The early days

Citation

Andrew, A.M. (2006), "The early days", Kybernetes, Vol. 35 No. 1/2. https://doi.org/10.1108/k.2006.06735aaf.001

The early days

The early days

It was not like that

I have to disagree with the assertion by Peter Krieg (Kybernetes vol. 34, no. 3/4, pp. 551-557) that cybernetics, in its early days, was strongly influenced by alleged military origins. On the contrary, its flourishing after the end of the war was an escape from the frustration of war-related activity to something that could be seen as contributing to human welfare.

Certainly, in Britain and USA there were many young people familiar with electronics, as a result of work on radar and other wartime developments. Some of them turned their attention to neuroscience, where the pulse techniques that had been pioneered in radar proved ideal for design of stimulators and recording equipment. Ways of making microelectrodes were devised, and it became possible to record from single neurons and also to stimulate selectively. The way seemed clear to analyse the working of the nervous system as one would an electronic device.

Of course, the new availability of electronic digital computers was a major influence, but the computer use mainly discussed by cybernetics pioneers in USA was the work in the Aberdeen Proving Grounds on gun control, rather than the Manhattan Project. This work was of course war-related, as was the famous study by Norbert Wiener of prediction for anti-aircraft gun control, but, unlike the atom bomb, effects of these would be targeted solely on combatants.

It can now be seen that many attempts to draw parallels between brains and computers were glib and misleading. The model neurons of Pitts and McCulloch were advanced by them as only extremely crude representations of real neurons, but this warning was not taken to heart as widely as it should have been. The magnitude of the problem of understanding the working of the brain was undoubtedly underestimated by many people, including Herbert Simon as shown by the excerpts quoted by Krieg.

The motivation, though, was entirely defensible and laudable. There are many distressing diseases of the nervous system and it was reasonable to suppose that the more that was known the better would be the chances of treatment. Also, there was response to what has been described as the greatest challenge faced by science, namely that of unravelling the working of the brain.

Artificial intelligence arose as an offshoot of cybernetics, and some of its results are of general applicability, so may be embodied in military as well as other systems. However, the assertion that the early work was supported by the military requires clarification. It is true that the group around Warren McCulloch, as well as Heinz von Foerster's BCL, received funding from the Office of Naval Research and the corresponding office of the US Air Force. Curiously, however, this was provided without any requirement that the research should have military applications. One of the reasons given by Heinz von Foerster for closing down the BCL was that the service agencies were starting to alter their policy in this respect.

It is true that the difficulty of the task of “understanding man's understanding” was grossly underestimated, and quite basic aspects of brain activity are still little understood. In certain areas, though, a great deal of progress has in fact been made, especially in examining visual and other sensory pathways. The findings tend to be presented under the heading of neuroscience rather than cybernetics. There are other ways in which cybernetics has left its mark but not its name, as for instance in the ready reference to feedback in numerous contexts, and awareness of the part played by information transfer, as in references to “messenger RNA”.

Just how the working of the brain might be understood is by no means clear, and here second-order cybernetics must enter. As has been said (by, I think, Warren McCulloch) a correct theory of the working of the brain would need to have been written by itself.

There is much more to cybernetics than brain theory, and under its heading principles inherent in biological control have been applied successfully in, for example, sociology and management. A few of the applications may have been of a military nature, but on the whole the topic is one of peace and goodwill.

Alex M. Andrew