---
name: Signifier
slug: signifier
category: interaction
status: published
created: 2026-08-21T00:00:00.000Z
modified: 2026-08-28T00:00:00.000Z
definition: The visible cue that advertises an affordance, such as a grip
  texture, an underline, or a shadow that says this can be pressed.
aliases:
  - name: perceptible signifier
    source: community
  - name: visual cue
    source: community
  - name: interaction cue
    source: community
tags:
  - perception
relations:
  contrastWith:
    - affordance
    - feedforward
    - hidden-gesture
  variantOf: []
  partOf: []
  seeAlso:
    - visual-hierarchy
    - overloaded-command
    - easter-egg
implementations: []
sources:
  - title: "Supercharge: UI glossary"
    url: https://supercharge.design/blog/ui-glossary-list-of-essential-ui-design-terms
demo: inline
exhibit: false
useWhen: the cue that tells you an action is possible
---

Don Norman introduced signifier to clean up a mess he had made. The Design of Everyday
Things borrowed affordance from perception psychology, designers took it to mean the
visual hint rather than the action possibility, and by the 2013 revision he had split the
two: the [affordance](/affordance) is what the thing allows, the signifier is what tells
you so. The distinction earns its keep because on a screen the possibilities are all
identical. Glass affords touching, everywhere, equally. Every difference a reader can
perceive between a button and a paragraph is drawn, which makes the signifier the entire
design surface.

A signifier is anything perceivable that communicates where the action goes: the raised
edge and fill of a button, the underline under a link, the grip dots on a drag handle, the
cursor that becomes a hand or a resize arrow, the partial row peeking at the bottom of a
list that says the content continues, the chevron that says this opens. Some are physical
inheritances (a door's flat plate says push), most are pure convention, and convention is
not a weaker kind of signifier. It is the strongest one available, because the reader
arrives already fluent in it. That is why redesigning a link to not look like a link is
expensive in a way that has nothing to do with taste.

Norman's own examples run the other way too. Worn paint on a door, a footpath beaten
across a lawn, a queue of people at a counter: none of these were designed to communicate
anything, and all of them tell you what to do. Interfaces have the same accidental
signifiers, and they are usually bugs. A cursor that stays as an arrow says a control is
not clickable. Text with no underline says it is not a link. Uniform padding says two
elements are the same kind of thing. Readers act on these signals whether or not anyone
intended them.

Which is what makes the failure worth naming precisely: an affordance without a signifier
is invisible, and a signifier without an affordance is a lie. Flat design spent years
rediscovering the first (borders and fills were doing work nobody had costed), and the
second turns up as a false affordance every time text is styled like a link but does
nothing, or a static image is dressed as a play button. Hover only signifiers are both
failures at once, since roughly half of readers have no pointer to reveal them with.