Conceptual Blending

Prime #

84

Origin domain

Cognitive Science

Also from

Linguistics & Semiotics

Related primes

Metaphor, Analogy, Abstraction, Schema

Core Idea

Conceptual blending is the cognitive operation of constructing a new integrated mental space — the "blend" — from two or more input mental spaces, by selectively projecting elements from the inputs into the blend and allowing emergent structure and meaning to arise in the blend that is not present in any single input. The essential commitment is that the blend is not a union, intersection, or analogical copy of the inputs but a genuinely new cognitive construction with its own structure, running pattern, and inferences, causally connected to the inputs through a generic space that captures what they share structurally. Every blending claim specifies (1) two or more input spaces with distinguishable elements and relations, (2) a cross-space mapping identifying correspondences between input elements, (3) selective projection from inputs into a blended space, and (4) emergent structure in the blend — composition, completion, and elaboration that yields meaning beyond what any input provides.

How would you explain it like I'm…

Mixing two ideas

If you mash up a horse and a horn in your head, you get a unicorn. The unicorn is not just a horse or just a horn, it is a new thing with its own story. That is what your brain does when it blends ideas to make something new.

Idea mash-up

Conceptual blending is when your mind takes two different ideas and combines pieces from each into a brand new third idea. The new idea has things that neither original had on its own. Think of a 'desktop' on a computer: you take the idea of a real desk with folders and the idea of a screen with files, and you blend them. The result lets you 'drag' a 'file' into a 'trash can' even though none of that really exists physically.

Building a new idea from two

Conceptual blending is a cognitive operation where your mind builds a new 'mental space' by pulling selected pieces from two or more existing mental spaces and combining them. The blend isn't a sum or overlap; it has emergent structure, meaning, and inferences that none of the inputs had. The computer desktop is a classic case: it borrows from physical desks (folders, files) and from screens (display, clicking), and the blend supports brand-new actions like dropping a file into a trash icon to delete it. The pieces stay connected to their originals through a shared 'generic' structure, which is why the blend still makes sense.

 

Conceptual blending is the cognitive operation of constructing a new integrated mental space — the blend — from two or more input mental spaces, by selectively projecting elements from the inputs and allowing emergent structure to arise in the blend that is not present in any single input. The blend is not a union, intersection, or analogical copy: it is a genuinely new construction with its own organizing logic, running pattern, and inferences. The inputs remain causally connected to the blend through a generic space that captures their shared structural skeleton. A complete blending account specifies (1) two or more input spaces with distinguishable elements and relations, (2) a cross-space mapping that identifies correspondences, (3) selective projection into the blended space, and (4) emergent structure arising in the blend through composition, completion, and elaboration — the source of meaning beyond what any input alone provides. Fauconnier and Turner's *The Way We Think* (2002) is the canonical treatment.

Structural Signature

A cognitive process qualifies as conceptual blending when each of the following holds:

• The input mental spaces. At least two distinct mental spaces are active, each with its own elements, relations, and frame structure (e.g., "computer" as one space, "desktop" as another; or "surgery" as one space, "butchery" as another).

• The cross-space mapping. Correspondences between elements of the inputs are established — often partial and analogical — so that elements in one space are understood to align with elements in another. This mapping is the precondition for selective projection.

• The generic-space abstraction. An abstract schema common to the inputs is available (implicitly or explicitly), capturing what they structurally share. The generic space provides the highest-level constraint on what can map across inputs and what can be blended.

• Selective projection. Elements and relations are projected from each input into the blend selectively; not every aspect of any input transfers, and the choice of what to project shapes the blend's character and licensed inferences.

• The blended space. A new integrated mental space distinct from any input, containing emergent structure — composition of projected elements, pattern completion from background knowledge, and elaborated inferences — that is not in any single input and often not in the inputs combined.

• Running-the-blend operation. The blend can be elaborated, reasoned within, and used to generate novel inferences; conclusions drawn in the blend can then be selectively mapped back (the back-projection) to the inputs or to action in the world, validating or constraining which blend-inferences are licensed.

