A reflexive dispositional analysis of mechanistic perception

Minds and Machines 16 (4):479-493 (2006)
  Copy   BIBTEX

Abstract

  The field of machine perception is based on standard informational and computational approaches to perception. But naturalistic informational theories are widely regarded as being inadequate, while purely syntactic computational approaches give no account of perceptual content. Thus there is a significant need for a novel, purely naturalistic perceptual theory not based on informational or computational concepts, which could provide a new paradigm for mechanistic perception. Now specifically evolutionary naturalistic approaches to perception have been—perhaps surprisingly—almost completely neglected for this purpose. Arguably perceptual mechanisms enhance evolutionary fitness by facilitating sensorily mediated causal interactions between an organism Z and items X in its environment. A ‘reflexive’ theory of perception of this kind is outlined, according to which an organism Z perceives an item X just in case X causes a sensory organ zi of Z to cause Z to acquire a disposition toward the very same item X that caused the perception. The rest of the paper shows how an intuitively plausible account of mechanistic perception can be developed and defended in terms of the reflexive theory. Also, a compatibilist option is provided for those who wish to preserve a distinct informational concept of perception

Author's Profile

John Dilworth
Western Michigan University

Analytics

Added to PP
2009-01-28

Downloads
710 (#18,853)

6 months
85 (#44,068)

Historical graph of downloads since first upload
This graph includes both downloads from PhilArchive and clicks on external links on PhilPapers.
How can I increase my downloads?