Responses to the first two editions of Theories of Visual Perception were generally very positive. Students in five countries told the author that the book helped them to grasp the essentials of some of the various theories of visual perception. Many lecturers have adopted the book. With two exceptions, the book received favourable reviews. A number of young researchers in artificial intelligence have reported that Theories of Visual Perception was their entry point into vision. All this has been very gratifying.
Any feelings of complacency were quickly checked by some general criticisms from the author's own students at Exeter University. In the main, these focused on two chapters: that on psychophysics, and that on Brunswik's theory. To argue that the essential nature of a sensory threshold has been a subject of theoretical debate for 100 years, and that a major branch of visual research had arisen because of this debate, did not convince these young readers. After much thought, the original chapter on psychophysics has been dropped from this volume.
Brunswik's probabilistic functionalism struck many students as a real oddity. They understood the chapter on his theory, but found his own writings difficult in the extreme. The chapter has been retained in this second edition for two reasons: first, because of the present author's hunch that Brunswik may have been right in his intuitions concerning the basic nature of perception; second, because of some recent publications concerning perception and the statistical nature of real world events. This new work has been described in the present chapter on Brunswik's contribution.
Since the second edition of this book appeared, there has been a torrent of recently published research findings concerning visual perception. At this point it must be stressed that Theories of Visual Perception is not a textbook on visual perception per se. Rather, the book is an attempt to describe how a number of general theories of visual perception developed: their backgrounds, their underlying assumptions, their strengths and weaknesses, and their current status. For those wishing to read fuller accounts of experimental findings in vision research, there exist a number of remarkably good textbooks, one example being Bruce, Green and Georgeson (1996), which is the present author's favourite. Readers should also enter the Web