Prelude
When I sent the draft of this essay to the editor, I think it would be fair to say he was unhappy with it. We exchanged views, and this exchange was very enlightening, and much in the spirit in which I undertook to write for the journal. I do not wish to publish the exchange as such. Instead, I write this prelude in order to clarify the areas of contention, most of which came from differences in background and in use of vocabulary. In doing this, I hope also to clarify certain points what may, otherwise, be leading to confusion and misunderstanding. I am also aware that I exceed the length that I had considered I should write to, and to which I hope to return in my next piece.
My aim in writing these essays is to consider cybernetics as a discipline in its own right. I stated this at the outset. I am concerned that cyberneticians have not cared properly for cybernetics, have been too willing to present it in terms of other disciplines and to give away its insights, rather than nurturing the subject. I may be hopelessly romantic in this view, but I hold it nevertheless. Cybernetics deserves to be considered in its own terms, and the logic and coherence that is within it deserves to be developed in those terms. One way it has got lost is in the almost pathetic need to show its relevance and utility (science dominated by utilitarianism) and that it is just like everything else: thus, it does indeed loose its distinction! (I would hold the same about that other main area I am concerned with, architecture, in which case practitioners seems to have so little regard for its own value that they have always to describe it in terms of theories borrowed from other fields, without realising that architecture is the theory in its own right).
In wishing to consider cybernetics from its own point of view, I want to consider the subject from within. (Later in this piece I mention stability from within and without: I shall return to this notion in a later essay). I am concerned with cybernetics. Just that. Self referential cybernetics, autonomous: the cybernetics of cybernetics. In asserting this I assert that cybernetics has value (our value) as well as values. In insisting cybernetics has value, I insist cybernetics can contain (our) experience. What I am interested to show is that cybernetics can contain as much as, if not more than, other subjects. Therefore, I am not interested in priority but in all embracingness. It could be argued that many, if not most, of the major developments in theory are not made from priority, but, rather, of all-embracingness, and, in the end, that all-embracingness gives extension into prediction. Thus Occam’s Razor is satisfied.
For this reason, I am not concerned to show the uniqueness of any particular discovery, but to show that cybernetics can contain much. I am not concerned with semiotics but with cybernetics. As someone who simply does not see symbols and for whom symbolism, having no meaning, seems totally arbitrary, I find I am not interested in what I know of semiotics, any more than the psychology of CJ Jung. It simply is not about a world I inhabit. Although this may be a tremendous failing on my part (as is my dislike of Mozart), but there is nothing I can do about it. It is just not how I experience. In contrast, I am interested in what happens when we push understanding of mechanism, for instance, to the limits where it breaks down: what are we left with, and what does this tell us of that particular mechanism and also of mechanism in general? I like to strip down to an abstracted minimum rather than deal with a world made in explicit interpretation and meaning: for me these are personal and implicit, and ineffable.
Thus, I am interested to consider circularity as a more general case than linearity. In my construction, the linear is the circular cut, the trace left on the sand by the wheel. In particular, I am interested in linear causality as seen in this light: where circular causality is seen, as the founders of cybernetics told us, as the general case, with linear causality as the special case. Where causality is circular, we loose some of the most powerful and, alternatively, least attractive aspects of causality: we need make no assumptions about blame, cause, truth and the agency of cause.
I like to find ways of constructing what I see as being the territory of cybernetics. But others see as other subjects such as linguistics, semiotics and method within, and even making up, the cybernetic corpus. In what I do, I am not interested in precedent, but in coherence and the value of the explicative powers of cybernetics. I show how other fields behave and belong within the cybernetic world, not how cybernetics fits into other fields.
What I write below concerns the way Gordon Pask handled the problem of communicating when coding is unreliable (as I believe, in principle, it always is). To some readers, the word code has a symbolical richness-coding: a rich means of interaction and interpretation. This is not so for me, in the manner I talk of Pask’s work below, or in which I argued against coding in communication. In this context, a code is seen as mathematical (not semiotic, symbolic or psychological). Coding is seen as what spies do, what Alan Turing showed us how to handle and crack/decipher.
