On the stages of psychological safety

Every manager who has run a good "tell us something real" exercise has felt the same disappointment. The room goes quiet, or worse, performs. People say something true-shaped rather than something true. The exercise was not wrong. It simply arrived before the team was ready for it.

That gap - between asking for honesty and a team actually being able to give it - is where the research on psychological safety begins, and where the shape of Guftagoo Works comes from.

What William Kahn observed

William Kahn first described psychological safety in 1990 while studying why people fully engage in some workplaces but withdraw in others. He found that people engage fully only when they feel three things are true - that the work is meaningful, that they have what they need to do it, and that it is safe to show up as themselves without fear of damage to how they are seen. That third condition is what he named psychological safety. It was never a wellness idea. It was an engagement idea. People do not withhold their real thoughts because they lack them. They withhold because showing them once felt costly.

What Amy Edmondson added

Kahn described the condition. Amy Edmondson, at Harvard, spent the next decade asking what it does inside teams specifically. Her research found that psychological safety is the strongest predictor of whether a team actually learns - whether people ask for help, admit an error, or flag a problem before it becomes expensive. Crucially, she found safety does not mean comfort or the absence of disagreement. It means the team believes that speaking up will not be held against them. Her later work brought the term into wide use in organisations, but the core finding is older and simpler than the popularity suggests: teams get smarter when it is safe to say "I don't know" or "I was wrong."

Dr Timothy Clark's four stages

Kahn and Edmondson each described a state - safe, or not safe. Timothy R. Clark's contribution was to describe safety as a sequence rather than a switch. In his account, a team moves through four stages: inclusion safety, the sense of simply being accepted; learner safety, the sense that it is safe to ask and to fail while learning; contributor safety, the sense that one's input is genuinely wanted; and challenger safety, the sense that it is safe to disagree with how things are done. Each stage depends on the one before it. A team cannot skip to comfortably disagreeing with its manager if it has not yet established that questions are welcome, and it cannot ask questions comfortably if it does not yet feel it belongs. This is the structure Guftagoo Works most directly borrows - not the words, but the order.

Where Guftagoo differs

We have adapted Clark's four stages into six months, not because the number four was wrong, but because two things needed more room than a single stage each. Inclusion, in our experience, splits into a genuinely playful opening month and then two separate months of belonging - an early, easier kind, and a deeper kind that asks a little more. And learner, contributor and challenger safety needed new names, in language that belongs to a weekly practice rather than a diagnostic model: Learning Together, Shared Contribution, Honest Challenge. The sequence is Clark's insight. The practice, the cards, the callback, and the six-month pacing are ours.

Why a weekly practice instead of a workshop

Almost none of this research recommends a single intervention. Edmondson's own work is explicit that safety is built through repeated, ordinary interactions - how a manager responds the first three times someone admits a mistake, not how convincingly a workshop explains the concept. A single good day in a room fades by Thursday. That is the practical reason Guftagoo Works is a weekly card, not a training day: the research does not describe an event. It describes a habit.