• The integration network. The whole system of input spaces, generic space, blended space, cross-space mappings, and back-projections that together form a single dynamic system for conceptual construction.

• The optimality principles. Principles governing what makes a blend apt: integration (the blend achieves unified structure), coherence (it respects causal and intentional relations), topology (input structure is preserved), and compression (the blend achieves economy).

What It Is Not

• Not metaphor. Metaphor is a one-way mapping from a source domain to a target domain (life IS a journey), with target-domain understanding structured by source-domain knowledge. Blending is bi- or multi-directional with a novel emergent space containing structure from both. Metaphors often ground blends but the blend does more — it creates new structure where metaphor typically maps existing structure. See metaphor.
• Not analogy. Analogy identifies structural similarities between two domains and licenses inference from source to target. Blending constructs a new integrated space in which both sources are re-expressed and combined; the analogical mapping is a precondition (the cross-space mapping) but the blend goes beyond mapping to emergent construction. See analogy.
• Not simple combination or union. A blend is not "input1 + input2" or their set-theoretic union; projection is selective, and emergent structure arises that is in neither input.
• Not creativity in general. Blending is a specific mechanism for novel-concept construction, not all creative cognition. Divergent thinking, incubation, and analogical leaps may or may not involve blending.
• Not a language-only phenomenon. Although identified through linguistic and conceptual examples, blending operates in visual imagination, design, music, mathematics, and social reasoning — wherever mental spaces are constructed and integrated.
• Common misclassification. Calling any novel association "blending"; using "blend" loosely for any cross-domain thinking; equating blending with metaphor because both involve cross-domain alignment.

Broad Use

• Cognitive linguistics and cognitive science
  • Fauconnier and Turner's conceptual integration theory; mental spaces; many-space networks (simplex, mirror, single-scope, double-scope); vital relations (time, space, identity, cause- effect); compression across vital relations.
• Semantics and pragmatics
  • Meaning construction in discourse; noun-noun compounds ("land yacht", "desk jockey"); counterfactuals; viewpoint-shifted expressions; frame-based meaning.
• Creativity, design, and innovation
  • Product concepts that combine unrelated categories (smartphone = phone × computer × camera); brand-concept construction; design fictions; creative problem-solving through forced combinations.
• Mathematics and scientific reasoning
  • The complex-number blend integrating real numbers and 2D geometry; the function-as- machine blend; kinetic-theory blends of particles and gases; thought experiments that blend idealized and realistic features.
• Art, literature, and film
  • Metaphorical paintings; literary conceits (Shakespeare's life-as-stage); surrealist juxtapositions; film montage creating emergent meaning; advertising images that blend product and aspiration.
• Moral and political reasoning
  • Ethical reasoning by blending familiar moral frames with novel situations; political slogans (war on poverty, war on drugs) that blend literal-war structure into non-war domains and import inferences from the source.
• Human-computer interaction and interface design
  • Desktop metaphor (folders, files, trash) blending physical office with computation; skeuomorphic interfaces; game mechanics that blend real-world referents with digital action.

Clarity

Conceptual blending clarifies by forcing articulation of the input spaces, cross-space mapping, and emergent structure behind a novel concept or expression. A claim like "the new product is a smartphone-wallet-key" resolves into "input 1 = smartphone (communication, computation, sensors); input 2 = wallet (payment, identification, money storage); input 3 = key (access control, physical actuation); generic space = personal-carried devices that grant capability; cross-space mapping aligns carrying, pocketability, and access-granting; selective projection omits smartphone weight and wallet thickness while projecting their capabilities; emergent structure in the blend includes unified device constraints (battery vs always-accessible), security models that inherit from all three, and failure modes (phone dies = can't pay or enter) not present in any single input." The clarifying force is to make visible the cognitive construction work that a fluent expression hides.

