One of Johnson’s most accessible pieces. Written for the business audience, it applies his theory of perception and “reflective participation” to the world of computing and office work.

In recent years, we have begun to hear more and more about the coming of the post-industrial age, but not so much about its actual complexion. Supposedly our information sciences will spur it into being and automation of some narrowly conceived kind might lie at its heart. In this article Avery Johnson argues for a wider vision of the possible uses of our existent technology, and describes not, as he says, a far-off future utopia but a world that is already at our fingertips, if we will simply accede to a new understanding.

Dr. Johnson is a Cyberneticist and Ecologist with the Ecology Tool & Toy Company.

A few of us believe that we have groped our way into a new territory which is going to require extensive exploration before even an adequate set of words is adapted to use in communicating about it; for in fact, that very communication lies at the heart of the matter.

In seminar teaching the instructor has the opportunity to confront and interact with the student directly and, short of dragging him off to the laboratory where he might be able to play with embodiments of the concepts themselves, the instructor can at least play-act and demonstrate the events for which he cannot find words. The optimum teaching tech­nique, as I like to describe it (but cannot live up to as often as I would like), is to arrange matters so that the student teaches me about his world, but does so in the contexts that I set for him, We can set up a dialogue where, moment by moment, our changes in point of view are exchanged immediately without recourse to question-and-answer and we become our own laboratory for the examination of this very process. I apologize for not being able tt> do the same thing for the reader, at least not this year. I cannot make the substantive material of my argu­ments change in response to your way of looking at it or reading it so that it would follow and lead your line of interest. If I could do so, my article would appear on one page, for that one page would, in the course of time, take on all the shapes and relationships necessary to convey its messages to you, much as I would act in person: talking, watching, listening, gesturing. But let’s proceed anyway in this older style of pre­sentation, and I will choose the explanations and examples which I hope can lead into sharing some delight of discovery in exploring the potential of the post-industrial age.

One point I hope to make clear as we go along is that I am not describing a far-off future utopia which must await astounding tech­nological breakthroughs. I am describing a world which is already at our fingertips. All we need to do is simply accede to a new understand­ing of the relationships that exist within it, and implement our presently available technology toward the active embodiment of those relationships.

I want to direct these arguments at managers of technology because I believe they hold the keys of change. The problems of communication faced by managers are virtually a microcosm of the ones our society faces everywhere, and the tools they have at hand with which they might explore alternative solutions are precisely those which may effectively be turned in radically new directions. As we proceed. I want to set up several contrasts between the kinds of solutions which the old industrial “think” found appropriate to its specific problems and those which our

new technologies have made possible. In doing so, I shall have something to say about questions of communication in general, in a way, what I want to talk about is wealth—real wealth—not money, but that which money ought to be able to buy. You are truly wealthy when the part of the world that matters to you is responsive to your wants: Where you do not have to adapt yourself to situations but can expect some agency, genie, or mechanism to act for you as congenial intermediary. The industrial age has provided us the tools but has not yet endowed them with the spark of life. That is our principal focus here.

Let’s try a cold plunge into a responsive, communicative world of the not-so-distant future: Imagine that tomorrow morning when you enter your office, something about the place looks very strange. There are no books or catalogues on your shelves, no stacks of correspondence in the baskets, in fact no paper in sight at all—except for a lonely memo pad which you will find has pages which vanish minutes after they have been used. Your filing cabinets have disappeared and there is no desk in sight; just a comfortable-looking chair with a side-table for telephones and a funny piece of furniture which you are told is a computer console that has been assigned to replace all that has been taken away.

Well, you may as well see what you can salvage out of this mess. Can’t do much worse than wreck that console-thing, and right now that doesn’t seem such a bad idea.

You sit down in the chair and instantly you notice something ani­mated about it, something responsive to your every movement. The thing comes alive and gives the impression that it is exploring you to adjust itself to your comfort. You get up. walk across the room, and come back to it again. This time it reacts differently and seems to “know” you better, anticipating your shifts of weight and muscle tensions. Strange, though, it isn’t trying to lull you to sleep the way loungers do. It’s as if it were being attentive to what you want it to be. Hmm. Not bad for a start.

