Brain EP. 12 / 64 By SENSMAN

The Subtitle of Common Sense — What Is Common Sense: Knowing That You Do Not Know, Metacognition, and If A, then B

Common sense is not brilliant. It is the ability to know that you do not know. A machine covers the blank with smooth words; human common sense stops in front of it.

The subtitle of common sense — judgment that stops the moment it knows it does not know

I stood behind a pharmacy counter for seventeen years.

The phrase I used most in those years was “Let me check that first.”

When a dose on a prescription looked unusually high, or a name could be confused with a similar drug, I stopped and called the clinic. Waiting customers grew irritated. Most of the time the prescription was right.

I stopped anyway. Korean pharmacy law requires it, but before the law my body had learned to move that way.

I do not know. So I check.

Looking back, that one line is all I know of common sense.

You cannot change the event, but you can change its meaning

What has happened cannot be undone. What it means can be chosen differently by different people.

This is not consolation; it has been measured.

In 1998 James Gross showed two groups the same unpleasant film with different instructions. One was told not to let their feelings show (suppression). The other was told to interpret the film’s meaning differently (reappraisal).

StrategyUnpleasantness feltPhysical cost
Suppression — push the feeling downnot reducedsympathetic arousal actually rose
Reappraisal — change the meaningreducedno increase

The suppressor is quiet only on the surface. The one who changes the meaning changes inside.

Ridicule sharpens the senses; defeat rewrites the procedure; a blocked road reassembles the next move. The same event becomes different material.

A lesson must be a reflex, not a sentence

“I’ll do better next time” is usually not kept next time either.

The lessons that last have a different form.

If A, then B. When signal A comes, do B.

The psychologist Peter Gollwitzer called this an implementation intention. Studies piled up comparing people who only set a goal with people who also fixed “when this happens, I will do that”.

In 2006 Gollwitzer and Sheeran pooled 94 studies. The effect of implementation intentions on reaching goals was d = 0.65, medium to large by psychology’s standards.

The reason is simple. Binding the situation to the action in advance removes the deliberation when the situation arrives. Judgment becomes reflex.

So what must be written down in advance is not a feeling but answers.

  • If the same thing happens again, how will I move?
  • When which signal comes, what will I do?
  • How long will I wait, and at what moment will I step back?
  • What will I confirm before I go back in?

Those answers collected together are a standard operating procedure (SOP).

One sheet of paper halved surgical deaths

There is an experiment that shows how strong a standard procedure can be.

In 2009 Haynes and colleagues, with the World Health Organization, introduced a 19-item checklist into operating theatres at eight hospitals in eight countries. The items were things everyone already knew: confirm the patient’s identity, mark the surgical site, share the expected blood loss.

BeforeAfter
Deaths1.5%0.8%
Complications11.0%7.0%

No new technology went in. It only made sure that what was already known was not skipped.

Even a brilliant surgeon skips a check when tired. Common sense left to individual talent collapses. It survives when it is fixed as procedure.

Common sense is the ability to know that you do not know

Here is this piece’s definition.

Not a genius’s flash, not an intuition others admire.

Common sense is the ability to know that you do not know.

Psychology calls knowledge about one’s own knowing metacognition.

There is a famous report on what happens when metacognition fails. In 1999 Kruger and Dunning found that people in the bottom quarter on tests of logic, grammar and humour — actually around the 12th percentile on average — placed themselves around the 62nd. Those who do not know also do not know that they do not know.

The study comes with a caveat. In 2020 Gignac and Zajenkowski, among others, argued that much of the effect is a statistical artefact of how it is measured. The direction broadly replicates; the size may be smaller than first reported.

The more famous a study, the more you should read the part still being checked alongside it. That too is common sense.

The machine pretends to know what it does not

Against this definition, the difference between people and machines becomes sharp.

Language models are trained to produce plausible sentences. Even to questions they cannot answer they offer confident ones — citing papers that do not exist, inventing figures. This is called hallucination.

Why? In 2025 Kalai and colleagues pointed to evaluation as one cause.

AnswerTest score
“I don’t know”0
A guesspoints when it happens to be right

On a test that penalises “I don’t know”, guessing pays even when you do not know. That is why a machine raised to score well covers the blank with smooth words.

Human common sense stops in front of the blank. While writing this piece every figure was re-checked against its original source — all the more so when a machine’s sentence sounded convincing.

Are children’s lies really caught so quickly?

People say children cannot lie without being caught: the body reacts first and the eyes cannot hide it.

It sounds right. Measured, it is wrong.

In 2017 Gongola, Scurich and Quas pooled 45 experiments: 7,893 adult judges and 1,858 children aged 3 to 15.

