Architectural atmosphere

Why can one room feel welcoming and another of identical size feel cold?

Two rooms, both six metres by eight, both painted white, both at twenty-one degrees. You want to stay in one and leave the other, and you knew which was which within about a second of stepping in.

That speed is the interesting part. Whatever your body responded to, it was not deliberating — so it was responding to something present and physical, not to taste or association. The temptation is to call that something "atmosphere" and treat it as ineffable. I would rather ask a duller question: what, measurably, differs between those two rooms?

Take the plainest candidate first, because it is the one people least expect.

Sound. Strip a room of soft surfaces and its reverberation time — how long a sound takes to decay by sixty decibels — goes up. Long reverberation smears speech, so conversation becomes effortful and everyone raises their voice, which raises the noise floor further. Rooms described as "cold" are very often rooms with long reverberation times — a number you can measure with a meter and change with a rug. This is the most solid item on the list: the acoustics are physics, and the effect on speech intelligibility is well established.

Surfaces. Put your hand on an oak rail and a steel rail in the same room and one feels warm. Neither is warmer — both are at room temperature. What differs is thermal effusivity, the rate at which a material draws heat out of your skin. Metal and stone pull heat fast and read as cold; wood, textile, cork pull it slowly and read as warm. So "warm materials" is not a metaphor.

Light. A window with something to look at through it is not the same as a window with a wall beyond it. The most-cited study here is Ulrich's 1984 finding that surgical patients in rooms facing trees had shorter postoperative stays and needed less strong pain medication than matched patients facing a brick wall — suggestive, and worth flagging as one small naturalistic study rather than a law.

Enclosure and outlook. People given a choice consistently take the seat with its back to a wall and a view over the room — the edge of a café before its middle. Appleton called this prospect-refuge: we prefer places that let us see without being exposed. As a description of what people do it is robust; as an evolutionary explanation it is much shakier, and I would not lean on it.

Ceiling height. Meyers-Levy and Zhu reported in 2007 that people under higher ceilings did better on tasks requiring abstract thought, and people under lower ceilings on detail-focused ones — suggestive, and it has not travelled entirely cleanly through replication. Less controversial is the relational point: height that is generous for a hall is oppressive in a corridor, so there is no good ceiling height, only a good ratio to the room's plan and its use.

Scale to the body. Doorways, sills, seat heights, handrails, the depth of a window reveal — sized to a body, they read as made for occupation; sized for monumental effect, as made for looking at.

Now notice the pattern in the list. Almost none of these are visible in a photograph, and none appear on a floor plan. Reverberation, effusivity, the depth of a reveal, whether the light is direct — all are things you meet with your body rather than your eye. Which begins to explain why two rooms can be identical on the drawing and unrecognisable in person, and why "atmosphere" seems ineffable: we describe rooms with the sense that is least involved in judging them.

Three refinements, and one caution I want to make loudly.

First, none of these levers acts alone, and they trade against each other. A high ceiling with hard plaster gives you both grandeur and a long reverberation time; the fix is soft surfaces, not a lower ceiling. Daylight from one side gives glare and a dark back half; the same quantity from two sides reads as evenly lit. The combination produces the impression, which is why single-variable advice disappoints.

Second, the response is fast but not fixed. Familiarity, culture, what you came to the room to do, and how many other people are in it all shift the judgement, sometimes more than the fabric does.

Third — the caution. This evidence base is uneven. The acoustics and the thermal behaviour of materials are hard physics. The daylight and view findings are moderately good, mostly from health settings, and entangled with everything else about a hospital. The ceiling-height and prospect-refuge work is a mixture of small laboratory studies and theory, some of it not replicating cleanly, most of it correlational. It is honest to say certain rooms are reliably experienced as cold; it is not honest to offer a tidy causal formula for warmth.

• How reverberation time is actually specified for different room uses, and why open-plan offices fight it.
• Whether the ceiling-height effects survive larger pre-registered replication, and what that would change.

Every term the collection defines is gathered in the glossary.