Conceptual Blending is a cognitive process in which different concepts and knowledge structures are partially aligned and selectively integrated to create new meanings and relationships that were not present in the original concepts. Also known as Conceptual Integration, it is involved in language comprehension, metaphor, imagination, creative thinking, scientific discovery, technological development, and many other intellectual activities.
For example, when organizing files on a computer screen, we use concepts such as “desktop,” “folder,” and “trash bin.”
These concepts connect our experience of organizing physical desks and documents with computer-based information processing.
However, a folder displayed on a computer screen is not an actual paper folder. Moving a file to the trash bin does not involve physically moving an object.
Nevertheless, we can intuitively understand how to manipulate digital information by drawing on our knowledge of organizing physical documents.
Furthermore, operations such as dragging a file into a folder or retrieving a deleted file from the trash bin form a system of interactions that cannot be explained entirely by either physical desk organization or computer data processing alone.
This is one example that can be understood through Conceptual Blending.
The essence of Conceptual Blending is not simply combining existing concepts, but selectively integrating elements and relationships from different conceptual worlds to create new meanings and possibilities for reasoning.
However, Conceptual Blending is primarily a cognitive theory explaining how humans construct meanings and form new concepts. It does not imply that combining different things will necessarily produce creative or useful results.
Origins of Conceptual Blending Theory
Conceptual Blending Theory was developed by Gilles Fauconnier and Mark Turner.
One of its foundations is Mental Spaces Theory, developed by Fauconnier beginning in the 1980s.
Mental Spaces Theory proposes that people construct temporary conceptual structures when understanding language, considering hypothetical situations, or recalling past events.
For example, when thinking, “If I had joined another company,” we can imagine a situation different from our actual circumstances and reason about what might have happened in that alternative world.
The conceptual structures used to represent such imagined or described situations are called Mental Spaces.
Fauconnier and Turner investigated how correspondences are established between multiple mental spaces and how new meanings emerge from their integration.
They published “Conceptual Integration and Formal Expression” in 1995 and “Conceptual Integration Networks” in 1998, establishing important elements of the theory.
Their 1998 paper presented Conceptual Integration as a fundamental cognitive operation related to analogy, categorization, and the construction of mental models.
In 2002, they published The Way We Think: Conceptual Blending and the Mind’s Hidden Complexities, exploring the role of Conceptual Blending in language, culture, mathematics, science, and imagination.
Fauconnier also published “Creativity in Everyday Language” in the Japanese journal Cognitive Studies in 1996, discussing Conceptual Blending as a creative cognitive operation involved in ordinary language use.
An important aspect of this theory is that creativity is not treated exclusively as an exceptional ability possessed by a few individuals.
Instead, Conceptual Blending is understood as a cognitive process that may also operate when people interpret everyday language, understand metaphors, or imagine hypothetical situations.
What Are Mental Spaces?
Understanding Conceptual Blending requires an understanding of Mental Spaces.
A Mental Space is a local, temporary conceptual structure constructed when we think about a particular situation.
Consider the following statement:
“Ten years ago, he was a student. If he had pursued a research career, he might now be working at a university.”
To understand this statement, we distinguish at least three situations: his present circumstances, his circumstances ten years ago, and a hypothetical situation in which he pursued a research career.
These situations are related, but not all of them describe events that actually occurred.
Mental Spaces do not refer to physical storage areas in the brain or independently existing locations within the mind.
They are conceptual representations of situations constructed through thought and language.
In Conceptual Blending, elements and relationships belonging to different Mental Spaces are mapped onto one another and selectively projected into a new space.
As a result, meanings and relationships that were not directly present in the original spaces may emerge.
The Four-Space Model of Conceptual Blending
A representative explanation of Conceptual Blending uses four Mental Spaces.
| Space | Role |
|---|---|
| Input Space 1 | Represents elements and relationships associated with one concept or situation |
| Input Space 2 | Represents elements and relationships associated with another concept or situation |
| Generic Space | Represents an abstract structure shared by the input spaces |
| Blended Space | Integrates selected elements from the inputs and forms new meanings and structures |
Consider the relationship between a physical desk and computer file management.
Input Space 1 includes a desk, paper documents, folders, a trash bin, and the physical movement of documents.
Input Space 2 includes digital data, file systems, storage locations, deletion operations, and graphical displays.