You begin to reach for the phone and the console screen flashes a question: “May I call someone for you?” Well, why not? You reach for the keyboard and start typing YES but the screen has the next question there immediately: “Name? I have your personal telephone list on file.” You start typing the first name, and as you proceed to hit the keys of the teletype, each becomes easier to depress, and toward the end, the key­board is running itself. The screen flashes a message to the effect that Henry is in the building and will be notified of your call. You suddenly realize that you really don’t want to disturb him if he is not in his office, and you start an abrupt movement of annoyance toward the keyboard, but the darned thing beats you to it. “OK. Shall I leave a note on his screen?” A gentle raising of your hand is read as approval and you are flashed: “Good. Will do.”

Perhaps, when you and machine have become better acquainted, a gentle softening of the cushion you are sitting on will be a more positive indication to you that your wishes have been understood, making un­necessary any verbal acknowledgment on the screen. Think about how you communicate such things to someone very close to you: a nod. a smile, a squeeze of the hand? Why should such richness of exchange be denied you in other matters of equal subtlety and importance?

This article is aimed at the suggestion that there are ways in which our present technology of data processing and control can be applied to interfacing man with his environment so as to lead him into a range of creative activities never before possible.

The idea is not to automate to replace the man but rather to augment him as a man, to turn him on to time and space frames, and to an ele­gance of complexity which formerly did not exist for him because he had no way to participate in them. If you can provide him with tools of ele­gance and complexity that endow him with greater effectiveness of inter­action with whatever part of his world matters to him, you will have served him well as a first step. What happens then, however, is that his interests broaden and allow much more to matter. His “mind” is no longer something that resides within his skull, but reaches to the farthest limits of this effectiveness—and beyond. He becomes an “instant ecologist.”

Two years of working together in the Environmental Ecology Lab­oratory in Boston has convinced us that we formerly had no awareness whatever of our environment, and that the development of this aware­ness is crucial to the kinds of communication we want to improve upon. If you can instrument your environment so that it responds to you— poking back when you poke it—then very soon you begin to command an awareness of complexities of relationship you never dreamed were there. Your understanding is expressed in terms of you and your be­havior instead of something which you might read on a page.

What about pages and hard copy and printouts? Do all of our answers to questions have to be written out on paper? Let’s dig deeper. Let’s consider the way in which the environment of our industrial age has conditioned us to think about the roles of man and of machine.

Heretofore, factories have housed machines where people have been employed to play the roles of the parts of machines difficult to design into the hardware because something had to be self-organizing enough to perform the informal tasks and to monitor them for novelty and change (bad things). The alternative we are often threatened with as a tech­nological future is that of automated factories in which still fewer people are employed to push buttons and watch dials. This is just plain silly. It’s like saying that a painter must be highly trained so that hand and eye can later do their tasks by themselves, cranking out artifacts without further creative intervention on his part. The fact is to the contrary. What his training has provided him is the ability to control the materials of his craft so that he may better express his artistic intentions—new expres­sions, novel metaphors that show the observer some aspect of his world in a new context. The control that the master has is so complete that his use of his medium may be an informal process, experienced directly. He has learned to think in terms of ultimate results rather than in terms of the steps necessary to achieve them, 

Think for a moment about the uses to which you are putting your computer facilities right now. Let’s try a sample list:

Accounting - Computer-Aided Design

Payroll - Production Control

Inventory - Catalogue Editing

Shipping Orders - Modeling Simulations

Budgets -Linear Programming

Sales Records -Forecasting

Mailing List - 3-D Graphics

Now look again at the list and try to characterize the entire content as if it described the useful behavior of a domestic pet, a sophisticated pet robot that does your bidding. Don’t worry about how it goes about a task once that task has been specified. Think about what transaction must occur between you and it to specify your bidding.