(Doing so in action rather than theory, he built the world’s first electronic computers (1943), the Colossus Machines or Bombes of Bletchley Park).
These codings are codings of literal connection and transformation in which one sequence may be represented exactly by another, which may in principle be de-coded provided the encryption key is to hand. They are, in effect, tautological - as Wittgenstein would lead us to believe.
Nor, in writing this essay, am I trying to write a learned paper. I wish to introduce certain concepts and to consider them so that even those who are not familiar with them may gain something from reading. I do not mean to be either learned or specialist. Thus, I am averse to giving either references or precedents.
Pask was concerned with circularity in communication. Conversation was what he took to be the form of circular communication. It is this that I present below, as a gem, as cybernetic and as cybernetics. What I am not talking about is language or linguistics, a semiotic system or study, or chronology. I would rather not make reference to the work of anyone (including myself), but some reference is made in order to aid those who feel the inadequacies of a presentation such as this. What I wish to achieve is to show a cybernetic approach, to show what cybernetics has to say, that it has something to offer, and that what it has to offer comes out of it being Cybernetics. Cybernetics has its own framework, its own value and values, which I hope to re-inforce.
Essay
I promised that I would discuss conversation, but the death of Gordon Pask, who did so much to develop and clarify our understanding of the power and value of conversation, intervened. And so I wrote about him and his late colleague, Robin McKinnon-Wood. In this article, I return as promised to conversation, especially the conversation as Pask meant it. As may be remembered, this promise was given as a response to my criticism of communication by coding.
Much of what is presented here will be found in greater detail in my recent publication "Communication without Coding: Cybernetics, Meaning and Language" (Modern Language Notes, vol. III no. 3, April 1996). The reader to whom this presentation seems superficial or trivial is referred to that essay.
Let us start with conversation as we might use the term in an ordinary, everyday - even conversational - manner.
A conversation is frequently about something, and normally involves at least two participants. For the purposes of exploration, we will assume these as conditions. There is also a (representational) medium of conversation - often called "language". However, the particularities of the medium are not important for this account. That a conversation begins and ends became very important (as a limitation) to Pask, but we will not be concerned with that, either. (For anyone wishing to look into Pask’s attitude to conversation, his last paper, "Heinz von Foerster’s Self Organisation, the Progenitor for Conversation and Interaction Theories" gives a moving account and testament. It was published in the issue of Systems Research (vol. 13 no. 3, 1996) dedicated to a festschrift for Heinz von Foerster). The bones of what happens can be presented thus:
One participant (A), having something (some idea/concept/observation) to communicate, presents it to the other participant (B) through the chosen representational medium. (In our universe, the chosen representational medium is frequently spoken language. But I wish to emphasise that I am not discussing language. If anything I am discussing conditions under which language might come to be). The other (participant B) makes whatever sense he will of this, creating his own idea/concept/observation (of the originator’s idea/concept/observation) responding with another presentation made to be sent back to the originator (A) of the communication, i.e. the participant who had the original something to communicate. It is very important that the presentation made back to the first participant by the second is not a repetitive "the same" as that made by the first. To simply repeat words, for instance, is not enough, because this does not present an alternative, merely a mirroring, and it therefore gives no possibility of error and correction. Apart from anything else, the inclusion of error and correction means that there need be no a priori: if these are built in, a structure ca be grown as it becomes necessary or desirable.
The first participant (A) makes his own sense of what is presented back to him, and, if he finds that it is near enough to what he originally wished to say, he can believe that he has communicated the something he wished, in the first instance, to communicate to the other participant (B). If not, he can try again with a modification or alternative to his original presentation - or even repeat the original. The tactics he uses to manage the "error" are up to him, decided by a combination of the context, his experience and so on.
(It should be apparent why the second participant’s (B’s) mirroring of the exact form of the presentation of the first participant (A) is not adequate: it leaves no possibility for the first participant to make an error comparison, and does not indicate that the second participant is making his own idea/concept/observation (of the originator’s idea, concept, observation).)
