Statistical Mechanics of Nervous Nets

Abstract

There are many deep and difficult problems to be solved before any adequate understanding is achieved, of the workings of the central nervous system. “Macro”-neurophysiology and neuroanatomy provide an image of the central nervous system as a system of ordered nets arranged in a multileveled hierarchy, with an immense number of circuits within and between nets. Feedback inhibition undoubtedly plays a very important role in maintaining the activity of such nets, which is controlled by internal and external signals, especially from nets in the lower levels of the hierarchy (the reticular systems), and by signals from the receptors. “Micro”-neurophysiology and anatomy fill in the details necessarily obscured by macroscopic analysis. The analysis of intercellular interactions and of the responses of cells to specific forms, has established that there is a specific organization of nets into functional columns maintained by nearest neighbor interactions, by spatial summation and by inhibition. There are topological maps between many nets and from receptive fields. There are differences between the responses of nets in anaesthetized and unanaesthetized animals In unanaesthetized animals, cells may exhibit a maintained or “steady-state” reponse in addition to the transient response normally found in the cells of anaesthetized animals. How do these tonic and phasic responses influence each other? What is the role of the functional columns, and of the multileveled hierarchies? These are key questions that must be answered before any real understanding of neural coding and information processing can be expected.

This research has been sponsored in part by the Physics Branch, Office of Naval Research, Washington, D.C., under Contract No. F 61052 67C 0061.