I have listened to programmers and managers talk about their uses of computers, and the simile they often share is to compare the data processor with a room full of secretaries (of no special form but having an extraordinary talent for speed, accuracy, patience, and obstinate obedience) to whom one states the task required and who then perform it with the data and the rituals at hand. If you listen carefully you would imagine that the secretaries in that room have eyes and ears to see and hear with, and hands to work with, but they never seem to be endowed with mouths. They have no way to ask you questions in return, to point out the ambiguities in your requests. In short, you are not allowed to approach that room with an ill-formed notion of what you want. The group therein has no way to be conversational with you. no way to ex­plore your wantings nor for you to explore its information space as you might explore a summer garden, the handshake of a friend, or the playful whims of a child. In effect, you have access only to those services for which you already know a name and whose execution is well defined.

“Well, so what?” you may want to ask. Computers are tools like any other, and tools don’t argue with you. You know how to use them or you don’t.

I’l agree that’s all there is to it if you are content that the tools you use should not be expected to adapt themselves to the task at hand or to your particular style of approach to it. All the technology of the industrial age has been directed toward making machines that perform defined operations faster, with more repetitive precision, and indefatigably doing, so to speak, what factory labor has been doing all along. Some mechanic or programmer had to set up the machine to make things that would meet the requirements of “a market” (probably an averaged composite of people). But if the products don’t meet your needs or appetites closely enough, you simply have the freedom not to buy!

Now, that’s a “first-order” freedom, we must admit. But it doesn’t give you much freedom to express what you want. There’s damned little information transfer between you and the system serving you when you simply say “no thanks” and walk away. If you were dealing with a craftsman plying his custom trade by hand, he would be intent on pleas­ing you personally. The information exchange between the two of you would be high and informal and would gradually converge upon the rele­vant issues. Your request to him might be vague at first, but eventually would result in a very specific embodiment: the crafted artifact.

Is there any convincing reason why our highly developed information technologies are incapable of a similar dialogue with you, directed toward the same sort of ends’? I am not talking about the problem of computer use of vernacular language. That is not the issue. Craftsmen can be deaf, dumb, and blind and still provide you with what you want if you are both willing to share the same information space for a while.

That’s the problem: the sharing of information space.

What do we mean by that?

For the sake of argument, let’s look at the kinds of interchange that people have with people. While doing so, let’s be especially attentive in watching for those places that science might label “difficult to analyze” and wonder whether perhaps therein might lie the key to what we are seeking: an informal language of direct experience.

Two lovers have not seen each other for some time. What kind of messages flow between them as they approach from afar? Listen:

’’I’ll be there about 8 tonight.

“Good. I’ll meet you. Which airline?”

“TCI. I think it’s Flight 201, but it gels in about 7:30.”

“I’ll be there...can’t wait...” and so on.

These words might just as well have been exchanged by teletype— except for the loss thereby of the eagerness of voice, the nuances of tension or joy that come across. If that can be postponed until the meet­ing, the conversation is inherently amendable to teletype or computer coding, because the context of the words is so fixed by the situation that their meaning is unambiguous. We do not have to wonder which of the myriad meanings of “meet” is to be inferred nor whether the number 8 is precise enough.

At the airport the lovers see each other while still 50 yards apart. As they move closer they begin to read each other’s walk or run, the flash of the eye, the smile, the disregard for surroundings. But this is truly a “reading” as if the message were written on a page. Only mildly is a dia­logue in process. Neither yet participates in responding to the other’s changes while the interval between is great.

For the analysis of the exchange we are still within the capabilities of present science practice and might imagine that the readings are being made by a polygraph: a bouquet of physiological monitoring equipment recording heart rate, skin temperature, pupil dilation, respiration, blood pressure, and so forth. Interpretation of the records does not have to be made by someone who is himself involved in the encounter, not even at the instant of its happening. Analysis might be performed at high speed by a “pattern recognition” program in a computer, and a verbal, numer­ical, or graphical result could be presented which would succinctly de­scribe the emotional set of the two rapidly converging lovers. The context is so fixed and the method of analysis so ritualized that the information is distilled out of the scenario and is reduced to a simple set of names.

And now at last they rush into each other’s fond embrace, and at this point we leave them.