Manages Complexity

• Structures product and service innovation: forced-combination methods, SCAMPER, design-by- analogy, and cross-domain blending exercises are blending-based methods that produce novel concepts by systematically varying the inputs and the selective-projection scheme.
• Frames linguistic meaning construction: rather than treating words as having fixed meanings, blending treats many expressions (especially noun-noun compounds, neologisms, and figurative language) as prompting online construction of blended spaces. This explains productivity and context-sensitivity of meaning.
• Supports thought-experiment design: well- constructed thought experiments blend idealized and realistic features (Schrödinger's cat blends quantum superposition with macroscopic biology, Maxwell's demon blends statistical mechanics with intentional agency), and careful specification of the blend's projection scheme reveals what the experiment does and does not license.
• Explains creativity that goes beyond analogy: many creative leaps are not pure analogical mappings but constructive blends — new structure emerges in the integration that was in neither source, providing a cognitive mechanism for genuine novelty within known psychological processes.
• Frames moral and political discourse: metaphor and frame choice in political discourse work through blending — "investment in education" blends finance and pedagogy, importing assumptions (returns, portfolios, risk) that shape reasoning. Recognizing blends in political language reveals hidden inferential commitments.

Abstract Reasoning

Conceptual blending trains a reasoner to ask:

• What input spaces are active in this concept, expression, or construction?
• What cross-space mapping aligns elements across the inputs?
• What generic-space schema captures what the inputs share?
• Which elements are projected from each input into the blend, and which are omitted?
• What emergent structure appears in the blend that is in neither input?
• What inferences can be run in the blend, and which of those should be exported back to the inputs or to action?
• Is the blend appropriate to the reasoning task, or does it smuggle in inferences from an input that do not belong in the target?

Knowledge Transfer

Role mappings across domains:

• Input space ↔ source domain / constituent concept / frame / mental space
• Cross-space mapping ↔ analogical correspondence / alignment / equivalence class
• Generic space ↔ shared schema / common abstraction / type-level structure
• Selective projection ↔ feature choice / mapping rule / what transfers
• Blended space ↔ new integrated concept / composite / conceit
• Emergent structure ↔ novel inference / new failure mode / unexpected property
• Running the blend ↔ reasoning within the composite / elaboration / inference generation

A designer conceiving a smartwatch, a mathematician constructing complex numbers, a politician crafting a metaphorical slogan, and a novelist developing a literary conceit are all doing the same structural work: identify two or more input spaces, establish cross-space mappings, selectively project elements into a blend, elaborate the emergent structure, and run the blend for inference. The same diagnostic — "what inputs, what mapping, what projection, what emergent structure, what licensed inferences?" — applies across their contexts, with the same failure modes (importing unlicensed inferences, projecting too generously, treating emergent structure as pre-existing) in each.

Examples

Formal, Conceptually Canonical

The "Buddhist monk" thought experiment analyzed by Fauconnier and Turner exemplifies how blending yields inferential structure absent from either input alone. ^[1] A Buddhist monk ascends a mountain path starting at dawn on day 1; the next day he descends the same path starting at dawn. Is there a time of day at which he is at the same point on both days? Naive reasoning struggles to answer. Blending resolves it: input 1 = the ascent (monk, path, dawn-to-dusk progression); input 2 = the descent (monk, path, dawn-to-dusk progression); the generic-space abstraction = a single agent traversing the path between dawn and evening. The blended space has both monks on the path simultaneously — structure in neither input. Running-the-blend operation: the two monks must meet somewhere on the path, since one ascends and one descends over the same stretch. Selective projection omits the two-day separation and maps both journeys into one temporal overlay. Back-projection to the original puzzle: there must be a time of day at which the monk (on both days) is at the same location. The blend compresses two times into one and two people into corresponding agents; the emergent structure — the necessary meeting — corresponds to the time-of-day answer in the inputs. This example demonstrates that blending does substantive inferential work beyond what either input alone, or analogy between them, accomplishes. ^[1]

Mapped back: The tensional moment where the blend's inference (the monks must meet) re-enters the original problem (is there a time?) shows how the integration network licenses a conclusion that neither "ascent alone" nor "descent alone" provides. Analogy between ascent and descent notes their structural similarity; blending goes further by constructing a unified scenario in which both happen, making their intersection inferentially mandatory. The example is canonical because it isolates blending from metaphor (no "life IS a journey" target-source asymmetry), from mere analogy (more than structural mapping), and from simple composition (the meeting is not predictable from ascent + descent).