The Generic Space captures shared relationships involving organizing objects, assigning locations, moving items, classifying them, and retrieving them when necessary.
In the Blended Space, icons displayed on a screen are treated as files, which can be dragged into folders as part of a new system of interaction.
| Element | Physical Desk | Computer System | Blended Interaction |
|---|---|---|---|
| Objects being managed | Paper documents | Digital data | Files displayed on the screen |
| Organizational locations | Desk, filing cabinet | Storage areas, directories | Desktop, folders |
| Movement | Physically carrying documents | Data management operations | Drag and drop |
| Disposal | Throwing documents into a trash bin | Deletion processing | Moving files to the trash bin |
| Recovery | Retrieving discarded documents | Restoring stored data | Recovering files from the trash bin |
This is a schematic analysis illustrating Conceptual Blending. It does not claim that the actual historical design of desktop interfaces can be explained solely through this cognitive theory.
The important point is that the Blended Space is not simply the combination of everything contained in the two original spaces.
Digital files are not physical paper, and dragging a file is not the same as physically moving a document.
Nevertheless, selectively integrating elements from these spaces produces a system of interactions that users can understand intuitively.
The four-space structure is a basic model used to explain the theory. Actual conceptual integration may involve more than two input spaces or interactions among multiple blends.
Cross-Space Mapping and Selective Projection
Conceptual Blending begins by establishing correspondences between different Input Spaces.
This is called Cross-Space Mapping.
For example, when comparing a physical desk with a computer system, paper documents can be mapped to digital files, storage locations to directories, and document disposal to data deletion.
However, not every characteristic of the original Input Spaces is carried into the Blended Space.
For example, the weight and thickness of physical paper are not directly transferred into digital file operations.
By contrast, relationships involving classification, movement, and retrieval are retained and used in the new interaction system.
The process of selecting certain elements and relationships from the inputs is called Selective Projection.
Conceptual Blending therefore depends not only on which elements are mapped to one another but also on which elements are included in the blend and which are excluded.
Rather than integrating everything, Conceptual Blending selectively incorporates elements and relationships needed to create meaningful new structures.
How Emergent Structure Arises
One of the most important concepts in Conceptual Blending Theory is Emergent Structure.
Emergent Structure refers to new relationships or structures formed within the Blended Space that were not present in precisely the same form in the original Input Spaces.
Fauconnier and Turner describe three processes involved in its formation.
Composition
Selected elements and relationships from different Input Spaces are combined, allowing new relationships to arise.
Completion
The combined elements are supplemented through existing knowledge and familiar conceptual frames, making the resulting situation meaningful and understandable.
Elaboration
The resulting blend is mentally developed to explore what might happen in the new situation and what consequences or possibilities it implies. This is sometimes described as Running the Blend.
Consider the concept of the computer desktop.
Integrating physical document organization with digital information management allows files to be treated as manipulable objects on a screen.
Existing knowledge about organizing documents helps users understand operations such as placing files in folders or discarding unnecessary items.
The resulting system of interactions can then be elaborated to include moving multiple files, creating nested folder structures, and recovering files from the trash bin.
However, Composition, Completion, and Elaboration should not necessarily be understood as three strictly separated chronological stages.
In actual thinking, these processes may interact continuously.
A Classic Example: The Buddhist Monk Problem
One of the best-known examples used to explain Conceptual Blending is the Buddhist Monk Problem.
A monk begins climbing a mountain at sunrise on a particular day and reaches the summit by sunset.
Several days later, the monk begins descending from the summit at sunrise and reaches the foot of the mountain by sunset, following the same path.
The monk may walk at different speeds and take different breaks during the ascent and descent.
Now consider the following question:
“Is there a location that the monk passes at exactly the same time of day during both journeys?”
When the two journeys are considered separately, the answer may not be immediately obvious.
Conceptual Blending allows us to imagine the two journeys taking place on the same day.
Imagine one monk walking uphill from the foot of the mountain while another monk walks downhill from the summit at the same time.
Because they travel along the same path in opposite directions, they must meet somewhere.
That meeting point corresponds to a location that the original monk passes at the same time of day during both journeys.
The problem can be represented using the four-space model.
| Space | Content |
|---|---|
| Input Space 1 | A monk climbs the mountain on one day |
| Input Space 2 | The same monk descends the mountain on another day |
| Generic Space | A person moves along a path, with relationships between position and time |
| Blended Space | Two monks travel along the same path in opposite directions on the same day |
Two journeys occurring on different days are integrated into a single imagined day.