Why? Why now, just when we are becoming captivated by the scene?

Not because of discretion, but because the complexity of their com­munication is so great and so immediate that our language of word sym­bols on this page is quite inadequate to convey anything of it but a frag­ment of suggestion.

This is where science cops out and art takes over. The writer of intimate scenes must rely heavily upon the reader’s own world of experience or of fantasy to supply the meanings to the words.

The communication sciences have as yet reneged when it comes to the description or simulation of conditions where the flow of information is so multichanneled, complexly interconnected, and mutually intervenient that one can point Io nothing fixed in time or space except statistics. Technology is concerned primarily with “doings” and “havings,” the transitive changes in the world that can be represented on paper and can be repeated in fact. It has left to art the “seemings” or the expressions of intention where the relation of observer to observed is the message of interest and where his frame of reference, his context, must be known before we can make any statement about what his observation meant to him.

Let’s go back again to your real world and your computer whose uses we are trying to examine. Would you consider it an improvement or a loss if the machine available to you could not be programmed explicitly to do things for you “right the first time,” but had to be taught in some sense—starting with a low-resolution, imperfect style of execution but rapidly learning the procedures that are relevant to your operation? Would you consider it more useful to have an interface between you and the machine rather than a teletype, a graphical display, and a high-speed printer? Suppose the machine were additionally sensitive to your postures or the tempos of your expression in different moods, and were also par­ticipating in the control of your environment; light levels, air temperature and movement, acoustical privacy, or communications routing? Let’s imagine that the system is “good enough” at its roles that you will feel neither mothered nor smothered, but rather served in a manner maxi­mally relevant to your style of doing and being. First, what might be possible with such an attentive servant at your elbow, and. second, what mysterious process is involved in achieving it?

The answer to the first part of the question is enormous in scope.

The answer to the second part is simpler than you might expect. But this is where the new think is required for which our formal educations have not trained us.

Let’s go at this second part first, for then the world of possibilities should open tip almost by itself.

You see, we have been led to believe that knowledge is useless or perhaps even nonexistent unless it can be set down formally: named, pointed at, made graphical, or otherwise symbolized in some final form. Where relationships appeared hard to manage or a scale of measurement

difficult to find, statistics were invented to lit descriptions to the scientific mold. The social sciences in particular have made great use of statistical procedures for hypothesis testing because very little is ever repeatable in any more explicit way.

The most serious difficulties arise in the attempt to understand sys­tems in which there are continuous complex interactions and no visible beginning or end: no separably identifiable causes or effects. For my part, these complex, recursive systems are the only ones that interest me: people and their communications with each other and with their world, management systems, the process of education, ecology, evolution.

Whenever the system to be studied is one where output and input cannot be identified separately, where the opening of the loop puts a stop to the behavior that is of interest, that interest is usually deferred and those measurements and descriptions amenable to analysis are deemed sufficient. Even though the most sophisticated uses to date of computer capabilities for the examination of complex relationships provide a tran­sient, graphical output on paper for visual observation, they never give me the opportunity to plunge my hands into a writhing mass of variables to play with them and with their interactions. I’m serious about this! Visually presented readouts of processes running from Time A to Time B, but in which f do not participate, simply cannot put across to me a deep understanding of their relationships and consequences—in short, of their ecology. Communication into and out of a computer by way of teletype, graphics terminal, and high-speed printer do not allow me the range of interaction of which I am capable, even if I am there all the time the program is running. If the events taking place inside that computer bear no signature of my presence in the room or of my shifts of attention, then my understanding of their message my “grasp,” literally of their interactions—will be only superficial and can be slated in “once-removed verbiage” that goes into a report. We have never yet asked our informa­tion-handling systems to participate with us in an intimate dialogue. Our computers are condemned to be gloriously fast and accurate basket eases, unable to discover meanings in the world around them arising from their own responses to that world. Likewise, we cut ourselves off from under­standing of the processes internal to the machine and to their relevance to our world.

