Amodal Ecological Perception: Constructing an Invariant World from Ephemeral Sensations

The ecological approach to perception considers how perceptual systems are shaped by and rely on contingent features of the environment in which the organism evolved.  

These considerations dramatically transform the epistemological questions about perception familiar from the traditional, passive computational paradigm.  In psychology, the ecological approach is most closely associated with the work of J. J. Gibson.  Many of Gibson’s insights have been rediscovered in recent philosophy of perception, but often without much connection to the rich psychophysical tradition that stems from his work.  Furthermore, neither literature makes much appeal to the closely related study of perception in evolutionary biology.  The primary aim of this project is to unite these three complementary strands of research on perception from psychology, philosophy, and biology.

A test case for the value of an interdisciplinary ecological approach to perception will be its ability to explain the role of perception in grounding our knowledge of the world.  How is it that transient, subjective sensations in several sensory modalities come to underwrite stable representations of an objective world?  These representations are either multimodal, integrating features from different sensory modalities, or amodal, abstracting away from the particularities of sensations from any single sensory organ.  Somehow in the process of combining information from different sensory modalities into a unified representation of the world, we are able to transcend the contingent features of our sensory apparatus and latch on to robust external regularities in the environment.  

Gibson characterized perception in terms of the detection of invariant patterns in the flow of information through our sense organs.  It is these invariant patterns that we use as the basis for amodal representations, and it is the origin of these invariants in external regularities that ensures the veridicality of these representations.  Most work on perceptual invariants so far has focused on the examples of color, pitch, and space.  The process of cross-modal interactions, especially between vision and hearing, has also been extensively studied.  This project will build on this tradition, moving beyond it to develop conceptual tools for understanding the transmission of invariants across modalities such as olfaction and touch, and their integration with other sensory information into richer a- and multimodal experiences.


Project Activity: Conference 'Ecological Perception: Amodal and Multimodal Trends' (May 2015)

Project team: 

Project Leader: Dr. Alistair Isaac

Funded by EIDYN/PPLS