UX Glossary · S
Signifier
The perceivable cue that tells a user an action is possible, as distinct from the possibility itself.
A signifier is the perceptible signal that communicates what can be done. Don Norman introduced the term to correct a confusion he had helped create: he had used affordance loosely to mean what a user could see was possible, where the original concept meant what was actually possible.
Separating them makes the design work clearer. The affordance is the possibility; the signifier is what advertises it.
Why the distinction is practical
A screen has no physical affordances at all. Everything is signified. A button is pressable because of a border, a fill, a shadow, a shape convention or its position, and there is nothing underneath those cues.
Remove the signifiers and the affordance is invisible, which is precisely what happened when flat design was taken to its limit. The function still existed; nobody could tell.
Good and bad signifiers
A door’s flat plate signifies push, and a vertical bar signifies pull. When those are wrong, people fail regardless of intelligence, which is Norman’s point.
In interfaces the equivalent failures are common: text coloured slightly differently as the only cue that it is a link, icons with no established meaning and no label, and areas that are clickable but look identical to areas that are not.
In practice
The test is prediction. Someone should know what will happen before they touch it. If they have to try it to find out, the signifier is missing.