Ask most any computer expert (the field has acquired the label of Artificial Intelligence) a question like: “Why can’t computers do...that I can do?” and the answer you are likely to receive will inform you that you must wait until either (a) the available machines are more complex and faster or (b) our programming techniques have become clever­er, or both, Such answers always characterize the future improvements in service as arising from a sort of generational or parent-child transforma­tion, a phylogenic evolution (that is, life changes from parent to child). Almost never will the answer convey a problem of an ontogenetic nature (that is, the development of an individual in his own lifetime as con­trasted with generation changes): the education of an already sufficiently complex system, but one not yet sufficiently sophisticated in the sense of having an ability to respond to metaphors in a diverse set of contexts which it has explored through its own self-referent behavior. Often the same attitude is taken toward children: pumping them full of something which they must observe passively—programming them to be good little computer-like quiz-show contestants. No wonder they don’t learn, and no wonder that we have mistakenly revered our computer systems that can “learn” that way! Friend, you and I were students like that ourselves once.

The future computerization of our environment does, of course, arouse uneasiness in us because of the threat of the loss of privacy and of the “big brother” image of computers trying to control us. But that arises out of the belief that computers are intended only to speed up, regulate, and proliferate cheaply the kinds of operation that the industrial age has held so dear. It is only now becoming clear that there are human behaviors at least as valuable as repetitive production skills which could be enhanced in their elegance and range through adequate interfacing with a high-speed data processor,

The computerniks overwork the acronym GIGO (Garbage In, Gar­bage Out) because they think of the machines as having single inputs and single outputs. They think that what they are paying for is a single transformation upon their data as they pass through. This is much like looking upon a saddle horse simply as a consumer of grass and producer of manure.

As for the loss of privacy, let us look at what one is trying to protect. Personally, I believe that in the years to come there is going to be a profound shift in the kind of information for which people wish to ob­serve discretion. Data about me such as my yearly income, bank balance, age, weight, race, schooling, marital status, or special hangups will simply be seen for the irrelevant, nondescriptive list that it is. Again, taxonomy, the naming of things, will take a back seat to the experience of interacting with them. What will be private is the manner of my dialogue with the system or with anyone else—a process that exists only during the dia­logue itself, and exists only for the two or more parties actually engaged in it. Taxonomic data is meaningless except for its context of listedness.

O.K., enough of my carping at what is. Let’s take a look at what the mysterious processes arc that can lead us to what could be. It had been apparent so far that I am leading up to the demand for a means of in­formal dialogue with a machine so that we may explore the other in a process of evolutionary change, Where is the gimmick that takes us be­yond what technology has been offering to date as an adequate interface with computers and other complex systems?

Consider a somewhat exaggerated and simplified description of what is involved in dialogue. 

A and B are talking and each has in mind an idea he wishes to con­vey to the other.

If we were to diagram the interchange of information according to our usual notions, there would be single arrows pointing each direction to signify the flow of information, and A would be thought of as con­structing gradually.

The situation thus described is no different than sending a letter or a tape recording, or entering a program by keypunch or teletype. We certainly expect that being face to face with a respondent makes the job of putting across an idea so much simpler than by mail or dictation. Why? Aha, you say, there is instant feedback in a direct con­versation! Yes, true, but of what? What does the other person send back to you to reassure you that he is grasping your idea correctly? If he just parrots words back verbatim, that’s not very reassuring. He might as well be a tape recorder himself!

In order for A to ascertain that B is comprehending, A must have begun to build a model for himself of “B-having-A’s-idea.”

That is, he must have a notion of what the idea would sound like if B had expressed it himself. 

For that to happen, A must have acquired sufficient knowledge of B to be able to anticipate B’s words and gestures and postures (the better A knows B, the less new information is needed).

Now comes the gimmick that makes it work—the little twist that makes this way of thinking different. Up to this point one might well have imagined all of the foregoing manipulations and model building to go on inside the head as “gedanken-experiments” or pure imaginings.

You are A; you build a model of B in your head; you express your idea; you watch B’s responses and you compare them with your model; and there you are. Right?

