Divergent Thinking vs. Convergent Thinking

The human mind navigates reality through two distinct yet deeply intertwined modes of thought. When confronted with an ill-defined problem, an open horizon, or a conceptual void, the intellect must break through the boundaries of established categories to generate a wide array of untethered possibilities. Conversely, when faced with an array of competing alternatives, strict operational constraints, or formal empirical problems, the intellect must reverse its vector, systematically eliminating errors, synthesizing evidence, and focusing on the single, verifiable truth. These two complementary dynamics are known in cognitive science as divergent thinking and convergent thinking.

Historically polarized in both popular culture and pedagogical theory, these cognitive trajectories have often been reduced to simplistic archetypes. Divergent thinking is frequently celebrated as the uninhibited wellspring of artistic inspiration and revolutionary disruption, while convergent thinking is relegated to the dry, procedural mechanics of standardized testing, formal logic, and bureaucratic calculation. Such dichotomies fundamentally mischaracterize the architecture of human cognition. Groundbreaking scientific insights do not emerge solely from an unstructured burst of generative fancy, nor are masterpieces of literature composed through unchecked spontaneity without rigorous editing and structural discipline.

Instead, genuine intellectual achievement and sustainable innovation demand an ongoing, highly calibrated dialectic between both modes of thought. Understanding the precise operational mechanics, neurobiological substrates, psychometric evaluations, and systemic interactions of divergent and convergent thinking is not merely an academic exercise. It is an urgent practical necessity for researchers, educators, organizational leaders, and knowledge workers navigating an increasingly complex, volatile, and technologically augmented world. This treatise offers an exhaustive analysis of both cognitive modes, tracing their origins, mapping their neurological pathways, contrasting their functional architectures, and articulating how they unite to drive intellectual synthesis.

1. Introduction to Cognitive Processing Modes

1.1 Defining Divergent and Convergent Paradigms

The formal taxonomy distinguishing divergent and convergent operations within human cognition traces its lineage directly to the mid-twentieth century, crystallized through the pioneering psychometric architecture of Joy Paul Guilford. In his landmark 1950 presidential address to the American Psychological Association and subsequent formalizations of the Structure of Intellect (SOI) model in 1956, 1959, and 1967, Guilford recognized that contemporary intelligence metrics were fundamentally incomplete. The prevailing psychometric paradigms of his era, dominated by Spearman’s concept of general intelligence (g) and operationalized via standard IQ tests, evaluated human intellectual capacity almost exclusively through a singular lens: the subject’s ability to converge upon a single, predetermined, logically necessary correct response.

Guilford argued that human cognition encompasses far broader operations, establishing a foundational distinction between “divergent production” and “convergent production.” Within Guilford’s theoretical framework, divergent thinking is defined as the cognitive capacity to generate an abundance of varied, multifaceted, and novel ideas or responses to an open-ended stimulus. It is a process that moves outward from a single point of origin into uncharted conceptual territory, thriving in ambiguous problem spaces where no singular, predetermined solution exists. Rather than seeking an absolute end, divergent processing measures ideational fertility through the production of cognitive variance.

Conversely, convergent thinking is defined as the deductive, inductive, and analytical integration of disparate pieces of information, aimed at identifying the singular, verified, or mathematically optimal solution to a well-structured problem. Convergent operations rely on the systematic application of logical rules, domain-specific heuristics, empirical constraints, and categorization frameworks. Where divergent thinking radiates outward toward multiplicity, convergent thinking focuses inward toward unity, coherence, and verification. In contemporary cognitive psychology, these paradigms are recognized not merely as testing parameters, but as foundational, universal modes of human information processing.

1.2 Why Divergent and Convergent Modes Are Frequently Conflated and Contrasted

Despite Guilford’s balanced theoretical formulation, the cultural and academic reception of these concepts experienced a pronounced bifurcation throughout the late twentieth century. A widespread misconception emerged in educational psychology and organizational management that equated divergent thinking entirely with “creativity” in its entirety. Under this simplified framework, divergent thinking was romanticized as an innate, mystical artistic intuition, while convergent thinking was branded as cold, analytical, and fundamentally uncreative logic. This false dichotomy established an adversarial conceptual relationship between two modes that are, in biological reality, fundamentally cooperative.

This misunderstanding is compounded by the neurological complexity of spontaneous human problem-solving. When an individual engages in high-level intellectual labor, the human brain does not segregate these processing modes into isolated hemispheres. The persistent popular myth of the “creative right brain” versus the “logical left brain” has been thoroughly refuted by contemporary functional neuroimaging. During real-world problem-solving, cognitive processing shifts fluidly across widely distributed neural substrates, obscuring the observable boundaries between divergent generation and convergent filtering. The spontaneous nature of insight—the sudden “eureka” moment—often blinds the conscious thinker to the intense, iterative convergent labor that preceded and succeeded the divergent breakthrough.

Furthermore, this conflation has been exacerbated by systemic institutional biases within global educational architectures. Modern industrial education systems, designed around scalable assessment metrics, have historically privileged convergent evaluation. Standardized examinations, multiple-choice architectures, and algorithmically scoreable curricula reward the rapid, error-free deduction of singular answers while actively penalizing divergent ideation, non-linear interpretations, and exploratory ambiguity. As a consequence, generations of learners have been trained to associate intellectual legitimacy strictly with convergent execution, casting divergent capacity into the margins of non-academic or purely artistic pursuits. This pedagogical bias creates profound cognitive vulnerabilities when individuals encounter real-world problems characterized by radical uncertainty and missing premises.

1.3 The Strategic Imperative in Innovation and Epistemology

In modern epistemology, organizational dynamics, and technological design, mastering the relationship between divergent and convergent thinking has evolved from a theoretical interest into a vital strategic asset. Complex, adaptive challenges—ranging from global macroeconomic volatility and climate engineering to synthetic biology and artificial general intelligence—cannot be resolved through purely convergent optimization of legacy systems, nor through unconstrained, unverified divergent brainstorming. Systemic progress requires structured methodologies that consciously alternate between both modes of cognition.

This imperative is embodied in foundational innovation frameworks, such as the Double Diamond design model developed by the British Design Council in 2004. The Double Diamond maps problem-solving into four distinct, sequential phases arranged across two continuous cycles of expansion and contraction: Discover (divergent exploration of context), Define (convergent focus on the root problem), Develop (divergent ideation of potential solutions), and Deliver (convergent prototyping, testing, and operational deployment). Without deliberate divergent phases, organizations risk executing the wrong solutions with exceptional analytical efficiency. Without disciplined convergent phases, organizations collapse into ideational chaos, squandering capital on ungrounded concepts that never reach operational maturity.

