Reclaiming Affordances

The next time you're about to use the word “affordance”, please stop and check if the word “cue” would work instead.Because if it would, then:

1. you are using an obscure technical term for something that already has a perfectly good plain English word, and

2. you are using that technical term incorrectly.

Yes, I know that languages are living entities. None other than the eminent Don Norman, despairing in his attempts to correct the misuse of “affordances”, has cited this as a reason to abandon the term to its abusers. Because after all, words can change their meanings. Generally, I celebrate this fact. But not this time.

Technical terms are different. People can start calling air “Oxygen”, but that does not mean that scientists should change the periodic table. And this is an important technical term. It describes something very specific for which there is no other word. When J.J. Gibson went to the trouble of making up the word “affordances” in 1977 he thought long and hard about it first, and he coined it carefully so that it  makes sense. An affordance is “something that is afforded”.

“The affordances of the environment are what it offers to the animal, what it provides or furnishes, either for good or for ill” [1]

Affordances are not the cues in the environment that let you know that something is afforded to you. Affordances can be hidden, or perceived. Designers can make them easier to perceive by adding cues. Cues can lie, but affordances cannot: they are what actually is possible. Importantly they are also relational – they exist in the relationship between an environment and its user. A matchbox affords support to an ant but not to me.

I've attached an image of a green child's sippy cup to this article. To a toddler the handles on either side afford gripping with both hands at once, making it easy to tilt the cup to their face. It affords drinking. It also has a screwtop which the toddler cannot operate – no affordance for them there. But to me it does afford opening, so that I can refill the cup. This is great design, using the relational quality of affordances to promote some behaviours and constrain others, depending on the capabilities and aims of the different users.

People who are thinking and talking about design need a word for what the environment actually makes possible for a particular user in the context being described, and affordances is that word.

Finally – there is no reason to misuse the term “affordances” when you really mean “cues”. If you hear someone else misusing it (and I don’t care if it is Jared Spool), then call them out. We need this word back, and if language is indeed made by its users, well that is us. We can make the choice to use this word well, or to waste it.

Thank you for your attention.

V.

[1] Gibson, J.J., 1977, “The Theory of Affordances” in Perceiving, acting, and knowing: toward an ecological psychology
eds. R. Shaw, J. Bransford, Lawrence Erlbaum Associates, Hillsdale, NJ.

Jared M. Spool says:
November 21, 2009 at 9:54 am

Hi Viveka,
I believe that the rules of lexicography still allow a single word to have multiple definitions.

Affordances in the Gibsonian sense are very important.

However, interface design seems to want a word that isn’t cue (or clue, which is the layman alternative I tend to use) to be specific to the use intention. A cue is from the designer side, whereas an affordance is from the user side — if the user doesn’t recognize the cue, it’s not a good affordance.
So, I’m not sure what all the fuss is about. I think there’s still room in the dictionary to allow both definitions to co-exist.

Even if I am Jared Spool.
:)
Jared

viveka says:
November 21, 2009 at 3:35 pm

Hi Jared, thanks joining in :)

I agree that there's no point appealing to rules of lexicography. English lexicographers (unlike French ones) are descriptive, not prescriptive. They no longer have any special claim to authority over the use of language; English is in the hands of its users. That’s fine with me. I’m not a lexicographer, I’m a user of language.

I’m an interaction design practitioner and researcher, and I use and need “affordance” often. The difficulty is that when I try to use this very specific, obscure technical term my meaning is further obscured by the fact that some of my colleagues are not clear on what it means; or if they are they don’t know if I am clear, so I have to define it every single time I use it. This is tiresome. Fields of endeavour have jargon and specialised terms in order to speed things up when we talk among ourselves. Jargon can also be used to bamboozle outsiders, but I consider that a misuse.

Technical terms survive if they are useful. To have two separate meanings for a neologism that was invented fairly recently for a specific purpose makes that neologism less useful within its community. It makes it fabulously more useful for bamboozling, but again I don’t think that’s a valid argument.

