The Act of Creation

by Arthur Koestler

  We shall see that Child's 'isolation' concept has a wide range of applicability. In all cases, isolation of the part from the whole leads to de-differentiation or other forms of regression; in some cases this is followed by re-differentiation and reintegration. Isolation leading to regression of an irreversible kind plays a considerable part in pathology, psycho-pathology, and social pathology.* On the other hand, regression followed by a progressive rebound releases creative potentials which are normally under restraint. Its magic can be observed on every level: from asexual reproduction to the repair of structural damage and functional disorder, and further up to psychotherapy, scientific discovery, and artistic creation. In the chapter which follows I shall briefly discuss the manifestations of 'super-flexibility -- of reculer pour mieux sauter -- on these various levels.

  NOTE

  To p. 453. Thus, for instance, Smithers in A Clinical Prospect of the Cancer Problem (1960) stresses the decisive influence which Child's "Physiological Foundations of Behaviour" had on the development of his ideas.

  IV

  RECULER POUR MIEUX SAUTER

  Structural Regenerations

  In primitive organisms such as the flatworm, and in the early embryonic stages of higher organisms, a physiologically isolated part tends in general 'to lose its characteristics as a part and to become or approach the condition of a new whole individual'. [1] Liberation of the part's previously restrained genetic potential 'involves a change in behaviour and structure from that of a part towards that of a whole organism'. [2] Such organisms could be said to live not only in dynamic equilibrium with their environment, but in a kind of 'regenerative equilibrium' which enables them to rise to virtually any challenge by means of these secondary adaptations.

  Some higher animals are still capable of regenerating lost organs or limbs. Let us have a closer look how it is done. When a salamander's leg has been amputated, the tissues near the wound surface de-differentiate and the cells acquire an embryonic appearance. This is the regressive or 'catabolic' phase. Around the fourth day begins the formation of the blastema -- the regeneration bud; and from here on throughout the 'anabolic' or synthetic phase the process follows closely the formation of limbs in normal embryonic development.* The blastema elongates into a cone and develops axially, the toes at its tip appearing first, and the rest of the limb gradually taking shape as it grows in length. When the organ is completed, central control is taken over by the nervous system, just as in the case of the embryo. The nervous system, however, also plays an indispensable part in initiating the first, catabolic phase. If no peripheral nerves are present in the amputation stump, regeneration does not occur.**

  The 'isolated part' in this case is the amputation stump. The blastema has in the beginning the multi-potential characteristics of the organ primordia, and its re-differentiation again proceeds stepwise: if the field is split in half, each half gives rise to a whole organ; if one half is removed, the remaining half will still develop into a complete limb.

  Although the isolated part, transformed into a new organ primordium (or its close equivalent), enjoys a high degree of independence and controls the formation of the new limb, its ties with the higher levels of the hierarchy are not completely severed. Its function in the whole has changed; its normal controls (through the nervous system and local chemical gradients) are out of action or even reversed; but the organism as a whole nevertheless assists the regenerating part by certain emergency measures -- a 'general alarm reaction' followed by a 'general adaptation syndrome', each stage indicated by metabolic changes and by the appearance of specific proteins and hormones in the circulatory system.

  Thus the isolation of the part is only temporary and relative; and when the process is completed, the regenerated limb assumes its normal function in the whole. The entire regressire-progressive sequence is the means by which the animal's 'regenerative equilibrium' enables it to adjust to traumatic experiences from the environment.

  Looked at from a different angle, one might say that the whole process is designed to prevent or correct malformations, i.e. faulty integrations. Without the initial nerve supply, the regressing amputation stump would be resorbed, the scar tissue would close over it, and the animal would achieve a modus vivendi as a cripple -- a faulty integration. On the other hand, a frog which will not normally regenerate a lost limb will do so if the nerve supply to the stump is artificially augmented, providing the initial stimulus to start the process. Traumatic challenges can only be met by the liberation of the organism's latent powers -- a temporary return to a more youthful or primitive condition.

  Reversed Gradients

  An important part in regeneration, as in morphogenesis, is played by axial gradients. The apical or 'head' end of the fertilized egg, the growing embryo, or the regenerating limb, are exposed to the highest degree of stimulation and show the highest rate of metabolism, protein synthesis, and RNA activity. Thus the anterior end becomes the dominant region, the 'head' in the literal and metaphorical sense, and exercises a restraining influence on the genetic potentials along the axial gradient -- so that activity is highest at the front and lowest at the tail end (higher organisms have of course a complex pattern of interacting gradients, some axial, some radial). But physical isolation by blockage or hyper-excitation (Child's third and fourth cause) of parts in previously subordinate positions can be shown to alter or reverse the gradient. In plants, where the dominant region is the growing tip of the shoot, pruning makes previously subordinate parts burst into activity. If in the marine polyp, tubularia, a piece of the stem is cut out, the frontal end of the fragment will normally regenerate the hydrant which is its 'head'; but if a ligature is applied to isolate the front end of the fragment, the gradient is reversed and the tail end, now the region of maximum excitation, becomes dominant and grows a head. Similarly, short pieces of planaria sometimes regenerate one head at the front and another at the tail end, if the metabolism of both cut surfaces is equally high.

