ABSTRACT
In the object recognition community, recent years have seen the growth of a movement based on the idea that visual object recognition is mediated by the activation of template-like views. This chapter reviews the evidence and motivation for the view-based account of object representation, describes in detail the nature of this account, and discusses its logical and empirical limitations. I argue that view-based theory is fundamentally and irreparably flawed as an account of human shape perception and object recognition. I then present the most plausible and common objections to my arguments and respond to each. Finally, I relate view-based theories of object recognition to other models based on formally similar types of knowledge representation and conclude with a discussion of why such accounts must necessarily fail as an account of human perception and cognition.
