Parallel thinking is a structured approach to group problem-solving in which everyone in a discussion explores the same perspective at the same time, rather than taking opposing sides and arguing them out. Coined by Edward de Bono in the early 1990s, it was designed as an alternative to the adversarial debate style that dominates most Western decision-making. The idea is deceptively simple: instead of one person attacking an idea while another defends it, the whole group looks at the problem through a single lens together, then moves on to the next lens, and the next, until a fuller picture emerges. Its best-known practical tool is de Bono’s Six Thinking Hats, which assigns each lens a color-coded role.
How Parallel Thinking Differs from Traditional Debate
De Bono’s central criticism was that Western intellectual culture inherited a deeply adversarial style of reasoning from what he called the “Gang of Three”: Socrates, Plato, and Aristotle. In this tradition, you discover truth by setting up opposing claims and arguing one against the other. One side wins; one side loses. De Bono argued that this approach, while powerful for exposing logical flaws, carries costs: it rewards ego over exploration, it polarizes groups, and it can shut down creative ideas before they develop.1Encyclopedia of Educational Theory and Philosophy. Creative and Lateral Thinking: Edward de Bono The adversarial model works well in courtrooms and formal debates. It works less well when a team of engineers, nurses, or teachers needs to cooperate on a complex problem where no single person holds the answer.
Parallel thinking replaces opposition with coordination. Rather than asking people to stake out positions and defend them, you ask everyone to look in the same direction at once. If the group is examining risks, everyone examines risks, including the person who proposed the idea. If the group is generating optimistic possibilities, the habitual skeptic participates too. The key distinction is that perspectives are explored sequentially and collectively, not split across individuals who then clash. De Bono described parallel thinking as generating multiple views from differing perspectives concurrently, with the sole purpose of moving overall comprehension of a topic forward.2The Palgrave Handbook of Educational Thinkers. Edward de Bono
This matters practically because, in most group settings, people adopt a position early and spend the rest of the meeting defending it. Parallel thinking short-circuits that instinct. You are not “the optimist” or “the critic” for the duration of the discussion. You wear each hat in turn. The result is that the group’s energy goes toward understanding the problem from all angles rather than winning the argument from one angle.
The Six Thinking Hats as a Practical Framework
The Six Thinking Hats is the most widely used implementation of parallel thinking, and for many people it is their only encounter with the concept. Each hat represents a distinct mode of thought:
- White hat: focus on available data, facts, and information gaps. What do we know, and what do we need to find out?
- Red hat: emotions, gut feelings, and intuitions. No justification required; participants simply share how they feel about the idea.
- Black hat: caution, risks, and weaknesses. What could go wrong? Where are the dangers?
- Yellow hat: optimism and value. What are the benefits? Why might this work?
- Green hat: creative alternatives and new ideas. What else could we try? What lateral possibilities exist?
- Blue hat: process control. The facilitator role: setting the agenda, calling for hat changes, summarizing progress.
The sequence matters less than the discipline. The facilitator (blue hat) decides which hat the group wears and when to switch. Everyone shifts together. That synchronized switching is what makes it parallel thinking rather than just “brainstorming with categories.” Nobody is permanently assigned to the black hat, so the chronic pessimist on your team gets to spend time in yellow-hat mode, and the relentless optimist has to sit with the risks for a few minutes.
Evidence from Education
The Six Thinking Hats technique has been studied most extensively in educational settings, where structured group thinking is easier to test in controlled conditions. In one study of eighth-grade students, applying the technique across five curriculum-related sessions improved students’ critical thinking skills, as measured by pre- and post-study questionnaires alongside student interviews and diaries.3Asos Journal. THE EFFECT OF SIX THINKING HATS TECHNIQUE ON CRITICAL THINKING SKILLS OF 8 TH GRADE STUDENTS The improvement makes intuitive sense: forcing a student to look at a topic through the lens of emotions, data, risks, and creativity in sequence builds a more layered understanding than simply asking them to “discuss.”
