> More than understand an inscrutable icon. Besides, anyone can look up Fuel in a dictionary. How do you look up an icon?
This makes sense. But I think that's why the unicode consortium adds common gestures/symbols/icons so actively. The icon-recognizing problem you address here is not impossible to fix with some app combining scanner and translator.
I appreciated your comments throughout this thread.
Symbol language works better in the longer term. When there's consensus about the icon functionality, they work. Look at traffic signs in Europe. They barely contain text. In the USA, they contain text all the time. In a large country where the main language is English this makes sense. In a continent consisting of many languages (and tourism), its important to have consensus on symbols instead.
What does suck is converting to symbols from text whilst you're used to text. That's the legacy of backwards compatibility.
You are oversimplifying. There is a valuable trade off between text on one end and abstract symbols on the other. Icons are in the middle.
On the text end of the spectrum you have high clarity, but it takes longer time to take in and find the one you are looking for. These are good if you rarely interact with the them and it's in situations that aren't urgent and you can't or don't have to train for.
Abstract symbols on the other hand are fastest to identify but require you to be trained. This makes sense for power users out urgent situations that are worth preparing for.
Icons are the middle ground. You can identify them faster than a walk of text and you might be able to guess pretty well what they mean on first encounter.
Using icons on a car this makes sense. You might have to do a little bit of trial and error the first time around, but will be much faster on future usage. This is a worthwhile trade off for a car, since you want the driver to quickly find the right control and then look at the road again. It works especially well given that the vast majority of icons are common between different cars.
I think it's even a better trade off for traffic signs, since drivers get trained anyway. All but new drivers will quickly start relying on the shape of the text on the sign, rather than read all the text on every speed limit sign. If you do need to read it, the amount of text is not practical.
> On the text end of the spectrum you have high clarity, but it takes longer time to take in and find the one you are looking for.
I just don't believe it takes a longer time to find the word "Defrost".
> These are good if you rarely interact with the them and it's in situations that aren't urgent and you can't or don't have to train for.
Take a look at aircraft instrument panels. They are very much designed for clarity and to be usable in emergencies, usually the hard way. They use words. Altimeter, Airspeed, Vertical Speed, etc.
> you might be able to guess pretty well what they mean on first encounter.
Not me. 3/4 of icons in new cars I don't know what they mean. Most of the ones on my iphone as well. Thankfully, sometimes they use words like Save, Edit, Library, etc., and I no trouble with them. The ones on the app screen thankfully have a word under them. After all, look at that weird stick icon, which has "App Store" written under it. Or the Photos icon. Apple has even given up on the trash can, reverting to the word "Delete".
> but will be much faster on future usage
Pilots and airplane cockpit designers evidently disagree.
> drivers
Quite a lot of concepts have no picture that makes any inherent sense for it. Like "60 miles per hour". Or "stop". or "yield". You can certainly get used to a red octagon meaning Stop, but for someone who is not trained on it, the red octagon might as well mean "pizza next exit".
I lived in Europe around 1970, when they were transitioning from the old signage with words to the new ones. They wanted to harmonize the signs. The trouble is, if they used words, it was a matter of national prestige which language was used. Using icons was a solution to the political issue, not a readability one.
The American stop sign was settled on because nobody could agree which European stop sign would be selected, so the compromise was none of them.
To sell this all to the public was the notion that icons are better, but it was really political.
> Take a look at aircraft instrument panels. They are very much designed for clarity and to be usable in emergencies, usually the hard way. They use words. Altimeter, Airspeed, Vertical Speed, etc.
People are trained for those instruments. The written names are not that useful, because you won't fly something before understanding most of the cockpit. The instruments often use shortened or genetic name which is you need to learn beforehand. Example:
What can you learn from reading "FD", or "INS", or "DH", or "M/CG"? Or from tens of completely unlabelled indicators. You either need to already know it, or learn from the manual.
Calling it an "altimeter" reinforces what it is, even "FD" does, an icon does not. Icons are not better, or instrument makers would switch to them and the FAA would mandate them.
Believe it or not, but pilots can get frazzled under stress and make mistakes. Having an actual label that says "Altimeter" is better than the pilot having to remember that squiggle means "Altimeter". Just call it "Altimeter" in the first place. One less level of indirection.
For a related UI case, pilots would learn the different sounds of all the warning horns in the cockpit. But pilots under stress would get confused about which sound meant what. The solution? Replace the sound with a voice saying what's wrong, like "pull up pull up". Works much better.
This makes sense. But I think that's why the unicode consortium adds common gestures/symbols/icons so actively. The icon-recognizing problem you address here is not impossible to fix with some app combining scanner and translator.