A doorway is wide enough for one person to walk through without turning sideways, but a person carrying a broad box may need to rotate. The doorway did not change; the relation between its opening and the actor’s effective width did. An affordance is an opportunity for action that exists in a relationship between an environment and an organism with particular capabilities. The idea helps explain why perception is often about what one can do, not merely about naming objects. It also explains why a design that works for one body, tool, or context may fail for another.
An action opportunity, not just an object property
James Gibson introduced affordances within an ecological approach to perception. A surface might afford standing for one animal, shelter for another, and neither for a third. The claim is not that a surface changes its physical composition with every observer. It is that the action-relevant relation depends on the organism’s body and abilities as well as the environment. A low step may be climbable for an adult, challenging for a toddler, and inaccessible to someone using a particular mobility aid. Calling the step simply “climbable” without naming the actor leaves the claim incomplete.
The relational view has several philosophical versions. Chemero argued for understanding affordances through organism-environment relations, while Stoffregen described them as properties of the animal-environment system. These accounts disagree on details of ontology and perception. For practical research, the shared demand is useful: specify the action and the capability. A vague statement that an app screen “has affordances” says little until it identifies what an intended user can actually perceive and do with it under realistic conditions.
An affordance is not the same as a command or an obligation. A bench may afford sitting, lying, placing a bag, or gathering with another person. Some possibilities are physically available but prohibited by social rules. Others may be available only at a certain time, with a particular skill, or when the bench is dry. The concept is richer when these conditions are named. It becomes misleading when one intended use is presented as the only real possibility or when the designer’s expectation is mistaken for the user’s experience.
Affordances can exist even when a person fails to notice them. Conversely, an object can appear to invite an action it cannot support. A painted door on a wall may look passable but cannot be opened. In design practice, the distinction between an actual action possibility and a cue that communicates it is crucial. Donald Norman’s account of everyday design emphasizes signifiers, the perceivable clues that tell people where and how to act. A visible handle may communicate pulling, but its shape cannot rescue a door whose mechanism requires pushing. The environment must both support the action and communicate it well.
What body-scaled experiments reveal
Warren’s stair-climbing study examined judgments and action in relation to body dimensions. The key question was not whether a step had one universally climbable height, but how the height related to a person’s leg length and movement capacity. This body-scaled framing makes a testable prediction: people with different bodies may make different judgments about the same stair. Such studies support the importance of relations between actor and layout. They do not show that every affordance can be reduced to one ratio or that people always judge perfectly.
Warren and Whang studied walking through apertures. Participants changed from straight walking to body rotation as the opening became narrow relative to shoulder width. The experiment gives a concrete example of perception guiding action without measuring an opening only in absolute centimeters. Yet a doorway can be navigated differently with a backpack, an injury, or a heavy object. The original ratio is not a timeless accessibility standard. It demonstrates a method for asking which bodily dimension and task constraints matter in a defined situation.
Action also depends on skill. An experienced climber may traverse a slope that a novice cannot safely use. A skilled wheelchair user and a new user may approach the same ramp differently. Skill can change the feasible set without altering the physical setting, but this should not be used to blame an individual for inaccessible design. There are costs, risks, and social requirements beyond bare physical possibility. A ramp that can technically be climbed with extreme effort may still be an unacceptable route for ordinary use.
Tools extend or alter the actor. A cane changes reach and feedback; a bicycle changes speed and turning radius; a screen reader changes how a digital interface is encountered. The relevant organism-environment relation then includes the tool and the learned way of using it. A doorway wide enough for a walking person may not afford passage with a mobility device. A button visible to a sighted user may be inaccessible without a meaningful accessible label. These are not merely differences in preference. They concern whether the action can be performed under actual conditions.
Perception and action are coupled
Affordance research is often interested in how people perceive possibilities while moving. Approaching a stair reveals information about its rise, surface, and handrail. Trying a grip may reveal whether an object can be lifted securely. Perception is therefore not a static photograph of the world followed by a wholly separate action decision. Movement can sample information, and new information can change the action. This does not mean every affordance is directly perceived without learning or inference. It means an explanation should consider the active relation between perceiver and environment.
A person can misjudge an affordance. They may attempt a gap that is too narrow or avoid a step they could climb. Lighting, fatigue, fear, prior experience, and urgency can affect the judgment. Researchers should distinguish the action possibility from the person’s estimate of it. A successful passage through a doorway indicates that passage was possible on that occasion, but it does not reveal whether the person was confident beforehand. A refusal to try may reflect accurate risk assessment, not a perceptual deficit.
Social affordances add complexity. A public square may afford conversation if people can sit and feel welcome. The benches alone do not ensure conversation; norms, surveillance, safety, and who is allowed to linger matter. Rietveld and Kiverstein describe a rich landscape of affordances shaped by sociocultural practice and skill. This account broadens the concept beyond locomotion, but it also requires careful boundaries. A claim that a space “affords belonging” needs evidence about people’s lived access and interactions, not merely a rendering of the physical design.
Temporal conditions also change what is possible. A crossing may be safe at a quiet hour and hazardous in heavy traffic. A park path may be accessible when maintained and impassable in winter. A digital service may allow an action during normal operation but block it after a timeout. Designing for an affordance requires examining the conditions under which people actually encounter it, including maintenance, weather, time pressure, and failure modes. A one-time usability session may not capture those variations.