Applied, Industry Exemplar

Product naming and positioning for a hybrid transportation device illustrates blending in design-thinking. ^[2] Input 1 = the input mental spaces (bicycle: pedal, rider, human-powered, road-legal as bicycle); input 2 = electric motor (throttle-controlled, battery-powered, speed capability); input 3 = scooter (stand-up posture, small wheels, urban commute use). The generic-space abstraction = personal-scale road vehicle for on-demand mobility. The cross-space mapping aligns standing from scooter with portability from bicycle, aligns throttle from motor with ease-of-use from both. Selective projection yields the "e-scooter": standing posture from scooter, motor and battery from electric, wheels sized as a composite; pedal-assist from bicycle omitted in most variants; human-powered efficiency not projected. The blended space is neither bicycle nor motorcycle but a new regulatory category. The emergent structure includes novel failure modes (battery depletion mid-commute, range anxiety), new business models (rentals, last-mile logistics), and novel security constraints (theft, lock design) not present in any single input. ^[3] Running-the-blend operation: market positioning reveals whether the blend targets commuters, teenagers, or delivery services; each running imports different inferences. Back-projection shows why the e-scooter fails in contexts where the inputs remain separated (a hill-climber prefers bicycle; a long commute demands motorcycle range).

Mapped back: The structural kinship with the Buddhist monk example is precise — the input mental spaces, the cross-space mapping, selective projection, the emergent structure, running-the-blend operation — despite the shift from thought experiment to commercial design. Both blends exhibit the optimality principles: the blend compresses information (one device instead of three), achieves multiple input roles simultaneously, and maintains topology (standing posture aligns with portability both inputs provide). The design blend reveals why "e-bike" (bicycle with motor) or "motorized scooter" (scooter with more power) are weaker blends: they project more elements from one input, yielding less emergent novelty and less market distinctiveness.

Structural Tensions and Failure Modes

• T1: Selective Projection Is Consequential and Often Tacit.

  • Structural tension: Which elements project from each input into the blend determines the blend's character and licensed inferences. Projection choices are usually tacit — reasoners do not articulate what they kept and what they left out — yet different projection choices yield different blends with different implications. The tension is between the cognitive economy of tacit projection and the rigor of explicit projection analysis.
  • Common failure mode: Blending with unexamined projection that imports unwanted inferences (e.g., "war on poverty" projecting combat tactics onto social policy); different parties using the same blend with incompatible projections and talking past each other; design blends that project all of an input's features including irrelevant or harmful ones.

• T2: Emergent Inference Is Seductive but Not Always Licensed.

  • Structural tension: The blend generates inferences not present in either input, and those inferences feel valid because they "run" smoothly in the blend. But not all emergent inferences in a blend are licensed for the inputs or for action. The tension is between the generative power of running-the-blend and the need to check which inferences survive back- projection.
  • Common failure mode: Treating blend inferences as automatically true of the inputs (computers "think", corporations "want"); metaphorical reasoning that inherits source-domain inferences without warrant; thought experiments whose conclusions depend on emergent blend structure that does not correspond to reality.

• T3: Blend Stability vs Productivity.

  • Structural tension: Blends can conventionalize — becoming stable concepts with fixed meaning (desktop, mouse, cloud) — or remain productive, with speakers and reasoners regenerating them online. Stable blends are efficient but can become opaque, hiding their originating mappings; productive blends remain visible as constructions but demand cognitive work each time.
  • Common failure mode: Frozen blends carrying outdated projections (treating "computer virus" as contagious in ways biological viruses are but software exploits are not); productive blends over-elaborated into incoherent composites; educational communication that assumes audiences share the speaker's conventionalized blends.

• T4: Cross-Cultural and Cross-Expertise Variation in Blending Resources.

