CYBERNETICS & HUMAN KNOWING

Subject-headings:
Mechanism ; Self-reference ; Cybernetics ; Machine ; Observer

A few weeks ago I lent my copy of Steve Joshua Heim’s invaluable book "The Cybernetics Group" (Cambridge, MIT Press, 1991) to an old, and by now dear friend. After reading it, he said "Thank God for second order cybernetics". This response is in stark contrast to his response of years ago. When I had just been examined for my PhD by Heinz von Foerster and Gordon Pask (on a thesis that was, as might be inferred from the examiners, very definitely second order cybernetics), he accused me, and all other cyberneticians, of being fascists who wanted to reduce the world to mechanism and destroy both quality and the human. I am glad to say he later changed his mind.The reason I expose my friend thus is that, in his assessment, there is something of the difficulty encountered by many in trying to understand cybernetics: and some of this difficulty is, I believe, related to how we see the role of the animal and the machine in Wiener’s descriptive phrase.

I have always understood the outlook of cybernetics to be essentially humanist. That is, I have always seen it as supporting notions such as quality, and the value of that which is outside conventional scientific description. That is one of the reasons I have been interested in cybernetics: it enables me to handle areas that were previously unhandleable, without a danger of severely distorting them. But I have, equally, always seen it as bringing mechanism into play in a profoundly powerful way. My introduction to cybernetics (in a consultation with Gordon Pask) showed me the incredible power of thinking in a cybernetic manner, of treating the world as extreme mechanism. My friend’s responses remind me of the possible paradox or contradiction in this, in the relationship between cybernetics and the duality animal/machine.

Perhaps a word of further explanation is needed here. For Wiener et al, the notion of animal and machine must initially have been one in which the animal is treated as mechanism. It is, for instance, made up of parts, and these parts may be treated as pulleys and pumps, etc. I think they proposed that the whole animal was greater that the sum of its parts. Nevertheless, the parts were enumerable and mechanical. This is, of course, the Cartesian metaphor and formed the context in which they were working.

Yet, at the same time, I believe they understood that this was not really and adequate description. Living at a time when the product of this type of thinking had taken horrendous form in atomic weaponry, they were also afraid. Heims talks of the beginnings of McCarthyism, of the social panic and insecurity of the time and the awful, fearful (and consequently rather un human) behaviour that often resulted. That which is outside the realm of rational thought took on a pressing presence in their minds (hence Wiener’s second book. ‘The Human Use of Human Beings’), and they were sharply aware of the value of their humanity. They lived in the paradox between a means of investigation that tended to deny, for instance, the reality of fear while, all the time, suffering it themselves.

And they were aware of the strangeness of the albeit necessary circular causality they were advocating which brings into play self-reference, that great bugbear excluded by classical logic, the supposed driving engine of and giver of value to the Cartesian way of understanding they were heirs to. For selfreference is specifically excluded from this view of classical logic.

I would suggest that it is useful, when thinking of the animal, to remember that there are aspects of the animal, at least as far as the human is concerned, that are an important part of our experience, which cannot and should not be excluded, and which Wiener et al also must have had in the back of their minds when they introduced the duality animal/machine. Perhaps it would have been better if he had used the word human instead of animal. Anyhow, it is in this sense that I talk of cybernetics as being essentially humanist. It is able to accommodate (some of) those areas of experience that are excluded by the mechanistic view of Descartes.

Nevertheless, reading Heims’ book and thinking back to Wiener’s "Cybernetics" and some of the cyclostyled and undated papers I have of McCulloch, I am surprised at the exclusiveness with which the early pioneers believed in the overriding power of mechanism. In the case of McCulloch, if Heims is correct, the belief was overwhelming: he was interested only in mechanism, and believed that proper science dealt with mechanism and that mechanism could explain everything that it should properly explain (to reformulate Turing’s machine). Put in terms of Wiener’s determining statement, cybernetics is concerned with control and communication in the animal and the machine, where the animal is treated as (if) a machine, that is, as mechanism. Heims, of course, sets this in the appropriate historical and social context.

Yet, what I see as the power of the cybernetic method of investigation, of the particular way that cybernetics allowed and encouraged exploration of the world we find our selves in (or, as we would now say, we make) is that it transcends mechanism. This is where my friend’s old criticism was ill-founded: cybernetics applied mechanism not only to the world it found/created, but also to that application and to that world. Its particular quality is that it pushes to the limit and then beyond. The cybernetician, in my image of him/her, asks not only the current, but also the next question.

