What is Metacognition, Really?

Metacognition is, crudely, thinking about thinking. But that phrasing flatters the term into uselessness. The version we care about is older, more specific, and far more operational.

In 1979, the developmental psychologist John Flavell proposed a framework that still survives scrutiny. He split metacognition into two tractable parts. The first is metacognitive knowledge: what you know about how minds work, including yours. That you tire in the afternoon. That certain problems collapse when you write them out. That your memory compresses meetings into a single mood. The second is metacognitive regulation: the moment-to-moment choices you make about how to think. Slowing down. Switching strategies. Checking your work. Noticing that the plan you've been building for ten minutes rests on an assumption you have not examined.

Flavell's original move was sharper still: he split metacognitive knowledge into three variables, and the distinction is worth doing work with. Person variables are what you know about your own cognitive wiring , that afternoon fade, that weakness for the last voice in the room, that way your memory compresses a meeting into a single mood. Task variables are what you know about what this specific situation demands , that a decision under uncertainty is not a calculation, that a negotiation is not a debate, that some problems want a diagram and others want more reading. Strategy variables are what you know about the tools available to deploy , write it out, steelman the opposite, set a timer, sleep on it, ask a peer to predict the outcome. Skill is knowing which strategy fits which task for which version of yourself today. Most operators are strong on one variable, thin on the other two, and have never noticed the asymmetry.

Schraw and Moshman sharpened the regulation side in 1995 by distinguishing three sub-skills: planning, monitoring, and evaluation. Planning is what you do before: predicting how long a task will take, choosing a strategy, deciding where to put your attention. Monitoring is what you do during: asking whether your current approach is actually producing understanding, whether your confidence is tracking your evidence, whether the emotion you're feeling is information or noise. Evaluation is what you do after: looking at the result and learning something about yourself that changes the next attempt.

That third piece, evaluation, is where most operators fail. They have plans. They sometimes monitor. They almost never keep honest records of what happened and why, so the next plan inherits all the same blind spots as the last one.

Four levels of practice

David Perkins, writing in 1992, gave us a ladder that is worth the effort of climbing. It is useful to name where a person currently stands, because the work required is different at each rung.

A tacit operator is unaware of their own thinking. They know things or they don't. They solve problems or they don't. When a strategy works, they cannot say why; when one fails, they cannot tell you what they would change. This is the default adult state on most tasks , not a deficiency, just the starting line.

An aware operator notices that thinking is happening. They can tell you they are weighing options, generating alternatives, checking a gut feeling. But the awareness is descriptive rather than deliberate: they see the weather of their mind without reaching for an umbrella. Most professionals arrive here by their mid-thirties and stay there. It is a genuine step up from tacit, and it is also where a great deal of otherwise capable thinking quietly stalls.

A strategic operator has a shelf of named strategies and reaches for them when a task is recognizable. They know to write it out. They know to sleep on it. They know to ask for a pre-mortem. What they mostly do not do is adjust while a task is in motion; the strategy is selected up front and then executed, even when the situation has quietly changed. Most professional training produces strategic operators and stops.

A reflective operator revises while the work is happening. Mid-meeting, mid-draft, mid-decision, they notice the approach isn't producing what they need and change it. This is the expensive rung. It cannot be taught by case study because the skill is live , it requires feedback on thinking that is still in motion, which is why the Academy is four days of facilitated practice instead of four days of lecture.

The same person can sit on different rungs in different domains. An executive who is reflective in capital allocation may be aware in board dynamics and tacit in their marriage. The point of naming the levels is not to classify people; it is to make the next move visible.

Why this matters for decisions

An executive's job is to make decisions under uncertainty faster and more frequently than anyone else around them. The bottleneck in that job is rarely information. It's almost always the quality of the thinking applied to the information, and a specific kind of quality: the ability to notice when the thinking is going badly and adjust while there is still time.

A leader who can only think is a leader who can be stampeded by the next plausible story. A leader with metacognition can, mid-story, say: I notice I'm being persuaded. Let me check what's actually in evidence. That pause, repeated a few hundred times a quarter, compounds into measurably better calls.

The research is blunt about this. Metacognitive skill correlates with performance across domains , medicine, negotiation, complex problem-solving , independently of raw intelligence or domain expertise. It is trainable. It responds to deliberate practice. And it degrades without maintenance, the way physical conditioning degrades, which is why a one-day offsite on "critical thinking" does almost nothing.

Three failure modes in the wild

Overconfidence in judgment. You trust your gut because it has worked before, without tracking the cases where it was wrong. A calibration log breaks the illusion within about thirty entries.

Emotional mislabeling. You feel something you call "stressed" and treat it as a signal to push harder. It was actually disappointment from the last meeting, still bleeding into this one. The mislabel produces a worse decision than the raw feeling would have.

Narrative capture. You form a story about what is happening, then collect only the evidence that fits the story. The monitoring skill is the one that catches this , a quiet internal alarm that says the last three things I noticed all point the same way; am I hunting?

What we mean when we say "the practice"

When we talk about a metacognitive practice, we do not mean meditation, though meditation touches it. We mean a short, specific set of operations a person performs daily, in the flow of real work, to build monitoring and evaluation into their default loop. A morning intent. A midday checkpoint on what's actually happening inside the decisions being made. An evening reflection that keeps the ledger honest. A decision journal with calibrated forecasts. A peer who will ask, at the right moment, what you might be wrong about.

None of this is exotic. All of it is uncommon. That gap , between what is known and what is practiced , is where the Academy, and this library, do their work.