Dr. Neil's Notes
General > People
Psychological Safety
Introduction
No framework for distributed work will function if people do not feel safe to speak, to challenge, and to raise problems. Psychological safety is not a soft concept. It is the precondition for everything else. A team that cannot surface bad news early, that does not challenge ideas before they become commitments, and that stays silent when something is going wrong, will eventually fail regardless of how well-designed its processes are.
Amy Edmondson's research on this is consistent: team psychological safety is most strongly predicted by whether the leader models fallibility, invites input, and responds non-punitively to mistakes. Policies and programmes do not create safety. Leader behaviour does.
Rewarding the Raising of Problems
One of the most common cultural failure modes I have observed is a team that celebrates the heroes who fix crises and says nothing about the people who raised the problem three months earlier when it was still fixable. If the culture only rewards the fix, people learn not to raise the flag until the situation has become dramatic enough to justify the attention.
The correction is straightforward. Make the raising of problems explicitly valued. Thank people for flagging things early. Treat early warning as a mark of good judgement, not as a failure report. Over time, the team learns that problems are safer to surface than to sit on.
Separating the Person from the Mistake
A blame culture does not only hurt the person blamed. It silences everyone watching. When someone sees a colleague punished for a mistake, they learn that mistakes are dangerous to admit to. That lesson tends to spread faster than any intended lesson about the mistake itself.
The alternative is not the absence of consequences. It is a clear separation between the error and the person who made it. Understanding what went wrong and preventing it from happening again does not require a narrative in which someone is at fault.
Making Dissent Safe
Before a decision is final, it is worth actively inviting the counterargument. The most useful question a leader can ask is not "does everyone agree?" but "what am I missing?". The difference is significant. One invites compliance. The other invites challenge.
After the decision is final, something different is needed; commitment. People can disagree with a decision and still commit to it. What is not workable is silent non-compliance, where someone who disagreed simply deprioritises the outcome without saying so. Encouraging people to dissent openly, and then to commit clearly, is one of the most valuable habits a team can build.
Belonging vs. Safety
Safety and belonging are related but not the same thing. A person can feel safe from punishment for speaking and still feel peripheral to the team. Belonging is knowing that your particular presence and perspective makes the team genuinely better. If you cannot articulate specifically why each person on the team improves it, inclusion work is incomplete.
In distributed teams this requires more design than in co-located ones. The informal moments that signal to someone that they matter, being included in a conversation, being asked their opinion before a decision is made, being mentioned by name when something goes well, are harder to replicate across screens and time zones, but they are not impossible. They just need to be deliberate.