  The Dangers of Regression

  Two more phenomena must be mentioned in this context: the first illustrates the flexibility, the second the vulnerability of regenerative processes.

  If the crystalline lens of a salamander-eye is removed, part of the iris de-differentiates, forms a vesicle, enters the cavity through the pupil, re-differentiates, and forms a normal lens -- whereas in embryonic development the lens is formed by the epidermis overlaying the eye-cup, without participation of the iris. Thus the morphogenetic skill of making a lens can make use of either of two different materials; the code is again invariant, the strategy adaptable.*

  On the other hand, the factors which, in a higher organism, determine whether a given trauma will lead to regenerative or pathological changes are of an extremely delicate nature. Thus Smithers [3] writes:

  The type of structure regenerated, or the kind of neoplasm formed, will depend on the level of the controlling field-gradient against which it is exerting itself, and the steepness of the gradient it can itself establish and promote as shown by its tendency towards undifferentiated cell-reproduction. The part which is physiologically isolated then produces an imperfect portion of a new whole, giving rise to whatever tissues it is capable of forming under the circumstances pertaining. This may result in malformations of all degrees, from simple overgrowth of adult tissues, through irregular mixtures of recognizable, well-differentiated cells, to the most rapidly growing, undifferentiated tumors.

  Pathogenic regulatory responses are reactions to stimulations which are 'outside the standardized range of normal experience of the species during its developmental peak period. They do not differ from the normal regulatory responses, however, in any fundamental particular. . . . Tissue overgrowth as a response to a long-continued external irritant, is of the same order as heat regulation, wound-healing or lactation. . . . Useless or harmful regulating mechanisms and tissue responses to isolation, injury, or stimulation are not fundamentally different in kind from those favourable ones which have become i
ncorporated into the inheritance of the species because they promoted survival through the period of reproductive activity. . . . The tissues most often called on for regeneration and repair, or most liable to recurrent stimulation into specialized activity, are those most prone to tumour formation.' [4]

  'Routine Regenerations'

  The last sentence that I have quoted leads into the borderland between regenerative and 'normal' processes: namely routine replacements. They range from the periodic moulting of feathers and shedding of the antlers, to the replacement of the whole human epidermis about once a month owing to wear and tear, and the replacement of red blood cells at the rate of 3 x 10^11 per day; not to mention the metabolic turnover on the molecular level which consumes about thirty per cent of our total protein intake. This type of routine (or so-called 'physiological') regeneration which goes on all the time is sometimes described as a constant 'renewal' or 'rejuvenation' of the body. It is often impossible to make a clear distinction between 'wear' and 'tear' -- for instance in minor abrasions of the skin. The differential factor is obviously the degree of stress, which, past a critical threshold, will bring general alarm reactions and 'adaptations of the second order' into play.

  Reorganizations of Function

  The transplanted salamander limb which functions normally in spite of its randomized nervous connections can be regarded as an example of both regeneration of structure and reorganization of function. The pathways leading into the limb all seem to be equipotential in their capacity as conductors of the excitation-clang. Without entering the old controversy about equipotentiality versus localization of functions in nervous tissues, it seems to be safe to say that in repetitive routines and local reflexes, equipotentiality has 'frozen up' into fixed local arrangements; whereas in case of injury to the pathways in question, the equipotentiality (or rather, multi-potentiality) of alternative 'canal-systems' is revived, and they take over the function of the injured system. To quote Lashley: 'The results indicate that when habitually used motor organs are rendered non-functional by removal or paralysis, there is an immediate, spontaneous use of other motor systems which had not previously been associated with, or used in, the performance of the activity.' [5] Nearly a century earlier Pflüger had shown that even the spinal reflexes of a frog are capable of 'crisis adaptations'. If a drop of acid is placed on the back of the left front limb of a decapitated frog, it will attempt to wipe it away with the left hind limb; but if prevented from doing so it will use the right hind limb -- which it normally never does in the exercise of the wiping reflex.

  Turning from the spinal level to the brain, Lashley's celebrated maze experiments have shown what astonishing regenerative adaptations the cerebral cortex is capable of. If a rat is trained to choose between two doors the one where a brighter light is shown, this habit (or at least part of it) must be localized in its optical cortex, for if this is extirpated, the habit is lost. But a rat with its optical cortex cut out can still be taught or re-taught the same skill. This means that some other cortical area has taken over the learning function after extirpation of the proper area -- just as in the morphogenetic field intact tissue will deputize for lost tissue. Moreover, if a rat has learned to run a certain maze, no matter what parts of its motor cortex are injured, it will follow its path -- even if it has to roll the whole way with paralysed legs; and if the injury makes it incapable of executing a right turn, it will achieve its aim by a three-quarter turn to the left. The rat may be blinded, deprived of its smell, partially paralysed in various ways -- each of which would throw the chain-reflex automaton, which the rat was supposed to be, completely out of gear. Yet: 'One drags himself through [the maze] with his forepaws; another falls at every step but gets through by a series of lunges; a third rolls over completely in making each turn, yet avoids rolling into a cul-de-sac and makes an errorless run. The behaviour presents exactly the same problem of direct adaptation of any motor organs to the attainment of a given end which was outstanding in my earlier observations on monkeys after destruction of the pre-central gyri. If the customary sequence of movements employed in reaching the food is rendered impossible, another set, not previously used in the habit, and constituting an entirely different motor pattern, may be directly and evidently substituted without any random activity.' [6]