Affordances in design
Designers often use “affordance” as shorthand for an obvious button or intuitive interface. This can obscure the original relational idea. A button’s visual style may suggest clicking, but whether a user can activate it depends on motor access, device input, language, network state, and authorization. The intended action may be clear yet impossible, or possible yet hidden. Norman’s distinction between affordances and signifiers helps separate functionality from communication. Good design checks both and tests them with the people for whom the system is intended.
Consider a form with a large, high-contrast “Continue” button. It may afford tapping for a person using a phone, but if the form silently rejects an unfamiliar name or requires a mouse-only calendar control, continuation is not truly available to all intended users. The visible button is a signifier for a workflow that may fail elsewhere. An affordance analysis follows the entire action, including prerequisites and consequences. It does not stop at the appearance of the first control.
In physical environments, a door handle may invite pulling while the door opens by pushing. This mismatch generates error that is often attributed to careless users. An affordance-informed analysis asks whether the form of the handle, hinges, signage, and expected movement agree. It also asks whether a person with limited grip strength can use the door, whether it remains operable in a crowd, and whether emergency conditions alter the action. The design problem is more specific than “make it intuitive.”
Design affordances should not be confused with persuasive triggers. A brightly colored notification can attract attention and suggest a tap, but that does not mean the action supports the user’s goals. An interface may make one action easy while hiding an equally important refusal. The relational question includes agency: what can the person do, at what cost, and with what information? A product that affords enrollment but obstructs cancellation is not simply more usable; it allocates action possibilities unevenly.
Measurement and interpretation
To test an affordance, researchers can manipulate dimensions of the environment and observe judgments and actions across participants with known capabilities. The stair and aperture studies are examples. A design study might record whether users discover a control, whether they can operate it, and whether the outcome matches their expectation. These are different measures. Discovery concerns perceptual access; operation concerns action; outcome concerns whether the environment did what the person intended. Combining them prevents a single completion rate from hiding where the interaction failed.
Thresholds should be described as distributions rather than universal cutoffs. Two people with the same shoulder width may rotate at different aperture widths because their clothing, speed, confidence, and preferred margin of safety differ. Repeated trials may also change behavior. A ratio estimated in a controlled study is an informative relation, not a license to design every doorway at that exact ratio. Accessibility standards and safety margins require broader evidence and public requirements.
Language is another measurement issue. When asked whether a surface is “climbable,” one participant may mean physically possible, another comfortable, and another safe to attempt while carrying a child. A study can reduce ambiguity by specifying the action, conditions, and risk tolerance, then comparing judgments with observed performance where ethical. Without those details, disagreement may reflect different interpretations of the question rather than different perception of the layout.
Affordance analysis can also uncover conflicting action possibilities. A wide open plaza may make walking direct for one person but make shelter or orientation difficult for another. Installing barriers may improve control of traffic while preventing spontaneous gathering. These are not cases where the environment has one true affordance and observers simply disagree. Different people pursue different actions under different constraints. Design decisions must therefore state whose action is being supported and which possibilities are reduced. Empirical observation can identify the tradeoff; democratic and ethical judgment must still decide how to handle it.
People can learn to perceive possibilities over time. A novice may see only a confusing array of controls in a new device, while an experienced user sees a small set of reliable operations. Training can make information more useful, but a design should not require unnecessary expertise for ordinary tasks. The key evaluation is whether the intended user can discover and safely perform the needed action within the expected context. If success appears only after repeated coaching from a researcher, the study has shown that the action can be taught, not that the environment already communicates it effectively.
Affordance language can help teams discuss redesign without pretending that one feature has only one purpose. For example, replacing a heavy entrance door with an automatic one may make independent entry possible for many people while creating new needs for power, sensors, maintenance, and emergency access. The new mechanism changes a network of action possibilities rather than simply adding a single benefit. Field testing should include ordinary days and failure conditions, because a control that works in a showroom may fail during a power outage or when a user approaches at an unexpected angle. A relational account asks whether the promised action remains feasible across those conditions, and for whom.
What the evidence does not show
The evidence does not show that all affordances are immediately or accurately perceived, that one body ratio captures every action possibility, or that a designer can declare an affordance from an object’s appearance alone. Philosophical accounts of affordances differ, and experimental results are tied to specific actors, tasks, and environments. Extending the concept to social settings can be productive, but it requires evidence about norms and access as well as material features. An affordance is not a universal instruction embedded in a thing.
The doorway example makes the practical point. Whether passage is possible depends on the opening, the person’s body, what they carry, how they move, and the surrounding situation. The concept of affordance directs attention to that fit. It helps researchers build more precise experiments and helps designers ask a better question than “Does this look usable?”: usable for whom, to do what, under which conditions, and with what consequences?
Research references
- Gibson, J. J. (1979). The Ecological Approach to Visual Perception. Houghton Mifflin. View source
- Warren, W. H. (1984). Perceiving affordances: Visual guidance of stair climbing. Journal of Experimental Psychology: Human Perception and Performance, 10, 683–703. View source
- Warren, W. H., & Whang, S. (1987). Visual guidance of walking through apertures: Body-scaled information for affordances. Journal of Experimental Psychology: Human Perception and Performance, 13, 371–383. View source
- Chemero, A. (2003). An outline of a theory of affordances. Ecological Psychology, 15, 181–195. View source
- Stoffregen, T. A. (2003). Affordances as properties of the animal-environment system. Ecological Psychology, 15, 115–134. View source
- Rietveld, E., & Kiverstein, J. (2014). A rich landscape of affordances. Ecological Psychology, 26, 325–352. View source
- Norman, D. A. (2013). The Design of Everyday Things (revised ed.). Basic Books. View source