PROCESS DUALISM:

Cybernetics and the Mind/Body Problem

by Robert Johannson

APPENDIX A: MATERIAL PROCESSES ARE ENERGY TRANSFORMS

The basic human experience is the experience of change. The basic sentence structure of noun and verb reflects the constant and the changing. The basic metaphysical question "What is real?" and its cognate "What is the really real?" has often been interpreted as "What is constant?"

The original inspiration for materialism was Aristotle's concept of "substance." All the different aspects of things were "accidents" and subject to change, but the accidents adhered in an unchanging substance.

Democritus is usually credited with the idea of atomism, the attempt to break down matter into its basic elements. This lead to the development of the periodic table of elements and ultimately to sub-atomic physics. But as the elements get smaller, the basic and constant "elementary particle" seems as elusive as ever.

No serious materialist has ever been concerned with matter simply as matter. We are concerned with change, with movement, with matter in process. The great developments of the last few hundred years have been in the development of the idea of "energy."

Aristotle thought that a static position was normal, and change of position had to be explained. Galileo saw that it was not the position of the object that was constant, but its velocity. That meant that it was change of velocity that had to be explained. Newton developed this idea further with the idea of force causing changes in velocity. Force could be calculated with the formula F=ma. Force is a vector having both magnitude and direction, thus it is symbolized by the arrow.

Since all change came to be conceived of as the result of forces the arrow became the common symbol of cause and effect. On this understanding of causality lay the materialist hope that a knowledge of the direction and quantity of all forces would lead to the ability to predict all future events, and to deduce all past events. All things were understood as determined by "forces."

This is the root of the materialist metaphysics. Reality is made up of atoms or particles. These atoms are in motion. Their relationship to other atoms is changing. This change is determined by forces. Therefore to understand any change it is only necessary to understand the forces at work and put them in the appropriate atoms in the appropriate space.

The idea of force, however, was found to be insufficient. It could not explain the new heat engines that were being developed. Forces did not simply interact they were transformed. Coal burned and produced heat, the heat applied to water created steam, the steam could then drive a steam engine. It was in attempting to understand the operation of these new engines that the concept of change as energy transform was developed.

Energy is force through space. To exert a pound of force through a foot of space is to use one foot-pound of energy. If we look at material processes we see that they can be interpreted as transforms of energy. Potential energy can be transformed into kinetic energy or motion. One kind of motion can be transformed into other kinds of motion. Motion can be transformed into electrical current. Electrical current can be transformed into heat and light.

The laws of energy transforms are the two laws of thermodynamics: conservation of energy and entropy. The law of conservation of energy affirms that in any energy transform the quantity of energy will remain constant. The law of entropy affirms that the transform process itself will consume energy.

Transforms are reversible. The energy can flow either way. The commonest experience of this is when driving a car. If you take your foot off the gas pedal, the engine will change from a source of energy and motive power to a source of braking power. The ability to reverse the process depends on the nature of the transform mechanism, and of course, entropy.

Some physicists argue that it is entropy, that accounts for the irreversible quality of time. Time has an irreversible direction towards a universe of uniform temperature also known as heat death. This involves the other interpretation of entropy that heat will not pass from a cooler to a hotter.

In an energy transform the quantity of energy remains constant, while the measures of force and distance change.

An example of a simple energy transform is a lever such as we see in this figure:

The basic idea of an energy transform is the conservation of energy. The energy in FORM A is equal to the energy in FORM B.

For the purposes of our analysis this is a sufficient description of materialist analysis.

Modern Physics has, of course, gone further. But it has moved in the direction of a greater concern with form. The very idea of an energy transform mechanism is a formal idea. A formal change in the mechanism of the transform has a profound effect on the structure of the transform. Thus in our example of the lever, a change in the position of the fulcrum will change the ratio of the transformation. A change in the form of the system is a change at a higher logical level.

It is worth quoting Einstein and Infeld (1966) on the wave.

"We have all seen the waves that spread in wider and wider circles when a stone is thrown into a pool of water. The motion of the wave is very different from that of the particles of water. The particles merely go up and down. The observed motion of the wave is that of a state of matter and not of matter itself. A cork floating on the wave shows this clearly, for it moves up and down in imitation of the actual motion of the water, instead of being carried along by the wave...The particles constituting the medium perform only small vibrations, but the whole motion is that of a progressive wave. The essentially new thing here is that for the first time we consider the motion of something that is not matter, but energy propagated through matter."