In medical education, the results are more striking. A study of oral pathology residents compared a traditional teaching approach with a method combining problem-based learning and the Six Thinking Hats. Diagnostic accuracy jumped from 40% under traditional teaching to roughly 87% with the combined method. Critical thinking and risk analysis scores rose from 30% to over 93%, and knowledge retention two weeks later remained substantially higher in the group that used the hats.4PubMed Central. Evaluating the effectiveness of problem-based learning combined with six thinking hats in oral pathology education These are large effect sizes, though the study involved a relatively small group of residents and used a specific combination of methods, so the Six Thinking Hats alone may not account for the entire difference. Still, the pattern across educational research is consistent: structured parallel exploration of perspectives tends to outperform unstructured discussion or one-directional lectures.
More recently, researchers have begun pairing the Six Thinking Hats with AI tools. A study of pre-service teachers found that groups using the Green Hat (creative thinking mode) in combination with ChatGPT showed measurably different patterns of co-regulation during instructional design tasks compared with groups that did not use the AI tool.5ScienceDirect. Fusing Six-Hat thinking with AI: How the Green Hat and ChatGPT Co-Regulate pre-service teachers’ instructional design—insights from epistemic network analysis The AI essentially served as an additional thinking partner during the creative phase, suggesting that the hat framework is flexible enough to absorb new tools without losing its structure.
Applications in Healthcare Decision-Making
Healthcare is one area where adversarial thinking can be genuinely dangerous. A patient’s care plan involves input from doctors, nurses, therapists, the patient, and family members, all of whom may hold different priorities and information. A nurse focused on daily comfort may clash with a surgeon focused on long-term outcomes, and in adversarial mode, the person with the most institutional authority tends to win regardless of whose insight is most relevant.
Using the Six Thinking Hats in collaborative care settings gives each stakeholder structured time to voice concerns, share emotional responses, and propose alternatives without the dynamic collapsing into hierarchy or conflict. A review in nursing practice concluded that the technique could augment collaborative decision-making by providing patients, families, and health professionals with opportunities to make informed decisions about care that considers key issues for everyone involved.6PubMed. Collaborative care: Using six thinking hats for decision making The red hat, in particular, fills a gap that traditional clinical reasoning ignores: the patient’s emotional reality. A treatment plan that is medically optimal but emotionally unbearable for the patient is not, in practice, a good plan. Giving emotions a named and legitimate slot in the process means they get heard without derailing the factual or risk-assessment phases.
Why It Works: Psychological Safety and Shared Mental Models
One of the strongest explanations for why parallel thinking improves group performance is that it lowers the psychological cost of speaking up. In adversarial settings, voicing a concern means attacking someone else’s position. Raising an emotional objection can feel unprofessional. Admitting you do not understand the data can feel like exposing a weakness. Parallel thinking sidesteps all of this because the structure itself legitimizes each mode. When the facilitator says “black hat,” everyone is supposed to raise concerns. You are not being negative; you are doing the exercise.
Research on team reflection in pediatric healthcare teams found that pre-action reflective behaviors, which share structural similarities with parallel thinking, fostered psychological safety, reduced barriers for sharing input, supported inclusive leadership, and enhanced the development of shared mental models. These behaviors also enabled more effective planning. Participants described how the reflective process both relied on and reinforced team dynamics like trust and familiarity, creating a self-amplifying loop.7PubMed Central / Wiley Online Library. ‘You can’t have an ego in this game’: A simulation primed qualitative inquiry of team reflection in paediatrics That self-amplifying quality is worth emphasizing: the more a team practices structured perspective-sharing, the more trust they build, which makes the next round of structured perspective-sharing work even better.
This also helps explain why some organizations report that their first attempt at Six Thinking Hats feels awkward but the fifth or sixth session runs smoothly. The tool is partly a trust-building exercise disguised as a thinking framework.
What Happens in the Brain During Perspective-Taking
Parallel thinking asks people to temporarily adopt a perspective that is not their default. When you put on the yellow hat and genuinely try to see the upside of an idea you dislike, you are engaging in cognitive perspective-taking. Neuroimaging research sheds some light on what this demands from the brain.
Taking a third-person perspective, as opposed to a first-person one, activates regions in the inferior parietal cortex, the medial posterior cortex, and the prefrontal cortex. In adults, this kind of reasoning relies primarily on the left inferior parietal cortex and the precuneus. Children doing the same task show broader activation, including the dorsolateral prefrontal cortex on both sides, suggesting that perspective-taking becomes more efficient and lateralized with practice and development.8PubMed. Learning to appreciate others: neural development of cognitive perspective taking In practical terms, adults who are experienced perspective-takers seem to need less brainpower to do it, which frees up cognitive resources for the actual problem at hand.
