audit-dialogue-subtext

Audit Dialogue Subtext

Audit what speakers are doing with language, not how literary the dialogue sounds.

Direct dialogue is not automatically weak. The target is dialogue where the literal content is the only thing happening.

Analyze Each Exchange Across Four Layers

Literal — what the line says.
Intent — what the speaker wants the line to accomplish.
Subtext — what is communicated indirectly.
Effect — what changes because the line was said.

Core Audit Dimensions

Goal clarity

Can the speaker's conversational objective be inferred?

Tactic

Is the line an attempt to persuade, threaten, test, conceal, shame, reassure, provoke, etc.?

Tactic progression

Do tactics change when they fail?

Subtext

Does the speaker have a reason not to state the full truth, desire, fear, or accusation directly?

Emotional directness

Direct emotional speech may be powerful when earned. Flag it only when it bypasses established defenses, relationship constraints, or dramatic pressure without explanation.

Information asymmetry

Does the line respect who knows what?

Power

Does the exchange reveal or change leverage, status, dependency, or willingness to walk away?

Exposition

Is information being stated because the characters need to communicate it, or because the audience needs to hear it?

Redundancy

Does a line repeat an established fact, emotion, accusation, or intention without changing the interaction?

Finding Types

On-the-nose emotion — emotional truth stated directly where stronger character evidence suggests evasion or indirectness.
Topic without tactic — line discusses the subject but does not pursue a goal.
Repeated tactic — multiple lines attempt the same thing without adaptation.
Exposition to audience — characters state information they already share for audience benefit.
Subtext collapse — hidden conflict becomes explicit without sufficient pressure or turn.
Missing response — a meaningful provocation receives no behavioral or tactical consequence.
Power stagnation — a conflict scene contains no meaningful change in leverage or commitment.
Generic agreement loop — lines confirm one another without changing state.
Strong directness — explicitly identify cases where blunt speech is dramatically earned and effective.

Procedure

Establish scene and relationship context.
Infer each speaker's goal.
Tag meaningful lines or beats with intent and tactic.
Identify what each speaker cannot comfortably say directly.
Track information and power changes.
Locate repeated or ineffective tactics.
Audit exposition and emotional directness.
Report only material findings.
Recommend repair categories rather than automatically rewriting.

Severity

High — dialogue's dramatic engine is substantially missing or information logic is broken.
Medium — exchange works but becomes explicit, repetitive, or tactically flat.
Low — local opportunity for stronger implication or sharper tactic.

Report confidence separately.

Output

Conversation verdict

Strong / mixed / flat / information-heavy / intentionally direct.

Speaker goals

Subtext map

Beat / line	Literal	Intent	Subtext	Effect

Use only for the most informative beats on long scripts.

Material findings

Location	Finding	Why it matters	Severity	Confidence	Smallest fix

Strong moments

Identify effective implication, tactical adaptation, or earned directness.

Failure Modes

Avoid:

assuming ambiguity is always better than clarity
forcing subtext into procedural or informational dialogue where none is needed
treating every explicit feeling as bad writing
rewriting lines instead of analyzing them
ignoring relationship norms
confusing clever wording with subtext
inventing secret motives unsupported by the scene

Final Quality Check

Every material finding should be grounded in intent, information, power, relationship, causality, or pacing rather than a preference for indirect dialogue.

About this skill