Albert Einstein and Leopold Infeld, The Evolution of Physics. pp 100-101

A wave cannot be perceived at the level of particle motion. It can only be perceived at a higher logical or holistic level. Relativity theory is a formal theory. This concern with form leads on to the modern study of chaos. Chaos is a purely formal term.

APPENDIX B: WHAT IS A CYBERNETIC SYSTEM?

Before we can begin to determine if cybernetics has in fact solved the mind/body problem, we first have to define what a cybernetic system is. This is not as simple as it would at first appear. There is no accepted standard description. Each writer presents his own definitions and analysis. It is necessary therefore to look at a number of different descriptions and see if we can derive the basic elements of a cybernetic system.

CYBERNETICS AS TRANSFORMATION

In Norbert Wiener's (1948) book Cybernetics, we find a diagram of a cybernetic system. It is shown in this figure:

Norbert Wiener was a mathematician who worked on radar guidance systems for anti-aircraft guns. His method of diagramming a cybernetic system is still used among electrical engineers.

Thus Lathi (1974) presents the schematic in this figure:

The process that both diagrams present is a process of transformation. A "transform" is a process where some aspect of a thing remains constant while other aspects of it change according to a system of correlation. Mathematicians call the correlation "mapping." Elements in Set A are said to be mapped "onto" the elements of Set B. The symbol that is used is an arrow. A "arrow" B indicates that Set A correlates to Set B.

A transform always has four parts. There is 1.) the variable in the first set, 2.) the process, 3.) the process variable, and 4.) the variable in the second set. In the formula x + 2 = y, the "x" is the first variable, addition "+" is the process, "2" is the process variable and "y" is the variable in the second set. In these diagrams the processors are symbolized by the boxes, and the variables by the lines that go into (input) and out of (output) these boxes.

Mapping theory obscures the processors, while the Wiener/Lathi diagrams highlight them. The processors need to be highlighted because the process is determined by them. They are the limits of the system, and a change to the processors is a change to the operating characteristics of the system. This is an important point. It means that you can change the whole system by changing a part. Thus we can begin to talk about the system as a whole.

To symbolize the double sided and matching quality of the transform we have come to use two wavy lines as illustrated in this figure:

This symbol allows us to indicate both variables and the processor, and thus separate out the four parts of the transformation process. But whether we uses boxes, arrows or wavy lines the same three part process is being symbolized.

If we look closely at the diagrams presented by Wiener and Lathi we see that in a cybernetic system there are basicly three processors and three variables. The processors are 1.) the controller, 2.) the effector (or the plant), and 3.) the feedback take-off. The feedback take-off is not symbolized by a box, but by a small circle. If we look at one of Lathi's exemplary diagrams we see that the feedback take-off is a processor. In this example, the feedback is one percent of the output.

The three variables are called input, output and feedback. The terms "input" and "output" raise basic boundary issues that are endemic to systems analysis. What can be said to be "inside" the system and what can be said to be "outside?"

CYBERNETICS AS INFORMATION PROCESS

Ludwig Von Bertalanffy was the founder of modern general systems theory. He saw cybernetics as included in the general theory of system. The basic concern he had was with the reality of the part-whole relationship. He was a biologist and it was clear to him from his study of biology that you could not think of a living being as simply a collection of parts. The nature of the whole determines the nature of the parts. The cat does not hunt because it has claws, it has claws because it hunts. The claws are only one part of a survival strategy dependent on hunting.

The relation of the part to the whole is the problem of holistic structures. Holistic structures develop a number of levels. Each level is a whole to the levels below it, but it is merely a part to the levels above it. This is diagrammed in this figure:

The problem of holism is related to the problem of boundaries. For the moment let us look at his diagram of a cybernetic system. In Von Bertalannffy (1968) there is the following diagram of a cybernetic system.

It is similar to the diagrams of Wiener and Lathi, but there is a significant difference. The processors are the same with one slight change. The controller and the effector are still there but the feedback take-off is called a "receptor," and the terms for the variables have been changed to stimulus, message, message, and response. He makes a distinction between information (messages) and action (stimulus, response). With this model it is not entirely clear what the control apparatus does to change one message into another, nor the difference between a stimulus and a response since both are inputs for the receptor, but we will take this up later. For the moment what is important is the distinction between information and action.

