Words, Images, and Concepts

and perspective-independent. So the first challenge for the imagistic model is to account for capacities like allocentric spatial navigation, which seems to draw on a different set of representations than those used in perception. This is not a matter of simply adding dimensions to the same perceptual space, but of mapping locations in a different kind of space. Putting the point a little differently, allocentric navigation may involve spatial images, but not perceptual images. This actually points to a slightly wider moral, which is that focusing on a division of all mental representations into either perceptual or conceptual is probably too simple. There is a wide range of nonconceptual representations that may be used by specialized nonperceptual subsystems such as the dead reckoning device.