Qualia Could Arise from Information Processing in Local Cortical Networks

Roger Orpwood

doi:10.3389/fpsyg.2013.00121

Abstract

Re-entrant feedback, either within sensory cortex or arising from prefrontal areas, has been strongly linked to the emergence of consciousness, both in theoretical and experimental work. This idea, together with evidence for local micro-consciousness, suggests the generation of qualia could in some way result from local network activity under re-entrant activation. This paper explores the possibility by examining the processing of information by local cortical networks. It highlights the difference between the information structure (how the information is physically embodied), and the information message (what the information is about). It focuses on the network’s ability to recognize information structures amongst its inputs under conditions of extensive local feedback, and to then assign information messages to those structures. It is shown that if the re-entrant feedback enables the network to achieve an attractor state, then the message assigned in any given pass of information through the network is a representation of the message assigned in the previous pass-through of information. Based on this ability the paper argues that as information is repeatedly cycled through the network, the information message that is assigned evolves from a recognition of what the input structure is, to what it is like, to how it appears, to how it seems. It could enable individual networks to be the site of qualia generation. The paper goes on to show networks in cortical layers 2/3 and 5a have the connectivity required for the behavior proposed, and reviews some evidence for a link between such local cortical cyclic activity and conscious percepts. It concludes with some predictions based on the theory discussed.

Highlights

An important development in work exploring mechanisms underlying consciousness has been a move from the important search for Neural Correlates of Consciousness (Rees et al, 2002; Koch, 2004) to a deeper look at mechanisms involved, a search for Explanatory Correlates of Consciousness (Seth, 2009), the neural processes that can account for fundamental properties of conscious experience
This paper explores the behavior of local cortical networks under the activation of top-down feedback from an information processing perspective, to see if there is anything in that behavior that could lead to the emergence of qualia
CORTICAL NETWORK MICROANATOMY How well does the theory outlined above map onto the known microanatomy and dynamics of local networks in the cerebral cortex? The theory revolves around the establishment of repeated cycles of activity through cortical networks following local feedback, and for this activity to be under top-down re-entrant facilitation

Summary

INTRODUCTION

An important development in work exploring mechanisms underlying consciousness has been a move from the important search for Neural Correlates of Consciousness (Rees et al, 2002; Koch, 2004) to a deeper look at mechanisms involved, a search for Explanatory Correlates of Consciousness (Seth, 2009), the neural processes that can account for fundamental properties of conscious experience. Taken together the ideas of re-entrant feedback and micro-consciousness seem to be arguing that re-entrant feedback activation is necessary for conscious percepts, but that the site of the neural activity from which they emerge could be quite a local one (Figure 1) This line of argument seems to be moving toward a conclusion about a set of neural activities that lead to the generation of consciousness. This paper explores the behavior of local cortical networks under the activation of top-down feedback from an information processing perspective, to see if there is anything in that behavior that could lead to the emergence of qualia. This paper takes an information-based perspective to explore the pattern recognition capabilities of networks of pyramidal cells It analyses the behavior of networks encouraged by top-down feedback to engage in extensive local feedback to themselves and to establish attractor states, and to examine how they process the fedback representations they generate as outputs. It develops further previous studies on self-reflective neural circuits (Orpwood, 2007, 2010)

INFORMATION PROCESSING IN LOCAL NETWORKS

EVIDENCE FOR THE BEHAVIOR DISCUSSED

CONCLUSION