is load-bearing. This exercise identifies which lines are structural and which are filler masquerading as texture.

The Exercise: Take a dialogue scene of at least 10 lines. Remove 30% of lines. Then remove another 20%. Each time, preserve only the lines that carry tension. If removing a line doesn’t hurt — that line was not working.

Step 1 — First Pass (remove 30%): Mark or cross out lines that don’t shift power, reveal information, or cost the speaker something.

(Work on a printed copy, a separate document, or in the margin of your draft.)

Step 2 — Second Pass (remove another 20%): Of what remains — which lines could be implied by what surrounds them?

After compressing, ask:

  • Does the scene still make sense? (If not, a structural element is missing — restore the function, not the filler)
  • Is the tension higher or lower after compression?
  • What do the removed lines tell you about what the scene was actually doing?

Section 7 — Consequence Test

If a line can be cut without changing what happens next — it is not doing structural work.

The Test: For each line of dialogue in your scene, ask: What changes — in the scene, in the relationship, in what each character knows — because this line was spoken? If the answer is “nothing,” cut or rewrite.

Line of DialogueWhat changes because this was said?

Note: A line doesn’t have to cause a dramatic event to carry consequence. It can shift tone, reveal a secret through what it avoids, or change what one character now knows about the other. Small shifts are still shifts.


Section 8 — Read-Aloud Audit

This is the final pass — and the only one that works at the sentence level. Read your revised scene aloud. Your ear will catch what your eye misses.

Mark, don’t rewrite — yet. Flag the problems first. Then return to each one with the section’s questions before revising.


1. Unnatural Rhythm

Lines that catch in your mouth when spoken aloud. Constructions no one would use in speech. Sentences that feel written, not said.

  • Where did you stumble or pause unnaturally?
  • Which lines required re-reading to understand?

Lines flagged for rhythm:


2. Identical Voice Patterns

Both characters sound like the same person. Same sentence length, same register, same vocabulary. If you removed the dialogue tags — would you know who was speaking?

  • Which exchanges could belong to either character?
  • Where does one character’s vocabulary bleed into the other’s?

Lines flagged for voice:


3. Emotional Flatness

Lines that are delivered, not felt. The words are technically correct but nothing is at stake in them.

  • Which lines could be said by a robot?
  • Where is the character speaking rather than fighting, hiding, or calculating?

Lines flagged for flatness:

Revise only the flagged lines. Untouched lines are working. Leave them alone. The goal is surgical — in and out — not a full rewrite.


A Final Note

Dialogue revision is really scene revision. The line-level polish matters — but it is the last thing, not the first. If you have worked through this workbook from Section 1 forward, you have already done the harder and more important work.

What you want is a scene where something is at stake, someone is resisting something, and the cost of speaking — or not speaking — is real. When those conditions exist, the dialogue finds its own tension. Your job becomes restraint, not invention.

“Subtext only works when something is being actively avoided.” Build the avoidance into the architecture. The rest follows.


This workbook is a free companion resource from Ford Mountain Publishing, LLC. If your scenes — or your dialogue — need deep structural work, our developmental editing and manuscript assessment services are designed for exactly this kind of work. Learn more at fordmountainpublishing.com/services.


WANT MORE?

SIGN UP TO RECEIVE THE LATEST UPDATES, BEHIND THE SCENES, PLUS SOME EXCLUSIVE GOODIES!

We don’t spam! Read our privacy policy for more info.


Discover more from Ford Mountain Publishing, LLC

Subscribe to get the latest posts sent to your email.

What do you think?