Consequently, high-level leadership and scientific governance now demand metacognitive awareness as a baseline operational competency. Leaders and researchers must possess the deliberate capacity to evaluate their own cognitive states, recognize when an intellectual inquiry is stalled due to premature convergence or unchecked divergence, and actively switch cognitive modes. Epistemologically, genuine breakthroughs occur at the dynamic frontier where expansive exploration is immediately held accountable to rigorous, unsparing verification.

2. Understanding Divergent Thinking: Foundations and Dynamics

2.1 Conceptual Architecture and Core Characteristics

Divergent thinking operates on a conceptual architecture characterized by horizontal exploration, semantic drift, and the active suspension of cognitive closure. While traditional logic funnels incoming data toward a definitive conclusion, divergent cognition radiates outward, seeking out remote semantic associations and unconventional connections across distant memory structures. To operationalize and evaluate this expansive cognitive mode, Guilford (1959) and subsequently Ellis Paul Torrance (1966) codified the four foundational psychometric pillars of divergent performance:

• Fluency: The sheer quantitative volume of non-redundant, task-appropriate ideas, associations, or solutions generated within a specified temporal window. Fluency serves as the baseline energetic indicator of ideational output.
• Flexibility: The number of distinct conceptual categories, domains, or structural shifts represented across the generated responses. High flexibility indicates a cognitive system capable of abandoning habitual pathways and jumping across semantic boundaries.
• Originality: The statistical infrequency, conceptual novelty, and qualitative distinctiveness of the responses relative to a normative population baseline. Originality captures the departure from the obvious, prepotent, and conventional.
• Elaboration: The depth, intricacy, structural detail, and contextual richness with which an idea is articulated, embellished, and mentally constructed beyond its raw core.

Mechanistically, divergent thinking relies on non-linear cognitive trajectories that actively bypass immediate semantic constraints. When exposed to a stimulus word, object, or situational problem, the human mind instinctively activates its most accessible, frequently reinforced semantic associations—a phenomenon known as the primary associative hierarchy. Divergent processing deliberately inhibits these immediate cognitive reflexes, allowing spreading activation to propagate through the deeper, more diffuse networks of semantic memory. This intentional delay of cognitive closure creates a mental environment where distant, low-probability conceptual connections can emerge.

Crucial to this process is the deliberate suspension of critical evaluation. In standard cognition, an internal editor continuously evaluates early ideas against real-world constraints, immediately discarding unusual proposals as illogical or impractical. Divergent processing temporarily silences this internal critic. By intentionally decoupling idea generation from quality control, the cognitive system increases psychological safety and widens its exploratory field. This prevents premature optimization and creates space for novel, structurally ambitious ideas that can later be refined through convergent analysis.

2.2 Neurobiological Mechanisms and Cognitive Substrates

Modern cognitive neuroscience has made significant strides in uncovering the neural architecture that supports divergent thinking, dismantling the old belief that creative insight is purely random or anatomically localized. Functional magnetic resonance imaging (fMRI) and electroencephalography (EEG) reveal that high-level divergent ideation depends on dynamic, coordinated interactions between two large-scale brain networks: the Default Mode Network (DMN) and the Executive Control Network (ECN), mediated by the Salience Network (SN).

The Default Mode Network—comprising the medial prefrontal cortex (mPFC), posterior cingulate cortex (PCC), precuneus, and inferior parietal lobule—serves as the primary engine of spontaneous ideation and associative retrieval. Historically recognized for its activation during task-negative states, mind-wandering, episodic memory recollection, and prospective self-projection, the DMN facilitates the uninhibited, spontaneous flow of thoughts. During divergent ideation, the DMN activates robustly, driving internal semantic search across distant associative networks without the immediate suppression of unconventional connections.

However, pure, unconstrained DMN activation results in daydreaming rather than productive innovation. Groundbreaking neuroimaging investigations by Beaty et al. (2014, 2016, 2018) demonstrated that creative divergent capacity is characterized by a unique, paradoxical co-activation and functional connectivity between the DMN and the Executive Control Network (ECN, anchored in the dorsolateral prefrontal cortex and posterior parietal cortex). While the ECN typically suppresses DMN activity during demanding analytical tasks, divergent thinking requires the ECN to temporarily loosen its top-down control. It allows the DMN to surface unusual associative combinations, intervening strategically to guide the search, evaluate ideas against broad creative goals, and suppress mundane, repetitive associations (Benedek et al., 2014; Silvia, 2015).

At the neurochemical level, this cognitive agility is heavily regulated by dopaminergic pathways projecting through the mesolimbic and mesocortical systems. Optimal divergent performance correlates with intermediate-to-high dopaminergic tone, which modulates cognitive flexibility by decreasing latent inhibition. Latent inhibition is the unconscious cognitive mechanism that filters out irrelevant environmental stimuli and peripheral semantic nodes. By lowering this filter, dopamine allows a wider array of internal associations and external cues to enter working memory, broadening the semantic search horizon and increasing the probability of unexpected, original associations.

2.3 Measurement and Psychometric Assessment

Quantifying an inherently open-ended, non-deterministic cognitive operation presents significant psychometric challenges. Standardized assessment instruments must balance subjective qualitative appraisals of creativity with objective, statistically validated measurement protocols. The gold standard for psychometric assessment of divergent capacity remains the Torrance Tests of Creative Thinking (TTCT), first formalized in 1966 and refined across multiple subsequent iterations.

The TTCT is bifurcated into two distinct batteries: the TTCT-Verbal and the TTCT-Figural. The verbal battery confronts participants with tasks such as “Ask-and-Guess” (interrogating a depicted scene and predicting antecedents and consequences), “Product Improvement” (suggesting unusual modifications to transform a mundane toy into an engaging one), and “Unusual Uses.” The figural battery, designed to evaluate non-linguistic ideation across spatial dimensions, challenges individuals with “Picture Construction” (using a basic curved or geometric shape as the foundation for an elaborate image), “Incomplete Figures” (completing fragmented line drawings in unexpected configurations), and “Repeated Parallel Lines” (generating distinct, meaningful visual artifacts from uniform, repetitive geometric anchors). Standardized scoring protocols evaluate responses across the established metrics of Fluency, Flexibility, Originality, and Elaboration, alongside affective and stylistic indicators like abstract title usage and emotional expressiveness.