The distinction means that there are two very interesting transformations going on in the cybernetic process. One is the transformation of activity into information, and the other is the corresponding transformation of information into activity.

MODELLING CYBERNETIC SYSTEMS

Now for something completely different. One of the problems that cybernetics has faced from the very beginning is the difficulty of expressing the cybernetic relationships in terms that would show the information transforms.

A diagram that indicates the information element more clearly is the diagram developed by Jay Forrester (1968) in Principles of Systems.

Forrester is remarkable for continuing the M.I.T. interest in systems and seeking to apply cybernetic theory to social phenomena.

In Forrester's diagram the two different levels of operation are clearly delineated. On the first level you have the materials flow, the world of particle movement, of material and efficient causes. This level is indicated by the solid lines. On the second level you have the world of information, of mind, of formal and final causes. This level is indicated by the dotted lines. (Not quite as solid as material flows.)

It is quite different from the diagrams we have been looking at. But we can recognize some of the familiar elements. The small circle on the edge of the box is the feedback take-off. The dotted line leading away from the take-off is the information variable. His diagram obscures the difference between the controller and the effector by combining them into one. But he does make clear what it is that they are doing.

His symbol for the controller/effector is a gate or valve. The effector takes material or energy from a "source." and directs it in specific ways towards a state, symbolized by the box. The symbol for the source is a blob with squiggly lines. This indicates that the source is the boundary of the system. It is the point at which materials enter the system. He uses the same symbol for a "sink," the point at which material or energy leaves the system. It is what Von Bertalanffy would call an "open" system. He thus stresses the importance of a collateral energy system. The other aspect that he adds is the time adjustment. The activity of the effector is a "rate," an activity over time.

Although he does not separate out the controller from this process, he does make clear what it is that the controller does. The controller takes the information about the state of the system and compares it to a goal. The goal is a specific value that is the desired value of the state. The dotted line connecting the goal to the Rate indicates that the "goal" is a form of information.

Although not entirely clear from the diagram the box also serves as a processor. It takes the product of the rate and combines it with the present situation. This is in fact a fourth processor that is an essential element of a cybernetic system.

Forrester was the first to try to see the whole cybernetic process as a series of information transforms. He wanted to develop a computer model of a cybernetic system. A computer program is by definition a series of information transforms. He accomplished this with the world model in Limits to Growth.

The advantage of this method of modelling a cybernetic system is the ease with which it can be translated into computer programs. If we were to code this model into BASIC it would look something like this.

This program would print a series of inventory figures for a a period from a specific date "d" to another date "c" at intervals of "a." Note that since it is a computer program there is no receptor or feedback take-off. The inventory is already in the form of data. This is a major limit to the Forrester method that he tries to overcome by the use of "co-efficients."

DIAGRAMMING A CYBERNETIC SYSTEM

At this point we can identify the four basic processes of a cybernetic system. It will have four basic processors:

1. a feedback take-off or receptor, which translates external states into internal information, a process I call SENSING;

2. a controller which compares the feedback information to a value or desired goal, a process I call EVALUATING;

3. an effector or plant or rate which transforms external material or energy into specific actions, a process I call CHOOSING;

4. a situation or environment that transforms the specific actions of the system into specific results or changes, a process I call ACTING.

Corresponding to the transformers there are four variables (elements of information) which are transformed:

1. The external changes that are sensed I call CHANGES;

2. The internal information that this produces I call SENSE;

3. The information after it has been evaluated I call EVALUATIONS;

4. The action of the system upon the situation, I call ACTION.

We can then diagram the processes of a cybernetic system in the following way.

In the Forrester example, the process of EVALUATING would be DI - I = E. I is the SENSE or feedback that tells us the level of the present inventory. DI is the VALUE or goal for the inventory. E is the EVALUATION. The process of EVALUATING would be subtracting I from DI.

This method of diagramming also brings out the process variables. These variables are the limits or constraints of the system. I have called them MODELS, VALUES, RESOURCES, and SITUATION.

