Talk:Affordance

Affordance as an uncountable noun
I disagree with the edits made by Roconnell2580 on 23 January 2007 at 04:59 (see changes here). The change note was "Corrected the use of the plural ('affordances') and added an explanation for usage of the term."

Roconnell2580 writes: "While the term 'affordance' is a noun, it is not a 'thing'—it is essentially an attribute of a thing. Therefore one would not refer to the 'affordances of a web page' when speaking of navigation or behavioral elements like links and buttons. One refers to the links and buttons as elements which have either poor or excellent affordance." I don't think this is correct. The uncountable noun "affordance", signifying an attribute, is an additional meaning for the word, derived from Norman's countable use of "an affordance" meaning "a perceived action possibility". Gibson's use as a synonym for "an action possibility" is also countable.

The paragraphs edited/added are now inconsistent with the section about Gibson's original definition. The section about Norman's term is also made incoherent because it seems like a global search-and-replace was used, without fixing the grammar of the sentences.

I'd like to suggest reverting those changes but adding a paragraph mentioning the uncountable use of the noun.

Any objections?

Forlornturtle 11:53, 20 February 2007 (UTC)
 * Sounds perfectly reasonable. The article certainly needs work.  Also, does anyone else find the last paragraph unclear? --Ronz 17:23, 20 February 2007 (UTC)


 * If you're referring to the paragraph that starts with the quote I used above, I see what the author is saying, but it only makes sense within the context of one of the meanings of the word "affordance". Also, I don't think it is very objective; the design advice given could be argued both for and against, depending on your design philosophy. Forlornturtle 17:54, 20 February 2007 (UTC)


 * I edited the article based on my above suggestions on 3 March 2007 at 19:22. Cheers. Forlornturtle 09:46, 28 March 2007 (UTC)

Too much a HCI perspective?
"However, the definition from his book has become established enough that both uses have to be accepted as convention." I do not think this statement is right in all fields. I think in cognitive science, the term affordance is perfectly acceptable, as referring to the term affordance as coined by Gibson. I would suggest changing it into: "... established enough in the field of Human Computer Interaction that both ...". Any better suggestions? Maybe it even better to have a subparagraph dealing with Affordances as something with a slightly different meaning in HCI. I would prefer to have the description here more closer to the original term, and less emphasize on Norman's error. - murb: 10:50, 26 March 2007 (UTC)


 * That makes sense. I've edited the sentence accordingly, and added a clarifying sentence at the end of Gibson's definition. However, I think it's important to keep equal focus on both meanings and to reflect on the discrepancy, in order to clear up any confusion and avoid misunderstandings. Forlornturtle 10:09, 28 March 2007 (UTC)

Norman's definition is a distortion needs to be acknowledged as such. Norman really should have used a different term to describe his notion of this concept, as it is very much NOT what Gibson was getting at. As well, and more crucially, there is a problem with aligning Gibson to any form of "cognitive" anything (cog sci). Gibson was devout anti-cognitivist (see "the textbook Gibson" paper https://www.semanticscholar.org/paper/The-%22textbook-Gibson%22%3A-The-assimilation-of-Costall-Morris/b0f95b68a375e1bc49d182905aae14c2debca538). So the reference to "cognitive science" is a misnomer here as Gibson was rejecting cognitivism whole cloth, and by extension, neurocentrism. All those who wish to understand Gibson's meaning should undertake a close and careful reading of his work, and not attempt to assimilate his ideas into others paradigms - often paradigms he explicitly rejected (which would be known from a close and careful reading of his work!). 104.232.202.201 (talk) 15:28, 5 January 2021 (UTC)

Edits by 158.36.83.173 on April 30, 2007.
The change made to the introductory sentence seems to ignore the dual-meaning of the word affordance, focusing only on Norman's adaptation. The previous, now-restored version uses an intentionally ambiguous definition in the first sentence, leaving the full explanation of both meanings until later. Please read the sections History & Definitions and Consequences of the duality of the term. If anyone can think of ways to make the situation clearer in the introduction, please do try to improve it. Thanks. Forlornturtle 12:22, 2 May 2007 (UTC)

Section "Example on analyzing using affordance"
I'd like to suggest removing this section, because, apart from its essay-like style, I don't think it adds much value to the article. The example under "History and definitions" involving the recliner and the softball, although not as elaborate, serves the same purpose and is, in my opinion, sufficient to illustrate the idea. It also covers both Gibson's and Norman's "affordances", while this section disregards the duality altogether. Any thoughts? Forlornturtle 21:41, 20 July 2007 (UTC)
 * I concur. -- Zim Zala Bim ( talk ) 21:50, 20 July 2007 (UTC)
 * Done. Forlornturtle 15:33, 30 July 2007 (UTC)