Parallel to the TTCT, Guilford’s Alternative Uses Task (AUT) serves as the most widely implemented experimental paradigm in modern cognitive laboratories. In the AUT, participants are given a common everyday object—such as a brick, a paperclip, or an automobile tire—and granted a constrained duration (typically between two and four minutes) to articulate as many non-traditional, functionally viable uses as possible. Responses such as “building a wall” for a brick are categorized as prepotent, zero-originality baseline associations. Conversely, suggestions such as “carving it into a printing stamp,” “grinding it into pigment for paint,” or “using it as a counterweight in an improvised water filtration system” reflect conceptual category shifts, demonstrating high flexibility and statistical originality.

In recent years, the psychometric evaluation of divergent thinking has been modernized through computational linguistics and natural language processing. Rather than relying entirely on subjective human raters, researchers use distributional semantic models to compute objective “semantic distance.” By analyzing vast corpora of human text, algorithms map words into high-dimensional vector spaces. Computational metrics then calculate the cosine distance between the target object and the participant’s proposed responses. This approach provides an automated, bias-free, continuous scale of semantic novelty, transforming the empirical measurement of ideational originality.

2.4 Strengths, Cognitive Biases, and Operational Limitations

The primary cognitive asset of divergent thinking is its unmatched ability to dismantle functional fixedness and break through entrenched mental sets. Functional fixedness, a cognitive bias first documented by Karl Duncker in 1945, is the tendency to perceive an object, concept, or system solely in terms of its customary, traditional utility. Divergent thinking systematically deconstructs these fixed conceptual categories, enabling researchers and problem-solvers to look past standard operational definitions. It allows them to uncover unexpected, non-obvious configurations that can resolve persistent conceptual impasses.

Furthermore, divergent thinking serves as a vital safeguard against premature cognitive closure—the psychological drive to quickly settle on a clear, definitive answer to relieve the ambiguity of uncertainty. In complex, dynamic environments where initial problem definitions are often incomplete or flawed, an early rush to closure can lead to costly solutions addressing the wrong core issues. Divergent processing encourages a sustained exploration of the problem space, uncovering latent variables, alternate perspectives, and systemic interdependencies that analytical deduction might otherwise overlook.

However, divergent thinking also carries distinct cognitive vulnerabilities and operational risks. When uncoupled from analytical reality, it is susceptible to unchecked ideational proliferation. The generative mind can become trapped in a loop of speculative conceptual production, crafting ideas that are statistically original but structurally unfeasible, economically absurd, or physically impossible. Novelty, when pursued purely for its own sake, can produce bizarre and impractical dead ends that consume valuable time and resources.

Additionally, sustaining an open, ambiguous divergent state requires considerable mental effort. Prolonged divergence creates cognitive fatigue, as the mind must continuously resist the natural human drive for structure, resolution, and closure. Without deliberate transitions into convergent evaluation, prolonged divergence can lead to paralysis by analysis, decision fatigue, and systemic disorganization. Unbounded ideation without execution remains an incomplete intellectual journey.

3. Understanding Convergent Thinking: Foundations and Dynamics

3.1 Conceptual Architecture and Core Characteristics

Convergent thinking represents the intellect’s capacity for analytical synthesis, logical integration, and systematic reduction. If divergent thinking is conceptualized as an expanding wave of possibilities, convergent thinking is the focused lens that gathers scattered rays of data and concentrates them into a precise, coherent point. Its primary goal is epistemic validation: taking a diverse set of observations, data points, or generated ideas, and filtering them down to identify the single, optimal, or logically necessary solution.

The operational framework of convergent processing is built on formal deductive, inductive, and abductive logic. It evaluates proposed hypotheses against established empirical facts, scientific laws, mathematical proofs, or domain constraints. Convergent thinking prioritizes accuracy, speed, structural coherence, and methodological rigor. It approaches problem-solving methodically, eliminating unviable branches on a decision tree and testing each potential answer against reality until only the most resilient and defensible option remains.

Crucially, convergent thinking does not operate in an intellectual vacuum; it relies heavily on established knowledge, rule systems, and domain-specific heuristics. When an engineer calculates the structural integrity of a suspension bridge, or an epidemiologist tracks the transmission vector of a pathogen, they do not invent a new physics or biology on the spot. Instead, they apply proven principles, historical data, and computational models to filter out unfeasible hypotheses. Convergent thinking provides the structural stability and intellectual discipline that allows conceptual insights to be converted into functional, real-world solutions.

3.2 Neurobiological Mechanisms and Cognitive Substrates

The neurobiological architecture of convergent thinking is characterized by intense, focused activation of the frontoparietal Central Executive Network (CEN), accompanied by the deliberate downregulation of the Default Mode Network. The brain’s main objective during convergent tasks is to filter out peripheral distractions, suppress irrelevant semantic associations, and direct mental energy toward evaluating specific constraints.

At the center of this focused network is the dorsolateral prefrontal cortex (dlPFC), working closely with the posterior parietal cortex (PPC). The dlPFC maintains working memory capacity, coordinates rule-based logical sequences, and exercises cognitive control over immediate impulses. Simultaneously, the dorsal anterior cingulate cortex (dACC) monitors performance, detects cognitive conflicts, and tracks errors. When an individual works through a complex logical syllogism, mathematical calculation, or convergent deduction, the dACC continuously evaluates progress against the target goal, signaling the dlPFC to adjust attention whenever errors or inconsistencies emerge.

A critical component of successful convergent thinking is the active inhibition of semantic interference, primarily orchestrated by the inferior frontal gyrus (IFG) and the anterior cingulate. Unlike divergent thinking, which widens associative horizons, convergent processing requires a high-threshold semantic filter. It must actively suppress irrelevant associations, metaphors, and tangential ideas that could derail the linear pursuit of a solution. This inhibitory mechanism ensures that working memory is reserved exclusively for task-relevant data and formal logical operations.

Neurochemically, convergent processing is heavily modulated by the noradrenergic system originating from the locus coeruleus, operating in harmony with stable, baseline prefrontal dopamine. The release of noradrenaline sharpens attentional focus, increasing the signal-to-noise ratio in cortical processing networks. This biological state suppresses ambient cognitive noise, stabilizes attention, and supports the sustained mental effort required to work through complex, structured, and rules-based challenges.

