Creativity is the process of producing possibilities, combinations or expressions that are sufficiently new and sufficiently useful, meaningful or appropriate for a particular context.
In one line: creativity works when knowledge supplies material, attention notices unusual relationships, the mind generates alternatives, constraints give those alternatives shape, and testing turns an interesting possibility into something that actually works.
Evidence boundary: Creativity research spans psychology, education, design, arts and innovation. There is no single universal creativity mechanism or test. The model below combines well-supported ideas—domain knowledge, divergent and convergent processes, constraints, iteration and feedback—into a public explanatory map rather than a formula for scoring a person’s creativity.
Creativity is often described as if ideas arrive from nowhere: inspiration strikes, the artist sees what nobody else can see, or the inventor suddenly has a breakthrough.
Breakthroughs do happen. But most creative work has a longer history underneath it. People collect knowledge, notice problems, compare examples, borrow structures, test combinations, reject weak attempts and gradually produce something that did not exist in that form before.
What Is Creativity?
A useful definition of creativity requires two things. The output must have some novelty, but novelty alone is not enough. It must also be useful, fitting, meaningful or valuable relative to the job.
A random sentence nobody has written before may be novel but useless. A perfectly correct sentence copied from a model may be useful but not very creative. Creative work lives in the tension between difference and fit.
Knowledge → notice → generate → combine → constrain → test → select → refine → share → receive feedback → revise.
1. Creativity Needs Material to Work With
People do not create from an empty mind. Writers use language, genre, memory and experience. Scientists use existing theories, methods and observations. Designers use materials, engineering knowledge and user needs. Students use vocabulary, examples, concepts and patterns they have previously encountered.
This is why knowledge and creativity are not opposites. Rich knowledge can increase the number of meaningful combinations available.
The danger is not knowledge itself. It is rigid knowledge that permits only the familiar arrangement.
2. Attention Notices Possibilities Others May Ignore
Creative work often begins by noticing an unusual feature, contradiction, unmet need or relationship.
Why does this method fail here? Why do children misunderstand this explanation? Why does one material bend while another breaks? Why does this story feel emotionally flat? Why does a process contain a step that seems unnecessary?
The question changes the search field. A different observation can generate a different creative route.
3. Divergent Thinking Generates Alternatives
Some phases of creativity benefit from producing several possibilities before judging them too quickly.
What else could this object do? How many explanations fit the evidence? How could the introduction begin? Which representations could make this idea easier to understand?
Generating options expands the solution space. But a large pile of ideas is not yet a finished creative product.
4. Recombination Creates New Structures From Existing Pieces
Many creative ideas are new combinations rather than entirely new components.
A metaphor connects two domains. A new lesson combines a familiar concept with a different representation. A technological product may combine existing sensors, software and distribution systems in a new arrangement.
Recombination becomes more powerful when a person can cross boundaries between domains rather than seeing every subject as an isolated box.
5. Constraints Can Increase Creativity
Unlimited freedom sounds creative, but too many possibilities can make selection difficult.
Constraints define the problem. Write within 500 words. Build with these materials. Explain the idea to a Primary 4 student. Use only evidence available before a certain date. Reduce the cost by half without reducing safety.
Good constraints remove irrelevant possibilities and force more interesting trade-offs. Bad constraints merely prevent useful routes without serving the job.
6. Incubation Lets the Search Continue Outside Direct Effort
People sometimes solve a problem after stepping away from it. A walk, sleep or unrelated task can be followed by a new connection.
This does not mean the mind magically creates while doing nothing. Often the problem has already been richly represented and explored. A break changes attention, reduces fixation and allows different associations to become available.
Incubation works best after genuine engagement, not as a substitute for it.
7. Externalising the Idea Makes It Testable
An idea in the head can feel more complete than it really is.
Sketch it. Write the paragraph. Build the prototype. Explain the concept aloud. Run the calculation. Once externalised, missing relationships and unintended consequences become visible.
Creative work becomes stronger when imagination meets a representation that can resist it.
8. Convergent Thinking Selects What Deserves to Survive
Creativity also requires judgement.