DIAGRAM OF A CYBERNETIC SYSTEM

For the sake of simplicity we came to diagram a cybernetic system like this:

Simplified Cybernetic Diagram

APPENDIX C: DEFINING INFORMATION

In order to understand the way that cybernetics looks at information we have to define information and communication. Mind has traditionally been understood to be the area of information and communication. But we have to look closely at the definitions to see if they are in fact the same, or whether we are calling things "information" and "communication" that are not what we usually mean by those terms.

The concept of ideas as the fundamentally constant in a world of change goes back to Plato's idealism. Plato felt that there is an ideal realm where the ideas are the reality. The earthly world since it was subject to change and decay was merely a shadow of this "real" world of ideas.

We have a tendency to think of mind in static terms, as an idea or a category. But Mind is best considered as an active process. Descartes did not say "I have ideas therefore I am." He said "I think, therefore I am." It is the processes of mind that we are primarily concerned with. The concept of ideas changing developed into the historical dialectic of Hegel. Another powerful analysis of ideas has seen the processes of mind as the processes of communication.

The Shannon and Weaver (1964) diagram of the communication process is illustrated in figure 3.

Shannon (1964) comments:

"The fundamental problem of communication is that of reproducing at one point either exactly or approximately a message selected at another point. Frequently the messages have meaning; that is they refer to or are correlated according to some system with certain physical or conceptual entities." p.31

Shannon uses a set of symbols similar to the ones used by Wiener and Lathi. The boxes represent the three processors: transmitter, receiver, noise addition. Boxes are also used to symbolize sources and destinations. (This is the recurring problem of symbolizing boundaries.) The lines between them indicate the variables. Shannon distinguishes three kinds of variables: message, signal and noise.

INFORMATION IS CHOICE

An essential part of the Shannon-Weaver model of communication was the definition of "information." Information is a formal term and thus involves the problem of defining form.

The approach that Shannon took was to define it in terms of the choices in a set. The basic concept of information is the concept of either...or, choice within a limited set of alternatives. Given a particular set of options, the information involved in one element is related to the number of choices available in the set. For example in the English alphabet there are twenty-six letters, therefore every letter in a word has 26 elements of information in it. (Twenty-seven if you count the space.) This number can then be manipulated mathematically. It can be translated into "bits" (binary digits) and used to calculate questions of redundancy, probability, and entropy.

Shannon's analysis allows us to see number as choice within a set of alternatives. This is comparable to the shift from Roman numerals to Arabic numerals. The Roman numerals began with the idea of counting fingers. Our word "digit" is from the Latin word for "finger." "V" was a whole hand and thus "IV" could be conceived as one less than a whole hand, and "VI" as one more than a hand. These are usually called "natural" numbers. In contrast the system of Arabic numerals was a system of choice within a positional set. The meaning of each numeral is determined by its position. Each position is a set: ones, tens, hundreds, thousands, etc. Each numeral is a choice within that set. For this system "zero" is essential for it allows for a null set. This translates easily into the "one or zero" of binary. With Roman numerals binary is inconceivable.

This binary definition of form is easy to see on a video screen. The presence of an illuminated dot is a choice between alternatives; either hitting the phosphor dot with an electron flow or not. It is either x or not x.

The system of "real" numbers which conceives of a number as a point on a continuum is a combination of the positional definition of arabic numerals with the continuum as the set within which choices are made, and the natural number system where each point is given a number. This combination of the "unit" within the "continuum" gives rise to the paradoxes of Zeno.

It is easy to see how these definitions shade into statistics, for statistics is concerned with the frequency of occurrence of alternatives within a set.

When we compare this definition to the usual definitions of information, we see that they are the same.

The study of form is called logic. Logicians tell us that the other logical operators can be derived from two: "v" either...or, and "-" not. These two operators create sets. This corresponds to Shannon's definition of information as "choice within a set." And Bateson's definition of information as "a difference that makes a difference."

If we look at the standard symbol of a category or set, the Venn diagram, we can see the basic disjunction implied by any set. The Venn diagram is also a binary form.

A set or a category is basicly a boundary. There can be null sets, because a set is simply a boundary, a distinction. All boundaries have a basic characteristic. There can be no inside, without an outside. There can be no {x} without a corresponding {-x}. Deductive logic is based on this assumption, usually called the law of the excluded middle. Even fuzzy logic does not challenge this, only the clarity of the distinctions or boundaries. If there are no distinctions then there is no information, no sets and no logic.

