Affordances, Signifiers, and Constraints
Every designed thing, an object, a process, an org chart, a policy, has an actual set of actions it makes possible (its affordances), cues that tell people where and how to act (its signifiers), and features that rule certain actions out (its constraints), and all three can point in different directions from what the designers intended.
What Is It?
Psychologist James J. Gibson developed the concept of affordances within ecological psychology, introducing it in “The Theory of Affordances” (1977) and elaborating it fully in “The Ecological Approach to Visual Perception” (1979). For Gibson, an affordance is a real, relational property between an environment and an actor: a set of stairs affords climbing for a person in a way it doesn’t for a mouse. It’s a relationship that depends on both sides, the environment and the actor, not a fixed property that belongs to the object alone, and it exists whether or not anyone notices it. Donald Norman adapted the concept for design in “The Psychology of Everyday Things” (1988). His early usage helped popularize the idea that good design should make action possibilities easy to perceive, but this also caused the term “affordance” to get stretched to cover something closer to a visual cue than Gibson’s original concept. Norman spent much of his later writing, including the revised “The Design of Everyday Things” (2013), clarifying the distinction: an affordance is what’s actually possible; a signifier is a perceptible cue that communicates where or how to act. The two aren’t the same thing, and a design can misalign them.
The classic example is a door. A door’s physical properties may genuinely afford pushing, or grasping and pulling. A flat plate mounted on the door is a signifier, it tells you to push. A handle is also a signifier, telling you to pull, and it typically affords grasping regardless of what the door actually requires. When a door has a handle but needs a push, the problem isn’t that the door is “lying” about its affordance, it’s that the signifier is pointing at the wrong action. That mismatch, common enough that “Norman door” has become shorthand for it, is precisely why keeping affordances and signifiers separate matters: a structure can genuinely make an action possible while its cues tell people something else entirely.
Constraints are the third piece: features of a design that narrow down the range of actions available, so the intended action is easier to find and the wrong one is harder or impossible to take. Norman identified four kinds. Physical constraints make an incorrect action literally impossible, a key that physically fits a lock only in the correct orientation. Semantic constraints rely on a design element’s meaning, a rear-view mirror’s shape implies how it should be mounted and read. Cultural constraints depend on shared convention, red conventionally means stop. Logical constraints use the relationship between parts to imply what’s wrong, leftover screws after an assembly imply a mistake somewhere in the process without needing a label to say so. Good constraints don’t necessarily feel effortless or invisible, a safety interlock is deliberately restrictive by design, but they reduce the space of possible actions on purpose, making errors harder and the appropriate action easier to discover or execute.
Why Does It Matter?
A process, an org chart, a tool, or a policy is not neutral scaffolding sitting underneath an organization’s stated intentions; it has its own affordances, signifiers, and constraints, regardless of what leadership says it’s for. A five-signature approval chain doesn’t just slow spending down: it affords repeated review, constrains unilateral spending, and increases the effort required to act quickly, caution is a plausible behavioral consequence of that design, not the affordance itself. A meeting in which speaking time goes to whoever jumps in first and interrupts most easily affords repeated intervention by assertive speakers while constraining contribution from people who wait for an explicit opening, whatever the meeting’s stated ground rules say. A policy might genuinely afford escalation, a real channel exists, while its constraints require three layers of approval and its signifiers, the process is labeled “exception handling”, quietly communicate that using it is unwelcome. Those are three different design effects worth distinguishing rather than collapsing into one vague sense that “the process discourages X.”
This is a distinct diagnostic lens from simply asking what a structure incentivizes: an affordance, signifier, or constraint operates on what a structure’s shape makes possible, visible, or ruled out, which can diverge entirely from formal incentives. A policy can technically reward a behavior that the actual workflow makes so effortful or so poorly signposted that almost no one takes it. Worth adding: the same structure doesn’t necessarily afford the same thing to everyone in the room. A senior leader and a new hire may sit in the identical meeting, but the social environment affords challenging a decision very differently to each of them, even though nothing about the room or the agenda changed.
What Changes Once You See It?
You start reading a process or a structure for what it actually makes possible, signals, and rules out, rather than taking its stated purpose at face value. You get a sharper way of diagnosing why a well-intentioned policy keeps producing the wrong behavior: often the honest answer isn’t that people are ignoring the policy, it’s that the surrounding structure makes the intended behavior technically possible but hard to discover, cumbersome to perform, or blocked by some other part of the system. You also start noticing your own organization’s Norman doors, places where the intended action and the structure’s signifiers, affordances, or constraints point in different directions, and where redesigning the structure itself will likely do more than restating the intention ever will.
Common Misunderstandings
- It isn’t the same as intentions or formal incentives. A structure can afford a behavior nobody intended, and can fail to afford a behavior that’s officially encouraged or even financially rewarded; affordances, signifiers, and constraints operate on what a structure’s design actually makes possible, visible, and ruled out, not on what a policy document says should happen.
- It isn’t a claim that constraints are bad and affordances are good. A well-placed constraint is often the entire point, preventing a costly error before it happens rather than relying on someone remembering a rule. The goal isn’t removing constraints, it’s making sure a structure’s affordances, signifiers, and constraints actually match the behavior it’s meant to produce.
- It doesn’t mean structure is destiny. Motivated people routinely act against what a structure affords, signifies, or constrains, going around a slow approval chain or speaking up despite a meeting format that discourages it. The pattern describes what’s possible, visible, or likely at scale, not a hard determinism about any individual’s behavior. It’s also worth keeping “possible,” “easy to discover,” and “encouraged” as distinct ideas rather than treating them as interchangeable, an action can be genuinely afforded while still being inconvenient or poorly signposted.
- It isn’t a separately validated organizational-research finding the way some entries in this Field Guide are. Affordances, signifiers, and constraints are well-established theoretical constructs, foundational in ecological psychology, design, and human-computer interaction, but they’re design principles, not a single replicated experimental effect. The translation into policies, meetings, and org charts preserves the design question, what does this arrangement make possible, discoverable, easy, difficult, or impossible, without claiming that Gibson or Norman’s research specifically established that any particular organizational structure produces caution, dominance, or any other behavioral outcome.
Diagnostic Question
What does this process or policy actually make possible, what does it tell people is expected, and what does it rule out, and how much do those three line up with what it was actually designed to do?
Explore Further
Field Notes
- “Your Org Has Norman Doors”
Related Field Guide
Origin
Psychologist James J. Gibson developed the concept of affordances within ecological psychology, introducing it in “The Theory of Affordances” (1977) and elaborating it in “The Ecological Approach to Visual Perception” (1979). Donald Norman adapted affordances for design in “The Psychology of Everyday Things” (1988), and later clarified the distinctions among actual affordances, perceived affordances, signifiers, and constraints. The revised “The Design of Everyday Things” (2013) treats affordances (possibilities for action) and signifiers (the perceptible cues that communicate those possibilities) as distinct concepts, alongside a four-part typology of constraints: physical, semantic, cultural, and logical.