Thus, from a very early stage, and certainly by the time of the Macy Conferences that Heims’ book documents, such a basic concept of control and communication as feedback (not, of course, new in itself) had been both understood and assimilated, and its consequences for our interpretation of control were also well understood. Causality no longer existed uniquely in the linear, conventional sense: for, in a system exhibiting feedback, cause must be seen as circular. No longer does the chicken "cause" the egg (or vice versa: an ambiguity that only points to the difficulty of the traditional position re causation), but they even "cause" the other through continuous, circular causation. Thus, not only were concepts - based in mechanism - developed and made usable, but they were then pushed to the limit, frequently in a sort of self-referential boot-strapping where the concepts of this new subject (cybernetics) were applied to these cybernetic concepts themselves, in order to clarify them.

This is how I came to view cybernetics when I first met Pask, in 1967. I saw this tremendous arsenal of concepts, but also, and much more impressively, I saw the power of the way the concepts were used to clarify not only what they were applied to, but also themselves. I saw the "machine" in cybernetics not only throwing light on human activities, but also being used to push concepts to the limit: not necessarily to confirm these limits, but to see beyond the limitations of scope and view that these very concepts gave us.

Thus, by going to the limit and beyond, we see not only what the concepts are, where they end (as well as how they help us, in use): we also see the areas that are outside their scope, the areas that remain unassailed (and, probably, unassailable) by the mechanical metaphor. For this reason, almost any investigative work in cybernetics is, of necessity, implicitly involved with the animal - or, at least, the non-mechanical - no matter how much it concentrates on the machine.

A look at the work of that time does, I think, bear this out. For me, the thought that epitomises this period is Ross Ashby’s. A look at his papers will show how often they are concerned with limits and that which is beyond the scope of analytic scientific thinking (so, for instance, he often wrote about Bremmermann’s limit to the computable). His "An introduction to Cybernetics" (London, Chapman and Hall, 1956) remains, for me, the classic text of cybernetics: powerful in mechanism, yet freely enjoying the limits, especially of the Black Box .

It could be only a matter of time before such an approach began to look at how the subject of cybernetics itself was made: the agency of cybernetics. That is to say, for a subject concerned with mechanism, and the limits of mechanism, it cannot be too long before the mechanism by which the subject is made comes under investigation: especially if the general approach of the subject is to go to these limits and then beyond. This time occurred in cybernetics between, as I like to bound it, 1968 and 1975, by the end of which period cybernetics had come to terms with its own making and transferred its foundation from linear-with-applied-circularity to circular-with-selected-linearity.

Included in mechanism, it was found, was the mechanism of observation that allowed the mechanism of the "system" (or whatever you like to call it) to be determined, and the mechanism that made such describing possible: as well as the mechanisms that permitted there to be observation and, indeed, an observer to make the observation. The observer (traditionally excluded in accounts based in mechanism) became an integral part of the system under consideration and observation became an inclusive and included process. Questions of what is inside and outside were no longer clearly answerable: the system no longer had an independent existence, which lead to a revision of the status of what was knowable.

To achieve this transfer, simple though it now appears, required thinking of the highest calibre, and equal bravery (or innocence) to cope with the very considerable personal risks being taken by those who did it: for they had to rigorously dissemble one mindset, a mindset with an enormous amount of cultural weight behind it, which had dominated their lives in science, and then dare hang in there without any support for an eternal instant while the new order began to be formed and to take shape. And, as this happened, to apply once again the same rigour but without the same supports and the old preconceptions and habits.

The metaphor changed: from being the animal treated as a machine, it became the machine treated as an animal. (Perhaps it is worth remembering that the source of the word animal lies in the ideas of spirit and breath: the anima. In the case at least of Pask, the world was literally populated by animae.) That is to say, there was a shift from a preoccupation with external control and causality, to an appreciation and study of autonomy, and the internal control that we come to believe is involved in that. The prefix "self" became more and more important, and with it all those old questions (such as are raised by Goedel’s Theorem) were approached with/in new ways and, in some cases, became dissolved as problems. This is the land of second-order cybernetics, the joint share-holder in the concerns of this journal.