  • Structural tension: Blending depends on input-space content — concepts, frames, background knowledge — that varies across cultures, eras, and expertise levels. A blend that works for an expert or insider audience may fail for others who lack the input-space content, or may be constructed differently. The tension is between blending's generativity and its dependence on shared cognitive resources.
  • Common failure mode: Expert communication that assumes audience access to input spaces the audience lacks (a scientific metaphor opaque to laypeople); translation and cross-cultural communication where source-culture blends have no target-culture equivalent; instructional materials built on blends whose inputs students haven't yet acquired.

• T5: The Blend as Creative Leap vs. Recognition Error.

  • Structural tension: ^[4] Some blends generate genuinely novel insights — the Buddhist-monk blend produces a solution the inputs alone could not; complex numbers blend reals and geometry into a unified field with new mathematical power. Other blends miscategorize or produce false equivalences — treating "artificial intelligence" as truly intelligent, or "love is a journey" as capturing actual romantic experience. The tension is that the cognitive mechanisms are identical; distinguishing creative blending from erroneous blending requires post-hoc validation against the target domain. There is no internal mark in the blend itself that identifies it as creative vs. misleading.
  • Common failure mode: Celebrating a blend as innovative without checking whether it licenses false inferences (a design blend that projects incompatible constraints from both inputs); treating all metaphorical blends as equivalent despite vast differences in their inferential validity; educators or scientists using pedagogical blends that later require unlearning because they are structurally convenient but factually wrong.^[5]

• T6: Conceptual Blending as Universal Cognition vs. Overstated Theoretical Reach.

  • Structural tension: ^[1] The Fauconnier-Turner thesis claims that blending is foundational to "almost all" human cognition — metaphor, analogy, grammar, meaning-making, creativity, reasoning. Yet critics (Pinker 2007) argue that blending is a powerful construct but applied inconsistently; many phenomena attributed to blending are better explained by direct analogy, category membership, or logical inference without invoking a novel blend. The tension is methodological: blending is so flexible and post-hoc that nearly any conceptual phenomenon can be described as a blend. The theory's power — its ability to cover diverse cases — is also its weakness; falsifiability is difficult because blending can be retrofitted to almost any example.
  • Common failure mode: Attributing all novel concepts to blending without first checking whether direct analogy or metaphor suffices; interdisciplinary teams that adopt blending as an explanatory framework and over-apply it, labeling every cross-domain transfer a "blend"; educational or design contexts where blending theory is invoked to justify unrigorous conceptual leaps ("it's a blend, so novelty is guaranteed").^[6]

Structural–Framed Character

Conceptual Blending is a hybrid on the structural–framed spectrum. Part of it is a bare pattern that means the same thing in any field; part of it is a frame—a vocabulary and a set of assumptions—inherited from cognitive science. It leans toward the structural side, with only a light frame riding along.

On the structural side, the operation has a clean shape that applies wherever two or more sources are merged: distinct input spaces, selective projection of elements from each, and emergent structure in the resulting blend that was in none of the inputs—a pattern visible in metaphor, in product design, and in mathematical analogy alike. On the framed side, the prime keeps speaking the language of its home discipline—"mental spaces," cognitive operations, the running of meaning inside a mind—which presupposes a mental subject doing the integrating, an assumption that is not purely formal and cannot be fully shed when the term is borrowed. Its origin is theoretical rather than institutional, and it carries little evaluative weight, but the cognitive vocabulary does not entirely fall away in transfer. Balancing a transferable combinatorial core against an inherited mentalistic frame, it lands toward the structural side of the mid-spectrum.

Substrate Independence

Conceptual Blending is among the most substrate-tethered entries in the catalog — composite 1 / 5 on the substrate-independence scale. Its vocabulary — mental spaces, projections, the blend itself — is fundamentally cognitive-science machinery rather than a substrate-neutral structure, and that imported framing is hard to strip away. Claims that it transfers to organizational innovation or coordination turn out to be borrowed language: calling a merger a 'blend' is not the same structural reasoning, just a reused word. With no demonstrated structural instance in physical, biological, or social systems, it reads as a cognitive-science catalog entry that does not lift cleanly off its home medium.

