The behavioural basis of culturalism – a general systems approach

10. The Principle of a Model

An essential part of a control mechanism is a model representing the system’s relationship with its environment. Every system must be endowed with such a model or its behaviour could not display any order. Take the field of genetics. Protein synthesis can only be explained in terms of a control mechanism exploiting such a model. Horowitz writes,

“It seems evident that the synthesis of an enzyme – a giant protein molecule consisting of hundreds of amino acid units arranged end to end in a specific and unique order – requires a model or set of instructions of some kind. These instructions must be characteristic of the species; they must be automatically transmitted from generation to generation, and they must be constant yet capable of evolutionary change. The only known entity that could perform such a function is the gene. There are many reasons for believing that it transmits information by acting as a model or template.” [12]

Kenneth Craik was undoubtedly the first person to provide a similar explanation for day-to-day behaviour. [13] He viewed the nervous system “as a calculating machine capable of modelling or paralleling external events”, and he considered this to be ‘the basic feature of thought and of explanation’.

The gene and the brain, however dissimilar they may appear to the outside observer, are functionally the same. They are both ‘cybernisms’, or control mechanisms, as is a gene-pool or that organisation of brains in which is organised the ‘worldview’ of a society.

11. The Coordination of Control Principle

Each of the constraints a system is subjected to are the constraints of the system as a whole and of the supra-system of which it is part. Only in this way can its response provide that compromise between competing environmental requirements which ensures its stable relationship with the environment (see 12. The Compromise Response Principle). Only in this way can it act as a unit of behaviour, i.e. as a self-regulating system.

This means that a behaviour pattern must be mediated by an integrated pattern of instructions alone provided by a cybernism functioning as part of a natural system. This means that with regard to a society’s behaviour pattern, if this is to constitute an integrated pattern of behaviour it must be mediated by a society’s cultural pattern, and not by any external or asystemic agency.

It is significant that the behaviour of human societies has been until very recently entirely self-regulating. Primitive societies had no dictators or bureaucracies. They were run by public ‘opinion’ which reflected their traditional law, ‘demouphemos’ as the Greeks called it in Homeric times, which only later was institutionalised into ‘demoukratos’, the latter without the former being but a façade or an empty shell.

12. The Compromise Response Principle

There must be an optimum value to every variable representing a strategy exploited by a system to achieve stability. No such strategy is desirable per se. Therefore there is no possible reason for maximising any one of them. On account of the principle of economy, the optimum will, in fact, coincide with the minimum required for dealing with environmental challenges. Only a self-regulating system which is linked by feedback-loops to all those parts of its environment whose behaviour can affect it is capable of maintaining itself along the optimum course that provides the best compromise between environmental requirements.

Consider the evolutionary development of a pig. It is designed to satisfy all supra-systemic requirements. The pig must be able to smell, see, dig for truffles, run away from predators, etc. The pigs bred by man, i.e. asystemically controlled pigs, are simply designed to provide the maximum amount of meat. The breeder does not care whether they can see, smell, dig for truffles, or run away from predators. If he could dispense with the legs altogether and put them on wheels, he would probably do so in the interest of ‘greater efficiency’. Such an animal is, needless to say, increasingly unadaptive and would have little chance of survival in its natural environment.

13. The Asystemic Control Principle

When a system disintegrates, it loses its capacity for self-regulation. In such conditions, its maintenance can only be ensured by the application of some sort of external control.

Thus a doctor administers drugs to a body incapable of controlling itself; a surgeon intervenes when its basic structure has so far diverged from the optimum as to be incapable of self-readjustment. Artificial fertilizers may have to be added to the soil when the self-controlling mechanisms, which in a balanced ecosystem ensure nitrogen fixation, are no longer operative. Insecticides may have to be sprayed over crops when natural biological controls cease to be effective.

At a social level, ‘welfare’, as unknown as it is unnecessary in a self-controlling society, becomes indispensable once the society has so disintegrated that people are no longer capable of looking after themselves. In the same way, coercive government becomes necessary when the cultural mechanisms, designed to ensure social control, have broken down. The choice in such societies is not between ‘democracy’ and dictatorship, but between chaos, thinly disguised as democracy, and dictatorship. As Disraeli wrote, “A country can either be governed by tradition or by force.”

