There is also a less obvious competition between holons within the organism in times of stress, when the exposed or traumarised parts tend to assert themselves to the detriment of the whole. The pathology of hierarchic disorder will be discussed in Part Three.

  Under normal conditions, however, when the organism or body social is functioning steadily, the integrative and self-assertive tendencies are in a state of dynamic equilibrium -- symbolised by Janus Patulcius, the 'Opener', with a key in his left hand, and Janus Clusius, the 'Closer', jealous guardian of the gate, with a staff in his right.

  To sum up, stable inorganic systems, from atoms to galaxies, display hierarchic order; the atom itself, formerly thought of as an indivisible unit, is a holon, and the rules which govern the interactions of sub-nuclear particles are not the same rules that govern the interactions between atoms as wholes.

  The living organism is not a mosaic aggregate of elementary physico-chemical processes, but a hierarchy of parts within parts, in which each holon, from the sub-cellular organelles upward, is a closely integrated structure, equipped with self-regulatory devices, and enjoys a degree of self-government. Transplant surgery and experimental embryology provide striking illustrations for the autonomy of organismic holons.

  The integrative powers of life are manifested in the phenomena of symbiosis between organelles, in the varied forms of partnership within the same species or between different species; in the phenomena of regeneration, in lower species, of complete individuals from their fragments; in the re-formation of scrambled embryonic organs, etc. The self-assertive tendency is equally ubiquitous in the competitive struggle for life.

  V

  TRIGGERS AND FILTERS

  All the time the Guard was looking at her, first through a telescope, then through a microscope, and then through an opera-glass. At last he said, 'You're travelling the wrong way . . .' Through the Looking-Glass

  Triggers

  You turn a switch or push a button on a machine, and this simple, effortless gesture releases the coordinated action of hundreds of wheels, pistons, levers, vacuum tubes or what have you. Such trigger mechanisms, where a relatively simple command or signal releases extremely complex, pre-set action-patterns, are a favourite device in biological and social organisation. By this means the organism (or body social) is able to reap the full benefits of the autonomous, self-regulating character of its subdivisions -- its holons -- on lower levels. When the Cabinet decides to raise the Bank Rate from six per cent to seven per cent, or to send troops to a trouble-spot in the East, the decision is worded in brief, laconic terms, which merely imply, but do not specify, the intricate sequence of actions that will follow. The decision triggers various department heads and experts into activity; these will provide the first set of more specific instructions, and so on, down along the branching hierarchy to the terminal units -- bank clerks or paratroopers. At each step on its downward journey, the signal releases pre-set action-patterns which transform the implicit message into explicit terms, from the general into the particular. We have seen analogous processes at work in the production of articulate speech: the non-verbal, inarticulate intent of conveying a message triggers off the phrase-structuring mechanisms, which in turn bring the rules of syntax into play, and so on, down to the spelling out of the individual phonemes.

  In the performance of manual skills we follow the same procedure: my conscious ego, at the apex of the hierarchy, gives out the laconic order: 'Light cigarette', and leaves it to the lower echelons in my nervous system to fill in the details by sending out a pattern of impulses, which activate sub-centres, which control the contractions of single muscles. This spelling-out process, from intention to execution, is rather like operating a series of combination locks, on different levels, in descending order. Every holon in the motor hierarchy has like a government department -- its rule-governed patterns for co-ordinating the motions of limbs, joints, muscles, according to the level which it occupies in the hierarchy; thus the command 'Light cigarette' does not have to specify what each of my finger muscles is supposed to do to strike a match. It merely has to trigger the appropriate centres into action, which will spell out the implicitly 'coded' command in explicit terms by activating their own sub-units in the appropriate strategic order, guided by local feedbacks. Generally speaking, a holon on the n level of the hierarchy is represented on the n + 1 level as a unit and triggered off as a unit.*

  * Or, to put it differently: the holon is a system of relations which is represented on the next higher level as a unit, i.e., a relatum.

  Like all our previous generalisations, this, too, is meant to apply to all types of hierarchies -- including, for instance, the hierarchic sequence of embryonic development. This starts with a rather remarkable kind of trigger action: pricking the unfertilised egg of a virgin frog with a fine platinum needle is sufficient to initiate the growth of that egg into a normal adult frog. It has been shown that even in higher mammals like rabbits and sheep, simple mechanical or chemical stimuli can produce the same effect. Sexual reproduction is indispensable for creating variety; for mere propagation a simple trigger releaser will do.

  The trigger is, of course, normally a sperm. The genetic code of the fertilised egg is said to contain the 'blueprint' of the future adult, but it would be more correct to say that it embodies a set of rules or instructions for manufacturing it. The rules are laid down in a chemical code, which comprises four letters: A, G, C and T (the initials stand for chemical substances whose long names are irrelevant to our purpose). The 'words' which these letters form on the long spirals of chromosomes in the cell nucleus contain the instructions which the cell has to follow. One of the main tasks of an embryonic cell is the manufacture of proteins required for growth. There are thousands of different proteins, but they are all made of the same building blocks: twenty different kinds of amino-acids, put together in different combinations; and each amino-acid corresponds to a 'word' of three letters in the genetic code. Thus the instructions of the implicit four-letter alphabet are 'spelled out' in the twenty-letter alphabet of amino-acids, which provides all the necessary combinations for the thousands of proteins which make an organism.