• Composite substrate independence — 1 / 5
• Domain breadth — 2 / 5
• Structural abstraction — 3 / 5
• Transfer evidence — 1 / 5

Relationships to Other Abstractions

Current abstraction Conceptual Blending Prime

Parents (1) — more general patterns this builds on

• Conceptual Blending presupposes Analogy Prime

  Conceptual blending presupposes analogy because cross-space projection into a blended space relies on prior structural mapping between the input domains.

Hierarchy paths (2) — routes to 2 parentless roots

• Conceptual Blending → Analogy → Comparison → Self Checking

Show alternative path (1)

Neighborhood in Abstraction Space

Conceptual Blending sits in a moderately populated region (47^th percentile for distinctiveness): it has near-neighbors but no dense thicket of synonyms.

Family — Representation, Composition & Mental Schemas (11 primes)

Nearest neighbors

• Comparison — 0.73
• Schema — 0.73
• Analogy — 0.72
• Concept Drift — 0.71
• Juxtaposition — 0.70

Computed from structural-signature embeddings · 2026-07-26

Not to Be Confused With

Conceptual blending must be distinguished from Composition, its nearest neighbor (similarity 0.688). The two both involve combining elements into wholes, but they operate at fundamentally different levels and produce fundamentally different outputs. Composition is the spatial or structural arrangement of existing, well-defined elements into a coherent larger form — a painting composed of brushstrokes, a collage composed of found images, a building composed of structural members. The elements entering composition retain their identity and function; composition's work is organizing them into spatial relationships and hierarchies that produce aesthetic or functional coherence. Blending, by contrast, does not arrange pre-existing elements; it constructs a genuinely new mental space by selectively projecting elements from multiple input spaces and allowing emergent structure to arise in the blend that is in neither input alone. The e-scooter blend of bicycle, motor, and scooter produces safety-constraint interactions and business-model implications (rental economics, battery anxiety) that are not present in any input and cannot be predicted by arranging the inputs spatially. Composition produces coherence through arrangement; blending produces novelty through integration and emergence. A designer using composition selects and arranges; a designer using blending identifies inputs, establishes mappings, and elaborates what emerges in the integrated space.

Blending is distinct from Metaphor, which is a one-directional mapping from source domain to target domain that structures the target's understanding in terms of the source. "Life is a journey" maps journey-domain properties (destinations, obstacles, waypoints) onto life-domain understanding, with the mapping flowing primarily from journey to life, not bidirectionally. Blending, by contrast, is multi-directional: both inputs are re-expressed in the blend, structure emerges in the blend that neither input provided, and the blend is often as novel for understanding the inputs as it is for understanding the target. The Buddhist-monk blend treats both ascent and descent as equally transformed—mapped into a single time-space where they must intersect. Neither ascent nor descent alone provides the inferential power; the integration does. Metaphor can ground blends (the "life is a journey" metaphor may anchor a more complex blending of life-stages with journey-milestones), but blending goes further by constructing integrated scenarios that produce inferences metaphor alone does not license.

Blending is not Cognitive Reframing, which is the substitution of one interpretive lens or frame for another while the underlying objective facts remain the same. Reframing a "glass-half-empty" situation as "glass-half-full" changes the affective valence and subsequent motivation while the actual glass contents do not change; the objective condition is reframed, not transformed. Blending is not reframing; it is constructing a new integrated space with its own structure, running pattern, and inferential behavior. A situation reframed from "cost" to "investment" is still fundamentally the expense being described; an expense blended with financial-return properties creates novel constraints (return-on-investment thresholds, portfolio-diversification logic) that are genuinely new, not merely reinterpreted.

Finally, blending is not Boundary, which marks demarcations between inside and outside with rules of selectivity and exclusion. Boundaries define what belongs to a category or space and what does not; they work through differentiation and exclusion. Blending works through selective integration; elements are projected from multiple inputs precisely because the mapping and projection scheme determines what the blend achieves. A boundary says "this is in, that is out"; a blend says "these elements from this input and those elements from that input are projected, and their combination produces emergent structure." The operative logic is constructive and emergent, not demarcatory.