The trouble with all forms of asystemic control is that:

They are grossly simplified – display minimum complexity (see 30. The Cybernismic Complexity Principle).
They are inefficient and over-centralised – (defying 4. The Law of Economy) and hence, are very expensive logistically.
Systems can no longer be kept on their optimum course through feedback interaction with their environment – their goal being determined arbitrarily by the external controller. This means that they must become unstable.
They will not constitute an integrated pattern of responses.
They will create further disequilibria increasing the need for further asystemic control (see 59. The Problem Multiplier Principle), thereby involving the system in a positive feedback course towards ever further instability.

In each case therefore, to maintain stability, asystemic controls must be reduced to the minimum, i.e. applied to special cases only – deviants in a society for instance; and their object should be simply to re-establish the proper functioning of the original (systemic) self-regulating mechanisms.

14. The Representation Principle

The relationship of a model to the situation it represents is one of representation for specific purposes – those of the system’s behaviour vis-a-vis its environment. It is usually considered that a model represents the environment. This is false. It must, in fact represent the relationship between the system and its environment.

Why this is so is clear from the consideration that if the model is to provide that information required to ensure the increased stability of the system in a given environment, it must represent both the former and the latter since changes in either one or the other will result in different behavioural requirements. The model will therefore be of the supra-system seen from the point of view of the system.

15. The Abstraction Principle

The purpose of a model is to permit a particular behaviour pattern. In accordance with the law of economy, the variety and complexity of the model and hence the number of variables used will be the minimum necessary for this purpose. The variables are abstractions, or specific aspects of the environment, chosen for their relevance to the model.

This must be true of the units of all models, objective as well as subjective. Men, dogs, houses, telephones are all abstractions. So is an atom, since the concept is reached by abstracting from an infinite number of atoms (all of which will differ in some respects) those that are relevant to our model, and ignoring the vast number of other aspects whose information value (vis-a-vis our model) does not justify their being taken into account. The term ‘atom’ is a very simple model of the infinity of slightly different ‘things’ it purports to represent – one that displays that degree of accuracy that is sufficient for present-day scientific purposes.

‘John Smith’ is also an abstraction of a particular ontogenetic process in which temporal differences, i.e. the differences between John Smith at the age of two and at the age of 46, are, for particular purposes, judged irrelevant and hence filtered out.

The world is made up of systems that to varying degrees are open. The number of factors that can influence a given situation are thus infinite and only a minute fraction will be represented by a model. This must impose a limitation on the accuracy of systemic interpretation and hence on that of the responses that they will give rise to.

If, in open defiance of the law of economy, it were decided to construct a model permitting interpretation and prediction with 100 percent accuracy, it would have to be an exact to-scale replica of the world and its surroundings, correct to the minutest detail. Short of attempting to undertake such a chimeric project, we must be satisfied with statistically probable predictions based on models that are simplifications of the environment they are designed to represent.

16. The Principle of Integrated Representation

Phylogenetically, one of the main features of the development of the nervous system has been its gradual centralisation. The nerve-net of the more primitive metazoa slowly gives rise to the central nervous system whose functioning then becomes more and more dependent on the brain. This process is known as encephalisation.

What is the philosophy behind this change? The answer is that, in the face of environmental disorder, the organism, if it is to survive, has to behave more and more as a unit, i.e. the number of those cases in which the separate parts can function on their own is correspondingly reduced.

This implies that, whereas previously a great many environmental changes were only relevant locally, i.e. to the behaviour of the parts, they are now taken to be relevant to the behaviour of the organism as a whole.

Let us take the case of the nervous system of the octopus. This strange animal has a large ganglion, or rudimentary, brain in each of its tentacles, as well as a smaller one in its head. [14] Such an organisation of the neurons is valid only insofar as the information derived by each ‘brain’ can be considered of so specialised a nature as to be relevant but to the corresponding tentacle, and of no consequence to the rest of the organism.

Behaviour, as we go up the ladder of life, can more and more be regarded as the effect of the whole environment on the whole organism. Significant in this respect is the fact that the centralised cybernism uses the same classificatory system for all parts of the environment. Thus there is no division of the information within a model in accordance with the different detecting devices used.

