In my work I often hear leaders describe walking into a room and knowing within a second or two whether it is safe to say what they actually think. I have always described that as fast processing, the brain moving quickly through information it already has. New research from Lisa Feldman Barrett at Northeastern University and Earl Miller at MIT, published this year in Nature Reviews Neuroscience, suggests I had the mechanism the wrong way round.
Their argument is that the brain does not take in a room and then categorise it. Categorisation happens first, and perception is built to fit around it. The anatomical case for this is the striking part. In the visual cortex, roughly 90 percent of the synaptic connections carry signals travelling backward, from memory and context outward toward the eyes, rather than forward from the eyes inward. The wiring is built so expectation shapes what is sensed, not the other way round. By the time the raw signal from the room arrives, the brain has often already assembled a category of safe or not safe to interpret it with.
Picture the last time you walked into a meeting and felt, before anyone had said a word, whether it was safe to raise the thing you had been rehearsing in the car. That was not your senses reporting back on what happened in the room. According to Barrett and Miller, that was prediction arriving ahead of the senses, built from what your brain already expected this kind of room, this kind of leader, this kind of meeting to hold.
We have spent years treating psychological safety as something people consciously weigh up once they are in the room. Apparently not. The category of safe or not safe is constructed before conscious thought catches up, and it does not arrive as a neutral label. Barrett and Miller describe categories as bundled together with the bodily response they are already preparing, so a not safe read does not sit there waiting for a decision. It comes with the readiness to defend, withdraw or shut down already switched on.
This is the strongest anatomical case I have seen for something I have argued since 2021 with Social Safety, that whether a person feels safe to take an interpersonal risk is assessed before it is consciously felt, not after.
What would change in your team’s meetings if you accepted that people are not choosing to hold back, their brain has already decided the room is not safe before they knew they were deciding anything?
Barrett, L. F. & Miller, E. K. (2026). Categorization is ‘baked’ into the brain. Nature Reviews Neuroscience.
#Neuroscience #PsychologicalSafety #SocialSafety #Leadership #NeuroscienceOfLeadership