No, there is more to it. It is necessary for you to play the role of B outwardly. That is, as A you express the idea, but simultaneously you must be playing the role of B so that you receive A’s transmission sim­ilarly to the way that B receives it and then compare your own responses in B-role to those actually expressed by B.

To put it another way, which is perhaps easier to think about, both parties to a dialogue are involved in self-referent behavior that passes through a part of the environment which they share in common. The intervention of each other’s complex feedbacks to himself produces the exchange of knowledge. Physiological organisms are not good at having “knowledge” of something in which they have not themselves participated. What I retain of the meaning of an object or event which has occurred in my world is not the object or event itself but my response to it.

Let me give you another example. I could take you into your own living room where the objects are very familiar to you. You have seen and touched and handled them many times. J now ask you to stand fac­ing away from me with your hand behind you, and I ask you to maintain your hand outwardly unresponsive—as a passive receiver of inputs; then I take one of those objects, touch it to you, move it over your hand, and rotate it so that eventually you have “felt” it in every way that you have ever felt it; and I bet that it will be a perplexing task for you to name the object for me. The task of identification is not a simple matter of pattern recognition when it is not you who is forming the pattern. If I now allow you one fleeting grasp, recognition will be immediate! The way in which the object intervenes in your act of grasping and thereby modifies that responsive act is your experience of it.

Consider love-making as an example of dialogue. If C and D are making love, the art is not to be found in C doing something to D and vice versa, even though that’s the way the sex-instruction books depict it. Love-making is an art which requires performance upon a very complex but subtle medium in order to convey the message. That medium is pro­vided actively by the other person. D knows how to enjoy D-sclf; C then intervenes as artist, using D’s own self-enjoyment as a medium for the en­hancement of that very enjoyment. The intervention is mutual and the complexity of the multiple loops of dialogue cannot be dissected or recorded or set down on paper for the communication is “immediate.” That part of the environment which the lovers share is so encompassing of each that no observer could possibly decipher the messages exchanged. The communication is private in the strictest sense.

Suppose we are investigating a complex system which has been modeled in a computer program, instead of a “display” being provided of the changing values of the variables as a set of graphs put out on paper, we are provided with an animated display of the variables and their interactions on an oscilloscope, graphics terminal, and in addition a “hand-control” of a rather unique nature. Think of it simply as a box containing a bunch of gas-filled bags with sensors interspersed between them. There is a hole through which you can plunge your hand and which allows you to feel your way around in the box to explore the changing shapes and consistencies of its contents, The pressure in each bag represents a variable in the system, but you are not initially aware of the specific assignment of variables to the bags nor of their relative mag­nitudes of response.

As you grope around and explore the hand-environment, you can watch the changes in the variables that you are producing because they appear on the animated display. You will find, in fact, that it is now really animated because it acts as an extension of your own perceptual system: It responds to your outputs and is no longer just a display that presents something to you as a passive receiver of inputs. The program will be engaged simultaneously in its own “playing” with the system and with you as it maintains the relationships that have been modeled within it. Your own curiosity about the relationships, their rates of change, their inflection points, and so on may be satisfied until you truly have a grasp of them.

Connecting two or more such devices to the machine allows a dia­logue to take place between people in which people do not have to talk in terms of equations, coefficients, or by pointing to graphs. They now can participate together and “converse” through simultaneous peturbations of the system. What might formerly have been an intended mean­ing that was difficult to express in words or on paper becomes: “There, feel that? That’s what I mean.” The language is an informal one, expres­sive of direct experience, and need not be set down in symbols in order to be known. The artificial organism with which the hand plays becomes a sort of dynamic sketch material of high complexity with a behavior of its own which you can shape by intervention. The dialogue does not re­quire equipment of a precise design; a tradeoff has been made where exactness of structure may be sacrificed for liveliness of interplay and control.

The hand control is not simply an output device to display changes which are responses to inputs entered elsewhere, nor is it an input device whose messages are to be transformed into graphical displays. It is at least both: a way to provide you with the opportunity to participate in an on-going, self-referent process that enables you to use your responses as the metaphors of its expression about what the relationships are between its parts.