Which idea actually solves the problem? Which explanation is accurate? Which story opening creates the intended effect? Which design can be produced safely and affordably?
Generating without evaluating produces novelty without discipline. Evaluating too early produces discipline without enough novelty. Strong creative work alternates between opening and narrowing the search.
9. Feedback Exposes What the Creator Cannot See Alone
The creator knows what they intended. The audience sees only what actually arrived.
Readers reveal where a text becomes confusing. Users reveal where a design creates friction. Experiments reveal whether a hypothesis survives contact with evidence. Teachers reveal whether an original answer still meets the assessment criteria.
Feedback does not mean surrendering creative ownership. It provides additional information for the next decision.
10. Expertise Changes the Search Space
Experts often appear creative because they can see possibilities novices cannot yet recognise. They know which rules are essential, which conventions can be broken and which apparent shortcuts will fail later.
But expertise can also create fixation if familiar solutions become too dominant. Creative expertise therefore needs both a deep domain model and the ability to question its default routes.
11. Creativity Is a Process, Not a Personality Verdict
Calling one child “creative” and another “not creative” can hide important differences among domains and conditions.
A student may generate unusual stories but struggle with visual design. Another may be inventive in mathematics but cautious in open-ended writing. Creativity depends partly on knowledge, confidence, task structure and whether the environment permits experimentation.
A more useful question is: Under which conditions can this learner generate, test and refine worthwhile alternatives?
12. AI Changes the Creative Environment Without Removing the Human Job
Generative AI can expand option generation, transform representations and accelerate drafts. That changes the cost of producing possibilities.
But lower generation cost increases the importance of selection, evidence, taste, purpose and accountability. Someone still has to decide what problem is worth solving, what is true, what is appropriate for the audience, what should be rejected and who remains responsible for the final output.
UNESCO’s 2026 ICT in Education Prize explicitly frames AI-era creativity around protecting human critical thinking, imagination, judgement and agency rather than allowing fluent machine output to substitute for them.
The Whole Creativity Chain
Build knowledge → notice a gap or possibility → generate alternatives → recombine → use constraints → externalise → evaluate → test → receive feedback → refine → share → learn from the result.
A Useful Metaphor: Creativity Is a Workshop
The workshop contains materials accumulated over time. Tools make some transformations possible. Constraints come from the job, the material and the user. Drafts can be cut apart and rebuilt.
A workshop with no materials cannot build much. A workshop that never experiments produces only copies. A workshop that never measures produces objects that may look interesting but fail when used.
Creativity at Three Zoom Levels
Micro: one variation
Can the learner produce another way to express, represent or solve this?
Meso: one creative project
Can several possibilities be generated, compared, tested and refined into a coherent output?
Macro: a creative culture
Does the environment provide knowledge, freedom to test, useful constraints, diverse perspectives and enough safety for unsuccessful attempts to produce learning rather than humiliation?
How Creativity Fails
- Empty novelty: the output is different but does not serve the job.
- Knowledge poverty: there are too few useful materials to recombine.
- Premature judgement: weak first ideas prevent exploration of better alternatives.
- No convergence: possibilities multiply but nothing is selected or finished.
- Constraint collapse: novelty ignores truth, safety, audience or purpose.
- Fixation: the first familiar solution dominates every attempt.
- Identity labelling: people are classified as creative or uncreative instead of developing the mechanisms.
How Creativity Is Repaired
Enrich the inputs. Change the representation. Generate several deliberately different alternatives before selecting. Add one meaningful constraint. Compare with examples outside the immediate domain. Build a rough prototype earlier. Ask for feedback on the job rather than approval of the person.
Most importantly, separate two phases that are often confused: make more possibilities and judge which possibilities deserve to survive.
What Parents and Students Should Notice
- Does the learner have enough knowledge to create with?
- Can they generate more than one possible route?
- Can they explain why one idea is better for the job?
- Do constraints improve the work or merely frustrate it?
- Can the student tolerate a weak draft without treating it as a verdict on ability?
- Can feedback change the product without erasing ownership?
- Can AI or other tools expand possibilities without replacing judgement?