In reality, there is only one monk.
In the Blended Space, however, we imagine two monks.
This hypothetical situation makes it easier to understand the relationship between the two journeys.
The example demonstrates that Conceptual Blending is not necessarily about physically combining existing objects. It is about integrating different situations to enable new reasoning.
The conclusion can also be established mathematically under the assumption of continuous movement along the same path. Conceptual Blending provides an intuitive representation of the underlying relationship.
Metaphor and Conceptual Blending
Metaphor is one of the important areas in which Conceptual Blending operates.
Consider the expression “computer virus.”
A biological virus is an infectious entity that uses a host cell’s machinery to replicate.
A computer virus is a type of malicious code capable of infecting and replicating through other programs or files.
The two are physically and technically different, but they share relationships involving infection, replication, and propagation.
These similarities allow people to understand certain software behaviors through concepts associated with biological infection.
| Biological Virus | Computer Virus |
|---|---|
| Uses host-cell mechanisms | Uses computer or program execution environments |
| Infects a host | Infects files or programs |
| Replicates | Copies its code |
| Spreads to other hosts | Propagates to other files or systems |
| Requires infection-control measures | Requires detection, protection, and removal measures |
However, a computer virus is not a biological organism.
Not every characteristic of biological infection applies to software.
Conceptual Blending selectively maps relationships between the domains to produce a new way of understanding the phenomenon.
The historical origin and technical definition of the term “computer virus” should be distinguished from analyzing the expression through Conceptual Blending Theory.
The theory can be used to explain how such metaphors are understood.
Differences Between Conceptual Blending and Analogy
Conceptual Blending and Analogy share the use of correspondences between different concepts or situations.
However, they emphasize different aspects of cognition.
Analogy uses knowledge from one domain to understand another or solve a different problem.
In particular, Gentner’s Structure-Mapping Theory emphasizes correspondences between relational structures rather than similarities in superficial attributes.
Conceptual Blending, by contrast, selectively integrates elements and relationships from multiple Input Spaces to create a new Blended Space.
For example, identifying a structural correspondence between a biological vascular system and an industrial fluid distribution network can be understood as analogical reasoning.
Using that correspondence to create the concept of “a self-healing material system that supplies repair agents to distributed damaged regions through internal pathways” can also be examined through Conceptual Blending.
However, Analogy and Conceptual Blending do not need to be treated as completely separate cognitive operations.
Structural mapping and the formation of new conceptual situations may occur together or in sequence.
The two theories can provide different perspectives on closely related cognitive phenomena.
Differences Between Conceptual Blending and Conceptual Combination
Conceptual Blending is closely related to Conceptual Combination.
Conceptual Combination broadly refers to the formation of new concepts or meanings by combining two or more existing concepts.
For example, combining “pet” and “fish” forms the concept of a “pet fish.”
Characteristics such as “kept in an aquarium” and “fed by an owner” may then be inferred.
Conceptual Blending Theory can explain such meaning construction through mappings and selective integration among different Mental Spaces.
However, research on Conceptual Combination also includes other theoretical approaches, such as attribute inheritance, relational interpretation, and constraint-based meaning construction.
| Aspect | Conceptual Combination | Conceptual Blending |
|---|---|---|
| Primary subject | Forming new concepts from multiple existing concepts | Integrating Mental Spaces to construct new meanings |
| Main operations | Attribute inheritance, property mapping, relational interpretation, conceptual restructuring | Cross-Space Mapping, Selective Projection, Emergent Structure |
| Basic explanation | Concepts A and B are combined to form concept C | Input Spaces and a Generic Space contribute to the formation of a Blended Space |
| Main research fields | Cognitive psychology, semantic interpretation, creativity research | Cognitive linguistics, cognitive science, meaning construction, creativity research |
| Theoretical position | A relatively broad cognitive phenomenon | A particular theoretical framework explaining meaning construction |
Conceptual Blending is an important theoretical approach to understanding Conceptual Combination.
However, not every act of combining concepts should automatically be described as Conceptual Blending.
For example, simply placing two different products in the same container does not necessarily generate a new conceptual structure.
The important question is whether the combination produces new relationships, meanings, or possibilities for inference.
