Brain as n-dimensional nonlinear state machine

Wolf Singer, Director at the Max Planck Institute for Brain Research
in Frankfurt, presents his view of how the brain works, from the cellular level to the level of consciousness.

The money quote:

To our intuition, it seems foreign that the neuronal correlate of that which we perceive as a solid, tangible object should be a highly abstract, spatially and temporally structured excitation pattern – and that, in this way, not only three-dimensional things, but also smells, feelings or intended actions are represented. And every such representation corresponds to one specific state out of a nearly infinite number of possible states. Or to put it differently, the cerebral cortex system continuously moves from one point to the next in an inconceivably high-dimension space. The trajectory – that is, the trail of this movement – depends on the entirety of all internal and external factors that impact the system.

He then explores the implications of this theory, touching on time, evolutionary adaptation, and social structures. The sense I get from this is that the brain at a given moment may be represented by a sufficiently large-dimensioned matrix of neural excitations, but that it may not be moved forward or backwards using that model due to the effect of the entirety of those internal and external factors.

The research done by Foster and Wilson at MIT seems to bolster this view. The rats they studied ran through their neural space-representation in reverse order when pausing in a maze run. This local reversing of patterns occurred when the rats were going the same route and when attempting new routes. Sounds like your brain plays Simon. This seems to implicate that the brain has hooks into specific patterns that repeat on call. Not the same as back-propagation algorithms.

Somebody pointed Singer link out, but I’ve lost track of whom. If you want kudos, feel free to leave a comment.

February 28th, 2006 | Alife | No comments