I sat through another lecture about the importance of using metaphors in interface design today, and I reached my breaking point. I’m gonna come right out and say it: I think metaphors get more credit than they deserve in interface design. Yes, I know that many design objects have names that map to real world objects (desktop, trash can, windows, tabs, etc.), but, to reference an idea in today’s lecture, just because the designer uses a real world object as inspiration for a design decision does not mean that the user reaps any benefit from the metaphor.
Let’s think about this logically for a second. What about a desktop metaphor suggests that we overlap windows on top of it? Papers, yes, but windows? What about a computer window mimics the real world object anyway? Are we really suggesting that, when a computer user sees a window on the computer, there’s some subconscious process that says, “Oh, because I’ve interacted with windows in the real world, I know how this computer object should work.”?!?! For that matter, when you see these things on internet forms:
do you really rely on your knowledge of one of these?
Despite the fact that the interaction is the same, I say no, and here’s why. According to Cognitive Psychology, if one object (the metaphor) relies on another to be interpreted, a concept called “spreading activation” will bring the second object to short-term memory (which is why, if I say “fly,” words like “bird,” “bug,” and “sky” might come to mind). Acknowledging that humans are bad at introspection, think back to when you saw the picture of the car radio. Was the correlation natural, or did you think about it? I don’t think it’s a natural leap, but I’m guessing you said it was natural because we are so used to calling the object on top a “radio button.” If you associate the two, I contend that its a learned association through computer use, not natural like metaphor enthusiasts would suggest.
I should say that I’m not against the use of metaphors all together. I happen to use them frequently if my user research implies that it would be helpful. I find that objects that are part of the work (whiteboards, diagrams, etc.) make better metaphors than objects that support work (desktops, file folders). If you want to try a fun exercise in support of my stance, take your laptop to an elderly person who has never used a computer. Let them explore a little bit. After 5 minutes, close or minimize all of the windows, and ask them what they are looking at. Do you really think they will say “a desktop”?
Maybe my stance is too extreme, but it’s in response to this complacence in the HCI discipline everywhere I go that metaphors are universally good things in interface design. Yet, both at UNO and here at CMU, the lectures on metaphors seem to be implying just that. With a bad metaphor, the best you can hope for is performance on the same level as if you hadn’t used a metaphor at all. At worst, you’ve screwed up the user’s mental model even further. To quote my esteemed Project professor, the lecturers should consider hedging.
It sounds like what you object to is less the metaphors themselves and more that they aren’t used effectively.
Metaphors are more than naming objects using real-world names, obviously. And they aren’t to be memorized and recalled by name later. An interface designer has an extremely difficult task - to use an interface to allow a user to accomplish a task and to use the same interface to communicate with the user as to how to accamplish that task. Sometimes it’s a matter of affordances or feedback, but aligning the user’s conceptual model with the designers or even the developers is difficult. I think metaphors can be an extremely effective way of accompishing this mapping.
If users need to keep track of things they have to do, what’s more natural than a task list?
When metaphors become unnatural and when they move beyond the point of communication, I agree with you that they are overrated. Natural metaphors, communicating with design, … all of these can be extremely useful. Not surprisingly, it’s the same line we see in the rest of our lives - moderation.
Then again, it could be my MHCI brainwashing here… :)
Naw, you’re exactly right. My trouble is the message I got from the SAUI lecture on metaphors:
“Metaphors are universally good. Just look at how the Desktop took off. Don’t be too realistic with metaphors, because then you’re going overboard.”
That seems exactly wrong, because (a) metaphors aren’t universally good, as you point out, and (b) realistic metaphors do a better job tapping into the user’s existing schema.
While any reference I make to “hedging” is half-sarcastic, I honestly wish there had been some in this lecture. It’s not enough to simply rely on metaphors to tap into existing knowledge, you need to know what’s at work cognitively. I will jump out of my seat and applaud when I see that in a lecture.
Stumbled across your blog looking for discussions on radio button metaphors…
I think you’re right, metaphors are overrated but that doesn’t stop designers from searching for the perfect one for their app.
But there’s another factor that often takes over, that is when the metaphor becomes the prevailing thought - example - ‘desktop’ to me is no longer the thing my monitor sits on, it’s that MS Windows thing with 200 icons littered all over it. So ‘desktop’ now has an imposed familiarity that in reality is divorced from the metaphor that inspired it in the first place.
I would argue that the desktop metaphor (and others - ‘drag n drop’, ‘plug n play’) took off due to brute force marketing rather than because it was ‘just right’.
One cutesy little one that I wish had taken off was from the 70’s (I’m an old bloke) at a conference where we were shown some new Apple stuff. A drag n drop hand would close into a fist when the button was pressed to drag a Thing. I thought that was pretty cool.
At the same demo, a file sitting on the finder (now, there’s a crazy metaphor) was represented as a cube. You would click in proximity to an edge or a corner and the cube would rotate to display various properties. After a while you came to know which way to turn it to see what you wanted.
One more thing about ‘brute force’. In Australia we don’t have ’shopping carts’. We have shopping trolleys. Well, we have shopping carts NOW - all over the internet. And I recently heard someone in a supermarket refer to shopping ‘cart’. Aaaahhh, the tides turn!
Radio Buttons… I do a fair bit of training and I always make a point of explaining that old cars used to have radios where a button popped out when you pressed another one. (Mind you, I have owned many such cars, they aren’t that old!)
My audiences almost universally relate to it, but I would certainly say that this relating definitely comes from mild interest, NOT from knowing how an old radio worked in the first place.