It is only by regarding the whole of our environment as classified in terms of the same neuro-cybernism, whose role it is to determine the optimum pattern to our environment, that it is possible to understand human behaviour.

Several illustrations can be given. Motivation Research in industry has cast considerable light on this subject. It is now current to talk of ‘the image’ of anything one is trying to sell, whether it be toothpaste or the Democratic candidate in the American presidential election. By ‘image’ one in fact means its status as a stimulus for our various behavioural tendencies, i.e. it will tell us what ‘instincts’ the product is appealing to and thus what should be our method of selling it.

Tinbergen shows how people will react in different ways to dogs with different head shapes. [15] A dog with a rather high forehead and small snout automatically stimulates maternal behaviour. A large powerful dog will undoubtedly appeal to totally different instincts. Undoubtedly, the sort of people who have a poodle as a pet are very different from those who have an alsatian, as the possession of these two types of dog satisfies totally different psychological requirements.

Similarly, behaviour towards political leaders can only be understood if the latter are interpretable as possible fathers, husbands, lovers, sons, grandsons; or, for men, rivals, brothers, etc.

Cut and dry stimuli, such as an excellent meal presented to a hungry man, will clearly affect the behavioural centres determining eating so strongly that the effect of this stimulus on other centres may be almost negligible, but never entirely so. More complex stimuli having lower information value, such as a violin concerto, will, on the other hand, affect a variety of centres and determine a more composite response.

If science is to be of any use whatsoever in guiding public policy, its division into specialised disciplines using distinct and unrelated classifications must be slowly abandoned. It is very unlikely, however, that this will happen as it is difficult to see the different branches of science being satisfactorily controlled (externally) from the centre in such a way that they provide a coordinated and integrated representation of the biosphere. That is why the only means of controlling a society’s relationship with its environment is by means of its culture which, like all undisturbed cybernisms in the natural world, must provide an integrated representation of it.

17. The Postulation of a Model Principle

Data are detected because they appear relevant to a model. This means that a model has already been built before detection occurs. Indeed, as empiricists maintain, the notion that the ‘mind’, by which I presume is meant the ‘cybernism’, is at birth a ‘tabula rasa’, is inconsistent with the nature of its role as an integral part of a control mechanism.

A rudimentary model of the system, reflecting the experience of the species as a whole must be inherited by each generation. This model will be postulated at birth as a representation of the supra-system, end data will be detected and cybernised to the extent that it might serve towards confirming or invalidating it.

The cybernisation of data, and the mediation of an appropriate response, will then serve to modify the model so that it will furnish an ever-more accurate representation of the system.

There is no reason to suppose that the collation of scientific data should occur in any other way. It may be objected that this is contrary to the very principle of experimentation that is considered to be at the basis of modern scientific method. This is not so. The role of experimentation is precisely to determine as objectively as possible the validity of a previously established hypothesis. What it is contrary to, however, is blind experimentation which occurs in financially over-endowed laboratories, and has no counterpart in the ‘natural’ world. If this is true, then scientific theories, rather than being reached ‘inductively’ in accordance with empiricist theory, must be regarded as postulated.

In this way, Le Verrier postulated, by purely mathematical means, the then unknown planet Neptune as an explanation of certain otherwise inexplicable disturbances of the other planets. [16] Later when the German astronomer Gelle directed his telescope to the spot in the night sky that had been figured out by Le Verrier, he saw there a tiny speck that changed its position slightly from night to night and the planet Neptune was discovered (1846). Dirac postulated the positron as the most elegant way of explaining certain atomic phenomena inexplicable in terms of existing variables.

Leucippus and his disciple Democritus postulated the atom, and Bohr postulated the modern version of this ancient hypothesis. Watson and Crick proceeded in the same manner when developing the genetic code, as is revealed in the latter’s book, The Double Helix. These discoveries are well-known. There is a quite unjustified tendency, however, to regard them as scientific curiosities – and as exceptions to the general rule that science develops inductively by the meticulous examination of impartially accumulated data, in accordance with the empiricist thesis.

One of the consequences of adopting this model of behaviour is that the notion of learning by random trial-and-error becomes untenable. If a rat is put into a maze, we know that it can be taught to find its way out. However, before hitting upon the correct route, it will have to make a series of unsuccessful trials. Now, if these trials were arranged to form a series, would it be possible to put order into this series, or must each trial be considered as random?