3.3 Measurement and Psychometric Assessment

Because convergent thinking pursues definitive, verifiable answers, its psychometric assessment is deeply integrated into classical intelligence and cognitive aptitude testing. The operational parameters of these assessments prioritize objectivity, statistical reliability, and clear-cut, binary scoring rubrics where answers are either demonstrably correct or fundamentally incorrect.

The most prominent instruments evaluating pure convergent analytical intelligence include the Wechsler Adult Intelligence Scale (WAIS) and Raven’s Progressive Matrices. In the WAIS battery, subtests such as Matrix Reasoning, Figure Weights, and Arithmetic assess an individual’s ability to extract underlying abstract rules from visual or numerical sequences, synthesize mathematical constraints, and deduce the single missing variable. Raven’s Progressive Matrices presents a completely non-verbal evaluation of abstract reasoning: participants analyze a two-dimensional grid of geometric patterns, infer the underlying visual logic, and select the only viable shape that correctly completes the pattern.

In creativity research, the most widely used psychometric tool for convergent processing is Sarnoff Mednick’s Remote Associates Test (RAT), introduced in 1962. The RAT occupies an intriguing position at the intersection of associative memory and convergent deduction. In a standard RAT item, participants are presented with three seemingly unrelated words (for example: “cream,” “skate,” and “water”) and tasked with discovering a single, fourth unifying word that forms a valid compound word or common phrase with all three (in this instance: “ice” — creating ice cream, ice skate, and ice water).

Successfully solving a RAT problem requires a distinctive two-stage cognitive motion. The mind must first cast outward to retrieve associative networks linked to each of the three prompt words (a divergent sweep). It must then rapidly cross-reference, filter, and converge upon the single linguistic node that satisfies all three constraints simultaneously. If an individual generates numerous creative associations for “cream” (such as “coffee,” “cheese,” or “pastry”) but cannot converge on the specific item that overlaps with both “skate” and “water,” they fail the task. The RAT thus serves as a precise psychometric measure of convergent integration, testing an individual’s ability to locate a mutually constrained solution within a complex semantic space.

3.4 Strengths, Cognitive Biases, and Operational Limitations

The primary advantage of convergent thinking is its exceptional execution efficiency, analytical precision, and systematic verification. Without convergent discipline, societies could not construct reliable bridges, design bug-free software, verify pharmacological safety, or establish consistent legal frameworks. Convergent thinking provides the rigorous filtering mechanisms that separate actionable, empirically sound solutions from wishful thinking and chaotic speculation.

Moreover, convergent thinking excels at optimizing well-defined, existing systems. When the boundaries, rules, and fundamental parameters of a domain are already established, convergent analysis methodically identifies inefficiencies, reduces waste, surfaces logical contradictions, and fine-tunes performance. It is the cognitive engine of technical mastery, iterative refinement, and scientific reproducibility.

However, this reliance on established frameworks also produces significant cognitive vulnerabilities. Convergent thinking is inherently susceptible to confirmation bias—the tendency to search for, interpret, and prioritize data that reinforces pre-existing assumptions, while dismissing disconfirming evidence. Because convergent cognition operates within established rules and heuristics, it often treats those rules as absolute, unalterable constraints rather than provisional models of reality.

This dynamic frequently leads to premature closure, where problem-solvers anchor to the first logically viable solution they discover, ignoring broader, non-linear alternatives that could be far more effective. In volatile, rapidly shifting environments undergoing paradigm shifts (what Thomas Kuhn described as scientific revolutions), convergent thinking can become a liability. Trapped within legacy axioms, an overly convergent intellect continues optimizing processes that have become obsolete, unable to imagine alternative frameworks that render old rules irrelevant.

4. Key Differences Between Divergent and Convergent Thinking

4.1 Teleological and Epistemic Goals

The philosophical and functional divergence between these two cognitive modes begins with their teleological architecture—their fundamental purpose and intended endpoint. Divergent thinking is teleologically expansive and horizontal. Its ultimate objective is the generation of cognitive variance, the exploration of uncharted conceptual spaces, and the discovery of novel perspectives. Epistemologically, divergent thinking embraces probabilistic pluralism: it accepts that multiple, conflicting hypotheses can coexist as valuable possibilities throughout the early stages of inquiry. Its value structure is driven by novelty, conceptual range, and structural originality.

In contrast, convergent thinking is teleologically reductive and vertical. Its ultimate objective is epistemic validation, accuracy, and the elimination of ambiguity. Convergent processing takes a field of options and narrows it down until only a unified, defensible, and empirically resilient conclusion remains. It is grounded in a commitment to verification, coherence, and actionable truth. Where divergent thinking asks, “What else could this be, and in how many different directions can we look?”, convergent thinking demands, “What is the single most accurate, feasible, and optimal path forward based on our constraints?” The table below outlines these core distinctions across key cognitive dimensions.

Cognitive Dimension	Divergent Thinking	Convergent Thinking
Primary Epistemic Objective	Generation of variance, novelty, and conceptual possibilities; expanding the problem space.	Deductive integration, verification, and discovery of the single optimal solution; reducing the problem space.
Processing Trajectory	Horizontal, radiating outward, associative, non-linear, and exploratory.	Vertical, linear, analytical, deductive, focused, and hierarchical.
Dominant Neural Networks	Default Mode Network (DMN) coupled dynamically with selective, flexible Central Executive Network (CEN) engagement.	Central Executive Network (CEN/ECN), frontoparietal control systems, and dorsal anterior cingulate cortex (dACC).
Attentional Filter Threshold	Low-threshold, diffuse sensory/semantic gating; reduced latent inhibition to welcome peripheral stimuli.	High-threshold, highly focused selective filtering; active suppression of semantic and environmental interference.
Tolerance of Ambiguity	Exceptionally high; deliberate suspension of judgment and comfort with unresolved cognitive tension.	Low-to-moderate; strong psychological drive toward epistemic closure, resolution, and structural clarity.
Primary Psychometric Tools	Torrance Tests of Creative Thinking (TTCT), Alternative Uses Task (AUT), semantic distance analysis.	Wechsler Adult Intelligence Scale (WAIS), Raven’s Progressive Matrices, Remote Associates Test (RAT).
Characteristic Cognitive Biases	Ideational proliferation, impracticality, over-valuing raw novelty, and resistance to operational constraints.	Confirmation bias, premature closure, anchoring effects, and susceptibility to functional fixedness.
Evaluative Stance	Deferred critique; non-judgmental acceptance to maximize psychological exploration.	Continuous, real-time critical appraisal; rigorous error checking and systematic boundary enforcement.
Underlying Neurochemistry	Modulated by flexible mesolimbic dopamine tone, promoting associative fluidity.	Modulated by noradrenaline from the locus coeruleus and prefrontal dopamine, reinforcing signal-to-noise clarity.
Environmental Preconditions	Psychological safety, open temporal horizons, broad multidisciplinary stimuli, and minimal acute stress.	Clear problem definitions, explicit metrics, rigorous structural parameters, and well-organized information.