(As to the question of how the something the first participant wishes to communicate and his presentation of it are related: that is another matter. But I like and use de Saussure’s arbitrary co-incidence reported in the "Course on General Linguistics" (McGraw-Hill, New York, 1966), the slicing of time constructing two streams into synchronicity. However, the way this happens is not essential to discussions of (Paskian) conversation).
The mechanism of conversation thus described is clearly essentially cybernetic. It accepts error, which it regulates; and it invokes feedback, being a thoroughly circular process. It is a clear example of circular causality. It may be taken to be stable (within) while (probably) appearing to wander dangerously and aimlessly when considered from the outside, veering towards extinction! I hope to deal with the concept of stability, especially internal and external stability, in a future essay.
But conversation also has other, very special qualities.
Paramount amongst these, the conversation says nothing about meaning. I did not use the word meaning once in the description I gave of the common-place conversation. It does not assume that meaning is communicated, or that the act of communication (between the participants) has, necessarily, anything whatsoever to do with meaning.
In a conversation we make our own meanings. Meaning is not part of the public domain. We no longer have to discuss meaning, for it is ineffable. Thus, there is no "correctness" (although there may well be meanings that we make which, when acted upon, lead to unsuccessful or destructive behaviours: making our own meanings says nothing of the appropriateness of such meanings). What we are given by conversation is a release from meaning-as-external-phenomenon-to-be-studied, and its replacement by the conditions under which it can be created: a mechanism for exchange, and the insistence that goes with this - we are responsible for the meanings we make for ourselves (and, for that matter, our other actions), But, in that we do come to appear to gain what may be thought of as "parallel" personal meanings, we come to agreements (including the agreement t disagree): the process of conversation is a process of the negotiation of agreements between ourselves. We agree it is "as if" (see R. Glanville "as if" (Radical Objectivism), in Trappl, R (ed) "Cybernetics and Systems ‘94", World Scientific, Singapore) we have certain understandings in common. This is what conversation offers us.
A secondary point, but one of great importance (permitting the use of conversation as a generator of novelty) is that the errors inherent in this form of communication and the fact that it is truly interactive, resulting from the two participant (with their own distinct and unique constructions of the world), makes the conversation a natural for the creation and development of novelty. Pask even argued, originally as early as 1969 ("The Architectural Relevance of Cybernetics", Architectural Design, September 1969), that design is a form of conversation (through time) with oneself through paper and pencil. This explains the power of the doodle on the nominal back-of-the-envelope as a device for producing and developing ideas. This is one of the special gifts that cybernetics offers (as I argued in an earlier issue) : it accepts error and does not try to get rid of it. Instead, it manages it.
Pask came to consider the conversation through his interest in learning, especially in the creation of what we now call "learning environments", and the notion of "knowledge" necessary to support this. Pask’s underlying concern, if one is to invent one, is the wish that learners should be allowed openly to learn within their own frameworks and learning styles: they have different experiences, come to a lesson (potential learning occasion) knowing different things, have different intentions in what they wish to learn (and why they wish to learn it) and different tactics and styles through which they learn, and, in the end, need to be able to do their learning themselves (creating their own meanings: no teacher can learn for the student, a remarkable fact often overlooked in "education" especially at the end of this millennium). To do this, the student needs to be able to investigate what is to be learnt in a manner that suits him (i.e., conversationally); and the teacher needs to be able to test that the student has learnt in some way that allows the student to present his own understandings (this is, in a conversation). In fact, it was the need to sensitively test what had been learnt that, I believe, brought conversation to the forefront in Pask’s mind and caused him to devote his intellectual powers together with the rigours of cybernetic analysis to an understanding and refinement of (ordinary) conversation that became his "Conversation Theory" (Hutchinson, London, 1975) - around which this essay is written. Not that he ever claimed the insight as exclusively his own. He acknowledged both his colleagues and students at System Research and Brunel University and the Psychiatrist RD Laing whose Interpersonal Measure is one origin of Pask’s Conversations (see, especially, Laing’s "Knots", Penguin, London, 1970) as contributing directly, as well as a tradition sweeping back to a Socratic dialectic.
