Flashes Upon an Inward Eye: A Wittgensteinian Study of Mental Imagery
Dissertation, The Johns Hopkins University (
1988)
Copy
BIBTEX
Abstract
A renewed interest in mental imagery has recently emerged in the field of psychology. Many current theories still presuppose a flawed model--one in which images are inner objects, constructed from stored parts. My goal is to demonstrate the incoherence of such apparently intuitive views of imagery, and to establish a Wittgensteinian theory. ;After indicating the scope and limits of the notion of mental imagery, I consider Wittgenstein's private language argument. Since mental images are easily taken for private objects, it is important to prove that images, like sensations, cannot be identified independently of public practice. This seems to raise a problem about accommodating the inner phenomenal qualities of images, which I address. ;I then attack the empiricist claim that seeing an image and seeing something in the external world involve the same processes. I argue that this supposition is fundamentally incoherent. ;Following my discussion of the conceptual problems in comparing vision and imagery, I focus on the kinship between seeing-as and imagery. From this discussion Wittgenstein's notion of an "internal relation" and its role in imagery emerges as crucial. I interpret internal relations as holding between the thought that constitutes an image and the object of the image. Then, employing this notion, I explore the interconnections among the will, intentionality, and imagery. In doing so, I derive a reading of Wittgenstein's central question for mental images, "What makes my image of him into an image of him," and claim that the answer involves the notion of intentionality, and not inner copies of outer objects. My internal relations account, I conclude, provides the most coherent theory of mental images. ;In the final chapter I turn away from explication of Wittgenstein to indicate how my Wittgensteinian study of the concept of mental imagery raises difficult questions for theoretical work on images in cognitive science . I show that Pylyshyn's program is more consistent with my approach than Kosslyn's