Gerald Ajam

The room is an interface

Interfaces organise possibility. They signal what can be done, make some actions easy, constrain others, reveal state, hide complexity, and direct attention. Rooms do many of the same things.

Status Position

Published September 2026

All notes

TL;DR

A room makes some actions obvious and others hard, the way a screen does, so it can be designed and questioned like one.

The distinction it turns on

The usual reading

The room as backdrop

A neutral container the lesson happens in.

The argument here

The room as interface

It signals, permits and hides, and it decides before anyone speaks.

01The situation

From classrooms to screens

I spent ten years teaching geography in classrooms, and I now spend my days on the user experience of a learning platform. The longer I do the second job, the more the first one looks like it too.

02The tension

My interpretation

What if a room were questioned like a screen?

A learning environment can be examined using many of the same questions we ask of a digital interface.

The habit is to treat the room as neutral: a container the lesson happens in. Read as an interface, it stops being neutral.

Every arrangement makes some actions obvious, some difficult, some visible and some owned by one person.

03The move

Six questions, none about furniture

Six questions do most of the work. None of them is about furniture.

  1. What actions does it make obvious?

    Gibson (1979) called the possibilities an environment offers its affordances. Norman (2013) added signifiers: the cues that tell you an action is available. A button with a shadow says press me. A flat label says nothing.

    Rooms signify too. Rows facing a board say that the important thing happens at the front and your job is to watch it. A cleared table with paper and pens in the middle says something different before anyone speaks. Learners read these cues within seconds of walking in, and they read them more reliably than they read the lesson objective.

  2. What does it make difficult?

    Good interfaces add friction on purpose, to the delete button and the payment screen. Bad ones add it by accident, to the thing you came to do.

    Rooms are full of accidental friction. The student who needs scissors has to cross the room and ask, in front of everyone. A group discussion in a room with hard walls and no soft surfaces becomes a shouting match. Talking to someone who disagrees with you means turning your chair round. None of this was designed. All of it shapes what happens.

  3. Who controls it?

    Every interface has a permission model. Some users can edit, some can only view, and the design makes the difference obvious or hides it.

    In a classroom, the permission model is often the projector remote. Whoever holds it decides what the room looks at. The same goes for the light switches, the whiteboard pens and the seating plan. If only the teacher can change any of these, the room says the learners are viewers. That may be right for a given lesson. It should be a choice, not a default.

  4. What information remains visible?

    An interface shows its state: what is saved, what is selected, where you are. Hide the state and people lose their place.

    The whiteboard is the room’s persistent shared state. A worked example left on it is a reference every learner can return to. A slide that advances after thirty seconds is state that vanished. Walls do the same work over longer periods: a timeline, a map, drafts in progress. Dillenbourg (2013) describes the teacher’s job as orchestration, and a visible record of where the class has got to is one of the few things that lets a teacher orchestrate thirty people at once.

  5. How easily can the configuration change?

    Software has switching costs. If changing a setting takes six clicks through a buried menu, people stop changing it.

    Rooms have switching costs measured in minutes and noise. Moving from rows to groups can take thirty seconds with a practised routine or five minutes without one. That cost decides which patterns get used. A room that costs five minutes to reconfigure will be reconfigured once a term, whatever the tables were bought for. Flexible furniture without a routine to move it is a conventional room at a higher price.

  6. What behaviour does the default arrangement encourage?

    Most people never change a default setting. They take the app as it arrived. That is why the default is the most powerful decision an interface designer makes.

    The classroom’s default is however it was left the night before, usually by the cleaner or the last teacher. That arrangement decides the first ten minutes of every lesson, and often the rest. A useful question in any lesson visit is not what the teacher did but what the room was already set up to do before they arrived.

04Where it breaks

Where the analogy breaks

A room is not a screen, and I should be honest about where the comparison fails. A room holds bodies, which get tired, cold and restless. It carries sound, so one group’s discussion is another group’s noise. And it carries social risk that a screen rarely does. Asking a question in front of thirty peers is not the same act as typing one into a box.

The other difference is evidence. An interface can be A/B tested with thousands of users in a week. A room cannot. You get one class, one term, and a hundred other things changing at once. Barrett et al. (2015) found that physical factors such as light, air quality and temperature are linked to pupils’ progress, and I take that seriously. Claims that one layout beats another are much weaker, and I treat them as hypotheses. The analogy gives me better questions. It does not give me answers.

05Where it lands

The decision that keeps deciding

The point of treating a room as an interface is not to import the vocabulary. It is to stop treating the room as neutral. Every arrangement makes some actions obvious, some difficult, some visible and some owned by one person.

The default arrangement is a design decision someone made. It keeps deciding, every lesson, until someone changes it.

The point

What transfers

The reason this page exists, rather than the story that produced it.

  1. Before planning a lesson, ask what the room is already set up to do. If it contradicts the activity, the room will usually win.
  2. Look at who holds the remote, the pens and the seating plan. That is the room’s permission model, whether or not anyone designed it.
  3. Measure the switching cost. If reconfiguring takes more than a minute, build the routine before buying the furniture.
  4. Keep the state visible. A worked example that stays on the board is worth more than one that scrolls away.
  5. Treat layout claims as hypotheses and environmental ones, such as light, air and acoustics, as the firmer ground.
Singapore