No measurement is to be made directly upon you, but rather the system senses its own movements within a changing environment. Tile changes are the changes you have produced by your playful intervention. No electrodes or needles or calipers are necessary.

There is no handbook required to tell you what to do. The system is, in a sense, adjusting its own knobs all the time and inviting you to par­ticipate in the adjustments.

Consider another application of computers to the enhancement of man’s reach into a process: computer-aided design. There is much value to be found in employing a system that can display a design as a graphi­cally completed object that can update changes automatically, straighten lines, iterate subcomponents, or show how a circuit or mechanism works based upon the algorithmic operations the computer can call forth to rep­resent the parts. All very good! It speeds things up and performs all of the drudgery and record keeping and repetitions. But it really is not much more than the automated factory all over again. The designer still has not been offered the means available to participate in the shaping of a system in form and behavior which all the while is shaping itself in its search for the best way to embody itself in its search for the best way to embody the designer’s intentions regarding the ultimate result. The pos­sibilities in this approach to man-machine dialogue for the purpose of teaching design as well as for the performance of it are tremendous!

Suppose that you want to teach someone about the complex relation­ships of an ecological system, or an economic system, or a microbiologi­cal system. The teaching process is usually a verbal one with a few lifeless diagrams thrown in to save a few thousand words, And if you press the teacher as to why the process must be verbal and not participatory, you will eventually be informed that the system under study is inaccessible. Its rates of change involve years or generations or microseconds; it exists in a macrosopic or a microscopic world—-unreachable. Must we really continue to believe that?

The point that I am trying to make is not simply that “nonverbal communication is a good thing,” but rather that if you can give a person an opportunity to participate in a process, it suddenly becomes real to him and he can begin to think about it in his own vernacular of experi­ence. Why is it that all of the world’s most astute historians are relatively old men while the prime age for innovative mathematicians is about fifteen or sixteen? Could it be that if the time frame of history could be compressed to a real-time participatory experience for a growing mind, that profession might also receive its greatest impetus from thinkers who have not yet themselves been trapped into the accepted structures of thought and expression? I would bet on it.

Heretofore it has appeared to be most axiomatic that in order to have come to know a system or a technique really well, you have built an ex­ample of it for yourself (made a piece of clothing, cooked a meal, wired a circuit, built a house, climbed a mountain). The opportunity to inter­act with the materials or distances or problems of detail was only to be found in being yourself the immediate and substantial creator of the object or event, no matter what physical duress or perceptual stretching are necessary for the feat. I contend that that is just not the case any­more. Direct experience in a distribution of time frames which may be varied over a continuous range is now available so that we may discover relationships within those experiences that might never have come to our attention in any direct confrontation with reality and its inflexible media.

It should also be pointed out that the use of technological devices for the demonstration that “turns the observer on” to relationships that were not there for him before need only be temporary until he has a “feel” for them. The technology takes the role of a ’’sketch material“ and should not be thought of as a crutch or prosthesis which must be used in per­petuity. Once you have been shown the pattern of fields and forests from an airplane, they are forever different for you on the ground. Once you have witnessed the workings of your own hand in a fluoroscope, the structure of its bones may be thought about and played with thereafter from a new standpoint of insight.

Now let’s try to zero in on a few “how to’s.” My purpose is not so much to suggest to you a way to put a project in motion within your own company starting with “off-the-shelf” componentry (some of the essen­tials aren’t on the shelf yet), but rather to give you enough of a picture of what to watch for so that you may be alert to the possible alternatives when they become available. They will soon.