Different Types of Conceptual Blending Networks
Fauconnier and Turner classify Conceptual Integration Networks into several types based on the relationships between their organizing frames.
Four representative types are Simplex, Mirror, Single-Scope, and Double-Scope networks.
| Network Type | Main Characteristics |
|---|---|
| Simplex Network | One input provides a frame of roles or relationships, while another supplies elements that fill those roles |
| Mirror Network | The input spaces and the blend share a common organizing frame |
| Single-Scope Network | The blend is organized primarily according to the frame of one input space |
| Double-Scope Network | Inputs with different organizing frames contribute elements and relationships to form a new structure |
The Buddhist Monk Problem, for example, is commonly discussed as an illustration of a Mirror Network.
In a Double-Scope Network, multiple inputs have different organizing principles that contribute to the blend.
Their assumptions or relationships may conflict.
However, reconciling those differences can lead to the formation of new conceptual structures.
Double-Scope Blending is particularly interesting in relation to creative thinking.
Nevertheless, a Double-Scope Network does not automatically produce original or useful ideas.
Furthermore, complex cognitive activities cannot always be classified neatly into only these four categories.
Compression of Vital Relations
Another important aspect of Conceptual Blending Theory is Compression.
For example, people can understand events occurring over long periods or relationships between distant objects as parts of a single conceptual situation.
We can think about technological developments spanning several decades as a single process of “technological evolution.”
We can also imagine what might happen if two scientists from different historical periods were able to debate one another.
Even though they never actually met, we can place them together in an imagined conceptual situation.
Such operations compress complex relationships involving time, space, causality, and identity into structures that are easier to reason about.
Fauconnier and Turner use the term Vital Relations for important relationships such as time, space, causality, and identity, examining how these are compressed through Conceptual Integration.
Compression in this context is not simply shortening information or producing a summary.
It involves transforming complex relationships into conceptual structures that humans can handle more intuitively.
This perspective is relevant to understanding complex systems and long-term processes of change.
Conceptual Blending and Creativity
Conceptual Blending provides an important framework for understanding creative thinking.
When generating new ideas, people do not necessarily discard existing knowledge and begin from nothing.
Instead, they may reinterpret familiar concepts through new relationships or integrate different situations into a new conceptual structure.
For example, applying a technology to an unfamiliar industry may require more than directly transferring knowledge from the original domain.
The constraints and requirements of the new domain may need to be integrated, producing a different technological configuration.
Conceptual Blending provides a framework for analyzing such processes.
However, Conceptual Blending Theory does not automatically evaluate the novelty or usefulness of an idea.
A blend that makes sense conceptually is not necessarily a scientific discovery or technological invention.
To produce a creative achievement, it is necessary to clarify what new relationships have been formed and examine the value they provide.
Conceptual Blending in Technological Development
Technological development often involves combining technologies or mechanisms originating in different fields.
Consider a company possessing equipment anomaly detection technology and another field in which systems record and retrieve historical events and their outcomes.
Anomaly detection technology observes equipment conditions and identifies abnormalities.
An event-recording system stores information about what happened in the past, what responses were taken, and what outcomes followed.
A simple combination might place anomaly detection results and historical records within the same user interface.
However, Conceptual Blending suggests the possibility of establishing new relationships between the two conceptual structures.
For example, the current equipment condition might be treated as an event pattern that can be compared with past situations.
Historical situations resembling the present condition could then be retrieved.
By examining which abnormalities occurred afterward and what responses were effective, the system might support decisions about the current situation.
| Space | Main Structure |
|---|---|
| Equipment anomaly detection | State observation, comparison with normal conditions, anomaly identification |
| Historical event records | Situations, events, responses, and outcomes |
| Shared abstract structure | Relationships among states, related events, and responses |
| Blended Space | A system that relates current conditions to past events and supports reasoning about possible developments and responses |
This example involves more than integrating two existing functions into one system.
It establishes the new relationship of interpreting a current equipment state through historical events.
This may lead to new ways of understanding abnormal conditions and evaluating possible responses.
However, this is an illustrative concept rather than a validated implementation.
Representing equipment states, calculating similarity, analyzing relationships among events, ensuring data reliability, and evaluating decision accuracy all require further investigation.
Example: Self-Healing Materials and Biological Supply Networks
Conceptual Blending can also be used to examine technological ideas inspired by other domains.
