The Computer Simulation of Behaviour

Michael J Apter, Michael J. Apter

• Publisher: Taylor & Francis
• Year: 2018
• eBook ISBN: 9781351021005
• Edition: 1
• Topic: Psychology
• Subtopic: Cognitive Psychology & Cognition
• Index: Psychology

This book, originally published in 1970, concerns the new technique of computer simulation in psychology at the time. Computer programs described include models of learning, problem-solving, pattern recognition, the use of language, and personality. More general topics are discussed including the evaluation of such models, the relation of the field to cybernetics, and the problem posed by consciousness. Today it can be read and enjoyed in its historical context.

I

BRAINS, COMPUTERS AND MODELS

THE MAN-MACHINE ANALOGY

There are two related ideas concerning men and machines. The first is that machines can be built with the same functional capacities as men; the second is that men are not only like, but really are, complex machines. Neither of these ideas are entirely new and they date from long before the advent of the computer.

The possibility of an artificial man has obsessed men through the ages, from the obscure medieval idea of a golem—a wooden image given life by the saints—to the horror-laden fantasy of a Frankenstein’s monster, the functional twentieth-century robot and the futuristic android (which, as readers of science fiction know, is a robot which is indistinguishable from a human being).

The idea that in any case man, in common with all organisms, is a machine—the mechanist or anti-vitalist hypothesis—is more difficult to trace being caught up as it has been with undercurrents of thought and feeling over the centuries in science, philosophy and religion. The idea that man is a machine implies that man is no more than a machine and that concepts like soul, will, sin and so on are at best irrelevant and at worst totally misleading. Naturally, scientists acting as scientists have methodically disregarded concepts of this kind which do not lead to objective measurement and which are non-parsimonious in explanation. Thus they have inevitably treated man as if he were a machine. In this light the computer simulation of behaviour lies in the main stream of scientific thought, although it perhaps implies a more blatantly mechanistic attitude to the human brain and human behaviour than is usually found in the biological sciences.

Oddly enough, the attitude of the modern scientist is in one respect similar to that of primitive man: men and machines are seen as essentially alike. Primitive man being animistic like a child elevates objects to the level of men by ascribing souls to them: stones feel pain if you kick them, clouds enjoy floating in the air, boats want to move through the water. But to the scientist they are the same not because both kinds of entity have souls but because the idea of a soul serves no purpose in the study of either. Between primitive man and the scientist lies a third view which is probably still that of the symbolic ‘common man’ and which has been associated in western civilisation especially with Christianity. This is the view that man has a soul while objects and machines do not.

If one studies the development of the biological and psychological sciences in modern times one finds a progression from this Christian view to the view that, whether or not man does possess a soul, he can be successfully studied as if he did not. This progression has been associated with various views about what kind of machine man can most appropriately be regarded as.

In the seventeenth century Descartes, hovering uneasily between strong commitment to both Christianity and the rapidly developing science of his time, argued that animals were wholly machines and had no soul while humans were partly machine and possessed a soul.

On the basis of observations carried out during dissections and even vivisections of animals, Descartes (in his tract De Homine) claimed that ‘vital functions’ of the animal body were the result of heat and motion of liquids within the body, especially within blood vessels and nerves. In particular, he saw the nerves as hollow tubes between brain and muscles in which flowed a substance: the movement of this substance in the nerves caused the movements of the muscles and hence brought about behaviour in much the same way as the performance of various moving toys of the period was based on the use of hydraulic tubes. Animals were said to be devoid of consciousness and even thought, and to Descartes the cries of a cat being vivisected represented no more than the noise of breaking machinery.

Man, however, was believed by Descartes to be different: although in many of his functions he was like an animal and therefore mechanical, he differed in that he possessed and was governed by a soul. On this view man was able to think, control his activities and possess consciousness. Descartes even went so far as to point to an anatomic seat for the soul: the pineal gland, which is situated centrally in the brain. He had a number of reasons for choosing this location, including the apparent absence of the pineal gland in animals: its centrality, the fact that it is not doubled (like the cortex, cerebellum, etc.), and his belief based on observation that it is the centre of a network of nerve endings which act upon it and which it in turn acts upon: ‘... the machine of the body is so formed that from the simple fact that this gland is diversely moved by the soul, or by such other cause, whatever it is, it thrusts the spirits which surround it towards the pores of the brain, which conduct them by the nerves into the muscles, by which means it causes them to move the limbs.’¹ Descartes also argued that the pineal gland exerted its influence not by altering the amount of motion of the ‘spirits’ in the nerves but by altering their direction.

The argument was spoiled not long after by the discovery by the Danish scientist Niel Stenson of the pineal gland in animals.² More significantly it appeared that Descartes’ notion of the activity of the soul contradicted Newton’s Third Law of Motion and the related Law of Conservation of Momentum. The effect of these laws is that the total quantity of motion in the universe in any given direction is constant, as well as its total amount being constant.

In the next century La Mettrie, the French philosopher, took Descartes’ argument to its logical conclusion: if animals are machines then men are machines too. In his book L’Homme Machine (1748) he argued that ‘... man is but an animal, or a collection of springs which wind each other up. . . . If these springs differ among themselves, these differences consist only in their position and in their degrees of strength, and never in their nature; wherefore the soul is but a principle of motion or a material and sensible part of the brain, which can be regarded, without fear of error, as the mainspring of the whole machine, having a visible influence on all the parts.’³ This materialist view is perhaps a direct antecedent of modern ‘behaviourism’ with its emphasis on what the organism does and its (at least methodological) disregard for consciousness. La Mettrie used the term ‘machine’ quite broadly and gave consideration to chemical as well as purely mechanical processes in the body.

