The author of this book has obviously spend a great deal of time in academia -- much to his detriment. The book reads like a college-level psychology textbook, complete with references to the myriad of other people upon whom his work is drawn.
For example, the author is not content to just define a term and move on. Instead, he reviews the entire history of the term, what other people have thought about the term, and then summarizes all of the thoughts. Geez, there are even examples at the end of each chapter.
If my review is hard to read, I blame it on the fact that I have been reading this book for the past hour.
But, if you can get past the overly-academic writing style, you will find a good deal of useful information. The concept of object-oriented user interfaces is often misinterpreted and/or mis-implemented. There is a detailed history of the object-oriented GUI, and good discussions on the human factors that lead to good GUI designs.
One note is that the book was published in 1995, so brace yourself for lots of examples from the leading GUI of that time: Windows 3.1! Windows NT is only mentioned as 'Cairo'. But as is true of all design models, the age of the book really has no relevance.