A team is asked how an empty shop could serve its neighborhood. Someone proposes a repair café, another a place to rehearse music, a third a shared kitchen, and a fourth a quiet study room. None of those suggestions is yet a workable plan. Each opens a different path for investigation. The group has widened the space of possible answers before selecting one. That movement from one prompt toward multiple plausible directions is the central idea of divergent thinking.
Divergent thinking is the generation of multiple ideas or responses to an open-ended question. Researchers often study it with tasks that permit many answers rather than one prescribed solution. The concept is narrower than creativity. A creative contribution also has to be judged in relation to a purpose, field, and setting; novelty without relevance may be merely unusual. Divergent thinking can supply candidate ideas, but evaluation, refinement, implementation, and reception still matter. It is therefore better understood as one part of a creative process or an indicator of creative potential than as a complete measure of creativity.
What diverges, and from what?
An open-ended prompt gives a starting point. Thinking diverges when a person explores more than one route away from it. The routes may differ in purpose, category, audience, material, or underlying assumption. Asked for ways to use a cardboard box, a person might imagine storage, a stage prop, a plant shelter, and a model building. That is different from producing four storage arrangements. Both responses increase the count of ideas, but only one crosses several conceptual categories.
The contrast with convergent thinking is useful if it is not made absolute. Convergent work narrows options under criteria, such as selecting a safe, affordable use of the box. Most real projects need both movements. A designer may open a range of ideas, examine constraints, generate a revised set, and then choose again. Even during idea generation, the person must understand the prompt, remember prior responses, and avoid exact duplicates. Divergence is not the absence of judgment or knowledge.
Runco and Acar describe divergent-thinking tests as indicators of creative potential. Their distinction matters: a high score can suggest a capacity to generate possibilities under test conditions, but it does not guarantee creative work in daily life. A person may have strong ideas but lack resources or an opportunity to develop them. Someone with an ordinary test score may make a valuable contribution through domain knowledge, sustained collaboration, or careful revision. The score and the achievement are different outcomes.
Nor is divergence identical to rebellion or originality for its own sake. A genuinely useful alternative can be a modest change in how a problem is framed. When a clinic asks how to reduce missed appointments, proposing a different reminder may be one answer; asking whether transport, appointment timing, or inaccessible communication is the larger issue opens different classes of answers. The quality of that move depends on evidence about the clinic and its patients, not on how surprising it sounds.
The alternate-uses task
A common research instrument is the alternate-uses task. Participants receive a familiar object, often a brick, paper clip, or bottle, and are asked to list uses other than its usual one within a set time. The task is attractive because it is simple to administer and yields multiple observable responses. The researcher can count answers, judge how unusual they are, and examine the range of categories represented. A prompt that seems trivial can reveal meaningful differences in strategies and scoring.
Imagine two participants each given three minutes to find uses for a brick. One lists fifteen construction variants. The other lists eight responses that include a doorstop, a weight, a drawing surface, and a percussion instrument. The first is more fluent if fluency means the number of admissible ideas. The second may appear more flexible if the responses span more categories. Whether either person’s ideas are more original depends on how originality is defined and scored. A vivid anecdote cannot replace an explicit scoring rule.
The task also depends on instructions. “List as many uses as possible” invites a different balance of speed and selectivity from “give your most creative uses.” A time limit, sample language, object familiarity, and permission to elaborate can change what participants do. A child, engineer, and artist may draw on different experiences. Their task scores should not be read as context-free amounts of a single personal substance called creativity.
Hass argues that standard summary scores obscure the dynamics of idea generation. Responses arrive in an order, and later ideas may move farther from common associations or reflect a new search path. Looking only at a final total can hide how a person shifted from one category to another. Semantic-distance methods offer one way to study that movement, but they too depend on a language model and on which kinds of distance matter to the task. They are analytical tools, not a transparent window into thought.
Fluency: how many admissible ideas?
Fluency is commonly the number of distinct, relevant responses produced under the task rules. Its attraction is clarity: researchers can usually agree on whether a response was offered, though they may still disagree about duplicates or whether an idea meets the prompt. Fluency can reflect search effort, verbal facility, task engagement, prior experience, and available time. It is a performance measure under specified conditions, not a direct reading of a permanent trait.
