Implicature Audit¶
Inference audit — instantiates Pragmatic Context Alignment
Examines a message for what it implies beyond the literal words — what's suggested by what's said, unsaid, emphasized, or softened — and flags the risky implications.
People routinely mean more than they say, and hear more than is said. Implicature Audit examines a message — not a particular reader — to catalog the meanings it generates over and above its literal content: what gets implied by word choice, by what's conspicuously omitted, by emphasis, ordering, or hedging. Its defining move is to set the literal content beside the implied content and list the gap, then flag which implications are load-bearing (an indirect request that the reader must act on) and which are hazards (an unintended suggestion the wording leaks). It is message-centric and inventory-producing: where a walkthrough simulates one recipient's full reasoning, the audit sweeps the text itself for everything it conversationally implies.[n1]
Example¶
A product team drafts a mobile push notification: "Your free trial ends in 2 days. Upgrade now to keep your data." The literal content is a factual deadline plus an action. The Implicature Audit sets that beside what users will infer: that "keep your data" implies data will be deleted if they don't pay (an implicature the team never intended — the data is only frozen, not erased); that "Upgrade now" plus the countdown implies scarcity and pressure; and that the notification is, in effect, an indirect demand for payment dressed as a helpful reminder.
The audit's output is a two-column list — literal claim on the left, likely implicature on the right — with two rows flagged red: the false deletion threat (a hazard the wording leaks) and the buried implicit action requirement (users must act before a hard deadline, stated softly enough that some will miss it). The fix follows directly: say plainly that data is preserved and retrievable after upgrading, and make the required action and its true deadline explicit rather than implied. Nothing in the literal text was false; the danger lived entirely in the implicature.
How it works¶
- Separate literal from implied. Write down what the message literally asserts, then, line by line, what a cooperative reader will take it to further mean.
- Read the silences and the emphasis. What is conspicuously not said, what is foregrounded, what is hedged — each generates its own implicature.
- Surface the implicit action. Identify any required next action that the wording only implies, and judge whether "implied" is strong enough for the stakes.
- Flag hazards and gaps. Mark implicatures that are unintended (leaks) or that carry an action too quietly, and rewrite to promote or defuse them.
Tuning parameters¶
- Implicature net — how wide to cast (only the obvious inferences versus subtle tonal ones). A wider net catches leaks but generates noise and can read paranoia into innocent phrasing.
- Cooperative baseline — which reader you audit against: a charitable, attentive reader or a skimming, adversarial one. Auditing against the harsher reader catches more but may over-correct friendly copy.
- Action-explicitness threshold — how strongly an implied action must be promoted to explicit before it counts as safe. Higher thresholds prevent missed asks but make copy blunter.
- Silence weighting — how much interpretive load to assign to omissions and emphasis versus stated words.
When it helps, and when it misleads¶
Its strength is with concise, high-traffic, or indirect messaging — notifications, marketing copy, terse instructions, polite hints — where the words are few and the inferences many, and where a leaked or missed implicature scales to every reader. It operationalizes Grice's insight that speakers convey a great deal by conversational implicature[1], meaning carried by the act of saying rather than by the words' literal content.
Its failure mode is that implicature depends on the reader's context, so an audit run against the author's assumed reader can miss the inferences a different audience will draw, and can also manufacture worries that no real reader would have. The classic misuse is weaponizing it — mining a message for deniable implicatures so a claim can be implied without being stated. The guarding discipline is to promote genuinely load-bearing implications to explicit statements (especially required actions) rather than trusting the hint, and to check the audit against a reader unlike the author when the audience is broad.
How it implements the components¶
inference_path_map— it maps the step from literal content to what the message conversationally implies, one implicature at a time.implicit_action_requirement— its signature output flags any required action the wording only implies, so it can be made explicit for the stakes.literal_content_record— it anchors on a written record of the literal claims, against which the implied meanings are read.
It does not model a specific recipient's background to predict their whole reading — that is Recipient Inference Walkthrough — and it does not tune the politeness_and_face_risk_profile behind a softener, which is Politeness/Directness Adjustment.
Related¶
- Instantiates: Pragmatic Context Alignment — supplies the audit of everything a message means beyond its literal words.
- Sibling mechanisms: Common Ground Checklist · Context Frame Header · Context Shift Test · Example and Boundary Case Pair · Politeness/Directness Adjustment · Pragmatic Breakdown Log · Readback for Situated Meaning · Recipient Inference Walkthrough · Scope and Audience Note
Editorial Notes¶
Form Classification¶
Form family: Assessment, Review & Assurance
Rationale: Implicature Audit operates as a bounded evaluation of existing evidence or work that produces a finding or disposition because it examines a message for what it implies beyond the literal words — what's suggested by what's said, unsaid, emphasized, or softened — and flags the risky implications
Independent corroboration: The frozen evidence defines Implicature Audit as 'Examines a message for what it implies beyond the literal words — what's suggested by what's said, unsaid, emphasized, or softened — and flags the risky implications', so its operative form is Assessment, Review & Assurance.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Linguistics & Semiotics
Origin pattern: Single lineage
Present-day reach: Universal
Rationale: The audit directly operationalizes Grice's theory of conversational implicature in linguistic pragmatics.
Related originating lineages:
- Philosophy — Philosophy of language materially developed the cooperative principle and speaker meaning.
- Rhetoric — Persuasive practice independently exploits what wording leads an audience to infer without literal assertion.
Review resolution: Both reviewers independently assign linguistics_semiotics as the primary originating domain, so that shared primary is retained. Alternate domains are the union of reviewer-identified formative or independently originating lineages; later application settings alone are excluded. The evidence describes one principal historical lineage. Its operational pattern is portable across essentially any subject domain. The encyclopedia entry makes that composition explicit.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] Conversational implicature — H. P. Grice's term for meaning a speaker conveys by the act of saying something in a given way, beyond the literal content, on the assumption that speaker and hearer are cooperating. "Some students passed" implicates "not all did," though it does not literally say so. The audit hunts exactly these carried-but-unstated meanings. ↩
References¶
[1] Grice, H. P. "Logic and Conversation". In P. Cole & J. L. Morgan (Eds.), Syntax and Semantics, Volume 3: Speech Acts, 41–58. Academic Press (1975). Explains conversational implicature as meaning conveyed through an utterance and conversational context beyond what its words conventionally say. registry ↩