Is an affordance really an action?
The article starts, "An affordance is an action that an individual can potentially perform in their environment." Is that true, that an affordance is an action? According to Gibson, an affordance is a possibility, and according to Norman, an affordance is a property. (See http://www.interaction-design.org/encyclopedia/affordances.html) Both of these are categorically different from the Wikipedia definition. ("Walking" would be an action; "walkability" would be an action possibility)   So it seems to me that Gibson and Norman have two different definitions, and the Wikipedia definition doesn't match either of them. Billgordon1099 (talk) 18:14, 28 November 2007 (UTC)
 * You're correct. Article changed accordingly.  --Lockley (talk) 20:52, 28 November 2007 (UTC)

In the edit of 23:42, 23 February 2010, Roberto the Benevolent has changed it back to the old and questionable definition: http://en.wikipedia.org/w/index.php?title=Affordance&action=historysubmit&diff=345982933&oldid=345982406 I've changed it back to the version before this edit. --118.109.96.176 (talk) 15:20, 3 April 2010 (UTC)

Yes Gibson and Norman are different. Norman's idea is a distortion and he does not acknowledge this and Gibson's is the true definition. The real problem is that folks are not reading Gibson's work AND they are not *understanding* it (either). Gibson coined the term and his own definition is the correct one. In a nutshell, an affordance is not given but CONSTITUTED and by many factors: the environment, the perceiver and the reciprocal *relationship* between them. Gibson argued that an affordance was neither an innate property (this is where Norman is fully wrong) nor is it something that is merely 'perceived.' Bear with me, this is just a hasty plain language summary of some of Gibson's main arguments from https://www.amazon.ca/Ecological-Approach-Visual-Perception-Classic/dp/1848725787 [find an online sample of the introduction if possible].

What makes something an affordance is the perceiver's consideration and use of it as such. Thus it is constituted in and through action. Not merely observation, although observation is a component of deducing the possibility for action. A bird may perceive a branch as a location to perch and fly to it with the intention of landing. If the branch is firmly attached to the tree it will afford perching. However if the branch is broken (and the bird does not perceive this) that does not mean it is a 'false' affordance (as another part of this article wrongly suggests), but merely that through the bird's actions and interactions it was not constituted as such. An elephant cannot perch on a branch, thus affordances are also relative to the animal who attempts to use it for some purpose. If it 'affords' something - even an action the perceiver may not have anticipated, it is thus constituted as an affordance. An example is sunlight (right from Gibson). Sunlight (or any illumination) affords seeing (Gibson was clear to argue that sight was not merely about optics or the physiology of vision, but contingent on the environmental affordance of light along with the physical position of the head and body). A cat might select a spot in front of the window to rest because it is illuminated by warm sunlight, based on a momentary choice to be in the sunlight. It may subsequently choose to relocate to a dark shaded location (under a couch) for homeostasis. These 'decisions' are contingent on the environment as well as the animal, in the moment, and its needs and abilities. If an animal or human is incapacitated - merely perceiving an affordance won't make it such if they cannot act. It has to be possible to use it and the use is entirely contingent on all of the factors above: perception/observation, environmental contingencies, the identity/situation of the animal/actor and the relations between all of these variables. Only if all of these variables are working together can anything be constituted as an affordance and Gibson was very clear on this. 104.232.202.201 (talk) 16:03, 5 January 2021 (UTC)

Deleted "Further reading" section (for now)
The only links that seemed both authoritative and general enough (by Gibson) were broken.

This section should not be a catalogue of affordance-related papers, which are often quite specialised and/or present novel ideas whose merit is yet to be proven. It should point to authoritative sources where you can read more about the general subject. Forlornturtle (talk) 11:19, 9 April 2009 (UTC)

Add a section with current formalisms
The first would be Turvey (1992) Affordances and Prospective Control: An Outline of the Ontology. The agreed on portion is that they are relational in the sense that an affordance is what refers to both the animal and it's surrounding environment. Affordance = <> in a sort of tuple. Turvey makes a fair start, and Chemero(2000&2003) and Stoffregen(2000) carry the torch. Alenarcic (talk) 23:38, 18 March 2010 (UTC)

Further Shift of Meaning
117.192.227.23 (talk) 01:28, 6 September 2011 (UTC):

The references supporting "The different meanings now associated with the word, although closely related, can be a source of confusion in writing and conversation if the intended meaning is not made explicit and if it is not used consistently. Even authoritative textbooks can be inconsistent in their use of the term." do not seem to be appropriate...