Quantity sometimes matters because more candidates create more chances for a useful one. It does not follow that quantity is sufficient. A list of twenty near-identical suggestions may offer less decision value than five well-differentiated options. Conversely, a team that stops at its first respectable idea may miss alternatives. A practical session can therefore separate the generation of a broad candidate set from a later examination of feasibility, value, and consequences.
Fluency also complicates other scores. If originality is calculated by adding points for every rare response, someone who gives more responses has more opportunities to earn points. If flexibility is counted by the number of categories visited, a longer list likewise has more chances to cross categories. This statistical dependence does not make originality or flexibility meaningless, but it means that a high score on one dimension may partly reflect high fluency. Scoring designs must decide whether and how to account for that relation.
In educational settings, encouraging many possibilities may be valuable as a practice, but one should not claim that any single classroom prompt increases general creativity. A useful learning question is what students did with the possibilities: Did they notice assumptions, use evidence, explain a surprising connection, and revise their answer? The count is one observation among several.
Originality: novel relative to what?
Originality concerns the unusualness or novelty of a response. A statistical method may label an answer original if few people in a reference group give it. Human judges may instead rate how surprising and apt it seems. These are not interchangeable procedures. An answer can be rare because it is clever, because it is obscure, or because it misunderstands the prompt. A response can be common in one community but fresh in another.
Silvia and colleagues studied subjective scoring methods for divergent-thinking tasks and examined their reliability and validity. One approach asks participants to nominate their most creative responses and has judges rate those selected ideas. This can reduce the simple advantage of producing a long list, because each person’s judged set is limited. Yet the result now depends on a person’s ability to recognize their best ideas and on raters’ criteria. The method solves some measurement problems while introducing others worth reporting.
Originality also needs to be distinguished from usefulness. A brick used as an improvised paperweight may be familiar but effective; a bizarre proposal may be statistically rare but unsafe or irrelevant. Creative evaluation generally asks whether novelty works with a purpose, even if different fields weigh the two differently. Divergent-thinking tasks often sample only a small part of this larger judgment. A scoring system can identify unusual responses without showing that someone could develop them into a contribution.
This is especially important when comparing people across languages or social settings. A judge’s sense of an “obvious” answer reflects experience and a reference group. Someone may generate ideas that are highly relevant locally but unfamiliar to outside raters, or use a term that translation makes sound ordinary. Fair assessment needs transparent prompts, scoring criteria, and evidence that they are appropriate for the population being assessed.
Flexibility: movement among categories
Flexibility usually refers to the variety of response categories or the ability to move between them. In an alternate-uses task, a response set that includes transport, shelter, art, and sound may cover more categories than one centered entirely on storage. However, categories are made by researchers or raters. Their width and boundaries affect the score. Is a brick used to prop a door in the same category as a brick used to hold down paper? The answer depends on the coding scheme.
Weiss and Wilhelm studied whether flexibility measured in divergent-thinking tasks captured something beyond fluency and originality. In their adult sample and latent-variable models, the flexible-thinking indicators shared substantial variance with those other measures; the residual flexibility factor did not show the expected links to several selected covariates. Their conclusion is limited to the tasks and operationalizations studied, but it is a strong reminder that naming three scores does not prove three independent abilities.
Flexibility in this measurement sense is also not identical to everyday adaptability. A person might move across categories in a timed word task yet struggle to revise a work plan when colleagues present new evidence. Another may offer few ideas in the task but adapt skillfully in a familiar domain. The task has its own demands, including language, speed, and understanding of the prompt. It should not be used as a shortcut diagnosis of a person’s general openness or competence.
In a group conversation, category movement can still be a useful practical observation. If every proposal assumes the same audience, cost structure, or technology, a facilitator can invite a new perspective. That invitation is a design move, not a claim that a flexibility score has been raised. The value lies in making neglected options visible enough to examine.
From task scores to real work
Erwin, Tran, and Koutstaal studied several divergent-thinking tasks alongside originality in a design-product ideation context. Such studies test whether task performance relates to a more applied outcome, a question that cannot be answered by the task itself. Their design context is still specific: university design work has its own brief, participants, raters, and criteria. A relationship in that setting should not be carried unexamined to scientific discovery, school assessment, or hiring.
