<p> The psychologist J. J. Gibson, and later the enactivists, espoused a view of perception emphasized active sensory exploration, and the biological functions perception serves. They tended to neglect the internal complexity of perceptual systems. Neuroscientists and computer vision researchers, on the other hand, focus on the complex structure and inner workings of perceptual systems, to the neglect of biological and behavioral context. Here I will formulate a version of ecological realism which reconciles and critiques these seemingly disparate approaches. </p><p> I argue that the objects of perception are relational invariant structures preserved within the changing flux of perceptual input. The function of perception is to enable appropriate behavior with respect to affordances, which are objective three-way relations between worldly features, animal abilities, and animal needs. The invariant relationships perceived tend to be those which signify affordance relationships for the species and individual in question. </p><p> The perception-action cycle is but one example of the circular dynamics of perceptual systems. The neural portions of such systems are also in a state of constant feed-forward and feedback dynamical interaction with one another. These dynamics confer an active autonomy on perceptual systems as manifested by phenomena like dreams, hallucinations, and perceptual illusions. Metaphorically, such systems may function to constantly formulate and test hypotheses about affordances based on perceptual evidence and prior categorical experience. Hierarchical predictive models of perception, in which perceptual systems consist of a hierarchy of Bayesian statistical predictors, represent a possible means by which this metaphor might be crafted into a testable scientific hypothesis. Perception, even if it involves actively autonomous perceptual systems coping with ambiguous input, is epistemically reliable most of the time, because it is constantly tested by action. Perceptual states are true or valid if they bear an appropriate relationship to objective affordance relationships, and false or invalid if they do not. These views require a reformulation of the venerable distinction between `direct' and `indirect' perception. Perception is ontologically direct in the sense of dealing in objective relationships in the world, but justificationally indirect in the sense of requiring an argument that perceptual beliefs are generally epistemically reliable.</p>
| Identifer | oai:union.ndltd.org:PROQUEST/oai:pqdtoai.proquest.com:1561403 |
| Date | 17 September 2014 |
| Creators | Patton, Paul |
| Publisher | Indiana University |
| Source Sets | ProQuest.com |
| Language | English |
| Detected Language | English |
| Type | thesis |
Page generated in 0.0151 seconds