What was judgedAdult accuracy
Truths and lies together54%
Lies alone47.5% — no different from a coin toss

Adults judging adults are no better. In 2006 Bond and DePaulo pooled 206 studies: average accuracy in detecting lies was 54 percent.

Believing you can tell and actually telling are different things. The confidence that “I can spot a lie” is exactly where common sense breaks.

The moment you call certain what is not certain, the subtitle of reason that was carrying your judgment blurs. What keeps that subtitle clear is not a sharp eye but the habit of checking.

The order of common sense

Put together, common sense moves in four steps.

  1. Admit that you do not know
  2. So check
  3. If you cannot check, move small
  4. If you cannot even move small, wait

Faced with grand words, loud announcements and endlessly repeated lines of hope, common sense asks a few things.

  • Did the speaker see it with their own eyes?
  • Where does the number come from?
  • Are the people who know best acting on what is said?

Before those questions, a feast of words loses its force.

What is still unknown

  • Can metacognition be trained? Many classroom studies report gains, but transfer to other domains appears small in several results
  • How large is the Dunning-Kruger effect really? Reanalyses with different measurement keep appearing
  • Can a machine be made to say “I don’t know”? Changing evaluation has been proposed but has not become standard

The definition this piece fixes

Set it down here.

Common sense is not the amount of knowledge but knowing where knowledge ends. I do not know → check → move small → wait.

That is why this piece’s hexagram is Bi, 否: heaven and earth do not meet, a blocked time, and the noble one keeps his virtue plain to avoid trouble. Not forcing a blocked road, not trying to win a game you do not understand — that is common sense.

This series has taken up intuition, imagination, emotion and common sense in turn. Intuition reads signals not yet in words; imagination draws roads that do not yet exist; emotion tells you through the body which way you lean. The final judgment is made by common sense.

One place remains: the place where people believe all of that judgment happens.

The mind.

The next piece begins with 一切唯心造 — everything is made by the mind.

Summary

  • The event cannot be changed, but its meaning can. Reappraisal reduces negative feeling more than suppression (Gross 1998)
  • A lesson lasts when fixed in advance as “If A, then B”. Implementation intentions: d = 0.65 across 94 studies
  • A surgical checklist cut deaths from 1.5% to 0.8% and complications from 11.0% to 7.0% (Haynes 2009)
  • Common sense is knowing that you do not know — metacognition
  • The Dunning-Kruger effect replicates in direction; its size is debated
  • Language models lean towards guessing because evaluation gives no credit for “I don’t know”
  • Adults detect children’s lies 47.5% of the time — chance level
  • The order of common sense: I do not know → check → move small → wait

This piece is rebuilt for search readers from “The Subtitle of Common Sense” of the serial essay The Great Corona Swindle. What happened at the pharmacy is the author’s own experience; the research findings and figures have been reset against primary literature. No individual health advice or diagnosis is offered here.

Definitions

Metacognition
Cognition about one's own cognition: judging what one knows and does not know.
Implementation intention
A plan fixed in advance in the form 'if situation A arises, I will do B'. Proposed by Peter Gollwitzer.
Cognitive reappraisal
Regulating an emotional response by reinterpreting what an event means rather than the event itself.

Frequently asked

What is common sense?
This piece defines common sense as the ability to know that you do not know — not the amount of knowledge but awareness of its edge. Psychology calls knowledge about one's own knowing metacognition. In practice it runs in order: when you do not know, check; if you cannot check, move small; if you cannot even move small, wait.
Is the Dunning-Kruger effect real?
In 1999 Kruger and Dunning reported that people with the lowest test scores most overestimated their own ability. Since the 2010s, however, critics have argued that much of the effect comes from statistical artefacts such as regression to the mean and the way it is measured. The direction — low performers misjudge their level — broadly replicates, but its size and interpretation are still debated.
Why does AI pretend to know what it doesn't?
Language models are trained to continue text plausibly. In 2025 Kalai and colleagues argued that evaluation adds to this: answering 'I don't know' scores zero while a guess is sometimes right, so a model built to do well on tests leans towards guessing even when it does not know. Confidently producing things that are not true is called hallucination.
Are children's lies really easy to spot?
No. In 2017 Gongola, Scurich and Quas combined 45 experiments with 7,893 adult judges and found an average accuracy of 54 percent in telling children's truths from lies. For lies alone it was 47.5 percent — a coin toss. Professionals such as police and teachers did only slightly better.

References

#common sense#metacognition#Dunning-Kruger effect#implementation intention#standard operating procedure#checklist#cognitive reappraisal#AI hallucination#lie detection#judgment

#상식#메타인지#더닝크루거효과#실행의도#표준행동절차#체크리스트#인지재평가#AI환각#거짓말탐지#판단력

Published October 28, 2026