There is a kind of person most of us have sat across from in a meeting, at a dinner table, or in the middle of an argument who speaks with a confidence so total that it shuts the room down. They do not pause. They do not hedge. When a question arrives they respond immediately, and the speed of the response carries its own authority, as if the quickness itself is evidence that they know. The room takes a moment to collect itself and then mostly defers, because the alternative is to challenge someone who seems so much more certain than you feel.
I used to find these people intimidating. Now I mostly find them legible. Because once you understand the mechanism underneath that certainty, it stops reading as intelligence. It starts reading as something older and more specific, a habit built to manage a feeling that most of them stopped being consciously aware of a long time ago.
What overconfidence actually signals
In 1999, psychologists Justin Kruger and David Dunning published a paper at Cornell that has since become one of the most cited in social psychology. Their research, published in the Journal of Personality and Social Psychology, found that people with limited knowledge in a domain consistently overestimated their own competence, while people with genuine expertise tended toward more calibrated, and sometimes more cautious, self-assessment. The mechanism they identified was metacognitive: the skills required to perform well in a domain are often the same skills required to recognize poor performance in that domain. If you lack the former, you also lack the latter. You do not know enough to know what you do not know.
This is the cognitive portrait of the most confidently wrong people in any room. They are not lying. They genuinely believe what they are saying with the conviction they are projecting. Their certainty is real. The problem is that their certainty is not calibrated to their accuracy, and without the metacognitive equipment to see that gap, they cannot close it. The more they talk, the more they generate what feels to them like confirmation.
But Kruger and Dunning’s finding, as important as it is, only explains part of the picture. It describes the cognitive structure of overconfidence. It does not fully explain why some people develop it so reliably while others in the same room do not. For that, you have to go earlier.
Where the certainty was learned
Carol Dweck has spent decades studying how children’s beliefs about intelligence shape their behavior, and one of her most important findings is about what happens when hesitation gets penalized. In research with Claudia Mueller, published in the Journal of Personality and Social Psychology, Dweck and Mueller found that children praised for being smart, rather than for effort or process, developed a specific pattern of behavior in response to difficulty. They avoided challenges. They became reluctant to show uncertainty. They interpreted hesitation and struggle as evidence that they were not, in fact, smart, and they prioritized protecting the smart label over engaging honestly with the difficulty.
Dweck’s broader body of work describes these as fixed mindset behaviors. When a child grows up in an environment where intelligence is treated as a fixed trait to be demonstrated rather than a capacity to be developed, the incentive structure around being wrong changes completely. Being wrong means you are not smart. Being slow means you are not smart. Hesitating, revising, saying I am not sure, all of these register as threats to the identity that is generating the approval. So the child learns to answer fast. To project certainty. To treat the confident delivery of an answer as the social product, regardless of whether the answer is correct.
That habit does not disappear when the child grows up. It becomes increasingly automatic, increasingly invisible to the person carrying it, and increasingly difficult to challenge because the person’s sense of competence is built on top of it. The adult in the room who never pauses before answering is not, in most cases, doing so because they are exceptionally well-informed. They are doing it because they learned somewhere, probably before they were ten years old, that pausing was dangerous.
What the hesitation was protecting against
The feeling that is being avoided is not complicated. It is the feeling of not knowing, in a room where knowing is the currency of worth. In the environments where this habit forms most reliably, being uncertain was not treated as an honest and temporary cognitive state. It was treated as evidence of a deficiency. Teachers moved on quickly from children who did not immediately have the answer. Parents who rewarded quick, confident responses trained the same pattern. Peer groups that mocked confusion made hesitation socially costly in a way that left marks.
Decades later, the person is still answering fast to avoid that feeling. Except the feeling is not usually being triggered by anyone in the current room. Nobody is waiting to mock them. Nobody is withdrawing approval because they paused to think. The social threat that trained the reflex has long since passed. But the reflex outlasted it, because reflexes always do. The behavior became the default before anyone was old enough to examine it, and examining it as an adult requires recognizing that it was a response to a problem that no longer exists.
Research on children’s calibration between confidence and accuracy, published in PLOS ONE, found that even young children can detect miscalibration in others, that there is something recognizable in the gap between someone’s certainty and the reliability of what they are certain about. Adults are even better at this. Which means the room usually knows, at some level, that the most confident voice is not necessarily the most accurate one. They just do not always know how to say so, because certainty carries social weight that accuracy does not.
What genuine intelligence looks like in the same room
The people I find most intellectually compelling in a room are almost never the loudest. They are the ones who pause before answering a genuinely difficult question, not because they are slow, but because they are taking the question seriously. They are the ones who say I am not sure, let me think about that in contexts where others would have already delivered a verdict. They are the ones whose confidence, when it does appear, lands differently, because it is clearly calibrated to something real.
I notice this in myself and in the people around me in Saigon. My wife is one of the most perceptive people I know. She does not answer questions quickly. She thinks visibly, sometimes in ways that create a pause that would make a lot of Western conversational styles uncomfortable. And then what she says is usually more accurate and more interesting than what I would have said if I had answered immediately. The pause is not weakness. It is the sound of genuine consideration. I am still learning to let my own pauses last long enough to do the same work.
In Buddhist thought, the concept of right speech, or samma vaca, includes the idea that speech should emerge from genuine understanding rather than from the need to fill a silence or perform a role. There is a quality of speaking that comes from actually knowing, and a different quality that comes from needing to seem like you do. Most people in most rooms can feel the difference. They just rarely say so out loud. I write about this kind of grounded, honest engagement in my book Hidden Secrets of Buddhism.
The loudest, most certain person in the room is not performing confidence in spite of knowing a lot. In many cases, they are performing it instead. The certainty is the product of a much older transaction, a deal struck with an environment that punished uncertainty, that is still being honored long after the other party left the room.
Understanding that does not make them less frustrating to sit across from. But it does make them less intimidating. And it makes the quiet person on the other side of the table, the one still thinking before they answer, easier to recognize for what they actually are.