Consider self-healing materials and biological vascular systems.
Some self-healing materials contain encapsulated repair agents that are released when damage occurs.
However, one-time local storage may limit the ability to supply repair agents repeatedly to the same location.
Biological vascular systems, by contrast, transport substances continuously to different regions of an organism.
Comparing these systems suggests the following structure.
| Space | Content |
|---|---|
| Input 1: Self-healing material | Damage, repair agents, repair reactions |
| Input 2: Supply network | Supply sources, transportation pathways, distribution, continuous supply |
| Generic Space | Relationships involved in delivering required substances to required locations |
| Blended Space | A self-healing system that repeatedly supplies repair agents to damaged regions |
The resulting problems extend beyond simply placing tubes inside a material.
For example, how would the system identify damaged regions and direct repair agents toward them?
How would it maintain functionality if its supply pathways were also damaged?
Such questions arise when attempting to integrate the conceptual structures into a practical system.
Vascular supply networks for self-healing materials have already been investigated in scientific research. This example is intended to explain cross-domain Conceptual Blending, not to claim a new invention.
Making new relationships and necessary conditions explicit can support the formulation of R&D hypotheses.
Relationship with Abstraction
Conceptual Blending is closely related to Abstraction.
Abstraction extracts characteristics, functions, structures, and relationships relevant to a particular purpose from concrete objects or phenomena.
Conceptual Blending identifies correspondences among different spaces and selectively integrates elements to form new conceptual structures.
For example, biological vascular systems and industrial fluid distribution networks differ in appearance, materials, and operating mechanisms.
However, abstracting the relationship “transporting substances from a supply source to distributed locations” provides a basis for comparing them.
Conceptual Blending can use this common structure while integrating selected characteristics from the different spaces.
Nevertheless, Abstraction and Blending are not the same operation.
Abstraction extracts relevant properties or relationships from an object.
Blending uses such representations to form new structures of meaning.
The Generic Space represents common structure among the inputs, but this structure is not necessarily consciously or explicitly formulated during ordinary thinking.
Relationship with Systems Thinking
Systems Thinking emphasizes relationships among system components and the behavior of the system as a whole.
Conceptual Blending similarly emphasizes the formation of new relationships rather than merely placing independent elements together.
The two approaches therefore share an interest in relational structures.
However, Systems Thinking is primarily concerned with understanding the structures and dynamics of systems.
Conceptual Blending is a cognitive theory explaining how different conceptual structures are integrated.
For example, imagining a new control system inspired by a biological regulatory mechanism may be analyzed through Conceptual Blending.
Determining whether the proposed control system will operate stably requires control theory, system dynamics, experimentation, or other technical methods.
Conceptual novelty must be distinguished from system feasibility.
Using Conceptual Blending as a Creativity Method
Conceptual Blending was originally developed as a theory explaining human cognition.
However, its framework can be adapted to support creative problem-solving.
A possible practical procedure is as follows.
- Clarify the problem and objective. Determine what needs to be understood or what new function or solution is being sought.
- Select two or more Input Spaces. Identify the original problem or technology and another concept from a potentially relevant domain.
- Describe the structure of each Input Space. Identify elements, functions, relationships, operating principles, and constraints.
- Search for common relationships. Examine which elements or structures can be mapped between the inputs.
- Integrate selectively. Determine which characteristics and relationships should be projected into a new conceptual structure.
- Elaborate the blend. Explore what would happen in the new situation and what additional functions would be required.
- Identify emergent features and contradictions. Clarify new relationships, necessary conditions, and unresolved problems.
- Concretize and verify. Develop the idea into a technology, product, service, or other application and examine its feasibility.
This is a practical adaptation of Conceptual Blending Theory for creative problem-solving.
It is not a standardized idea-generation procedure prescribed by Fauconnier and Turner.
Nor does following these steps guarantee that a new or useful concept will emerge.
In practice, the selection of Input Spaces, identification of shared structures, and interpretation of the blend may need to be repeated.
Evaluating Blended Concepts
Concepts formed through Conceptual Blending should be evaluated separately for meaningfulness, novelty, and practical usefulness.
