Tacit knowledge is rooted in the idea that human beings know far more than they can put into words. The philosopher Michael Polanyi introduced this concept in 1966, arguing that much of what we know exists as personal, experiential insight rather than facts we can write down or explain to someone else. The classic example is riding a bicycle: you can do it fluently, but try describing the exact sequence of muscle adjustments and balance corrections to someone who has never ridden, and you’ll find it nearly impossible. That gap between what you can do and what you can articulate is the foundation of tacit knowledge.
The Philosophical Roots
Polanyi’s core claim, laid out in his book The Tacit Dimension, was that “we can know more than we can tell.” This wasn’t a minor observation. It was a direct challenge to the scientific tradition that treated all legitimate knowledge as something that could be made fully explicit, written into rules, and verified through skepticism. Polanyi argued the opposite: that tradition, inherited practices, implied values, and personal judgment are not obstacles to knowledge but crucial parts of it. Even scientific discovery, he insisted, depends on intuitions and unspoken assumptions that researchers absorb through experience rather than from textbooks.
This philosophical position has two dimensions. The first is technical: the informal “know-how” that a master craftsman develops over years, the kind of skill that resists being broken into steps. A seasoned surgeon’s ability to sense when tissue feels wrong, or a chef’s instinct for when a sauce has reduced enough, falls into this category. The second dimension is cognitive: the beliefs, values, mental models, and perceptions so deeply embedded in a person’s thinking that they’re taken for granted. These shape how someone interprets a situation without conscious deliberation. Together, these two dimensions explain why tacit knowledge is personal, context-dependent, and stubbornly resistant to being written into a manual.
How the Brain Stores Unspoken Skill
Neuroscience has mapped tacit knowledge onto specific brain structures that operate differently from the regions responsible for facts and conscious recall. The brain essentially runs two parallel memory systems. Declarative memory, the system for facts and events you can consciously retrieve, depends on a structure deep in the brain’s temporal lobe called the hippocampus. Tacit knowledge, by contrast, lives primarily in the basal ganglia, a cluster of structures involved in habit learning and procedural skill.
The basal ganglia specialize in learning gradual stimulus-response associations, the kind of pattern recognition that builds up slowly through repetition rather than through a single “aha” moment. Studies of people with Parkinson’s disease, which disrupts the brain’s dopamine supply to the basal ganglia, illustrate this clearly. These patients struggle with tasks requiring incremental, pattern-based learning but perform normally on tests of factual recall. The reverse pattern appears in patients with hippocampal damage, who can still learn skills but can’t remember new facts.
As a skill becomes more automatic, brain activity shifts from the front of the basal ganglia to the back, reflecting a transition from effortful learning to deeply ingrained habit. The cerebellum, a structure at the base of the brain, also plays a role in certain types of motor learning and coordination. This neural architecture explains something important: tacit knowledge isn’t just knowledge you haven’t gotten around to explaining yet. It’s stored in brain circuits that are fundamentally separate from the ones that handle language and conscious reasoning, which is part of why verbalizing it feels so difficult.
Two Thinking Systems and Where Tacit Knowledge Fits
Cognitive psychology offers another lens through dual-process theory, the idea that human thinking operates in two modes. The first mode, often called fast thinking, is automatic, effortless, and association-based. It produces the snap judgments and gut feelings that guide most of your daily decisions. The second mode, slow thinking, is deliberate, effortful, and rule-based. It’s what you engage when you work through a math problem or carefully weigh the pros and cons of a decision.
Tacit knowledge operates almost entirely through fast thinking. When an experienced nurse glances at a patient and senses something is wrong before any test results come back, that’s fast thinking drawing on years of accumulated pattern recognition. Slow thinking can sometimes override these intuitions, and that’s useful when gut feelings are wrong. But in domains where someone has deep experience, fast thinking is often remarkably accurate. It processes cues that the person couldn’t list if asked, drawing on the kind of knowledge Polanyi described: knowledge held instinctively, in the unconscious mind and even in muscle memory.
How Tacit Knowledge Is Acquired
You don’t learn tacit knowledge from a textbook. It forms through two primary pathways. The first is socialization: learning directly from another person through observation, imitation, and shared experience. This is the apprenticeship model, where a newcomer watches an expert work, absorbs their habits and judgment, and gradually develops similar instincts. No document changes hands. The knowledge transfers through proximity and practice.
The second pathway is internalization, where you take something explicitly taught, like a set of guidelines or a procedural manual, and practice it so thoroughly that it becomes second nature. A medical student who memorizes diagnostic criteria and then spends years applying them in clinical settings eventually stops consciously consulting the criteria. The knowledge has moved from explicit to tacit, from something looked up to something felt. Research on expertise in fields like public health and emergency management confirms that professional skill is often characterized by actual field experience and cannot be transmitted solely through classroom training.
Both pathways share a common requirement: time. Tacit knowledge is inherently incremental. The brain’s habit-learning circuitry builds associations slowly, strengthening neural connections through repeated exposure to real situations with real consequences.
Why Tacit Knowledge Resists Being Made Explicit
Several features make tacit knowledge fundamentally hard to capture. It is context-dependent, meaning the same expert may respond differently in subtly different situations based on cues they process unconsciously. It is personal, shaped by an individual’s specific experiences and the mental models those experiences have built. And it is often social, embedded in relationships and networks that provide information and trust developed over years.
Research on professional decision-making during public health incidents found that experienced practitioners consistently had difficulty describing the judgment processes they relied on. Their descriptions pointed to knowledge of local contexts developed over years, accumulated experiential judgment, and trust in professional networks, all things that were largely taken for granted by the people who possessed them. When asked to generalize their decision-making approach across different situations, they often couldn’t. The knowledge was too tightly bound to specific circumstances.
Researchers have developed methods to surface at least some of this hidden expertise. One approach, the critical incident technique, uses structured interviews to walk an expert through a past challenging event step by step, asking them to reconstruct not just what they did but what they were thinking and perceiving at each moment. By anchoring the conversation in a specific, vivid memory rather than asking for general rules, interviewers can pull out insights that the expert wouldn’t volunteer unprompted. It doesn’t fully codify tacit knowledge, but it reveals fragments that can be taught to others.
The Challenge Tacit Knowledge Poses for AI
Polanyi’s insight created a lasting problem for computer science, sometimes called Polanyi’s paradox. If people can’t articulate what they know, then programmers can’t write those rules into software. Tasks that demand flexibility, judgment, and common sense, like developing a hypothesis or organizing a cluttered closet, have proved especially difficult to automate precisely because the knowledge behind them is tacit.
This paradox has shaped entire labor markets. Jobs that rely heavily on tacit knowledge, both high-skill roles requiring expert judgment and low-skill roles requiring physical adaptability, have grown in recent decades, while middle-skill routine jobs have been more vulnerable to automation. Modern machine learning attempts to sidestep the paradox by letting systems learn from human examples, inferring the rules people tacitly apply but don’t explicitly understand. It’s a workaround, not a solution. The gap between what humans can do and what they can explain remains one of the central challenges in artificial intelligence.