A team trying to design an accessible service needs far more than a list of options. It needs to understand users, legal and practical constraints, costs, and the effect of a choice on different people. Divergent thinking can help prevent early closure, especially when the first proposal reflects a narrow assumption. Later work must test candidates against evidence. A process that rewards novelty while ignoring affected people can generate harm as readily as innovation.
The same distinction applies to individual work. A writer may generate five structures for an essay, then use disciplinary knowledge to see which structure supports the argument. A clinician may consider several explanations for a symptom, then narrow them using history, examination, and appropriate tests. In both cases, multiplying possibilities is useful only as part of a responsible decision process. The professional standard is not “always diverge”; it is knowing when to open the question and when to evaluate it.
Groups bring another complication. The number of ideas spoken aloud is not the number of ideas people had. Status, turn-taking, language confidence, and fear of judgment can alter who contributes. A facilitator may collect ideas individually first, then compare them, so quieter members have a route into the discussion. Whether that procedure improves the final outcome depends on the setting and should be evaluated. It is a plausible design response to participation barriers, not a universal formula.
What the evidence does not show
Divergent-thinking tasks do not provide a complete creativity score. Runco and Acar treat them as indicators of potential, with predictive validity that depends on the criterion. Hass’s analysis likewise warns against treating a task summary as a full cognitive account. A person who performs well on alternate uses has shown something under those test conditions, not demonstrated future artistic, scientific, or organizational achievement.
The evidence does not establish that fluency, originality, and flexibility are always separable traits. Scoring formulas can make the dimensions statistically dependent, and Weiss and Wilhelm found limited distinct value for their flexibility indicators after accounting for other factors. A study that reports three labels without explaining its prompts, coding, and relationships among scores leaves readers unable to judge what was measured.
Nor does the research prove that a brief exercise, an online game, or a single brainstorming session permanently increases creative capacity. The effect of an activity on one test can reflect familiarity with the format or a changed strategy. Transfer to other tasks, durable change, and valuable real-world products require separate evidence. This page offers a way to interpret and use the construct, not a training prescription.
Finally, these studies do not justify high-stakes judgments about a person’s intelligence, worth, or employability from one alternate-uses task. Performance is shaped by instructions, culture, language, time, domain knowledge, and scoring. Even a well-designed measure is an estimate with uncertainty. It should be used for the question it was validated to answer, with consequences proportional to that evidence.
An Institute lens for possibility and judgment
The Institute’s practical synthesis is to treat divergent thinking as the deliberate opening of an option space, followed by careful evaluation. In a workshop, that might mean stating the problem clearly, generating different kinds of responses, asking whose perspective is missing, and only then applying constraints. The sequence is a working framework, not a validated four-stage test. Its purpose is to keep a team from mistaking the first plausible answer for the only one.
Return to the empty shop. A repair café, rehearsal room, shared kitchen, and study space are not equivalent. Each serves different people, has different costs, and would need different permission and support. Divergent thinking helped the group see that “a community space” was not one fixed idea. The next task is to learn from residents and test what the site can actually sustain. The strength of divergence lies in making that later judgment more informed, not in replacing it.
Research references
- Runco, M. A., & Acar, S. (2012). Divergent thinking as an indicator of creative potential. Creativity Research Journal, 24(1), 66–75. View source
- Silvia, P. J., Winterstein, B. P., Willse, J. T., Barona, C. M., Cram, J. T., Hess, K. I., Martinez, J. L., & Richard, C. A. (2008). Assessing creativity with divergent thinking tasks: Exploring the reliability and validity of new subjective scoring methods. Psychology of Aesthetics, Creativity, and the Arts, 2(2), 68–85. View source
- Hass, R. W. (2017). Tracking the dynamics of divergent thinking via semantic distance: Analytic methods and theoretical implications. Memory & Cognition, 45(2), 233–244. View source
- Weiss, S., & Wilhelm, O. (2022). Is flexibility more than fluency and originality? Journal of Intelligence, 10(4), 96. View source
- Erwin, A. K., Tran, K., & Koutstaal, W. (2022). Evaluating the predictive validity of four divergent thinking tasks for the originality of design product ideation. PLOS ONE, 17(3), e0265116. View source