What a disjunctive basis implies is that choice is the primary act in the creation of information. The set is a result of the choice. Without a choice there are no sets. Every category is the expression of a choice that creates a distinction. To choose is to create alternatives. To choose is to establish boundaries.

This implies that all boundaries, sets, classes and categories are arbitrary, since all categories are a product of choice. Boundaries are real, because choice is real. The boundaries can be true. They can correspond to something beyond themselves.

If categories are a product of choice then categories are historical. There is a time at which they are chosen. Choices create history.

The experience that convinces me that choice is the basis of experience is the experience of the Gestalt shift. We have the familiar picture that can be looked at in two different ways. In the first instance we do not see two different pictures. In the first instance we see one picture. When we finally see the second picture it comes with an "aha" of surprise. We are not aware of the perceptual choice until it is made.

The neurological basis of this is the neuron. The neuron makes a choice, either it fires or it does not fire. It was McCullough's exposition of the neurological basis of perception that lead to Von Neumann's development of the digital computer with its switches that are either on or off.

INFORMATION IS IMMATERIAL

Bateson (1972) saw information as a difference, or more precisely a difference that makes a difference. Bateson was concerned to show that the formal characteristics of information are not material. A difference does not have "extension."

The point here is so important that it is worth quoting Bateson (1979) on the nature of the switch.

"The entities or variables that fill the stage at one level of discourse vanish into the background at the next-higher or -lower level. This may be conveniently illustrated by considering the referent of the word switch, which engineers at times call a gate or relay. What goes through is energized from a source that is different from the energy source which opens the gate...

We do not notice that the concept "switch" is of quite a different order from the concepts "stone," "table," and the like. Closer examination shows that the switch, considered as part of an electric circuit, does not exist when it is in the on position. From the point of view of the circuit, it is not different from the conducting wire which leads to it and the wire which leads away from it. It is merely "more conductor." Conversely, but similarly, when the switch is off, it does not exist from the point of view of the circuit. It is nothing, a gap between two conductors which themselves exist only as conductors when the switch is on.

In other words, the switch is not except at the moments of its change of setting, and the concept "switch" has thus a special relationship to time. It is related to the notion "change" rather than to the notion "object." p.116

The switch has a formal or informational character, and like all formal distinctions it raises the ontological question: "In what sense can the choices not chosen be said to exist? Clearly they are real, but they do not in the strict sense "exist."

INFORMATION IS HOLISTIC

Once we accept the formal nature of material events, that they involve choices within a set of possibilities, we are operating at a holistic level.

It is with the set or category that we have the possibility of a holistic analysis, and the development of logical levels. We must make a distinction between two different kinds of parts. The set allows us to conceive of a whole that is made up of parts that are elements. The elements all share the same distinguishing characteristics, and thus are part of the same category. This is the quantitative analysis much like marbles in a jar. But there is a second kind of part that is formed by the intersection of two or more sets. Each of the two sets is a part of the intersecting set.

Our concept of an element or a unit is not the set itself but the intersection of two sets. The point where something enters our discussion is the point where it can be named. The point where anything can be named is the point at which at least two choices have been made. The traditional form of a definition is the "simuls et differens." (What category does it belong to and what makes it different from the other members of that category?)

This is the reason that all named sets are fundamentally fuzzy. No matter how absolute and abstract the distinction there is always a second quality that creates a continuum. Analysis is the search for the ultimate distinctions that will eliminate the fuzzy quality of names. But that distinction is probably nameless.

Shannon's definition of information as "choice within a set" is equivalent to the treatment of ideas in deductive logic. This definition also implies that information is immaterial. And it is the formal dimension that allows us to look at things in a holistic way.

APPENDIX D: DEFINING MEANING AS A TRANSFORM

The Shannon-Weaver analysis also draws out the essential presence of an information transform in any kind of communication.

In an information transform it is the information that remains constant. "The fundamental problem of communication is reproducing at one point either exactly or approximately a message selected at another point." In other words the fundamental problem of communication is the conservation of information. If information cannot be conserved then there cannot be communication. (If there cannot be communication then there cannot be Mind.)