The trial-and-error theory appears to assume the latter hypothesis. Strictly speaking, random behaviour does not occur in a self-regulating system whether it be an organism, a society, or an ecosystem. These all display varying degrees of order and hence of limitation of choice, or of non-randomness.

Looked at slightly differently, if a rat finds its way through a maze as a result of a series of trials, what will determine its first trial? If a hundred possible moves are open to it why should it make one rather than any other?

All the actions of an animal, whether it be an earthworm, a rat, or a man, as Craik was the first to show, must be regarded as based on hypotheses as to the nature of their environment. Thus, supposing that the first move made by the rat, rather than lead it towards the opening, on the contrary led it into a further cul-de-sac, the model responsible for this error would have to be modified.

Thus, when the rat made its next move, the situation would be interpreted in the light of a new model – one that had taken into account the failure of its predecessor to interpret correctly the short-term systemic situation in which the rat had found itself, and one furnishing an ever more precise representation of the system for the purpose of finding its way out of the maze.

In other words, each action would be regarded as a correction of an error, and if these were taken as forming a series, then the latter would be ‘damped’ in the sense that the errors would be progressively reduced, i.e. the sub-system would be tending towards ever-higher stability. The rate at which this process would be occurring would depend on the system’s ability to postulate and monitor hypotheses leading to adaptive behaviour.

18. The Probability Principle

If a bridge player is to maximise his chances of winning, he must be able to calculate the most probable distribution of the cards yet to be played among the three other players. A little reflection will reveal that this must apply to all behavioural processes. In each case, the hypothesis that is postulated and on which a behavioural response is based must be the one that has the highest probability. The fact that this is not evident at first sight, as it is in a game of chance, is because what constitutes the most probable hypothesis to explain a given situation will be different in each case since the information in terms of which probabilities must be calculated will, like Heraclitus’s river, be modified with each experience.

Are we not over-estimating the capacity of biological organisms in suggesting that they are capable of making such precise calculations? I do not think so. If relatively simple forms of life are capable of complex behavioural feats executed with the most extraordinary precision, then those must be based on equally precise calculations.

Consider the little fish (Gymnarchus niloticus) that lives in the Nile. [17] It is capable of darting in muddy water after the small fish on which it feeds, and never bumps into anything, in spite of the fact that its eyes are quite degenerate and only sensitive to extremely bright light. Lissman, who spent 12 years experimenting with this fish, found that it owed its capacity for finding its way around so skilfully to its ability to discriminate between minute differences in the conductivity of the objects in its immediate environment.

This skill is so developed that the Gymnarchus can tell the difference between mixtures of different proportions of tap water and distilled water entirely on the basis of their different conductivity. If salts or acids are added to the distilled water so that its electrical conductivity matches that of the tap water, it can no longer discriminate between them. Here again a complicated calculation must be made.

Similarly, Noel-Martin noted the extraordinary mathematical ability of bees:

“honeycombs are built according to maximum efficiency principles. Being hexagonal, the cells make use of available space in the most economic and symmetrical way possible, and the angle between adjoining cells is such that the smallest possible amount of wax is required for their construction.” [18]

It may be thought that these examples are simply curiosities of nature. However, if our thesis be correct, they are but striking examples of a principle in terms of which we must explain ‘perception’ and ‘thought’ at all levels of organisation. In each case it provides the most probable hypothesis in terms of the model of the system’s relationship with its environment to explain a given situation.

Thus, when I look out of my window and see a tree, a road, and people walking about, I am in fact formulating that hypothesis as to the nature of the environmental data isolated by my detecting mechanisms that has the highest probability in the light of my model of the environment. The same is true when I identify one of the passers-by as John Smith, and also when I assume that he is going home for dinner. And so it is when I guess that his dinner will consist of shepherd’s pie and bananas and custard.

As we proceed, these hypotheses become increasingly improbable in the sense that they constitute one possibility out of an ever-increasing number of different possibilities. In terms of information theory, they have higher information value. In each case, however, the possibility selected is the most probable one in terms of my model. In objective terms, the probability of the hypotheses being correct depends on the extent to which my model represents the situation (see 20. The Continuity of Information Principle).