In almost every group I have watched make a bad decision, someone in the room knew better. That is the part that stayed with me. The failure was rarely a failure of knowledge. The knowledge was present, sitting in someone’s head, and it never made it into the room.
I started noticing this during fieldwork, in meetings where a community was deciding whether to adopt some new practice. Afterward, talking to people one at a time, I would hear the doubts. The person who had farmed that particular soil for thirty years and suspected the recommendation would not survive the first heavy season. The one who remembered a similar scheme that had collapsed a decade earlier. They had known. In the meeting, they had said nothing, or said it once, quietly, and let it drop when no one picked it up.
For a long time I filed this under shyness or deference. I have come to think it is something more structural, and more troubling, because it means the problem will not fix itself.
The Gap Between Knowing and Saying
There is a distance between what a person knows and what they are willing to say out loud, and the size of that distance depends almost entirely on the room. The same person who will tell you the truth over tea will go silent in a meeting, because the knowledge stayed the same while the cost of voicing it rose.
Disagreeing in a group is expensive. You risk being wrong in front of people whose regard you depend on. You risk irritating whoever proposed the idea, who often holds more status than you. You break a consensus that everyone else seems comfortable with, which makes you the obstacle, the difficult one, the person slowing things down. And you do all of this for a benefit that, if it arrives at all, is shared by everyone in the room equally, including the people who stayed silent and risked nothing.
That asymmetry is the heart of it. The cost of dissent is personal and immediate. The value of dissent is collective and delayed. A person weighing whether to speak is weighing a certain, private cost against an uncertain, shared benefit, and the rational choice, most of the time, is to say nothing and let the decision happen. Multiply that across a room and you get a group that is confidently wrong while containing, distributed across its members, everything it would have needed to be right.
Why Welcoming Dissent Is Not Enough
The usual remedy is to say that groups should welcome disagreement. Leaders are told to be open, to invite challenge, to create a safe space. I have watched this fail often enough to distrust it.
The trouble is that welcoming dissent does nothing about its cost. A leader can sincerely invite challenge and the arithmetic facing the potential dissenter does not change. It is still their reputation on the line, still their standing that takes the hit if they are wrong, still their name attached to the objection that annoyed everyone. An invitation to speak freely, extended into a room where speaking freely is expensive, is mostly decorative. People read the real incentives, not the stated ones.
David Perkins gave this its clearest image. In his book on how organisations think together, he built the argument around King Arthur’s round table: a table with no head, where a person’s place does not decide whether their words carry. The round table was a claim about structure rather than manners. A group reasons well only when its arrangement keeps rank from filtering what gets heard.
What changes behaviour is making dissent cheap, which is a harder and more practical task. It means the person with authority holding their opinion back until others have spoken, so that no one is contradicting the boss. It means gathering views privately before the group converges, so a doubt can enter the record without its author having to stand up alone. It means treating a wrong objection as a normal event rather than a mark against someone, so that being mistaken in public is survivable. These are structural choices about how a decision gets made, and they matter far more than any amount of encouragement.
The Knowledge Is Already There
What I find hopeful in this, and I do find it hopeful, is that it reframes the problem. Groups that make poor decisions are usually treated as groups that lacked information or reasoned badly. Often they had the information and reasoned fine. What they lacked was a way for their quietest, least powerful, most easily overruled members to put what they knew on the table without paying too much for it.
That is a fixable problem. You cannot conjure knowledge that a group does not possess. You can change the conditions that keep the knowledge it already possesses trapped inside the people least able to afford the cost of saying it.
So when I watch a group now, I have stopped asking only whether the smart people are present. I ask whether the room is built so that the person about to say something unwelcome can afford to say it. Usually someone knows. The question is whether the group has made it safe enough for them to be the one who does.