4.2 Cognitive Mechanics and Information Flow

The mechanical flow of information reveals sharp differences between these two cognitive strategies. In divergent thinking, information propagation is driven by spreading activation throughout the semantic network. In this framework, concepts are stored as nodes within an expansive, multidimensional web. When a specific node is activated, that activation ripples outward, exciting adjacent concepts.

In a divergent state, this threshold of activation is deliberately lowered, allowing the signal to travel across remote semantic connections. For instance, the node “tree” might rapidly spread beyond “leaf” or “wood” to activate distant concepts like “genealogy,” “fractals,” “respiratory systems,” or “hierarchical data structures.” Divergent thinking embraces these distant associations, deliberately integrating anomalies and conceptual contradictions rather than filtering them out.

Conversely, the information flow of convergent thinking resembles a funneling mechanism governed by strict cognitive control. When a concept node is activated during a convergent task, the central executive network instantly evaluates its relevance against the task’s rules and constraints. Remote associations that lack logical utility are actively suppressed via top-down inhibitory control.

If an engineer is calculating the load-bearing requirements of a beam, metaphorical connections to branching vascular networks are suppressed as irrelevant cognitive noise. The convergent funnel accepts only data that conforms to explicit physical laws and mathematical properties. Contradictions are not embraced as creative sparks; they are flagged as operational errors that must be resolved, balanced, or discarded to maintain systemic coherence.

4.3 Psychological Mindset and Affective States

Divergent and convergent thinking foster distinctly different psychological mindsets and affective experiences. Divergent exploration requires a high tolerance for ambiguity and uncertainty. To inhabit a divergent state, a thinker must feel comfortable not knowing where the process will lead, resisting the natural psychological urge to seize on an early answer to relieve mental tension. This demands a mindset of epistemic curiosity and risk tolerance. It treats errors, unexpected failures, and bizarre connections not as mistakes, but as informative signals that expand the conceptual map.

Convergent thinking, on the other hand, is driven by a desire for resolution, clarity, and structural order. Its psychological state is defined by disciplined focus, analytical skepticism, and risk management. The convergent mindset views errors not as exploratory play, but as structural vulnerabilities that must be identified and corrected.

This approach values precision and accountability, prioritizing reliability and real-world viability over raw novelty. When individuals switch between these two modes, they often experience an internal shift: moving from the exploratory freedom and open horizons of divergent ideation to the focused, critical accountability of convergent execution.

5. Synergies and Overlapping Cognitive Dimensions

5.1 The Dialectical Interdependence of Innovation

While isolating divergent and convergent thinking into distinct operational categories is valuable for psychometric study and experimental analysis, treating them as mutually exclusive silos is an intellectual error. Groundbreaking discoveries, artistic masterpieces, and foundational technological advances do not emerge from divergence or convergence in isolation. Rather, they are the direct product of a sustained dialectic—a rhythmic, continuous cycle moving between these two cognitive engines.

Divergent thinking without convergent evaluation inevitably descends into chaotic, unanchored speculation. The history of innovation is filled with generative thinkers who continuously produced eccentric, novel concepts, yet failed to leave a lasting impact because they refused the disciplined, analytical labor required to evaluate, ground, and test their ideas against real-world constraints. Unchecked divergence produces endless sketches, unfinished manuscripts, and wild prototypes that collapse under the demands of physical execution or logical scrutiny.

Conversely, convergent thinking without divergent input leads directly to sterile, incremental stagnation. A cognitive system or organization operating purely through convergent analysis excels at optimizing existing models, eliminating inefficiencies, and maintaining rigorous quality control. However, it cannot generate the conceptual leaps required to rewrite obsolete rules. It continues calculating the absolute best path forward along a road that may no longer be relevant.

This creative dynamic is articulated theoretically in the Geneplore model of creativity, established by Ronald Finke, Thomas Ward, and Steven Smith in 1992. The Geneplore framework illustrates that creative cognition operates via a continuous, cyclical feedback loop comprised of two major phases: the Generative phase and the Exploratory phase. In the generative phase, divergent mechanisms construct crude, nascent, novel mental representations termed “pre-inventive structures.” Once generated, these structures immediately enter the exploratory phase, where convergent mechanisms systematically analyze their utility, test their physical and logical properties, and eliminate internal contradictions.

If the pre-inventive structure proves flawed, it is returned to the generative phase for divergent modification. If it meets the required constraints, it is refined, finalized, and executed. Innovation is fundamentally this iterative dance: expand, filter, iterate, and finalize.

This dynamic feedback loop is reinforced by the Dual-Pathway Model of Creative Ideation proposed by Carsten De Dreu, Bernard Nijstad, and Matthijs Baas (2010). Their empirical research proves that exceptional, innovative ideation is systematically achieved through two distinct, yet equally viable pathways: the cognitive flexibility pathway and the cognitive persistence pathway. The flexibility pathway operates through divergent, broad associative searches across many distinct conceptual categories. The persistence pathway operates through convergent, deeply focused, and sustained analytical labor within a narrow set of categories. Breakthrough thinking occurs when these two pathways work in tandem—either across collaborative, multidisciplinary teams or within a single, metacognitively agile mind capable of fluidly cycling between expansive associative exploration and laser-focused analytical persistence.

5.2 Shared Cognitive Substrates and Memory Systems

The operational divide between divergent and convergent processing obscures their shared dependence on foundational cognitive architectures and memory systems. Neither mode creates information out of nothing; both draw from the same deep well of long-term semantic and episodic memory networks.

To generate an original metaphor for a complex scientific problem (a divergent act), or to construct a rigorous formal proof disproving a mathematical conjecture (a convergent act), an individual requires an extensively structured, highly integrated domain-specific knowledge base. Deep knowledge is the essential raw material for both processing modes. The more sophisticated and organized an individual’s semantic memory architecture, the more effectively divergent thinking can trace remote connections, and the more rapidly convergent thinking can spot subtle errors and logical contradictions.