So, let's try the experiment I suggested at the top of my post – replacing every use of affordance with “cue”:

You wrote: “if the user doesn’t recognize the cue, it’s not a good affordance”
Let’s try: “if the user doesn’t recognize the cue, it’s not a good cue”.
Yes! That makes perfect sense, it doesn’t require you to explain a new word to laypeople or to argue to experts that you don’t mean “Gibsonian” affordances when you use the word that Gibson made up, first used in an article by Gibson called “The Theory of Affordances” which contains the phrase “I made it up”.

Interface design wants a word that isn’t cue, or clue? Why not? Because we want to seem important? You can say “a cue for an affordance” if you must; sometimes it’s fun to point out that we know cool words that our clients don’t. But I have never yet found a sentence where “cue” wasn’t sufficiently clear.

So to be perfectly clear where I’m coming from here: get off my damn lawn ;)

Jared M. Spool says:
November 22, 2009 at 3:40 am

You seem to be having an issue I don’t have.

I’ve heard this argument since Norman started using the term. And yet, I’ve never once fallen victim to the confusion you describe.

So, feel free to use affordances in a different way than I use them. And, when we meet, you can feel free to correct my misuse.

After all, it’s not the first word that Australians use differently than Americans.

:)
Jared

viveka says:
November 23, 2009 at 12:20 pm

I think I get where you're coming from. If you talk mostly to lay people or novices who have never heard the word before, and you (mis)define the word for them at the start of the conversation, then I can see how you'd be sanguine about this. I run into the problem because I work with designers (who have picked up the "cue" misdefinition) and cognitive psychologists (who stick with Gibson's original meaning).

In the 2005 article I link to above though, you imply that cognitive psychologists say that “Affordances are clues”. They really don’t.

As well as HCI practice I am doing cross-disciplinary research. “Affordances” is used by people in perceptual psychology, cognitive psychology, environmental psychology, industrial design, human/computer interaction (HCI), interaction design and artificial intelligence. I have to talk to people in all these fields about my research. All of them use the word as defined by Gibson in 1977. Only a subset of HCI practitioners and early career researchers use it in the way that you prefer.

The difference is stark and yet it’s subtle. Affordances are real and underlying; cues are the surface indicators of those affordances. It’s easy to get essence and surface confused, but if you care about good design then it’s vital to separate them.

By using a word that refers to something essential as though it referred to the surface indication, HCI practitioners reveal ourselves to be shallow, concerned only with appearances and not with how things actually work. This is not something to be meekly accepted; it is something to fight.

Jared M. Spool says:
November 24, 2009 at 12:50 am

I’ve gone ahead and updated my Drag N Drop post to acknowledge your objection.

I still don’t care enough to change my usage of the word, but appreciate that you’re out there fighting the good fight for, um, something…

Jared

Clifton says:
November 25, 2009 at 7:58 pm

Love it. Completely agree, “affordance” is generally overused when “cue” is what we’re trying to say.

If a user doesn’t recognize a cue, the affordance remains. I recently discovered that the vent on my dad’s air conditioner can be lifted up to shift the angle of the vent. In the 20 years he’s owned the unit, he never knew that was possible. The affordance–the ability to be moved and manipulated–has been there for 20 years. But there wasn’t a sufficient cue for my dad to recognize the affordance. By explaining it to him, I provided another cue, in the form of words and a demonstration.

I love how many UI stories involve a product owned by the author’s parent.

viveka says:
November 25, 2009 at 9:33 pm

Thank you Clifton. And thanks to Jared for being so gracious under fire.

One nice thing about affordances in interaction design is that sometimes their function is apparent, so all you have to do is point them out or reveal them. The examples given in Norman’s Psychology of Everyday Things [alternative kickback-free link] are like this: slots, handles. If you can see the slot, then you can perceive that it affords insertion of something slot-sized. If you can see the handle you can perceive that it affords grasping, especially if it has a grip texture. On the other hand to put a grip texture on an on-screen scroll thumb is quite different: the designer is taking advantage of the cultural convention that arose from our perception of the true affordance in the real world.

