Neurophysiology and Digital Computing

The nervous system is the most obvious and complex form of communication ‘‘in the animal,’’ although the endocrine system also operates in a broadcast or ‘‘to whom it may concern’’ fashion and other forms are playing a part.

McCul­loch looked to neurophysiology for answers to the question he posed, and 5 years before the appearance of the book by Wiener, he and Walter Pitts published a mathematical proof of the computational capabilities of networks of model neurons that have some correspondence to the real kind. Much speculation was aroused by the apparent correspon­dence between the all-or-nothing response of neurons and the use of binary arithmetic and two-valued logic in digital computers.

In this and other ways, interest in cybernetics arose from advances in electronics, especially those stimulated by World War II. Young scientists who had worked on military projects wanted to turn their skills to something to help humanity, such as biological research. Ways were found of making micro electrodes that allowed recording from single neurons and stimulating them, and it looked as though the nervous system could be analyzed like an electronic device.

A great deal has been learned about the nervous system using microelectrodes, not least by a group around Warren McCulloch in the Massachusetts Institute of Technology in the 1950s and 1960s. Further insight came from theoretical treatments, including the idealization of neural nets as cellular automata by John von Neumann and the mathe­matical treatment of morphogenesis (the development of pattern or structure in living systems) pioneered by Alan Turing.

Nevertheless, much of the working of the central nervous system remains mysterious, and in recent decades, the main focus of cybernetics has shifted to a higher-level view of thought processes and to examination of inter personal communication and organizations.

A recurring theme is that of self-organization, with several conferences in the 1950s and 1960s nominally devoted to discussion of self-organizing systems. The aim was to study learning in something like a neural net, associated with the spontaneous emergence of structure or at least a major structural change.

Von Foerster treated the topic in terms of thermodynamics and pointed out that spontaneous emergence in an isolated system would be contrary to the second law. More recently, self-organization has been discussed rather differently with emphasis on the emergence of structure as such, independent of learning or goal-seeking.

Ilya Prigogine and his followers have resolved the contradiction between the apparent implications of the second law of thermodynamics, that entropy or disorder can increase only, and the observed increase of order in biological evolution.

The contradiction is resolved by obser­ving that the second law applies to systems close to equili­brium, and living systems only exist far from equilibrium and can be termed ‘‘dissipative structures’’ as they receive energy and pass it to their environments. It has been shown that spontaneous emergence of structure is natural in such conditions. Implications for psychology and social systems, as well as cosmology, are drawn.