  The differentiation of structures and their shaping into form in the growing embryo is a stepwise affair which has been compared to the way a sculptor carves a statue out of a piece of wood -- but also to the child's acquisition of articulate and coherent speech. At each successive step, from the fertilised egg to the finished product, the overall instructions contained in the four-letter alphabet of the genetic code are first roughed in, then sketched in, and finally spelt out in elaborate detail; and each step is initiated by biochemical triggers (enzymes, inducers, hormones, and other catalysts).

  How to Build a Nest

  I shall have more to say about hierarchic order in embryonic development in Chapter IX; for the moment let us turn to the instinctive activities of the adult animal.* The growing organism is governed by its genetic code: in the adult organism a different type of code takes over, located in the nervous system. It incorporates the fixed 'rules of the game' which control the stereotyped rituals of courting, mating, duelling, and the much more flexible skills of building nests, hives or webs. Each of these skills can again be hierarchically 'dissected' into subskills, that is, functional holons, down to the level of 'fixed action-patterns' -- to use Konrad Lorenz' term. In all these activities the trigger principle plays a dominant and conspicuous role. The triggers are certain stimulus patterns in the environment -- sights, smells, sounds, which the ethologist calls 'releasers' or 'sign-releasers'. Thus, for instance, the nuptial colours of the stickleback (a freshwater fish) are blue eyes and a red under-belly; and any object, regardless of its shape, that is red underneath, when brought near the territory of a male stickleback will act as a releaser for attack. The stickleback has five different methods of threatening and attacking, each triggered by a slightly different releaser. Similarly, animal species which engage in ritual tournaments -- w
here the adversary acknowledging defeat is spared -- have each a limited repertory of fighting moves, rather like the lunges, thrusts and ripostes of fencers.

  * Most activities which we call 'instinctive' are in fact partly acquired, or modified, by early learning.

  W.H. Thorpe has made a detailed analysis of the functional holons which enter into the nest-building activity of the long-tailed tit. He enumerated fourteen different action-patterns (such as 'searching' and 'collecting' building materials; 'weaving', 'pressing', 'trampling', 'lining', etc.), each of them consisting of simpler patterns, and triggered by at least eighteen different releasers. Instead of endlessly watching rats endlessly pressing the bar in the Skinner box, students of psychology would be well advised to study Thorpe's description, of which the following is a much abbreviated version.

  The tit uses four different building materials: moss, spider's silk, lichens and feathers, each of which has a different function and requires a different kind of skilled manipulation. The activity starts with the search for a convenient site, a branch which forks in the right way. When the site is found, moss is collected and placed on the fork. Most of it falls off, but the bird persists until a few pieces have stuck. When this stage is reached, the bird switches from collecting moss to collecting spider's silk, which is rubbed on the moss until it sticks, then stretched and used for binding. These activities continue until a platform has taken shape. Now the bird switches back to moss and starts constructing the cup around it, fast by 'sidewise weaving', later by 'vertical weaving' in a sitting position, steadily rotating its body as the curved rim of the cup begins to take shape. At this stage, new action-patterns make their appearance: 'breast-pressing' and 'trampling' with the feet. When the cup is about one-third complete, the bird starts collecting the third building material, lichens. These are used to cover the outside only of the nest, 'by stretching out over the rim from inside the nest and by hanging on the outside in various more or less acrobatic attitudes'. When the cup is about two-thirds completed, the building routine is changed in such a way as to leave a neat entrance-hole at the most convenient point of approach. The wall around the hole is strengthened, the dome of the nest completed, and now the furnishing can begin, using the fourth building material, feathers. Thorpe comments:

  So much for simplicity! But perhaps the most significant point of all is the evidence provided that the bird must have some 'conception' of what the completed nest should look like, and some sort of'conception' that the addition of a piece of moss or lichen here and here will be a step towards the 'ideal' pattern, and that other pieces there and there would detract from it. . . . Its actions are directional and it 'knows when to stop' . . . . [1]

  By comparing this description with Watson's description of how Patou makes a gown ('Has he a picture in his mind? He has not'), or with Skinner's method of conditioning pigeons, one gets an idea of the contrast between the flat-earth view of Behaviourism and living reality. Where, for instance, is the indispensable 'reinforcement' -- the stick and the carrot which, according to the Behaviourist, would be required at each step to make the bird persist in activities that include thirteen different types of construction jobs? And yet the tit persists, without any reward, until it has finished the nest. And how could it be maintained that the tit is 'controlled by the contingencies of the environment' when it has to search the environment, now for moss, now for spider's silk, now for lichen and feathers; yet, however varied the 'contingencies of environment', it succeeds in building the same kind of nest? Or, take as another example, the common spider, who will suspend its web from three, four or more points of attachment, according to the lie of the land, but will always arrive at the same familiar symmetrical pattern, where the radial threads bisect the laterals at equal angles, according to the fixed canon of rules which controls its activities. How to apply these rules to a particular environment -- whether to make a pentagonal or hexagonal web is a matter of flexible strategy.

