The Role of Conscious Experience in the Study of Perception and Cognition

Cognitive science, and particularly vision science, has had a deeply ambivalent relation with the phenomenon of conscious experience. On one hand, the way things appear or what they look like has always been an important, if not the primary, source of data, at least for vision science. When one thing looks bigger in one condition than in another or looks to be moving faster under one condition than another, or when colors appear different under one lighting condition than another, these are considered primary data to which theories of vision are expected to respond. On the other hand, the content of a person's experience has also proven to be one of the most misleading sources of evidence, because it is not neutral with respect to the theories that the subject holds, be they scientific or folk theories. Moreover, which explanations appear most natural is highly sensitive to the way we describe our experiences, and conversely, the way we describe our experiences (even to ourselves) depends to a large extent on what tacit theories we hold.

The way we describe our perceptual experiences often caries with it the implication that the content of the experience itself explains observed phenomena—that the occurrence of experience X causes experience Y which then explains some ensuing behavior. There are more or less benign versions of this sort of what might be called intentional causation (also sometimes referred to as psychological determinism;Hochberg 1968). An important and essentially irreproachable version of this thesis is the appeal to the tight coupling that holds between how a part of a scene is experienced and how other parts of the scene tend to be experienced (see Epstein 1982;Rock 1997). For example, if you see the edge of a Necker cube marked x (in figure 4.1) as being the top front edge of the figure, then you are likely to see the face of which it is a part as the top face (as in the second panel); but if, instead, you see the edge marked y as being the top front edge, then you will see the face that it bounds as the top edge of the figure and the appearance of all the other edges will change so the interpretation of the figure remains coherent. When the percept changes (as it does in ambiguous figures such as in the first panel) the couplings force the interpretation of related parts to change accordingly. This fact has been the basis for a successful technique in computer vision called constraint propagation (see the

Figure 4.1

The way we see edges is intimately connected with the way we see the faces those edges bound. These appearances (or ''ways of seeing'') form a tightly coupled system. If our percept of one part changes, the appearance of other parts change systematically in order to maintain the coherence of the whole. Panels 2 and 3 show the two versions of the ambiguous figure in the first panel. Notice how the interpretation of an edge is connected to the interpretation of the faces it bounds and that both interpretations change together when the ambiguous percept in the first panel ''flips.''

Figure 4.1

The way we see edges is intimately connected with the way we see the faces those edges bound. These appearances (or ''ways of seeing'') form a tightly coupled system. If our percept of one part changes, the appearance of other parts change systematically in order to maintain the coherence of the whole. Panels 2 and 3 show the two versions of the ambiguous figure in the first panel. Notice how the interpretation of an edge is connected to the interpretation of the faces it bounds and that both interpretations change together when the ambiguous percept in the first panel ''flips.''

work on the ''blocks world'' that culminated in the successful system devised by David Waltz, described in, among other places, Pylyshyn 2003, chap. 3;Waltz 1975). (It is also the basis of an approach to models of reasoning using constraint satisfaction;Tsang 1993.)

When we speak of labels on the representation of a scene, or indeed when we speak of what a pattern is seen as, we are speaking of the contents of a perceptual representation. In most cases such contents are assumed to be conscious;hence we are speaking of the contents of a perceptual experience. But what exactly is the content of a conscious perceptual experience? The content of a belief is relatively clear, because beliefs are individuated in part by their contents—that is, we identify a belief by what it is about, or we may treat beliefs that are about the same thing as the same belief (I am ignoring for now the fact that beliefs can differ in ways other than in their content, e.g., their form). But what about the content of a conscious experience? The situation here is not at all straightforward.

Was this article helpful?

0 0

Post a comment