You approach a glass door. A flat metal plate covers the surface at hand height. You push. The door doesn’t move. You pull. It opens. You mutter something under your breath and walk through, vaguely embarrassed by a failure so minor that you’ll forget it within seconds — and repeat it at the next badly designed door you encounter. The embarrassment, by the way, is misplaced. The failure belongs to the door.

Norman Doors and the Man Who Named Them

Cognitive scientist Don Norman identified the problem in his 1988 book The Design of Everyday Things, and the concept has since acquired his name: a “Norman door” is any door whose design gives incorrect signals about whether to push, pull, or slide. The essential argument is simple. If a door requires a sign to explain how to use it, the door’s design has failed. The physical form of the door — its handles, plates, hinges, and frame — should communicate the correct action without any text.

A flat plate says “push.” Your palm lands on it, and the natural motion is forward. A vertical handle says “pull.” Your fingers wrap around it, and the natural motion is toward your body. When a door has a pull handle on the push side, the hardware is issuing an instruction that contradicts the function. You’re not wrong for pulling. The door lied to you.

Why Doors Get It Wrong

If the solution is so obvious — flat plates on push sides, handles on pull sides — why do so many doors get it wrong? Three factors converge.

Aesthetic uniformity. Architects and interior designers frequently specify identical hardware on both sides of a door to maintain visual symmetry. A building’s entrance might feature matching brushed-steel pull handles on the exterior and interior, even though one side should push and the other should pull. The visual coherence of the facade takes priority over the functional legibility of the interaction.

Fire code complications. In many jurisdictions, fire codes require doors on escape routes to open outward — in the direction of egress. But commercial buildings often want entrance doors to open inward, preventing them from swinging into pedestrian pathways. When the aesthetic choice and the code requirement point in opposite directions, the resulting hardware compromise frequently produces a door that communicates nothing clearly.

Cost and standardisation. Door hardware is manufactured in standard configurations. Ordering custom combinations — a flat plate for one side and a pull bar for the other — increases cost and complicates installation. Many builders default to the cheapest option that meets building code, regardless of usability implications.

The Affordance Problem

Norman introduced the concept of “affordance” to describe the relationship between an object’s physical properties and the actions it invites. A chair affords sitting. A button affords pressing. A horizontal bar at waist height affords pushing. When a door’s affordances align with its function, nobody notices the door. When they don’t, everyone notices — they just blame themselves instead of the design.

The self-blame is significant. Most people who push a pull door experience a flash of social embarrassment, not a critique of the object. The assumption is “I should have known” rather than “this door should have communicated.” Design researcher Hisham Abushaala has noted that poorly designed interfaces train users to distrust their own instincts, even when those instincts responded correctly to the physical signals the object presented. You saw a flat plate. You pushed. That was the right instinct. The door was the wrong object.

Revolving Doors and Automatic Slides: The Bypass Solutions

Revolving doors and automatic sliding doors eliminated the push-pull problem entirely by removing the interaction altogether. You don’t have to decide anything — you walk forward, and the door accommodates you. Their adoption in commercial buildings, hospitals, and airports was driven primarily by energy efficiency and accessibility, but a secondary benefit is the complete elimination of Norman door confusion.

The persistence of manual swing doors in most residential, retail, and institutional settings means the design problem remains relevant. A 2019 observational study conducted across 50 commercial buildings in a European capital found that approximately 38 percent of entrance doors displayed push-pull hardware inconsistencies — handles where plates should be, or identical hardware on both sides. One in three doors actively misled the people trying to use them.

The Cost of Getting It Right

Fixing a Norman door is trivially easy. Replace the pull handle on the push side with a flat plate or a horizontal crash bar. Total cost: 20 to 80 euros in hardware and a half-hour of installation time. The fact that millions of buildings worldwide have not made this change reflects something deeper than laziness or ignorance. It reflects a built-environment culture that does not consider user interaction a design priority.

Architects design buildings to be looked at. Engineers design buildings to stand up. Building codes ensure buildings are safe. But nobody in the standard construction workflow is formally responsible for ensuring that a person approaching a door for the first time can operate it without confusion. The usability of the most fundamental interaction in architecture — entering and exiting — is treated as self-evident, even when the evidence clearly shows it is not.

Every day, in every city, millions of people push doors that should be pulled and pull doors that should be pushed. Each one assumes the mistake was theirs. It wasn’t. The door told them the wrong thing, and nobody in the process of designing, manufacturing, specifying, or installing it prioritised telling them the right thing. The Norman door is not a quirky design curiosity. It is a daily, global, quietly universal failure of the built world to communicate one of its simplest possible messages: which way this opens.