ABSTRACT
My comment has to do with what I think slightly distinguishes the approaches followed by Roger Shepard and by Steve Kosslyn. While both are strongly arguing for and I think convincingly demonstrating the existence of a visual or analogical mode of reasoning, they emphasize different aspects of it. And maybe the differences of emphasis show up in the different kinds of experiments they favor. Steve I think is mostly interested in the process
· of imagery-the relation between storage of images in long term memory and retrieval of particular images in short term memory. Roger Shepard seems to be after more abstract principles-principles of motions and transformations that preserve the rigid three-dimensional structure of objects: as if he was looking for some abstract characterization of the "competence" underlying the visual or analogical capacity. For him, it seems that this competence has to do with the class of possible abstract motions or transformations. And I was interested to hear Roger Shepard's reference to the "missing (or empty) elements" of generative linguistics. He compared those empty elements to the mental paths created by the mind in perception.