A separate line of research on role-playing offers an intriguing complement. When participants stepped into a role during psychodrama exercises, activity in the left prefrontal cortex decreased compared with natural conversation. This implies that self-referential processing, the internal monologue of “what do I think about this,” quiets down when you commit to thinking from someone else’s vantage point.9The Arts in Psychotherapy. Decreased activation in left prefrontal cortex during role-play: An fNIRS study of the psychodrama sociocognitive model That is essentially what the hats are asking you to do: temporarily set aside your own default stance and think from a designated perspective. The neuroscience suggests this is not just a metaphor. The brain genuinely shifts its processing patterns when you adopt a role.
Cognitive Flexibility and Its Limits
Switching between hats requires cognitive flexibility, which is the ability to shift your thinking style from one mode to another. Research on task switching shows this is not costless. People with higher working memory capacity tend to switch more efficiently, spending less time preparing for each new task set.10PubMed Central. Task Switching: On the Relation of Cognitive Flexibility with Cognitive Capacity This means parallel thinking may come more naturally to some people than others, and facilitation speed should account for the group’s capacity. Rushing through hat changes could leave some participants still mentally stuck in the previous mode.
There is also an interesting tension between flexibility and depth. Research on adolescents found that while set-shifting, the broad ability to move between mental categories, supported divergent thinking, rapid cue-driven task switching actually predicted lower originality. Quick switching seemed to interfere with the deeper exploration required to produce genuinely novel ideas.11PubMed Central. Set-shifting and task-switching make differential contributions to divergent thinking in adolescence The implication for parallel thinking practice is that facilitators should give the group enough time in each hat to go deep. If switches happen too rapidly, the group gets the structure without the substance. A green-hat phase that lasts two minutes is unlikely to produce the kind of original ideas the framework is designed to surface.
Developmental research adds another layer. Studies of children show that executive flexibility and semantic flexibility (the ability to shift between conceptual categories) are strongly correlated, suggesting shared control mechanisms. But the interplay between conceptual knowledge and cognitive control grows with age, meaning younger participants may need even more time and scaffolding to shift effectively between hats.12PubMed Central. A cross-sectional study of executive and semantic flexibility in childhood using task-switching Educators using the Six Thinking Hats with younger students should expect longer transitions and more facilitator guidance than they would need with adult professionals.
Parallel Versus Serial Idea Generation
The word “parallel” in parallel thinking can cause confusion with a related but distinct concept in brainstorming research: parallel versus serial subgroup structures. In brainstorming, “parallel” sometimes means splitting a large group into subgroups that work simultaneously on the same problem, while “serial” means passing ideas from one subgroup to the next in sequence. A field study of large groups found that parallel subgroups produced more unique ideas, while serial subgroups produced more elaborations on existing ideas.13PubMed. Exploring asynchronous brainstorming in large groups: a field comparison of serial and parallel subgroups
De Bono’s parallel thinking is a different animal. It is not about splitting groups; it is about synchronizing them. Everyone thinks in the same direction at the same time. But the brainstorming research is still relevant because it highlights a useful practical point: if your goal is a large volume of unique ideas, parallel structures (whether in the de Bono sense or the subgroup sense) tend to outperform approaches that funnel ideas through sequential critique. If your goal is deeper elaboration of a few strong ideas, a serial or adversarial approach might actually be more productive. The best facilitators know when to use each.
Team Composition and Communication Style
Not every team responds to parallel thinking in the same way. Research on small networked teams found a surprising result: teams skewing toward extraversion did not communicate as efficiently as more introverted teams when using text-based collaboration, which is the default mode for many remote and distributed teams today. The researchers noted that complex mission environments require a high degree of parallel information processing, which is more naturally performed through text chat than through the voice-based communication that extraverted teams tend to prefer.14International C2 Journal. Effects of Individual and Team Characteristics on the Performance of Small Networked Teams
This has practical implications for how you set up a parallel thinking session. In a room where everyone speaks aloud, dominant voices naturally take up more airtime, even during a structured hat exercise. Moving part of the process to written channels, such as having everyone type their black-hat concerns into a shared document simultaneously, can level the field. Introverted team members, who often do their best thinking in writing, get equal real estate. And the parallel nature of the exercise becomes literal: everyone is generating input at the same time, not waiting their turn to speak.