I must beg off from the task of listing any directly pertinent biblio­graphy. Frankly, aside from the influences which my colleagues and I at the Environmental Ecology Laboratory (EEL) have felt from previous workers (the salient force among them having been Warren S. Mc­Culloch), we know of no other similar work currently under way. The problem may simply be our ignorance, for we have felt no embarrass­ment at working in a vacuum. We do watch constantly in some of the more ubiquitous journals for indications of a trend in this direction, but they are few and somewhat disappointing, Therefore, please take what I say not as “truth” to be set down and pointed at, but as a set of notions to be played with because they workl

The problem might appear to be twofold: (a) machines must be given ways in which to explore their environments (the inside of a print­ing press, the spaces surrounding an airplane to a radius of miles), and (b) the user must be provided a means to start his dialogue with the machine without having first to master a complicated symbolic language or being required to engage in any prior training. However, if you can allow yourself to go beyond the necessity of thinking about those two parts of the problem as separate, then you will have made a long jump into this new way of thinking. Why not look upon the user as simply another environment which the machine is called upon to explore and to “satisfy” will) dialogue? The selection of an appropriate mechanism to transact communication between the machine and any of its environ­ments will be found to afford many similarities among them, but perhaps the most challenging one is the man-machine interface. The user’s ca­pacity for the exploration of information spaces and for the extraction of meaning out of redundancy and noise has never been matched, nor has it probably ever been allowed full flower. What follows is not a descrip­tion of devices or programs, but some of the more useful rules of thumb that we at EEL have come across in our work.

In order to allow a participant a perceptual reach into the system he is dealing with, you must provide him with multiple channels among which (N.B.) at least one sensor-effector pair are in a common modality. What I mean by the latter is this: The difficulty with visual or auditory displays generally is that the eye of the observer does not give out any light of its own nor the ear any sound. More simply stated, the way the perceiver looks or listens has no direct effect upon the information being presented to him. In such cases he must be a passive observer. In our culture we are highly trained to be passive observers and we find it diffi­cult later to imagine why the brain cannot be described adequately as a complicated computer with lots of inputs and outputs that may be used irrespective of their lack of interaction with each other. Our tendency to isolate sensory and motor experience allowed our abilities in symbolic language to become highly developed. But this has served to degrade our capacity for the more informal processes of dialogue. An obvious result is that we are apparently content that computers be manipulators of sym­bols incapable of direct participation in events around them.

One of the routes to man-machine participation in a “common sen­sorimotor modality” is to allow the computer to be aware of where and how the man is looking at the graphical information being presented to him. Another is the manual or whole-body interface in which the in­formation is conveyed through touch and movement. Such is the com­munication common to lovers or to mother and child: that which touches also moves and vice versa; no symbolic transformations need to be per­formed between sensation and response.

I want to make it clear that I am not espousing wholly non-verbal communication between man and machine or man and man but rather that the alternative modes of direct experience may serve to provide the kind of contextual emphasis or shifting which can speed up the information transfer rate immensely. McLuhan’s observations on the “cool” of TV versus the “hot” of radio are an example of the fact that with both sensory modalities present, contextual shifts in either may be made far more rapidly by the use of the other than they could through one alone. Both visual and aural display may then be of higher speed and lower resolution than formerly. Remember in the old silent movies how ex­plicitly every action had to be carried out to get its purpose across? Or when you describe over the telephone something that you have right there in your hand, don’t you feel confined by a strictly verbal medium? At least in a letter you could draw a sketch!

If a salesman comes to you with peripheral hardware for your machine to interface you better with it, I am advising you to look for playfulness rather than precision in its operation, controls that move and change under your hand and have an active life of their own rather than just sitting there awaiting passively to be pushed or turned; displays that take account of your attention to them and change accordingly: and readouts that are of a less permanent and symbolic nature but which start to demand that your machine console be something you can interact with when you want information. Printouts are for wrapping fish and starting fires. Why, when that machine is there to give you access to in­formation, should you have to find some remote column in yesterday’s printout so that you can check something out?

Let us all conspire together to try to rescue the computer from its congenital amputations by demanding that the machines be allowed sen­sitive extremities that can explore for themselves the contexts of the tasks they are expected to perform. Their potential service to us can be virtually unlimited if we recognize the value of informal dialogue and learn not to ask the usual scientific questions about those tasks. We are not looking for an algorism of human behavior nor for a facsimile of “thinking” in our machines. Rather, let us think of them as having a potential for enhancing our reach and grasp by interfacing us in active dialogue with a world of relationships delightful in their complexity.