Creativity Begins Before Idea Generation: Someone Has to Find the Right Problem
Many creative tasks are framed as “find a solution”. But some of the most important creative work happens earlier: which problem deserves attention, and how should it be framed?
A weak framing can make excellent idea generation useless. “How do we make students revise longer?” produces one search space. “Why does revision time fail to become retrievable knowledge?” produces another. The second framing may reveal retrieval, feedback, spacing or method-selection problems that the first question hides.
Observe the world → notice a mismatch → frame the problem → test whether the framing explains the real constraint → only then generate solutions.
Creative Search Has a Space, and Knowledge Changes Its Shape
A creative problem contains possible moves, combinations and representations. Novices may search broadly but shallowly because they do not know which possibilities are meaningful. Experts can often exclude impossible routes quickly and search more deeply inside promising regions.
This means creativity is not simply “more ideas”. A useful search balances breadth — enough variation to escape the obvious — with structure — enough domain knowledge to distinguish a promising possibility from noise.
Analogy Lets Structure Travel Between Domains
Analogy is one of creativity’s strongest transfer mechanisms. A person notices that a relationship in one domain resembles a relationship in another and asks whether some structure can be carried across.
Good analogy preserves the relevant relationship rather than copying surface appearance. A teacher may borrow feedback loops from engineering, a designer may borrow branching from biology, or a writer may borrow narrative structure from an historical event.
But analogy can mislead when the structural differences matter more than the similarities. Creative transfer therefore needs a boundary check: what survives the analogy, and what does not?
Fixation Is Often a Search Problem, Not a Lack of Imagination
The first example, familiar method or dominant representation can anchor later thinking. Once one route becomes mentally available, alternatives may be generated only as small variations around it.
Useful anti-fixation moves include changing representation, delaying exposure to a model answer, generating solutions from deliberately different categories, using an extreme case, reversing an assumption, or asking what someone from another discipline would notice.
If every new idea resembles the first idea, widen the search mechanism before demanding more effort.
Constraints Have a Sweet Spot
Too little constraint creates an enormous search space with weak direction. Too much constraint removes the degrees of freedom required for novelty.
The useful question is therefore not whether constraints help creativity in general, but which constraints define the real job and which constraints merely preserve habit or authority.
A strong brief protects truth, safety, audience and resource limits while leaving enough freedom for the form of the solution to change.
Incubation Is Useful Because Attention and Fixation Change
Stepping away can improve creative performance in some situations, especially after the problem has already been encoded and explored. A break may weaken fixation, allow forgotten associations to reactivate, or let attention return with a different weighting.
This is not evidence that every difficult task should be abandoned until inspiration arrives. Incubation is most useful as part of a larger loop: engage deeply → reach a meaningful impasse → disengage temporarily → return and test what changed.
Originality Needs a Reference Class
“Original” always means original relative to something. An idea may be new to the student, new to the classroom, new to a professional field or new to recorded history. These are different achievements.
This matters in education. A child independently rediscovering a known mathematical pattern may be doing genuinely creative cognitive work even though the mathematics itself is centuries old.
At research or commercial scale, provenance matters more. Claims of originality require wider search, prior-art awareness and clearer evidence that the contribution differs materially from existing work.
Novelty, Usefulness and Truth Can Pull in Different Directions
A creative scientific explanation must still answer to evidence. A creative engineering design must still obey physical constraints. A creative essay may deliberately bend convention while still needing enough coherence for a reader to reconstruct the intended effect.
Different domains therefore use different acceptance tests. Creativity does not exempt an output from the standards of the domain in which it claims to work.
Collaborative Creativity Requires Independent Input Before Social Convergence
Groups can combine more experience than one person, but early group discussion can also cause anchoring. The first confident suggestion becomes the centre of the search, status shapes whose ideas survive, and quieter members stop contributing alternatives.
One useful architecture is:
Independent generation → visible pool of alternatives → clarify without defending → evaluate against explicit criteria → combine where useful → prototype → test.
This preserves diversity long enough for the group to benefit from it.
