The most expensive belief in culture work is that belonging is a messaging problem. An organization drafts a values statement, books a training day, and hangs the poster in the break room. The people it was written for keep running a much older system underneath: a nervous system that quietly scores every meeting, every correction, and every silence for one thing. Safety. That system does not read posters.

What the scanner showed

The evidence behind this is more than two decades old and still underused. In 2003, Naomi Eisenberger and her colleagues at UCLA published a study in Science with a simple design. People played a virtual ball-tossing game while lying in a brain scanner, and partway through, the other players quietly stopped passing to them. Nothing dramatic. No insult, no confrontation. Just exclusion.

The result became one of the most cited findings in social neuroscience. Brain regions that track the distress of physical pain became more active during the exclusion, and that activity tracked how much the exclusion hurt. The interpretation has been refined and debated in the years since, as good findings are, but the core observation holds up well as a working frame: being left out is not a bruised ego. The body treats it as a form of injury.

That single frame changes what culture work is. If exclusion registers as injury, then belonging is not a branding exercise. It is closer to a safety program.

Where belonging is actually decided

It follows that belonging is not decided at the offsite. It is decided in small, repeated signals, most of them cheap to send and expensive to fake.

How correction lands is one. Respectful, private correction signals that a mistake is survivable here. Correction delivered as public exposure signals threat, and everyone who watched it learns the same lesson as the person who received it.

What follows a mistake is another. Repair, or quiet exile. The room watches which one happens and updates its verdict about how safe it is to try anything.

And then the small courtesies that barely seem worth naming: who gets interrupted, whose name gets said right, whose idea gets credited to the person who actually had it. None of these appear in a values deck. All of them are read, in real time, by the safety system described above.

Cultural humility as a working practice

If the signals of safety were universal, a checklist would solve this. They are not. What reads as respect, directness, warmth, or threat varies by person, by culture, and by history, and in a genuinely multicultural workplace, that variation is a daily operating condition rather than a footnote.

That is the case for cultural humility, and for treating it as a practice rather than a module. A module ends. A practice is a standing posture: assume you do not already know what safety looks like for each person on your team. Ask. Watch. Adjust. And when you get it wrong, repair quickly, because repair lands in the same physiology the injury did.

This is also why the practice cannot be delegated to a poster. The signals that decide belonging come from the people with power in the room, in the moments when something is at stake. A leader who corrects with respect after a costly mistake does more for belonging in ninety seconds than the values statement does in a year.

Why this is worth a leader’s attention

The payoff is not warm feelings. People who feel safe think better, speak up earlier, surface problems while they are still cheap, and stay. People who feel scored and provisional spend real cognitive capacity on self-protection, and that capacity comes straight out of the work. The teams that feel safe are not the ones with the best poster. They are the ones where the small signals keep coming back safe.

Belonging is built. It is built in small signals, sent by people who stayed curious. If this is territory you are working on with your own team, here is how we work with teams.