Mark Boulton

Users don’t read. I’m not just talk­ing about web users either. Ever spent more than 1 second read­ing a road sign? Ever spent more than 1 second read­ing a dir­ec­tion sign in a pub­lic build­ing? Ever spent more than 1 second try­ing to use a web­sites nav­ig­a­tion? That’s my point. 

Design­ers for the web need to look more at sys­tems design, semi­ot­ics and way­find­ing for cues for their interfaces. 

Take icon­o­graphy for example. Icon­o­graphy, espe­cially in com­put­ing, has arrisin with the advent of more com­plex GUI’s, BUT it has risen primar­ily because of a series of com­mon tasks which need to be illus­trated in some ‘real world’ way. 

This Image shows a num­ber of icons dis­played which show a num­ber of com­mon tasks. As you can see the design of these icons vary, but only subtly. There are some in each set which ‘feel’ right how­ever, these are the suc­cess­ful icons which tap into the uncon­sious cues asso­ci­ated with semi­ot­ics. I ques­tion icon design and it’s valid­ity within design. My exper­i­ence of ‘icon’ design (and i’m not talk­ing brand­ing or logo­types here, just icons) is thay are a) Are not thought about in enough detail b) They are almost always dec­or­a­tional, there­fore their func­tion is often sec­ond­ary to how they look. c) Most icons are so badly designed they need words with them in order to decifer their mean­ing. Not good.

Here’s a good essay on icon­o­graphy and semi­ot­ics. Have a read, it makes a lot of sense. 

Talk­ing of sys­tem icon design, this is a great resource for com­par­ing oper­at­ing sys­tems and their iconography. 

Prob­ably more on this later, when i’ve thought a bit more about it.