Possible criteria include:
| Evaluation Criterion | Key Question |
|---|---|
| Validity of Mapping | Are the correspondences between Input Spaces meaningful? |
| Appropriateness of Selective Projection | Are suitable elements and relationships incorporated into the blend? |
| Emergent Structure | Are new relationships or functions formed that were not present in the original inputs? |
| Semantic Coherence | Can the resulting concept be understood consistently? |
| Novelty | How does the concept differ from existing concepts or technologies? |
| Usefulness | Does it contribute to new understanding or problem-solving? |
| Feasibility | Can the necessary technical and physical conditions be satisfied? |
| Verifiability | Can the inferred functions or effects be tested? |
A clearly articulated Emergent Structure does not necessarily imply high technical feasibility.
Similarly, the presence of common features between Input Spaces does not guarantee that those similarities are relevant to the problem being solved.
When Conceptual Blending is applied to R&D, the resulting concept must be examined against concrete technological conditions.
Limitations of Conceptual Blending Theory
Conceptual Blending Theory has several important limitations.
First, it explains cognitive processes but does not guarantee that a new concept will be useful.
Second, the identification of Input Spaces and Blended Spaces may depend on the interpretation of the analyst. Different analyses of the same example may be possible.
Third, explaining a phenomenon through Conceptual Blending after it has occurred is different from accurately predicting the phenomenon in advance.
Fourth, it may be difficult to specify precise rules determining which concepts are selected and why particular blends are formed.
Fifth, the formation of emergent meaning in cognition is different from the actual emergence of new physical or technological properties.
Sixth, it may be difficult to establish how explanations based on Conceptual Blending differ empirically from simpler cognitive theories or alternative theories of analogy.
Seventh, much of the theory has been developed through the analysis of language comprehension and meaning construction. It is not an established method for directly predicting the success of engineering inventions.
The 2024 paper “Integrational Creativity: From Combining and Blending to Transforming and Resonating” also examines questions concerning the falsifiability of Conceptual Blending Theory, comparisons with alternative explanations, and the processes through which new concepts become established socially.
These issues do not invalidate the theory.
They indicate the importance of distinguishing its explanatory value from its empirical limitations.
When applying Conceptual Blending to creative problem-solving, theoretical interpretations should therefore be distinguished from demonstrated practical effectiveness.
Conceptual Blending with AI
LLMs can represent features and relationships from multiple concepts or situations in natural language and propose new combinations.
For example, an LLM can organize the functions, structures, and operating principles of two technologies and identify possible correspondences between them.
It can then selectively combine characteristics to generate a new system concept.
However, producing an output resembling Conceptual Blending does not demonstrate that an LLM internally performs the same cognitive operations as humans.
Even if an output can be analyzed using four Mental Spaces, this does not necessarily imply that the model internally constructs those same structures.
Nevertheless, the Conceptual Blending framework may be useful as an external method for structuring AI-assisted idea generation.
For example, a prompt simply asking an LLM to “combine two technologies and propose a new idea” may produce familiar combinations or suggestions that merely place existing functions together.
A more structured approach could explicitly examine Input Spaces, shared structures, Selective Projection, emergent relationships, and conditions for feasibility.
However, whether such an approach produces better results than conventional AI-based idea generation requires empirical comparison and evaluation.
Structuring AI-Assisted Blending
When AI is used to support Conceptual Blending, it can be useful to record conceptual mappings separately from inferences about the resulting blend.
A possible analysis structure is shown below.
| Analysis Element | Description |
|---|---|
| Input Space A | Elements, functions, and structures of the original technology or problem |
| Input Space B | Elements, functions, and structures of a technology or phenomenon from another domain |
| Correspondences | Elements and relationships that can be mapped between the inputs |
| Generic Space | Abstract structure shared by the inputs |
| Selective Projection | Characteristics and relationships incorporated into the new concept |
| Blended Space | New concept formed through integration |
| Emergent Structure | New functions or relationships inferred through integration |
| Constraints and Unresolved Issues | Conditions required for practical realization |
| Verification Methods | Research or experiments needed to test the resulting hypotheses |
Organizing results in this way makes it possible to examine how a concept was formed rather than merely generating an appealing name or description.
It also helps distinguish characteristics inherited from existing technologies from those newly hypothesized through the combination.
However, Emergent Structures described by an LLM remain conceptual inferences.
They do not establish scientific or technological feasibility.
Relationship with Concept Exploration
Conceptual Blending can be connected with Concept Exploration, an approach proposed by Mindware Research Institute.