Creative Evaluation Needs Criteria Before Attachment Becomes Too Strong
Creators naturally become attached to ideas they have invested in. Evaluation becomes more reliable when important criteria are stated before one candidate becomes emotionally dominant.
- Does it solve the intended problem?
- Is it accurate enough for the domain?
- Is it meaningfully different from existing options?
- Can the receiver use it?
- What does it cost?
- What risks or externalities does it create?
- What evidence would make us reject it?
Criteria do not eliminate taste or judgement. They make part of the judgement inspectable.
Creative Culture Is an Allocation System for Attention, Risk and Permission
Organisations often say they want creativity while punishing unsuccessful experiments, allocating no time for exploration, rewarding only predictable output or requiring every idea to be approved before it can become concrete.
A genuinely creative culture decides where experimentation is safe, how much failure can be tolerated, which standards cannot be relaxed, how ideas receive resources, and how weak experiments are stopped without humiliating the people who proposed them.
AI Makes Generation Cheap, So Selection Becomes More Expensive in Relative Terms
Generative AI can produce dozens of plausible variants almost instantly. This changes the bottleneck. The scarce resource may no longer be the first draft; it may be problem definition, verification, originality checking, taste, integration and responsibility for what is finally released.
AI-assisted creative work therefore benefits from preserving provenance: what came from the human, what came from external evidence, what was generated by the model, and which transformations produced the final output.
A High-Resolution Creativity Audit
- Problem: Are we solving the right problem or only the first visible symptom?
- Reference class: New to whom — learner, group, field or world?
- Knowledge base: What materials, examples and principles are available to recombine?
- Representation: Does the current framing hide useful relationships?
- Search breadth: Are alternatives genuinely different or cosmetic variants?
- Analogy: Which structure can transfer from another domain, and where does the analogy break?
- Fixation: What first example or convention is dominating the search?
- Constraints: Which limits serve the job and which merely preserve habit?
- Incubation: Has the problem been engaged deeply enough for a break to be useful?
- Externalisation: Can the idea be inspected as a draft, model or prototype?
- Criteria: How will novelty, usefulness, truth, cost and risk be judged?
- Independent input: Did group members generate alternatives before social convergence?
- Provenance: What existing work, source or AI generation contributed to the output?
- Test: What world receipt tells us whether the idea works?
- Learning: What did the rejected attempts teach that should change the next search?
Connect Creativity to the Wider eduKateSG Mechanism Estate
- How Knowledge Works — the organised material creative search recombines.
- How Problem Solving Works — how search spaces, representations and constraints become usable routes.
- How Evidence Works — why creative claims must still meet the standards of their domain.
- How AI Works — how generative systems expand option production while shifting the bottleneck toward selection and verification.
- How Innovation Works — what has to happen after a creative possibility exists for it to become useful change.
Continue Through eduKateSG
- How Knowledge Works
- How Thinking Works
- How Problem Solving Works
- How Culture Works | The Role of Culture
Evidence and Further Reading
UNESCO’s Framework for Culture and Arts Education implementation guidance, updated in January 2026, identifies creativity and critical thinking among the capabilities culture and arts education can nurture. UNESCO’s 2026 ICT in Education Prize goes further in the AI context, emphasising critical thinking, imagination, ethical judgement and human agency as technologies become more capable of generating content.
Frequently Asked Questions
Can creativity be taught?
Conditions and processes that support creative work can be taught and practised: domain knowledge, generating alternatives, changing representations, using constraints, testing ideas and learning from feedback. No single exercise makes every person creative in every domain.
Does memorisation reduce creativity?
Not inherently. Well-organised knowledge can supply material for creative work. Problems arise when learning stops at reproduction and never asks students to recombine, apply, question or transform what they know.
Is AI creative?
AI systems can generate outputs that humans judge novel or useful, but the educational question remains broader: who defines the purpose, evaluates truth and value, accepts consequences and decides what should enter the world? Human judgement remains central to those responsibilities.
Final compression: Creativity works when existing knowledge is opened into new combinations, alternatives are allowed to grow, and disciplined testing turns novelty into something that is genuinely useful, meaningful or true to the job.