Furthermore, both processing modes are fundamentally constrained by working memory capacity (WMC) and the coordinating influence of the Salience Network. While divergent thinking was once assumed to require an entirely uninhibited, structure-free mental space, empirical research by Benedek et al. (2014) demonstrated that working memory capacity actively supports divergent performance. Sustaining multiple divergent branches in mind, keeping track of emerging associations, and intentionally avoiding repetitive, conventional answers requires significant cognitive energy and working memory control.

In this shared neurological ecosystem, the Salience Network—anchored in the anterior insula and the dorsal anterior cingulate cortex—serves as the brain’s internal conductor. The Salience Network detects crucial changes in the environment and internal cognitive states, dynamically arbitrating between the spontaneous, associative retrieval of the Default Mode Network and the focused, analytical discipline of the Central Executive Network. The two modes do not fight for control; they are systematically deployed by the brain’s regulatory networks to meet the shifting demands of the task at hand.

5.3 Integrative Thinking and Creative Synthesis

At the highest levels of human intellectual performance, the boundaries between divergent and convergent processing dissolve into unified states of integrative thinking and creative synthesis. This synthesis is clearly demonstrated in the psychological concept of “simultaneous constraint management.” True creative mastery rarely takes place in an environment of total, unconstrained freedom.

Instead, as seen in complex architectural design, spacecraft engineering, or the composition of a musical fugue, the thinker must generate deeply original forms (divergence) while simultaneously adhering to unforgiving physical, structural, or acoustic laws (convergence). The creative challenge is not to alternate between freedom and restriction, but to use the constraints themselves as the creative catalyst, generating novel variations that satisfy strict operational parameters.

This integrative synthesis was captured by Arthur Koestler in his concept of “bisociation,” detailed in his 1964 work The Act of Creation. Koestler argued that creative insight does not arise from simple, linear associative chains. Instead, it occurs when two previously independent, internally coherent, but completely divergent frames of reference are suddenly synthesized by a single convergent insight.

This concept was further expanded by Gilles Fauconnier and Mark Turner in their theory of conceptual blending. Integrative thinkers pull conceptual structures from two wildly different domains and project them into an entirely new, blended mental space. This third space inherits specific, logical rules from each source domain (convergence) while generating entirely unprecedented structural relationships (divergence).

Finally, this cognitive synthesis is evident in expert intuition. When a grandmaster makes an unexpected chess move, or an experienced physician accurately diagnoses a rare condition within moments of seeing a patient, their mind is not engaged in random guessing or painstaking, step-by-step deduction. Instead, their intuition represents compiled convergence operating over years of divergent immersion.

Through decades of rigorous practice, their brain has integrated thousands of disparate patterns and conceptual relationships into deep memory structures. In a fraction of a second, their mind scans this vast associative space (a lightning-fast divergent sweep) and immediately converges upon the single optimal decision. What appears to be an unexplainable spark of insight is actually the seamless, instantaneous integration of both processing modes operating at the peak of human capability.

6. Contextual Applications: Deployment Frameworks and Workflows

6.1 Optimal Domains and Conditions for Divergent Thinking

Recognizing the distinct strengths of divergent thinking allows organizations, researchers, and individuals to deploy it deliberately in domains where linear analytical deduction reliably hits an impasse. The natural home for divergent processing is in the early stages of greenfield research and development, open-ended scenario planning, and systemic problem formulation. When an industry faces rapid technological disruption or an academic field encounters an empirical contradiction that existing theories cannot explain, applying standard convergent logic to old assumptions simply reproduces the problem.

Divergent thinking is essential whenever an organization must break through functional fixedness and overcome cognitive impasses. It is particularly powerful in speculative futures research, blue-sky product ideation, exploratory qualitative science, and complex diplomatic negotiations where conventional binary choices lead to stalemate. In these environments, the fundamental problem definition itself is often incomplete or misleading. Divergent thinking allows teams to ask entirely new questions, reframe challenges from counter-intuitive angles, and explore multiple, competing problem definitions before committing resources to an analytical path.

To cultivate effective divergent thinking, specific environmental and cognitive conditions must be maintained:

• High Psychological Safety: Participants must feel absolute confidence that generating wild, unconventional, or provisionally absurd associations will not result in ridicule, social penalties, or career risk. If social anxiety is present, the executive control network automatically suppresses distant semantic retrieval in favor of safe, conventional ideas.
• Temporal Decoupling: Ideation sessions must be completely separated from immediate analytical evaluation and administrative oversight. The mind cannot engage in open associative drift if it is constantly interrupted by real-time feasibility audits.
• Broad Cross-Domain Exposure: Providing conceptual cross-training, interdisciplinary reading, and physical environments rich in diverse sensory and conceptual stimuli supplies the associative memory network with the raw material needed to forge unexpected connections.

6.2 Optimal Domains and Conditions for Convergent Thinking

Just as open-ended exploration requires divergent freedom, high-stakes operational execution, risk management, and formal scientific verification demand the unsparing rigor of convergent thinking. Once a problem has been clearly defined and candidate hypotheses have been framed, continued divergence becomes a dangerous distraction. Convergent processing is the non-negotiable operational standard in structural engineering, aviation safety analysis, regulatory compliance, pharmacological dosing, and critical software testing.

In a medical intensive care unit, an aerospace trajectory calculation, or a formal legal audit, ambiguous associations and novelty for its own sake are entirely out of place. The primary imperative in these domains is the relentless elimination of error, the mitigation of risk, and the deductive validation of safety-critical systems against unforgiving standards. Convergent thinking is equally essential during crisis management and emergency response operations. When a crisis strikes—such as an active cybersecurity breach or a major supply chain failure—open-ended deliberation can lead to catastrophe. Teams must rapidly analyze incoming real-world data, apply proven diagnostic heuristics, converge upon an unambiguous, decisive action plan, and execute it with operational discipline.

The optimal conditions for sustaining convergent processing require deliberate cognitive ergonomics:

• Strict Attentional Hygiene: Because convergent analytical thinking relies on working memory and prefrontal control, teams must be protected from constant digital pings, multitasking, and environmental interruptions that degrade cognitive focus.
• Clear Constraint Definitions: Convergent thinking thrives on precise, unambiguous parameters. Providing clear performance metrics, physical tolerances, budgetary realities, and logical bounds enables the mind to methodically eliminate unfeasible options.
• Transparent Evaluative Rubrics: Teams must understand the explicit criteria used to evaluate competing hypotheses. Clear guidelines ensure that choices are driven by objective evidence, logical consistency, and empirical efficacy, rather than subjective bias or office politics.

