Volume is a poor test for confidence. A quiet person can be certain and wrong; an expressive person can be accurate and generous. The more reliable distinction appears when ability is tested, another person contributes and evidence changes. Grounded confidence tolerates those moments because self-worth does not depend on being superior. Arrogance protects a rank: it overstates the self, discounts other people and treats correction as a threat. That makes behaviour over time more informative than posture, eye contact or a single bold claim.

Separate confidence, accuracy and status

Confidence is a person’s felt certainty about an answer, ability or course of action. Accuracy is whether that judgement matches reality. Status is how the person ranks relative to others. The three can move independently. A novice can be confident and mistaken; an expert can be accurate but cautious; a leader can hold authority without claiming superior worth. Calling all visible certainty arrogance collapses these distinctions.

A broad research review defines arrogance as an inflated view of one’s abilities or importance that is tied to presumption, excessive self-confidence or superiority. Its framework joins overconfidence in the self with underestimation of others. The relational component matters: error alone is not arrogance, and high competence alone does not protect against it. Review: Foundations of Arrogance

This is why confidence should be calibrated rather than minimised. Someone making a safety decision may need to speak decisively. The question is whether the certainty tracks evidence, names uncertainty where it matters and remains open to better information. A person who always sounds tentative can still ignore evidence; a person who says “I am confident because these three checks agree” has made the claim testable.

Four behaviours reveal the difference

First, watch how the person explains the basis for a claim. Confidence can show its work; arrogance often substitutes position, charisma or contempt. Second, watch what happens when evidence conflicts. A confident person may defend the conclusion vigorously and still update it. An arrogant response changes the subject, attacks the messenger or raises the certainty after disconfirmation.

Third, notice how credit and competence are distributed. Grounded confidence can recognise another person’s skill without losing identity. Arrogance treats another person’s success as a reduction in its own standing, so it appropriates credit, withholds information or dismisses contributions. Fourth, look at repair. Everyone becomes defensive. The useful signal is whether the person returns, corrects the record and changes behaviour.

A single encounter is weak evidence. Culture, language, neurodiversity, stress and power affect eye contact, directness and conversational style. Someone who interrupts may be anxious, rushed or following a different norm. Someone who avoids self-promotion may be confident but constrained by social penalties. Judge repeated choices in relevant contexts and describe the behaviour rather than assigning a fixed identity.

Intellectual humility is not low self-esteem

A major review of intellectual humility finds a common core across competing definitions: the metacognitive ability to recognise limitations in one’s knowledge and beliefs. The literature links that quality with learning, discovery, reduced polarisation and better interpersonal outcomes, while also noting that measures and causal claims remain imperfect. Nature Reviews Psychology: predictors and consequences of intellectual humility

Humility does not require saying “I am probably wrong” when the evidence is strong. It requires matching certainty to evidence and leaving a route for revision. A pilot, surgeon or engineer may need to make a timely decision with incomplete information. Intellectual humility appears in the conditions attached to the decision: what is known, what remains unknown, what would trigger a change and whose expertise is needed.

False modesty can be another form of status management. Habitually denying skill may solicit reassurance, shift emotional labour to colleagues or hide accountability. A calibrated statement is clearer: “I have done this successfully before; I have not handled this particular failure mode.” It preserves competence and exposes the boundary where help is useful.

Confidence grows through evidence, not slogans

Choose one bounded skill and define observable performance. Instead of “be more confident in meetings,” try “state one recommendation, give its reason and answer one question without apologising for speaking.” Repetition then supplies evidence about preparation, communication and recovery. The result can support confidence or reveal a training need; both are more useful than trying to feel certain first.

Keep a calibration record for consequential judgements. Write the prediction, probability, reasons and outcome. Over time, compare 70% confidence claims with how often they were right. This catches both inflated certainty and chronic underconfidence. It also separates the emotional discomfort of deciding from the quality of the decision itself. Use a reasonable set of comparable decisions rather than treating two wins as proof of excellent judgement, and record outcomes you would otherwise prefer to forget.

Seek feedback that is specific enough to act on. “You need more confidence” often means the observer wants a recommendation earlier, a louder voice, less qualification or clearer ownership. Ask what behaviour created the impression and what a better version would look like. Feedback about style should not erase justified uncertainty or reward bluffing.

How to respond when another person seems arrogant

Move from labels to consequences. “When you interrupted twice, we lost the test result Anna was reporting” is discussable; “you are arrogant” invites a dispute about character. Ask for the evidence behind the claim, the confidence level and what would change the conclusion. In a team, use routines that distribute airtime, record decisions and make dissent possible before rank settles the issue.

Boundaries are appropriate when behaviour becomes demeaning or unsafe. Expertise does not excuse humiliation, and politeness does not require endless debate. State the standard, document repeated incidents and use formal routes when power prevents repair. At the same time, distinguish an unwelcome answer from abusive delivery. Someone can accurately reject an idea and still owe colleagues respect.

If the relationship matters, test for teachability. Offer one concrete example and a credible alternative. A confident person may dislike the feedback but can work with it. Persistent retaliation, contempt and refusal to examine evidence are stronger warnings than self-promotion alone. The goal is not to win a diagnosis; it is to decide whether collaboration can become safe and productive.

Leadership makes calibration visible

A meta-analysis of humble leadership describes the construct as accurate self-view, appreciation of others’ strengths and teachability. Across the included research, humble leadership was associated with outcomes such as engagement, voice and citizenship behaviour, although the authors caution that measures and study designs limit strong causal conclusions. Meta-analysis: humble leadership and outcomes

Leaders can make confidence useful by separating decision authority from certainty. They can say who decides, invite disconfirming evidence, set a review point and explain why the decision stands. After an error, correcting the record without theatrical self-abasement models the behaviour expected from others. Sharing credit signals that competence is not a zero-sum resource.

The practical line between confidence and arrogance is therefore not thin in the way a mood or facial expression is thin. It is a pattern. Confidence is willing to be measured, can name its limits and leaves other people fully human. Arrogance needs superiority, so it becomes brittle around correction and contribution. Look at what the person does when reality or another person refuses to stay beneath the story they tell about themselves.

Sources