Cross-Cultural Teams and Adapted Frameworks
One persistent challenge in distributed and multicultural teams is that thinking and communication styles vary across cultures. What counts as appropriate criticism, how emotions are expressed in professional settings, and how hierarchy shapes who speaks first all differ from one cultural context to another. Researchers have adapted the Six Thinking Hats framework specifically for this challenge, creating what they call “Cultural Dimensions Thinking Caps.” The method uses Hofstede’s Cultural Dimensions as a theoretical basis and de Bono’s hats as a practical basis, aiming to help multicultural distributed development teams solve problems creatively by making cultural assumptions explicit and structured rather than implicit and collision-prone.15The Design Society. DESIGNING A METHOD FOR IMPROVED (DISTRIBUTED) CULTURAL UNDERSTANDING – THE CULTURAL DIMENSIONS THINKING CAPS
The idea is that if a team member from a high power-distance culture is reluctant to contradict a senior colleague, the hat structure gives them permission and a mandate to do so during the black-hat phase. The structure externalizes the act of criticism, attaching it to the process rather than to the person. This is one of the subtler strengths of parallel thinking generally: it depersonalizes disagreement. You are not arguing with your boss. You are both wearing the black hat right now, and the black hat’s job is to find problems.
When Parallel Thinking Is Not the Right Tool
Parallel thinking has real limitations, and advocates sometimes oversell it by treating adversarial reasoning as inherently inferior. Adversarial debate exists for good reasons. When a decision has clear right and wrong answers, when the facts are disputed and need to be stress-tested, or when a bad idea needs to be killed quickly, structured opposition can be faster and more decisive than a full Six Thinking Hats cycle. A courtroom cross-examination is adversarial for the same reason surgery uses a scalpel instead of a paintbrush: precision and accountability matter more than creative exploration.
Parallel thinking also struggles when the facilitator is weak. Without firm blue-hat guidance, groups tend to revert to their defaults. The loudest voice dominates, the hat switches get sloppy, and the process becomes a veneer over the same old power dynamics. The technique depends on genuine buy-in from all participants. If half the room thinks the exercise is silly, the hats become costumes rather than cognitive tools.
There is also a tempo problem. In fast-moving crises, there may not be time to walk methodically through six perspectives. Emergency medicine, military operations, and financial trading floors all require snap judgments where a decision delayed by structured deliberation is a decision made too late. Parallel thinking is best suited to problems where time allows reflection: strategic planning, product development, policy design, care planning, and educational inquiry. For situations demanding rapid action, other frameworks, including hierarchical command structures, exist for good reason.
Parallel Thinking in the Age of AI-Assisted Teams
An emerging question is how AI tools fit into a parallel thinking framework. Large language models can generate text from virtually any perspective on demand, which makes them natural candidates for filling hat roles. Want a devil’s advocate analysis? Ask the AI to think in black-hat mode. Need creative alternatives? Green hat. Some early research, including the study pairing ChatGPT with the Green Hat in teacher training, suggests this can work, though the dynamics change when one of the “thinkers” at the table is not human.5ScienceDirect. Fusing Six-Hat thinking with AI: How the Green Hat and ChatGPT Co-Regulate pre-service teachers’ instructional design—insights from epistemic network analysis
The risk is that AI-generated perspectives may feel thorough without actually being grounded in the team’s specific context. A language model can produce a convincing-sounding black-hat critique of any idea, but it does not know the organizational politics, the budget constraints visible only to the CFO, or the patient’s unspoken fear that the nurse noticed during a quiet moment. Parallel thinking’s power comes partly from the fact that real humans with real, different knowledge bases are forced to share that knowledge through a structured process. If AI substitutes for that human knowledge rather than supplementing it, the framework loses something essential. The most promising use is probably as a preparation tool: let AI draft hat-specific prompts or pre-populate data for the white-hat phase, then let the humans do the actual thinking together.