6.3 Operationalizing the Sequence in Teams and Systems

To institutionalize both processing modes without letting them undermine each other, organizations must implement structured methodologies that manage the sequence and timing of divergent and convergent workflows. The most common organizational failure is allowing these modes to collide in the same meeting. When leaders assemble a team for an “open brainstorming session” and then instantly evaluate, critique, or run cost-benefit analyses on every idea shared in real time, the process collapses. The analytical voice shames the exploratory voice, psychological safety evaporates, and the session retreats into safe, incremental suggestions.

To prevent this friction, high-performing innovation teams use strict temporal decoupling. They divide workflows into dedicated, chronologically separated phases where each cognitive mode is pursued exclusively, as illustrated in the following workflow sequence:

• Phase 1: Divergent Problem Reframing (Exploration): The team initiates the project by interrogating assumptions, interviewing diverse stakeholders, and brainstorming dozens of alternative interpretations of the underlying problem without judging their practicality.
• Phase 2: Convergent Problem Definition (Focus): The team switches gears, applying analytical rigor, empirical evidence, and strategic boundaries to synthesize their findings into a single, unambiguous design brief.
• Phase 3: Divergent Solution Ideation (Generation): Working from that unified brief, the team re-enters a divergent state. Using structured creative heuristics—such as the SCAMPER model (Substitute, Combine, Adapt, Modify, Put to another use, Eliminate, Reverse) or TRIZ theory—they generate a wide array of potential solutions, actively pushing for volume, flexibility, and original approaches.
• Phase 4: Convergent Evaluation and Prototyping (Execution): Ideation is formally closed. The team changes hats, applying quantitative decision tools such as a Pugh Matrix, failure mode and effects analysis (FMEA), and rigorous cost-feasibility rubrics to systematically rank, filter, and stress-test the concepts, advancing only the most viable candidates to physical prototyping.

This sequential approach is supported by established operational methodologies, such as Edward de Bono’s Six Thinking Hats. By assigning specific cognitive modes to distinct, deliberate phases—wearing the “Green Hat” for uninhibited, non-judgmental divergent ideation, and later donning the “Black Hat” for critical, unsparing convergent risk evaluation—teams can direct their collective cognitive energy without triggering interpersonal conflict. Managing cognitive diversity requires deliberate leadership: ensuring that generative thinkers are protected during exploratory phases, while analytical specialists are valued as the essential arbiters of structural integrity during execution.

7. Frequently Asked Questions

7.1 Can an Individual Excel at Both Divergent and Convergent Thinking?

Yes. Although individuals often lean toward one cognitive mode due to personality traits, educational training, or vocational habits, divergent and convergent thinking are not mutually exclusive capabilities. The belief that a person is permanently locked into being either an ungrounded creative or an analytical processor is a psychological myth. Empirical psychometrics consistently demonstrates that individuals who cultivate high cognitive flexibility can achieve advanced proficiency in both modes.

The key to excelling at both is metacognitive agility: the conscious ability to monitor one’s own cognitive state and deliberately switch mental strategies to match the task at hand. History’s most impactful intellectual figures—such as Leonardo da Vinci, Richard Feynman, and John von Neumann—exemplified this ambidextrous cognition. Feynman, for instance, could drift into deep, unconventional conceptual associations (divergent exploration) and then seamlessly transition into uncompromising, rigorous mathematical formalism (convergent validation). Ambidextrous thinking is a skill that can be developed through intentional practice, multidisciplinary study, and the cultivation of metacognitive awareness.

7.2 Is Divergent Thinking Truly Synonymous with Creativity?

No. Divergent thinking is a primary metric of creative potential, but it is not synonymous with creativity itself. This common conflation has caused widespread confusion in education and management. In modern cognitive psychology, creativity is formally defined by a two-criterion standard: to be judged creative, an output must be both novel (original, unexpected) and effective (useful, appropriate, structurally viable, or contextually meaningful).

Divergent thinking is exceptionally good at satisfying the first criterion: generating novelty, variance, and unconventional associations. However, it cannot satisfy the second criterion on its own. An idea that is entirely novel but completely unfeasible or incoherent is not creative; it is simply bizarre or chaotic. Convergent thinking is the mechanism that evaluates, refines, and shapes raw novelty into functional utility, ensuring the idea meets real-world constraints. Genuine creativity exists exclusively at the intersection where divergent ideation is brought to fruition by convergent discipline.

7.3 How Do Modern AI Systems Perform in Divergent Versus Convergent Thinking?

Modern artificial intelligence architectures, particularly Large Language Models (LLMs), emulate both divergent and convergent thinking through distinct computational mechanisms, though they operate differently from biological cognition. LLMs emulate divergent thinking through stochastic token prediction modulated by temperature and top-p sampling parameters. When the temperature is elevated, the model increases the probability of selecting lower-frequency, semantically distant tokens, generating unusual metaphors, varied scenarios, and novel text combinations that can closely mimic human divergent ideation.

Conversely, artificial intelligence achieves convergent thinking through techniques like Chain-of-Thought (CoT) prompting, automated theorem proving, symbolic logic integrations, and reinforcement learning from human feedback (RLHF). These methods guide the system to methodically work through logical steps, eliminate unviable reasoning branches, and converge on a single, mathematically verified answer. However, unlike the human brain—which uses the Salience Network to fluidly and autonomously balance the Default Mode and Executive networks in response to real-time anomalies—current AI systems lack true metacognition. They depend on human engineers and users to prompt, orchestrate, and guide the switch between divergent generation and convergent verification.

7.4 How Does Cognitive Stress Impact Divergent Versus Convergent Performance?

Acute psychological stress exerts a profound, asymmetric impact on human cognitive processing, primarily mediated by the systemic release of cortisol and catecholamines (adrenaline and noradrenaline). Under acute stress, evolutionary survival mechanisms prepare the brain for immediate threat response by increasing prefrontal noradrenaline and constricting the attentional spotlight. This physiological reaction sharply impairs divergent thinking. Stress diminishes latent inhibition, suppresses the Default Mode Network’s associative wanderings, and severely reduces the cognitive flexibility required to explore remote semantic networks. Under pressure, the mind defaults to safe, prepotent, and highly familiar responses, making genuine creative breakthroughs nearly impossible.