My own favourite example is the sippy cup, showing how the same physical property of an object can be an affordance or a constraint (the very opposite) depending on who is interacting with it.

David Hamill says:
November 25, 2009 at 9:44 pm

I can’t see what the fuss is about. Unless use of the word is leading to miscommunication of a point then there isn’t a problem. Language is what we use to communicate with each other. If using the word ‘affordance’ has had this effect then job done. I don’t care if you start using the word ‘strawberry’ in its place, as long as people understand.

A ‘clanger’ is a mistake, but to UK adults over a certain age it’s also a small pink mouse-like creature that feeds off blue string pudding.

viveka says:
November 26, 2009 at 11:11 pm

I agree that words can mean whatever we want them to mean. There is no objective truth about what a word should mean. That doesn’t invalidate all value judgements though: we can judge the worth of a definition on whether it makes it easier or harder to communicate.

In this case, the existence of the extra similar-but-crucially-different use is indeed leading to miscommunication. For example, I am studying the effect of place on collaborative creativity. In my research I have found good evidence that the surface appearance of a place is not as important for some kinds of creative work as the affordances of that place. I want to use the term “affordances” because it is far quicker than saying “the things that the environment provides or makes possible for each participant, for good or ill, and dependent on the capabilities of that participant and their relationship with the environment”. However if people think that “affordances” means “surface appearance”, then my point makes no sense to them.

Elizabeth Buie says:
November 27, 2009 at 1:19 am

In talking about the flexibility of meanings, we need to consider whether we are talking about language in general or terminology in particular. Degrading a technical term is much more harmful to clear communication than is the ordinary evolution that happens to everyday language.

In 2002 I co-wrote an article for <interactions> called “What’s in a Word: The Semantics of Usability”, in which we argued exactly this. Here’s a quote from our article:

========

Word meanings can change over time; that’s all well and good. “Awful” used to mean “full of awe”, and seven centuries ago “nice” meant “foolish” (some might say it still does!). The rich garden of English vocabulary has grown from the endless planting of new words from foreign sources, jostling for their place in our prose and poetry. A living language is always on the move – good thing for us.

There’s a bit of a problem, though, when we want a word to conjure up something more concrete than a poetic image in the reader’s mind. When we want to use a word as terminology. Terminology is to vocabulary as bulldog is to kennel: It demands a certain kind of care. We want a term to hold its value. It has to say the same thing to everyone who needs to read or hear it.

========

Michael says:
November 25, 2009 at 9:59 pm

viveka,

I was brought here by tweet from Jared. I think he should be commended for bringing this discussion to over 9000 people.

I had no idea there was such a history behind this one word but have always felt “cue” was an underutilized word when describing, well “cues”. I’ve had many discussions when people who say affordance and needed to say cue. Now I have a better grasp for the argument.

Thanks for the lesson.

Andrew Ingram says:
November 26, 2009 at 12:51 am

It’s a common occurrence in any environment for people to adopt words without using their correct meaning, and it’s incredibly frustrating – to the point that you have to clarify definitions at the start of each meeting to avoid confusion.

I think it’s very important to only introduce technical terminology when everyday language is insufficient, and to use technical terminology consistently.

Your cue/affordance example is very good, for one thing it made the sentence much easier to understand.

Bob Kerns says:
November 26, 2009 at 3:40 am

Viveka,

I hope this doesn’t spread the focus of the debate too thinly, but here goes:

So tell me, if you substitute the word “feature” for “affordance”, what happens? Except there are also “misfeatures”. Which would correspond to your “constraint”, I think, except the original definition of “affordance” would appear to encompass both. Clearly, I’m using the term “feature” here in the sense it has come to have in the software world.

I’m just wondering if the drift in meaning isn’t a result of a lack of a strong need for the original meaning, to most people.

The distinction may be more useful in its original context, of things in the physical environment. In the software world, we are more concerned with people getting specified tasks done. We’re less likely to use a wrench as a hammer because somebody borrowed the hammer and didn’t put it back — even though a heavy wrench affords a hammer-like action.

