/flux

Analogical Thinking & Structural Mapping Reference

Purpose: Disciplined cross-domain analogy generation grounded in Dedre Gentner's structural-mapping theory (1983), which distinguishes surface similarity (objects look alike) from structural similarity (relational systems align). Far analogies — those crossing distant domains — yield the most innovation but are the hardest to find. Biomimicry, cross-industry pattern hunting, and engineering analogies all fall under this lens. Analogy fires best when the problem has clear relational structure (causes, constraints, flows) that can be mapped to a source domain.

Scope Boundary

• flux analogy: structural mapping from a source domain into the target problem. Produces aligned-relations table with novelty, transferability, and breakdown points.
• flux reframe (default): full DEEP pipeline — analogy is one optional stage inside COMBINE.
• flux cross: Bisociation-style cross-domain combine. Cross treats two domains as raw material to fuse; analogy aligns relations and tests transfer.
• flux scamper: intra-artifact modification. Lens A (Adapt) is shallow analogy; analogy is the deep version.
• flux challenge: assumption reversal. Use first if the source domain itself encodes hidden assumptions.
• flux inversion: invert-the-goal. Inversion asks "what guarantees failure"; analogy asks "what shape solved it elsewhere".
• flux (elsewhere): iterative dialogue. Analogy is a single mapping pass; Flux can develop one analogy across turns.
• field (elsewhere): empirical validation. Analogies predict; research confirms transferability.

Workflow

ENTER    →  state target problem as a relational structure
         →  list entities, relations, constraints, flows, goals

DISTANCE →  decide near vs far analogy budget
         →  near = same industry; mid = adjacent; far = distant domain / nature

HUNT     →  search candidate sources (catalog + free-form)
         →  generate ≥5 candidates across the chosen distance

MAP      →  align relations source → target (not objects → objects)
         →  mark which relations transfer cleanly, which break

TEST     →  identify breakdown points — where the analogy lies
         →  rate transferability (high / partial / metaphor-only)

DELIVER  →  hand to Spark (feature ideas), Atlas (architecture
            patterns), Magi (cross-option decision)

Near vs Far Analogy

Innovation research (Gentner; cross-industry studies of financial-services lateral thinking) shows far analogies yield the most viable novelty when relational alignment is rigorous — and the most empty rhetoric when it is not.

Biomimicry Catalog (selected starting points)

Cross-Industry Pattern Hunting

Question Bank for Structural Mapping

• What is the relation between A and B in the source? Does that relation exist in target?
• Which entities in source map to which in target? Multiple candidates? Pick by relational role.
• What constraint in source is absent in target? That absence often invalidates transfer.
• What constraint in target is absent in source? That absence often gives the analogy power.
• Where does the analogy break? Naming the breakdown is half the insight.
• Is the surface similarity (looks alike) or structural (works alike)? Surface alone is decoration.

Anti-Patterns

• Surface-similarity hunting — "this looks like X" without aligning relations. Produces metaphors, not insight.
• One-analogy commitment — locking onto first analogy and forcing fit. Generate ≥5, kill 4.
• Far-analogy theater — citing biomimicry without testing transfer. "Like a beehive" is rhetoric until relations are mapped.
• Ignoring breakdown points — every analogy breaks somewhere. Where it breaks is often the most useful signal.
• Confirming the preferred reframing — searching only for analogies that support the existing hypothesis. Deliberately seek disconfirming analogies.
• Mapping objects instead of relations — Gentner's core insight: align the relational system, not the entities.
• Treating analogy output as evidence — analogy generates hypotheses. Validate empirically before betting.
• Skipping near analogies for the glamour of far ones — near analogies often transfer cleanest. Far is high-variance, not always high-EV.

Handoff

• To Spark: high-transferability mappings as feature concept candidates.
• To Atlas: structural patterns that map to architecture decisions (hub-and-spoke, succession, mycorrhiza).
• To Magi: when multiple analogies suggest competing actions — decision needed.
• To Flux: when one analogy looks promising but the mapping is incomplete — iterate.
• To Omen: breakdown points as failure-mode candidates for pre-mortem.
• To inversion: when the analogy suggests a path, run inversion on that path to surface failure modes.
• To Field: high-novelty far analogies needing empirical validation before commitment.