‘Everybody knows’ that we read by recognising the shapes of words, right? It’s common knowledge in the design and typography worlds, I thought.
Turns out, the designers and typographers either misinterpreted early psychology work in the field, or the psychology work was plain flawed. This excellent overview article explains all, and I think it has me convinced that I, for one, have been parroting rubbish for years.
A few comments: first, I wish I was more surprised that some psychologists had adopted the ‘Not A therefore B’ fallacy early on – there’s some genuinely bad science on show in there, happily exposed by later work but more specifically by clearer thinking. I’m reminded of a truly dreadful paper I once saw presented at the British Association, which purported to study comprehension rates when listening to human speakers. The study concluded that comprehension did not fall off with speaker distance, which would have been surprising and interesting. However, a minute’s back-of-the-envelope examination of the size of their video screen, the likely line resolution of their dodgy camera, the resolving power of the human eye, and the distance range of their study… The casual assumption that the greater the distance, the less information the subjects had to work with was not, in that study, the case. There was, in essence, no difference between their tests.
Secondly, I’ve said before that I admire Microsoft’s efforts in typography and font design, and this is no exception. Given this work, I find it mystifying that Word continues to set type so badly, but I guess that’s backward-compatability for you. At least we have Verdana and Georgia, two of the very best screen fonts available.
As for ClearType, Microsoft’s type anti-aliasing system that uses the arrangement of red, green and blue dots on LCD screens to increase the effective horizontal resolution: it’s not something I’ve had the opportunity to use in practice. Anyone care to comment? I have used Apple’s similar rendering options (System Preferences → Appearance → Font smoothing style, for all you Mac OS X heads), but find the colour fringing distracting. Thus, I turn it off (‘Best for CRT’).
2 thoughts on “Things we thought we knew but didn’t”
Cleartype is amazing. It makes screen text almost as easy to read as laser printouts. Turning it on and off, even between regular anti-aliasing and Cleartype, is night and day for many fonts. (CRT at 1600×1200) My only complaint is that not all fonts work well with it, although I’ve only had trouble with a few custom fonts for specific software packages.
If I recall the flurry of patent talk when it came out it is based on very old Apple technology.
Hey now, watch what you say about cogntive psychologists! 🙂
There is some bad science, but that’s generally true about any discipline (and you’re also talking about psychology back before it really was even acknowledged as a discipline, let alone a science).
Most cognitive psychologists today believe in multiple mechanisms for many cognitive processes. The trick is that we try to isolate individual processes to get a better idea of how each works apart before trying to understand how they work together. Unfortunately, I think this often gets interpreted by the news and other people outside the field as being a single mechanism argument.
I suppose there are some extremists out there that do propose single mechanism theories for complex processes, but I like to believe that they are either just being inflammatory or just trying to get published (or both).