Simile Scaffold Prompt¶
Generative prompt — instantiates Literal-vs-Figurative Boundary Preservation
A guided prompt that helps a writer generate and choose among candidate source domains, tuning the comparison to the audience before any sentence is fixed.
The Simile Scaffold Prompt is a guided sequence of questions that helps a writer generate and choose a comparison before committing to any wording: What exactly needs illuminating? Which familiar sources share that feature? For this particular audience, which source's baggage is least dangerous? It works entirely upstream of the sentence. Its product is a decision — a chosen source plus a named transferable feature, tuned to the reader — not a finished clause. Where a clause template formats a comparison already decided, the scaffold is the mechanism that decides it, and its distinctive concern is the one the template never touches: selecting the right source from a set of live alternatives for a specific audience.
Example¶
A management consultant must explain a company's competitive position in a board memo after two rivals merged. She runs the scaffold instead of grabbing the first image that comes to mind. Target: our position in a suddenly more concentrated market. Candidate sources: a battlefield (vivid, but imports zero-sum framing and violence that could push a nervous board toward defensive overreaction); an ecosystem (captures interdependence, but implies no agency, as if the company just drifts); a crowded marketplace (captures competition and buyer choice, with light baggage). Audience: risk-averse directors who will act on whichever framing lands. The scaffold surfaces that battlefield would prime a war footing, while marketplace transfers the crowding-and-differentiation feature with the least misleading entailment.[1] Output: source = crowded marketplace, feature = differentiation under crowding. The memo's actual sentence gets written afterward — by a template — but the hard choice is already made and defensible.
How it works¶
The scaffold imposes a fixed order: clarify the target; brainstorm three or more candidate sources; score each on the strength of its transferable feature against its most dangerous entailment for the named audience; select the winner (or a shortlist). It is deliberately generative and selective, and it stops at the decision — it never writes the marker or the scope clause. Because it is just a structured prompt, the same scaffold drives a human writer or an LLM, and its reasoning can be saved as the rationale a mapping note would formalize.
Tuning parameters¶
- Candidate count — how many rival sources to generate before choosing. More candidates reduce the first-thought trap but slow the writer down.
- Audience specificity — a vague "general reader" versus a sharply modeled one. Sharper models catch audience-specific misreadings but risk over-fitting to an imagined reader.
- Entailment-scrutiny depth — how hard each candidate's baggage is interrogated. Deeper scrutiny avoids sneaky over-imports but can stall on diminishing returns.
- Stop point — commit to one winner versus hand a shortlist downstream. A shortlist defers the call to a reviewer with more context.
When it helps, and when it misleads¶
Its strength is preventing the first-source-that-comes-to-mind trap and making audience-fit explicit — the same simile misleads different readers differently, and the scaffold is where that gets caught. It is the highest-leverage step for aptness, the one thing form-level tools cannot supply. Its failure mode is over-generation and decision drag, and, more subtly, that an audience model can simply be wrong — a beautifully-scaffolded choice for a misimagined reader still misleads. The classic misuse is running the scaffold to rationalize a source the writer had already fallen in love with, generating throwaway alternatives to lose. The guarding discipline is to keep the candidate set honest, then hand the winner to a clause template for wording and to a real-audience check for verification.
How it implements the components¶
shared_feature_selection— its core act is naming which feature is meant to transfer, chosen against each candidate source's baggage.audience_interpretive_context— every candidate is scored for a specific audience, since the same source is safe for one reader and dangerous for another.comparison_alternative_set— it explicitly generates and weighs rival sources rather than defaulting to the first, and records why the losers lost.
It does not implement explicit_figurativeness_marker or transfer_scope_constraint — the Like/As Clause Template fixes those in the final wording; the scaffold stops the moment the comparison is chosen and never writes the sentence.
Related¶
- Instantiates: Literal-vs-Figurative Boundary Preservation — the scaffold is the generative front end that selects a safe, apt comparison.
- Sibling mechanisms: Like/As Clause Template · Source–Target Mapping Note · Non-Literality Disclaimer · Figurative Language Review Checklist · Comprehension Backcheck · Contractual Figurative-Language Scrub · Translation Marker Preservation Instruction
Editorial Notes¶
Form Classification¶
Form family: Interface, Display & Cue
Rationale: Simile Scaffold Prompt operates as a user-facing prompt, display, template, or perceptual cue that shapes attention and action at the point of use because it a guided prompt that helps a writer generate and choose among candidate source domains, tuning the comparison to the audience before any sentence is fixed.
Independent corroboration: The frozen evidence defines Simile Scaffold Prompt as 'A guided prompt that helps a writer generate and choose among candidate source domains, tuning the comparison to the audience before any sentence is fixed', so its operative form is Interface, Display & Cue.
Nearest alternative: Analysis, Modeling & Optimization — Simile Scaffold Prompt includes features of an analytical, modeling, inference, comparison, or optimization procedure that derives insight or a solution, but its defining operation is a user-facing prompt, display, template, or perceptual cue that shapes attention and action at the point of use.
Review outcome: Independent reviewer agreement; medium confidence.
Origin Attribution¶
Primary origin: Rhetoric
Origin pattern: Convergent development
Present-day reach: Universal
Rationale: Selecting a source domain and tuning a comparison to an audience is the rhetorical craft of simile. Aristotle treats simile as a rhetorical figure, and Silva Rhetoricae defines its comparative form; literary theory enriches source-domain selection.
Related originating lineages:
- Communication & Media Studies — Communication and media research supplies a parallel or contributing lineage for the mechanism's defining operation: a guided prompt that helps a writer generate and choose among candidate source domains, tuning the comparison to the audience before any sentence is fixed.
- Education & Pedagogy — A scaffold externalizes steps novices can practice.
- Linguistics & Semiotics — Conceptual mapping connects source and target domains through selected features.
- Literature & Literary Theory — literature_literary_theory contributes literary comparison, source domains, narrative sequence, and textual interpretation to this mechanism's defining operation—A guided prompt that helps a writer generate and choose among candidate source domains, tuning the comparison to the audience before any sentence is fixed—without displacing the selected primary historical lineage.
Review resolution: The blind reviewers disagree on primary lineage (literature_literary_theory versus rhetoric). Authoritative or primary research supports rhetoric as the best historical origin: Selecting a source domain and tuning a comparison to an audience is the rhetorical craft of simile. Aristotle treats simile as a rhetorical figure, and Silva Rhetoricae defines its comparative form; literary theory enriches source-domain selection. The cited Aristotle, Rhetoric Book III; Silva Rhetoricae, Simile directly supports the mechanism's defining operation. All independently supported contributing domains are retained without an arbitrary cap. origin_mode=convergent records lineage, while domain_reach=universal records later applicability separately from provenance.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Researched adjudication after independent review; high confidence.
Sources consulted:
References¶
[1] Lakoff and Johnson's Metaphors We Live By (1980) argues that the source we reach for silently structures the reasoning that follows — argument as war makes us defend and attack; a different source would make us cooperate or trade. The scaffold's whole reason to weigh alternatives is that the choice of source is never neutral; it shapes what the audience will infer next. registry ↩