(I chose to write this review only after reading both Emotional Design
and The Design of Everyday Things
. The wait was worthwhile.)The Design of Everyday Things
focuses on the usability of things used daily. First introduced in the late 1980s, under the title "The Psychology of Everyday Things" (POET), this book describes usability as combination of four main factors, speed of use, ease of use, background knowledge to use, and errors of use. This book marks a shift in the beliefs of computer scientists about design, from engineering task (focusing mostly on performance and manufacturing ease) to aesthetic (focusing mostly on appearance and marketability). For Norman
, the design, here understood as "the successive application of constraints until a unique product is left" (words attributed by Norman
to Richard Pew, leading cognitive scientist and one of the firm believers in human performance modeling), has to balance usability with aesthetics. Norman also argues, sometimes off-topic and perhaps not always logically, that professional designers should be used to design most of the things we use everyday, instead of the industry practice of using the time of regular employees such as engineers, programmers, computer scientists, and even managers.
The content is broad, yet reasonably well-structured and easy to grasp (use). There's too much to really comment, and perhaps spoilers are a bad idea; this is a book about design, and ideas about how to change everyday life for the better abound (some have already been implemented, such as the electronic display on each key of a computer keyboard, allowing for the visualization of the key mapping). Three of my favorites are the argument for digital instead of analog watch displays, the ideas for designing computer Dungeons and Dragons (role-playing adventure) games, and the keen observation of how the writing tool's speed influences the style and quality of the writing.
On the positive side, there are many principles or notions that can be readily applied in everyday design, such as deliberately making things difficult to increase secrecy/privacy, focusing on the trade-off number of controls/complexity of use, designing for multiple audiences, the perils of standardization, etc. Two rules of good design [for usability], as proposed by Norman, remain the same today: good conceptual model and visibility of system status. There's also an excellent yet short list of suggested reading (a bit dated by now).
On the negative side, Norman argues for a number of things that require much more detail to be believable, among them, the rather naive treatment of (human) failures and how to reduce them through better design. I did not like much the parts that relate to cognitive sciences; the theories of the 1980s have been replaced (extended, sometimes overturned) by the theories of the 2000s
Overall, a wonderful read and a great introduction on usability as a major focus of design. Rec: For everyone who designs for real users.