Metaphorical Blend Prompt¶
Prompt — instantiates Conceptual Blending for Innovation
Uses a single chosen metaphor as one input space, mapping its structure into the problem while marking exactly where the metaphor stops holding.
The Metaphorical Blend Prompt seeds a blend with one deliberately chosen metaphor — "treat X as if it were Y" — and then does the disciplined work of walking the metaphor's entailments into the problem. Its defining feature, and what separates it from a random provocation, is that the second input is a coherent, chosen source with rich internal structure (an immune system, a garden, a marketplace), selected because its logic promises useful transfer. The prompt maps that structure across, selects which entailments actually project, and — crucially — writes an explicit limit note marking where the metaphor breaks, because a good metaphor is dangerous: it keeps generating implications, and the untested ones drift into false design claims. This mechanism lives or dies on the boundary between "entailment we tested and adopted" and "entailment the metaphor merely suggests."
Example¶
A city's traffic operations team is stuck treating congestion as a plumbing problem — add lanes, widen pipes — and it isn't working. A facilitator poses a Metaphorical Blend Prompt: treat the road network as if it were an immune system. The chosen source is coherent and structured, and its entailments are walked one by one. An immune system detects local anomalies early (a fever, an infection) → map to sensors that flag an incipient jam before it spreads. It mounts a proportionate, local response rather than a citywide one → map to targeted signal changes and ramp metering around the emerging jam. It remembers past threats → map to learning which intersections recur as trouble spots.
Selective projection keeps those three entailments and drops others (the immune system destroys invaders; the city does not want to "attack" drivers). Then the limit note is written explicitly: the immune metaphor implies a self/non-self distinction and an inflammatory overreaction risk that have no safe analog in traffic — an "autoimmune" over-response that throttles healthy flow is exactly the failure to avoid, and the metaphor must not be extended to justify aggressive, blanket restriction. The blend — a decentralized, early-detection, proportionate-response traffic control scheme — carries only the tested entailments, with the boundary written down so no one later argues "the immune system would attack, so should we."
How it works¶
The prompt runs as a bounded projection. First, choose the metaphor for structural promise, not surface charm — a source whose internal logic plausibly holds transferable structure. Second, enumerate entailments: list what is true of the source by its own logic (detects, responds locally, remembers, overreacts), turning a vague image into a checklist of candidate transfers. Third, map and selectively project: for each entailment decide whether it transfers, translates, or is left behind, so the blend imports structure rather than vibes. Fourth, write the limit note: state the entailments the metaphor generates that must not be applied, naming the boundary where the source's logic diverges from the target's reality. The signature discipline is that every adopted entailment is a decision, and every rejected one is recorded — the prompt is as much about what the metaphor may not imply as what it may.
Tuning parameters¶
- Metaphor richness — a thin comparison versus a source with deep internal structure. Richer sources yield more entailments to mine but more boundary to police.
- Entailment coverage — mining a few obvious implications versus exhaustively listing them. Fuller enumeration surfaces both better transfers and more drift risks.
- Projection strictness — how demanding the "does this actually transfer?" test is per entailment. Strict projection prevents overreach but can prune promising transfers early.
- Limit-note explicitness — a casual caveat versus a written boundary list. Explicit boundaries are what stop metaphor drift downstream; they cost authoring effort.
- Metaphor count — one sustained metaphor versus swapping several. A single deep metaphor keeps the blend coherent; swapping generates breadth but no bounded structure.
When it helps, and when it misleads¶
Its strength is structured transfer with a fence around it: a chosen metaphor supplies a whole coherent logic to draw on, and the limit note lets a team exploit it without being captured by it. It works in the tradition of conceptual metaphor, where a source domain's structure systematically shapes reasoning about a target through its entailments.[n1]
Its failure mode is metaphor drift — the source keeps generating implications, and an untested one ("the immune system attacks invaders, so we should aggressively block drivers") gets treated as a design conclusion rather than a suggestive analogy. A related misuse is borrowed authority, where the metaphor's vividness lends false confidence to a claim the target never earned. The guarding discipline is the limit note itself: no entailment becomes a design claim until it is deliberately projected and tested, and the boundary of the metaphor is written down so later reasoning can't quietly cross it.
How it implements the components¶
This prompt realizes the bounded-metaphor face of the archetype:
input_space_a— the chosen metaphor is the structured source space whose internal logic seeds the blend.cross_space_mapping— it maps the metaphor's entailments onto elements of the problem (detection → sensors, local response → targeted control).selective_projection— each entailment is deliberately transferred, translated, or dropped rather than imported wholesale.blend_limit_note— it explicitly records the entailments that must not be applied, fencing the metaphor against drift.
It does not inject an arbitrary input_space_b, mine a tension_or_mismatch_map, or lean on a novelty_filter and blunt usefulness_test — that random-stimulus, discard-heavy approach belongs to its prompting twin Forced Connection Exercise; this prompt uses one chosen, coherent metaphor and bounds it rather than filtering many disposable sparks.
Related¶
- Instantiates: Conceptual Blending for Innovation — supplies the single-metaphor, entailment-bounded route into a blend.
- Sibling mechanisms: Forced Connection Exercise · Concept Blend Canvas · Blend Coherence Review · Hybrid Prototype · Design Mashup Workshop · Cross-Domain Innovation Sprint · Business Model Pattern Mixing · Interdisciplinary Model Synthesis
Editorial Notes¶
Form Classification¶
Form family: Interface, Display & Cue
Rationale: Metaphorical Blend Prompt operates as a user-facing prompt, display, template, or perceptual cue that shapes attention and action at the point of use because it uses a single chosen metaphor as one input space, mapping its structure into the problem while marking exactly where the metaphor stops holding.
Independent corroboration: The frozen evidence defines Metaphorical Blend Prompt as 'Uses a single chosen metaphor as one input space, mapping its structure into the problem while marking exactly where the metaphor stops holding', so its operative form is Interface, Display & Cue.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Cognitive Science
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Universal
Rationale: Conceptual blending and input-space mapping were formalized in cognitive science.
Related originating lineages:
- Linguistics & Semiotics — Conceptual metaphor theory and semantic mapping materially shaped the technique.
Review resolution: Both independent reviews place the primary provenance in cognitive_science. The queued differences (domain_reach_disagreement) concern secondary metadata, not primary lineage. The final retains linguistics_semiotics only where a reviewer supplied a formative-lineage rationale; downstream use or broad applicability by itself is not treated as origin. origin_mode=cross_disciplinary_synthesis because the supplied rationales identify formative contributions that are composed in the mechanism's present form. domain_reach=universal records established application breadth separately from provenance. confidence=high preserves the more cautious evidence assessment. encyclopedia_synthesis=false records whether either reviewer identified deliberate corpus-level composition.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] In Lakoff and Johnson's account of conceptual metaphor, a source domain systematically structures reasoning about a target through a set of entailments — which is exactly why a metaphor must be bounded: the same mapping that supplies useful structure also generates implications the target never licensed. ↩