  In human beings, structural regeneration -- of skin, bone, muscle, and peripheral nerves -- is confined to tissue-outgrowth: that is to say, the new structures are derived from cells of their own kind, not from de-differentiating tissues. But though we have lost the amphibian's enviable powers of replacing a lost limb, we have gained a unique super-flexibility of functions in our nervous system.

  On its lowest level it is manifested in certain secondary or crisis-adaptations of neuro-muscular mechanisms. The artificial limbs invented by the Austrian surgeon Sauerbruch are flexed by muscles in the stump which formerly acted as extensors, and extended by flexors. [7] Other reversals of function are obtained by grafting operations. When the musculo-spiral nerve which activates the extensors of wrist and fingers is severed with resulting paralysis, the damage can be repaired by grafting one of the flexor-tendons from the inner side of the wrist on to the extensor-muscle. After a while the flexor will deputize for his former antagonist, although it remains attached to the 'wrong' (median) nerve -- which thus carries opposite orders to the remaining flexors and to the transplanted one, along the same common path.* [8]

  In these and similar grafting operations the first step is the undoing of a fixed neuro-muscular connection; thereby the nerve to be grafted becomes 'de-specialized' as it were -- the functional analogue to the de-differentiation of structures -- and regains its multi-potentiality to function in more than one way. The second step, when the grafting operation is completed, is 're-specialization' of the nerve in its new role, and the reintegration of the new neuro-muscular unit into the whole. Similar considerations apply to Lashley's rats or Bethe's insects -- except, of course, that in their case, the reorganization of functions is spontaneous.

  Thus there is a close parallel between the regeneration of structures and the reorganization of functions after traumatic challenges; and both are continuous with the regulative principle in morphogenesis. The antithesis between 'localization of functions, fixed pathways' on the one hand, and 'mulitpotential pathways, selective responses' on the other, reflects the earlier antithesis between the mosaic character and the regulative character in embryonic development. But the seemingly opposite principles turn out to be in fact complementary aspects of development. With each successive step in the differentiation of the embryo, of the nervous system, and of adult behaviour-patterns, the regulative powers decrease, and the mosaic-character of structures and functions increases: tissues become specialized, responses localized, habits automatized -- up to a point. For all matrices of structure and behaviour display varying amounts of flexibility even while the organism lives in dynantic equilibrium with its environment; but the often unsuspected amount of its regenerative potential becomes only manifest when a severe challenge induces it to retrace its steps on the genetic gradient, as it were, and make a fresh start.

  Reculer sans Sauter

  Hyper-excited organs or organ-systems tend to get out of control. During the repair of physical injuries, the injured part tends to monopolize the attention of the whole organism; in periods of starvation, the digestive system asserts itself to the detriment of other parts; in rage and panic, the sympathico-adrenal apparatus tyrannizes the whole; and when sex is aroused, reason (as the Austrian proverb has it) 'descends into the testes'. The over-excited part behaves as if it were in a temporary state of 'physiological isolation' (pp. 452 f.), released from its restraints; it asserts its autonomy and sometimes tends to usurp the functions of the whole.

  Analogous situations occur on the cognitive level, where the 'hyper-excited part' appears in the guise of the idée fixe, or a 'closed system' of beliefs. Both the achievements and aberrations of human thought are to a large extent due to ob
sessional preoccupations with religious and scientific theories, or political ideologies, more or less closely knitted around some central idea, around a part-truth usurping the role of the whole truth.

  We have seen in the previous volume how an obsessional preoccupation can force the whole mental organization into its service during the period of incubation, and give birth to a new system of thought. But these are the glorious exceptions; in the vast majority of cases, the 'over-valued idea' (to use Kretschmer's [9] term) will become segregated from the rest of the mental field, and assert itself in harmful ways. The results are all too familiar: personalities whose whole outlook is dominated by prejudice and biassed values; the compulsive rituals of neurotics; the devouring obsession of the crank; and so on to the major psychoses in which large chunks of the personality have been 'split off' and become permanently isolated from the rest. The intrusion of magic causation; inability to distinguish between fact and fantasy; delusions of grandeur, or persecution by invisible powers, are symptoms of regression to earlier levels, of the de-differentiation of thought-matrices -- of reculer sans sauter.