Concept Exploration emphasizes searching for unfamiliar objects through their characteristics, functions, structures, relationships, and patterns rather than relying only on established terminology or disciplinary categories.
Conceptual Blending focuses on forming new meanings and relationships by integrating conceptual structures from different domains.
Combining the approaches suggests the following process.
First, relevant characteristics, operating principles, functions, and structures are extracted from a known technology.
These are then abstracted to identify functions and relationships that may also exist in different industrial fields.
The abstract structures are used to explore unfamiliar technologies and phenomena.
The discovered technologies are compared with the original technology to identify common structures and important differences.
Selected elements and relationships are then integrated to create new technological concepts or R&D hypotheses.
Finally, the necessary functions, technical constraints, and feasibility of the resulting concepts are examined.
This process involves interactions among Abstraction, Analogy, Structure Mapping, Conceptual Combination, and Conceptual Blending.
However, Concept Exploration and Conceptual Blending are not the same activity.
Concept Exploration places particular emphasis on discovering previously unknown subjects.
Conceptual Blending focuses on forming new conceptual structures by integrating different bodies of knowledge.
Concept Exploration provides an approach to discovering unfamiliar knowledge and technologies, while Conceptual Blending offers a theory for understanding how different knowledge structures can be integrated into new meanings and relationships.
Applications to ConceptMiner R&D
ConceptMiner R&D extracts technological elements from documents and uses their characteristics and patterns to support Cross-Domain Concept Exploration.
This process can identify technologies and phenomena from domains different from that of the original R&D problem.
However, discovering an unfamiliar technology does not necessarily produce a sufficiently developed R&D theme.
For example, structural similarity between the original technology and a technology from another domain does not automatically indicate how a new technological configuration should be developed.
The Conceptual Blending framework could therefore be used to represent the original technology and the discovered technology as separate Input Spaces.
Their shared structures and correspondences could then be examined.
Selected elements and relationships could be integrated to form a new Blended Space.
Particular attention should be given to the functions and structures inferred through this integration.
For example, if a proposed combination requires a function that neither original technology can provide, the missing function may become a new R&D challenge.
Similarly, conflicting operating conditions may reveal a technical contradiction that must be resolved.
Recording this information could make it possible to trace which technological elements contributed to a hypothesis and which mappings and inferences supported its formation.
However, whether explicitly applying Conceptual Blending produces better R&D hypotheses remains a question requiring validation.
For example, conventional LLM-based hypothesis generation, Structure-Mapping-based exploration, and hypothesis generation explicitly using Conceptual Blending could be compared.
Their results could be evaluated according to novelty, usefulness, technical feasibility, and other appropriate criteria.
The resulting hypotheses should also remain traceable to their source documents and undergo technical verification.
Conceptual Blending in ThinkNavi
ThinkNavi regards Conceptual Blending as a thinking-support approach for forming new meanings and structures by establishing correspondences between different conceptual worlds and selectively integrating their elements.
ThinkNavi Free could provide a worksheet based on the four-space model of Conceptual Blending.
For example, users could enter two concepts or technologies and describe the elements and relationships associated with each.
They could then identify correspondences and common abstract structures between the two inputs.
Next, they could select which characteristics should be incorporated into the new concept and examine the functions and relationships formed in the Blended Space.
Such a worksheet should do more than simply place two concepts side by side.
It should clarify the correspondences through which integration becomes meaningful.
It would also be useful to distinguish newly generated conceptual meanings from the conditions required for their practical realization.
An LLM-assisted version could support structuring Input Spaces, extracting shared relationships, suggesting Selective Projections, and developing new inferences within Blended Spaces.
It could also generate multiple alternative blends from the same inputs so that users can compare different possibilities.
For example, combining self-healing materials with supply networks need not lead only to the concept of transporting repair agents.
Different blends could emphasize communication of damage information, selection of repair locations, or maintenance of transportation pathways.
The resulting R&D questions would differ depending on which relationships are integrated.
Furthermore, technologies discovered through ConceptMiner R&D’s Cross-Domain Concept Exploration could be used as Input Spaces.
This would provide a structured approach to developing discovered knowledge into new technological concepts.
However, performing Conceptual Blending should not become an objective in itself.
After forming a new conceptual structure, it is still necessary to examine what problem it solves and under which conditions it could be realized.