Solution Archetypes

Solution archetypes in the catalog that build on this prime — directly (this prime is a source ingredient) or as a related prime.

Built directly on this prime (1)

• Conceptual Blending for Innovation: Combine elements from distinct conceptual spaces to create a new solution space with emergent possibilities.
  ▸ Mechanisms (10)

Also a related prime in 3 archetypes

• Dialectical Synthesis: Use opposing positions to generate a stronger synthesis rather than merely choosing one side.
• Metaphor Mapping and Interpretive Control: Transfer selected structure from a familiar source domain to a target while controlling which inferences travel, which do not, and what the audience does with them.
• Multi-Dimensional Solution Space Exploration: Before narrowing, deliberately vary independent design dimensions—such as function, form, user context, cost, risk, sustainability, material, channel, governance, and time horizon—so convergence selects from a genuinely broad solution space rather than from the first visible family of options.

References

[1] Fauconnier, G., & Turner, M. The Way We Think: Conceptual Blending and the Mind's Hidden Complexities. New York: Basic Books, 2002. The canonical statement of conceptual integration theory: integration networks, the four-space model, optimality principles, and extended analysis of the Buddhist-monk riddle. SUPPORTS markers 180 & 181 (the Buddhist-monk example and blending's inferential work beyond input/analogy) and 186 (T6: the claim that blending is foundational to 'almost all' human cognition — this is exactly the F&T thesis being characterized). Link is the archive.org copy. ↩

[2] Turner, M. The Origin of Ideas: Blending, Creativity, and the Human Spark. New York: Oxford University Press, 2014. Applies blending to creativity, innovation, and the evolutionary emergence of abstract thought. PARTIALLY SUPPORTS marker 182 — the marker sits on a specific e-scooter product-naming/design example; Turner grounds blending-in-design/creativity generally but does not analyze that example. Existence verified; link is the OUP product page. See flag. ↩

[3] Coulson, S., & Oakley, T. "Blending Basics". Cognitive Linguistics 11, no. 3-4 (2000): 175-196. Technical introduction to blending: reference, partitioning, mapping, structure projection, and how blends generate novel inferences while maintaining topology. SUPPORTS marker 193 (its own bibliography self-citation). PARTIALLY SUPPORTS marker 183 — the marker sits on the e-scooter's 'novel failure modes, new business models, novel security constraints'; Coulson & Oakley support the general 'blends produce emergent inferences' principle but do not treat that example. DOI verified (journal issue dated 2001). See flag on 183. ↩

[4] Hofstadter, D. R. Gödel, Escher, Bach: An Eternal Golden Braid. New York: Basic Books, 1979. Explores self-reference, recursion, and cross-domain analogy/blending across mathematics, art, and music. SUPPORTS markers 184 (T5: some blends generate genuinely novel insight) and 190 (its own bibliography self-citation) at the level of blending-as-insight; note GEB predates formal blending theory and is an illustrative rather than technical source. Link is the archive.org copy. ↩

[5] Coulson, S. Semantic Leaps: Frame-Shifting and Conceptual Blending in Meaning Construction. Cambridge: Cambridge University Press, 2001. Experimental/psycholinguistic account of frame-shifting and blending as online meaning construction; distinguishes blending from metaphor with linguistic evidence. PARTIALLY SUPPORTS marker 185 — the marker sits on 'educators or scientists using pedagogical blends that later require unlearning because they are structurally convenient but factually wrong'; the book grounds blending-as-online-construction and false-inference risk but does not make the specific pedagogical-unlearning claim. DOI verified. See flag. ↩

[6] Pinker, S. The Stuff of Thought: Language as a Window into Human Nature. New York: Viking, 2007. Argues for conceptual semantics and critiques conceptual-blending/metaphor theory as overstated, holding that direct logical inference and category membership explain much that is attributed to blending. SUPPORTS markers 187 and 191 (T6 critique that blending is applied inconsistently / over-extended — Pinker is the named critic). Link is the archive.org copy. NOTE: marker text says 'Pinker 2007' for the critique; correctly sourced. ↩

[7] Fauconnier, G. Mental Spaces: Aspects of Meaning Construction in Natural Language. Cambridge, MA: MIT Press (Bradford Books), 1985. Founds mental-spaces theory: meaning arises through constructions and mappings across mental spaces. Tier C (bibliography only). Link is the archive.org copy.

