Why the surest voice in the room is often the least informed

Understanding a subject means discovering its complications, and complications make people hedge.

There is a rough inverse relationship between certainty and knowledge, and the mechanism behind it is not mysterious. When you know a little about a subject, you have a simple model of it, and simple models generate clean answers. As you learn more, you meet the exceptions, the cases where the obvious rule fails, the arguments on the other side that turn out to be better than you assumed. Your model becomes more accurate and much less tidy. The answers it produces come with conditions attached, which is not a retreat but a more honest description of what is known.

The result is that the person with the cleanest position is frequently the one who has looked at the subject the least, and they are not being dishonest either. Their confidence is sincere. It reflects a model that has not yet been complicated by contact with the difficult cases. Someone further in cannot go back to that simplicity without pretending, and pretending is precisely what the audience often rewards.

This does not mean confidence is always a bad sign. In fast-feedback domains an experienced practitioner may be flatly certain about something, and be right, because they have seen it hundreds of times. The distinguishing feature is scope. Earned confidence is specific and bounded: they are certain about this, in these conditions, and will tell you where the certainty ends. Unearned confidence is broad and unconditional, covering the whole subject at once, with no visible edges. Listen for the edges. Their presence or absence tells you more than the tone of voice ever will.