I don't normally write book reviews, but I thought this one merited special attention. I've just finished reading Mark Changizi's remarkable book, The Vision Revolution. As someone with a stake in the vision sciences, I'm inherently interested in these sorts of topics. Most books written about visual perception are dry compendia of what we already know. Changizi's book answers questions that I had never thought to ask. But should have...
The book adopts an evolutionary approach to explain why, for example, few languages have color terms specific to skin color, why forward facing eyes are useful (and it's not the reason you think), why all writing systems tend to have object-like properties, and why our visual system is always guessing about the immediate future. Changizi examines how anomalies of the visual system (e.g., why two of our three cone types respond to wavelengths that are so close to each other when spacing the cone sensitivities equally would better represent color space) follow neatly from his theoretical claims about the purpose and evolution of human vision. My favorite chapter examined the evolutionary advantage of forward facing eyes, particularly for larger animals in forests.
This book represents the best of theoretical vision science, taking strong positions and following their implications to their natural conclusions. In reading this book, I was reminded of the sort of vision science that James Gibson did, not in substance, but in the style of thought. Gibson challenged some of the fundamental assumptions about how we see, reconceptualizing vision in ecological terms. Although Changizi doesn't attempt to overhaul the entire enterprise in the way that Gibson did, he instead asks and answers big questions that, for some reason, are typically neglected in discussions of vision. These are the sorts of question that should be central to the work of vision researchers, but most empirical vision research is too deep in the trenches to see the big picture in the way that Changizi does. They are the sorts of issues that should be addressed in undergraduate perception courses, but almost never are. The book represents a style of thinking that I wish were more common in cognitive psychology and perception today. Although I don't agree with all of Changizi's claims, I found his arguments and evidence compelling — reading his book forced me to revise some of my own misconceptions and preconceptions about the visual system.
For anyone who studies vision, this book is a must read -- it will change the way you think about the vision sciences. It is an accessible read for anyone interested in human cognition. Read it to see how a masterful theorist revisualizes one of the oldest subdisciplines of psychology.