There’s also the phrase “better affords”. Does this mean, it does it better, or it better communicates? I do NOT like the term cue as a substitute for the (mis)use of affordance. It has the serious problem of blurring the boundary between “telling the user what to do”, and “communicating the existence of an affordance (your usage)”. Both are things that are more commonly in need of discussion than your meaning of affordance — especially the communication aspect.

That’s because, if there’s no communication, the affordance is undiscovered, theoretical, futile. A lot of design activity is around communicating or efficiently presenting the affordances that are there. The affordances are often a given; the design activity neither adds nor removes them.

I think the need to discuss affordances largely centers around FAILED affordances. A lot of Norman’s book is on that, as I recall. The affordance is there, waiting to be used — but nobody uses it. Why?

A truly successful affordance often becomes synonymous with the name of the object. A hammer “hammers” things. Or the object is named for its primary affordance.

My intent here isn’t to argue for one side or the other, actually. I’m not a French lexicographer! I just think these are reasons you’re fighting an uphill battle.

BTW, I like your approach the the kickback-link issue. It clearly provides and communicates the non-kickback affordance, without needing a separate disclaimer, explanation, etc. I’m going to steal the idea. Not that I’ve ever made a dime from such links, but I can dream…

:)

Disclaimer: I used to work with Jared decades ago.

viveka says:
November 26, 2009 at 11:40 pm

Hi Bob,

Thanks for the thoughtful input, I’m enjoying this discussion immensely.

>There’s also the phrase “better affords”. Does this mean, it does it better,
>or it better communicates?

I think the beauty of the term “affordance” is that it gives us a way to talk about what’s really there: whether it’s helpful or unhelpful, visible or invisible. Accordingly I would use “better affords gripping” for a handle that has better grips in actual use. I would say “better communicates that it affords gripping” for a handle that has more visible grips.

Yes, communication of function is important – an invisible affordance is useless. However I argue that keeping our terms for communication separate from our terms for underlying affordance helps us think more clearly about exactly what is going on. Is the affordance broken, or is it poorly communicated? Is it the wrong affordance? Do we want to highlight one affordance rather than another?

On “feature” for “affordance” – I think they’re different as well. A feature connotes something useful that you want to highlight for users. Something you want to “feature” in the marketing materials. Some affordances may be features. Some constraints (e.g. password protection or parental controls) might also be features. A bad affordance can be a misfeature, as can a bad constraint.

James Landay says:
November 26, 2009 at 9:52 pm

I appreciate your effort at correcting our misuse of this term. But, “Only a subset of HCI practitioners and first-year researchers use it in the way that you prefer” is a bit strong… no, make that very strong. I have been a researcher in this field for almost 20 years, teaching introductory and advanced graduate courses for the last 13, and I have been defining it incorrectly all this time! I got my definition from a professor I assisted in teaching HCI at Carnegie Mellon. So, this means there are quite a few people running around using it incorrectly… Uh, sorry?

viveka says:
November 26, 2009 at 11:21 pm

The apology should be mine. Since I wrote this I’ve found a number of long-standing HCI practitioners who have been using “affordances” to mean “cues” or (close to correct but still not quite right) “perceived affordances”.
I still think it’s a misuse for the reasons I’ve outlined, but I can certainly understand how it came about. I’ve struck through the overstatement in my above comment accordingly.

I’m singularly impressed with people such as James who are happy to reconsider a long-held belief when presented with a reasonable counter-argument. I hope that I am as gracious when, inevitably, the same happens to me.

Gerard says:November 27, 2009 at 2:47 am

This is certainly a long-standing debate. For an argument of why the Gibsonian definition of affordance is still important, and why changing the definition is unfortunate for our profession, check out:

Torenvliet, G. L. (2003). We can’t afford it: The devaluation of a usability term. interactions, 10(4 (July/August 2003), 12-17.

Email gerard /at / torenvliet / dot / ca for a PDF.