"In Human–Computer Interaction, Preece et al. (1994, p. 6) explicitly define perceived affordances as being a subset of all affordances, but the meanings are intermingled later in the same paragraph by talking about "good affordance". - This actually seems to support the last point of the previous section - "However, the definition from his book has become established enough in HCI that both uses have to be accepted as convention in this field."


 * The point is that they're being inconsistent: they're defining the term to mean one thing (Gibson's action possibilities) but then use it to refer to a very different concept (the discoverability of action possibilities), without having explained the multiple meanings. Forlornturtle (talk) 19:31, 17 April 2013 (UTC)

''In Universal Principles of Design, Lidwell, Holden & Butler. (2003, p. 20) first give the example of round wheels better affording (i.e., allowing) rolling than square ones, and later that of a door handle "affording" (i.e., suggesting) pulling. - I have not read the book, but the use of the term affordance from the line cited in "the example of round wheels better affording rolling than square ones"'' can also imply 'suggest' (i.e. round wheels better suggest rolling than square ones) rather than 'allow'


 * Thanks for pointing that out. The full quote is actually "Round wheels are better suited than square wheels for rolling; therefore, round wheels are said to better afford rolling." I've updated the footnote to clarify this. Forlornturtle (talk) 19:31, 17 April 2013 (UTC)

Section "The space of affordances"
I can't say that I really understand what the point of this section is. I think it's pretty incoherent and doesn't fit into the flow of the rest of the article. At the very least its language is too technical. It was inserted by an anonymous user and has never been edited.

Unless anyone objects, I'll remove this section soon. If you see its value and want to rewrite it, feel free to dig it out again.

Forlornturtle (talk) 21:06, 16 April 2013 (UTC)
 * The references seem a good addition. If you remove the section, please place a copy of it here at the talk page. Diego (talk) 05:46, 17 April 2013 (UTC)


 * OK, here it is:

"Previous studies have shown a motor recruitment during the observation of graspable objects. This recruitment has been considered crucial in encoding the observed objects in terms of one or more potential motor acts. However, an agent can actually act upon an object only when the latter is close enough to be reached. In a behavioural study, Costantini et al (2010) have recently demonstrated that an objectual feature (i.e. the oriented handle of a mug) might suggest or even demand a specific motor act (i.e. a reach-to-grasp motor act) only when the object is presented within the observer’s own reaching space. Subsequently, in a TMS study, Cardellicchio et al. (2011)  provided the neural counterpart of previous behavioural data. Their results shed new light on the functional role of the motor system in encoding visually presented objects. Indeed, they indicate that its recruitment is spatially constrained, as it depends on whether the object falls within the actual reaching space of the onlooker providing that the recruitment of such motor representations does not exclusively depend on the visual presentation of the affording features of an object, rather it primarily relies on the actual reachability of the object, i.e. when it is literally ready-to-her/his-own-hand. Furthermore, another TMS study of Cardellicchio et al. (2012) assessed whether and what extent the presence of another individual might impact on the observer's processing of affording features, as measured by the excitability of the primary motor cortex. The result showed that the mere sight of an object located outside the reaching space of the observer but within the reaching space of another individual evoked a suitable motor response similar to that afforded by an object falling within the observer's reaching space."


 * Forlornturtle (talk) 19:00, 17 April 2013 (UTC)

June 2016 cleanup
I've made some major edits because this article has become overly convoluted and technical over the last few years. There have been many additions that sound too academic, whose value is unclear or which weren't well integrated with the surrounding text.

I made the introduction super-simple, focusing on the basic definition and pointing out that there are different meanings. I'd encourage everyone to keep specifics to the other sections. There were a couple of statements which were too technical and detailed for an introduction, and didn't seem to add much information compared to what's in the rest of the article, so I removed them. If you feel they contained useful additional information, please try to work it into the rest of the article in a readable way. Several references were deleted as part of this, but they just seemed to offer additional definitions, which I feel are already well covered.

In the section As action possibilities, I removed two paragraphs about academic work. I think the points these were making are secondary to the topic and not needed in this article. They were also not well integrated into the text. Again, if you feel they added value, please work it into the text in a simple and clear way, or let's discuss it here.

I also reintroduced a part about the uncountable use of the noun and about a different meaning for the verb. These had been deleted (fairly) because they lacked references, which I've now added. I personally don't like these new meanings, but in reality people do use them, and I think the article should capture this.

Forlornturtle (talk) 10:43, 5 June 2016 (UTC)