There are many counterexamples in the book… the one I liked is the remote control that has a loop antenna on it. The user, a computer science prof, points the loop antenna to his stereo, but still the stereo didn’t seem to receive the signals from the remote very well. Later on, he found out that the “loop antenna” was merely a hook for hanging it on the wall, and he had actually been pointing transmitter of the remote at himself!
It’s interesting in a few ways: First of all, weren’t there any writings on the remote that if you point it to yourself you’ll be reading upside down? Secondly, 20 years ago people might assume that an antenna could be on a remote, but nowadays I would think most users would have seen enough remotes to not assume to see an antenna.
So definitely there’s a cultural aspect to usability design: 20 years ago a drop-down list might not be intuitive, but most computer users nowadays would have seen it enough, and these standard GUI widgets has become part of the “culture” for computer users.