Effectively, there was a change in focus and interest from causality to autonomy; from chain to circle; from action and reaction to interaction; from mechanical to animal; from exclusion (observing from without) to inclusion (observing from within); from realist to constructivist; and from coded messages to conversation. And then the earlier metaphor became re-incorporated, so both metaphors were held simultaneously, co-existing, each throwing its own light, which is where we are now.

Given that this is the case, that cybernetics is the field which explores not only its own application, but its limits, its basis and its operations (its self) using these same concepts of cybernetics, why did the pioneers not jump to this position/conclusion, immediately (or at least rapidly), instead of seeming to take position as extreme mechanists?

I suggest there are several reasons. One is the people they were and the time they lived in, when mechanist thinking was producing the goods without us seeing and understanding the damage that was simultaneously being created through side effects: what British Prime Minister Harold Wilson, optimistically, later referred to as the "White Heat of Technology".

The second, of course, is that it requires a special sort of thinking to move a field, or at least to move how we appreciate it, from the first to the second order. Maybe not all of those who formed cybernetics could think like this. Maybe they were tired by the initial act of creation. Maybe some tried and may, even, have had some partial success. But not enough! (It seems that Gregory Bateson, in particular, and to some extent Margaret Mead and her entourage, understood the need for this shift.)

But the third is, I believe, the most telling. When I asked Gordon Pask (who, after all, knew all the Macy Conference crew) what Wiener et al would have thought about second-order cybernetics, especially since it seemed to contradict, undermine or dispose of much of what they had achieved, he told me they would have been delighted and approved. When I asked him how this could be, he said that Wiener, in particular, realised there was another step to take, but did not know how to do so. He was waiting for others to pick up the baton and run with it, to complete the forming of the subject he had begun. If you look at his second (and, perhaps, better) cybernetics book, "The Human Use of Human Beings" (1954) you can see this attitude seeping through.

It seems that it is hard for us to let go of our old views. Pioneers and revolutionaries in many fields can often only point the way. They indicate, they strain in the direction they are pointing, but in the end they are too tied to the place that generated the need for the pioneering changes to be able to move themselves. After they have pointed the way, others must make the running.

And it is amazing how widespread this phenomenon is. Recently, I dropped in on a research seminar at one of the smaller Universities in Melbourne, Australia, to find the early cybernetic texts being referenced along with French deconstruction and post-modernism in a presentation concerning the nature of art, the artist, and especially the art work as we leave the second millennium. What should they be? The presenter was straining to transcend the limits of both the theoretical work he was using to shed understanding and his "old" concept of art, the artist and the art work, without quite knowing how: he was stuck in first order cybernetics, when second order cybernetics would have helped him!

The way in which the phrase "in the animal and the machine" is important in Wiener’s determination of cybernetics may only now be becoming apparent. The inclusion of this pair, without an ordering or hierarchy of control, without the metaphor being expressed one way or the other (eg, as the animal dominating the machine and/or vv), permitted the development of cybernetics so that it could start with a primacy of the mechanical metaphor, but, by extending that into the absurd and then transcending this absurdity, allow a science and rigour to be developed that gives access to areas previously beyond any but the most extremely metaphysically metaphorical of investigations, while also humanising how we understand and do science. Who could be satisfied that, before the concept of autopoiesis, we examined life by killing it; that before conversation we insisted on universal coding; that before Objects we had no "structures" to allow the attachment of unique views; and that before distinctions we had no logics that permitted genesis and supported the act?

References

Heims, S.J. (1991): The Cybernetics Group, The MIT Press, Cambridge, Massachusetts.

Wiener, N. (1948): Cybernetics, control and communication in the animal and the machine, Hermann, Parts

Notes

1. There is a wonderful collection (in 5 volumes) of McCulloch’s papers that Heinz von Foerster assembled and edited, published by InterSystems, Seaside California, but sadly no longer available. The determined researcher might be able to access this collection.

2. And I saw the astonishing power of Gordon thinking with them!

3. I am aware of those who argue that all is metaphor, etc. This is not a position I wish to take issue with. Nor do I want to debate its applicability. I hope I will be allowed to use this expression here without being required to argue a position relative to various interpretations in which metaphor is given particular significance.

4. Remember: humanist and humane!