Neural Codes and Fields at the Microscopic, Mesoscopic, Macroscopic and Symbolic Levels

Abstract

This paper makes two self-confessedly ambitious proposals. One is a theory of mind and world with an inventory of possible relations between the two of such generality that sensorimotor behaviour, potentially conscious cognition, and quantum mechanics fall out s special cases.  The second is that the variety of neural codes is as multifarious as that of the domains in which mind functions; alternatively put, each cognitive "context" can be viewed as a field. Where  cognitive "context" is lacking - a la quantum mechanics - the result is the quantum field theory of researchers like Schwinger.

This paper makes  also makes the radical claim that dynamical systems theory provides solutions to problems plaguing neuroscience, rather than simply attractive models. It starts with the microscopic level, that of single neurons. A biologically realistic neuron model as a harmonic oscillator is shown to allow neurons do pseudo-Fourier transforms. While it is already known that spike timing becomes naturally causal in this model, we have also implemented a C++ simulation showing that it can operate on a raw power spectrum, and learning can be formulated as adjustment of delays. In short, the neural code at the microscopic level is, as Karl Pribram thought, the Fourier transform.

The mesoscopic and macroscopic (EEG) levels, which are at times connected   in Freeman's writing, cater for the missing piece of "intentionality" ie how mind "intends - points to - things in the world. It is argued that nonequilibrium thermodynamics provides a good model here. The vocabulary of dynamical systems, starting as we already have with the periodic attractor of the harmonic oscillator qua pendulum, is proposed as a first approximation for what we need to do at the mesoscopic level.

That will finally bring us to the symbolic level, at which we experience, talk to each other, and do math. It is argued that formalisms that cater for co-ordinate free flows are more appropriate here than any others. Clearly, tensor calculus and lie groups will prove useful. We also consider physicists who have eschewed cognitive neuroscience as a failure and, with some brilliance, argue that physics ideas like pilot waves will prove crucial. While this will be the most speculative part of the paper, this area is developing rapidly and quietly like all successful revolutions.