Shannon makes a clear distinction between the concept of "information" and the concept of "meaning." Anything that involves choices within a set has "information." But meaning implies that "they (the messages) refer to or are correlated according to some system with certain physical or conceptual entities."

For example. "the cat on the mat" contains information. It involves choices within a set. Speaking quantitatively it contains twenty-seven to the eighteenth power elements of information. "ypez ayzgiypeday" also contains information. Exactly the same amount of information, twenty-seven to the eighteenth power. But "the cat on the mat" has meaning, and "ypez ayzgiypeday" does not. "the cat on the mat" is correlated according to some system with a set of English words. Those words are themselves "correlated according to some system with certain physical or conceptual entities."

Each correlation is an information transform or transmitter. The transform is the basic unit of communication because it is the transform that determines the meaning of the information.

The "meaning" of any given item of information will depend on the set that it is correlated with, and the system of correlation. Identical information can have a variety of different interpretations depending on the set that it is correlated with.

live cat - NAMING - French word "la chat"

live cat - NAMING - English word "cat"

The basic insight of gestalt psychology was the importance of this correlation for the interpretation of experience. Meaning also depends on the form of the correlation. We may wish to correlate some people with the set good guys and bad guys. But how they are correlated? Some people see Americans as good guys and others see them as bad guys. The sets are the same but the mapping is different.

live dog - NAMING (STANDARD) - English word "dog"

live dog - NAMING (BABY TALK) - English word "bow-wow"

live dog - NAMING (TECHNICAL) - English word "canine"

Consider the classic question, "What ere the stars, Joxer? What ere the stars?" The astronomer looks at the stars and sees energy transforms. The navigator looks at the stars and sees a guidance system. The astrologer looks at the stars and sees guidance of a different kind. Depending on their training, people see the same things in different ways.

One aspect of the Shannon-Weaver model that is implicit is the idea that an information transform always involves two different sets. In the Shannon-Weaver model there is no direct communication. Communication always requires a medium.

In mathematical mapping theory the element in the second set is often called the "image" of the element in the first set. Another way of looking at it is to consider the first set the referent. It is this referential quality that is behind the "subsidiaries" and "focus" of Polanyi. The second set acts as a subsidiary because our focus is on its meaning, i.e. its relation to the first set. This tends to obscure the second set in our perception. If you focus on the letters on this page you will not see the ink that is their subsidiary. Similarly if you focus on the words you will not see the letters that are their subsidiaries. Many people will have been surprised by the earlier example where I used the example of the word "cat," and was referring to the letters as printed on this page.

The referent quality of information transforms allows for holistic structuring. Naming is one aspect of this.

Information transforms are always "about" something. They express a correspondence between ideas, a relationship. It is this quality of correspondence in the information transform that allows it to create different levels of meaning.

The referenial quality of information creates different logical levels. The word that philosophers usually use is the prefix "meta". Another logical level is a level which is about the original level. The original level is the referent of the meta-level. Thus language allows us to talk about things, a meta-language allows us to talk about language.

To use a map as an example.

First there is a territory.

One logical level above this is the map which is about the territory. The territory is the referent.

One logical level above the map is a catalogue of maps, which is about maps. The map is the referent.

One logical level above this is the library's card catalogue which has a list of map catalogues. The map catalogue is the referent.

As well as the logical levels there are categorical levels. The referent does not have to be one thing. It is usually a category of things. The process of jargon development, and technical language is a process by which the technical word comes to imply a wide array of meaning.

It is this categorical quality that allows for the development of language both in humans and in the computer. In the computer we talk about third and fourth generation computer languages. Each computer language is a coding system. As you travel up the generations of the coding system, each coding element comes to imply more in the way of action at a more basic level. Press the "Print" instruction and a whole series of events is coded into that one press of a button.

Similarly when I decide to get up and get a cup of coffee, this sets in motion a whole series of sub-activities, such as standing up, walking, reaching, pouring, etc., that are all part of the one activity of "getting a cup of coffee."

The information transform is not reversible. The process that translates the symbols of English into the symbols of morse code, is different from the process that translates the symbols of morse code into English. The decoding process is complementary to the coding process.

As an example of a information transform we can use a simple cypher wheel. The alphabetical symbols on one set are correlated one to one with the alphabetical symbols on the other set.

