Permission Architecture is a framework for reading human action and its absence. It begins with a single observation: action is not only chosen. Action can be permitted by the system, shaped by what the system predicts will happen, blocked by a protection response, or released by a change in the environment.
This matters because almost all of the ways we currently think about behaviour begin in the wrong place. We begin with the behaviour that is visible and name it as if it were the cause. A person avoids something, and avoidance is named as the problem. A person cannot continue working, and burnout is named. But behaviour is late. By the time something is visible, a whole sequence of invisible decisions has already taken place inside the system.
The four questions
Permission Architecture asks four questions of any action or absence of action:
Does the person have the capability for this action? Capability is the knowledge, skill, or physical capacity to perform it. If capability is absent, the question is about learning or recovery, not permission.
If capability is present, does the person have access? Access means the capacity is available right now, in this context, under these conditions. A person may be capable of speaking in front of a group and have no access to that capability in a context that carries a high threat prediction.
If access is present, is there permission? Permission is the signal from the whole system that the action may proceed. Permission can be blocked by a protection response, by a prediction of negative consequence, by sensory load, by hormonal state, by the presence of a specific other person, or by a hundred factors the person may not be consciously aware of.
Finally: who or what holds the permission architecture? In coercive contexts, another person or system holds the architecture. In autistic burnout, the masking system has consumed the available permission. In clinical burnout, the system has withdrawn permission from the routes that cost the most.
Why this replaces judgement
When you read a person's actions through Permission Architecture, you cannot reach a judgement. You can only read a route. The question shifts from whether the person was willing to whether the permission was available. This shift is what makes it possible to investigate rather than evaluate, and to find the route back rather than assign a cause.