Gaver’s technology affordances article reminded me about Norman’s quite recent rants on how the concept of affordance has been greatly misunderstood in the HCI and design communities. Thus, I reread the short Interactions articles:
- Donald A. Norman. 1999. Affordance, conventions, and design. interactions 6, 3 (May 1999), 38-43.
- Donald A. Norman. 2008. THE WAY I SEE IT: Signifiers, not affordances. interactions 15, 6 (November 2008), 18-19.
Norman is the one who officially introduced the term/concept affordance to HCI and design fields in his book The Psychology of Everyday Things so he’s saying on the topic might have more weight than others’ opinion. Norman states that designers should mainly think about perceived affordances, i.e. what actions users perceive to be possible. In practice perceived affordances are visual feedback that advertise the affordances. By this definition Norman separates affordances and perceived affordances from each other. This is also where Norman and Gaver take the different paths. Gaver has no problems in stating that an icon or other user interface component in a display can have affordances while Norman explicitly rejects this idea. Actually Norman and Gaver utilize the same example of scrollbars in GUIs and end up with opposite opinions. To Gaver a good scrollbar affords scrolling while to Norman using a scrollbar is a cultural, learned convention. Because of this difference Norman and Gaver also see the usefulness of the concept in design quite differently.
In the more recent column Norman goes even further and suggests that instead of affordances the designers should think of signifiers. Norman defines a signifier as “some sort of indicator, some signal in the physical or social world that can be interpreted meaningfully”. Norman’s signifiers are broader than perceived affordances since for example a lack of something (Norman gives an example of lack of people on a train platform signifying that the train has already left) can be a signifier while perceived affordance is a property of a physical object. Here again the mismatch between Norman’s and Gaver’s thinking is visible. Gaver emphasizes that affordances link action and perception and thus one thing can afford different actions to different people (e.g “climbability” of stairways). Norman suggests that affordances are more universal properties of physical objects and instead of affordance we should focus on conceptual models, constraints, conventions and signifiers.