The significance of Conceptual Blending lies not in simply combining different concepts, but in using their underlying structures to create new relationships and possibilities for reasoning that were not previously apparent.
A process that extracts common structures through Abstraction, identifies cross-domain correspondences through Analogy, forms new conceptual structures through Conceptual Blending, and then concretizes and verifies the results represents one possible approach to systematizing creative problem-solving and R&D theme exploration.
However, its practical value should be evaluated through actual exploration cases and research outcomes, not solely through theoretical explanations.
Primary English Terms: Conceptual Blending / Conceptual Integration
Related Terms: Mental Spaces, Input Space, Generic Space, Blended Space, Cross-Space Mapping, Selective Projection, Emergent Structure, Composition, Completion, Elaboration, Double-Scope Network, Vital Relations, Compression
Related Thinking Methods:
- Abstraction | Extracting Features, Functions, and Structures from Concrete Phenomena to Discover New Possibilities
- Conceptual Combination | Creating New Meanings, Functions, and Possibilities by Combining Concepts
- Concept Formation | How Categories Are Made
- Analogy — Connecting Knowledge Across Domains to Generate New Ideas and Hypotheses
- Structure-Mapping Theory | Understanding Relational Correspondences and Generating New Inferences Across Domains
- Metaphorical Thinking,
- SCAMPER | Generating New Ideas Through Seven Perspectives
- TRIZ | A Systematic Approach to Inventive Problem Solving Through Technical Contradictions
- Morphological Analysis | Systematically Exploring New Possibilities Through Combinations of Elements
- Systems Thinking — Understanding Interdependence, Feedback, and Change Over Time
- Concept Maps — Representing Relationships Between Concepts and Understanding Knowledge Structures
- Concept Exploration | Discovering Emerging Concepts and Unknown Possibilities
- Cross-Domain Concept Exploration,
- R&D Theme Exploration and Generation
References and Further Reading
- Mark Turner and Gilles Fauconnier, “Conceptual Integration and Formal Expression,” Metaphor and Symbolic Activity, Vol. 10, No. 3, pp. 183–204, 1995. Research on the integration of multiple Mental Spaces to construct new meanings.
- Gilles Fauconnier, “Creativity in Everyday Language,” Cognitive Studies: Bulletin of the Japanese Cognitive Science Society, Vol. 3, No. 3, pp. 14–27, 1996. A discussion of creative cognitive operations in everyday language comprehension and hypothetical reasoning.
- Gilles Fauconnier and Mark Turner, “Conceptual Integration Networks,” Cognitive Science, Vol. 22, No. 2, pp. 133–187, 1998. A foundational paper systematizing Conceptual Blending, Selective Projection, and Emergent Structure.
- Mark Turner and Gilles Fauconnier, “A Mechanism of Creativity,” Poetics Today, Vol. 20, No. 3, pp. 397–418, 1999. An examination of Conceptual Integration as a cognitive mechanism for generating new meanings.
- Gilles Fauconnier and Mark Turner, The Way We Think: Conceptual Blending and the Mind’s Hidden Complexities, Basic Books, 2002. A comprehensive presentation of Conceptual Blending Theory.
- Gilles Fauconnier and Mark Turner, “Conceptual Blending, Form and Meaning,” Recherches en communication, No. 19, 2003. A discussion of Conceptual Integration and meaning construction.
- Dedre Gentner, “Structure-Mapping: A Theoretical Framework for Analogy,” Cognitive Science, Vol. 7, No. 2, pp. 155–170, 1983. A foundational study of Structure-Mapping Theory.
- Fintan J. Costello and Mark T. Keane, “Efficient Creativity: Constraint-Guided Conceptual Combination,” Cognitive Science, Vol. 24, No. 2, pp. 299–349, 2000. Research on constraints and creative meaning construction in Conceptual Combination.
- Hung-Hsiang Wang, “A Case Study on Design with Conceptual Blending,” International Journal of Design Creativity and Innovation, Vol. 2, No. 2, pp. 109–122, 2014. A case study applying Conceptual Blending to design.
- “Integrational Creativity: From Combining and Blending to Transforming and Resonating,” Inquiry, 2024. A discussion of Conceptual Blending, creativity, and theoretical challenges.
- Mark Turner, “Blending and Conceptual Integration.” A collection of research publications and resources on Conceptual Blending.