  All instinctive activities consist of hierarchies of sub-skills -- in the spider's case the judging of angles and weaving of the thread -- controlled by fixed rules and guided by adaptable strategies. It is this dual characteristic which justifies us in calling a sub-skill a 'functional holon'. As such, it also has the various other characteristics of holons previously discussed. A skill can be exercised in the service of some larger activity and as part of it; but virtually any skill can also become a habit which brooks no interference and may be pursued for its own sake. In the first case, the functional holon serves the integration of behaviour; in the second case, it can display very marked self-assertive tendencies -- the proverbial 'stubbornness of habits'. Whatever clever 'strategies' you use to disguise your handwriting, you cannot fool the expert and get away with it in court. The same goes for your gait, accent of speech, the use of favourite turns of phrase. Habits are behavioural holons, governed by rules which mostly operate unconsciously. Taken together, they constitute what we call personality or style. But each holon also has a margin of strategic choices, and that margin of choice increases in ascending order with the increasing complexity of higher levels. And if we ask what determines the conscious choices at the apex, we again find ourselves in a regressing series.

  Filters

  So far we have been concerned with 'output': the spelling out of intent into action, including the 'intent' of the fertilised egg to grow into an adult, and of a fertile idea to grow into articulate language. Before we turn to the 'input' side -- sensations and perceptions -- it might be useful to revert for a moment to the analogy of a military operation in old-fashioned, classical warfare.

  The General in Command issues an order which contains the plan of action in broad outlines; this is transmitted from Divisional Headquarters to Brigade Headquarters to Battalion Headquarters, and so on; at each successive echelon in the hierarchy the plan is more elaborated until the last detail is filled in. The reverse process takes place in collecting information about the movements of the enemy and the lie of the land. The data are collected on the lowest, local levels by patrols reconnoitring the terrain. They are then stripped of irrelevant detail, condensed, filtered and combined with data from other sources at each higher echelon, as the stream of information flows upward along converging branches of the hierarchy. Here we have a very simplified model of the working of the sensory-motor nervous system.

  On the motor side, we had a series of 'triggers'. On the perceptual side we have instead a series of 'filters' or 'scanners', through which the vital input traffic must pass on its ascent from sense-organ to cerebral cortex. Their function is to analyse, de-code, classify and abstract the information that the stream carries, until the chaotic multitude of sensations, which constantly bombard the senses, is transformed into meaningful messages.

  Of most of these input-processing activities we are blissfully unaware. They are performed by a whole hierarchy of processing agencies built into the apparatus of perception. On the lowest level, there is the screening, or filtering out, of sensations that are irrelevant to the activity in hand or the mood of the moment. One is normally not aware of the pressure of the chair against one's backside, nor of the contact between skin and clothing. The eye and the ear are also equipped with such selective screening contrivances ('lateral inhibition', 'habituation', etc.).

  The next stage in processing is very striking -- once one starts thinking about it. If you hold the index finger of the right hand ten inches, the same finger of the left hand twenty inches, in front of your eyes, you see them as being of equal size, although the image on the retina of one is twice as large as the other. People moving about in a room do not seem to shrink or grow in size -- as they should -- because we know that their size remains constant, and this knowledge somehow interferes with the visual input at some level of the nervous system, and falsities it in the noble cause of making it conform to reality. The photographic lens has no such built-in mechanism; it will honestly show the left index finger twice as large as the right, and a s
unbathing girl's foot stretched out towards the camera as a case of elephantiasis. 'Even our elementary perceptions', wrote Bartlett, 'are inferential constructs' [2]; but the inferential process functions on unconscious levels of the hierarchy.

  The tendency to see a familiar object as of its actual size, regardless of distance, is called by psychologists the 'size constancy phenomenon'. Not only the size but also the colour and shape of the retinal image of a moving object is all the time changing with its distance, illumination and angle of vision; yet we are mostly unaware of these changes. Accordingly, to the phenomenon of size constancy we have to add those of colour and shape constancy.

  The constancies are only a part of our repertory ofperceptual skills, which form the grammar of vision, and provide the 'rules of the game' that enable us to make sense out of the ever-changing mosaic of our sensations. Though they operate automatically and unconsciously, they can be modified by learning. When a subject in a psychological laboratory puts on inverting glasses which turn the world, including his own body, upside down, he is at first completely lost, unable to walk, and may also feel seasick. After a few days of constantly wearing the glasses he readjusts himself to living in a visually upside down world. The adjustment requires at first great conscious effort, but in the end the subject seems hardly aware that the world is upside down. The retinal image remains inverted, and so of course is its projection in the brain, but his mental image -- there is no other word for it -- is now the right way up; and when at this stage the glasses are taken off, it takes him some time to readjust to normality.*