[8] Fauconnier, G. Mappings in Thought and Language. Cambridge: Cambridge University Press, 1997. Develops cross-space mapping and blending as a distinct operation; treats compression, topology, and back-projection constraints. Tier C (bibliography only). DOI verified.

[9] Grady, J. E., Oakley, T., & Coulson, S. "Blending and Metaphor". In Metaphor in Cognitive Linguistics, ed. R. W. Gibbs & G. J. Steen, 101-124. Amsterdam: John Benjamins, 1999. Clarifies that metaphor often grounds blends but blending yields emergent structure metaphor alone does not. Tier C (bibliography only). DOI verified. NOTE: the prime's entry MIS-CITES the venue — it gives editors 'G. Steen & J. Gavins' and book 'Cognitive Poetics: Goals, Definitions, and Directions, Edinburgh University Press'; the actual publication is in Gibbs & Steen (eds.), Metaphor in Cognitive Linguistics (Benjamins, 1999), pp. 101-124. See flag.

[10] Sweetser, E. "Blended Spaces and Performativity". Cognitive Linguistics 11, no. 3-4 (2000): 305-333. Analyzes performativity as blending in which structure transfers from a representing space to the space represented. Tier C (bibliography only). DOI verified. NOTE: the prime's entry is GARBLED — it gives title 'Properties, Parts, and Prototypes...', editors 'D. Blank & P. Koch', and book 'Espace et Temps en Langage', none of which match Sweetser's real 2000 paper. Corrected here. See flag.

[11] Lakoff, G., & Johnson, M. Metaphors We Live By. Chicago: University of Chicago Press, 1980. Founds conceptual-metaphor theory; metaphor as a mechanism of mind grounded in embodied experience. Tier C (bibliography only; tagged CROSS-DP-22 in source). Link is the archive.org copy.

[12] Lakoff, G. "The Contemporary Theory of Metaphor". In Metaphor and Thought, 2^nd ed., ed. A. Ortony, 202-251. Cambridge: Cambridge University Press, 1993. Systematizes conceptual-metaphor theory and its relation to cross-domain mapping. SUPPORTS marker 188 (metaphor constrains but does not fully determine blended-space construction) at a general level. DOI (Cambridge chapter) verified. NOTE: marker FACT-D15-188 is malformed in the source — it sits at the END of the lakoff-1993 reference-definition line inside the References section (line 569), not in body prose. See flag.

[13] Gentner, D., Bowdle, B., Wolff, P., & Boronat, C. "Metaphor Is Like Analogy". In The Analogical Mind: Perspectives from Cognitive Science, ed. D. Gentner, K. J. Holyoak, & B. N. Kokinov, 199-253. Cambridge, MA: MIT Press, 2001. Empirically compares metaphor, analogy, and blending; argues metaphor works via alignment-and-projection like analogy, with domain asymmetry. SUPPORTS marker 189 (its own bibliography self-citation). DOI (MIT Press chapter) verified.

[14] Mithen, S. The Prehistory of the Mind: The Cognitive Origins of Art, Religion and Science. London: Thames and Hudson, 1996. Uses 'cognitive fluidity' (leakage/integration across modular cognitive domains) to explain the emergence of symbolic thought, art, and science. SUPPORTS marker 192 (its own bibliography self-citation; cross-domain integration as a distinctive human capability). Link is the archive.org copy.

[15] Boroditsky, L. "Metaphoric Structuring: Understanding Time Through Spatial Metaphors". Cognition 75, no. 1 (2000): 1-28. Empirically shows abstract time is structured by mappings from concrete space, with cross-linguistic variation. SUPPORTS marker 194 (its own bibliography self-citation; space-time blends shape reasoning across languages). DOI verified.

[16] (No separate key — see lakoff-1993 above. The reference list defines lakoff-1993 once; marker 188 maps to it.)