Nor does this understanding, based on negotiation and the construction of personal meanings, mitigate against the development of normal language. Language, in this view, is seen as an interpersonal social construct much as Luhmann, for stance - and working in a different culture - would have it, derived from negotiation but frozen, to which one subscribes. As I recently wrote in Modern Language Notes "In making the negotiatory pragmatic shortcut that is language, we agree not to disagree [with meanings becoming frozen in words, in a social space"].
In the first instance, it makes no assumptions about a priori meanings. It assumes only that meanings can be made, and that they are personal and therefore private and, in the strictest sense, non-communicable. Words do not have meanings, and what we communicate, whatever that is, is not meaning. We build our own meanings in the framework of the conversation (and in other frameworks). The making of sounds and gestures, for instance, is just that. I may make a sound or gesture to communicate an idea, but it is up to you what you do with the sound or gesture, how you choose to understand it, the meaning you make, whether you ever hear it and call it a sound. Nevertheless, using the form of the conversation, it is possible for us to develop meanings that, at least in so far as we can judge from behaviours, are more-or-less equivalent.
As I have argued in the Prelude to this essay, this is a simpler assumption than the assumption made in a view in which coding is primary because there are fewer preconditions and preconditional structures. It does not need a code to be made, or even the notions that code exists and of coding, or meanings to exist. The question of how whatever is in the code is understood by each of us, different as we are (and it is time we examined difference rather than dismissing it in favour of enforced similarity), does not arise. Conversation requires less assumptions. We don’t even have to ask how coding comes about, how it is understood to be a code. Of course, we may make and use (the preferred verb is usually discover) codes. We do not need to rely on them. But we make them, and we negotiate that they are codes and how they will be used as codes. This way, we do not even have to carry out the sort of metaphorical lobotomy that the military, for instance, carry out on new recruits, to ensure they obey commands blindly and without the ambiguity of interpretation. Occam, in the expression of his Razor, is on the side of conversation, not coding.
You don’t have to ask how meaning, as well as the participants, got into the conversation: it didn’t. You don’t have to ask how something is recognised as a code, how we know it’s a code in the first instance - it isn’t - or how it can become decoded - it doesn’t. It seems to me that the only alternative to this understanding is to return to naive and discredited realism. We might wish to return to the pre Vienna Circle notion (for instance) the Nature is waiting for us to discover her Laws, a notion still clung to by a number of scientists, especially, for instance, neo-Darwinians. The code. And the Meaning of Life in the deciphering of the Code. Perhaps some linguists and semiologists also hold such notions, in their attempts to define what is in meaning and language and other sign systems. I cannot see how anyone can hold such views nowadays. This is my blind spot. Their holding such views may well be theirs! For they are, of course, confusing the "as if" with the as.
Second order cybernetics teaches us several things. One of them is that the observer is in the system. That the observer matters. That the observer observes, and that what he observes - his observations - are his observations: they depend on him and they are his. Because he is himself and no one else, they are necessarily distinct and different, and, when they are "communicated", what is communicated is not them but the opportunity to create, for another, his version of what we may later come to share as "them" (as he understands them).
(This is not to say that we cannot so construct our knowledge that we encourage regularities to appear, both repeatable and constant; or that an agreed language may not seem to carry meanings. The difference is between what we agree to "standardise", to treat as if it were shared and constant, to create procedures for, such that they will generate "knowledge", and where we understand this to have come from. That is, the limits that we push our understandings to and through).
By involving the observer as conversationalist; by not relying on prior, agreed meaning (how ever could we have agreed it?); by insisting that meaning is personal and is not communicated; by keeping the form of communication separate from the content, conversation provides a much more basic, much more general, much less demanding, much more cybernetic understanding of how communication can happen in the world of today. And it positively supports the creation of novelty and the semblance of life. The assumption of conversation allows that we can communicate when we see differently (with all that that entails and portends) and what its is that we communicate: what its status is. It is better than notions based in coding for Occam’s reasons. But there are difficulties. I shall return to these, all things being equal, in the next issue.
Return to the content of this issue
Return to the Cybernetics and Human Knowing Homepage