Convergent thinking, by contrast, shows a more complex, resilient response to moderate stress, often following an inverted-U curve described by the Yerkes-Dodson Law. Moderate autonomic arousal can actually enhance convergent performance on well-learned, structured, and analytical tasks by sharpening focus, increasing vigilance, and filtering out peripheral environmental distractions. However, if stress becomes chronic or overwhelming, prefrontal cortical control degrades entirely, impairing working memory and compromising both convergent logical deduction and divergent exploration. Establishing psychological safety is therefore not a soft organizational luxury, but a biological prerequisite for sustaining open, divergent problem-solving under pressure.

8. Conclusion: Synthesizing the Dual Engines of Human Intellect

8.1 Summary of the Core Dichotomy and Complementarity

Divergent thinking and convergent thinking represent the dual engines of human intellect, two complementary trajectories of cognitive computation. Divergent thinking is the engine of variance, unconstrained exploration, horizontal association, and possibility. It ventures into ambiguous conceptual territories, dismantling functional fixedness and surfacing unconventional combinations from the depths of semantic memory. Convergent thinking is the engine of validity, deductive discipline, analytical focus, and error correction. It gathers scattered observations and hypotheses, testing them against real-world constraints, physical laws, and logical requirements until only the most defensible solution remains.

Neither cognitive mode holds an inherent moral, intellectual, or operational superiority over the other. The historical tendency to elevate one while diminishing the other—whether by overvaluing unanchored artistic divergence or prioritizing rigid convergent testing—creates a false dichotomy that misinterprets the architecture of human thought. Divergence without convergence produces disorganized, ungrounded speculation that never materializes into functional value. Convergence without divergence produces sterile, incremental optimization trapped inside obsolete frameworks. Genuine intellectual breakthroughs emerge exclusively from the dynamic, ongoing equilibrium between them: the courage to explore the unknown, coupled with the discipline to hold every hypothesis accountable to reality.

8.2 The Trajectory of Cognitive Cultivation

As human civilization navigates an era marked by systemic complexity and the rapid automation of routine cognition, the deliberate cultivation of our thinking processes becomes an educational and professional imperative. Computational systems are rapidly mastering routine convergent tasks: running algorithms, analyzing vast datasets, and checking formal legal and mathematical proofs with super-human speed and reliability. Simultaneously, automated generative models can simulate broad divergent text generation at the click of a button. In this emerging technological landscape, the unique value of human intellect does not lie in acting as a slow calculator, nor as an ungrounded ideator.

Instead, the true hallmark of advanced human intellect is metacognitive agility: the conscious, intentional orchestration of the cognitive cycle. It is the ability to recognize when an intellectual inquiry requires stepping back to diverge, and when it requires stepping forward to converge. It means learning to welcome ambiguity without fear during exploratory phases, and then pivoting to apply unsparing analytical critique during execution. Educational institutions must move beyond their historical over-reliance on standardized convergent testing, redesigning curricula to intentionally nurture both modes through iterative, hands-on, problem-based learning.

Ultimately, cognitive agility is the definitive intellectual competency of our time. By understanding the neurobiological, psychometric, and operational mechanics of both divergent and convergent thinking, we can move past simplistic dichotomies and tap into the full, integrated capacity of the human mind. The future will not belong to the undisciplined dreamer, nor to the rigid technician, but to the ambidextrous thinker who can move fluidly between them—envisioning bold new worlds with divergent freedom, and bringing them safely into reality through convergent discipline.

References

• Beaty, R. E., Benedek, M., Silvia, P. J., & Schacter, D. L. (2016). Creative cognition and brain network dynamics. Trends in Cognitive Sciences, 20(2), 87–95. https://doi.org/10.1016/j.tics.2015.10.004
• Beaty, R. E., Kenett, Y. N., Christensen, A. P., Rosenberg, M. D., Cote, A. M., Bodick, M. B., Kane, M. J., & Silvia, P. J. (2018). Robust prediction of individual creative ability from brain functional connectivity. Proceedings of the National Academy of Sciences, 115(5), 1087–1092. https://doi.org/10.1073/pnas.1713532115
• Benedek, M., Jauk, E., Sommer, M., Arendasy, M., & Neubauer, A. C. (2014). Intelligence, creativity, and cognitive control: The common and differential involvement of executive functions in intelligence and creativity. Intelligence, 46, 73–83. https://doi.org/10.1016/j.intell.2014.05.007
• De Dreu, C. K., Baas, M., & Nijstad, B. A. (2008). Hedonic tone and activation level in the mood-creativity link: Toward a dual pathway to creativity model. Journal of Personality and Social Psychology, 94(5), 739–756. https://doi.org/10.1037/0022-3514.94.5.739
• Finke, R. A., Ward, T. B., & Smith, S. M. (1992). Creative Cognition: Theory, Research, and Applications. MIT Press.
• Guilford, J. P. (1950). Creativity. American Psychologist, 5(9), 444–454. https://doi.org/10.1037/h0063487
• Guilford, J. P. (1956). The structure of intellect. Psychological Bulletin, 53(4), 267–293. https://doi.org/10.1037/h0040755
• Guilford, J. P. (1959). Traits of creativity. In H. H. Anderson (Ed.), Creativity and Its Cultivation (pp. 142–161). Harper & Row.
• Guilford, J. P. (1967). The Nature of Human Intelligence. McGraw-Hill.
• Koestler, A. (1964). The Act of Creation. Hutchinson & Co.
• Mednick, S. A. (1962). The associative basis of the creative process. Psychological Review, 69(3), 220–232. https://doi.org/10.1037/h0048850
• Nijstad, B. A., De Dreu, C. K., Rietzschel, E. F., & Baas, M. (2010). The dual pathway to creativity model: Creative ideation as a function of flexibility and persistence. European Review of Social Psychology, 21(1), 34–77. https://doi.org/10.1080/10463281003765323
• Silvia, P. J. (2015). Intelligence and creativity are more similar than different: Evidence from cognitive, psychometric, and training studies. Journal of Intelligence, 3(2), 59–69. https://doi.org/10.3390/jintelligence3020059
• Torrance, E. P. (1966). Torrance Tests of Creative Thinking: Norms-Technical Manual. Personnel Press.