The correlation is contained in the matching of the choices in one set with the choices in the other set. In this example they are correlated by position. The fourth letter in Set A corresponds to the fourth letter in Set B, d=n. This kind of correlation is called by mathematicians a one-one mapping of A into B.

For communication to occur there must be a transmitter. Just as the energy transform requires a transformer. So an information transform requires a transmitter, a medium and a receiver. There must be some energy process involved.

To take this correlation and turn it into a transmission we can place it on a typewriter. We could use the correspondence between the letters of Set A and the letters of Set B to design a template so that when the letter "b" on the keyboard was pressed, the letter "s" would be typed. The typewriter would then act as an information transform. A message keyed in English would be printed as code. To reverse the process a decoding template could be placed over the keys. Then a message keyed in code would be printed in English.

English - TYPING (CODING TEMPLATE) - Code

Code - TYPING (DECODING TEMPLATE) - English

Using this coding transform the expression "ypez ayzgiypeday" which was meaningless, now has a meaning.

words - SPELLING - the cat on the mat - CODING - ypez ayzgiypeday

It is the correlation established by the coding transfomr that gives meaning.

Entropy in the information transform process is the loss of information due to the process. In this example the entropy is very low.

DO COMPUTERS REALLY UNDERSTAND?

This analysis now allows us to go one step further. Searle (1980) observes that "symbol manipulation by itself couldn't be sufficient for understanding." We have distinquished between information and meaning. We must also distinguish between meaning and understanding.

The information may have a meaningful correlation, but that meaning can only be understood if the message can be decoded. The expression "ypez ayzgiypeday" did have a meaning, it was correlated with the phrase "the cat on the mat." But the expression could only be understood when it could be decoded.

Thus a computer does not understand the data that it manipulates because it cannot decode it. The process that we have called "understanding," is the same as the process described by Shannon and Weaver (1964) as the process of "communication."

"The fundamental problem of communication is that of reproducing at one point either exactly or approximately a message selected at another point." For there to be understanding there must not only be meaning, but that meaning must be communicated.

But there are degrees of understanding. The computer does not understand the human data that it processes because it cannot decode it into human experience. But it can understand a message from another computer since it can decode it into computer messages that are meaningful for its operation as a computer. Thus the understanding of a computer is a difference in degree rather than a difference in kind.

BIBLIOGRAPHY

Bateson, Gregory 1972. Steps to an Ecology of Mind. San Francisco: Chandler Publishing Co.

1979. Mind and Nature: A Necessary Unity. Toronto: Bantam Books.

von Bertalanffy, Ludwig 1968. General System Theory: Foundations Development Applications. Harmondsworth: Penguin Books.

Einstein, Albert & Leopold Infeld 1966. The Evolution of Physics. New York: Simon and Schuster.

Forrester, Jay W. 1968. Principles of Systems. Cambridge, Mass.: Wright-Allen Press Inc.

Johannson, Robert & Joan 1990. Empowerment: A Systems Approach to People and Groups. Winnipeg: Parent Empowerment Group Inc.

Koestler, Arthur 1967. The Ghost in the Machine. London: The Hutchinson Publishing Group Ltd.

Lathi, B.P. 1974. Signals, Systems, and Controls. New York: Intext Educational Publishers.

McCulloch, Warren S. 1965. Embodiments of Mind. Cambridge, Mass.: The M.I.T. Press.

Meadows, Donnella H. et.al. 1974. The Limits to Growth (Second Edition). New York: Universe Books.

Polanyi, Michael and Harry Prosch 1975. Meaning. Chicago: The University of Chicago Press.

Rosenblueth, Arthuro and Norbert Wiener and Julian Bigelow 1943. "Behaviour, Purpose and Teleology," Philosophy of Science, X. 18-24.

Samuel, Arthur L. 1960. "Some Moral and Technical Consequences of Automation - A Refutation," Science, Vol. 132. 741-2.

Searle, John R. 1980. "Minds, Brains, and Programs," The Behavioral and Brain Sciences, Vol. 3. 286-297.

Shannon, Claude and Warren Weaver 1964. The Mathematical Theory of Communication. Urbana Ill.: University of Illinois Press.

Wiener, Norbert 1948. Cybernetics: or Control and Communication in the Animal and the Machine. New York: The Technology Press.