Two people walk into the same room. High ceiling, tall windows, warm light, a view onto green, furniture kept spare and modern. One of them feels free in it. The same room leaves the other unable to settle. I have not built this exact room; both reactions are invented. Anyone who has spent time in different rooms will still recognise the split: the same walls, at the same time, can affect two people in opposite ways.

Architecture usually thinks from the building outward. Proportions, materials, light, acoustics, all the elements a designer can actually shape. Architecture treats the room as the cause and the person as its effect.

I no longer think that is right. Most of what someone feels in a room is decided before they walk through the door. It arrives with the person: memory, expectation, the role they are playing that day, whatever happened to them an hour earlier, fear, sometimes relief.

Rooms do not arrive with a fixed function either. An office only becomes an office once someone connects it to a place where they have to deliver. A church works the same way. One visitor remembers being made to sit still there as a child. Another first found quiet there and kept it. The same building holds several different inner histories at once, and it is the visitor’s own expectation that turns into the feeling inside.

It works from the other direction too. The brain does not take in a room neutrally. It filters, compares and judges, in milliseconds, before consciousness even switches on. What someone “sees” is never quite the room as it is. It is the room as read through the filter of experience, mood and existing neural pattern.

If the person finishes building the room this way, optimising the room is not enough on its own. Whoever designs it has to understand the person who is going to walk in. A user with a checklist of functional needs is only part of that person. The larger part is someone carrying a history that decides what architecture can actually do to them.

I built myself what I thought was the perfect office. I could not work in it. The desk, the light and the quiet were all right. I had built the room for somebody else, a version of myself I am not.

My functional needs were easy to name: a desk, light and quiet. The actual ones took longer, and I only found them by accident. I need people near me who ask nothing of me. Silence isolates me. A loud café does more for me than a quiet office.

No questionnaire would have found this, and no architect asked. I could not have said it myself, because I did not know it.

Architects build for too simple a picture of the person, one where needs can be listed on a form: light, temperature, acoustics and floor area. Useful information, but it never reaches the layer underneath, where consciousness, memory and identity decide what a room can do to somebody.

Rooms that work well do something to a person that is hard to name. They take the person in before making any demand of them. Put plainly, this is architecture’s central task.

I know architects, some of them personally, who shake their head at this.

A different starting point follows. Architecture should begin with the person walking through the door: what they will do in the room, and who they are in the moment they step inside. The floor plan comes after that.

There are few tools built for that question. Architecture handles what a room is for. Psychology looks for what is wrong with someone, while neuroscience measures how a body reacts to a signal. The split is a simplification, but it makes the gap visible. None of these fields treats the person as a whole, and none asks what someone already brings to a room before they step inside.

If the task is to build better rooms, it depends on a harder task nobody has settled yet: understanding the person who is about to walk through the door, before they decide for themselves what the room means.