Not long afterwards another French philosopher, Abbé de Condillac,¹ took up the man-machine idea by considering a statue which is, in fantasy, given one after another all the senses that man possesses. Condillac used this fiction, which would probably have been inconceivable at an earlier date, to consider the way in which an entity can come to have knowledge of its own body and of objects outside its body, and what the limitations of knowledge are. In a sense he is a forerunner of cybernetics (see Chapter 2) since he examines the logical properties of ‘systems’ without regard to whether they are ‘living’ or not.

Although the machine model of man is not paraded so obviously in the nineteenth century² the human organism, including the brain, is already being treated scientifically by physiologists as a machine. It has been pointed out³ that the major emphasis of this physiological study during the nineteenth and well into the twentieth century is on questions of energy utilisation, the organism being regarded as an engine for converting food into more living matter or into activity. The activity of the nervous system is also seen mainly in this context; that is, the body is studied in terms of physics and chemistry, but not in terms of its logical organisation or its information utilisation.

Early in the present century another way of looking at the brain became popular in the newly emerging discipline of scientific psychology: this was the idea that it acts rather like a self-operating telephone switchboard connecting up sensory inputs with motor outputs in different patterns. In this view, learning involves connecting inputs and outputs in new combinations (or in some sense strengthening connections already in existence) and forgetting involves disconnecting inputs and outputs that have previously been connected (or in some sense weakening the connection). This analogy between the brain and a telephone switchboard was stated explicitly by writers such as Pearson⁴ and was at least implicit in the ideas of major figures such as Sherrington in England, Thorndike in America and Pavlov in Russia. This theory had its roots in the associationist philosophy so beloved by generations of British philosophers and is still a force in modern psychology.

Finally, in the progress of the man-machine analogy to date, the development of the modern computer has led to the conception of the nervous system as a type of general-purpose computer.

It could be said, cynically, that our attempt to understand the brain is based at any time on the characteristics of the most advanced machines in use at that time. Descartes, La Mettrie and other thinkers of the seventeenth and eighteenth centuries conceptualised the nervous system in terms of clocks, music boxes, water toys, and automata of the period like dancing ballerinas. Scientists of the nineteenth century thought in terms of steam engines, and early in the twentieth century in terms of the telephone networks then coming into existence. It could be argued that it is only to be expected that the current excitement about computers should lead to a fashion for comparing the brain to a computer, the implication being that it will be overtaken in due course by another more ‘modern’ and up-to-date analogy. The reader will be in a better position to judge for himself whether or not this analogy is something more fundamental than a temporary fashion after reading the next section.

THE COMPUTER-BRAIN ANALOGY

Let us, then, look more closely at the analogy which is now so often drawn between the brain and a general-purpose digital computer.¹ The two kinds of system may be said to resemble each other in the following ways:

(1) Both are general-purpose devices. That is to say, they are both capable of dealing with a wide variety of problems whose exact nature cannot be anticipated, but they can temporarily assume special-purpose forms when they are dealing with particular problems. Unlike the brains of most other animals which are specialised to carry out certain particular functions, the human brain is highly flexible, especially through its ability to abstract and use language. One might say that the brain and the general-purpose computer are the two most general-purpose devices known to man. Of these two, the human brain is at present effectively more general-purpose than the computer. This book is concerned to a large extent with the question: ‘How do we release the full general-purpose potential of the computer?’

(2) Both are information processing devices. That is, at the most general level they may both be described as receiving environmental information through input devices, manipulating this information in accordance with information already in the system (in computer terms the sequence of instructions comprising the program together with other information held in store) and returning information to the environment through output devices. For this reason computer models of particular brain processes are often called ‘information processing models’. Interestingly enough, the information processed in both cases appears to be coded in binary form. In one case it is represented in terms of two-state electronic components and in the other in terms of neurons (the cells of the nervous system) which either fire (relay a pulse) or do not fire at a given moment (although recent neurophysiological research has shown that the situation may after all be more complex than this).

(3) Both may incorporate models within themselves. As Kenneth Craik,¹ who anticipated many of the ideas of cybernetics, pointed out, one of the fundamental functions of the brain may be to build up and use models of its environment. Among contemporary scientists, J. Z. Young² in particular has made much use of the idea in his work on the octopus brain. The interesting situation arises that not only can the computer model the brain, the brain can also model the computer.

(4) Both achieve their intellectual excellence by carrying out large numbers of primitive operations. As Wooldridge has said: ‘. . . what is really fundamental about the general types of machines that have been too narrowly described as “electronic digital computers” is that they get their amazing results by the performance of a very large number of very simple processing steps. This would also appear to be a valid description of the essence of brain function.’¹ In both cases, too, the primitive operations are carried out at high speed.

At the same time, there are a number of obvious differences between the brain and the present generation of digital computers. Despite recent advances, the size of components in the computer is still much greater than in the brain, and there are fewer of them, but the speed with which they function is far greater. Also, the brain is, unlike the digital computer, not entirely digital in its operation; thus even if the neurons are digital (discret...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Table of Contents
• Acknowledgments
• What this book is about
• 1 BRAINS, COMPUTERS AND MODELS
• 2 CYBERNETICS
• 3 DIGITAL COMPUTING
• 4 LEARNING
• 5 PROBLEM SOLVING
• 6 PATTERN: RECOGNITION AND CREATION
• 7 LANGUAGE
• 8 PERSONALITY
• 9 CONSCIOUSNESS
• Further reading
• Index